Zeolites and Catalysis: Synthesis, Reactions and Applications (2 Volume set)

Zeolites and Catalysis Synthesis, Reactions and Applications Edited by ˇ Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones ...

Author: Jiri Cejka | Avelino Corma | Stacey Zones

102 downloads 1835 Views 11MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Zeolites and Catalysis Synthesis, Reactions and Applications

Edited by ˇ Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones

Zeolites and Catalysis Edited by ˇ Jiˇrı´ Cejka, Avelino Corma, and Stacey Zones

Related Titles Blaser, H.-U., Federsel, H.-J. (eds.)

Swiegers, G.

Asymmetric Catalysis on Industrial Scale

Mechanical Catalysis

Challenges, Approaches and Solutions 2010 ISBN: 978-3-527-32489-7

Mizuno, N. (ed.)

Modern Heterogeneous Oxidation Catalysis

Methods of Enzymatic, Homogeneous, and Heterogeneous Catalysis Hardcover ISBN: 978-0-470-26202-3

Jackson, S. D., Hargreaves, J. S. J. (eds.)

Metal Oxide Catalysis 2009

Design, Reactions and Characterization

ISBN: 978-3-527-31815-5

2009

Ding, K., Uozumi, Y. (eds.)

ISBN: 978-3-527-31859-9

Barbaro, P., Bianchini, C. (eds.)

Catalysis for Sustainable Energy Production 2009 ISBN: 978-3-527-32095-0

Crabtree, R. H. (ed.)

Handbook of Green Chemistry Green Catalysis Three Volumes

Handbook of Asymmetric Heterogeneous Catalysis 2008 ISBN: 978-3-527-31913-8

Ertl, G., Kn¨ozinger, H., Schuth, ¨ F., Weitkamp, J. (eds.)

Handbook of Heterogeneous Catalysis Eight Volumes 2008 ISBN: 978-3-527-31241-2

2009 ISBN: 978-3-527-31577-2

Ozkan, U. S. (ed.)

Design of Heterogeneous Catalysts New Approaches based on Synthesis, Characterization and Modeling 2009 ISBN: 978-3-527-32079-0

Zeolites and Catalysis Synthesis, Reactions and Applications

Edited by ˇ Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones

The Editor ˇ Prof. Dr. Jiˇrı´ Cejka Academy of Sciences of the Czech Republic Heyrovsk´y Institute of Physical Chemistry Dokjˇskova Dolejskova 3 182 23 Prague 8 Czech Republic

All books published by Wiley-VCH are carefully produced. Nevertheless, authors, editors, and publisher do not warrant the information contained in these books, including this book, to be free of errors. Readers are advised to keep in mind that statements, data, illustrations, procedural details or other items may inadvertently be inaccurate.

Prof. Dr. Avelino Corma University Politecnica de Valencia Institute de Tecnologia Quimica Avenida de los Naranjos s/n 46022 Valencia Spain

Library of Congress Card No.: applied for

Prof. Dr. Stacey I. Zones Chevron Texaco Energy Research and Technology Company 100 Chevron Road Richmond, CA 94802 USA

Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliograﬁe; detailed bibliographic data are available on the Internet at http://dnb.d-nb.de.

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.

 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – by photoprinting, microﬁlm, or any other means – nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not speciﬁcally marked as such, are not to be considered unprotected by law. Composition Laserwords Private Limited, Chennai Printing and Bookbinding strauss GmbH, Mo¨ rlenbach Cover Design Formgeber, Eppelheim Printed in the Federal Republic of Germany Printed on acid-free paper ISBN: 978-3-527-32514-6

V

Contents to Volume 1

Preface XIII List of Contributors 1 1.1 1.2 1.2.1 1.2.2 1.2.3 1.2.4 1.2.5 1.2.6 1.2.7 1.2.8 1.2.9 1.2.10 1.2.11 1.2.12 1.3 1.3.1 1.3.2 1.3.3 1.4 1.4.1 1.4.1.1 1.4.1.2 1.4.1.3 1.4.2 1.4.2.1 1.4.2.2

XVII

Synthesis Mechanism: Crystal Growth and Nucleation 1 Pablo Cubillas and Michael W. Anderson Introduction 1 Theory of Nucleation and Growth 3 Nucleation 3 Supersaturation 3 Energetics 4 Nucleation Rate 5 Heterogeneous and Secondary Nucleation 5 Induction Time 6 Crystal Growth 6 Crystal Surface Structure 6 2D Nucleation Energetics 8 Spiral Growth 9 Interlaced Spirals 10 Growth Mechanisms: Rough and Smooth Surfaces 10 Nucleation and Growth in Zeolites 11 Overview 11 Zeolite Nucleation 13 Crystal Growth on Zeolites and Zeotypes 14 Techniques 15 The Solid Crystal 15 AFM 15 HRSEM 16 Confocal Microscopy 16 Solution Chemistry – Oligomers and Nanoparticles 17 Nuclear Magnetic Resonance 17 Mass Spectrometry 19

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

VI

Contents

1.4.2.3 1.4.3 1.4.3.1 1.5 1.5.1 1.5.1.1 1.5.1.2 1.5.2 1.5.3 1.5.4 1.5.4.1 1.5.4.2 1.5.5 1.5.5.1 1.5.5.2 1.5.6 1.6

Cryo-TEM 20 Modeling 21 Monte Carlo Modeling of Crystal Growth 21 Case Studies 23 Zeolite A 23 Thompson Synthesis 24 Petranovskii Synthesis 26 Silicalite 28 LTL 33 STA-7 35 {001} Faces 38 {001} Faces 38 Zincophosphates 43 ZnPO4 -Sodalite 43 ZnPO4 -Faujasite 47 Metal Organic Frameworks 47 Conclusions and Outlook 49 References 50

2

Synthesis Approaches 57 Karl G. Strohmaier Introduction 57 Aluminophosphates 58 Mineralizers 59 Dry Gel Conversion Syntheses 61 Low Water Syntheses 62 Germanium Zeolites 63 Isomorphous Substitution 65 Structure-Directing Agents 67 SDA Modeling 70 Co-templating 72 Layered Precursors 73 Nonaqueous Solvents 77 Summary and Outlook 79 Acknowledgments 80 References 80

2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13

3 3.1 3.2 3.3 3.4

Ionothermal Synthesis of Zeolites and Other Porous Materials 87 Russell E. Morris Introduction 87 Hydrothermal, Solvothermal, and Ionothermal Synthesis 89 Ionothermal Aluminophosphate Synthesis 90 Ionothermal Synthesis of Silica-Based Zeolites 92

Contents

3.5 3.6 3.7 3.8 3.9 3.10 3.11

4

4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8 4.9 4.10

5 5.1 5.2 5.2.1 5.2.2 5.3 5.3.1 5.3.2 5.4 5.4.1 5.4.2 5.4.3 5.5

Ionothermal Synthesis of Metal Organic Frameworks and Coordination Polymers 92 Ambient Pressure Ionothermal Synthesis 93 The Role of Cation-Templating, Co-Templating, or No Templating 95 The Role of the Anion – Structure Induction 97 The Role of Water and Other Mineralizers 99 Unstable Ionic Liquids 101 Summary and Outlook 101 References 102 Co-Templates in Synthesis of Zeolites 107 Joaquin P´erez-Pariente, Raquel Garc´ıa, Luis G´omez-Hortig¨uela, and Ana Bel´en Pinar Introduction 107 Templating of Dual-Void Structures 108 Crystallization of Aluminophosphate-Type Materials 113 Combined Use of Templating and Pore-Filling Agents 116 Cooperative Structure-Directing Effects of Organic Molecules and Mineralizing Anions 117 Cooperative Structure-Directing Effect of Organic Molecules and Water 119 Control of Crystal Size and Morphology 122 Membrane Systems 123 Use of Co-Templates for Tailoring the Catalytic Activity of Microporous Materials 123 Summary and Outlook 125 Acknowledgments 127 References 127 Morphological Synthesis of Zeolites 131 Sang-Eon Park and Nanzhe Jiang Introduction 131 Morphology of Large Zeolite Crystals 132 Large Crystals of Natural Zeolites 132 Synthesis of Large Zeolite Crystals 133 Morphology Control of MFI Zeolite Particles (of Size Less than 100 µm) 138 Dependence of Structure-Directing Agents (SDAs) 139 Dependence on Alkali-Metal Cations 141 Morphological Synthesis by MW 142 Examples of MW Dependency 142 Morphological Fabrication by MW 143 Formation Scheme of Stacked Morphology 146 Summary and Outlook 149

VII

VIII

Contents

Acknowledgments References 150 6 6.1 6.2 6.3 6.3.1 6.3.1.1 6.3.1.2 6.3.1.3 6.3.1.4 6.3.1.5 6.3.2 6.3.2.1 6.3.2.2 6.3.2.3 6.3.2.4 6.4

7 7.1 7.2 7.2.1 7.2.2 7.2.3 7.2.4 7.2.5 7.2.6 7.2.7 7.3 7.3.1 7.3.2 7.3.3

150

Post-synthetic Treatment and Modiﬁcation of Zeolites 155 Cong-Yan Chen and Stacey I. Zones Introduction 155 Direct Synthesis of Zeolites 155 Post-synthetic Treatment and Modiﬁcation of Zeolites 157 Aluminum Reinsertion into Zeolite Framework Using Aqueous Al(NO3 )3 Solution under Acidic Conditions 158 Experimental Procedures 158 One-Step Method versus Two-Step Method 159 Effects of the Ratio of Al(NO3 )3 to Zeolite 160 Effects of pH, Time, Temperature, and Other Factors 161 Applicable to Medium Pore Zeolite? 161 Synthesis of Hydrophobic Zeolites by Hydrothermal Treatment with Acetic Acid 162 Experimental Procedures 162 Highly Crystalline Pure-Silica Zeolites Prepared via This Technique 163 Effects of Type of Acid, pH, Temperature, and Other Factors 163 Experimental Results from Our Lab 164 Summary and Outlook 166 Acknowledgments 167 References 167 Structural Chemistry of Zeolites 171 Paul A. Wright and Gordon M. Pearce Introduction 171 Zeolite Structure Types Exempliﬁed by Those Based on the Sodalite Cage 172 Introduction 172 The Framework: Secondary Building Units in Zeolite Structural Chemistry 175 Assembling Sodalite Cages: Sodalite, A, Faujasites X and Y, and EMC-2 177 Faujasitic Zeolites X and Y as Typical Examples 178 Key Inorganic Cation-Only Zeolites Pre-1990 179 Structures Templated by Simple Alkylammonium Ions 182 Lessons from Nature 184 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features 185 Introduction 185 Novel Structures and Pore Geometries 187 Expansion of the Coordination Sphere of Framework Atoms 191

Contents

7.3.4 7.3.5 7.3.6 7.3.7 7.3.8 7.4 7.4.1 7.4.2

The Current Limits of Structural Complexity in Zeolites 193 Chirality and Mesoporosity 195 Ordered Vacancies and Growth Defects 197 Zeolites from Layered Precursors 198 Substitution of Framework Oxygen Atoms 199 Summary and Outlook 201 Summary 201 Outlook 202 References 204

8

Vibrational Spectroscopy and Related In situ Studies of Catalytic Reactions Within Molecular Sieves 209 Eli Stavitski and Bert M. Weckhuysen Introduction 209 Acidity Determination with IR Spectroscopy of Probe Molecules 211 Zeolite Synthesis Processes 218 Selection of Zeolite-Based Catalytic Reactions 221 Catalytic Decomposition of Nitric Oxides 221 Methanol-to-Oleﬁn Conversion 225 IR Microspectroscopy 231 Concluding Remarks and Look into the Future 232 Acknowledgment 234 References 234

8.1 8.2 8.3 8.4 8.4.1 8.4.2 8.5 8.6

9 9.1 9.2 9.2.1 9.2.1.1 9.2.1.2 9.2.1.3 9.2.2 9.2.2.1 9.2.2.2 9.2.3 9.3 9.3.1 9.3.2 9.3.3 9.3.4 9.3.4.1 9.3.5 9.3.5.1 9.3.5.2

Textural Characterization of Mesoporous Zeolites 237 Lei Zhang, Adri N.C. van Laak, Petra E. de Jongh, and Krijn P. de Jong Introduction 237 Methods for Generating Meso- and Macropores in Zeolites 239 Postsynthesis Modiﬁcation 239 Dealumination 239 Desilication 241 Detitanation 242 Templating Method 243 Hard Template 243 Soft Template 244 Other Methods 245 Characterization of Textural Properties of Mesoporous Zeolites 246 Gas Physisorption 246 Thermoporometry 251 Mercury Porosimetry 255 Electron Microscopy 256 SEM and TEM 256 NMR Techniques 266 129 Xe NMR Spectroscopy 266 PFG NMR 269

IX

X

Contents

9.3.6 9.4

In situ Optical and Fluorescence Microscopy 271 Summary and Outlook 273 Acknowledgments 274 References 274

10

Aluminum in Zeolites: Where is it and What is its Structure? 283 Jeroen A. van Bokhoven and Nadiya Danilina Introduction 283 Structure of Aluminum Species in Zeolites 284 Reversible versus Irreversible Structural Changes 285 Cautionary Note 286 Development of Activity and Changing Aluminum Coordination 286 Where is the Aluminum in Zeolite Crystals? 289 Aluminum Zoning 289 Aluminum Distribution Over the Crystallographic T Sites 292 Summary and Outlook 296 Acknowledgment 298 References 298

10.1 10.2 10.2.1 10.2.2 10.2.3 10.3 10.3.1 10.3.2 10.4

11 11.1 11.2 11.2.1 11.2.2 11.2.3 11.2.4 11.3 11.4 11.5 11.6 11.7

12 12.1 12.2 12.2.1 12.2.2 12.3 12.3.1

Theoretical Chemistry of Zeolite Reactivity 301 Evgeny A. Pidko and Rutger A. van Santen Introduction 301 Methodology 302 Ab initio Methods 303 DFT Methods 303 Basis Sets 304 Zeolite Models 306 Activation of Hydrocarbons in Zeolites: The Role of Dispersion Interactions 307 Molecular-Level Understanding of Complex Catalytic Reactions: MTO Process 316 Molecular Recognition and Conﬁnement-Driven Reactivity 321 Structural Properties of Zeolites: Framework Al Distribution and Structure and Charge Compensation of Extra-framework Cations 326 Summary and Outlook 330 References 331 Modeling of Transport and Accessibility in Zeolites 335 Sof´ıa Calero Diaz Introduction 335 Molecular Models 336 Modeling Zeolites and Nonframework Cations 336 Modeling Guest Molecules 337 Simulation Methods 338 Computing Adsorption 339

Contents

12.3.2 12.3.3 12.3.4 12.4 12.4.1 12.4.1.1 12.4.1.2 12.4.1.3 12.4.2 12.4.2.1 12.4.2.2 12.5

13 13.1 13.2 13.2.1 13.2.2 13.2.3 13.2.4 13.2.5 13.3 13.3.1 13.3.2 13.3.3 13.3.4 13.4 13.5 13.6

Computing Free Energy Barriers 341 Computing Volume-Rendered Pictures, Zeolite Surface Areas, and Zeolite Pore Volumes 343 Computing Diffusion 344 Molecular Modeling Applied to Processes Involving Zeolites 346 Applications in Technological Processes 346 Molecular Modeling of Conﬁned Water in Zeolites 346 Molecular Modeling of Hydrocarbons in Zeolites 348 Molecular Modeling of Separation of Mixtures in Zeolites 349 Applications in Green Chemistry 351 Carbon Dioxide Capture 351 Natural Gas Puriﬁcation 352 Summary and Outlook 353 Acknowledgments 354 References 354 Diffusion in Zeolites – Impact on Catalysis 361 Johan van den Bergh, Jorge Gascon, and Freek Kapteijn Introduction 361 Diffusion and Reaction in Zeolites: Basic Concepts 362 Importance of Adsorption 364 Self-Diffusivity 364 Mixture Diffusion 365 Diffusion Measurement Techniques 365 Relating Diffusion and Catalysis 366 Diffusion in Zeolites: Potential Issues 368 Concentration Dependence of Diffusion 368 Single-File Diffusion 370 Surface Barriers 372 The Thiele Concept: A Useful Approach in Zeolite Catalysis? 374 Pore Structure, Diffusion, and Activity at the Subcrystal Level 375 Improving Transport through Zeolite Crystals 379 Concluding Remarks and Future Outlook 382 References 383

Contents to Volume 2 14

Special Applications of Zeolites 389 V´ıctor Sebasti´an, Clara Casado, and Joaqu´ın Coronas

15

Organization of Zeolite Microcrystals 411 Kyung Byung Yoon

XI

XII

Contents

16

Industrial Potential of Zeolites 449 Giuseppe Bellussi, Angela Carati, and Roberto Millini

17

Catalytically Active Sites: Generation and Characterization 493 Michael Hunger

18

Cracking and Hydrocracking 547 Marcello Rigutto

19

Reforming and Upgrading of Diesel Fractions 585 Carlo Perego, Vincenzo Calemma, and Paolo Pollesel

20

Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites 623 ˇ Sulaiman Al-Khattaf, Mohammad Ashraf Ali, and Jiˇr´ı Cejka

21

Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves for the Conversion of Natural Gas to Fuels and Chemicals 649 Agust´ın Mart´ınez, Gonzalo Prieto, Andr´es Garc´ıa-Trenco, and Ernest Peris

22

Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG) 687 Michael St¨ocker

23

Metals in Zeolites for Oxidation Catalysis 713 Takashi Tatsumi

24

Environmental Catalysis over Zeolites 745 Gabriele Centi and Siglinda Perathoner

25

Zeolites as Catalysts for the Synthesis of Fine Chemicals 775 Maria J. Climent, Avelino Corma, and Sara Iborra

26

Zeolites and Molecular Sieves in Fuel Cell Applications King Lun Yeung and Wei Han Index 863

827

V

Contents to Volume 1

1

Synthesis Mechanism: Crystal Growth and Nucleation Pablo Cubillas and Michael W. Anderson

1

2

Synthesis Approaches 57 Karl G. Strohmaier

3

Ionothermal Synthesis of Zeolites and Other Porous Materials 87 Russell E. Morris

4

Co-Templates in Synthesis of Zeolites 107 Joaquin P´erez-Pariente, Raquel Garc´ıa, Luis G´omez-Hortig¨uela, and Ana Bel´en Pinar

5

Morphological Synthesis of Zeolites Sang-Eon Park and Nanzhe Jiang

6

Post-synthetic Treatment and Modiﬁcation of Zeolites 155 Cong-Yan Chen and Stacey I. Zones

7

Structural Chemistry of Zeolites 171 Paul A. Wright and Gordon M. Pearce

8

Vibrational Spectroscopy and Related In situ Studies of Catalytic Reactions Within Molecular Sieves 209 Eli Stavitski and Bert M. Weckhuysen

9

Textural Characterization of Mesoporous Zeolites 237 Lei Zhang, Adri N.C. van Laak, Petra E. de Jongh, and Krijn P. de Jong

10

Aluminum in Zeolites: Where is it and What is its Structure? Jeroen A. van Bokhoven and Nadiya Danilina

131

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

283

VI

Contents

11

Theoretical Chemistry of Zeolite Reactivity 301 Evgeny A. Pidko and Rutger A. van Santen

12

Modeling of Transport and Accessibility in Zeolites Sof´ıa Calero Diaz

13

Diffusion in Zeolites – Impact on Catalysis 361 Johan van den Bergh, Jorge Gascon, and Freek Kapteijn

335

Contents to Volume 2 Preface XIII List of Contributors 14 14.1 14.2 14.2.1 14.2.2 14.2.3 14.3 14.4 14.4.1 14.4.2 14.5 14.5.1 14.5.2 14.5.3 14.6

15 15.1 15.2 15.2.1 15.2.1.1 15.2.1.2 15.2.1.3 15.2.1.4

XVII

Special Applications of Zeolites 389 V´ıctor Sebasti´an, Clara Casado, and Joaqu´ın Coronas Introduction 389 Zeolite Membranes 389 Membrane Reactors and Microreactors 390 Zeolite-Based Gas Sensors 392 Mixed-Matrix Membranes 394 Host–Guest Interactions 396 Medical and Veterinary Applications 399 Medical Applications 399 Veterinary Applications 400 Other Applications 401 Racemic Separations 401 Magnetic Zeolites 402 Hydrogen Storage 403 Summary and Outlook 404 References 406 Organization of Zeolite Microcrystals 411 Kyung Byung Yoon Introduction 411 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly 411 Monolayer Assembly on Solid Substrates 412 Types of Linkages 412 Types of Substrates 415 Methods 415 Characteristic Points to Monitor the Quality of the Monolayers 416

Contents

15.2.1.5 15.2.1.6 15.2.1.7 15.2.1.8 15.2.2 15.2.3 15.2.4 15.2.5 15.2.6 15.2.7 15.3 15.4 15.5

16 16.1 16.2 16.2.1 16.2.2 16.3 16.3.1 16.3.2 16.3.3 16.4 16.5 16.6 16.6.1 16.6.2 16.6.3 16.7

17 17.1 17.2 17.2.1 17.2.2

Four Key Processes Occurring during Monolayer Assembly 417 Effect of Method on Rate, DCP, Coverage, and Binding Strength 424 Factors Affecting Binding Strengths 426 Driving Forces for Uniform Orientation and Close Packing 427 Patterned Monolayer Assembly on Substrates 429 Multilayer Assembly on Substrates 431 Organization into 2D Arrays on Water 431 Organization into Surface-Aligned Zeolite Microballs 434 Self-Assembly of Substrate-Tethering Zeolite Crystals with Proteins 435 In Situ Self-Organization of Zeolite Crystals into Arrays during Synthesis 437 Monolayer Assembly of Zeolite Microcrystals by Dry Manual Assembly 438 Current and Future Applications 441 Summary and Outlook 442 Acknowledgments 444 References 444 Industrial Potential of Zeolites 449 Giuseppe Bellussi, Angela Carati, and Roberto Millini Introduction 449 Application of Zeolites in Slurry Processes 450 TS-1 Based Catalyst for Liquid-Phase Oxidation Processes 451 New Advance in Slurry Phase Reaction with Zeolitic Catalysts 453 Rebalancing the Reﬁnery Products Slate 455 Bottom Cracking Conversion 457 LCO Upgrading 459 Oleﬁns Oligomerization 461 Advanced Separation Technologies 462 Zeolites and Environmental Protection: Groundwater Remediation 467 New Materials for Emerging Applications 471 Zeolites 471 Hierarchical Zeolites 473 Silica-Based Crystalline Organic–Inorganic Hybrid Materials 479 Summary and Outlook 484 References 485 Catalytically Active Sites: Generation and Characterization 493 Michael Hunger Introduction 493 Acid Sites in Zeolites 494 Nature of Acid Sites 494 Formation of Brønsted and Lewis Acid Sites 496

VII

VIII

Contents

17.3 17.3.1 17.3.2 17.3.3 17.3.4 17.3.5 17.3.6 17.4 17.4.1 17.4.2 17.5 17.5.1 17.5.2 17.6 17.6.1 17.6.2 17.7 17.7.1 17.7.2 17.8

18 18.1 18.1.1 18.1.2 18.2 18.2.1 18.2.2 18.2.3 18.2.4 18.2.5 18.3 18.3.1 18.3.2 18.3.3 18.3.4 18.3.5 18.3.6 18.4

Characterization of Acid Sites 498 Catalytic Test Reactions 498 Titration with Bases 500 Temperature-Programmed Desorption of Bases 501 Microcalorimetry 504 FTIR Spectroscopy 508 NMR Spectroscopy 514 Base Catalysis 521 Nature of Base Sites 521 Formation of Base Sites 522 Characterization of Base Sites in Zeolites 523 Test Reactions 523 Analytical and Spectroscopic Methods 525 Metal Clusters in Zeolites 529 Nature of Metal Clusters 529 Formation of Metal Clusters 530 Characterization of Metal Clusters in Zeolites 532 Test Reactions 532 Analytical Methods 534 Summary and Outlook 535 References 535 Cracking and Hydrocracking 547 Marcello Rigutto Introduction 547 The Oil Reﬁnery – Where to Find Zeolites in It, and Why – and the Place of Hydrocracking and Catalytic Cracking 547 The Changing Environment for Reﬁning 549 FCC 551 The FCC Process 551 The FCC Catalyst, and Catalytic Chemistry 555 Residue Cracking and the Effect of Deposited Metals on the Catalyst 558 Light Alkenes by Addition of ZSM-5 559 Potential Use of Other Zeolites in FCC 561 Hydrocracking 561 The Hydrocracking Process 561 Feedstocks and Products 563 Hydrocracking Catalyst Systems, and Catalytic Chemistry 566 Zeolite Y in Hydrocracking 570 New Catalyst Developments 575 Residue Conversion – Some Notes 576 Summary and Outlook 576 References 578

Contents

19 19.1 19.2 19.2.1 19.2.2 19.2.3 19.2.3.1 19.2.3.2 19.3 19.3.1 19.3.1.1 19.3.1.2 19.3.2 19.3.2.1 19.4

20

20.1 20.2 20.3 20.3.1 20.3.2 20.3.3 20.3.4 20.3.5 20.4 20.4.1 20.4.2 20.5 20.6 20.6.1 20.6.2 20.6.2.1 20.6.3 20.7 20.8

Naphtha Reforming and Upgrading of Diesel Fractions 585 Carlo Perego, Vincenzo Calemma, and Paolo Pollesel Introduction 585 Catalytic Reforming 587 Process 587 Reforming Chemistry 591 Catalyst 595 Zeolite Catalysts 598 Commercial Catalysts 600 Upgrading Diesel Fractions: Catalytic Dewaxing 601 Shape Selectivity 602 Catalytic Dewaxing via Shape Selective Cracking 605 Dewaxing via Isomerization 607 Commercial Applications 609 Commercial Processes 610 Summary and Outlook 618 References 619 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites 623 ˇ Sulaiman Al-Khattaf, Mohammad Ashraf Ali, and Jiˇr´ı Cejka Introduction 623 Zeolites under Study 623 Toluene Disproportionation 625 Zeolite Modiﬁcation by Silicon Deposition 626 Zeolite Modiﬁcation by Precoking 627 Zeolite Modiﬁcation by Dealumination 627 Zeolite Modiﬁcation by Metal Deposition 628 Factors Affecting Toluene Disproportionation 629 Ethylbenzene Disproportionation 630 Effect of Crystal Size and Surface Modiﬁcation 631 Kinetic Investigations of Ethylbenzene Disproportionation 631 Disproportionation and Transalkylation of Trimethylbenzene 633 Alkylation of Aromatics 635 Ethylation of Benzene 635 Methylation of Toluene 636 Modiﬁcation of the External Surface of Zeolites 638 Ethylation of Toluene and Ethylbenzene 640 Miscellaneous 642 Summary and Outlook 643 Acknowledgments 644 References 644

IX

X

Contents

21

21.1 21.2 21.2.1 21.2.2 21.2.3 21.3 21.3.1 21.3.2 21.3.2.1 21.3.2.2 21.3.3 21.3.3.1 21.3.3.2 21.3.3.3 21.4

22 22.1 22.2 22.3 22.3.1 22.3.2 22.3.3 22.3.4 22.3.5 22.4 22.4.1 22.4.2 22.4.3 22.5 22.6 22.7 22.7.1 22.7.2 22.7.3

Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves for the Conversion of Natural Gas to Fuels and Chemicals 649 Agust´ın Mart´ınez, Gonzalo Prieto, Andr´es Garc´ıa-Trenco, and Ernest Peris Introduction 649 Direct Conversion of Methane 651 Oxidative Conversion: OCM and Methylation Processes 651 Nonoxidative Methane Homologation and Alkylation Processes 654 Nonoxidative Methane Dehydroaromatization (MDA) 655 Syngas Conversion Processes 659 Selective Synthesis of Short-Chain (C2 –C4) Oleﬁns 659 Fischer–Tropsch Synthesis (FTS) 663 Conventional FTS 663 Modiﬁed (Bifunctional) FTS 668 Synthesis of Oxygenates 670 One-Step Synthesis of Dimethyl Ether (DME) from Syngas 670 Syngas to Higher (C2+ ) Oxygenates 674 Carbonylation of MeOH and DME 676 Summary and Outlook 678 Acknowledgments 680 References 680 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG) 687 Michael St¨ocker Introduction 687 Mechanism and Kinetics of the MTO and MTG Reactions 690 Methanol to Oleﬁns (MTO) 697 Catalysts and Reaction Conditions 697 Deactivation 697 Process Technology and Design 699 Commercial Aspects/Economic Impact 700 Future Perspectives 701 Methanol to Gasoline (MTG) 702 Catalysts and Reaction Conditions 702 Deactivation 702 Process Technology 703 Methanol to Propene (MTP) 703 TIGAS Process 705 Mobil’s Oleﬁn-to-Gasoline and Distillate Process (MOGD) 706 Catalyst and Process Operation 706 Thermodynamic Considerations 706 Technical Process 707

Contents

22.8 22.9

Summary and Outlook 707 Outlook 708 References 708

23

Metals in Zeolites for Oxidation Catalysis 713 Takashi Tatsumi Introduction 713 Titanium-Containing Zeolites 715 TS-1 715 Ti-Beta 722 Ti-MWW 724 Other Titanium-Containing Zeolites 731 Solvent Effects and Reaction Intermediate 732 Other-Metal-Containing Zeolites 736 Conclusion 738 References 739

23.1 23.2 23.2.1 23.2.2 23.2.3 23.2.4 23.2.5 23.3 23.4

24 24.1 24.2 24.3 24.4

25 25.1 25.2 25.2.1 25.2.2 25.2.3 25.2.4 25.2.5 25.2.6 25.2.7 25.2.7.1 25.2.7.2 25.2.7.3 25.3 25.3.1 25.3.2 25.3.3

Environmental Catalysis over Zeolites 745 Gabriele Centi and Siglinda Perathoner Introduction 745 A Glimpse into Opportunities and Issues 746 Fields of Applications 756 Summary and Outlook 769 References 770 Zeolites as Catalysts for the Synthesis of Fine Chemicals 775 Maria J. Climent, Avelino Corma, and Sara Iborra Introduction 775 Acid-Catalyzed Reactions 775 Friedel–Crafts Acylation 775 Hydroxyalkylation of Aromatic Compounds 780 Diels–Alder Reactions 783 Acetalization of Carbonyl Compounds 787 Fischer Glycosidation Reactions 789 Isomerization Reactions: Isomerization of α-Pinene and α-Pinene Oxide 792 Oxidation and Reduction Reactions 795 Epoxidation Reactions 795 Baeyer–Villiger Oxidations 799 Meerwein–Ponndorf Verley Reduction and Oppenauer Oxidation (MPVO) 803 Base-Catalyzed Reactions 808 The Knoevenagel Condensation 809 Michael Addition 813 Aldol Condensations 816

XI

XII

Contents

25.4

Summary and Outlook 819 References 819

26

Zeolites and Molecular Sieves in Fuel Cell Applications 827 King Lun Yeung and Wei Han Introduction 827 Zeolites in Electrolyte Membrane 827 Zeolite Conductivities 829 Zeolite/Polymer Composite Membranes 833 Zeolite/PTFE Composite Membranes 839 Zeolite/PFSA Composite Membranes 839 Zeolite/Chitosan Composite Membranes and Others 840 Self-Humidifying Composite Membranes 841 Zeolite and Mesoporous Inorganic Membranes 841 Zeolite Electrocatalysts 842 Zeolites and Molecular Sieves in Fuel Processing 844 Removal of Sulfur Compounds in Fuel 845 Hydrogen Production and Puriﬁcation 845 Reforming of Hydrocarbons 845 Steam Reforming of Alcohols 849 Decomposition of CH4 and NH3 849 CO Removal from H2 -Rich Gas 849 Hydrogen Storage 850 Summary and Outlook 856 Acknowledgment 856 References 856

26.1 26.2 26.2.1 26.2.2 26.2.2.1 26.2.2.2 26.2.2.3 26.2.2.4 26.2.3 26.3 26.4 26.4.1 26.4.2 26.4.2.1 26.4.2.2 26.4.2.3 26.4.2.4 26.4.3 26.5

Index 863

XIII

Preface One can safely say that the impact of zeolites in science and technology in the last 50 years has no precedents in the ﬁeld of materials and catalysis. Although the ﬁrst description of zeolites dates back up to 250 years ago, the last ﬁve decades experienced an incredible boom in zeolite research activities resulting in the successful synthesis of almost 200 different structural types of zeolites, numerous excellent scientiﬁc papers on the synthesis of zeolites, characterization of their properties, and applications of zeolites in adsorption and catalysis that have revolutionized the petrochemical industry. In addition, based on the knowledge of zeolites several other areas of porous materials have recently emerged including mesoporous materials, hierarchic systems, metal-organic frameworks (cationic-periodic polymers) and mesoporous organosilicas. All these materials have substantially increased the portfolio of novel porous materials possessing new interesting properties, but this topic is not covered in this book. This book consists of two volumes. The ﬁrst one is mostly concentrated on recent advances in the synthesis of zeolites and understanding of their properties while the second volume describes recent achievements in the application of zeolites mostly in catalysis. More speciﬁcally, the ﬁrst volume starts with a chapter by P. Cubillas and M.W. Anderson (Chapter 1) discussing mechanisms of the synthesis of zeolites and zeotypes, including nucleation and crystal growth, employing various microscopic techniques. This is followed by a chapter of K. Strohmaier (Chapter 2) providing a detailed survey on the synthesis of novel zeolites and different layered precursors incorporating different metal ions into the framework, and applying ever increasing number of structure-directing agents. A new approach to the synthesis of zeolites and other porous materials by ionothermal synthesis combining ionic liquids as the solvent together with the structure-directing agent is presented by R. Morris (Chapter 3). Zeolite synthesis can also be controlled by a simultaneous use of two different templates providing new tool for creative chemistry Nas discussed by the group of J. P´erez-Pariente (Chapter 4). Morphological control of zeolite crystals is one of the key issues to understand the mechanism of zeolite crystallization as well as to control the performance of zeolites in various applications as it is outlined by S.-E. Park and N. Jiang in Chapter 5. Introduction of other elements than silicon into the zeolite framework can be done not only via synthesis but also in the Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

XIV

Preface

postsynthesis steps as highlighted for deboronation followed by realumination as described by C.Y. Chen and S.I. Zones (Chapter 6). P.A. Wright and G.M. Pearce show how the individual zeolite structures are built from basic secondary building units. The authors focus not only on general aspects of zeolite structures but also on the description of structures of zeolites determined very recently (Chapter 7). Structural and textural characterization of zeolites starts in Chapter 8, written by E. Stavitski and B.M. Weckhuysen, providing good examples of application of vibrational spectroscopy under static conditions that can drive into in situ catalytic investigations. The group of K. de Jong (Chapter 9) makes an effort to evaluate different physicochemical methods used for textural characterization of zeolites. Gas physisorption, mercury porosimetry, electron microscopy (including 3D experiments), various NMR techniques up to in situ optical and ﬂuorescence microscopy are discussed in detail. The location, coordination, and accessibility of framework aluminum are of key importance for acid-catalyzed reactions in zeolites and these issues are addressed by J.A. van Bokhoven and N. Danilina in Chapter 10. Theoretical background of zeolite reactivity employing different computational approaches and models is covered in Chapter 11 by E.A. Pidko and R.A. van Santen. S. Calero Diaz presents an overview of current developments in modeling of transport and accessibility in zeolites showing some recent models and simulation methods that are applied for systems of environmental and industrial interests (Chapter 12). The ﬁnal chapter of the ﬁrst volume is written by the group of F. Kapteijn (Chapter 13), in which diffusion in zeolites starting from basic models of diffusion up to the role of diffusion in adsorption and catalytic processes is discussed. The second volume starts with a chapter of the group of J. Coronas concentrating on special applications of zeolites including green chemistry, hybrid materials, medicine, veterinary, optical- and electrical-based applications, multifunctional fabrics, and nanotechnology (Chapter 14). After that K.B. Yoon presents the opportunities to organize zeolite microcrystals into two- and three-dimensionally organized structures and the application of these organized entities in membranes, antibacterial functional fabrics, supramolecularly organized light-harvesting systems, and nonlinear optical ﬁlms (Chapter 15). The remaining chapters are exclusively devoted to the application of zeolites in catalysis. G. Bellussi opens this part with a broad overview of current industrial processes using zeolites as key components of the catalysts and further challenges in this area (Chapter 16). Generation, location, and characterization of catalytically active sites are discussed in depth by M. Hunger showing different aspects of shape selectivity and structural effect on the properties of active sites (Chapter 17). M. Rigutto (Chapter 18) stresses the importance of zeolites and the main reasons for their application in cracking and hydrocracking, the largest industrial processes employing zeolites as catalysts. Further, C. Perego and his coworkers focus on reforming and upgrading of diesel fractions, which with gasoline are by far the most important and valuable key fractions produced by petroleum reﬁneries (Chapter 19). Transformation of aromatic compounds forms the heart of petrochemical processes ˇ with zeolites as key components of all catalysts. S. Al-Khattaf, M.A. Ali, and J. Cejka

Preface

highlight the most important recent achievements in application of zeolites in various alkylation, isomerization, disproportionation, and transalkylation reactions of aromatic hydrocarbons (Chapter 20). With decreasing supply of oil, natural gas obtains more and more importance. A. Martinez and his coauthors discuss in some detail different ways of methane upgrading into valuable fuels and chemicals (Chapter 21). Methanol, which can be obtained from natural gas, could be one of the strategic raw materials in future. Novel processes transforming methanol into oleﬁns or gasoline are covered in Chapter 22 by M. St¨ocker. Incorporation of catalytically active species into zeolite frameworks or channel systems for oxidation reactions is covered in Chapter 23 by T. Tatsumi. The main attention is devoted to Ti-silicates. G. Centi and S. Perathoner focus on increasing applicability of zeolites in environmental catalysis with a particular attention to conversion of nitrogen oxides (Chapter 24). K.L. Yeung and W. Han describe the emerging ﬁeld of application of zeolites in fuel cells for clean energy generation. The authors show that zeolites can play an important role in hydrogen production, puriﬁcation, conditioning, and storage (Chapter 25). The ﬁnal chapter by the authors from the group of A. Corma presents possibilities of application of zeolite as catalysts in the synthesis of ﬁne chemicals. The examples discussed include, for example, acylation, hydroxyalkylation, acetalization, isomerization, Diels–Alder, and Fischer glucosidation reactions. Bringing together these excellent chapters describing the cutting edge of zeolite research and practice provides an optimistic view for the bright future of zeolites. The number of new synthesized zeolites is ever increasing and particularly novel extra-large pore zeolites or even chiral zeolitic materials will surely be applied in green catalytic processes enabling to transform bulkier substrates into desired products. In a similar way, application of zeolites in adsorption or separation is one of the most important applications of this type of materials saving particularly energy needed for more complex separation processes if zeolites were not available to do the job. Fast development of experimental techniques enables deeper insight into the structural and textural properties of zeolites, while particularly spectroscopical methods provide new exciting information about the accessibility of inner zeolite volumes and location and coordination of active sites. Catalysis is still the most promising area for application of zeolites, in which novel zeolitic catalysts with interesting shape-selective properties can enhance activities and selectivities not only in traditional areas such as petrochemistry but also in environmental protection, pollution control, green chemistry, and biomass conversion. Last but not least, novel approaches in the manipulation and modiﬁcation of zeolites directed to fuel cells, light harvesting, membranes, and sensors clearly evidence a large potential of zeolites in these new areas of application. The only limitation in zeolite research is the lack of our imagination, which slows down our effort and attainment of new exciting achievements. It was our great pleasure working with many friends and excellent researchers in the preparation of this book. We would like to thank sincerely all of them for their timely reviews on selected topics and the great effort to put the book together. We believe that this book on zeolites will be very helpful not only for experienced

XV

XVI

Preface

researchers in this ﬁeld but also students and newcomers will ﬁnd it as a useful reference book. ˇ Jiˇr´ı Cejka Avelino Corma Canos Stacey I. Zones Prague Valencia Richmond October 2009

XVII

List of Contributors Sulaiman Al-Khattaf King Fahd University of Petroleum & Minerals (KFUPM) Chemical Engineering Department Research Institute Dhahran Saudi Arabia Mohammad Ashraf Ali King Fahd University of Petroleum & Minerals (KFUPM) Center of Excellence in Reﬁning & Petrochemicals Research Institute Dhahran Saudi Arabia Giuseppe Bellussi Enitecnologie San Donato Milanese Research Centre Reﬁning & Marketing Division Research & Technological Development Via F. Maritano 26 20097 San Donato Milanese (MI) Italy Vincenzo Calemma Eni S.p.A. Reﬁning & Marketing Division 20097 San Donato Milanese (MI) Italy

Angela Carati Enitecnologie San Donato Milanese Research Centre Reﬁning & Marketing Division Research & Technological Development Via F. Maritano 26 20097 San Donato Milanese (MI) Italy Clara Casado Universidad de Zaragoza Chemical and Environmental Engineering Department and Nanoscience Institute of Arag´on Mar´ıa de Luna 3 50018 Zaragoza Spain ˇ Jiˇrı´ Cejka Academy of Sciences of the Czech Republic Heyrovsk´y Institute of Physical Chemistry Dokjˇskova Dolejskova 3 182 23 Prague 8 Czech Republic

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

XVIII

List of Contributors

Gabriele Centi Universita di Messina Dip. Di Chimica Industriale ed Ingegneria dei Materiali Salita Sperone 31 98166 Messina Italy Maria J. Climent Universidad Polit´ecnica de Valencia Instituto de Tecnolog´ıa Qu´ımica UPV-CSIC Avda. de los Naranjos s/n 46022 Valencia Spain Avelino Corma University Politecnica de Valencia Institute de Tecnologia Quimica Avenida de los Naranjos s/n 46022 Valencia Spain Joaquı´n Coronas Universidad de Zaragoza Chemical and Environmental Engineering Department and Nanoscience Institute of Arag´on Mar´ıa de Luna 3 50018 Zaragoza Spain Andr´es Garcı´a-Trenco UPV-CSIC Instituto de Tecnolog´ıa Qu´ımica Avenida de los Naranjos s/n 46022 Valencia Spain

Wei Han The Hong Kong University of Science and Technology Department of Chemical and Biomolecular Engineering Clear Water Bay Kowloon Hong Kong PR China Michael Hunger University of Stuttgart Institute of Chemical Technology 70550 Stuttgart Germany Sara Iborra Universidad Polit´ecnica de Valencia Instituto de Tecnolog´ıa Qu´ımica UPV-CSIC Avda. de los Naranjos s/n 46022 Valencia Spain Agustı´n Martı´nez UPV-CSIC Instituto de Tecnolog´ıa Qu´ımica Avenida de los Naranjos s/n 46022 Valencia Spain Roberto Millini Enitecnologie San Donato Milanese Research Centre Reﬁning & Marketing Division Research & Technological Development Via F. Maritano 26 20097 San Donato Milanese (MI) Italy

List of Contributors

Carlo Perego Eni S.p.A. Istituto Eni Donegani Via Fauser 4 28100 Novara Italy Ernest Peris UPV-CSIC Instituto de Tecnolog´ıa Qu´ımica Avenida de los Naranjos s/n 46022 Valencia Spain Siglinda Perathoner Universita di Messina Dip. Di Chimica Industriale ed Ingegneria dei Materiali Salita Sperone 31 98166 Messina Italy Paolo Pollesel Eni S.p.A. Reﬁning & Marketing Division 20097 San Donato Milanese (MI) Italy

Vı´ctor Sebasti´an Universidad de Zaragoza Chemical and Environmental Engineering Department and Nanoscience Institute of Arag´on Mar´ıa de Luna 3 50018 Zaragoza Spain Michael St¨ ocker SINTEF Materials and Chemistry Department of Hydrocarbon Process Chemistry P.O. Box 124 Blindern 0314 Oslo Norway Takashi Tatsumi Tokyo Institute of Technology Chemical Resources Laboratory Division of Catalytic Chemistry 4259-R1-9 Nagatsuta-cho Midori-ku Yokohama 226–8503 Japan

Gonzalo Prieto UPV-CSIC Instituto de Tecnolog´ıa Qu´ımica Avenida de los Naranjos s/n 46022 Valencia Spain

King Lun Yeung The Hong Kong University of Science and Technology Department of Chemical and Biomolecular Engineering Clear Water Bay Kowloon Hong Kong PR China

Marcello Rigutto Shell Technology Centre Amsterdam Grasweg 31 1031 HW Amsterdam The Netherlands

Kyung Byung Yoon Sogang University Department of Chemistry Center for Microcrystal Assembly Seoul 121–742 Korea

XIX

1

1 Synthesis Mechanism: Crystal Growth and Nucleation Pablo Cubillas and Michael W. Anderson

1.1 Introduction

Crystal growth pervades all aspects of solid-state materials chemistry and the industries that rely upon the functionality of these materials. In the drive toward greener, more efﬁcient processes crystal engineering is an increasingly important requirement in materials such as catalysts; semiconductors; pharmaceuticals; gas-storage materials; opto-electronic crystals; and radio-active waste storage materials. In order to impart this desired functionality it is crucial to control properties such as crystal perfection, crystal size, habit, intergrowths, chirality, and synthesis cost [1]. The issues that concern crystal growth for nanoporous materials are similar to those that concern all crystal growths. Crystal habit and crystal size are of vital importance to the efﬁcient functioning of these, and any other crystals, for real application. In the extreme case, single-crystal nanoporous ﬁlms will require substantial skewing of both habit and size from normal bounds – this is currently impossible for zeolites but is being realized to some extent for metal organic framework (MOF) materials. Less extreme is the modiﬁcation of crystal aspect ratio, for example, in hexagonal crystal systems where the pore architecture is often one-dimensional, growth of tablet-shaped crystals is usually preferred over more common needle-shaped crystals, particularly when molecular diffusion is important. All crystals incorporate both intrinsic and extrinsic defects, but whereas the presence of the latter may be easily controlled through purity of synthesis conditions, control of the former requires a deep knowledge of the underlying crystal growth mechanism. By defect we mean a well-deﬁned aperiodic interruption in the periodic crystal structure. First, it is important to understand the nature of the defect, which normally requires a form of microscopy. Transmission electron microscopy (TEM) is the principal method used for this, but scanning probe microscopy is also useful. Owing to the structural complexity of framework crystals, each crystal system tends to display a unique defect structure that must be individually characterized. An extension of the same phenomenon is the incorporation of intergrowth and twin structures. Such defects are introduced during the crystal growth stage usually as a result of competing crystallization pathways that are near Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

2

1 Synthesis Mechanism: Crystal Growth and Nucleation

energy equivalent. By understanding the growth mechanism it should be possible to identify the crucial step that controls this fork in the crystal growth, determine the energetic considerations, and predict modiﬁcations to growth conditions so as to enhance the probability of forming one particular crystal over another. This is crucial, for instance, for the preparation of chiral crystals that are assembled from a spiral stacking of achiral units [2, 3]. The advent of atomic force microscopy (AFM) (Figure 1.1) has opened up new possibilities to investigate the molecular events that occur during crystal growth and dissolution/recrystallization. The technique can be used both in situ and ex situ with each method suited to particular problems. Ex situ operation allows a vast array of synthetic parameters to be varied without concern for the delicacies of the AFM operation. In this respect, careful quenching experiments whereby the state of the nanoscopic features at the crystalline surface may be frozen rapidly before transfer to the AFM can be performed. This is crucial to prevent secondary processes caused by changing growth conditions upon crystal cooling and extraction from the mother-liquor. Rates and energies of crystal growth processes can be determined via such ex situ experiments through modeling both crystal topology and habit. In situ AFM gives a more direct approach to determining growth and dissolution rates. Further, surface structures that are inherently less stable may not be seen in ex situ analysis. Consequently, where possible, in situ analysis is preferred. The

(a)

(b)

(c)

(d)

(e)

(f)

Figure 1.1 (a) Interlaced spiral on aluminophosphate STA-7; (b) zeolite A reducing supersaturation; (c) metal organic framework ZIF-8; (d) in situ ZnPO4 -FAU growth structure; (e) interlaced spiral ZnPO4 -SOD growth structure; and (f, g) in situ dissolution of zeolite L.

(g)

1.2 Theory of Nucleation and Growth

structural details leading to the observed crystal growth, defect, and intergrowth structure can also be probed using electron microscopy, and by slicing crystals open we can look at the consequences of structural growth decisions in the heart of the crystals. To probe the solution chemistry from which the crystals have evolved, a combination of the speciation delineation of nuclear magnetic resonance (NMR) with the speed and sensitivity of mass spectrometry is increasing our knowledge substantially. Both these techniques also probe the extent of oligomerization in the buildup to nucleation that can be further probed using cryo-TEM methods.

1.2 Theory of Nucleation and Growth 1.2.1 Nucleation

The formation of a new crystalline entity from a solution starts through the nucleation process. Nucleation is deﬁned as the series of atomic or molecular processes by which the atoms or molecules of a reactant phase rearrange into a cluster of the product phase large enough as to have the ability to grow irreversibly to a macroscopically larger size. The cluster is deﬁned as nucleus [4] or critical nuclei. Nucleation can be homogeneous, in the absence of foreign particles or crystals in the solution, or heterogeneous, in the presence of foreign particles in the solution. Both types of nucleation are collectively known as primary nucleation. Secondary nucleation takes place when nucleation is induced by the presence of crystals of the same substance. 1.2.2 Supersaturation

The driving force needed for the nucleation and growth of a crystal is referred to as supersaturation and is deﬁned as the difference in chemical potential between a molecule in solution and that in the bulk of the crystal phase: µ = µs − µc

(1.1)

where µs is the chemical potential of a molecule in solution and µc is the chemical potential of the molecule in the bulk crystal. Following thermodynamics Eq. (1.1) can be expressed as µ = kT ln S

(1.2)

where k is the Boltzmann constant, T is the absolute temperature, and S is the supersaturation ratio. When µ > 0 the solution is said to be supersaturated, meaning that nucleation and/or growth is possible, whereas when µ < 0 the solution will be undersaturated and dissolution will take place. The form of

3

1 Synthesis Mechanism: Crystal Growth and Nucleation

the supersaturation ratio will change depending on the system considered (i.e., gas/solid, solution/solid, melt/solid). For nucleation and growth from solutions it takes the following form: n

S=

ai i

(1.3)

n

ai,ei

where ni is the number of ith ions in the molecule of the crystal, and ai and ai,e the actual and equilibrium activities of the i molecule in the crystal. 1.2.3 Energetics

According to nucleation theory, the work necessary to form a cluster of n number of molecules is the difference between the free energy of the system in its ﬁnal and initial states [4, 5] plus a term related to the formation of an interface between nucleus and solution. This can be expressed by (assuming a spherical nucleus): GT = −nµ + 4π · r 2 σ

(1.4)

where r is the radius of the nucleus and σ is the surface free energy. If each molecule in the crystal occupies a volume V, then each nucleus will contain (4/3)π · r 3 /V molecules. Eq. (1.4) will then take the following form: r3 4 GT = − π · µ + 4π · r 2 σ 3 V

4pr 2s

∆G

+

r* 0

−

r −4/3pr 3/v∆ µ

∆G T

(a)

J

4

(b)

0

∆ µc

∆µ

Figure 1.2 (a) Total free energy versus cluster size. (b) Nucleation rate as a function of supersaturation (showing the critical supersaturation).

(1.5)

1.2 Theory of Nucleation and Growth

Figure 1.2a shows a plot of GT as a function of r; it can be seen how the function reaches a maximum, which represents the energetic barrier that needs to be surpassed to achieve nucleation (G∗ ). The value of r at this maximum (r ∗ ) is deﬁned as the critical radius or nucleus size [4, 5]. Its value is deﬁned by r∗ =

2σ · V kT1nS

(1.6)

It has been proved that the value of r ∗ decreases (as well as that of G∗ ) as the supersaturation increases [6], meaning that the probability of having nucleation in a given system will be higher, the higher the supersaturation. 1.2.4 Nucleation Rate

The rate of nucleation (i.e., the number of nuclei formed per unit time per unit volume) can be expressed by an Arrhenius-type equation [5]: J = A exp

−G∗ kT

(1.7)

where A also depends on supersaturation. A typical plot of J as a function of supersaturation (S) is depicted in Figure 1.2b. It can be seen in this plot that the nucleation rate is virtually zero until a critical value of supersaturation is achieved, after which the rate increases exponentially. This critical supersaturation (µc ) deﬁnes the so-called metastable zone where crystal growth can proceed without concomitant nucleation taking place. 1.2.5 Heterogeneous and Secondary Nucleation

Equations (1.5) and (1.6) shows that both G∗ and r ∗ depend heavily on the surface free energy (σ ), so any process that modiﬁes this value will have an effect on the possible viability of the nucleation process. It has been proved that in the presence of a foreign substrate the decrease in the value of σ therefore reduces the values of G∗ and r ∗ at constant supersaturation [6], that is, making nucleation more favorable. A decrease in σ will also decrease the value of the critical supersaturation (µc ), since the nucleation rate is also dependent on the surface energy (Eq. (1.7)). This will make heterogeneous nucleation more viable than homogeneous nucleation at low supersaturation conditions. The reduction of the surface energy will be the highest when the best match between the substrate and the crystallizing substance is achieved. This situation is created, of course, when both the substrate and the crystallizing substance are the same, referred to as secondary nucleation. This mechanism will be more favorable than both heterogeneous and homogeneous nucleation and thus produced at lower supersaturation.

5

6

1 Synthesis Mechanism: Crystal Growth and Nucleation

1.2.6 Induction Time

Induction time is deﬁned as the amount of time elapsed between the achievement of a supersaturated solution and the observation of crystals. Its value will thus depend on the setting of t = 0 and the technique used to detect the formation of crystals. The induction period can be inﬂuenced by factors such as supersaturation, agitation, presence of impurities, viscosity, and so on. Mullin [5] deﬁned the induction time as t i = tr + tn + tg

(1.8)

The induction time is separated into three periods: tr is the relaxation time, required for the systems to achieve a quasi-steady-state distribution of molecular clusters; tn is the time required for the formation of a nucleus; and tg is the time required for the nucleus to grow to a detectable size. 1.2.7 Crystal Growth

Crystal growth is the series of processes by which an atom or a molecule is incorporated into the surface of a crystal, causing an increase in size. These different processes can be summarized into four steps [7, 8] illustrated in Figure 1.3: 1) transport of atoms through solution; 2) attachment of atoms to the surface; 3) movement of atoms on the surface; 4) attachment of atoms to edges and kinks. The ﬁrst process is the so-called transport process, whereas 2–4 are referred to as surface processes (and may involve several substeps). Since these different steps normally occur in series, the slowest process will control the overall crystal growth. Therefore, growth can be transport (when step 1 is the slowest) or surface controlled (when steps 2–4 are the slowest). 1.2.8 Crystal Surface Structure

Crystal growth theories are based on considerations of the crystal surface structure. One of the most commonly used models was that provided by Kossel [9]. This model envisions the crystal surface as made of cubic units (Figure 1.4) which form layers of monoatomic height, limited by steps (or edges). These steps contain a number of kinks along their length. The area between steps is referred to as a terrace, and it may contain single adsorbed growth units, clusters, or vacancies. According to this model, growth units attached to the surface will form one bond, whereas those attached to the steps and kinks will form two and three bonds, respectively. Hence, kink sites will offer the most stable conﬁguration. Growth will then proceed by

1.2 Theory of Nucleation and Growth

(4)* (2) (7) (3)

(6)

(5)

(4) (a) (1) Energy

(3) (5)

(2) (4)

(b)

(7)

(6)

Figure 1.3 (a) Schematic representation of processes involved in the crystal growth: (1) Transport of solute to a position near the crystal surface; (2) diffusion through boundary layer; (3) adsorption onto crystal surface; (4) diffusion over the surface;

(4*) desorption from the surface; (5) attachment to a step or edge; (6) diffusion along the step or edge; (7) Incorporation into kink site or step vacancy. (b) Associated energy changes for the processes depicted in (a). Figure modiﬁed from Elwell et al. [7].

the attachment of growth units to kink sites in steps. The kink will move along the step producing a net advancement of the step until this step reaches the face edge. Then, a new step will be formed by the nucleation of an island of monolayer height (or two-dimensional (2D) nucleus) on the crystal surface. This mechanism of growth is normally referred to as layer growth or single nucleation growth and is represented in Figure 1.5. A variation of this growth mechanism occurs when the nucleation rate is faster than the time required for the step to cover the whole crystal surface. In this case, 2D nuclei will form all over the surface and on top Terrace Surface vacancy

Steps Edge vacancy

Kink Growth unit Figure 1.4

Kossel model of a crystal surface.

7

8

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

(b)

(c) Figure 1.5 Schematic representation of layer growth. (a) Incorporation of growth units into step. (b) The step has almost advanced to the edge of the crystal. (c) Formation of 2D nucleus.

of other nuclei. These nuclei will spread and coalesce forming layers. This growth mechanism is normally referred to as multinucleation multilayer growth or birth and spread [10]. 1.2.9 2D Nucleation Energetics

The total free energy change due to the formation of a 2D nucleus of height h and radius r can be calculated by using Eq. (1.5): GT−2D = −π ·

hr 2 µ + 2π · rhσ V

(1.9)

The maximum of this function deﬁnes the value of the critical radius which is given by r2D ∗ =

σ ·V kT ln S

(1.10)

1.2 Theory of Nucleation and Growth

It can be seen that the value of r2D ∗ is half of the nucleus size for homogeneous nucleation (Eq. (1.6)). 1.2.10 Spiral Growth

The energetics of layer growth predict that growth takes place at relatively high supersaturation (needed to overcome the energy barrier associated with 2D nucleation). Nevertheless, it has been observed that crystals can still grow at lower supersaturation than predicted [11]. This dilemma was solved by Frank [12] who postulated that crystal surfaces are intercepted by dislocations. These dislocations will create steps in the surface, obviating the necessity for 2D nucleation. Figure 1.5 shows a schematic diagram on the formation and development of spiral growth. In the initial stage the dislocation creates a step (Figure 1.6a). Growth units attach to the step making it advance and thus generating a second step (Figure 1.6b). This second step will not advance until its length equals 2r2D ∗ ; this is because any growth of a step with a smaller size is not thermodynamically favored. Once the second step starts advancing, it will generate a third step which in turn will not start moving until its length equals 2r2D ∗ (Figure 1.6c), then a fourth step will appear, and so on (Figure 1.6d). This will generate a spiral pattern around the dislocation core, and a self-perpetuating source of steps where growth requires less energy than a layer mechanism (therefore, it can proceed at smaller supersaturation). In the case of a curved step, the spiral will be rounded and its curvature will be determined by the r2D ∗ value at the speciﬁc supersaturation conditions in which 2r2D*

(a)

(b)

(c)

(d)

Figure 1.6

Development of a polygonal spiral.

9

10

1 Synthesis Mechanism: Crystal Growth and Nucleation

the crystal grows. The theory of crystal growth by spiral dislocation was further reﬁned by Burton, Cabrera, and Frank [13], giving rise to what is known as the BCF theory. 1.2.11 Interlaced Spirals

Interlaced spirals are the result of a periodic stacking of differently oriented growth layers, each having a different lateral anisotropy of step velocity [14]. In other words, in this type of spiral a step with unit cell may dissociate in substeps symmetrically related but crystallographically different. The dissociation is the result of a different growth anisotropy by each step due to its different crystallography. Figure 1.7 shows a schematic diagram showing the formation of a spiral of this type according to van Enckevort [14]. The surface in the ﬁgure is produced by two distinct types of steps (I and II), of height 1/2 dhkl , emanating from a central point O. Layers of type I are bound by steps a and b, whereas layers from type II are bound by steps c and d. Steps a and d move fast and steps b and c move slowly. This results in steps a of layer I catching up with steps c from layer II, producing a double step of unit-cell height. The same process is observed in steps d joining the slow steps b. The result is a pattern of unit-cell height steps with interlaced crossovers formed by lower steps of height 1/2 dhkl . Interlaced spirals have been observed in numerous systems, including barite [15], molecular crystals [16, 17], silicon carbide [18], GaN [19], and sheet silicates [20]. 1.2.12 Growth Mechanisms: Rough and Smooth Surfaces

The growth mechanisms can be classiﬁed into three types depending on the interface structure. If the surface is rough the growth mechanism will be of adhesive type, whereas if the surface is smooth growth will take place by either

a O b d

a

a

c

b d

d

Type I

a d

Type II

a c

c

b

b

c

c a a

d

d

a b d Figure 1.7

b d

b d

d

Interlaced spiral formation. Figure modiﬁed from van Enckevort et al. [14].

1.3 Nucleation and Growth in Zeolites

Growth rate

Smooth

Spiral growth

Rough

Birth and spread

Adhesive growth

Supersaturation Figure 1.8

Mechanisms of growth as a function of supersaturation.

birth and spread, or spiral growth. A surface will transform from smooth to rough at high driving force conditions (high supersaturation). Figure 1.8 shows the different growth mechanisms as a function of supersaturation. At a low supersaturation, the interface is smooth and spiral growth is the mechanism of growth. After reaching a critical supersaturation for 2D nucleation, birth and spread dominates the growth. In these two domains crystals are bound by crystallographically ﬂat faces with polyhedral morphologies. At high supersaturation the surface transforms to a rough interface, and adhesive-type growth dominates. In the adhesive-type regime the energetics of growth unit attachment are the same regardless of the crystallographic direction, giving rise to crystals bounded by rounded noncrystallographic surfaces, producing spherulitic, fractal, and dendritic patterns.

1.3 Nucleation and Growth in Zeolites 1.3.1 Overview

Zeolite and zeotype synthesis is well known to be a complex process. The rate of crystallization, types of products formed, and their particulate properties (habit, morphology, and crystal size distribution) depend on a large number of parameters [21]. These parameters encompass the crystallization conditions (temperature, stirring, seeding, and gel aging) as well as composition-dependent parameters (pH, water content, and the ratio between framework-forming elements, template concentration, and ionic strength). Nevertheless, a typical zeolite/zeotype synthesis will involve the following steps [22]: 1) A mixture of amorphous reactants which contains the structure-forming ions (such as Si, Al, P, Ga, Zn, etc.) in a basic medium (although a few zeolite synthesis can also take place in acidic medium [23]). This leads to the formation

11

12

1 Synthesis Mechanism: Crystal Growth and Nucleation

2)

3)

4)

5)

of a heterogeneous, partly reacted phase, which has been referred to as the primary amorphous phase [22, 24]. The nature of this amorphous phase ranges from gel-like to colloidal in the so-called clear solution synthesis [25, 26]. Heating of the reaction mixture (above 100 ◦ C) at autogenic pressures in metal autoclaves. Prior to this the reaction mixture may be left for aging for a period of time (hours to a few days). Formation of a ‘‘secondary amorphous phase’’ at pseudo-equilibrium with a solution phase [22]. Evidence exists that this phase possesses some short range order due to the structuring effect of cations in the solution [24, 27–30]. After an induction period the formation of nuclei takes place. This induction time can be related to the deﬁnition for simple systems given in Eq. (1.8) [31]. The relaxation time (tr ) would be the time required for steps 1–3 to take place, that is, for the formation of the quasi-steady state amorphous solid, whereas tn and tg have the same meaning. Growth of the zeolite material at the expense of the amorphous solid.

These steps are well deﬁned for a multitude of zeolite and zeotype syntheses, but in many cases it may be difﬁcult to differentiate them. This could be because some of the steps overlap or because the experimental difﬁculties in studying the synthesis are excessive [22]. Steps 1–3 have been studied in the last several years by many authors. It is not the purpose of this chapter to deal in detail on this subject but the reader is directed to the review by Cundy and Cox [22] for additional information and references. Figure 1.9 shows the typical shape of a crystallization curve for a zeolite synthesis where both the nucleation rate and the evolution of the crystal-length or crystallinity in the system are plotted as a function of the synthesis time. It can be seen that nucleation only takes place after an induction time, that is, after steps 1–3 have taken place. The rate of nucleation increases rapidly but then decreases to zero. After a certain number of nuclei have formed crystal growth takes place. Initially the Crystal size

Nucleation rate

Supersaturation Time Figure 1.9 Schematic representation of the zeolite synthesis process showing the evolution of nucleation and growth rates, as well as supersaturation, as a function of time.

1.3 Nucleation and Growth in Zeolites

growth rate increases exponentially but rapidly achieves a steady state before ﬁnally decreasing to zero when the nutrients are exhausted. The synthesis process can also be followed according to the theoretical supersaturation curve (superimposed in Figure 1.9). The supersaturation increases initially giving rise to the nucleation and growth phase, then it levels off, as the growth rate achieves a steady state, and ﬁnally decreases to zero as all the nutrients in the solution are incorporated into the growing phase. 1.3.2 Zeolite Nucleation

Zeolite nucleation is a complex problem, since it implies the transformation of an initially amorphous or random structure into a crystalline framework. As observed before, during the formation of the secondary amorphous phase there is an increase in the structural order although of very short range. From this step a random number of structured areas may achieve the size of a nucleus and start to grow into a macroscopic crystal. The use of traditional nucleation theory for studying zeolite nucleation has been employed in the past, for example, in calculating the nucleation rate as the inverse of the induction time [32]. Nevertheless it has been observed that there may be important differences between zeolite crystallization and that of more condensed phases. One of these differences may stem from the high internal surface area of zeolites [33]. The process of zeolite nucleation has proved to be difﬁcult to study and analyze owing to the experimental difﬁculties in making in situ measurements. Various authors have obtained information by using the size distribution and using mathematical models [34] to infer the growth and nucleation rate [21, 35, 36]. Other studies have looked at the effect of aging and seeding on zeolite nucleation. Aging of the initial solution has proved to have inﬂuence on the ﬁnal crystal distribution [22], and hence it can provide valuable information on the nucleation mechanism as has been demonstrated for zeolite A [37] and silicalite [38]. The use of seeds can be the factor used to differentiate between primary and secondary nucleation [22]. Some of the most heated debate on the study of nucleation on zeolite and zeotype synthesis has been centered on the nucleation mechanism, whether this is homogeneous [36, 39] or heterogeneous [40] (primary nucleation) or even secondary [41] (crystal induced). Difﬁculties in discerning one mode or other stem in part due to the very nature of the gel phase, especially in the so-called clear solution system where it is physically difﬁcult to separate colloid-sized gel particles from the aqueous phase [22]. Nevertheless, there appears to be a growing body of studies supporting the idea that nucleation occurs mainly in a gel phase, speciﬁcally at the solution–gel interface [29, 42] where the nutrient concentration gradients are probably the highest. Recent studies on clear solution synthesis have also demonstrated that nucleation actually takes place inside the colloid-sized gel particles [26, 43–45].

13

14

1 Synthesis Mechanism: Crystal Growth and Nucleation

The mechanistic aspects of the nucleation process have also been extensively discussed and it is accepted that the process of progressive ordering inside the gel is conducted through a reversible mechanism of breaking and remaking the chemical bonds in the framework catalyzed by hydroxyl ions [46, 47]. Cations and organic-structure-directing agents also have a crucial role in the nucleation process by surrounding themselves with metal-oxide species in preferred geometries owing to electrostatic and van der Waals interactions [48, 49]. 1.3.3 Crystal Growth on Zeolites and Zeotypes

A multitude of studies have been carried out to study how zeolite and zeotype crystals grow. Surprisingly, in spite of the vast number of structure types, framework composition, and synthesis procedures, it has been generally found that zeolite growth increases linearly during most of the crystallization process. This is true for both gel synthesis [1, 50–55] and clear solution synthesis [56–59], although studies on the latter system have shown a dependence of the growth rate on the crystal size for values below 15–20 nm [60]. Zeolite growth has been found to be affected by a multitude of parameters such as temperature, gel composition, agitation, and aging, and many studies have been devoted to these topics [1, 22]. In general, measured growth rates of zeolites are consistently lower than those of more dense phases (such as ionic crystals), which has been postulated as due to the more complex assembling mechanism of the open-polymeric structure of zeolite and zeotypes [22]. The fact that the growth rate proceeds linearly almost to the end of the synthesis has been used to support the idea of a surface-controlled mechanism [21, 61]. This idea has been supported by the measured values of activation energies for the growth process, which vary between 45 and 90 kJ mol−1 [21, 62–64]. These values are much higher than those corresponding to a diffusion-controlled mechanism [21]. The quest for understanding the true growth mechanism and its fundamentals has been recently aided by the development of new high-resolution surface-sensitive techniques such as AFM [65–67], high-resolution scanning electron microscopy (HRSEM), [68], and high-resolution transmission electron microscopy (HRTEM) [69, 70]. Additionally, new developments in liquid- and solid-state NMR as well as mass-spectrometry techniques shed new light on the physical and chemical nature of the growing units [70]. Furthermore, advanced modeling techniques and theoretical studies have been used to further validate the experimental observation and to provide more insight into the molecular aspects of crystal growth [67]. AFM studies on zeotypes and zeolites were initially limited to the study of natural zeolites [71–75]; nevertheless, in the last several years, numerous studies on synthetic materials have been published [65–67, 76–82], including some on zeotypes [83–85]. Most of these studies have been ex situ where crystals were removed from the solution used for growth before analysis but a few in situ dissolution studies have also been performed [86, 87].

1.4 Techniques

Initial AFM studies were focused on natural zeolites obtaining high-resolution images of the zeolite surface, so its porous structure could be observed at the surface termination [71, 72, 75]. Nevertheless, in 1998, Yamamoto et al. [73] published images of natural heulandite crystals that showed the presence of steps suggesting a possible birth-and-spread mechanism. On the synthetic front, Anderson et al. [88] published the ﬁrst AFM study of zeolite Y, which showed also the formation of steps and terraces on the crystal surface. These studies have been followed by others dealing mainly with zeolite A [67, 77, 87, 89], and X/Y [78, 90] and silicalite [81]. From these investigations, detailed information on the surface termination, the mode of growth, dissolution, and the possible identity of the growth units has been inferred. Until recently, most AFM studies have only revealed the presence of steps and terraces, making some authors to conclude that birth and spread may be the preferential mode of growth for these materials [22]. Nevertheless, a recent study on zeolite A [91] and stilbite [74] shows the formation of spirals. Also, results highlighted in this chapter show that spiral growth in zeotype structures may be more prevalent than originally thought.

1.4 Techniques 1.4.1 The Solid Crystal 1.4.1.1 AFM The AFM is a surface-scanning technique invented by Binning et al. [92] in 1986 as a development of the scanning tunneling microscope (STM) [93]. The AFM provides three-dimensional images of surfaces by monitoring the force between the sample and a very sharp tip (a few nanometers wide). This is in contrast to the STM, which relies on the formation of a tunneling current between the tip and the sample. Therefore, the AFM can be used to scan the surface of virtually any kind of material. Typically, the sample is mounted on a piezoelectric scanner, which moves the sample in x–y and z. The lateral resolution is limited by the tip radius, which normally varies between 10 and 30 nm (although it may be as ˚ making it ideal to low as 3 nm), whereas the vertical resolution is around 1 A, observe small surface details such as steps or 2D nuclei. The end of the tip is attached to a cantilever, which bends when the force between the sample and the tip changes. The deﬂection of the cantilever is monitored by shining a laser on its top surface, which is reﬂected back to a photodiode detector. The output signal of the photodiode is then transmitted to a computer. A feedback control system informs the piezoelectric scanner of any changes in force between the tip and the sample, allowing it to alter the tip–sample separation to maintain the force at a constant value. There are different imaging modes available when using AFM depending on the motion of the tip over the sample. In contact mode, the tip is raster scanned

15

16

1 Synthesis Mechanism: Crystal Growth and Nucleation

over the sample while the cantilever deﬂection is kept constant using feedback control. Intermittent contact mode utilizes an oscillating tip and monitors the phase and amplitude of the cantilever [94]. In this mode the contact between the tip and sample is minimized, and hence may be advantageous for softer samples (e.g., biological samples). Some AFMs can monitor both the vertical and lateral forces (and friction) [95], known as friction force microscopy (FFM). In this case, information on the adhesion, friction, or other mechanical properties of the sample can be obtained [96]. The use of AFM has revolutionized the study of crystal growth in the last several years, not only due to its high vertical resolution, but because of its ability to scan in ﬂuids, thus making it possible to monitor in situ dissolution and growth processes. AFM has been extensively used in crystal growth studies of macromolecular crystals [97], minerals [98–100], ionic crystals [101, 102], organic semiconductors [103], thin ﬁlms [104], hierarchical porous materials [105], and many other crystal systems. The use of AFM in studying synthetic zeolites and zeotypes has been slow to come, mainly due to two reasons: (i) the limitations of studying micrometer-sized crystals using a conventional top-head low magniﬁcation (< 20×) optics makes locating good crystals very time consuming; (ii) the fact that most zeolites crystallize at temperatures above 100 ◦ C. Nevertheless, the development of tip-scanning AFMs coupled with high-magniﬁcation (up to 100×) inverted optical microscopes plus new developments in temperature-controlled ﬂuid cells have immensely increased the range of observations that can be made using these systems. This is further illustrated in the six case studies presented in this chapter. 1.4.1.2 HRSEM Scanning electron microscopy is a well known and used technique for the characterization of microporous materials. It has long been used in conjunction with X-ray diffraction (XRD) as one of the main tools in the studies of crystal growth in zeolites [1, 22]. Its use though has been limited to study the habit, morphology, and size of the synthesized materials but not to characterize the surface detail. This is due to resolution limitations owing to excessive charging on steps at the crystal surface. In recent years, the development of low voltage ﬁeld emission electron sources for SEMs (FE-SEM) has signiﬁcantly reduced this issue. For example, in 2005 Wakihara et al. [89] reported SEM images of zeolite A where steps could be observed. This new kind of SEM has been dubbed HRSEM and has been shown to be able to observe steps as small as 1.2 nm with ease [68]. The use of this technique promises to open a new chapter in the study of crystal growth of nanoporous materials, complementing the AFM and offering nanometer-scale resolution in areas where an AFM tip cannot have easy access (such as twins, intergrowths, and rough surfaces). Some examples of the full potential of this technique are highlighted in some of the following case studies. 1.4.1.3 Confocal Microscopy During crystal growth macroscopic defect structures often form, leading to intergrowths, twins, and other more exotic extended structures. Knowledge of these

1.4 Techniques

internal structures of the crystals gives a lot of information about the growth mechanism. Very often these macroscopic growth features can be observed with the resolution of an optical microscope. By operating with confocal optics, often with additional ﬂuorescence from probe molecules selectively adsorbed within the nanoporous crystal, these macroscopic features can be illuminated [106, 107]. 1.4.2 Solution Chemistry – Oligomers and Nanoparticles 1.4.2.1 Nuclear Magnetic Resonance In order to understand how crystals grow, it is necessary not only to understand how the solid phase grows but also to understand the chemistry occurring in the solution phase. Unlike molecular crystals, such as those used in the pharmaceutical industry, the building units for nanoporous materials are in constant interchange in solution. The ephemeral nature of the species makes the task of unraveling, not only what is present in solution but also which are the rate-determining steps in the kinetics of this constant interchange, a very daunting task. The most powerful tool for speciation in the solution state is NMR, and this can be used to good effect to monitor 29 Si in silicates [108–119], 31 P in phosphates, and 19 F in ﬂuorides [120]. These are spin one-half nuclei and, as a consequence, tend to yield highly resolved spectra that are amenable to both one- and two-dimensional spectroscopy. By determining connectivities by INADEQUATE and COSY NMR, a large number of species have now been identiﬁed (Figure 1.10). The experiment must be carefully set up in order to ensure quantitation [121] and by inserting chemical probes the course of crystallization may be followed by monitoring pH by NMR [122]. 29 Si is only about 4% abundant and therefore two-dimensional spectra can take some time to acquire, precluding the ability to follow rapid temporal changes. This problem is further compounded, as spin half nuclei in solution often have long relaxation times that also substantially slow data acquisition. Nonetheless, 29 Si spectra of silicate solutions in particular can reveal an enormous amount of information on a plethora of species. Operated in a multinuclear fashion the complete chemistry of a crystallization such as silicoaluminophosphate (SAPO)-34 [123] may be followed. Quadrupolar nuclei, such as 27 Al [124] and 17 O [125], can also play a role in understanding solution chemistry. The spectra, although often less well resolved in terms of revealing multiple speciation, can be collected very rapidly and therefore dynamic information may be extracted. Care must also be taken regarding the degree of condensation of species, because as soon as nanoparticulates/colloids are formed the restricted motion of species starts preventing spectral averaging, which is vital for high-resolution spectra. Not only does the lack of motion broaden spectra but also a continuum of complex species with slightly different T–O–T angles will result in a continuum of chemical shifts causing a broadening that cannot be removed even by magic-angle spinning. However, this apparent downside of NMR can also be used to good advantage. For spin one-half nuclei such as 29 Si the spectral broadening is often not so substantial to render the resonances invisible. It is therefore very

17

1 Synthesis Mechanism: Crystal Growth and Nucleation

18

a

a •

1

2

3A

3B

a

4A

4B

b

5A

a

a 5D

5E

c

6D

6E

e d

a

b

b

a

c

e

d

d

6J

a

c e

f

a

b

c

d

b

7B c

7C

b

e

g b g a

c d b b

7H

d

f

d

a

c f

d

a 7I

c

e

d

b a e

8B

b

a 8C

c

f

7D

6H

a

b g

f

e

d

a

b

a 6L′

a

c

b

b c

b

a

b

c 6C

a

6L

a c

d

6G f

d c

b

e c

a

7A

b

c

d 6K

b

c

d

6B

d

b 6F′

f

b a

e d

c

d e

6F

a

a

a 5C b

c

a

a

b

a

8A′

6A′

b

b

c

b

d

7G

6A

c

5B

b

a

6I

b

b

c

b

b

c

c

5A′

a c

c

a

a 4D

4c

b

b

6M

b 7E

b d c

a

7F

c

a

c

7J

c a

c

b

c

b

d

a

7K

e d

a

b

a

e 8D

a 8A b

9A

a

i

f

b g b d 9B d h c

c

a

Figure 1.10 Aqueous silicate structures identiﬁed in concentrated alkaline solution by29 Si– 29 Si COSY NMR [116].

easy to distinguish between small oligomers, giving rise to sharp resonances, and nanoparticles giving broad resonances. Further, because for spin one-half nuclei quantitation of spectra is straightforward, it is a simple exercise to determine the relative concentration of oligomeric and nanoparticulate species [118, 119] (Figure 1.11).

1.4 Techniques

(f)

(e)

(d)

(c)

(b)

(a) −60

−70

−80

−90

−100 −110 (ppm)

−120

−130

−140

Figure 1.11 Normalized 29 Si NMR spectra of clear solutions with TEOS : TPAOH : H2 O molar ratio of 25 : 9 : x where (a) x 1/4 152; (b) x 1/4 400; (c) x 1/4 900; (d) x 1/4 1900; (e) x 1/4 4000; and (f) x 1/4 9500 [119].

1.4.2.2 Mass Spectrometry In order to increase sensitivity and temporal resolution, mass spectrometry is becoming increasingly used in the study of solution speciation. Modern mass spectrometers utilizing soft ionization procedures such as electrospray ionization can readily yield parent ions in relatively high mass/charge ratio. Although the speciation is not as unique as that determined by NMR, through isotopic distribution analysis the possibilities may be narrowed considerably. The enormous advantage over NMR is the sensitivity and rapid data collection that permits in situ analysis of dynamic events [126–132]. The power of this technique is most aptly demonstrated in a recent work on the interconversion between silicate oligomers [132]. In a clever experiment, a solution enriched with 29 Si and a solution of naturally abundant 28 Si containing cubic octamers were mixed. The mass spectrometry clearly revealed that the ﬁrst exchange species contained equal amounts of the two isotopes indicating a concerted exchange mechanism involving four silicon nuclei. A similar experiment with triangular prismatic hexamers showed a concerted exchange of three silicon nuclei (Figure 1.12). Such concerted exchange mechanisms are probably omnipresent in the chemistry of silicates and are likely to play an important role in zeolite crystal nucleation and growth.

19

20

1 Synthesis Mechanism: Crystal Growth and Nucleation 28Si

−

6O15H5

413

413

29Si

−

6O15H5

419 416

419

413

416

419

m /z

(a) 28Si

413 (d)

416

413

413

419

m /z 28Si 29Si O H − 3 3 15 5

−

6O15H5

Si6O15H5− 419

29

416

419

m /z

(b)

416

413

413 (e)

416

419

m /z

419 416

413 (c)

416

419

m /z

Figure 1.12 Temporal developments of mass spectra after mixing equally concentrated solutions containing the prismatic hexamer with naturally abundant silicon and 29 Si-enriched silicon [132].

1.4.2.3 Cryo-TEM The transition from solution to nuclei and ﬁnally to crystals is a complicated process to monitor experimentally. Scattering techniques such as dynamic light scattering and X-ray or neutron scattering are able to determine dynamically the presence of important nanoparticulates during these crucial early stages of the birth of crystals. However, recently a new and powerful technique has been added to the arsenal, namely, cryo-TEM. By rapid freezing of the growth medium the nucleation and growth processes may be stopped and the sample transferred, while it remains cold to be analyzed by the electron microscope. This permits high-resolution electron microscopy to be performed on more-or-less unperturbed crystallization media. Recent results on the silicalite system [133] prepared from TEOS and TPA show clearly that the initially formed 5-nm nanoparticulates are amorphous in nature before they agglomerate and crystallize by an intraparticulate re-organization. By performing these experiments in the electron microscope as opposed to utilizing x-ray scattering techniques it is possible to discern the presence or absence of structural order even at these very early stages.

1.4 Techniques

1.4.3 Modeling 1.4.3.1 Monte Carlo Modeling of Crystal Growth Since the advent of AFM and high-resolution scanning electron microscopy we have a new window to follow the nanoscopic details of crystal growth. In order to interpret these data new tools are required to model not only the crystal morphology but also the details of the surface topology. Monte Carlo techniques provide a possible route to simulation of crystal morphology and topology whereby the structure is developed according to a set of thermodynamic rules. The ﬁrst problem is to decide which unit to choose as the growth unit. For a molecular crystal the answer is straightforward as the indivisible element in the growth is a single molecule, and Monte Carlo techniques have been successfully used, for instance, for urea [134–136]. For nanoporous materials such as zeolites it is necessary to coarse grain the problem in a way that makes the calculation manageable. This is readily done by realizing that the rate-determining steps in the crystal growth process are related to closed-cage structures capped with Q3 groups at the surface. Open-cage structures exposing Q2 and Q1 groups are highly susceptible to dissolution and as a consequence do not persist at the surface. Having selected the coarse grain as a closed-cage structure, a network of closed cages is constructed and the probabilities for growth and dissolution determined according to the energetics at each site. The methodology chosen is essentially that of Boerrigter et al. [137] whereby each site at the surface of a crystal is assigned an energy relative to the bulk phase (see Figure 1.13 for description of energy levels). Growth from solution then occurs through an activated complex, essentially desolvation followed by adsorption, but the energy which deﬁnes the relative rate, growth etch or probability P, of growth and dissolution, Pi2j /Pi2j , is a combination of the energy of the site (termed site i2j in Meekes terminology), relative to the bulk phase, (Ui2j − U)and the supersaturation, µ (Eq. (1.11)) [138–140]. The ordering of the energy levels to the ﬁrst approximation follows the order of the nearest neighbor connectivity, with second-order connectivity resulting in smaller energy differences. growth

Pi2j

etch Pi2j

= exp(β(U12j − U) + βµ)

(1.11)

By treating this problem numerically by computer the real meaning of supersaturation and equilibrium becomes immediately apparent. The Monte Carlo treatment is conducted so that growth and dissolution events occur according to Eq. (1.11). The supersaturation may be treated as a constant or variable. By setting up a virtual solution phase which increases in concentration for an etch event and decreases in concentration for a growth event, the supersaturation is allowed to drop freely as the crystal grows. Equilibrium is established when the crystal stops growing and the number of growth and etch events over time are equal. At this point the value of µ is determined. Conventionally, µ is zero

21

22

1 Synthesis Mechanism: Crystal Growth and Nucleation

U moth U moth,effi2j For growth at this site ∆U i2j = U moth,effi2j – U i2j

U i2j

U cryst

∆U = Umoth – U cryst ∆U i2j – ∆U

Figure 1.13 Energy level descriptions for growth of crystal from solution. At left is the solvated growth unit in solution. At top is a fully desolvated growth unit in vacuum and at the bottom, the growth unit

fully condensed in the bulk crystal. Energy nomenclature corresponding to [137] is given, which relates to the probabilities for growth and dissolution given by Eq. (1.11).

at equilibrium as the conventional deﬁnition of supersaturation is the difference between solution concentration and that recorded at equilibrium. However, a much more powerful deﬁnition of equilibrium is achieved by considering the value of µ relative to the energies of each growth site, (Ui2j − U). At equilibrium some sites will be in undersaturation (the low-coordinate sites) and some sites will be in supersaturation (the high-coordinate sites). Equilibrium is just the balance point, the center of gravity, of energies of all the sites taking into consideration the number of each of those sites. The Monte Carlo approach ﬁnds this center of gravity equilibrium state that always lies within the range of the kink sites. It is not a problem that has a ready analytical solution because the number of sites of each type depends upon the speciﬁc connectivity of the crystal – in essence it depends on the crystallography of the crystal. For studies of nanoporous materials, this gives a route to study the problem experimentally without worrying about the extremely difﬁcult problem of determining conventional supersaturation of the solution. Crystals may be taken to the equilibrium condition and then compared against the Monte Carlo calculation for the same condition. Figure 1.14 shows examples of crystals simulated under different conditions.

1.5 Case Studies

2

3 Crystal size

1

Supersaturation (a)

(b)

Time

(c)

(d)

Figure 1.14 Typical crystal size and supersaturation evaluation as a function of time for zeolite A synthesis. Inset SEM picture (image size 2.5 µm) shows typical zeolite A crystal produced in the synthesis. b–d show about 0.3 × 0.3 × 0.3µm3 sized zeolite A crystals simulated at time interval 1, 2, and 3, respectively [138].

1.5 Case Studies 1.5.1 Zeolite A

Zeolite A is one of the most widely used zeolites due to its ion-exchange capabilities [141]. Zeolite A has a framework structure known as Linde type A (LTA) [142] consisting of sodalite (SOD) cages linked via four rings, and an Si to Al ratio of 1 : 1. The SOD units join to produce an α-cage with a diameter of 11.4 A˚ (the large cavity in the center of the structure), and two channel systems which are connected to allow motion of the Na+ ions and water molecules. Its empirical formula is Na12 [Al12 Si12 O48 ]·216H2 O. Investigations on zeolite A have been numerous and varied with a multitude of studies focusing on the effect of different parameters in its growth, such as the addition of organic molecules to the synthesis mixture [143, 144], seeding [145], aging [37], and clear solution [25]. Zeolite A has also been the subject of various AFM studies [67, 146, 147], including the ﬁrst in situ dissolution study in a zeolite [87]. These observations were limited to the study of the {100} face in crystals extracted at the end of the synthesis (i.e., a low supersaturation condition), with no information available on

23

1 Synthesis Mechanism: Crystal Growth and Nucleation

the crystal growth mechanisms under other conditions. A more thorough study has been carried out to study the evolution of the surface features of zeolite A at different synthesis times by means of AFM and HRSEM. Zeolite A crystals were prepared following the preparation of Thompson and Huber [148]. The synthesis was carried out at 60 ◦ C and at the following synthesis times: 2.5, 4, 8, 20, 30, and 50 hours. A second preparation following the method by Petranovskii et al. [144] includes the addition of diethanolamine (DEA), an organic molecule whose main effect is to increase the size of the crystals and which also has an effect on the morphology of the crystals. In this experiment, the reaction was carried out at 90 ◦ C and the synthesis times were 12, 16, 20, 24, 28, 32, 36, 40, 44, 48, 72, 96, 120, 168, 336, and 504 hours. The reason behind the much longer experimental times was to observe any possible change in surface topography at long ‘‘equilibrium times’’ in order to allow for surface and habit rearrangement. 1.5.1.1 Thompson Synthesis Figure 1.15 shows the evolution of zeolite A synthesized as a function of time. It can be seen that after 8 hours the crystallization is almost complete. At 2.5 and 4 hours the growth rate of the crystals is at its maximum, which also corresponds to the highest supersaturation achieved in the system (Figure 1.9). At 8 hours, however, the growth rate has probably started to decrease, reaching almost zero at 20 hours. Figure 1.16 shows the corresponding HRSEM of crystals extracted after 2.5, 4, 8, and 20 hours, whereas Figure 1.17 shows the AFM images after 4, 8, and 20 hours. When comparing both images a more detailed picture of the crystallization process emerges. At 2.5 hours (Figure 1.16a) it can be seen that crystals are small (a few nanometers to 400 nm) and rounded, and there is also strong evidence of intergrowth formation and aggregation of crystals. Although no AFM images could be obtained on these round crystals, the HRSEM resolution allows us to see that the crystal surfaces are very rough. All these features are evidence of an adhesive mechanism of growth typical of high supersaturation conditions. At 4 120 100 % Zeolite A

24

80 60 40 20 0

0

5

10

15

20

Time (h) Figure 1.15 Percentage of zeolite A crystallized as a function of time. Superimposed is a theoretical growth curve for this kind of system.

25

1.5 Case Studies

(a)

SU - 7401F

(b)

LEI

1.0kV ×75,000

100nm WD 8.0mm

(d)

SU - 7401F

SU - 7401F

1 V ×20,000 1µm

LEI

WD 8.0mm

(c)

LEI

1.0kV ×15,000

1µm

WD 8.0mm

SU - 7401F

LEI

1.0kV ×15,000

1µm

WD 8.0mm

Figure 1.16 HRSEM photomicrographs of zeolite A crystals after (a) 2.5 hours; (b) 4 hours; (c) 8 hours; and (d) 20 hours of synthesis.

hour HRSEM (Figure 1.16b) and AFM (Figure 1.17a) images show a similar picture. At this point of time, crystals are much bigger (up to 1 µm) and still somewhat rounded but they start to show the formation of facets. This can also be seen in the AFM image (Figure 1.17a) where steps start to be visible, although evidence of active 2D nucleation is also present. Eight hours into the synthesis the crystals have grown even more, reaching sizes up to 1.5 µm. HRSEM images (Figure 1.16c) show that the crystals are more faceted than before, although still with rounded edges. The corresponding AFM image (Figure 1.17b) shows well-deﬁned steps and terraces with nucleation limited to the central terrace. This indicates that supersaturation has now dropped and growth takes place by a layer growth or single nucleation mechanism (Section 1.2.12). At 20 hours the crystals are fully developed (Figure 1.16d) and show the typical faceted morphology for this synthesis [149] bounded by {100} and {110} and {111} faces. AFM taken on these samples on the {100} face (Figure 1.16d) agrees very well with what has been published before [67, 89] and displays single square terraces at the center of the face that grow by step advancement toward the edge of the crystal. The steps are almost perfectly rectilinear as well. Figure 1.16c shows the AFM image of a {110} face; it can be seen that in this face the shape of the terrace is rectangular, displaying a faster growth along the <100> direction and slower on the <110>. This situation is indicative of a very low supersaturation and has been successfully replicated in Monte Carlo simulations.

25

26

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

(b)

0.5 µm (c)

0.2 µm (d)

0.2 µm

0.2 µm

Figure 1.17 AFM deﬂection images of zeolite A crystals after (a) 4 hours; (b) 8 hours; and (c and d) 20 hours of synthesis. a, b, and d show {100} faces, whereas (c) shows the {110} face.

1.5.1.2 Petranovskii Synthesis Figure 1.18 shows two SEM micrographs of the end product of the synthesis of zeolite A in the presence of DEA at 90 ◦ C. The end product in this case displays a much larger size (up to 15 µm) and the crystals are bound only by the {100} and {110} faces. Also the relative size of the {110} face in comparison to the {100} face is higher for this synthesis. A detailed AFM study was undertaken to analyze the surface topography of all the samples synthesized. Figure 1.19 shows eight AFM images taken on crystals extracted at 44, 48, 72, 96, 120, 168, 336, and 504 hours after the start of the synthesis. At 44 hours (Figure 1.19a), it can be seen that the crystal surface is still quite rough, indicating that 2D nucleation is still happening at a faster rate. Still SEM analysis on the evolution of the length of the crystals with synthesis time shows that at this time the reaction is almost complete, so supersaturation has already started to decrease. This is reinforced by the fact that at 48 hours, 2D nucleation rate has decreased to the point that the steps and growing nuclei can easily be resolved with the AFM (Figure 1.19b), although nucleation still occurs on narrow terraces. At 72 hours, AFM analysis (Figure 1.19c) shows fewer nuclei at

1.5 Case Studies

(a)

HV Mag WD Det Spot Pressure --30.0 kV 5000x 8.3 mm ETD 3.0

(b)

20.0µm J111

HV Mag WD Det Spot Pressure --30.0 kV 12000x 9.8 mm ETD 3.0

10.0µm J112

Figure 1.18 SEM photomicrographs of zeolite A crystals extracted after 120 hours of synthesis.

the surface in accordance with the gradual decrease in supersaturation expected at this point of the synthesis. Note that all the terraces are bounded by straight steps with sharp corners. Figure 1.19d–f show the central area of different {100} faces. In the three images a similar situation is observed, where nucleation has almost stopped but the terraces still possess sharp corners. Further into the synthesis, at 336 hours, there is a change in the shape of the terraces, as can be seen in Figure 1.19g. Now the corners of the square terraces are not sharp but curved. At 504 hours, terraces were found to be also curved at the corners (Figure 1.19h). This effect can be explained by means of the Monte Carlo simulation and corresponds to the equilibrium situation explained in Section 1.4.3.1. Since the ‘‘center of gravity’’ equilibrium state is located around energies of the kink sites, steps would ‘‘evolve’’ toward a shape composed mainly of these sites, resulting in some dissolution and rounding of the terrace corners. Zeolite A has been dissolved in situ using a mild sodium hydroxide solution [87, 150]. Under in situ conditions, it is often possible to capture the less-stable surfaces, which tend to be absent if the crystal is removed from the mother-liquor. All ex situ measurements on zeolite A have revealed a terrace height of 1.2 nm. However, these terraces dissolve in two steps according to different mechanisms and on different timescales. Figure 1.20 shows an AFM image selected from an in situ series which captures both structures. A 0.9-nm terrace is observed dissolving by terrace retreat and a 0.3-nm terrace remains undissolved on the same timescale. The explanation of these results is that SOD cages, 0.9-nm high, which are interconnected, dissolve in a correlated fashion by terrace retreat. One cage must dissolve before the next one can. Single four rings, 0.3 nm, dissolve on a different timescale in an uncorrelated fashion as they are not connected to each other. These results further reveal the importance of closed-cage structures, two of which are seen in this experiment, the SOD cage, and the double four ring.

27

28

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

(b)

1 µm (c)

0.5 µm (d)

0.5 µm (e)

0.5 µm (f)

1 µm (g)

1 µm (h)

0.5 µm

0.5 µm

Figure 1.19 AFM deﬂection (a, b) and height images (c–h) of zeolite A crystals after (a) 44 hours, (b) 48 hours, (c) 72 hours, (d) 96 hours, (e) 120 hours, (f) 168 hours, (g) 336 hours, and (h) 504 hours.

1.5.2 Silicalite

Silicalite is often used in fundamental studies of zeolites owing both to the importance of the MFI structure for catalysis through ZSM-5 and also because of the simpliﬁcation afforded by having a pure silica framework with no charge-compensating

1.5 Case Studies

Figure 1.20 4.3 × 4.3 mm2 deﬂection AFM micrograph of a zeolite A crystal under static solution of 0.5 M NaOH from in situ measurement after 33 minutes. Inset blue area shows 0.9 nm terrace and white area 0.3 nm terrace.

cations. From a crystal growth point of view it also presents an opportunity to investigate (i) the role of templates as the tetrapropyl ammonium cation is such a strong structure director and (ii) the role of intergrowth formation to the MEL, silicalite-2, structure. Both MFI and MEL structures are composed of connected pentasil chains that can be connected either via a mirror plane or an inversion center. Different choices result in the two structures and this is most readily controlled via structure directing agents that register with the resulting channel system. All AFM measurements to date, on either system, show that the fundamental growth step height is 1 nm, consistent with the pentasil chain unit, thereby conferring an important status upon this closed structure on the growth mechanism. In a study on the effect of supersaturation, a series of silicalite samples were prepared in a semicontinuous synthesis [151]. Untwinned seed crystals in their spent mother-liquor were placed in a continuous feed reactor. After the slurry reached equilibrium, the nutrient feed was switched on and varied to maintain a constant crystal growth of 0.4 µm h−1 . The reaction was continued for a total of 64 hours. During the reaction time, the nutrient feed was stopped for periods of about 16 hours each at 4.3, 9.6, and 16.4 hours, and the supersaturation allowed to drop such that the growth of the crystal more-or-less ceased. The AFM was recorded on each of these crystals, and this is shown in Figure 1.21. During the periods of constant growth the supersaturation level is relatively high and the surface of the crystal is characterized by having a high density of growth nuclei. The crystal grows exclusively via a birth-and-spread mechanism. As the supersaturation drops, when the nutrient feed is turned off the terraces continue to spread but surface

29

30

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a) 0

(b) 0.3

(c) 1.7

(d) 2.8

(e) 3.5

(f) 5.1

(g) 7.6

(h) 9.6

(i) 9.6*

(j) 11.8

(k) 13.3

(l) 14.9

(m) 16.4

(n) 16.4

Figure 1.21 AFM deﬂection images of {010} faces of the silicalite samples. Crystal shown in (a) corresponds to the seed crystal; (b–h) and (j–m) are crystals that were recovered from the reactor under continuous feed. Crystals shown in (i) and (n) were recovered

from the reactor after periods of 16 hours when the nutrient feed was switched off and are denoted by 9.6* and 16.4* hours. Time is expressed in hours. The scale bars represent 1 mm [151].

1.5 Case Studies

nucleation more-or-less ceases. The reason for this is that surface nucleation requires the highest energy, see Eq. (1.11), or highest value of supersaturation. growth etch , will be Near equilibrium the relative rate of growth to dissolution, Pi2j /Pi2j less than unity, and consequently as soon as an entity grows on the surface, it will immediately dissolve back into solution. Conversely, terrace spread is dependent growth etch upon lower coordinate, edge and kink sites for which Pi2j /Pi2j will be closer to or above unity permitting the terraces to continue spreading. The process is reversible as surface nucleation and crystal growth continue when the supersaturation is subsequently raised. This supersaturation control experiment demonstrates the ability to switch on and off speciﬁc crystal growth processes. Such a phenomenon could be used to control defects and intergrowths in zeolites. Zeolite intergrowths in zeolites often result from layer growth, whereby a new layer has more than one choice with similar energy (e.g., an ABA stacking rather than an ABC stacking). If there is a high density of surface nuclei present on any given surface, then there occurs a high probability that some of these are of the slightly less favored stacking sequence. Now, one given layer will have a number of C layers mixed in with the A layers, and as the terraces spread a defect will result when they merge as they will be incompatible. A possible route to overcome such defects would be to lower the rate of nucleation, by lowering the supersaturation such that terrace spreading is still very rapid, but the low nucleation density means that the probability of having nuclei of a different stacking sequence is minimized. Hence, the defect density should decrease. Conversely, by working at very high supersaturation the nucleation density will be high and the probability for stacking sequence incompatibilities will also be high leading to a high defect density. Nanoporous materials such as faujasite (FAU), BEA, and ETS-10 grow according to such rules. In silicalite, stacking sequence problems arise when pentasil chains attach by an inversion symmetry rather than a mirror symmetry. This will result in a switch from the MFI to the MEL structure when the mistake is on the (100) face, but on the (010) face the pentasil chain cannot connect to the ensuing crystal. This slows crystal growth, and large terrace fronts stack up along the foreign pentasil chain until eventually the defect is overgrown leaving a high density of undercoordinated Q3 silicons within the structure [81]. Lessons on the role of templates can be learned through competitive templating. TPA is known to be a strong structure director for the MFI structure. By increasing the length of the hydrocarbon chain by one unit tetra butylammonium (TBA) cations will be directed toward the MEL structure. However, the TBA cation is too large to ﬁt at every channel intersection, and consequently there is a topological dilution of the template at the growing crystal surface. The effect of this is immediately apparent in mixed TPA/TBA preparations. Greater than 90% TBA is required in the synthesis before there is a substantial incorporation of TBA over TPA with a corresponding substantial switch from MFI to MEL. Indeed even with 98% TBA and 2% TPA there is still a substantial frustration in the growth of the MEL structure (Figure 1.22). In order to have a smoother transition from MFI to MEL, a smaller template is required and N, N-diethyl 3,5-dimethyl piperidinium iodide

31

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

2% TPA 98% TBA

(b)

2% TPA

2 µm

400 nm

5% TPA 95% TBA

(c)

(d)

5%

2 µm (e)

1 µm 10% TPA 90% TBA

(f)

2 µm

1 µm 10%

(g)

(h)

0 −50 Height / nm

32

1 µm

−100 −150 −200 −250 0

1

Figure 1.22 HRSEM and AFM images of MFI/MEL intergrowths showing frustration of MEL growth by the incorporation of as little as 2% TPA in the growth swamped with TBA.

2 3 Length /µm

4

5

1.5 Case Studies

(DEDMPI) serves this role. DEDMPI can be accommodated at every intersection and produces MEL crystals that are much less susceptible to stacking sequence problems and the inherent defects. 1.5.3 LTL

Zeolite L has a unidimensional channel structure that typically grows as long hexagonal prismatic crystals with the channels running along the long axis. From a catalytic point of view this is undesirable as the intracrystalline path length for reactants and products is maximized resulting in restricted diffusion. As a consequence, there have been a number of attempts to modify the normal habit to yield tablet-like crystals with a short c-dimension. AFM studies of zeolite L [152] reveal the reasons for this crystal habit. Figure 1.23 shows the AFM image recorded on both the hexagonal (001) face and the sidewall (100) face of the crystal. On the top (001) facet the crystal grows via a layer growth, and the smallest height of the terrace is equivalent to the height of one cancrinite cage. The sidewalls show more interesting behavior. Long, thin, straight terraces are observed which are elongated in the c-direction of the crystal. Two signiﬁcantly different heights are measured for these terraces. The narrowest terraces are always 1.2 nm in height

0.0

b

1(i)

0.5

1.0

(ii)

(iii)

(iv)

(ii)

(iii)

(iv)

0.0

a

1.0 2(i)

c

Slow (µm)

1.0

0.5

0.0 0.0

0.0 Fast (µm)

Figure 1.23 Error signal AFM images of zeolite L with different aspect ratios. (1) The hexagonal face down the [001] direction of the crystal and (2) the sidewalls down the [100] direction of the crystal. For each face, (i) shows a schematic framework of the crystal, and (ii), (iii), and (iv) crystals with aspect ratios 1.5, 2.3, and 5.1, respectively.

33

34

1 Synthesis Mechanism: Crystal Growth and Nucleation

and the wider terraces are always 1.6 nm in height. The reason for this is that all the narrowest terraces correspond to a single cancrinite column, which grows very rapidly along the c-direction of the crystal but is highly frustrated to growth across the side-wall in the a- or b-direction. Wider terraces are 1.6-nm high because neighboring cancrinite columns are connected by a further cancrinite column that acts as a bridge across the large 12-ring channel. This result illustrates a generic problem to grow crystals with large pores. Circumventing the large pores can result in very unfavorable processes which are more-or-less akin to fresh nucleation. The bridging cancrinite column is more likely to dissolve back into the solution rather than persist until it is secured by the cancrinite column at the other side of the bridge – again according to Eq. (1.11). The problem lies in the fact that two, not one, cage structure is required to circumvent the large 12-ring pore. Growing both cancrinite column structures before one dissolves is an unlikely event, and therefore the kinetics are slow. The traditional way to build nanoporous structures over large void spaces is to add an organic templating agent which facilitates the process; in other words the kinetics for the process are improved. Zeolite L, however, is an example of a wide-pore zeolite that grows readily in the absence of organic-structure-directing agents. Consequently, zeolite L is a good demonstration of how large pore structures can be incorporated into a structure without expensive organic additives, and at the same time this system illustrates where the kinetics are severely frustrated in the process. This frustrated growth also results in the typical long prismatic crystals and the kinetics would need to be adjusted in order to successfully grow low-defect tablet-shaped crystals. Most methods reported to create shorter c-dimension zeolite L crystals operate under high supersaturation conditions, where the crystals have a high density of defects and the aspect ratio is only altered as a result of the interrupted growth at macro-defects. Careful adjustment of crystal aspect ratio, while maintaining a low defect density, is yet to be achieved. Zeolite L is also a very interesting system to investigate, in situ by AFM, the mechanism of dissolution. Figure 1.24 shows a series of images recorded as a function of time as the crystal dissolves under mild basic conditions. The micrographs have been recorded in lateral deﬂection mode, which monitors lateral twist of the cantilever during scanning. This mode is normally used for nanotribology studies as local friction will cause a twist of the cantilever. Three things are immediately apparent from the images. First, the terraces dissolve very rapidly along the c-direction of the crystal and very slowly in the lateral direction. Second, the place where the crystal dissolves is bright white, indicating a high degree of lateral twist. In essence, the AFM illuminates where the chemistry on the crystal occurs. This twist is also observed during the growth of nanoporous crystals. Third, it is observed that the bright white region on the dissolving terrace is substantially larger at the top of the terrace when the AFM tip is scanning down the crystal and vice versa. This demonstrates that the tip is aiding the dissolution. The tip in effect warms the crystal. By recording the images at different cantilever loads and temperatures it is possible to make a series of Arrhenius plots, which when extrapolated to zero load (i.e., no effect of the tip) yield the activation energy

1.5 Case Studies (i)

2.0

(ii)

(iii)

(iv)

(v)

(vi)

(vii)

(viii)

(ix)

Slow (µm)

1.5

1.0

0.5

0.0 0

500 Fast (nm)

Figure 1.24 Atomic force micrographs of the (100) face of zeolite L during dissolution in 0.2 M NaOH. The image in (i) shows a vertical deﬂection micrograph at the beginning of the experiment and ii–ix show lateral deﬂection micrographs on subsequent scans over the crystal surface. The time between

each image was approximately 4 minutes. The white ‘‘lights’’ indicate a change in friction experienced by the AFM tip, shown as a high contrast change on the image. The white arrows indicate the direction of scanning.

for this fundamental dissolution process, which is 23 ± 6 kJ mol−1 . The cause of the high lateral force is not entirely clear. It could be due to a change in friction when the crystal is undergoing rapid dissolution. However, it also might be due to high local energy being imparted to or from the tip as a result of a large energy change during dissolution. This latter explanation would be consistent with the fact that the phenomenon is observed during both growth and dissolution. 1.5.4 STA-7

STA-7 is a SAPO [153] with the structure type SAV [142]. SAV belongs to a group of four framework structures, composed of double 6-rings (D6Rs), which also includes the frameworks CHA, AEI, and KFI. The only difference between these structures is the arrangement of the D6R along the x, y, and z axes [154]. The SAPO STA-7 structure belongs to the space group P 4/n and contains two types of cages,

35

36

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

Y X (b)

Y Z

X

Figure 1.25 (a) STA-7 structure on the {001} face showing the position of the cyclam and TEA molecules inside the two different cages. The four different orientations of the D6R are also highlighted in different colors. (b) 3D view of the STA-7 structure. The four possible orientation of the D6R are

highlighted in red, green, black, and blue. The unit cell is highlighted in bold. It can be seen how the units alternate orientation along x and y directions but not along the z direction. A simpliﬁed color model of the structure is also shown for comparison.

1.5 Case Studies

A and B, connected three dimensionally by eight-ring windows. The larger cage, B, is templated by cyclam and the smaller one, A, by the co-template tetraethylammonium (TEA), as shown in Figure 1.25. In the structure, the D6R units have four different orientations (highlighted in different colors in Figure 1.25) and are related to each other by fourfold symmetry axes and an n-glide plane perpendicular to the <001> direction. Hence, alternating D6R units can be found along the <001> and <010> directions. In contrast, D6Rs form chains along the <001> direction. Consequently, a unit cell consists of two D6Rs along x and y axes, and one along the z axis, as can be seen in Figure 1.25b (highlighted in bold). For simplicity the STA-7 structure can be represented (hereinafter) as a series of cubes of different colors, each one representing a D6R of different orientation (Figure 1.25b). STA-7 crystals were prepared from SAPO-based gels treated hydrothermally (190 ◦ C for 3 and 10 days). Further details on the characterization can be found in Castro et al. [153], but it is important to point out that the formation of STA-7 has proven to be dependent on the presence of the two templates (cyclam and TEA). SEM observations at the end of the synthesis reveal the formation of crystals with a well-deﬁned tetragonal prismatic morphology and with a typical size of 30–35 mm (Figure 1.26a). Therefore, the crystals are bound by two crystallographically distinct (a)

50 µm (b)

(c)

x

y z

x

Figure 1.26 (a) Scanning electron micrograph of STA-7 crystals after the end of the synthesis. (b) Optical micrograph of an STA-7 crystal showing the {100} face. (c) Optical micrograph of the {001} face.

37

38

1 Synthesis Mechanism: Crystal Growth and Nucleation

faces, {100} and {001} (Figure 1.26b and c, respectively). Both types of faces were characterized by ex situ AFM. 1.5.4.1 {001} Faces Figure 1.27a shows a representative AFM image of a {001} face. It can be seen that the surface is covered by multiple nearly isotropic spirals, with all of the scanned surfaces showing a similar dislocation density of approximately 1–2 dislocations per 10 µm2 . No evidence of 2D nucleation was observed, indicating that the system was near to equilibrium when the crystals were extracted. The isotropic morphology of the spirals in this face indicates that there are no preferential growth directions at low supersaturation. Height analysis shows that the step height at the dislocation core (Burgers vector) is always 0.9 ± 0.1 nm (Figures 1.27b,c), which corresponds to the d001 spacing, that is, to the height of a D6R along the <001> direction. Figure 1.8d shows a simpliﬁed block diagram of the STA-7 structure, with the different orientations of the D6R highlighted in different colors as described in Figure 1.25b. Rows of similarly oriented D6Rs run parallel to the z direction. The original step structure produced by the dislocation is shown by the lightly colored blocks, assuming that the dislocation also runs parallel to the z axis with a Burgers vector of 0.9 nm. It can be seen that the sequence of D6Rs on both sides of the dislocation does not change with respect to the normal sequence (green–red–green in the diagram), that is, the bonding through the dislocation between the D6R units is the same as in the undisturbed (nondefective) crystal. Therefore, growth units (assumed to be a D6R) could attach to the step created without modifying the alternation sequence of D6R units along any direction, as represented by the darker colored blocks, perpetuating the STA-7 structure through spiral growth. 1.5.4.2 {001} Faces AFM observations on the {100} faces reveal the formation of two very distinct types of spirals. The ﬁrst, which is the more numerous one, has an elongated shape, with the long axis, or faster growth direction, parallel to the <001> direction. Height analysis shows that this type of spiral is produced by a dislocation with a Burgers’ vector of ≈ 0.9 nm (Figure 1.28a). The second type corresponds to an interlaced spiral (Figure 1.28b,c). The step splitting in this interlaced spiral produces the characteristic ‘‘saw tooth’’ pattern (Section 1.2.11) extending from the dislocation, which in this case is parallel to the <100> direction (highlighted by the white box in Figure 1.28c). Height analysis along this pattern reveals that the height of the steps is half a unit cell, that is, 0.9 ± 0.1nm (red line). On the contrary, a cross section from the dislocation center but parallel to the <001> direction (blue line) reveals that the most common step height is one unit cell, that is, 1.8 ± 0.1 nm (i.e., two monolayers). Figure 1.28b shows a higher resolution image of the spiral center, where it can be seen that two single steps (0.9 ± 0.1 nm) emanate from the dislocation, hence the Burgers vector of the dislocation is equal to a unit cell, that is, 1.8 ± 0.1 nm. As explained in Section 1.2.11, an interlaced spiral is produced when each of the different monolayers emanating from the dislocation possesses a different speed anisotropy. In the case of STA-7 this anisotropic growth

1.5 Case Studies

(a)

<100>

2 µm

<010> (b)

(c)

3

3

2

2

nm

nm

826.9 nm

1

1

0

0 0

400 nm

800

0

200

400

nm

(d)

Figure 1.27 (a) AFM amplitude image of a {001} face of an STA-7 crystal. Isotropic spirals in the surface are clearly visible. Inset shows the corresponding optical microscopy image. (b) AFM height image of the lower spiral and cross section. (c) Height image

of the upper spiral and cross section. The cross sections were taken along the red and blue lines in the images. (d) Simpliﬁed spiral structure after the two new layers have grown. The structure of the newly grown layers matches that of the underlying substrate.

39

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

(b)

<001>

0.2 µm

0.5 µm

3 2 1 0

3 2 nm

nm

40

0

300 nm

600

1 0 0

nm

500

(c)

<001>

<100>

Figure 1.28 (a) AFM height images of an elliptical spiral with cross section. (b) Height image showing details of the central area of the interlaced spiral shown in (c). Two substeps emanating from the dislocation are clearly seen. The cross section conﬁrms

1 µm

the single layer (i.e., 0.9-nm high) nature of these steps. (c) Amplitude image of an interlaced type spiral. Highlighted box contains the characteristic ‘‘sawtooth’’ pattern observed in these types of spirals.

1.5 Case Studies

is symmetry-induced due to the presence of an n-glide plane perpendicular to the {100} face. Recently, van Enckevort and Bennema [14] demonstrated that interlacing will be expected when a screw axis and/or a glide plane are perpendicular to the growing surface. The presence of the n-glide plane also determines the shape of the spiral to be symmetrical at both sides of the <100> direction, which marks the intersection of the n-glide plane with the {100} face. Also, the spiral is symmetrical along <001> direction owing to the symmetry axis. By taking into account these symmetry constraints, it is possible to deconstruct the growth anisotropies of the individual substeps. In the case studied, the two substeps show a difference in the growth rates along [001] and [001] directions. One substep will grow faster along the [001] direction than along [001], whereas the other will be opposite. This situation is summarized in Figure 1.29. Figure 1.29a,b shows a simpliﬁed diagram of the two substeps as they would grow if no interference occurred. The anisotropic growth along <001> direction for each substep is clearly evident. In Figure 1.29c, the trajectories of the two spirals are superimposed, demonstrating clearly how they can form the observed interlacing pattern (Figure 1.28c). The cause of the anisotropic growth can be explained by the tilting of the D6R units with regard to the <001> direction. This tilt can be seen more clearly in Figure 1.30a, which shows a cross section of the STA-7 structure perpendicular to <100> direction. In one layer, D6R units are tilted toward [001] direction, whereas on the following they will all be tilted in the opposite direction, [001]. This tilt creates two different step geometries, one acute (red box) and one obtuse (blue box), one of which may favor the attachment/docking of the template preferentially over the

(a)

(b)

Figure 1.29 Simpliﬁed diagrams showing the formation of an interlaced spiral. a and b The assumed shape of each substep if they could grow freely. (c) The overlapping of the two substeps. Here the interlaced pattern is readily observable.

(c)

41

42

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

(b)

Figure 1.30 (a) Simpliﬁed cross section perpendicular to <100> direction, showing the step structure in a {100} face. The two distinct step geometries, acute and obtuse, are highlighted in red and blue. (b) Cross section of the acute step showing the most stable docking conﬁguration for the cyclam molecule after simulation.

other and hence favor the growth along one direction. A similar situation has been observed in calcite crystals where steps are also nonequivalent and ion sorption depends on their geometry [155]. To test this hypothesis, template adsorption at the surface was simulated using an adapted version of the ZEBEDDE program [156] to perform a Monte Carlo simulated annealing (MCSA) [157]. The template adsorption energies (nonbonding) from the simulations suggest that while TEA can adsorb at all adsorption sites with approximately equal energy, the larger cyclam can only adsorb favorably into the large cage site on the ‘‘acute’’ side of the step. Therefore, it is presumed that growth will be favored on the acute step side, where the cyclam accelerates the rate of growth unit attachment relative to the TEA-only mechanism on the obtuse step. As the position of the acute steps alternates between layers as the {100} face grows, the fastest growth direction follows, creating the interlaced pattern. Figure 1.30b shows a detailed cross section of the step structure with the cyclam molecule attached in its more favorable position.

1.5 Case Studies

1.5.5 Zincophosphates

Zincophosphate open-framework materials are a class of zeotype materials. In some cases, they show framework types same as those found on zeolites, such as SOD [158] and FAU [159]. However, some possess unique framework types, such as the chiral zincophosphate (CZP) framework, which have no aluminosilicate analog. The zinc phosphates with SOD and FAU structures were ﬁrst synthesized by Nenoff et al. [158] under very mild pH and temperature conditions. However, the synthetic conditions are much milder in the case of the zinc phosphates [160]. These milder conditions are particularly well suited for performing in situ experiments on the AFM, as compared to aluminosilicate zeolites. Following results from in situ growth experiments on ZnPO-SOD and ZnPO-FAU are discussed. 1.5.5.1 ZnPO4 -Sodalite SOD zinc phosphate, which is synthesized at temperatures ranging from room temperature to 50 ◦ C [158], has a primitive cubic framework, with a unit cell constant a = 0.882 nm and belongs to the P-43n space group. The system contains a 1 : 1 mixture of tetrahedral zinc and phosphorous units that alternates within the framework, giving a stoichiometry of Na6 (ZnPO4 )6 ·8H2 O [158]. ZnPO-SOD was synthesized following the original room temperature recipe by Nenoff [158]. This produced highly intergrown crystals which were not suitable for AFM experiments. To solve this issue, the synthesis was performed at 6 ◦ C with the goal of decreasing growth and nucleation rates, which may produce better quality single crystals. The synthesis did produce single crystals, as well as intergrowth and the ZnPO-CZP. SOD crystals had dimensions ranging from a few microns up to 15 µm. Crystals were bound by {100}, {110}, and {111} faces. These crystals were attached to a resin and brought in contact with low supersaturated solutions. In situ experiments performed on the {100} face in contact with low supersaturated solutions revealed the formation of spirals. Figure 1.31a shows one of these spirals. The angles between steps are slightly distorted since the growth rate is too fast for the scan speed to catch up. The overall morphology observed should be a square. Still it can be clearly observed that the spiral formed is of the interlaced type, such as those observed on STA-7, with two monolayers spreading out from the dislocation. In this case, however, the splitting has a fourfold symmetry, which, of course, agrees with that of the ZnPO-SOD crystal studied. Also, it has a polygonal shape, contrary to the rounded contours on the STA-7. Height analysis reveals that each monolayer height is about 0.45 nm, which corresponds to half a unit cell of ZnPO-SOD. Figure 1.31b shows the simpliﬁed 3D structure for ZnPO-SOD where a monolayer (highlighted) has started to grow. The real insight on the crystal growth process that can be achieved by monitoring the process in situ is highlighted in Figure 1.32. In this ﬁgure, a sequence of lateral force AFM images shows two interlaced spirals growing, and the interlacing-inducing growth anisotropy for each substep can be clearly observed. In Figure 1.32a the two spiral centers are highlighted by white circles. Because of the

43

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

y

x 0.25 µm

Height (nm)

44

1.8 1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 0.00

0.05

0.10

0.15

0.20

0.25

0.30

Offset (µm)

z

(b)

y x

Figure 1.31 (a) AFM deﬂection image showing interlaced spiral growth on a ZnPO-SOD crystal and associated cross section. (b) Simpliﬁed 3D structure of ZnPO-SOD showing the 1/2 unit cell step (highlighted in bold).

1.5 Case Studies

(a)

(b)

y x

0.25 µm (c)

(d)

(e)

(f)

Figure 1.32 Sequence of lateral force AFM images showing the growth of two interlaced spirals.

way in which both spirals interact; two clear square areas are created in between. A white arrow in Figure 1.32a signals the position of a step which can be seen advancing in the following images. The advancing speed of this step is much higher along <010> direction than along <100>. On the contrary, the next step to growth on top (highlighted with white arrow in Figure 1.32c) advances much faster along <100> direction than along <010>, as can be seen on Figure 1.32d–f. Figure 1.33 shows the two interlaced patterns of each step as if they could grow freely (in a similar fashion as in Figure 1.29 for STA-7). It can be seen that the shape of each step would be rectangular because of the anisotropic growth, and how successive

45

46

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

y x (b)

z x (c) Figure 1.33 (a) Theoretical free development of steps in the absence of interference. (b and c) Simpliﬁed structure of ZnPO–SOD showing three different steps (highlighted in different colors) at various stages of growth.

steps will alternate between fast and slow directions. The interference pattern created by two such substeps is the actual pattern observed in the experiments. The reason behind this anisotropic growth is not fully understood, but it may have to do with the alternation of Zn and P positions in the ZnPO-SOD and the rates of condensation of the two different elements into the structure. Figure 1.33b shows a schematic ZnPO-SOD structure where three monolayers have grown (highlighted in three different colors). The red and blue colors represent Zn and P tetrahedra, respectively. Figure 1.33c shows a cross section of the structure, highlighting the fact that each monolayer is of half unit-cell height. Looking at the top monolayer (shaded in blue) in Figure 1.33b, it can be seen that the top Zn tetrahedra in the monolayer (inside white circles) are oriented in a line along <100> direction, whereas in the underlying monolayer (shaded in pink) they lie in a direction parallel to <010>. Correspondingly, the top P tetrahedra in each monolayer also alternate position. If the rate-determining step in the formation of a half SOD cage (necessary for step advancement) depends on the identity of the

1.5 Case Studies

atom in the structure to which it bonds a difference in growth rates as a function of direction can be envisioned. 1.5.5.2 ZnPO4 -Faujasite FAU zincophosphate was synthesized for the ﬁrst time by Gier and Stucky [161]. ZnPO-FAU belongs to the space group Fd-3, and has a unit cell constant ˚ Its formula is Na67 TMA12 Zn(ZnPO4 )·192H2 O, where TMA stands a = 25.1991 A. for tetramethylammonium. For this investigation, crystals were prepared using the original synthesis at 4 ◦ C [161]. The crystals produced have the typical octahedral morphology [83] and a size of a few microns. The solution used for crystal growth in the experiments was the clear mother-liquor produced as the synthesis takes place and the crystals form. It was taken after just 4 hours of synthesis. In situ observations of FAU-type ZnPO crystals at low supersaturation conditions showed a ‘‘birth and spread’’ growth mechanism (Figure 1.34). The shape of the 2D nuclei formed is triangular, in accordance with previous ex situ observations [147]. At high supersaturation conditions, growth takes place by the advancement of macrosteps with a height of tenths of nanometers. 1.5.6 Metal Organic Frameworks

Nanoporous MOF crystals have bonding halfway between that of weak hydrogen bonding in molecular crystals and strong covalent bonding in zeolites. The ﬁrst in situ AFM images of growing nanoporous crystals were reported on an MOF as the conditions for growth are less aggressive for the microscope than for zeolites. Figure 1.35 shows a series of AFM images of the important copper trimesate, Cu3 (C9 H3 O6 )2 (H2 O)3 (HKUST-1) [162], which is a signiﬁcant crystalline nanoporous MOF [163] built from Cu2 (H2 O)2 units and benzene-1,3,5-tricarboxylate (BTC) groups and used to form a cubic framework with a three-dimensional nanoporous channel system. The crystals exhibit only (111) facets and the sample has been prepared by growing under ambient conditions in an oriented manner on gold substrates functionalized with self-assembled monolayers (a)

(b)

1 µm Figure 1.34 Sequence of AFM deﬂection images showing the growth of nuclei in a ZnPO–FAU crystal.

(c)

47

48

1 Synthesis Mechanism: Crystal Growth and Nucleation

(a)

(f)

(b)

(g) (112) (211)

(c)

(h)

(d)

(i)

(e)

(j)

Figure 1.35 Real-time deﬂection AFM images of the growing {111} facet of a HKUST-1 crystal at (a) 56, (b) 77, (c) 79, (d) 82, (e) 85, (f) 88, (g) 91, (h) 94, (i) 97, (j) 108 minutes, after injection of the growth solution. (Times refer to the end of each scan). Image sizes are 0.763 × 0.613 µm2 .

1.6 Conclusions and Outlook

(SAMs) of 16-mercapto-1-hexadecanol [162]. This provides a unique platform for in situ AFM studies, since the crystals are ﬁrmly anchored by direct attachment to a gold-coated glass substrate, but more importantly, the orientation of the crystals can be tuned by using different functional groups for surface functionalization, such that the growth of the {111} face can be monitored directly. The crystal growth could be clearly monitored 56 minutes after the injection of the solution used for growth. The image at 56 minutes (Figure 1.35a) reveals an extremely ﬂat and relatively defective-free crystal surface exemplifying the utility of this synthetic protocol to produce high-quality crystal surfaces. In the subsequent images, growth of the surface is seen to proceed by a 2D crystal growth mechanism in which each new crystal layer nucleates at the same point on the crystal surface, indicated by an asterisk in Figure 1.35b. It is likely that a defect at this point on the crystal facet is acting as a nucleation center. Cross-sectional analyses of height images at each time during the growth reveals that the vast majority of growth steps have heights of 1.5 ± 0.1 nm corresponding to the 1.5 nm d111 crystal spacing of the HKUST-1 structure, but also half-step d222 crystal spacings are observed. Interestingly, the triangular terraces exhibit a linear growth until the apex of the terrace reaches the edge of the crystal. At that point the growth slows considerably, illustrating that the abundance of kink sites near the apex of the triangle is dominant for the propagation of the terrace. The results suggest that a layer grows by initial attachment of BTC and copper species onto a stably terminated crystal surface to form a small volume of a d 222 step with metastable termination. The attachment of additional reagent to the crystal occurs more rapidly at this newly created metastable termination, creating a new d111 step with a stable terminated surface.

1.6 Conclusions and Outlook

By applying a number of novel techniques to the problem of crystal growth in nanoporous materials, it is now possible to understand the mechanism at the molecular scale. In particular, the advent of AFM has opened a new window on the subject. Crystals are observed to grow by birth-and-spread mechanism as well as by spiral growth. The effects of supersaturation, temperature, chemical speciation, and structure are becoming apparent at this fundamental scale. It can be hoped that in the near future it should be possible to control crystal shape and habit, defects, and intergrowths through careful control of these growth parameters. In particular, it can be expected that scanning probe microscopies will develop apace. Most AFMs operate at 1 atm pressure with modest variations of temperature under solution conditions. Hydrothermal AFM has been realized to operate at P = 10 bar and T = 150◦ C [164], however, current designs do not permit the location of micron-sized crystals via optical microscopy techniques, and consequently some technique is required to be developed in order to realize this goal. Nevertheless,

49

50

1 Synthesis Mechanism: Crystal Growth and Nucleation

we can expect such breakthroughs in the near future that will extend the range of applicability of AFM to more zeolite systems. In AFM, the cantilever deﬂection is determined optically using a laser light source that is reﬂected off the back of the cantilever surface. For crystal growth measurements this poses many difﬁcult choices. The most obvious is that in a solution it is imperative that there is no turbidity present. The solution may be colored but should not contain any particulates that scatter light. Many crystallizations operate in a chemistry regime where substantial light scattering may be expected. Also, in order to monitor crystallization on micron-sized crystals, it is essential to combine the AFM with a high-resolution optical microscope in order to locate the cantilever on the desired crystal facet. Modern AFM design goes to some lengths in order to organize the geometry of the AFM/optical microscope tandem arrangement in order to accommodate the laser path of the AFM and the optical path of the optical microscope. These complications can be overcome if it was possible to determine the tip displacement or force by measuring the resistance change of a piezoelectric material on the cantilever. There are a number of groups who are developing such techniques since the ﬁrst images were recorded by Tortonese et al. [165] If it becomes possible to realize nonoptical methods for cantilever detection under crystal growth conditions, this would considerably expand the application of AFM in this ﬁeld. Very recently, the ﬁrst ever video-rate AFM recorded under solution conditions has been reported [166]. This work is based around a resonant scanning system. Topographic information is then obtained by deﬂection of the cantilever determined optically. Video-rate AFM until this work has been conﬁned to samples in air, but this new development is particularly interesting for the study of crystal growth or dissolution where it is difﬁcult to bracket the kinetics within a typical AFM frame rate. Finally, there has been considerabe progress recently to improve the lateral resolution of AFM to atomic, or even subatomic resolution [167]. Applied to the problem of crystal growth, this might permit the direct observation of template molecules at surfaces. References 1. Subotic, B. and Bronic, J. (2003) in

Handbook of Zeolite Science and Technology (eds S.M. Auerbach, K.A. Carrado, and P.K. Dutta), Marcel Dekker, New York, p. 1184. 2. Anderson, M.W., Terasaki, O., Ohsuna, T., Philippou, A., Mackay, S.P., Ferreira, A., Rocha, J., and Lidin, S. (1994) Nature, 367, 347–351. 3. Newsam, J.M., Treacy, M.M.J., Koetsier, W.T., and De Gruyter, C.B. (1988) Proc. R. Soc. London Ser. A: Math. Phys. Eng. Sci., 420, 375–405.

4. Kashchiev, D. (2000) Nucleation:

5. 6.

7.

8.

Basic Theory with Applications, Butterworth-Heinemann, Oxford. Mullin, J.W. (2001) Crystallization, 4th edn, Butterworth-Heinemann, Oxford. Kashchiev, D. and van Rosmalen, G.M. (2003) Cryst. Res. Technol., 38, 555–574. Elwell, D. and Scheel, H.J. (1975) Crystal Growth from High-temperature Solutions, Academic Press, New York. Lasaga, A.C. (1998) Kinetic Theory in the Earth Sciences, Princeton University Press, Princeton, p. 811.

References 9. Kossel, W. (1934) Annal. Phys., 21, 10.

11. 12. 13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

457–480. Ohara, M. and Reid, R.C. (1973) Modeling Crystal Growth Rates from Solution, Prentice-Hall International Series in the Physical and Chemical Engineering Sciences, Prentice-Hall, p. 272. Volmer, M. and Schultz, W. (1931) Z. Phys. Chem., 156, 1–22. Frank, F.C. (1949) Discuss. Faraday Soc., 5, 48–54. Burton, W.K., Cabrera, N., and Frank, F.C. (1951) Philos. Trans. R. Soc. London, A243, 299–358. van Enckevort, W.J.P. and Bennema, P. (2004) Acta Crystallogr., Sect. A, 60, 532–541. Pina, C.M., Becker, U., Risthaus, P., Bosbach, D., and Putnis, A. (1998) Nature, 395, 483–486. Astier, J.P., Bokern, D., Lapena, L., and Veesler, S. (2001) J. Cryst. Growth, 226, 294–302. Aquilano, D., Veesler, S., Astier, J.P., and Pastero, L. (2003) J. Cryst. Growth, 247, 541–550. van der Hoek, B., van der Eerden, J.P., and Tsukamoto, K. (1982) J. Cryst. Growth, 58, 545–553. Zauner, A.R.A., Aret, E., van Enckevort, W.J.P., Weyher, J.L., Porowski, S., and Schermer, J.J. (2002) J. Cryst. Growth, 240, 14–21. Baronnet, A., Amouric, M., and Chabot, B. (1976) J. Cryst. Growth, 32, 37–59. Barrer, R.M. (ed.) (1982) Hydrothermal Chemistry of Zeolites, Academic Press, London, p. 360. Cundy, C.S. and Cox, P.A. (2005) Microporous Mesoporous Mater., 82, 1–78. Flanigen, E.M. and Patton, R.L. (1978) Silica polymorph 76-726744 4073865, 19760927. Nicolle, M.A., Di Renzo, F., Fajula, F., Espiau, P., and Des Courieres, T. (1993) 1, 313–320. Mintova, S., Olson, N.H., Valtchev, V., and Bein, T. (1999) Science, 283, 958–960. Mintova, S. and Valtchev, V. (1999) Stud. Surf. Sci. Catal., 125, 141–148.

27. Walton, R.I. and O’Hare, D. (2001)

J. Phys. Chem. Solids, 62, 1469–1479. 28. Yang, H., Walton, R.I., Antonijevic, S.,

29.

30.

31.

32. 33. 34.

35. 36. 37. 38.

39.

40. 41.

42. 43.

44.

Wimperis, S., and Hannon, A.C. (2004) J. Phys. Chem. B, 108, 8208–8217. Kosanovic, C., Bosnar, S., Subotic, B., Svetlicic, V., Misic, T., Drazic, G., and Havancsak, K. (2008) Microporous Mesoporous Mater., 110, 177–185. Wakihara, T., Kohara, S., Sankar, G., Saito, S., Sanchez-Sanchez, M., Overweg, A.R., Fan, W., Ogura, M., and Okubo, T. (2006) Phys. Chem. Chem. Phys., 8, 224–227. Cundy, C.S. and Forrest, J.O. (2004) Microporous Mesoporous Mater., 72, 67–80. Cundy, C.S. and Cox, P.A. (2003) Chem. Rev., 103, 663–701. Pope, C.G. (1998) Microporous Mesoporous Mater., 21, 333–336. Bransom, S.H., Dunning, W.J., and Millard, B. (1949) Discuss. Faraday Soc 83–95. Giaya, A. and Thompson, R.W. (2004) AIChE J., 50, 879–882. Thompson, R.W. and Dyer, A. (1985) Zeolites, 5, 202–210. Gora, L. and Thompson, R.W. (1997) Zeolites, 18, 132–141. ˇ zmek, A., Suboti´c, B., Kralj, D., Ciˇ Babi´c-Ivanˇci´c, V., and Tonejc, A. (1997) Microporous Mesoporous Mater., 12, 267. Brar, T., France, P., and Smirniotis, P.G. (2001) Ind. Eng. Chem. Res., 40, 1133–1139. Bronic, J. and Subotic, B. (1995) Microporous Mesoporous Mater., 4, 239–242. Warzywoda, J., Edelman, R.D., and Thompson, R.W. (1991) Zeolites, 11, 318–324. Cundy, C.S., Lowe, B.M., and Sinclair, D.M. (1990) 100, 189–202. Kajcsos, Z., Kosanovic, C., Bosnar, S., Subotic, B., Major, P., Liszkay, L., Bosnar, D., Lazar, K., Havancsak, H., Luu, A.T., and Thanh, N.D. (2009) Mater. Sci. Forum, 607, 173–176. Erdem-Senatalar, A. and Thompson, R.W. (2005) J. Colloid Interface Sci., 291, 396–404.

51

52

1 Synthesis Mechanism: Crystal Growth and Nucleation 45. Mintova, S., Fieres, B., and Bein, T.

46. 47.

48. 49. 50.

51.

52.

53. 54. 55.

56.

57. 58.

59.

60. 61.

62.

63.

(2002) Stud. Surf. Sci. Catal., 142A, 223–229. Chang, C.D. and Bell, A.T. (1991) Catal. Lett., 8, 305. Flanigen, E.M., Bennett, J.M., Grose, R.W., Cohen, J.P., Patton, R.L., Kirchner, R.M., and Smith, J.V. (1978) Nature, 271, 512–516. Burkett, S.L. and Davis, M.E. (1995) Chem. Mater., 7, 920–928. Wakihara, T. and Okubo, T. (2005) Chem. Lett., 34, 276–281. Zhdanov, S.P. and Samulevich, N.N. (1980) Proceedings of the 5th International Conference on Zeolites, pp. 75–84. Bosnar, S. and Subotic, B. (1999) Microporous Mesoporous Mater., 28, 483–493. Iwasaki, A., Hirata, M., Kudo, I., Sano, T., Sugawara, S., Ito, M., and Watanabe, M. (1995) Zeolites, 15, 308–314. Cundy, C.S., Henty, M.S., and Plaisted, R.J. (1995) Zeolites, 15, 353–372. Cundy, C.S., Henty, M.S., and Plaisted, R.J. (1995) Zeolites, 15, 400–407. Bosnar, S., Antonic, T., Bronic, J., and Subotic, B. (2004) Microporous Mesoporous Mater., 76, 157–165. Gora, L., Streletzky, K., Thompson, R.W., and Phillies, G.D.J. (1997) Zeolites, 18, 119–131. Kalipcilar, H. and Culfaz, A. (2000) Cryst. Res. Technol., 35, 933–942. Schoeman, B.J. (1997) Progress in Zeolite and Microporous Materials, Parts A-C, pp. 647–654. Caputo, D., Gennaro, B.D., Liguori, B., Testa, F., Carotenuto, L., and Piccolo, C. (2000) Mater. Chem. Phys., 66, 120–125. Schoeman, B.J. (1997) Zeolites, 18, 97–105. Cundy, C.S., Lowe, B.M., and Sinclair, D.M. (1993) Faraday Discuss., 95, 235–252. Schoeman, B.J., Sterte, J., and Otterstedt, J.E. (1994) Zeolites, 14, 568–575. Kacirek, H. and Lechert, H. (1975) J. Phys. Chem., 79, 1589–1593.

64. Kacirek, H. and Lechert, H. (1976)

J. Phys. Chem., 80, 1291–1296. 65. Anderson, M.W. (2001) Curr. Opin.

Solid State Mater. Sci., 5, 407–415. 66. Anderson, M.W., Ohsuna, T.,

67.

68.

69. 70.

71. 72.

73.

74.

75.

76.

77.

78.

79.

80.

Sakamoto, Y., Liu, Z., Carlsson, A., and Terasaki, O. (2004) Chem. Commun., 907–916. Anderson, M.W., Agger, J.R., Meza, L.I., Chong, C.B., and Cundy, C.S. (2007) Faraday Discuss., 136, 143–156. Stevens, S.M., Cubillas, P., Jansson, K., Terasaki, O., Anderson, M.W., Wright, P.A., and Castro, M. (2008) Chem. Commun (Camb.), 3894–3896. Terasaki, O. and Ohsuna, T. (2003) Top. Catal., 24, 13–18. Slater, B., Ohsuna, T., Liu, Z., and Terasaki, O. (2007) Faraday Discuss., 136, 125–141. Komiyama, M. and Yashima, T. (1994) Jpn. J. Appl. Phys., 33, 3761–3763. Komiyama, M., Tsujimichi, K., Oumi, Y., Kubo, M., and Miyamoto, A. (1997) Appl. Surf. Sci., 121-122, 543–547. Yamamoto, S., Sugiyama, S., Matsuoka, O., Honda, T., Banno, Y., and Nozoye, H. (1998) Microporous Mesoporous Mater., 21, 1–6. Voltolini, M., Artioli, G., and Moret, M. (2003) Microporous Mesoporous Mater., 61, 79–84. Yamamoto, S., Matsuoka, O., Sugiyama, S., Honda, T., Banno, Y., and Nozoye, H. (1996) Chem. Phys. Lett., 260, 208–214. Sugiyama, S., Yamamoto, S., Matsuoka, O., Honda, T., Nozoye, H., Qiu, S., Yu, J., and Terasaki, O. (1997) Surf. Sci., 377, 140–144. Sugiyama, S., Yamamoto, S., Matsuoka, O., Nozoye, H., Yu, J., Zhu, G., Qiu, S., and Terasaki, I. (1999) Microporous Mesoporous Mater., 28, 1–7. Wakihara, T., Sugiyama, A., and Okubo, T. (2004) Microporous Mesoporous Mater., 70, 7–13. Anderson, M.W., Agger, J.R., Hanif, N., and Terasaki, O. (2001) Microporous Mesoporous Mater., 48, 1–9. Agger, J.R., Hanif, N., and Anderson, M.W. (2001) Angew. Chem. Int. Ed., 40, 4065–4067.

References 81. Agger, J.R., Hanif, N., Cundy, C.S.,

82.

83.

84.

85.

86.

87.

88.

89.

90.

91.

92. 93.

94.

95.

96.

Wade, A.P., Dennison, S., Rawlinson, P.A., and Anderson, M.W. (2003) J. Am. Chem. Soc., 125, 830–839. Dumrul, S., Bazzana, S., Warzywoda, J., Biederman, R.R., and Sacco, A. Jr. (2002) Microporous Mesoporous Mater., 54, 79–88. Singh, R., Doolittle, J., George, M.A., and Dutta, P.K. (2002) Langmuir, 18, 8193–8197. Meza, L.I., Agger, J.R., Logar, N.Z., Kaucic, V., and Anderson, M.W. (2003) Chem. Commun., 2300–2301. Warzywoda, J., Yilmaz, B., Miraglia, P.Q., and Sacco, A. Jr. (2004) Microporous Mesoporous Mater., 71, 177–183. Yamamoto, S., Sugiyama, S., Matsuoka, O., Kohmura, K., Honda, T., Banno, Y., and Nozoye, H. (1996) J. Phys. Chem., 100, 18474–18482. Meza, L.I., Anderson, M.W., Slater, B., and Agger, J.R. (2008) Phys. Chem. Chem. Phys., 10, 5066–5076. Anderson, M.W., Agger, J.R., Thornton, J.T., and Forsyth, N. (1996) Angew. Chem. Int. Ed., 35, 1210–1213. Wakihara, T., Sasaki, Y., Kato, H., Ikuhara, Y., and Okubo, T. (2005) Phys. Chem. Chem. Phys., 7, 3416–3418. Agger, J.R. and Anderson, M.W. (2002) Impact of Zeolites and Other Porous Materials on the New Technologies at the Beginning of the New Millennium, Parts A and B, pp. 93–100. Walker, A.M., Slater, B., Gale, J.D., and Wright, K. (2004) Nat. Mater., 3, 715–720. Binnig, G., Quate, C.F., and Gerber, C. (1986) Phys. Rev. Lett., 56, 930–933. Binnig, G., Rohrer, H., Gerber, C., and Weibel, E. (1982) Phys. Rev. Lett., 49, 57–61. Zhong, Q., Inniss, D., Kjoller, K., and Elings, V.B. (1993) Surf. Sci., 290, L688–L692. Mate, C.M., McClelland, G.M., Erlandsson, R., and Chiang, S. (1987) Phys. Rev. Lett., 59, 1942–1945. Szlufarska, I., Chandross, M., and Carpick, R.W. (2008) J. Phys. D Appl. Phys., 41, 123001.

97. McPherson, A., Malkin, A.J., and

98.

99.

100.

101.

102.

103.

104. 105.

106.

107.

108.

109.

110.

111.

112.

113.

Kuznetsov, Y.G. (2000) Annu. Rev. Biophys. Biomol. Struct., 29, 361–410. Bose, S., Hu, X., and Higgins, S.R. (2008) Geochim. Cosmochim. Acta, 72, 759–770. Higgins, S.R., Boram, L.H., Eggleston, C.M., Coles, B.A., Compton, R.G., and Knauss, K.G. (2002) J. Phys. Chem. B, 106, 6696–6705. Teng, H.H., Dove, P.M., and De Yoreo, J.J. (2000) Geochim. Cosmochim. Acta, 64, 2255–2266. Maiwa, K., Nakamura, H., Kimura, H., and Miyazaki, A. (2006) J. Cryst. Growth, 289, 303–307. Radenovic, N., van Enckevort, W., Kaminski, D., Heijna, M., and Vlieg, E. (2005) Surf. Sci., 599, 196–206. Moret, M., Campione, M., Caprioli, S., Raimondo, L., Sassella, A., Tavazzi, S., and Aquilano, D. (2007) J. Phys., 61, 831–835. Richter, A. and Smith, R. (2003) Cryst. Res. Technol., 38, 250–266. Loiola, A.R., da Silva, L.R.D., Cubillas, P., and Anderson, M.W. (2008) J. Mater. Chem., 18, 4985–4993. Karwacki, L., Stavitski, E., Kox, M.H.F., Kornatowski, J., and Weckhuysen, B.M. (2008) Stud. Surf. Sci. Catal., 174B, 757–762. Kox, M.H.F., Stavitski, E., Groen, J.C., Perez-Ramirez, J., Kapteijn, F., and Weckhuysen, B.M. (2008) Chem. Eur. J., 14, 1718–1725. Harris, R.K., Knight, C.T.G., and Hull, W.E. (1981) J. Am. Chem. Soc., 103, 1577–1578. Harris, R.K. and Knight, C.T.G. (1983) J. Chem. Soc. Faraday Trans. 2: Mol. Chem. Phys., 79, 1525–1538. Harris, R.K. and Knight, C.T.G. (1983) J Chem. Soc. Faraday Trans. 2: Mol. Chem. Phys., 79, 1539–1561. Knight, C.T.G. (1988) J. Chem. Soc. Dalton Trans. Inorg. Chem., 1457–1460. Kinrade, S.D., Knight, C.T.G., Pole, D.L., and Syvitski, R.T. (1998) Inorg. Chem., 37, 4278–4283. Kinrade, S.D., Donovan, J.C.H., Schach, A.S., and Knight, C.T.G.

53

54

1 Synthesis Mechanism: Crystal Growth and Nucleation

114.

115.

116.

117. 118.

119.

120.

121.

122.

123.

124. 125.

126.

127.

128. 129.

(2002) J. Chem. Soc., Dalton Trans., 1250–1252. Knight, C.T.G. and Kinrade, S.D. (2002) J. Phys. Chem. B, 106, 3329–3332. Knight, C.T.G., Wang, J., and Kinrade, S.D. (2006) Phys. Chem. Chem. Phys., 8, 3099–3103. Knight, C.T.G., Balec, R.J., and Kinrade, S.D. (2007) Angew. Chem. Int. Ed., 46, 8148–8152. Haouas, M. and Taulelle, F. (2006) J. Phys. Chem. B, 110, 22951. Aerts, A., Follens, L.R.A., Haouas, M., Caremans, T.P., Delsuc, M.-A., Loppinet, B., Vermant, J., Goderis, B., Taulelle, F., Martens, J.A., and Kirschhock, C.E.A. (2007) Chem. Mater., 19, 3448–3454. Follens, L.R.A., Aerts, A., Haouas, M., Caremans, T.P., Loppinet, B., Goderis, B., Vermant, J., Taulelle, F., Martens, J.A., and Kirschhock, C.E.A. (2008) Phys. Chem. Chem. Phys., 10, 5574–5583. Serre, C., Corbiere, T., Lorentz, C., Taulelle, F., and Ferey, G. (2002) Chem. Mater., 14, 4939–4947. Gerardin, C., Haouas, M., Lorentz, C., and Taulelle, F. (2000) Magn. Reson. Chem., 38, 429–435. Gerardin, C., In, M., Allouche, L., Haouas, M., and Taulelle, F. (1999) Chem. Mater., 11, 1285–1292. Vistad, O.B., Akporiaye, D.E., Taulelle, F., and Lillerud, K.P. (2003) Chem. Mater., 15, 1639–1649. Shi, J., Anderson, M.W., and Carr, S.W. (1996) Chem. Mater., 8, 369–375. Egger, C.C., Anderson, M.W., Tiddy, G.J.T., and Casci, J.L. (2005) Phys. Chem. Chem. Phys., 7, 1845–1855. Bussian, P., Sobott, F., Brutschy, B., Schrader, W., and Schuth, F. (2000) Angew. Chem. Int. Ed., 39, 3901–3905. Schuth, F., Bussian, P., Agren, P., Schunk, S., and Linden, M. (2001) Solid State Sci., 3, 801–808. Schuth, F. (2001) Curr. Opin. Solid State Mater. Sci., 5, 389–395. Pelster, S.A., Schrader, W., and Schuth, F. (2006) J. Am. Chem. Soc., 128, 4310–4317.

130. Pelster, S.A., Kalamajka, R., Schrader,

131.

132.

133.

134. 135. 136. 137.

138. 139.

140.

141.

142.

143. 144.

145. 146.

147.

W., and Schuth, F. (2007) Angew. Chem. Int. Ed., 46, 2299–2302. Pelster, S.A., Schueth, F., and Schrader, W. (2007) Anal. Chem., 79, 6005–6012. Pelster, S.A., Weimann, B., Schaack, B.B., Schrader, W., and Schueth, F. (2007) Angew. Chem. Int. Ed., 46, 6674–6677. Kumar, S., Wang, Z., Penn, R.L., and Tsapatsis, M. (2008) J. Am. Chem. Soc., 130, 17284–17286. Piana, S., Reyhani, M., and Gale, J.D. (2005) Nature, 438, 70–73. Piana, S., and Gale, J.D. (2005) J. Am. Chem. Soc., 127, 1975–1982. Piana, S., and Gale, J.D. (2006) J. Cryst. Growth, 294, 46–52. Boerrigter, S.X.M., Josten, G.P.H., van de Streek, J., Hollander, F.F.A., Los, J., Cuppen, H.M., Bennema, P., and Meekes, H. (2004) J. Phys. Chem. A, 108, 5894–5902. Umemura, A. (2009) PhD thesis, The University of Manchester, Manchester. Anderson, M.W., Meza, L.I., Agger, J.R., Attﬁeld, M.P., Shoaee, M., Chong, C.B., Umemura, A., and Cundy, C.S. (2008) Turning Points in Solid-State, Materials and Surface Science, 95–122. Umemura, A., Cubillas, P., Anderson, M.W., and Agger, J.R. (2008) Stud. Surf. Sci. Catal., 174A, 705–708. Breck, D.W. (1974) Zeolite Molecular Sieves, John Wiley & Sons, Ltd, New York. Baerlocher, C., Meier, W.M., and Olson, D.H. (2007) Atlas of Zeolite Structure Types, 7th edn, Elsevier, Amsterdam. Charnell, J.F. (1971) J. Cryst. Growth, 8, 291–294. Petranovskii, V., Kiyozumi, Y., Kikuchi, N., Hayamisu, H., Sugi, Y., and Mizukami, F. (1997) Stud. Surf. Sci. Catal., 105A, 149–156. Gora, L. and Thompson, R.W. (1995) Zeolites, 15, 526–534. Agger, J.R., Pervaiz, N., Cheetham, A.K., and Anderson, M.W. (1998) J. Am. Chem. Soc., 120, 10754–10759. Wakihara, T. and Okubo, T. (2004) J. Chem. Eng. Jpn., 37, 669–674.

References 148. Thompson, R.W. and Huber, M.J. 149.

150.

151.

152. 153.

154.

155.

156.

157.

(1982) J. Cryst. Growth, 56, 711–722. Yang, X.B., Albrecht, D., and Caro, E. (2006) Microporous Mesoporous Mater., 90, 53–61. Meza, L.I., Anderson, M.W., and Agger, J.R. (2007) Chem. Commun (Camb.), 2473–2475. Meza, L.I., Anderson, M.W., Agger, J.R., Cundy, C.S., Chong, C.B., and Plaisted, R.J. (2007) J. Am. Chem. Soc., 129, 15192–15201. Brent, R. and Anderson, M.W. (2008) Angew. Chem. Int. Ed., 47, 5327–5330. Castro, M., Garcia, R., Warrender, S.J., Slawin, A.M.Z., Wright, P.A., Cox, P.A., Fecant, A., Mellot-Draznieks, C., and Bats, N. (2007) Chem. Commun., 3470–3472. Cubillas, P., Castro, M., Jelfs, K.E., Lobo, A.J.W., Slater, B., Lewis, D.W., Wright, P.A., Stevens, S.M., and Anderson, M.W. (2009) Crys. Growth Des. Paquette, J. and Reeder, R.J. (1995) Geochim. Cosmochim. Acta, 59, 735–749. Lewis, D.W., Willock, D.J., Catlow, C.R.A., Thomas, J.M., and Hutchings, G.J. (1996) Nature, 382, 604–606. Jelfs, K.E., Slater, B., Lewis, D.W., and Willock, D.J. (2007) Zeolites to Porous MOF Materials, Vol. 170B, Elsevier B.V., pp. 1685–1692.

158. Nenoff, T.M., Harrison, W.T.A., Gier,

159.

160.

161. 162.

163.

164.

165.

166.

167.

T.E., and Stucky, G.D. (1991) J. Am. Chem. Soc., 113, 378–379. Harrison, W.T.A., Gier, T.E., and Stucky, G.D. (1991) J. Mater. Chem., 1, 153–154. Nenoff, T.M., Harrison, W.T.A., Gier, T.E., Calabrese, J.C., and Stucky, G.D. (1993) J. Solid State Chem., 107, 285–295. Gier, T.E., and Stucky, G.D. (1991) Nature, 349, 508–510. Shoaee, M., Agger, J.R., Anderson, M.W., and Attﬁeld, M.P. (2008) Cryst EngComm, 10, 646–648. Chui, S.S.Y., Lo, S.M.F., Charmant, J.P.H., Orpen, A.G., and Williams, I.D. (1999) Science, 283, 1148–1150. Higgins, S.R., Eggleston, C.M., Knauss, K.G., and Boro, C.O. (1998) Rev. Sci. Instrum., 69, 2994–2998. Tortonese, M., Barrett, R.C., and Quate, C.F. (1993) Appl. Phys. Lett., 62, 834–836. Picco, L.M., Bozec, L., Ulcinas, A., Engledew, D.J., Antognozzi, M., Horton, M.A., and Miles, M.J. (2007) Nanotechnology, 18, 044030-1–044030-4. Giessibl, F.J. and Quate, C.F. (2006) Phys. Today, 59, 44–50.

55

57

2 Synthesis Approaches Karl G. Strohmaier

2.1 Introduction

Zeolites were ﬁrst identiﬁed as natural minerals having unique physical absorption properties. Early researchers tried to reproduce the geological conditions in an attempt to make them synthetically. From the years 1845 to 1937 many researchers investigated the hydrothermal conversion and synthesis of silicates. Although there were a number of claims that zeolites had been made synthetically, they were unsubstantiated as identiﬁcation was based upon chemical analysis and optical observations. Later attempts at reproducing these early experiments did not give identiﬁable zeolites. It was not until the 1940s that Professor Richard M. Barrer found the necessary conditions for making zeolites synthetically and he identiﬁed them by powder X-ray diffraction. By reacting the powder minerals leucite and analcime with aqueous solutions of barium chloride at temperatures of 180–270 ◦ C for two to six days, a synthetic chabazite was obtained. In 1948, Professor Barrer was able to synthesize zeolites entirely from synthetic solutions of sodium aluminate, silicic acid, and sodium carbonate [1]. A dry powdered gel was ﬁrst obtained by drying the mixture at 110 ◦ C, and then crystallizing it to a synthetic mordenite by reacting the gel with water at higher temperatures. In the 1950s, Donald Breck and researchers at the Union Carbide Company began to synthesize zeolites from reactive aluminosilicate gels prepared from sodium aluminate and sodium silicate solutions. This led directly to the discovery of the ﬁrst two new synthetic zeolites, Linde A [2] and Linde X (Si/Al = 1.0–1.5) [3], both having new framework structures, LTA and FAU, respectively. Linde A was found to have excellent ion exchange and gas separation properties. Later, a new, higher silica containing composition of FAU, designated Linde Y [4] (Si/Al = 1.5–3.0), was synthesized and was found to be an excellent cracking catalyst for the conversion of distilled crude oil to gasoline products. The discovery of these two industrially important zeolites caught the attention of many scientists and kick started the ﬁeld of zeolite synthesis research.

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

58

2 Synthesis Approaches

Until this time, synthetic zeolites were made entirely from inorganic reagents and many experiments were performed to determine the effects of various alkali and alkali earth metals. The next major advancement in zeolite synthesis came in 1961 when Professor Barrer began experimenting with substituting part of the alkali and alkali earth cations with organic cations such as tetramethylammonium (TMA+ ) [5]. Using TMA+ , he was able to make the zeolite N-A (Si/Al = 1.2), the high silica composition of Linde A (Si/Al = 1.0). Chemists at the Mobil Oil Company also began to utilize organic cations in zeolite syntheses and, in 1967, they became the ﬁrst research group to make new zeolite structures with organic cations, which are also referred to as templates or structure-directing agents (SDAs). Two new commercially important zeolites, beta and ZSM-5, were made with tetraethylammonium (TEA+ ) and tetrapropylammonium (TPA+ ) cations, respectively. Both beta and ZSM-5 are the ﬁrst synthetic zeolites prepared with Si/Al ratios greater than 5. Because most natural zeolites have Si/Al ratios of one to about ﬁve, earlier researchers did not thoroughly investigate high-silica-containing gels. This new approach of utilizing organic SDAs and high-silica-containing gels led to the discovery of a large number of new zeolite structures by Mobil scientists and other groups over the next two decades. Some of the more important ones are ZSM-11 (MEL), ZSM-12 (MTW), ZSM-22 (TON), ZSM-23 (MTT), ZSM-48 (∗ MRE), ZSM-57 (MFS), EU-1 (EUO), and NU-87 (NES). Even to this day the use of new SDAs remains the primary strategy for discovering new zeolite structures.

2.2 Aluminophosphates

Another major breakthrough in zeolite synthesis occurred with the discovery of aluminophosphates (AlPOs) and, shortly thereafter, silicoaluminophosphates (SAPOs) molecular sieves. A number of scientists recognized that aluminum phosphate framework structures were known in nature in which phosphorus and aluminum atoms had tetrahedral coordination similar to that of aluminum and silicon in zeolites. While several groups worked on being the ﬁrst to make synthetic AlPOs, it was the researchers at Union Carbide in the early 1980s who found the right reagents (pseudoboehmite alumina and phosphoric acid with organic SDAs) and conditions (typically no inorganic cations) necessary for easily making them [6]. Technically, the microporous AlPOs are not zeolites because they do not have aluminosilicate compositions, but nonetheless they have similar or identical frameworks to zeolites. The synthesis of SAPOs, metalloaluminophosphates (MeAPOs, where Me = a metal or combination of metals such as Mg, Ti, Cr, Mn, Fe, Co, Ni, Zn, B, and Ge capable of tetrahedra coordination) and metallosilicoalumino-phosphates (MeAPSOs) quickly followed. The structures of AlPOs and SAPOs are such that there are alternating aluminum-phosphorus tetrahedra throughout the framework. As with zeolites, Lowenstein’s rule prevails in AlPO and SAPO materials also, such that aluminum does not have aluminum as a nearest neighbor. It was also found that phosphorus

2.3 Mineralizers

does not have phosphorus or silicon as a nearest neighbor. For this reason, it was believed that all phosphate structures have only even number of rings in their frameworks, that is, 4-, 6-, 8-, and 12-member rings. This results in the absence of AlPO materials having structures with odd-numbered rings such as ZSM-5, mordenite, beta, and many others. Other structures with only even-member rings such as ERI, FAU, and CHA frameworks exist in both the aluminosilicate and phosphate compositions. It was not until the discovery of ECR-40 that it was realized that it was possible to synthesize odd-numbered ring containing structures having phosphate compositions. By adding silicon to an AlPO synthesis with (C2 H5 OH)x (CH4 )4−x NOH (x = 2, 3) as SDAs, it was found that small amounts of a SAPO material having the MEI structure could be obtained at relatively long crystallization times (three to four weeks) [7]. The MEI framework contains three-, ﬁve-, and seven-member rings, which, as stated above, are not attainable due to alternating aluminum and phosphorus atoms in the structure because a ﬁve-membered ring always has either Al–Al or P–P nearest neighbors. In the ECR-40 structure, it was found that the Al, P, and Si were ordered such that there where no Si–P neighbors and an anti-Lowenstein Al–O–Al bond. For this to be possible, the SAPO had to have the speciﬁc composition Al16 P12 Si6 O72 , unlike all other SAPO materials, which can usually be prepared with a range of Si levels down to Si/Al = 0, that is, an AlPO. With this information, it was now straightforward to optimize the syntheses by using the structural composition, Al16 P12 Si6 O72 , in the gel synthesis to prepare the material in high yield [8]. This example represents the use of structural information to develop a strategy for optimizing the synthesis of a zeolitic material.

2.3 Mineralizers

An important feature of conventional zeolite synthesis is the use of high hydroxide concentrations to assist the mineralization of silicate and aluminate species in the reactant gel. It is important to have the right concentration of hydroxide ion as there is equilibrium between solution species and solid species in the gel. A very high concentration shifts the equilibrium very far toward the solution and prevents crystallization because the solubility of the silica and alumina is too high. On the other hand, a very low concentration does not solubilize the species enough to cause their transport to the growing crystals, resulting in an amorphous product. The concentration of hydroxide can also effect the transformation of metastable phases to dense phases. In most of the cases, the desired zeolite phase is metastable and the ﬁrst phase that crystallizes in a synthesis. At longer times, these metastable phases can transform in to undesirable dense phases such as zeolites P, sodalite, analcime, and quartz. Therefore, it is important to determine the correct hydroxide ion level so that the desired phase can be recovered before denser impurities begin to form.

59

60

2 Synthesis Approaches

In addition to the aforementioned importance of hydroxide concentration, the level of hydroxide can also have a large effect on the crystallization rate, crystal size, the Si/Al ratio of the product, and the ultimate product formed. An excellent example of the latter is a recent study that observes the effects of hydroxide concentration on syntheses performed using 1,4-bis(N-methylpyrrolidinium) butane as an SDA [9]. At a constant Si/Al ratio of 30, ZSM-12 (MTW) was produced from gels having a OH− /Si ratio ≤ 0.6, TNU-9 (TUN, a new 10-ring zeolite having a three-dimensional channel framework similar to ZSM-5) was produced from gels having a OH− /Si ratio = 0.73 and TNU-10 (a new high silica form of stilbite, STI) was produced from gels having a OH− /Si ratio = 1.0. The synthesis of IM-5 (IMF) is another example of a zeolite prepared with a relatively high OH− ratio (OH− /Si = 0.6). A recent study by Jackowski et al. [10] looked further at the effect of OH− levels in zeolite syntheses with a series of heterocyclic-substituted diquaternary ammonium compounds as SDAs. These studies illustrate the importance of evaluating a large OH− /Si range when trying to synthesize new zeolites. Besides hydroxide, ﬂuoride can also be used as an effective mineralizer for the synthesis of zeolites and phosphate materials. Flanigen and Patton were the ﬁrst to use ﬂuoride in a zeolite synthesis to prepare a highly siliceous MFI material called silicalite [11]. The incorporation of ﬂuoride into zeolites was found to reduce the number of defect sites in highly siliceous zeolites and improve their overall stability. Before the general use of ﬂuoride to prepare zeolites was undertaken, it was used ﬁrst as a mineralizer in phosphate synthesis by Guth and Kessler [12]. A number of new materials were discovered by the use of ﬂuoride in phosphate systems. Cloverite (-CLO) was synthesized with ﬂuoride in the gallium phosphate system using quinuclidine as an SDA [13]. Another gallium phosphate material having the LTA framework was also prepared with HF [14]. It was found that GaPO4 -LTA could not be prepared in the absence of ﬂuoride and that the ﬂuoride was located at the center of the double 4-ring (D4R) in the structure [15]. Although other gallium phosphate materials, such as ULM-[16] and Ga5 (PO4 )5 F4 2[N2 C4 H12 ] [17], have been prepared in the presence of ﬂuoride, they usually are found to contain GaO4 F trigonal bipyramids and GaO4 F2 octahedra in their structures. In addition to synthesizing new GaPO4 structures, HF was also found to be effective in synthesizing new AlPO compositions and structures such as AlPO4 -CHA [18, 19], UiO-6 (OSI) [20], and UiO-7 (ZON) [21]. The use of ﬂuoride was also found to improve the distribution of silicon substitution into SAPO materials. The way silicon substitutes for phosphorus or aluminum-phosphorus pairs in AlPO synthesis is very important to the acidity of the ﬁnal product, as silicon is the source of its Brønsted acidity [22]. At low silicon levels, isolated silicon (Si-Al4 Si0 ), having four aluminum nearest neighbors, is present. As Al has relatively low electronegativity, the isolated Si atoms generate weak acid sites. As more silicon substitutes in the framework, silicon islands begin to form where Si can have one silicon neighbor (Si-Al3 Si1 ) at the corners of the islands, two Si neighbors (Si-Al2 Si2 ) along the edges of the islands, and three (Si-Al1 Si3 ) on the inside corners of the islands. Because of the higher electronegativity of silicon, Si sites with more Si nearest neighbors have stronger acid site strength. Si atoms

2.4 Dry Gel Conversion Syntheses

located in the interior of the islands have four Si neighbors (Si-Al0 Si4 ) and generate no acid sites, similar to the absence of acid sites in a pure-silica zeolite. Therefore, to maximize the acidity (both concentration and strength) of SAPO materials, it is desirable to introduce as much silicon into the structure as possible to form a large number of small Si islands having Si sites along the edges and corners, but not too much silicon where most of the Si ends up in the center of the islands to give nonacidic Si-Al0 Si4 sites. The difﬁculty in doing this arises from the fact that it is not easy to regulate the silicon incorporation at near neutral pH conditions. Much work has been done to ﬁnd the best way to introduce silicon into the framework of AlPO materials. One strategy for introducing Si into SAPO materials is the use of ﬂuoride as the mineralizing or complexing agent. G.H. Kuehl at Mobil was the ﬁrst to use ﬂuoride in SAPO syntheses as a way to control the silicate concentration to gradually release silicate to the growing crystals as it is used up during crystallization [23]. Using ammonium ﬂuoride, he was able to prepare a SAPO-20 (SOD) material with high silicon substitution into the framework, which was conﬁrmed by 29 Si NMR.

2.4 Dry Gel Conversion Syntheses

Conventional zeolite syntheses are performed from a reaction medium containing an aqueous phase of dissolved reagents and solid species, which may exist as gel particles and be suspended in solution. Recently, new methods of zeolite syntheses have been studied in which the solid species are kept separated from the aqueous phase such that the solids never come in direct contact with water. These methods are generally referred to as dry gel conversion (DGC) syntheses [24] which include the vapor phase transport [25] (VPT) and steam-assisted conversion (SAC) [26] methods. Xu et al. [27] were the ﬁrst to make ZSM-5 from a dry gel by contact with vapors of water and volatile amines. In the VPT method, amorphous powders were obtained by drying sodium aluminosilicate gels of various compositions. They were transformed into ZSM-5 by contact with vapors of amines and water, or water alone. Amines were found to participate in the crystallization process by absorbing into the reacting, hydrous phase and elevating the pH. Whereas in the VPT method the SDA is physically separate from the aluminosilicate gel, in the SAC method the gel is prepared containing the SDA before the drying step. The dry gel is crushed and then suspended in a specially designed autoclave above a small reservoir of water. Using this method, Rao et al. [28] were able to make self-bonded pellets of high silica beta (Si/Al up to 365) after only 12 hours of crystallization. The SAC method was recently used to make pure-silica MCM-68 (MSE) [29]. After heating for ﬁve days at 150 ◦ C, a dry gel containing the MCM-68 SDA was transformed into a siliceous product, designated YNU-2P. The XRD pattern of this material showed it to be MCM-68, but it was found to lose crystallinity upon calcination. Rietveld reﬁnement of the powder diffraction data showed that some of the tetrahedral framework sites were not fully occupied resulting in defects, which was conﬁrmed by 29 Si MAS NMR. After a simple

61

62

2 Synthesis Approaches

postsynthesis silylation step using tetramethyl orthosilicate and HCl, the new material, now designated YNU-2, was found to have signiﬁcantly fewer defect sites. The repaired material was now stable to calcination to remove the organic SDA. 2.5 Low Water Syntheses

Although the use of ﬂuoride to prepare silica-MFI has been known for some time [11], it was not until years later that its use was exploited to its full potential after an important discovery was made at the laboratories of Prof. Corma at the Polytechnic University of Valencia. By combining the use of HF with highly concentrated gels, Prof. Corma and his colleagues were able to prepare a number of low-framework density zeolites in their purely siliceous form. In this technique, the gel was not separated from the aqueous phase as in the DGC techniques. Conventional aluminosilicate zeolite syntheses use water levels having the ratios of H2 O : SiO2 > 25 to allow easy homogenization and handling of the synthesis gels. By using very low levels of water, that is, H2 O : SiO2 < 10, in the presence of ﬂuoride it was found that siliceous zeolite beta (BEA) could be prepared without the use of seeds [30]. Using the synthesis method described in the original patent, zeolite beta could only be prepared with Si/Al = 5–100 [31]. It was found by 29 Si MAS NMR that siliceous beta had no connectivity defects. Until this time, low-framework density silica zeolites (<15 T-atoms per cubic angstrom) were known only in materials having high aluminum framework content, such as zeolites chabazite, faujasite, and zeolite A. A few years after the low-water, ﬂuoride synthesis of beta was disclosed, the synthesis of pure-silica chabazite was published [32]. Unlike silica beta, silica chabazite was found to contain a small number of Si(OSi)3 OH defects. A review by Camblor et al. describes the syntheses of a number of all-silica and high-silica zeolites from ﬂuoride-containing, low-water syntheses [33]. The effect of water level on the synthesis was found to have a profound effect on the product that was formed. With the same gel, merely varying the water level can produce a different zeolite. The general trend that lower H2 O/SiO2 ratios in the synthesis gel form zeolites with lower framework densities was observed. Using the trimethyladamantammonium SDA in the presence of HF, low-framework-density CHA was made at H2 O/SiO2 = 3, medium-density SSZ-23 (SST) was made at H2 O/SiO2 = 7.5, and high-density SSZ-31 (∗ STO) [34] made at H2 O/SiO2 = 15 (Figure 2.1). Zones et al. at Chevron Research Laboratories have also used this approach to prepare many other frameworks with all-silica compositions in the presence of ﬂuoride [35, 36]. Besides the usefulness of this new synthesis technique to prepare the all-silica form of known aluminosilicate and borosilicate zeolites, it has also led to a number of new zeolite frameworks which are tabulated in Table 2.1. Zeolite A (LTA) was ﬁrst synthesized over 50 years ago, and remained until recent times the only aluminosilicate that contained D4Rs in its structure. While there are a number of phosphate-containing frameworks, such as ACO, AFY, AST, -CLO, and DFO, that have D4Rs in their structure [45], for a considerable period of time, no new

H3C

CH3 CH3 N+

H2O:SiO2

FD

FD/ H2O:SiO2

2.6 Germanium Zeolites

20 15 10 5 0 CHA

SSZ-23

SSZ-31

Figure 2.1 The effect of water level on the framework density (FD) of the zeolite formed [33]. FD units are in T-atoms per cubic angstrom and the water level is the H2 O/SiO2 ratio. The structure of the trimethyladamantammonium template is shown above the chart legend. Table 2.1

New silica frameworks discovered from the ﬂuoride syntheses route.

Zeolite

Framework code

Framework density

Octadecasil ITQ-3 ITQ-4 ITQ-7 ITQ-12 ITQ-13 ITQ-27 SSZ-74

AST ITE IFR ISV ITW ITH ITV -SVR

16.7 16.3 17 15.4 16.3 17.8 15.7 17.1

Synthesis H2 O/SiO2

References

8.3 7.7 15 5.4 7 7 3 3–7

[37] [38] [39] [40] [41] [42] [43] [44]

silica or aluminosilicate frameworks containing D4Rs were made. The discoveries of octadecasil (AST), ITQ-7, ITQ-12, and ITQ-13 zeolites represent the ﬁrst examples of high-silica-containing frameworks with D4Rs. The synthesis of all these materials was made possible by the use of ﬂuoride as discussed above. Subsequent characterization studies by X-ray diffraction and MAS NMR suggest that the ﬂuoride anion exerts a structure-directing effect toward frameworks containing D4Rs and other small-cage secondary building units (SBUs) containing a high density of 4-member rings (4MRs) [33]. In some materials, it was found that ﬂuoride strongly interacts with framework silica to give pentacoordinate [SiO4 F]− species [46]. In D4R frameworks, the ﬂuoride anion was often found at the center of the D4R cube.

2.6 Germanium Zeolites

A new synthesis strategy, ﬁrst utilized by Yaghi to synthesize ASU-7 [47] and then exploited by researchers at Stockholm University, Polytechnic University

63

64

2 Synthesis Approaches

of Valencia, and University de Haute Alsace, utilizes germanium to promote D4R-containing structures. As discussed above, the addition of ﬂuoride to a synthesis gives it the ability to crystallize high silica frameworks and frameworks containing D4Rs. Analysis of pure-silica zeolite frameworks have found the most common Si–O–Si bond angle value to be 148◦ , slightly less than the mean value of 154 ± 9◦ [48]. In silica frameworks, the D4R interatomic bond angles are inclined to be smaller because of the 90◦ Si–Si–Si angle of the D4R cube. The presence of germanium in zeolite syntheses increases the likelihood of forming D4R-containing frameworks due to the smaller Ge–O–Ge angle (130◦ ) compared to the larger Si–O–Si angle. Theoretical energy calculations have shown that the substitution of Ge for Si in the D4R stabilizes this SBU [49, 50]. Using this new strategy, Prof. Corma’s laboratory and other groups have synthesized a large number of new frameworks, many having desirable multidimensional, extra-large pores. These new frameworks are summarized in Table 2.2. Although some have argued that the D4R in all-silica frameworks are strained, force-ﬁeld calculations have shown that the strain is minor [51], suggesting the possibility that these materials may be synthesized without germanium or ﬂuoride. Germanium is recovered as by-product from zinc ores and is very expensive. Its worldwide production was about 100 t in 2007 [66]. Its cost has varied between US$400 and US$1400 kg−1 in the past decade, making its use in a petrochemical catalyst highly unlikely. Not only is germanium prohibitively expensive, its presence in a framework reduces the zeolite’s stability. Corma’s laboratory has been successful, in some cases, in removing Ge from the synthesis by seeding. Germanium-free ITQ-24 has been prepared by seeding with Ge-containing ITQ-24 crystals in a boron-containing gel [67], but the ﬁnal products always contain a Table 2.2

New D4R zeolite frameworks discovered with the use of germanium.

Zeolite

Framework code

Framework density

Channels

ASU-7 ITQ-17/FOS-5 ITQ-15/ITQ-25/IM-12 ITQ-21 ITQ-22 ITQ-24 ITQ-26 ITQ-33 ITQ-34 SU-15 SU-32 IM-10 IM-16

ASV BEC UTL – IWW IWR IWS – ITR SOF STW UOZ UOS

17.9 13.9 15.5 13.6 16.6 15.6 14.4 12.9 17.4 16.4 15.2 16.7 17.9

1D 12 ring 3D 12 ring 14 × 12 ring 3D 12 ring 12 × 10 × 10 ring 12 × 10 × 10 ring 3-D 12 ring 18 × 10 × 10 ring 10 × 10 × 9 ring 10 × 9 × 9 ring 2D 8 ring 6 ring 1D 10 ring

Reference

[47] [52, 53] [54–56] [57] [58] [59] [60] [61] [62] [63] [63] [64] [65]

2.7 Isomorphous Substitution

signiﬁcant amount of germanium from the seeds. Because many of these new D4R frameworks have desirable multidimensional, large, and extra-large pore structures that do not exist in other frameworks, the challenge is to synthesize them without germanium. One approach to synthesizing Ge-free D4R structures may be to choose a suitable template. Zeolite A is a D4R-containing structure synthesized at low Si/Al ratios. By using supramolecular self-assembled molecules as SDAs, a pure-silica LTA material, designated ITQ-29, was prepared [68]. It was found that π –π-type interactions cause the assembly of two julolidine-derived cations with proper rigidity and polarity properties to template the LTA framework. The key to the success of this synthesis was also the realization that the LTA framework has sod cages [45], which could be templated by the addition of a second SDA, TMA+ .

2.7 Isomorphous Substitution

Researchers have also investigated the substitution of other metals, monovalent, divalent, and trivalent ones, for trivalent aluminum. A number of elements in the periodic table have been claimed to substitute, at least in small amounts, for aluminum in many known zeolite frameworks, particularly MFI. While a detailed discussion of these materials is beyond the scope of this chapter, the use of isomorphous substitution of aluminum to discover new frameworks is brieﬂy discussed. Elements other than aluminum (Al–O bond length is about ˚ have different tetrahedral atom–oxygen bond lengths, T–O–T angles, 1.74 A) and/or charges, which could stabilize other SBUs and induce the formation of new structures. Gallium is right below aluminum in the periodic table and easily substitutes for it in most of the zeolite structures. The notable exception is zeolite A, likely because it has D4Rs, which may be too strained with the larger Ga–O bond ˚ A study by Newsam and Vaughan [69], on the other hand, length (about 1.82 A). showed a general decrease in T–O–T bond angles with gallium substitution in a number of zeolite frameworks, which would likely stabilize the D4R as germanium does. Two new frameworks have been discovered with the use of gallium, CGS and ETR. The gallosilicate TsG-1 (CGS), having no aluminosilicate analog, was synthesized in 1985 by Krutskaya et al. from potassium gallosilicate gels [70]. In 1999, it was determined that TsG-1 had the same structure as CoGaPhosphate-6, whose structure was determined the year before [71]. A second gallosilicate zeolite, ECR-34 (ETR) having no aluminosilicate analog was discovered by Strohmaier and Vaughan [72]. Its synthesis requires three cations, Na+ , K+ , and TEA+ . ECR-34 and ITQ-33 [61] are the only silicate zeolites to have 18-ring pore openings. It is interesting to note that both the CGS and ETR frameworks can be built from the same SBU and that this SBU is not seen in any other aluminosilicate frameworks, indicating a structure-directing effect of gallium and potassium for this particular SBU (Figure 2.2).

65

66

2 Synthesis Approaches

ETR

CGS

Figure 2.2 Chain of connected open hexagonal prisms – the secondary building unit found only in the gallosilicate ETR and CGS frameworks.

Boron is known to substitute for aluminum in zeolites to form borosilicates (sometimes referred to as boralites). The boron–oxygen tetrahedral bond length ˚ is signiﬁcantly shorter than the aluminum–oxygen bond length (about 1.46 A) ˚ Also, the B–O–Si bond angle is found to be, on average, smaller (about 1.74 A). than that of either Si–O–Si or Al–O–Si bond lengths in tetrahedral framework materials [73]. It would therefore be expected that boron substitution for aluminum would induce the formation of new zeolite structures. While early work did not show any signiﬁcant substitution of boron into zeolite frameworks [74], later studies showed that it could be inserted into high-silica frameworks such as MFI, ∗ BEA, MWW (as ERB-1) MTW, and others [75]. Zeolite Nu-1, synthesized by Imperial Chemical Industries (ICI) in 1977 [76], was ﬁrst made as an aluminosilicate, but it was found that it could be made more reproducibly by using boron in the synthesis [77]. Its structure (RUT) was solved later by Gies and Rius [78]. Zones et al. prepared an aluminosilicate, designated SSZ-26, using the hexamethyl[4.3.3.0] propellane-8,11-diammonium SDA [79]. Later they used the N,N,N-trimethyl-8-ammonium tricylclo[5.2.1.0]decane SDA to prepare a borosilicate called SSZ-33, having a similar powder XRD pattern to that of SSZ-26 [80]. Sometime later, Lobo and Davis used a third template, N,N,N-trimethyl cis-myrtanyl ammonium, to prepare the borosilicate CIT-1 [81]. CIT-1 was found to have a 12 × 12 × 10 ring pore structure of the pure end-member polymorph B (CON), while the SSZ-26 and SSZ-33 structures, were found to be intergrowths of two closely related polymorphs, A and B [82]. Using boron instead of aluminum, scientists from Chevron, Ruhr University, ExxonMobil, and Stockholm University have discovered a number of new frameworks in the past 10 years. These new borosilicate zeolites are listed in Table 2.3. The synthesis of a new borosilicate, SSZ-82, was recently published, but its framework has not yet been determined [83]. Millini et al. reviewed the synthesis and characterization of borosilicate zeolites in 1999 [75]. Besides the trivalent boron and gallium analogs of aluminosilicates, divalent zinc, beryllium, and monovalent lithium are known to substitute into siliceous frameworks. Zinc was found to direct the synthesis of three new silicates, VPI-8 (VET) [93], VPI-9 (VNI) [94], and RUB-17 (RSN) [95], all three zeolites being prepared ˚ and, while with the TEA+ SDA. The Zn–O bond length is very long (about 1.94 A) zinc is claimed to promote the formation of low density, 3R (three-ring)-containing framework structures, only two of the three zincosilicate frameworks (VNI, RSN) have 3Rs. Like zinc, beryllium is a divalent cation capable of tetrahedral coordination

2.8 Structure-Directing Agents Table 2.3

Borosilicate zeolites having new frameworks.

Code

Type material

Framework composition

Reference

CON RTH RUTa

CIT-1 RUB-13 RUB-10 MCM-70 SSZ-48 SSZ-58 SSZ-53 SSZ-59 SSZ-56 SU-16 SSZ-65 SSZ-60

B2 Si54 O112 B2 Si30 O64 B4 Si32 O72 B0.6 Si11.4 O24 Si13.8 B0.2 O28 Si72 B2 O148 Si62.4 B1.6 O128 Si15.6 B0.4 O32 Si54.7 B1.3 O112 B8 Ge16 O148 B1.5 Si52.5 O108 Si27 BO56

[81] [84] [78] [85] [86] [87] [88] [88] [89] [90] [91] [92]

b

SFE SFG SFH SFN SFS SOS SSF SSY a The

RUT framework was ﬁrst discovered as the aluminosilicate, NU-1. – The structure of MCM-70 has been determined, but the framework has not yet been approved by the Inter. Zeolite Assoc. Structure Commission. b Note

and there are a number of beryllium-containing natural zeolites such as lovdarite and nabesite that have 3R in their structures. Two 3R-containing zeolites, OSB-1 (OSO) and OSB-2 (OBW) [96], have been prepared with beryllium. The ionic ˚ is about the same as Zn2+ , but its substitution in radius of lithium (0.73 A) a silicate would generate an even higher charge than divalent tetrahedral atoms. Nonetheless, two lithosilicates, RUB-23 [97] and RUB-29 [98], have been synthesized to form tetrahedral framework structures with lithium-containing strain-free 3Rs. No lithosilicate structure has yet been approved by the International Zeolite Association Structure Commission.

2.8 Structure-Directing Agents

As mentioned before, the use of organic SDAs is one of the primary approaches to discovering new zeolites. It is therefore important to understand how to design new SDAs that will give new porous frameworks. The size, geometry, rigidity, and hydrophobicity are very important features of SDAs that determine their ability to form zeolite structures. SDAs that are very small and spherical in shape tend to form small cages and give nonporous clathrasil products, while larger ones tend to form porous, large-pore zeolites. Linear molecules such as long diquaternary ammonium cations typically give one-dimensional medium pore zeolites such as EU-1 (EUO) and ZSM-48 (∗ MRE). Even though they are large, these linear

67

68

2 Synthesis Approaches

molecules are quite ﬂexible and tend to give one-dimensional zeolites. More rigid templates have been found to be an important factor in forming large-pore three-dimensional frameworks such as CIT-1 (CON) [99] and MCM-68 (MSE) [100]. As large- and extra-large-pore zeolites are desirable, they can be made with even larger SDAs. As the size of the organic molecule increases, its hydrophobicity also increases, which limits its solubility in aqueous medium and also its ability to form solvated cations. The interaction of solvated cations and condensing silicate species in solution has a profound effect on the crystalline product that can form and is the basis for the structure-directing effect of organic cations. Zones et al. [101] have discussed the effect of the carbon to nitrogen ratio (C/N+ ) of organic molecules on their effectiveness to crystallize porous zeolites. They concluded that C/N+ ratios between 11 and 15 were optimal for the formation of high-silica molecular sieves. Molecules that are moderately hydrophobic give the best match between the SDA and the silica precursors to induce the formation of zeolitic-building units. When the SDA has excessive hydrophilicity, the silicate species are not strong enough to interrupt the extensive hydration sphere around the hydrophilic SDA [102]. To study the effect of the hydrophobic/hydrophilic character of a number of SDAs, Zones et al. have determined the ability of charged organic molecules to partition between an aqueous solution and an organic chloroform phase. The molecules that partition well between both aqueous and organic phases had the right balance between hydrophobicity and hydrophilicity to function as good SDAs [101]. By surveying the relationship of SDA characteristics and the type of silicates that they can form, Lobo et al. [73] have summarized the important trends of high silica syntheses (Table 2.4). The simple organic molecules such as TMA+ , TEA+ , TPA+ , and TBA+ (tetrabutylammonium) were the ﬁrst to be thoroughly investigated as promoters to synthesize new zeolite structures. These molecules were readily available commercial reagents. Later on, simple linear diquaternary cations such as hexamethonium (HM) (N,N,N,N ,N ,N -hexamethylhexane-1,6-diammonium) were investigated and Table 2.4

Trends between SDAs and zeolite structure types [73].

SDA type

Zeolite type

None Small globular molecules Excess alkali metal cations Linear molecules Branched molecules Large polycyclic molecules Large globular molecules Large globular molecules + Al Large globular molecules + B Large globular molecules + Zn

Dense phases Clathrasils Layered structures, mordenite One-dimensional medium-pore zeolites Three-dimensional medium-pore zeolites Three-dimensional large-pore zeolites One-dimensional large-pore zeolites Three-dimensional large-pore zeolites Three-dimensional large-pore zeolites VPI-8 a one-dimensional large-pore zeolite

2.8 Structure-Directing Agents

were also found to be excellent SDAs. To design and make more complex SDAs, early researchers began to make their own organic molecules by taking readily available primary, secondary, and tertiary amines and performing a simple alkylation to quaternize them. For example, by alkylating substituted piperidines and adamantanes, Nakagawa et al. [103] have discovered a large number of new zeolites such as SSZ-23, SSZ-35, SSZ-39, and SSZ-44. To make more complex linear diquaternary molecules, workers began to substitute pyrrolidine and piperidine on the end of the linear diamines instead of methyl, ethyl, and so on, to discover some new structures and compositions such as IM-5, TNU-9, and TNU-10, as discussed above. In addition to pyrrolidine and piperidine, Jackowski et al. [10] also investigated the use of n-methyl homopiperidine, m-methyl tropane, and quinuclidine as end-group substituents on linear diamines. With the addition of HF as mentioned above (Table 2.1), they were able to synthesize zeolite SSZ-74, a new two-dimensional medium pore zeolite, using the hexamethylene-1,6-bis(N-methyl-N-pyrrolidinium) cation as the SDA. Alkylation of substituted imidazoles has also been investigated and has been found to give mainly one-dimensional zeolites such as MTW, TON, and MTT [104]. Later, substituted imidazoles were found to make three new zeolites, ITQ-12 (ITW) in the presence of HF [41], SSZ-70 (structure unknown) in the presence of boron [105], and IM-16 (UOS) in the presence of germanium [65]. The CoAPO, SIZ-7 [106] (SIV), was also prepared with an imidazolium template. In an effort to discover new zeolites, more complex SDAs were designed to prepare organic molecules more diverse than those obtainable by the simple alkylation of readily available amines. By using Diels–Alder chemistry, Zones et al. [107] prepared a family of tricyclodecane derivatives to make MOR, MTW, CON, SSZ-31, and SSZ-37 zeolites. Calabro et al. [100] used the Diels–Alder reaction to prepare the N,N,N ,N -tetraalkyl-exo,exo-bicyclo[2.2.2]oct-7-ene-2,3:5,6-dipyrrolidinium cation, which was used to prepare MCM-68 (MSE), the ﬁrst zeolite to have an intersecting 12 × 10 × 10 ring channel system [108]. Using the Beckmann rearrangement reaction, Lee et al. [109] made a family of polycyclic SDAs to make a number of zeolites including SSZ-50, which has the RTH framework. Other multistep organic syntheses have been used, which include reduction of alkyl nitriles to make the SSZ-53 (SFH) and SSZ-55 templates [110], catalytic hydrogenation of substituted quinoline to give the bicyclic SDA for making SSZ-56 [111] (SFS), reductive amination of ketones to give the SSZ-57 [112] and SSZ-58 (SFG) templates [87], and the amination of an acyl halide to make the SSZ-59 (SFN) template [88] and SSZ-65 (SSF) template [91]. Both SSZ-53 and SSZ-59 are notable because their structures contain extra-large 14-ring pores. While these complex organic molecules are very proﬁcient in making new and interesting zeolite frameworks, their cost may be initially too expensive for them to be used as commercial catalysts. The challenge remains to ﬁnd ways to make these zeolites with cheaper SDAs. Besides using organic tetraalkylammonium cations as SDAs, other types of molecules have been used to make porous zeolites. Small amines have been found to be useful templates for many zeolite syntheses, although they tend to be nonspeciﬁc for a given framework [113]. Cyclic ethers have also been employed

69

70

2 Synthesis Approaches

by a number of researchers to make zeolites. Dioxane was used by De Witte et al. [114] to prepare MAZ zeolite and trioxane was used to prepare ECR-1 (EON) [115]. Larger cyclic crown ethers, 15-crown-5 and 18-crown-6, which are able to efﬁciently coordinate to alkali metal cations, have been used to prepare both high silica FAU and EMC-2 (EMT) zeolites, respectively [116]. Like the complex tetraalkylammonium SDAs described above, these crown ethers are expensive to use as templates and other routes to high-silica FAU and EMT are known, which use the more economical TPA [117] and triethylmethylammonium [118] SDAs, respectively. Zeolite UTD-1 was the ﬁrst 14-ring zeolite to be synthesized and that used the organometallic bis(pentamethylcyclo-pentadienyl) cobalt(III) cation as a template [119]. Tetraalkylphosphonium templates have also been shown to be effective SDAs. Both ITQ-26 (IWS) [120] and ITQ-27 (IWV) [43] are new multidimensional large-pore zeolites prepared with phosphonium templates. Because both structures contain D4R, they require either ﬂuoride or germanium in their syntheses as discussed above.

2.9 SDA Modeling

In addition to the efforts put into the syntheses of organic molecules, many studies have been undertaken to understand the energy of interaction between the SDA molecule and the structure of the zeolite. The reason for these studies is to rationalize which organic molecules promote the formation of a given zeolite framework or, better yet, to predict which SDA makes a hypothetical structure. Using Monte Carlo and energy minimization calculations, Lewis et al. have calculated the stability and location of the SDAs in a number of zeolites [121]. The effects of aluminum content on the effect of SDA–zeolite interactions and the SDA–SDA interactions in the zeolite pores have also been studied by Sastre et al. [122] using the General Utility Lattice Program (GULP) energy minimization code. They were able to explain the role the SDA has in determining the Si/Al range in the synthesis of some zeolite structures. They also used GULP to determine the effect on SDA–framework interactions by germanium incorporation in the EUO, ITH, IWW, and IWR frameworks. All four of these zeolites are made with the HM SDA. They concluded that the low rigidity of the HM and relatively low SDA–framework interaction resulted in the lower selectivity of this template to promote the formation of one structure over another. The zeolite that formed was therefore determined by the effects of germanium and aluminum on the stability of a given framework. The EUO framework was the subject of another study to understand the catalytic differences between EU-1 materials made with different templates. By using molecular modeling to dock the HM and dibenzyldimethylammonium (DBDMA) SDAs in the EUO framework, they were able to locate the position of these two templates in the pores. By rationalizing that the electrostatic interactions would be most favorable when the positive charge of the SDA was located near the aluminum atoms in the structure, they were then able to determine the location of

2.9 SDA Modeling

the active Al sites. With this information, they were able to explain the selectivity of products in the n-decane cracking test. The active sites for HM-prepared EU-1 were determined to be in both the 10-member ring channels and the larger side pockets, while the active sites were only located in the side pockets of the DBDMA-prepared sample. ZSM-18 (MEI) is a one-dimensional 12-ring zeolite originally prepared with an expensive triquaternary SDA. By using molecular dynamics, Schmitt et al. [123] were able to predict potential ZSM-18 SDAs that were easier to synthesize. This was the ﬁrst time that molecular modeling was successfully used to design a template for a speciﬁc framework. In the previous paragraph, the importance of being able to predict which SDA would promote a known or hypothetical structure was introduced. Hypothetical structures were ﬁrst generated manually by early zeolite researchers who observed the chains, layers, and polyhedra found in known zeolite structures. By connecting these units in different manners, or by applying symmetry operators, several thousand new structures were generated. Joe Smith from the University of Chicago enumerated and began to keep a database of these structures [124]. Recently, several groups have developed strategies for automating the generation of new structures. With the recent advancement of computational speed, millions of hypothetical structures have been identiﬁed. Using a symmetry-constrained intersite bonding search method, Treacy et al. [125] have generated over 100 000 plausible zeolite frameworks. Simulated annealing was used by Earl and Deem [126] to generate over 4 million hypothetical structures, of which 450 000 were found to be plausible structures by comparing their lattice energies, calculated by the GULP program, to the energies of known zeolite structures [127]. Other researchers have used the tiling theory to generate new zeolite structures [128] and have systematically enumerated 1761 both simple and quasi-simple uninodal and simple binodal and trinodal frameworks. Of these, 176 were found to be chemically feasible by using calculated lattice energy, framework density, and other structural parameters [129]. These databases can now be searched for structures that could be candidates as useful absorbents for separating molecules or as shape selective catalysts for industrially important chemical processes. In particular, multidimensional, extra-large pore (pore openings with greater than 12 tetrahedral atoms) frameworks can be identiﬁed. One way to directly generate these type of structures with deﬁned pore geometry is the method of constrained assembly of atoms [130]. By deﬁning a forbidden zone, where the framework atoms cannot reside, zeolite frameworks can be generated with a given pore size and dimensionality. While these new structures have generated considerable attention, they remain a Gedanken experiment until someone discovers a way to synthesize them. The syntheses of inorganic zeolites have been thoroughly investigated for decades, and it is now rare that a new zeolite is discovered without the use of a SDA. A more recent notable example is MCM-71, a 10-ring zeolite of the mordenite group, obtained in the K–Al–Si system [131]. The ultimate goal is to be able to design an SDA to promote the crystallization of a given hypothetical structure. A close match between the size and shape of the organic molecule and a zeolite pore can help stabilize the growing crystal. As discussed above, molecular modeling

71

72

2 Synthesis Approaches

may help if a database of organic molecules is available for evaluation of the energy of interaction between the SDA and the zeolite framework. This strategy is limited by the ability to be able to visually propose an SDA, which has a good ﬁt between its energy minimized conformation and the shape of the pores and/or cages of the target framework. A de novo strategy has been developed by Lewis et al. [132] in which organic molecules are compositionally grown right in the pore to be modeled. Starting from a seed molecule placed inside the targeted zeolite host, organic fragments from a library are subsequently attached to systematically build up the size and shape of the developing template to match the size and shape of the pore. A cost function based on the overlap of van der Waals spheres was used to control the growth of the new template. While the authors were able to predict a suitable SDA for the known LEV structure, it still remains to be seen if this new computer algorithm, called ZEBEDDE, can predict a suitable organic SDA to template one of the many hypothetical structures discussed above.

2.10 Co-templating

The idea of cotemplating has been in existence for some time. The concept involves using two SDAs, each one templating a different SBU in the target structure. This technique was ﬁrst used by Flanigen et al. to synthesize SAPO-37 (FAU) [133]. It was found that SAPO-37 could only be prepared in the presence of two SDAs, TPA+ and TMA+ . Apparently the large TPA+ cation stabilizes the large FAU supercage and the smaller TMA+ cation stabilizes and promotes the formation of the smaller sod cages. The TMA+ cation was also used to help stabilize the sod cage in the synthesis of ITQ-29 as discussed previously. In a more recent study, Wright et al. at the University of St. Andrews used molecular modeling to rationalize the synthesis of two new SAPO materials, STA-7 (SAV) and STA-14 (KFI) [134]. Both frameworks have two types of cages, which are templated by two different SDAs. In the former case, the SAPO composition of SAV was prepared for the ﬁrst time by using cyclam and TEA+ as SDAs. In the latter case, the KFI framework was prepared for the ﬁrst time with a phosphate composition by using a mixture of azaoxacryptand template and TEA+ . The TEA+ cation was found by single crystal structure determination to be located in the smaller mer cage and the larger azaoxocrypd and molecule was believed to be in the larger alpha cage. The authors of this work suggested that this cotemplating approach may be a way of rationalizing the synthesis of new hypothetical structures that contain more than one cage. Besides promoting the synthesis of structures having more than one cage, the dual templating approach may also be used to economize the synthesis of a zeolite that uses an expensive template. Zones and Nakagawa developed a synthesis method where a small amount of expensive SDA cation is utilized to nucleate the formation of a given structure and a second inexpensive amine is added to ﬁll up and stabilize the cages or pores of the structure during crystallization. Using this approach, they were able to synthesize a number of zeolites requiring complex

2.11 Layered Precursors

templates, such as SSZ-25, SSZ-28, SSZ-32, and SSZ-35, by using a smaller amount of the expensive SDA and by adding the second amine such as isopropyl amine or isobutylamine. The key feature is that the amine is smaller than the organic SDA and the organic SDA is used in an amount less than that required to crystallize the zeolite in the absence of the amine [135]. Later, Zones and Huang found that the mixed template system can accelerate the crystallization rate of zeolite syntheses and also nucleate the synthesis of a new zeolite named SSZ-47 [136]. A more thorough discussion of cotemplating in zeolites is given in Chapter 13 of this book. One type of cotemplating strategy was developed by UOP is called charge density mismatch (CDM) [137]. In this technique, a precursor gel is ﬁrst prepared in which there is a charge density mismatch between the organic cations (low density) and the potential aluminosilicate (high density at low Si/Al ratios) material that is expected to form. One part of this strategy is to use readily available and cheap SDAs to see if they could be induced to make new zeolites and another part of this strategy is to use a mixture of the two SDAs. The mixture is heated for some time at a suitable temperature, but crystallization from this initial CDM solution is difﬁcult or impossible. These solutions were found to be stable at 100–150 ◦ C for a few days to a few weeks. After a given amount of time, a second high charge density cation, such as sodium and/or potassium, or even a small organic SDA such as TMA+ , was added, which then allowed the mixture to crystallize into a zeolite. The one new zeolite discovered using the CDM approach using both TMA+ and TEA+ templates was UZM-5 (UFI) [138], a new two-dimensional framework with D4R and alpha cages like those in the LTA framework. Other materials made with the CDM method so far are compositional variants of known zeolites and it is yet to be shown that this method is a productive approach for discovering new framework materials. However, there have been a number of new compositions of known frameworks that have allowed these materials to be studied as adsorbents. For example, new compositions with the MEI and BPH frameworks have been found that are stable to calcination (Figure 2.3).

2.11 Layered Precursors

Whereas zeolites are three-dimensional crystalline materials containing tetrahedral framework atoms, clays have two-dimensional structures containing both octahedral and tetrahedral coordination. The two-dimensional building units can be regarded as layers that are aligned along the third dimension by ionic or hydrogen bonds. Researchers in the late 1970s realized that the interlayer inorganic cations of CT

CDM Si/Al = 1

Forced Dual templating cooperative templating Charge density mismatch (CDM)

Figure 2.3

Traditional organic templating Charge density match Si/Al = ∞

Deﬁned templating regions for the CDM approach to zeolite synthesis [137].

73

74

2 Synthesis Approaches

smectite clays could be exchanged and replaced with larger oligomeric polycations such as the Keggin-like chlorohydrol cation, [Al13 O4 (OH)24 (H2 O)12 ]7+ [139, 140]. Upon calcination, the cations are converted to alumina pillars that keep the layers separated and provide for an interlayer microporous network. Large polycationic zirconia and titania species were also found to be effective for pillaring clays. Some zeolite structures can be described as being built from layers of tetrahedral aluminum and silicon oxide. While these layers are fully connected with covalent Si–O and/or Al–O bonds in the third dimension, it has been sometimes found that during its synthesis a precursor is formed where the layers of the zeolite are not connected by covalent bonding, but are connected by hydrogen bonding and/or kept propped open by the organic SDA. This precursor can, in some cases, be isolated as a stable solid, which can, in many cases, irreversibly transform to the fully tetrahedrally connected zeolite upon calcination. The ﬁrst unequivocally identiﬁed zeolite precursor was for the structure MWW and was designated MCM-22(P) [141]. This was exploited for the preparation of the ﬁrst pillared zeolite MCM-36, which is a high-activity micro/mesoporous composite, and which launched the discovery of the extensive family of distinct zeolites based on different packings of MWW layers [142, 143]. By using the ﬂuoride synthesis method and a bulky 4-amino-2,2,6,6-tetramethylpiperidine template, Schreyeck et al. [144] were able to prepare PREFER, a layered microporous aluminosilicate with (100) layers of ferrierite. When the layers in PREFER condense to form the FER structure, they are aligned in a manner to form a mirror plane parallel between the layers to give Immm symmetry. It has been found that other SDAs can force the FER layers to arrange in a different manner so as to form a new framework structure designated CDO. MCM-65 was synthesized using both quinuclidinium+ and TMA+ as cotemplates in the presence of sodium [145]. The powder XRD pattern of the as-synthesized material gave a C-centered orthorhombic structure. Upon calcination, the XRD pattern showed a reduction of 4.5 A˚ of the longest axis, indicating that a new structure had formed, possibly by condensation of a layered precursor. The framework of calcined MCM-65 was determined by model building after examination of the electron diffraction patterns. It has Cmcm symmetry and is built from the same layers as those found in FER, but with no mirror plane parallel to the layers [146] (Figure 2.4). Examination of the powder XRD and electron diffraction patterns of ZSM-52, ZSM-55, both prepared with choline SDA, and MCM-47, prepared with bis(N-methylpyrrolidinium)-1,4 butane, indicated that these materials have very similar structures to as-synthesized MCM-65. The structure of as-synthesized MCM-47 was determined by Burton et al. [147] from powder XRD data using the FOCUS and ZEFSaII algorithms. The location of the template was found to be located between the FER layers. While MCM-65 can be readily calcined to form a condensed crystalline structure, MCM-47 cannot. Two other materials, CDS-1 [148], prepared with TMA+ and dioxane, and UZM-25 [149], prepared from ethyltrimethylammonium or dimethyldiethylammonium, were also found to have the CDO structure. Other zeolite structures were also found to be formed through layered precursors. Nu-6(2), having the NSI framework, was formed from the Nu-6(1) precursor made with the 4,4 -bipyridyne SDA [150]. Another related material, called EU-20b, formed

2.11 Layered Precursors

(a)

Ferrierite (FER)

(b) MCM-65 (COD)

Figure 2.4 Frameworks of (a) ferrierite and (b) MCM-65, showing the different manner in which the layers are connected.

from calcining its precursor, EU-19 prepared with a piperazinium SDA, was found to form an intergrowth material composed of NSI and CAS framework layers [151]. Acetic acid treatment was found to decrease the distance between the layers of ˚ allowing the conversion of the material to the pure-silica RUB-15 from 14 to 7.7 A, SOD framework upon calcination [152]. The key to the success of the acetic acid treatment was that it allowed the layers to shift one half unit cell along the c-axis to align the layers, which does not occur by treatment with hydrochloric acid. The RUB-18 layered material was found to transform into RUB-24 zeolite having the two-dimensional eight-ring RWR framework [153]. Although RUB-24 was not fully crystalline and contained some structural defects, its crystallinity was good enough to determine its structure. The two-dimensional zeolite, RUB-41, having a 10 × 8 ring RRO framework channel system, was formed by calcining the precursor, RUB-39, prepared with the dimethyldipropylammonium template [154]. One of the more important zeolites to be prepared from a layered precursor is MCM-22, which was ﬁrst prepared in 1984 by BASF as the impure phase, PSH-3 [155]. The conditions for making it without impurities were later found by Mobil Oil Corporation. [156], who designated their material, MCM-22. The structure of the calcined material was determined by Leonowicz et al. [157] in 1994, and was found to have the MWW framework, which has two-dimensional pores of 10 rings, but with 12-ring pockets on the surface of the crystal. It was found that MCM-22 is a good catalyst for the alkylation reaction of benzene and propylene to make cumene, which was determined to occur on the surface of the crystals in the 12-ring pockets [158]. The realization of the importance of surface chemistry on the MCM-22 catalyst performance and the fact that this and a number of other zeolites are formed from layered precursors led to the development of new synthesis approaches aimed at maximizing the surface chemistry of these materials. By increasing the size of the pore opening to the site, or by exposing the active sites that were originally inside the pores to the surface of the crystal, larger molecules could gain access

75

76

2 Synthesis Approaches

to the catalytically active sites. One approach was to ﬁrst swell and then pillar the layers in a manner similar to that used to make pillared clays. Kresge et al. [159] were able to swell the MCM-22 layered precursor, MCM-22(P), by heating with a solution of cetyltrimethylammonium and TPAOH. The swollen material, MCM-36, could then be pillared with the D4R TMA silicate or tetraethylorthosilicate (TEOS) to give a porous material after calcination, having greatly increased surface area and activity [160]. Later, in a similar manner, Chica et al. [161] were able to swell and pillar the PREFER material to give ITQ-36, a pillared ferrierite. A new method of postalkoxysilyation on zeolitic lamellar precursors was recently developed by Wu et al. [162]. By treatment with diethoxydimethylsilane, they were able to directly expand the pore apertures between the layers of MWW, FER, CDO, and MCM-47 structures under low pH conditions without ﬁrst swelling the layers. The expanded structures containing Si–O–Si linkages between the layers were stable to calcination or reﬂux in water. These new materials, designated IEZs (interlayer expanded zeolites), showed catalytic activities indicative of structures having larger pores than those obtained by direct calcination of their nonsilylated precursors. A second approach for maximizing the surface area of zeolites formed from layered precursors is by delamination of the layers to give very thin sheets of the crystals. By ﬁrst using the method of Kresge et al. [159], researchers from Polytechnic University of Valencia, PQ Corporation and Shell International Chemicals, B. V., swelled the MWW layers of an MCM-22 material and then delaminated them by a simple one-hour ultrasonic treatment [163]. The delamination was conﬁrmed by TEM imaging, nitrogen absorption, and by the increased presence of surface hydroxyl groups compared to the MCM-22 zeolite as determined by IR. This new material, called ITQ-2, was found to have greatly increased the cracking activity for diisopropylbenzene and vacuum gas oil. Using the same technique, Corma et al. [164] were able to obtain similar results with PREFER. This material, called ITQ-6, showed diisopropylbenzene cracking activity four times that of conventional ferrierite. The titanium form of ITQ-6 was found to be more active than either TiFER or Ti-Beta for the epoxidation of 1-hexene with hydrogen peroxide. Likewise, NU-6(2) was delaminated to give MCM-39-Si [165] and ITQ-18 [166]. Delamination of NU-6(2) was conﬁrmed by nitrogen absorption, 29 Si MAS NMR, and IR spectroscopy. Yet another material, ITQ-20 [167], was prepared by the delamination of its layered precursor, PREITQ-19. Zeolite PREITQ-19 apparently gives the CDO structure upon calcination. By modiﬁcation of the synthesis of MCM-22, a new material called MCM-56 was made [168]. The structure of this zeolite is believed to be similar to MCM-22, but with layers that are not connected in a regular manner [143]. Evaluation of MCM-56 showed surface area and catalytic activity intermediate between ITQ-2 and MCM-22. A summary of these layered precursors and their related materials are given in Table 2.5. The continuing discovery of new layered zeolite precursors and especially of the long-known structures like FER and SOD led to a postulate of a new fundamental principle applicable to many, if not all, microporous structures, that is, that they can be generated by two pathways: a direct assembly in 3D or, indirectly, via the 2D layered precursor route [142].

2.12 Nonaqueous Solvents Table 2.5

Zeolites formed from layered precursors.

Material Layered precursors PREFER MCM-22(P) SSZ-25 as syn ERB-1 as syn NU-6(1) EU-19 RUB-15 RUB-18 RUB-39 PREITQ-19 MCM-65 as syn PLS-1 UZM-13 UZM-17 UZM-19 HLS [F, Tet-A]-AlPO-1 SAPO-34 prephase Delaminated zeolites ITQ-2 MCM-56 ITQ-6 ITQ-18 ITQ-20 Pillared/expanded zeolites MCM-36 ITQ-36 IEZ Others PSH-3 ZSM-52 ZSM-55 MCM-47 MCM-49

Description

Reference

Ferrierite (FER) precursor MCM-22 (MWW) precursor SSZ-25 calcined (MWW) precursor ERB-1 calcined (MWW) precursor NU-6(2) (NSI) precursor EU-20b (CAS-NSI intergrowth) precursor Silica SOD precursor RUB-24 (RWR) precursor RUB-41 (RRO) precursor ITQ-19 (CDO) precursor MCM-65 calcined (CDO) precursor CDS-1 (CDO) precursor UZM-25 (CDO) precursor UZM-25 (CDO) precursor UZM-25 (CDO) precursor Ga-SOD precursor AlPO-41 (AFO) precursor SAPO-34 precursor

[144] [156] [169] [170] [150] [151] [152] [153] [154] [171] [145] [148] [149] [149] [149] [172] [173] [174]

Delaminated MWW Disordered/delaminated MWW Delaminated FER Delaminated NSI Delaminated CDO

[163] [175] [164] [166] [167]

Pillared MWW Pillared FER Expanded MWW, FER, CDO, and MCM-47

[159] [161] [162]

Impure MWW CDO precursor that does not fully condense CDO precursor that does not fully condense CDO precursor that does not fully condense MWW directly synthesized

[155] [176] [177] [147] [178]

2.12 Nonaqueous Solvents

Water is the solvent of choice for synthesizing zeolites because of its ability to solubilize the components needed to form porous crystalline materials. Other solvents, such as alcohols, amines, ammonia, and ionic liquids have been used to make zeolites and MeAlPOs with limited success. Bibby and Dale [179] were the ﬁrst to

77

78

2 Synthesis Approaches

report the synthesis of zeolite in an organic solvent in 1985. Using ethylene glycol or propanol, they were able to synthesize silica-sodalite for the ﬁrst time by using TMA as SDA. Later, the pentasils silicalite (MFI), ZSM-39 (MTN), and ZSM-48 (∗ MRE) were made using ethylene glycol, butanol, and glycerol as solvents [180]. Kuperman et al. [181] made giant crystals of MTN, FER, and MFI zeolites using pyridine/HF and triethylamine solvent systems. Using ethylene glycol as a solvent, Huo and Xu were able to prepare the AFI, AEL, and AWO AlPO materials [182]. A few years later, Huo et al. prepared a new material, designated JDF-20, with glycol solvents [183]. The structure of this material was found to have 20 T-atom pore openings like those of cloverite (-CLO), and an interrupted framework with an Al : P ratio of 5 : 6. These and other nonaqueous zeolite syntheses were reviewed by Morris and Weigel in 1997 [184]. Ethylene glycol was found to be a particularly useful solvent for making new gallium phosphate materials such as the CGF and CGS frameworks [71]. Researchers at Mobil used a two-liquid (aqueous and immiscible organic) reaction system to slowly introduce Si to the growing crystals of SAPO materials [185]. Phase-transfer of the organic silicon reagent, tetraethylorthosilicate, from the organic hexanol phase and its hydrolysis at the interphase with the aqueous phase are suggested as critical steps in the substitution of silicon into the framework of the SAPO MCM-9 (VFI) [186]. More recent work has shown the advantage of using a soluble organic cosolvent over the insoluble hexanol reagent. Highly active SAPO-11 (AEL) material, named ECR-42, could be prepared using ethanol as a cosolvent with an organic source of silicon, such as tetramethylorthosilicate or TEOS. It was proposed that a monomeric source of silica is the optimum way to introduce silicon into the framework of AlPO materials [187]. Increased isomerization rates of n-decane and cracking rates of hexane and hexene along with 29 Si NMR characterization conﬁrmed the increased activity and better Si distribution within the crystals of ECR-42 over conventionally prepared SAPO-11 [188]. Because liquid and aqueous ammonia are known to solubilize alkali metal cations, its use in zeolite syntheses were investigated by Vaughan and Strohmaier [189]. While no crystalline materials were formed in pure ammonia, aqueous ammonia yielded high-silica cancrinite (CAN) with Si/Al ratios up to 1.96 using preformed silica–alumina gels with sodium hydroxide. The high-silica-containing framework was conﬁrmed by 29 Si MAS NMR. Lithium cations gave ABW, EDI, and PHI/MER products, while potassium gave EDI, GIS, and PHI/MER materials. More open framework materials were not observed. Later, they were able to prepare ˚ with aqueous ammonia using potassium hydroxide a nanocrystalline LTL (<200 A) and a silica–alumina cracking catalyst as a starting reagent [190]. A short time later, Garces et al., from Dow Chemical Co., used liquid ammonia (NH3 /H2 O = 9) to prepare Na-P1(GIS), dodecasil 3C (MTN), octadecasil (AST), and DSM-8, a pure-silica beta (∗ BEA) [191]. More recently, ionic liquids and eutectic mixtures have been investigated by the University of St. Andrews as alternate solvents for the synthesis of AlPO zeolite analogs [192]. Ionic liquids are salts that are ﬂuid at ambient temperatures and contain ionic species, while eutectic mixtures are higher melting organic salts to

2.13 Summary and Outlook

which other compounds are added to suppress their melting point temperature. Common ionic liquids are dialkylimidazolium salts, and a mixture of choline chloride with urea is an example of a eutectic mixture. Ionic liquids have low vapor pressures and the advantage that, at common crystallization temperatures (100–200 ◦ C), the reaction can be performed in open vessels as compared to the sealed autoclaves normally employed in the hydrothermal syntheses of zeolites and AlPOs. It was found that ionic liquids not only solubilize the starting reagents but can also act as a template. Using these new solvent systems, the group at St. Andrews prepared three known AlPO frameworks, SIZ-3 (AEL), SIZ-4 (CHA), and SIZ-5 (AFO). In the cobalt AlPO system, they were able to prepare two known frameworks, SIZ-8 (AEI) and SIZ-9 (SOD), and one new framework, SIZ-7 (SIV) [106]. The structure of SIZ-7 is similar to that of the related three-dimensional, eight-ring frameworks PHI, MER, and GIS. Ionothermal synthesis, or the use of ionic liquids and both solvents and templates, for the synthesis of zeolite analogs and metal-organic frameworks (MOFs) has been recently reviewed by Parnham and Morris [193]; a more detailed description of the approach is discussed in Chater 6 of this book.

2.13 Summary and Outlook

The primary approaches for discovering new zeolite materials have been discussed. While the earlier synthesis work focused on the effects of alkali metals, later work shifted to utilizing organic SDAs. Later, new framework T-atoms, such as B, Ga, Ge, P, Zn, and Be, were evaluated and were found to be successful in ﬁnding many new zeolite frameworks. While many of the these new frameworks have unique pore structures, especially the germanium-containing ITQ materials, their compositions are such that they are expensive and not as thermally stable as conventional aluminosilicates. This may limit their use as catalysts; therefore, a method has to be found to synthesize these new and interesting structures as aluminosilicate compositions. The realization that some zeolites form layered precursors has been exploited and new, active analogs have been prepared by pillaring, expansion, and delamination techniques. Furthermore, it may be a general phenomenon that many, if not all, frameworks can be assembled by two pathways: direct 3D growth or via a layered precursor. Several such pairs are known and are being discovered. In recent years, the effects of using ﬂuoride as a mineralizer, low water ratios, and new solvent systems, such as ionic liquids, have been investigated. These techniques have generated a number of new materials, especially new high-silica compositions of known zeolites and new low-density frameworks with multidimensional pores. Much research has also been carried out in supramolecular templating to prepare zeolites with meso and macropores; however, these researches have not been useful for discovering new frameworks and hence have not been discussed. Stirring and seeding effects are also important variables in crystallization, having a large effect

79

80

2 Synthesis Approaches

on purity, crystal size, and morphology, but they are generally used to optimize the synthesis of a zeolite after it has been discovered. There are a large number of variables to explore in a zeolite synthesis program and many experiments (typically thousands) have to be performed before a new zeolite is discovered. In the past 10 years, new, automated, high-throughput techniques, which allow many more experiments to be performed in a given amount of time, have been disclosed. Researchers at Sintef [194], Purdue University [195], Max Planck Institute [196], Jilin University [197], Polytechnic University of Valencia [198], and others have all described high-throughput equipment (HTE) for synthesizing microporous materials. Although many groups now have HTE, the rate of zeolite discovery has not increased signiﬁcantly in the past 10 years. The discovery of new zeolites and porous materials has always been initiated by new compositional ideas and synthesis approaches. There will deﬁnitely be new ideas to exploit in the future, which will lead to new frameworks and new compositional and structural variants of known materials. High-throughput techniques will allow these new ideas to be exploited at a much faster rate. As we continue to more thoroughly understand how zeolites crystallize, characterize their structures, and ﬁnd new methods to modify their properties, we will continue to increase their utility and ﬁnd new applications for these fascinating materials.

Acknowledgments

The help of Wieslaw J. Roth in preparing this manuscript is greatly appreciated. The support of ExxonMobil Research and Engineering Company, Corporate Strategic Research Laboratories is also acknowledged.

References 1. Barrer, R.M. (1948) J. Chem. Soc., 2. 3. 4. 5. 6.

7.

2158. Reed, T.B. and Breck, D.W. (1956) J. Am. Chem. Soc., 78, 5972–5977. Milton, R.M. (1959) US Patent 2,882,244. Breck, D.W. (1974) US Patent 3,130,007. Barrer, R.M. and Denny, P.J. (1961) J. Chem. Soc., 971–982. Wilson, S.T., Lok, B.M., Messina, C.A., Cannan, T.R., and Flanigen, E.M. (1982) J. Am. Chem. Soc., 104, 1146–1147. Vaughan, D.E.W. (1999) US Patent 5,976,491.

8. Afeworki, M., Dorset, D.L., Kennedy,

9. 10.

11. 12.

G.J., and Strohmaier, K.G. (2004) Stud. Surf. Sci. Catal., 154, 1274–1281. Hong, S.B. (2008) Catal. Surv. Asia, 12, 131–144. Jackowski, A., Zones, S.I., Hwang, S.-J., and Burton, A.W. (2009) J. Am. Chem. Soc., 131, 1092–1100. Flanigen, E.M. and Patton, R.L. (1978) US Patent 4,073,865. Guth, J.L., Kessler, H., Higel, J.M., Lamblin, J.M., Patarin, J., Seive, A., Chezeau, J.M., and Wey, R. (1989) in Zeolite Synthesis, ACS Sympoiusm Series, Vol. 398 (eds M. Occelli and H. Robson), Oxford Univ. Press, New York, p. 176.

References 13. Estermann, M., McCusker, L.B.,

14.

15.

16.

17.

18.

19.

20.

21.

22. 23. 24.

25.

26.

27.

Baelocher, Ch., Merrouche, A., and Kessler, H. (1991) Nature, 353, 320–323. Merrouche, A., Patarin, J., Kessler, H., and Anglerot, D. (1992) European Patent Application 0 497 698. Simmen, A., Patarin, J., and Baerlocher, Ch. (1993) in Proceedings of 9th International Zeolite Conference (eds R. von Ballmoos, J.B. Higgins, and M.M.J. Treacy), Butterworth-Heinemann, Boston, pp. 443–440. Loiseau, T., Retoux, R., Lacorre, P., and F´erey, G. (1994) J. Solid State Chem., 111, 427–436. Bonhomme, F., Thomas, S.G., Rodriguez, M.A., and Nenoff, T. (2001) Microporous Mesoporous Mater., 47, 185–194. Guth, F. (1989) PhD thesis, Synthesis and Characterizations of Crystallized Microporous Solids Containing AL, P and Sl, University de Haute Alsace, Mulhouse. Halvorsen, E.N. (1996) PhD thesis, Synthesis and Characterization of Aluminophosphate Molecular Sieves, University of Oslo, Norway. ˚ H., Akporiaye, D.E., Fjellvag, Halvorsen, E.H., Haug, T., Karlsson, A., and Lillerud, K.P. (1996) Chem. Commun., 1553. ˚ H., Akporiaye, D.E., Fjellvag, Halvorsen, E.H., Hustveit, J., Karlsson, A., and Lillerud, K.P. (1996) Chem. Commun., 601. Barthomeuf, D. (1994) Zeolites, 14, 394–401. Keuhl, G.H. (1988) US Patent 4,786,487. Matsukata, M., Nishiyama, N., and Ueyama, K. (1993) Microporous Mesoporous Mater., 1, 219. Kim, M.H., Li, H.X., and Davis, M.E. (1993) Microporous Mesoporous Mater., 1, 191. Matsukata, M., Ogura, M., Osaki, T., Rao, P.R.H.P., Nomura, M., and Kikuchi, E. (1999) Top. Catal, 9, 77. Xu, W., Dong, J., Li, J., Li, W., and Wu, F. (1990) J. Chem. Soc., Chem. Commun., 755.

28. Rao, P.R.H.P., Ueyama, K., and

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

44.

Matsukata, M. (1998) Appl. Catal. A: Gen., 166, 97–103. Koyama, Y., Ikeda, T., Tatsumi, T., and Kubota, Y. (2008) Angew. Chem. Int. Ed. Engl., 47, 1042–1046. Camblor, M., Corma, A., and Valencia, S. (1996) Chem. Commun., 2365. Wadlinger, R.L., Kerr, G.T., and Rohrbaugh, W.J. (1967) US Patent 3,308,069. D´ıaz-Caba˜ nas, M.J., Barrett, P.A., and Camblor, M.A. (1998) Chem. Commun., 1881. Camblor, M.A., Villaescusa, L.A., and D´ıaz-Caba˜ nas, M.J. (1999) Top. Catal., 9, 59–76. Lobo, R.F., Tsapatsis, M., Freyhardt, C.C., Chan, I., Chen, C.-Y., Zones, S.I., and Davis, M.E. (1997) J. Am. Chem. Soc., 119, 3732–3744. Zones, S.I., Hwang, S.-J., Elomari, S., Ogino, I., Davis, M.E., and Burton, A.W. (2005) C. R. Chim., 8, 267–282. Zones, S.I., Burton, A.W., Lee, G.S., and Olmstead, M.M. (2007) J. Am. Chem. Soc., 129, 9096–9079. Caullet, P., Guth, J.L., Hazm, J., Lamblin, J.M., and Gies, H. (1991) Eur. J. Solid State Chem., 28, 345. Camblor, M.A., Corma, A., Lightfoot, P., Villaescusa, L.A., and Wright, P.A. (1997) Chem. Commun., 2659–2661. Barrett, P.A., Camblor, M.A., Corma, A., Jones, R.H., and Villaescusa, L.A. (1997) Chem. Mater., 9, 1713–1715. Villlaescusa, L.A., Barrett, P.A., and Corma, A. (1999) Angew. Chem. Int. Ed. Engl., 38, 1997–2000. Barrett, P.A., Boix, T., Puche, M., Olson, D.H., Jordan, E., Koller, H., and Camblor, M.A. (2003) Chem. Commun., 2114–2115. Corma, A., Puche, M., Rey, F., Sankar, G., and Teat, S.J. (2003) Angew. Chem. Int. Ed. Engl., 42, 1156–1159. Dorset, D.L., Kennedy, G.J., Strohmaier, K.G., D´ıazCaba˜ nas, M.J., Rey, F., and Corma, A. (2006) J. Am. Chem. Soc., 128, 8862–8867. Zones, S.I., Burton, A.W., and Ong, K. (2007) International Patent WO 2007/079038.

81

82

2 Synthesis Approaches 45. Baerlocher, Ch., McCusker, L.B., and

46.

47. 48. 49.

50.

51. 52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

Olson, D.H. (2007) Atlas of zeolite framework types, IZA Structure Commission, 6th edn, Elsevier, Amsterdam. 16–111 Koller, H., W¨olker, A., Eckert, H., and Panz, C. (1997) Angew. Chem. Int. Ed. Engl., 36, 2823. Li, H. and Yaghi, O.M. (1998) J. Am. Chem. Soc., 120, 10569–10570. Wragg, D.S., Morris, R.E., and Burton, A.W. (2008) Chem. Mater., 20, 1561. Blasco, T., Corma, A., D´ıaz-Caba˜ nas, M.J., Rey, F., Vidal-Moya, J.A., and Zicovich-Wilson, C.M. (2002) J. Phys. Chem. B, 106, 2634–2642. Zwijnenburg, M.A., Bromley, S.T., Jansen, J.C., and Maschmeyer, T. (2004) Microporous Mesoporous Mater., 73, 171. Sastre, G. and Corma, A. (2006) J. Phys. Chem. B, 110, 17949–17959. Corma, A., Navarro, M.T., Rey, F., Rius, J., and Valencia, S. (2001) Angew. Chem. Int. Ed. Engl., 40, 2277–2280. Conradsson, T., Dadachov, M.S., and Zou, X.D. (2000) Microporous Mesoporous Mater., 41, 183–191. Corma, A., D´ıaz-Caba˜ nas, M.J., Rey, F., Nicolopoulus, S., and Boulahya, K. (2004) Chem. Commun., 1356–1357. Corma, A., D´ıaz-Caba˜ nas, M.J., and Rey, F. International (2005) Patent Application 2005/108526. Paillaud, J.-L., Harbuzaru, B., Patarin, J., and Bats, N. (2004) Science, 304, 990–992. Corma, A., D´ıaz-Caba˜ nas, M.J., Martinez-Triguero, J., Rey, F., and Ruiz, J. (2002) Nature, 418, 514–517. Corma, A., Rey, F., Valencia, S., Jorda, J.L., and Rius, J. (2003) Nat. Mater., 2, 493–497. Castaneda, R., Corma, A., Fornes, V., Rey, F., and Ruiz, J. (2003) J. Am. Chem. Soc., 125, 7820–7821. Dorset, D.L., Strohmaier, K.G., Kliewer, C.E., Corma, A., D´ıaz-Caba˜ nas, M.J., Rey, F., and Gilmore, C.J. (2008) Chem. Mater., 20, 5325–5331. Corma, A., D´ıaz-Caba˜ nas, M.J., Jord´a, J.L., Mart´ıez, C., and Moliner, M. (2006) Nature, 443, 842.

62. Corma, A., D´ıaz-Caba˜ nas, M.J.,

63.

64.

65.

66.

67.

68.

69. 70.

71.

72.

73.

74.

75. 76. 77.

78.

Jord´a, J.L., Rey, F., Sastre, G., and Strohmaier, K.G. (2008) J. Am. Chem. Soc., 130, 16482–16483. Tang, L., Shi, L., Bonneau, C., Sun, J., Yue, H., Ojuva, A., Lee, B.-L., Kritikos, M., Bell, R.G., Bacsik, Z., Mink, J., and Zou, X. (2008) Nat. Mater., 7, 381–385. Mathieu, Y., Paillaud, J.-L., Caullet, P., and Bats, N. (2004) Microporous Mesoporous Mater., 75, 13–22. Lorgouilloux, Y., Dodin, M., Paillaud, J.-L., Caullet, P., Michelin, L., Josien, L., Ersen, O., and Bats, N. (2009) J. Solid State Chem., 182, 622–629. U.S. Geological Survey (2008) Germanium-Statistics and Information, U.S. Geological Survey, Mineral Commodity Summaries. Cantin, A., Corma, A., Diaz-Cabanas, M.J., Jorda, J.L., and Moliner, M. (2006) J. Am. Chem. Soc., 128, 4216–4217. Corma, A., Rey, F., Rius, J., Sabater, M.J., and Valencia, S. (2004) Nature, 431, 287–290. Newsam, J.M. and Vaughan, D.E.W. (1986) Stud. Surf. Sci. Catal., 28, 457. Krutskaya, T.M., Kolyshev, A.N., Morozkova, V.E., and Berger, A.S. (1985) Russ. J. Inorg. Chem., 30, 438–442. Chippendale, A.M. and Cowley, A.R. (1998) Microporous Mesoporous Mater., 21, 271–279. Strohmaier, K.G. and Vaughan, D.E.W. (2003) J. Am. Chem. Soc., 125, 16035–16039. Lobo, R.F., Zones, S.I., and Davis, M.E. (1995) J. Inclusion Phenom. Mol. Recognit. Chem., 21, 47–78. Breck, D.W. (1974) Zeolite Molecular Sieves, John Wiley & Sons, Ltd, New York, p. 370. Millini, R., Perego, G., and Bellussi, G. (1999) Top. Catal, 9, 13–34. Whittam, T.V. and Youll, B. (1977) US Patent 4,060,590. Bellussi, G., Millini, R., Catati, A., Maddinelli, G., and Gervasini, A. (1990) Zeolites, 10, 642. Gies, H. and Rius, J. (1995) Z. Kristallogr., 210, 475–480.

References 79. Zones, S.I., Santilli, D.S., Ziemer,

80. 81. 82.

83. 84.

85. 86.

87.

88.

89.

90. 91.

92. 93.

94.

J.N., Holtermann, D.L., Pecoraro, T.A., and Innes., R.A. (1990) US Patent 4,910,006. Zones, S.I. (1990) US Patent 4,963,337. Lobo, R.F. and Davis, M.E. (1995) J. Am. Chem. Soc., 117, 3766–3779. Lobo, R.F., Pan, M., Chan, I., Li, H., Medrud, R.C., Zones, S.I., Crozier, P.A., and Davis, M.E. (1993) Science, 262, 1543–1546. Burton, A.W. (2009) US Patent Application 2009/0060813. Vortmann, S., Marler, B., Gies, H., and Daniels, P. (1995) Microporous Mesoporous Mater., 4, 112–121. Dorset, D.L. and Kennedy, G.J. (2005) J. Phys. Chem. B., 109, 13891–13898. Wagner, P., Terasaki, O., Ritsch, S., Nery, J.G., Zones, S.I., Davis, M.E., and Hiraga, K. (1999) J. Phys. Chem. B, 103, 8245–8250. Burton, A., Elomari, S., Medrud, R.C., Chan, I.Y., Chen, C.-Y., Bull, L.M., and Vittoratos, E.S. (2003) J. Am. Chem. Soc., 125, 1633–1642. Burton, A., Elomari, S., Chen, C.-Y., Medrud, R.C., Chan, I.Y., Bull, L.M., Kibby, C., Harris, T.V., Zones, S.I., and Vittoratos, E.S. (2003) Chem. Eur. J., 9, 5737–3748. Elomari, S., Burton, A., Medrud, R.C., and Grosse-Kunstleve, R. (2009) Microporous Mesoporous Mater., 118, 325–333. Li, Y. and Zou, X. (2005) Angew. Chem. Int. Ed. Engl., 44, 2012–2015. Elomari, S., Burton, A.W., Ong, K., Pradhan, A.R., and Chan, I.Y. (2007) Chem. Mater., 19, 5485–5492. Burton, A. and Elomari, S. (2004) Chem. Commun., 2618–2619. Freyhardt, C.C., Lobo, R.F., Khodabandeh, S., Lewis, J.E. Jr., Tsapatsis, M., Yoshikawa, M., Camblor, M.A., Pan, P., Helmkamp, M.W., Zones, S.I., and Davis, M.E. (1996) J. Am. Chem. Soc., 118, 7299–7310. McCusker, L.B., Grosse-Kunstleve, R.W., Baerlocher, Ch., Yoshikawa, M., and Davis, M.E. (1996) Microporous Mesoporous Mater., 6, 295–309.

95. R¨ ohrig, C. and Gies, H. (1995) Angew.

Chem. Int. Ed. Engl., 34, 63–65. ˚ H., Gier, 96. Cheetham, A.K., Fjellvag,

97.

98.

99.

100.

101. 102.

103.

104. 105. 106. 107.

108.

109. 110. 111.

112. 113.

T.E., Kongshaug, K.O., Lillerud, K.P., and Stucky, G.D. (2001) Stud. Surf. Sci. Catal., 135, 158. Park, S.H., Daniels, P., and Gies, H. (2000) Microporous Mesoporous Mater., 37, 129–143. Park, S.-H., Parise, J.B., Gies, H., Liu, H., Grey, C.P., and Toby, B.H. (2000) J. Am. Chem. Soc., 122, 11023–11024. Kubota, Y., Helmkamp, M.M., Zones, S.I., and Davis, M.E. (1996) Microporous Mesoporous Mater., 6, 213–229. Calabro, D.C., Cheng, J.C., Crane, R.A. Jr., Kresge, C.T., Dhingra, S.S., Steckel, M.A., Stern, D.L., and Weston, S.C. (2000) US Patent 6,049,018. Zones, S.I., Nakagawa, Y., and Rosenthal, J.W. (1994) Zeoraito, 11, 81. Goretsky, A.V., Beck, L.W., Zones, S.I., and Davis, M.E. (1999) Microporous Mesoporous Mater., 28, 387–393. Nakagawa, Y., Lee, G.S., Haris, T.V., Yuen, L.T., and Zones, S.I. (1998) Microporous Mesoporous Mater., 22, 69–85. Zones, S.I. (1989) Zeolites, 9, 483. Zones, S.I. and Burton, A.W. Jr. (2006) US Patent 7,108,843. Parnham, E.R. and Morris, R.E. (2006) J. Am. Chem. Soc., 128, 2204–2205. Zones, S.I., Nakagawa, Y., Yuen, L.T., and Haris, T.V. (1996) J. Am. Chem. Soc., 118, 7558–7567. Dorset, D.L., Weston, S.C., and Dhingra, S.S. (2006) J. Phys. Chem. B, 2045–2050. Lee, G.S. and Zones, S.I. (2002) J. Solid State Chem., 167, 69–85. Elomari, S.A. and Zones, S.I. (2001) Stud. Surf. Sci. Catal., 135, 479. Elomari, S.A., Burton, A., Medrud, R.C., and Grosse-Kunstleve, R. (2009) Microporous Mesoporous Mater., 118, 325–333. Elomari, S. (2003) US Patent 6,616,911. Rollman, L.D., Schlenker, J.L., Lawton, J.L., Kennedy, C.L., Kennedy, G.J., and Doren, D.L. (1999) J. Phys. Chem. B, 103, 7175–7183.

83

84

2 Synthesis Approaches 114. De Witte, B., Patarin, J., Guth, J.L., and

115. 116.

117. 118. 119.

120.

121.

122.

123. 124. 125.

126. 127.

128.

129.

130.

131.

Cholley, T. (1997) Microporous Mater., 10, 247–257. Keijsper, J.J. and Mackay, M. (1994) US Patent 5,275,799. Delprato, F., Delmottte, L., Guth, J.L., and Huve, L. (1990) Zeolites, 10, 546–552. Vaughan, D.E.W. and Strohmaier, K.G. (1990) US Patent 4,931,267. Vaughan, D.E.W. (1989) US Patent 4,879,103. Freyhardt, C.C., Tsapatsis, M., Lobo, R.F., Balkus, K.J., and Davis, M.E. (1996) Nature, 381, 295–298. Dorset, D.L., Strohmaier, K.G., Kliewer, C.E., Corma, A., D´ıaz-Caba˜ nas, M.J., Rey, F., and Gilmore, G.J. (2008) Chem. Mater., 20, 5325–5331. Lewis, D.W., Freeman, C.F., and Catlow, C.R.A. (1995) J. Phys. Chem. B, 99, 11194. Sastre, G., Leiva, S., Sabater, M.J., Gimenez, I., Rey, F., Valencia, S., and Corma, A. (2003) J. Phys. Chem. B, 107, 5432–5440. Schmitt, K.D. and Kennedy, G.J. (1994) Zeolites, 14, 635. Han, S. and Smith, J.V. (1999) Acta Crystallogr., A55, 332–382. Treacy, M.M.J., Rivin, I., Balkovski, E., Randall, K.H., and Foster, M.D. (2004) Microporous Mesoporous Mater., 74, 121–132. Earl, D.J. and Deem, M.W. (2006) Ind. Eng. Chem. Res., 45, 5449–5454. Deem, M.W. (2008) ICMR Workshop on Design and Synthesis of New Materials, August 1-2, Santa Barbara. Friedrichs, O.D., Dress, A.W.M., Huson, D.H., Klinowski, J., and Mackay, A.L. (1999) Nature, 400, 644. Majda, D., Almeida Paz, F.A., Friedrichs, O.D., Foster, M.D., Simperler, A., Bell, R.G., and Klinowski, J. (2008) J. Phys. Chem. C, 112, 1040–1047. Li, Y., Yu, J., Liu, D., Yan, W., Xu, R., and Xu, Y. (2003) Chem. Mater., 15, 2780–2785. Dorset, D.L., Roth, W.J., Kennedy, G.J., and Dhingra, S.S. (2008) Z. Kristallogr., 223, 456–460.

132. Lewis, D.W., Willock, D.J., Catlow,

133.

134.

135. 136. 137.

138.

139.

140. 141.

142. 143. 144.

145. 146. 147.

148.

C.R.A., Thomas, J.M., and Hutchings, G.J. (1996) Nature, 382, 604. Lok, B.M., Messina, C.A., Patton, R.L., Gajec, R.T., Cannan, T.R., and Flanigen, E.M. (1984) J. Am. Chem. Soc., 106, 6092–6093. Castro, M., Garcia, R., Warrender, S.J., Slawin, A.M.Z., Wright, P.A., Cox, P.A., Fecant, A., Mellot-Draznieks, C., and Bats, N. (2007) Chem. Commun., 3470–3472. Zones, S.I. and Nakagawa, Y. (1998) US Patent 5,785,947. Zones, S.I. and Hwang, S.-J. (2002) Chem. Mater., 14, 313–320. Lewis, G.J., Miller, M.A., Moscoso, J.G., Wilson, B.A., Knight, L.M., and Wilson, S.T. (2004) Stud. Surf. Sci. Catal., 154, 364–372. Blackwell, C.S., Broach, R.W., Gatter, M.G., Holmgren, J.S., Jan, D.-Y., Lewis, G.J., Mezza, B.J., Mezza, T.M., Miller, M.A., Moscoso, J.G., Patton, R.L., Rohde, L.M., Schoonover, M.W., Sinkler, W., Wilson, B.A., and Wilson, S.T. (2003) Angew. Chem. Int. Ed. Engl., 23, 1737–1740. Vaughan, D.E.W., Lussier, J., and Magee, J.S. (1979) US Patent 4,176,090. Pinnavaia, T.J. (1983) Science, 220, 365–371. Lawton, S.L., Fung, A.S., Kennedy, G.J., Alemany, L.B., Chang, C.D., Hatzikos, G.H., Lissy, D.N., Rubin, M.K., and Timken, K.C. (1996) J. Phys. Chem., 100, 3788. Roth, W.J. (2007) Stud. Surf. Sci. Catal., 168, 221. Roth, W.J. (2005) Stud. Surf. Sci. Catal., 158, 19–26. Schreyeck, L., Caullet, P., Mougenel, J.C., Guth, J.L., and Marler, B. (1996) Microporous Mesoporous Mater., 6, 249. Dhingra, S., Kresge, C.T., and Casmer, S.G. (2005) US Patent 6,869,587. Dorset, D.L. and Kennedy, G.J. (2004) J. Phys. Chem. B, 108, 15216–15222. Burton, A., Accardi, R.J., Lobo, R.F., Falconi, M., and Deem, M.W. (2000) Chem. Mater., 12, 2936. Ikeda, T., Akiyama, Y., Oumi, Y., Kawai, A., and Mizukami, F. (2004)

References

149.

150.

151.

152.

153.

154. 155. 156. 157.

158. 159.

160.

161. 162.

163.

164.

165. 166.

Angew. Chem. Int. Ed. Engl., 43, 4892–4896. Lewis, G.J., Knight, L.M., Miller, M.A., and Wilson, S.T. (2005) US Patent Application 2005/0065016. Zanardi, S., Alberti, A., Cruciani, G., Corma, A., Forn´es, V., and Brunelli, M. (2004) Angew. Chem. Int. Ed. Engl., 43, 4933–4937. Marler, B., Camblor, M.A., and Gies, H. (2006) Microporous Mesoporous Mater., 90, 87–101. Moteki, T., Chaikittisilp, W., Shimojima, A., and Okubo, T. (2008) J. Am. Chem. Soc., 130, 15780–15781. Marler, B., Str¨oter, N., and Gies, H. (2005) Microporous Mesoporous Mater., 83, 201–211. Wang, Y.X., Gies, H., and Lin, J.H. (2007) Chem. Mater., 19, 4181–4188. Puppe, L. and Weisser, J. (1984) US Patent 4,439,409. Rubin, M.K. and Chu, P. (1990) US Patent 4,954,325. Leonowicz, M.E., Lawton, J.A., Lawton, S.L., and Rubin, M.K. (1994) Science, 264, 1910–1913. Degnan, T.F. Jr. (2003) J. Catal., 216, 32–46. Kresge, C.T., Roth, W.J., Simmons, K.G., and Vartuli, J.C. (1993) US Patent 5,229,341. Roth, W.J., Kresge, C.T., Vartuli, J.C., Leonowicz, M.E., Fung, S.B., and McCullen, S.B. (1995) Catal. Microporous Mater. Stud. Surf. Sci. Catal., 94, 301. Chica, A., Corma, A., Forn´es, V., and D´ıaz, U. (2003) US Patent 6,555,090. Wu, P., Ruan, J., Wang, L., Wu, L., Wang, Y., Liu, Y., Fan, W., He, M., Terasaki, O., and Tatsumi, T. (2008) J. Am. Chem. Soc., 130, 8178–8187. Corma, A., Forn´es, V., Pergher, S.B., Maesen, Th.L.M., and Buglass, J.G. (1998) Nature, 396, 353. Corma, A., Diaz, U., Domine, M.E., and Forn´es, V. (2000) Angew. Chem. Int. Ed. Engl., 38, 1499–1501. Kresge, C.T. and Roth, W.J. (1993) US Patent 5,266,541. Corma, A., Fornes, V., and Diaz, U. (2001) Chem. Commun., 2642–2643.

167. Corma, A., D´ıaz, U., and Fornes, V.

(2006) US Patent 7,008,611. 168. Fung, A.S., Lawton, S.L., and Roth,

W.J. (1994) US Patent 5,363,697. 169. Zones, S.I., Hwang, S.J., and Davis,

M.E. (2001) Chem. Eur. J., 7, 1990. 170. Millini, R., Perego, G., Parker, W.O.

171. 172.

173. 174.

175. 176. 177. 178.

179. 180.

181.

182. 183.

184. 185.

186.

Jr., Bellussi, G., and Carluccio, L. (1995) Microporous Mesoporous Mater., 4, 221. Corma, A., D´ıaz, U., and Fornes, V. (2006) US Patent 7,008,651. Kiyozumi, Y., Ikeda, T., Hasegawa, Y., Nagase, T., and Mizukami, F. (2006) Chem. Lett., 35, 672. Wheatley, P.S. and Morris, R.E. (2006) J. Mater. Chem., 16, 1035. Vistad, Ø.B., Akporiaye, D.E., and Lillerud, K.P. (2001) J. Phys. Chem. B, 105, 12437–12447. Fung, A.S., Lawton, S.L., and Roth, W.J. (1994) US Patent 5,362,697. Chu, P., Herbst, J.A., Klocke, D.J., and Vartuli, J. (1991) US Patent 4,985,223. Rubin, M.K. (1991) US Patent 5,063,037. Bennett, J.M., Chang, C.D., Lawton, S.L., Leonowicz, M.E., Lissy, D.N., and Rubin, M.K. (1993) US Patent 5,236,575. Bibby, D.M. and Dale, M.P. (1985) Nature, 317, 157–158. Huo, Q., Feng, S., and Xu, R. (1988) J. Chem. Soc., Chem. Commun., 1486–1487. Kuperman, A., Nadimi, S., Oliver, S., Ozin, J.A., Garc´es, J.M., and Olken, M.M. (1993) Nature, 365, 239–242. Huo, Q. and Xu, R. (1990) J. Chem. Soc., Chem. Commun., 783. Huo, Q., Xu, R., Li, S., Ma, Z., Thomas, J.M., Jones, R.H., and Chippindale, A.M. (1992) J. Chem. Soc., Chem. Commun., 875. Morris, R.E. and Weigel, S. (1997) J. Chem. Soc. Rev., 26, 309–317. Derouane, E.G., Valyocsik, E.W., and Von Ballmoos, R. (1990) European Patent 146384. Derouane, E.G., Maistriau, L., Gabelica, Z., Tuel, A., Nagy, J.B., and Von Ballmoos, R. (1989) Appl. Catal., 51, L13–L20.

85

86

2 Synthesis Approaches 187. Strohmaier, K.G. and Vaughan, D.E.W. 188.

189.

190. 191.

192.

(2001) US Patent 6,303,534. Strohmaier, K.G., Afeworki, M., Chen, T.J., and Vaughan, D.E.W. (2007) 15th International Zeolite Conference, Bejing, Recent Research Reports. R-09–03. Vaughan, D.E.W. and Strohmaier, K.G. (1992) in Proceedings of 9th International Zeolite Conference (eds R. von Ballmoos, J.B. Higgins, M.M.J. Treacy), Butterworth-Heinemann, Boston, pp. 197–206. Vaughan, D.E.W. and Strohmaier, K.G. (1994) US Patent 5,318,766. Garces, J.M., Millar, D.M., and Howard, K.E. (1996) US Patent 5,589,153. Cooper, E.R., Andrews, C.D., Wheatley, P.S., Webb, P.B., Wormald, P., and

193. 194.

195.

196.

197.

198.

Morris, R.E. (2004) Nature, 430, 1012–1016. Parnham, E.R. and Morris, R.E. (2007) Acc. Chem. Res., 40, 1005–1013. Akporiaye, D.E., Dahl, I.M., Karlsson, A., and Wendelbo, R. (1998) Angew. Chem. Int. Ed. Engl., 37, 609–611. Choi, K., Gardner, D., Hilbrandt, N., and Bein, T. (1999) Angew. Chem. Int. Ed. Engl., 38, 2891–2894. Klein, J., Lehmann, C.W., Schmidt, H.-W., and Maier, W.F. (1998) Angew. Chem. Int. Ed. Engl., 37, 3369–3372. Song, Y., Yu, J., Li, G., Li, Y., Wang, Y., and Xu, R. (2002) Chem. Commun., 1720. Moliner, M., Serra, J.M., Corma, A., Argente, E., Valero, S., and Botti, V. (2005) Microporous Mesoporous Mater., 78, 73–81.

87

3 Ionothermal Synthesis of Zeolites and Other Porous Materials Russell E. Morris

3.1 Introduction

Innovation in zeolite synthesis remains an important aspect in the search for new framework materials with potential applications. The driver for the search for new zeolites and related solids is not only the need to provide materials for new and emerging applications [1] but is also the desire to understand how these fascinating materials are made, and ultimately how to control their architectures. Given that the applications of zeolites (and other porous solids such as metal organic frameworks) are intimately connected with their architecture, new synthetic methods that aim to understand how their structure can be controlled are very important. The core strategy that has been exercised over recent years has been the development of new organic compounds that can be used as structure directing agents (SDAs or templates). Simply preparing new SDAs has led to a signiﬁcant increase in the numbers of zeolite structures over recent years. This is still a method that produces some remarkable new materials, such as IM-12 [2]. However, other more innovative methods have also made their impact, both in terms of ﬁnding new ways to recycle templates [3] and in completely new synthesis concepts such as the use of ﬂuoride mineralizers and charge density mismatch solutions [4–7]. The use of ﬂuoride as a mineralizing agent to improve the solubility of the starting reagents and to catalyze the formation of bonds in the target frameworks has been exploited by several groups to produce several new materials over recent years [4–6]. Charge density mismatch solutions, developed by workers at UOP, have also provided routes to new solids [7]. In this process, stable solutions of the inorganic starting materials are prepared by using organic cations that do not make good SDAs because their charge density does not match that of the chemical composition of the inorganic framework which will be formed. Crystallization of the framework is then initiated by addition of another SDA, often in quite small amounts. High-throughput methods have also been applied with some distinct success, particularly by Corma’s group in Valencia [8]. In our laboratory, we have pioneered the use of ionic liquids (ILs) as both the solvent and SDA simultaneously [9]. The change from a molecular solvent, such as water or organic molecules, to Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

88

3 Ionothermal Synthesis of Zeolites and Other Porous Materials

an ionic one changes the chemistry of the system markedly. We have given the name ionothermal synthesis to this method to delineate it from hydrothermal or solvothermal synthesis. Over the last few years, ILs have received great attention in many ﬁelds [10]. Most of the studies have focused on the ‘‘green’’ chemistry [11] potential of these compounds, with particular emphasis on the drive to replace organic solvents in homogeneous catalysis [12]. The particular property of ILs that makes them environmentally suitable for these purposes is their low vapor pressure [13], which has signiﬁcant advantages when replacing highly volatile organic solvents. However, there are many other uses of ILs in diverse areas of technology ranging from electrolytes in batteries and fuel cells [14], as electrodeposition solvents [15] to the use of supported IL as catalysts [16]. In some reactions, the ILs act only as inert solvents and in others the liquid plays a more active role in the reactions that take place. The broadest deﬁnition of an IL is any material in the liquid state that consists predominantly of ionic species [10]. Any ionic salt that can be made molten can therefore be classiﬁed as an ‘‘IL,’’ always assuming that the ionic components of the solid remain intact on melting. There are many examples in the literature of molten salts being used as the medium in which inorganic materials have been prepared [17]. Usually, these synthetic procedures take place at highly elevated temperatures, producing dense phase solids. For example, alkali metal hydroxide molten salts can be used as the molten phase, often contained in sealed inert (such as silver) vessels in the synthesis of many inorganic solids. In general, such molten salt synthesis methods have been used as direct replacements for traditional solid-state synthesis techniques [17]. However, the modern deﬁnition of ILs tends to concentrate on those compounds that are liquid at relatively low temperatures and that contain organic components [18]. Room temperature ionic liquids (RTILs) are, as the name suggests, liquid at room temperature, while near room temperature ionic liquids (nRTILs) are often deﬁned as being liquid below a certain temperature, often 100 ◦ C, although this varies depending on the application envisaged for the liquids. For ionothermal synthesis, nRTILs are often deﬁned as being liquid below about 200 o C, the temperatures traditionally used in hydrothermal synthesis [9]. In modern usage, the term ionic liquid is almost exclusively reserved for liquids that contain at least one organic ion. The organic components of ILs tend to be large and often quite asymmetric, which contribute to their low melting points by making efﬁcient packing in the solid state more difﬁcult [10]. ILs show a range of properties that make them suitable for use as media for the preparation of inorganic and inorganic–organic hybrid materials. They can be relatively polar solvents, ensuring reasonably good solubility of inorganic precursors [19, 20]. Many (but not all) ILs have good thermal stability, enabling them to be used at elevated temperatures. Deep eutectic solvents (DESs) are a related class of IL, produced as a mixture of two or more compounds that has a lower melting point than either of its constituents [21]. Eutectic mixtures display unusual solvent properties that are very similar to those shown by the ILs. High solubility can be observed (depending on

3.2 Hydrothermal, Solvothermal, and Ionothermal Synthesis

the eutectic mixture used) for inorganic salts, salts that are sparingly soluble in water, aromatic acids, amino acids, and several metal oxides [22]. Advantages of eutectic mixtures over other ILs are their ease of preparation in pure state and their relative nonreactivity with water. Many are biodegradable and the toxicology of the components maybe well characterized. Eutectic mixtures based on relatively available components such as urea and choline chloride are also far cheaper than some other ILs. Fundamentally, there is of course no real difference between an IL and a molten salt, except perhaps that the organic nature of the components of an IL introduces much more scope for introducing functionality into the solvents. In the following feature chapter, the focus is on the use of the nRTILs containing organic components as the media for materials synthesis, consistent with modern usage of the terminology. In particular, the focus is on the synthesis of templated crystalline materials such as zeolites and metal organic frameworks where the IL cation acts to direct the structure of the resultant inorganic or inorganic–organic hybrid material.

3.2 Hydrothermal, Solvothermal, and Ionothermal Synthesis

Broadly speaking, the synthesis of crystalline solid-state materials can be split into two main groups; those where the synthesis reaction takes place in the solid state and those where it takes place in solution. The solid-state method usually requires rather high temperatures to overcome difﬁculties in transporting the reactants to the sites of the reaction. The high temperatures of solid-state reactions also tend to provide routes to the thermodynamically more favored phases in the systems of interest. Typically, this method is used to prepare solid-state oxides. Transport in the liquid phase is obviously much easier than in solids, and syntheses require much lower temperatures (often less than 200 o C). The most commonly used of this type of preparative technique is hydrothermal synthesis, where the reaction solvent is water [23]. The most common method of accomplishing hydrothermal synthesis is to seal the reactants inside Teﬂon-lined autoclaves so that there is also signiﬁcant autogenous hydrothermal pressure produced, often up to 15 bar. The lower temperatures required for hydrothermal synthesis often lead to kinetic control of the products formed, and it is much easier to prepare metastable phases using this approach than it is using traditional solid-state approaches. The important reaction and crystallization processes in hydrothermal synthesis do not necessarily take place in solution (although, of course, they can) but can occur at the surfaces of gels present in the mixtures. Solvothermal synthetic methods refer to the general class of using a solvent in the synthesis of materials [24]. Of course water is by far the most important solvent, hence the special usage of the term hydrothermal to describe its use. However, there are many other possible solvents. Alcohols, hydrocarbons, pyridine, and many other organic solvents have all been used with varying degrees of success [24].

89

90

3 Ionothermal Synthesis of Zeolites and Other Porous Materials

As with water, these molecular solvents produce signiﬁcant autogenous pressure at elevated temperatures. The solvents used in solvothermal synthesis vary widely in their properties, from nonpolar and hydrophobic to polar and hydrophilic. The solvents used in hydrothermal and solvothermal synthesis differ fundamentally from ILs in that they are molecular in nature. The ionic nature of ILs imparts particular properties, including low vapor pressures [25] (and very little, if any, autogenous pressure is produced at a high temperature).

3.3 Ionothermal Aluminophosphate Synthesis

Many ILs used currently often have chemical structures that are very similar to the structures of commonly used SDAs (sometimes also known as templates) in the hydrothermal synthesis of zeolites and other porous materials [26]. This realization led to the ﬁrst attempts to prepare zeotype frameworks using ILs as both the solvent and the template provider at the same time. The potential advantage of this approach is that the competition between the solvent and template for interaction with any growing solid is removed when both the solvent and the template are the same species. In principle, this may lead to improved templating of the growing zeolite crystal structure. The ﬁrst work in this area, published in 2004, used 1-ethyl-3-methyl imidazolium bromide (EMIM Br) and urea/choline chloride DESs to prepare several different materials depending on the conditions [9]. Since the ﬁrst breakthroughs in this area, there have been many further attempts to prepare zeotype materials. The ionothermal synthesis of aluminophosphate zeolites has been by far the most successful. Many common ILs are suitable solvents for the preparation of these materials, with both known [27–30] and previously unknown [31] structure types, as well as related low-dimensional materials [32, 33] being synthesized successfully. It is interesting to note that more than simply preparing the base aluminophosphate structure, the ionothermal method is also suitable for incorporating the dopant metal atoms that give the frameworks their chemical activity. Silicon (to make so-called SAPOs) [34] and many different tetrahedral metals (Co, Mg, etc.) can all be incorporated into the ionothermally prepared aluminophosphate zeolites, and aspects such as their catalytic activity [35] and the use of additional templates [36] show some very promising results. A discussion of some of the unusual concepts seen in AlPO synthesis is discussed in the remaining sections of this chapter. Figure 3.1 illustrates several of the SIZ-n(ST.Andrews Ionothermal Zeolite) materials that can be prepared from one particular IL–EMIM Br. Several of these materials have known frameworks but several others were previously unknown. The structure of SIZ-1 consists of hexagonal prismatic units known as double six rings joined to form layers that are linked into a three-dimensional framework by units containing four tetrahedral centers (two phosphorus and two aluminum) known as single four rings. The formula of the material is Al8 (PO4 )10 H3 ·3C6 H11 N2 but the Al–O–P alternation is maintained. The framework is therefore interrupted

3.3 Ionothermal Aluminophosphate Synthesis

91

SIZ-4

SIZ-3 Br −

SIZ-5

N +N

SIZ-1

SIZ-9

SIZ-6 SIZ-7

SIZ-8

Figure 3.1 Representatives of the SIZ-n series of aluminophosphate zeotype structures prepared using 1-ethyl, 3-methyl imidazolium bromide ionic liquids as both the solvent and structure directing agent.

with some unusual intraframework hydrogen bonding. The negative charge present on the framework (caused by the existence of terminal P–O bonds) balances the charge on the 1-methyl-3-ethyl imidazolium templates that are present in the pores. The overall structure of SIZ-1 shows a two-dimensional channel system parallel to the a and b crystallographic axes. SIZ-3, SIZ-4, SIZ-5, SIZ-8, and SIZ-9 all have known framework structures (AEL, chabasite (CHA), AFO, AEI, and sodalite (SOD) frameworks respectively). The structure of SIZ-7 is also a novel cobalt aluminophosphate material, given the International Zeolite Association (IZA) code SIV. However, SIZ-7 is a novel framework structure, which joins a family of related zeolites that includes the PHI, GIS, and merlinoite (MER) structure types. This family can be described as consisting of the double-crankshaft chain. In SIZ-7, these chains run parallel to the crystallographic a axis in the structure and are connected to form a one-dimensional small-pore zeolite structure with windows into the pores delineated by rings containing eight tetrahedral atoms (known as eight-ring windows). The repeat unit in the a direction is 10.2959 (4) A˚ and equals one repeat unit of the double-crankshaft chain. These chains are linked via four rings in both the b and c directions to form the eight-ring windows. The relative orientation of neighboring chains means that there are two types of eight-ring channels. The two different windows are of similar size (3.66 × 3.26 A˚ ˚ but are different in shape. In the b direction, the same type of and 3.40 × 3.52 A) eight-ring channel is repeated, leading to a repeat unit in this direction of 14.3715

92

3 Ionothermal Synthesis of Zeolites and Other Porous Materials

˚ while in the c direction the two types of channel alternate, leading to an (5) A, ˚ approximate doubling of the unit cell dimension in this direction to 28.599 (1) A. The overall structure of SIZ-6 is also shown in Figure 3.1. This is a very unusual material comprising 13.5-A˚ thick anionic aluminophosphate layers of chemical composition Al4 (OH)(PO4 )3 (HPO4 )(H2 PO4 )− . The layers themselves consist of rings containing four, six, and eight nodes (aluminum or phosphorus atoms). The eight-ring windows are large enough to make the layers potentially porous to small molecules. The layers are held together via some relatively strong hydrogen bonding. This occurs because two H2 PO4 groups, one each from two adjacent layers, forming dimeric units with O–O distances across the hydrogen bond of ˚ In addition, the negative charge on the layers is compensated for by one 2.441 A. 1-ethyl-3-methylimidazolium (EMIM) cation, which occupies the interlayer space.

3.4 Ionothermal Synthesis of Silica-Based Zeolites

The ionothermal synthesis of AlPOs is relatively straightforward. Silicon-based zeolites have, however, been much more of a challenge for ionothermal synthesis, although there has been more success in the synthesis of mesostructured silica using ILs [37]. The problem with zeolite synthesis is primarily the solubility of silica starting materials in the commonly used ILs, which is not sufﬁciently good to allow silicate and aluminosilicate materials to be prepared. Before 2009, there was only one report of a silica polymorph being prepared from an IL [38] and one report of the synthesis of a sodalite [39]. Successful synthesis of zeolites requires the preparation of ILs more suited to silicate dissolution. Recently, in our laboratory, we were successful in preparing ILs comprising mixed halide and hydroxide anions that are suitable solvents for the preparation of purely siliceous and aluminosilicate zeolites. The presence of hydroxide increases the solubility of the silicate starting materials and allows the zeolites to crystallize on a suitable timescale (Figure 3.2, Wheatley and Morris, manuscript in preparation). However, despite this proof of concept work, there is still much to be done to more fully understand the chemistry of silica in ILs, and it is likely that task-speciﬁc ILs will need to be developed before silica zeolites can be prepared routinely using ionothermal synthesis.

3.5 Ionothermal Synthesis of Metal Organic Frameworks and Coordination Polymers

Similar to the synthesis of zeolites, ILs can be used as solvents and templates to prepare many other types of solids. One of the most interesting and important class of materials that has been recently developed is that of metal organic frameworks (also known as coordination polymers) [40, 41]. These materials offer great promise for many different applications, particularly in gas storage [42–45]. Normally these materials are prepared using solvothermal reactions, with organic solvents such as

3.6 Ambient Pressure Ionothermal Synthesis

EMIM Br/OH mixed IL

a c TON

MFI

Figure 3.2 The ionothermal synthesis of pure silica zeolites (TON and MFI) using 1-butyl, 3-methyl imidazolium–based ionic liquids with mixed bromide-hydroxide counteranions. The BMIM cation can be clearly seen from the single-crystal X-ray diffraction structure of MFI.

alcohols and dimethyl formamide. Ionothermal synthesis has been used extensively over the last few years to prepare these types of solid, and there are now many examples in the literature [46–55]. Unlike zeolites, however, the lower thermal stability of coordination polymers leads to several issues regarding removal of ionic templates from the materials to leave porous materials. Often removing the IL cation is not possible without collapsing the structure. However, it is possible to prepare porous materials using DESs, and Bu has recently proven this very elegantly [56]. A great many of the materials prepared ionothermally are relatively lowdimensional solids, and this is clearly a very productive method for the preparation of such materials. It is very clear that in these systems changing the chemistry of the solvent to ionothermal leads to great possibilities in this area.

3.6 Ambient Pressure Ionothermal Synthesis

Perhaps the most striking feature of ILs is their very low vapor pressure. This means that, unlike molecular solvents such as water, the ILs can be heated to relatively high temperatures without the production of autogenous pressure. High-temperature reactions therefore do not have to be completed inside pressure vessels such as Teﬂon-lined steel autoclaves but can be undertaken in simple containers such as round-bottomed ﬂasks. The absence of autogenous pressure at high temperature also makes microwave heating a safer prospect as hot spots in the liquid should

93

3 Ionothermal Synthesis of Zeolites and Other Porous Materials

not cause excessive increases in pressure with their associated risk of explosion, assuming of course that the IL is stable and does not breakdown into smaller components during heating [57, 58]. Figure 3.3 shows the measured pressure during the synthesis of an aluminophosphate molecular sieve (SIZ-4) using a microwave heating experiment [59]. Figure 3.3a is the pure IL solvent, and it is clear that no autogenous pressure is produced. Figure 3.3b, however, shows that, even when only modest amounts of water are added to the system, signiﬁcant pressures are evolved. One of the most interesting potential uses of ambient pressure synthesis of zeolite coatings is for anticorrosion applications. Yushan Yan has shown that ionothermally prepared zeolite ﬁlms make excellent anticorrosion coatings for several different types of alloys [60, 61]. Given that current coatings technology is based on the use of environmentally unfriendly chromium, there is interest in ﬁnding more acceptable alternatives. Sealed zeolites are one such option. However, hydrothermal synthesis of zeolites inside sealed vessels is impractical for 30

Pressure (bar)

25 20 15 10 5 0 0

20

40

0

20

40

60

80

100

120

60 80 Time (min)

100

120

30 25 Pressure (bar)

94

20 15 10 5 0

Figure 3.3 The evolution of pressure (in bar) in the microwave synthesis of aluminophosphate SIZ-4 from (a) a pure ionic liquid solvent with no water added and (b) the same solvent system with 0.018 ml of

water added. The maximum temperature is 200 o C and the duration of heating is 60 minutes. There is almost no pressure evolution in the pure ionic liquid.

3.7 The Role of Cation-Templating, Co-Templating, or No Templating

large, oddly shaped, and cut pieces of metal. Yan contends that ambient pressure ionothermal synthesis eliminates the need for unwieldy sealed vessels, and, given the excellent coatings that can be prepared using this approach, offers an interesting and potentially important alternative technology.

3.7 The Role of Cation-Templating, Co-Templating, or No Templating

The original concept behind ionothermal synthesis was to simplify the templating process that occurs in traditional zeolite hydrothermal synthesis by making the solvent and the template the same species. The template molecules normally involved in zeolite synthesis are usually cationic as the resultant framework has a negative charge. The commonly used templating cations are very similar in chemistry to IL cations. It is not surprising therefore that the IL cations are often occluded into the ﬁnal structures of the materials, in exactly the same way as in traditional zeolite synthesis [9]. In an exactly analogous fashion, metal organic frameworks can also be synthesized using the ILs as both the solvent and the template. Most solvothermally prepared MOFs have neutral frameworks, but when the template is a cation the framework must, for charge balance, have a negatively charged framework, in exactly the same way as zeolites. Of course, the overall goal of all templating-based synthesis is to have control over the architecture of the ﬁnal material by changing the size of the templating cation. It is well known, however, that apart from rough correlations with the size of the cation, the templating interaction is not really speciﬁc enough to yield very precise control over the reaction. Figure 3.4 shows that the same general features hold for ionothermal synthesis. In this work, changing the size of the IL cation does have some effect on the ﬁnal structure – the larger cations form more open frameworks with the extra space needed to accommodate the large template. However, this is not particularly speciﬁc in this type of MOF synthesis, indicating that templating is more likely to be by simple ‘‘space ﬁlling’’ rather than any more speciﬁc or directed template–framework interactions (Lin and Morris, Unpublished work). In hydrothermal synthesis, there is also the possibility of adding alternative cations to act as templates. Of course, the situation is exactly analogous in ionothermal synthesis, and added templates offer equally great opportunities. Recently, Xing et al. [62] have shown that methylimidazolium (MIA), when added to an EMIM Br IL leads to a cooperative templating effect, occluding both MIA and EMIM in the same solid. The intriguing feature of this solid is that it seems, at least on ﬁrst inspection, that the material is made of two distinct layers. The MIA is located close to one layer and the EMIM close to the other – perhaps indicating that each cation plays a speciﬁc role in directing the structure of each part of the material. It is, of course, impossible to say this for certain until the full mechanism of synthesis is elucidated, something that is very difﬁcult in practice. However, further circumstantial evidence for this maybe the fact that the previously prepared

95

96

N

3 Ionothermal Synthesis of Zeolites and Other Porous Materials

Br + N

Br + N

Br + N

N

Br + N

N

Br + N

Br + N

Figure 3.4 The effect of changing the size of the IL cation on the resulting metal organic framework structure. The materials prepared in this study are Ni (blue) or cobalt (purple) terephthalate MOFs.

EMIM-templated materials SIZ-1 and SIZ-4 have closely related structures to the EMIM-‘‘templated’’ layer in this material. Up to now, the cation in the IL has only acted as a template in the synthesis. However, like any other solvent, including water, there is also the possibility of bonding interactions with the frameworks. Most of the ILs that are based on di-alkylated imidazolium cations have no obvious sites through which to coordinate to the metal sites in the way that water does. However, some ILs, under speciﬁc conditions, can breakdown to leave the monoalkylated imidazole species that can coordinate to metals [63]. As in hydrothermal synthesis, controlling how the solvent interacts with the framework materials is therefore important in determining the exact nature of the ﬁnal material. A similar example where the solvent can coordinate to the metal in a metal organic framework comes when using choline chloride/urea-based DES ILs. Normally, this type of solvent is regarded as being relatively unstable, especially the urea portion which can break up and deliver smaller templates into the reaction. However, under conditions where the urea is stable, it is possible to keep this intact, and in the case of ionothermally prepared lanthanum-based MOFs the urea coordinates to the metal [64]. In addition to the templating cations, ILs also contain an anion, and these turn out to be extremely important in controlling the properties of the solvents (Section 3.8). The anions can, in certain circumstances, also be occluded in the structure as a template, most often in combination with the IL cation. Bu et al. recently showed that in a series of MOFs (called ALF-n) the IL displayed several different types of behavior, including templating by only the cation and templating

3.8 The Role of the Anion – Structure Induction

by both the cation and anion simultaneously, illustrating the multiple functions that ILs can play even in the same systems [56]. Finally, of course, the ILs can only play the role as solvent and not be occluded in the ﬁnal structure at all. For species like aluminophosphate zeolites and MOFs where the chemistry of the cation is similar to that of commonly used templates one would expect them to be occluded in the ﬁnal structure. However, there are certain situations where this does not happen. Perhaps the most striking of these is when a very hydrophobic IL is used. In the case of aluminophosphate and MOF synthesis, the more hydrophobic the IL used the less likely the IL cation is to be occluded [32]. Of course, as the chemistry of the system is changed (e.g., by trying to make different types of inorganic material), the balance between the solvent and templating actions of ILs also changes.

3.8 The Role of the Anion – Structure Induction

As we have seen above, the common organic cations in ILs are chemically very similar to zeolite templates. However, ILs also contain an anion, and the nature of the anion plays an extremely important part in controlling the nature of the IL. Figure 3.5 demonstrates this dependence of property on anion very clearly. Two low melting ILs that are solid at room temperature can be prepared from the same cation (EMIM) but with two different anions – bromide and triﬂimide (NTf2 ). The two ILs have very different properties, especially when it comes to their interaction with water. Figure 3.6 shows what happens when the two compounds are left out in the air for 20 minutes. EMIM NTf2 is a relatively hydrophobic material and there is no change in its properties on exposure to the moisture in the air. EMIM Br, on the other hand, is highly hygroscopic and turns liquid on reaction with moisture in the air. Clearly, this change in IL chemistry on alteration of the IL anion is bound to have a signiﬁcant effect on the products of any reaction carried out in such solvents. One example of this is given in Section 3.7, where in the synthesis of aluminophosphates EMIM Br solvents lead to incorporation of the EMIM cation to form zeotype materials, whereas the use of the EMIM NTf2 IL leads to no occlusion of the IL cation [32]. More interesting, however, and potentially extremely useful, is the possibility of mixing the two types of liquid to form solvents with different chemistries from the end member liquids. Figure 3.6 illustrates this for the synthesis of cobalt bezenetricarboxylate MOFs [65]. The two end member ILs, EMIM Br and EMIM NTf2 , form two different types of material, while a 50 : 50 mixture of the two ILs, which are miscible, forms a third structure type. This type of result opens up the possibility of mixing ILs to form solvents whose chemistry is different from the end members, giving rise to much more control over the properties of the solvent. In a similar example, a mixed anion IL (50% bromide 50% triﬂimide) leads to the formation of coordination polymers containing ﬂuorinated

97

98

3 Ionothermal Synthesis of Zeolites and Other Porous Materials

A

B

A

B

Figure 3.5 The effect of moist air on hydrophobic EMIM triﬂimide (sample A) and hydrophilic EMIM Br (sample B). After 20 minutes exposure to normal air at room temperature, the EMIM Br has absorbed enough moisture from the atmosphere to turn from a solid into a liquid.

ligands when ILs containing only one anion (either bromide or triﬂimide) does not produce any crystalline solid [66]. It is clearly the nature of the anion that determines the ﬁnal material in these examples. However, the anions themselves are not generally occluded into the structure, and so this is an induction effect rather than a templating of the structure directing effect. It is perhaps not too surprising that changing the chemistry of the solvent will change the type of product in such a manner. In the example illustrated in Figure 3.6, there is no obvious correspondence between the nature of the anions and the nature of the ﬁnal material. However, in 2007, we published an example of an anion induction using a chiral anion as part of an IL to induce a chiral coordination polymer that contains only achiral building blocks [67] (Figure 3.7). In this example, a chiral IL prepared from the butyl methyl imidazolium (BMIM) cation in combination with l-aspartate as the anion, when used to prepare a cobalt bezenetricarboxylate MOF produced a chiral structure, with all indications that the bulk solid produced was homochiral. Where some speciﬁc property of the IL anion manifests itself in the resulting material, despite the fact that it is not actually occluded, the potential for ‘‘designer’’ structure induction becomes very attractive, and one would hope that such properties of ionothermal synthesis will be explored and exploited more thoroughly in the near future.

3.9 The Role of Water and Other Mineralizers

N

Br + N

100% bromide

50% bromide 50% triflimide

100% triflimide

Figure 3.6 The effect of the anion on the ﬁnal structure of the material produced in an ionothermal synthesis. The top reaction shows uses EMIM Br as the solvent, and produces one particular cobalt-trimesic acid

MOF. A 50 : 50 mixture of EMIM Br and EMIM triﬂimide produces a different MOF, while using only the EMIM triﬂimide produces yet another material.

3.9 The Role of Water and Other Mineralizers

One of the very ﬁrst questions asked about ionothermal synthesis was whether the ILs used were sufﬁcient in their own right to promote the synthesis of zeolites and other inorganic materials, particularly those oxides where water might catalyze the condensation reactions needed to form the required bonds. One of the ﬁrst things noted about ionothermal synthesis was that too much water was detrimental to the formation of zeolites. At low concentrations of water, zeolites were the main products, but as more water was added to the IL solvents so that they were about equimolar in concentration only dense phases could be prepared. Wragg and coworkers studied this effect in more detail and conﬁrmed through several hundred high-throughput reactions that larger amounts of water did indeed lead to dense phases [63]. The origin of this effect is still under investigation but it is known that the microstructure of water in ILs changes with concentration. At low concentrations, the water is hydrogen bonded relatively strongly to the anion, and exists either as isolated water molecules or as very small clusters [68]. However, as the concentration of water increases, larger clusters and eventually

99

100

3 Ionothermal Synthesis of Zeolites and Other Porous Materials O N

+

N

HO

O O

−

NH2

Figure 3.7 The use of an ionic liquid with a chiral anion induces a chiral MOF structure. Use of an achiral anion produces achiral structures.

hydrogen-bonded networks start to appear, which change the properties of the liquid markedly. Eventually, of course, as more and more water is added, it becomes the dominant chemical component (and therefore the solvent) and the system becomes hydrothermal rather than ionothermal. The strong binding of isolated water molecules in ILs leads to another interesting effect that can be used in ionothermal synthesis – so-called water deactivation. At low concentrations of water, this strong hydrogen bonding leads to water being less reactive than similar amounts in other solvents. This effect is so strong that highly hydrolytically sensitive compounds such as PCl3 can be stored for relatively long periods, whereas they react quickly, and often violently, in other ‘‘wet’’ solvents [69]. Such water deactivation is probably the reason why some of the materials prepared using ILs can have unusual features. For instance SIZ-13, a cobalt aluminophosphate material, has a layered structure that is closely related to a zeolite, but has Co–Cl bonds. Normally such bonds are hydrolytically unstable and, under hydrothermal conditions, it is unlikely that this material would be stable [27]. In zeolite (and other) synthetic procedures, mineralizers, such as ﬂuoride or hydroxide ions, added to the reaction mixtures in the correct quantities are often vital for crystallization of the desired molecular sieve products. Fluoride in particular has recently been an extremely useful mineralizer for aluminophosphate [70] and silicate [71, 72] synthesis. In addition to helping solubilize the starting materials under the reaction conditions, there is evidence that ﬂuoride itself can play a structure directing role [73] and is intimately involved in template ordering in certain materials [74, 75]. In ionothermal synthesis, the addition of ﬂuoride also seems to be important in determining the phase selectivity of the reaction [9]. It may also help catalyze the bond-forming reactions in zeolite synthesis, as suggested by Camblor and coworkers [71]. For instance, in the synthesis of aluminophosphates,

3.11 Summary and Outlook

the addition of ﬂuoride leads to the formation of SIZ-4 and SIZ-3, which are both fully four-connected zeolite frameworks, SIZ-1, which is an interrupted structure with some unconnected P–OH bonds. Tian and coworkers have recently completed an extremely useful kinetic study of the effect of both water and ﬂuoride added to ionothermal systems in zeolite synthesis [76]. It is clear from their results that both small amounts of water, and particularly ﬂuoride, increase the crystallization rate. If the reactions are carried out carefully to exclude as much water as possible, the crystallization of the zeolites becomes very slow indeed, suggesting that for all practical purposes a small reactant amount of water (probably in the IL) is vital if ionothermal synthesis is to be successful. 3.10 Unstable Ionic Liquids

In many publications, one often sees the mention of the high thermal and chemical stability of ILs. Bearing in mind of course that it is difﬁcult to generalize across all the possible ILs, this is true under many conditions. However, under ionothermal conditions, some quite common ILs can breakdown. Even some that are often relatively stable such as BMIM bromide can breakdown, especially in the presence of ﬂuoride ions [77]. One possible reaction is the transalkylation reaction that swaps the alkyl groups, leading to the formation of dimethylimidazolium cations, which then templates a zeolite structure [77]. DES ILs based on choline chloride/urea mixtures are also unstable under ionothermal conditions. The urea portion of the IL breaks up to release ammonium ions into the mixture, which then templates the SIZ-2 aluminophosphate material. This type of instability in the ILs is actually extremely repeatable. Deep eutectic ILs made from functionized ureas all break down in the same way to produce the expected functionalized ammonium or diammonium cations that then go on to template many different structures [78]. Such reproduction ability in the reactions of these ILs opens up interesting possibilities for the delivery of small amounts of template to the reaction mixture, as opposed to having the whole IL made up of the template. 3.11 Summary and Outlook

Normally, ILs are classed as ‘‘green’’ chemicals because they are most often used to replace volatile organic solvents. However, when preparing the materials, this perspective has been discussed and, in particular, all inorganic framework solids such as zeolites and ILs are more often than not replacing water. In these situations, ionothermal synthesis cannot be called a green technology compared to that which it replaces. When replacing organic solvents in, for example, the synthesis of metal organic frameworks there is more justiﬁcation for using the ‘‘green’’ tag.

101

102

3 Ionothermal Synthesis of Zeolites and Other Porous Materials

However, even in these syntheses, the success of the methodology has to rely on the ILs introducing new chemistry into the system that is not possible using other systems. Fortunately, over recent years, ionothermal synthesis has been recognized as a highly ﬂexible methodology that does indeed bring new chemistry to the system. Features like water deactivation and chiral induction offer many possibilities for the preparation of materials that are unlikely or even impossible to make in other solvents. One of the most interesting features of ILs for ionothermal synthesis is the sheer number of possible liquids available. There are an estimated 1 million binary ILs available, compared to only a few hundred molecular solvents. The wide range of accessible properties of the liquids provides huge opportunities for matching the chemistry of the solvent system to that of the reactants. However, this also presents huge challenges – it is at the moment extremely difﬁcult to predict a priori the properties of the solvent and how they will behave in combination with the reactants. Up to now, only a few of the easily available ILs have been studied, leaving many potentially interesting solvents completely unexplored. One particularly interesting feature of ionothermal synthesis is the use of mixed ILs to tailor the solvent toward a particular reaction chemistry by mixing two different miscible ILs to produce a new solvent with different properties (Section 3.8). Once again the issue of predicting the properties of the mixed ILs is a problem. However, this type of approach is particularly suited to high-throughput methodologies because new solvents can be prepared simply by mixing two ILs in various amounts, and the ‘‘brute force’’ approach afforded by high-throughput instrumentation can at least identify areas of interest in the compositional ﬁelds. The use of ILs in the synthesis of solids has, of course, not been limited to new hybrid and inorganic framework solids. Work in the nanomaterials area and increasingly in other areas, such as the organic solid state, has increased steadily over the last few years. However, there is still much scope to develop the synthesis methodology further. In the ﬁeld of zeolite science, the challenges are clear, particularly for the synthesis of silica-based zeolites. Here the plethora of possible ILs is both a blessing and a challenge as we really need to understand more fully the speciation of silicate ions in particular when they are dissolved in ILs. It is clear that the change from molecular to ionic solvents signiﬁcantly affects the chemistry, and that new zeolite-type structures will inevitably arise from ionothermal preparations. We hope that as we discover ever more about the interesting properties of ILs the ﬁeld of ionothermal synthesis will develop into an even more useful addition to the armory of synthetic zeolite chemists. References 1. Davis, M.E. (2002) Nature, 417,

3. Lee, H., Zones, S.I., and

813–821. 2. Paillaud, J.L., Harbuzaru, B., Patarin, J., and Bats, N. (2004) Science, 304, 990–992.

Davis, M.E. (2003) Nature, 425, 385–388. 4. Caullet, P., Paillaud, J.L., Simon-Masseron, A., Soulard, M.,

References

5.

6.

7.

8.

9.

10. 11.

12. 13.

14.

15. 16. 17.

18.

19.

20.

and Patarin, J. (2005) C. R. Chim, 8, 245–266. Villaescusa, L.A., Lightfoot, P., and Morris, R.E. (2002) Chem. Commun, 2220–2221. Villaescusa, L.A., Wheatley, P.S., Bull, I., Lightfoot, P., and Morris, R.E. (2001) J. Am. Chem. Soc, 123, 8797–8805. Blackwell, C.S., Broach, R.W., Gatter, M.G., Holmgren, J.S., Jan, D.Y., Lewis, G.J., Mezza, B.J., Mezza, T.M., Miller, A.M., Moscoso, J.G., Patton, R.L., Rohde, L.M., Schoonover, M.W., Sinkler, W., Wilson, B.A., and Wilson, S.T. (2003) Angew. Chem, 42, 1737–1740. Corma, A., Diaz-Cabanas, M.J., Jorda, J.L., Martinez, C., and Moliner, M. (2006) Nature, 443, 842–845. Cooper, E.R., Andrews, C.D., Wheatley, P.S., Webb, P.B., Wormald, P., and Morris, R.E. (2004) Nature, 430, 1012–1016. Rogers, R.D. and Seddon, K.R. (2003) Science, 302, 792–793. Blanchard, L.A., Hancu, D., Beckman, E.J., and Brennecke, J.F. (1999) Nature, 399, 28–29. Cole-Hamilton, D.J. (2003) Science, 299, 1702–1706. Earle, M.J., Esperanca, J., Gilea, M.A., Lopes, J.N.C., Rebelo, L.P.N., Magee, J.W., Seddon, K.R., and Widegren, J.A. (2006) Nature, 439, 831–834. Chou, S.L., Wang, J.Z., Sun, J.Z., Wexler, D., Forsyth, M., Liu, H.K., MacFarlane, D.R., and Dou, S.X. (2008) Chem. Mater, 20, 7044–7051. Abbott, A.P. and McKenzie, K.J. (2006) Phys. Chem. Chem. Phys, 8, 4265–4279. Miao, W.S. and Chan, T.H. (2006) Acc. Chem Res, 39, 897–908. Mugavero, S.J., Bharathy, M., McAlum, J., and zur Loye, H.C. (2008) Solid State Sci, 10, 370–376. Wasserscheid, P. and Welton, T. (2003) Ionic Liquids in Synthesis, Wiley-VCH Verlag GmbH, Weinheim. Reichert, W.M., Holbrey, J.D., Vigour, K.B., Morgan, T.D., Broker, G.A., and Rogers, R.D. (2006) Chem. Commun, 4767–4779. Nockemann, P., Thijs, B., Pittois, S., Thoen, J., Glorieux, C., Van Hecke, K., Van Meervelt, L., Kirchner, B., and

21.

22.

23. 24. 25.

26.

27.

28.

29.

30.

31. 32. 33.

34.

35.

36.

37.

38.

Binnemans, K. (2006) J. Phys. Chem. B, 110, 20978–20992. Abbott, A.P., Capper, G., Davies, D.L., Rasheed, R.K., and Tambyrajah, V. (2003) Chem. Commun, 70–71. Abbott, A.P., Boothby, D., Capper, G., Davies, D.L., and Rasheed, R.K. (2004) J. Am. Chem. Soc, 126, 9142–9147. Cundy, C.S. and Cox, P.A. (2003) Chem. Rev, 103, 663–701. Morris, R.E. and Weigel, S.J. (1997) Chem. Soc. Rev, 26, 309–317. Luo, H.M., Baker, G.A., and Dai, S. (2008) J. Phys. Chem. B, 112, 10077–10081. Lobo, R.F., Zones, S.I., and Davis, M.E. (1995) J. Mol. Incl. Phen. Mol. Rec. Chem, 21, 47. Drylie, E.A., Wragg, D.S., Parnham, E.R., Wheatley, P.S., Slawin, A.M.Z., Warren, J.E., and Morris, R.E. (2007) Angew. Chem. Int. Ed, 46, 7839–7843. Liu, L., Kong, Y., Xu, H., Li, J.P., Dong, J.X., and Lin, Z. (2008) Microporous Mesoporous Mater, 115, 624–628. Han, L.J., Wang, Y.B., Li, C.X., Zhang, S.J., Lu, X.M., and Cao, M.J. (2008) AIChE, 54, 280–288. Hu, Y., Liu, Y.J., Yu, J.Y., Xu, Y.P., Tian, Z.J., and Lin, L.W. (2006) Chin. J. Inorg. Chem, 22, 753–756. Parnham, E.R. and Morris, R.E. (2006) J. Am. Chem. Soc, 128, 2204–2205. Parnham, E.R. and Morris, R.E. (2006) J. Mater. Chem, 16, 3682–3684. Parnham, E.R., Wheatley, P.S., and Morris, R.E. (2006) Chem. Commun, 380–382. Xu, Y.P., Tian, Z.J., Xu, Z.S., Wang, B.C., Li, P., Wang, S.J., Hu, Y., Ma, Y.C., Li, K.L., Liu, Y.J., Yu, J.Y., and Lin, L.W. (2005) Chin. J. Catal, 26, 446–448. Wang, L., Xu, Y.P., Wang, B.C., Wang, S.J., Yu, J.Y., Tian, Z.J., and Lin, L.W. (2008) Chemistry, 14, 10551–10555. Wang, L., Xu, Y.P., Wei, Y., Duan, J.C., Chen, A.B., Wang, B.C., Ma, H.J., Tian, Z.J., and Lin, L.W. (2006) J. Am. Chem. Soc, 128, 7432–7433. Wang, T.W., Kaper, H., Antonietti, M., and Smarsly, B. (2007) Langmuir, 23, 1489–1495. Parnham, E.R. and Morris, R.E. (2007) Acc. Chem. Res, 40, 1005–1013.

103

104

3 Ionothermal Synthesis of Zeolites and Other Porous Materials 39. Ma, Y.C., Xu, Y.P., Wang, S.J., Wang,

40.

41. 42. 43.

44.

45.

46. 47.

48.

49. 50.

51. 52.

53.

54. 55.

56.

B.C., Tian, Z.J., Yu, J.Y., and Lin, L.W. (2006) Chem. J. Chin. Univ, 27, 739–741. Kitagawa, S., Kitaura, R., and Noro, S. (2004) Angew. Chem. Int. Ed, 43, 2334–2375. Ferey, G. (2008) Chem. Soc. Rev, 37, 191–214. Morris, R.E. and Wheatley, P.S. (2008) Angew. Chem. Int. Ed, 47, 4966–4981. Rosi, N.L., Eckert, J., Eddaoudi, M., Vodak, D.T., Kim, J., O’Keeffe, M., and Yaghi, O.M. (2003) Science, 300, 1127–1129. Banerjee, R., Phan, A., Wang, B., Knobler, C., Furukawa, H., O’Keeffe, M., and Yaghi, O.M. (2008) Science, 319, 939–943. Xiao, B., Wheatley, P.S., Zhao, X.B., Fletcher, A.J., Fox, S., Rossi, A.G., Megson, I.L., Bordiga, S., Regli, L., Thomas, K.M., and Morris, R.E. (2007) J. Am. Chem. Soc, 129, 1203–1209. Chen, S.M., Zhang, J., and Bu, X.H. (2008) Inorg. Chem, 47, 5567–5569. Hogben, T., Douthwaite, R.E., Gillie, L.J., and Whitwood, A.C. (2006) CrystEngComm, 8, 866–868. Ji, W.J., Zhai, Q.G., Hu, M.C., Li, S.N., Jiang, Y.C., and Wang, Y. (2008) Inorg. Chem. Commun, 11, 1455–1458. Liao, J.H. and Huang, W.C. (2006) Inorg. Chem. Commun, 9, 1227–1231. Liao, J.H., Wu, P.C., and Bai, Y.H. (2005) Inorg. Chem. Commun, 8, 390–392. Liao, J.H., Wu, P.C., and Huang, W.C. (2006) Cryst. Growth Des, 6, 1062–1063. Lin, Z.J., Li, Y., Slawin, A.M.Z., and Morris, R.E. (2008) Dalton Trans, 3989–3994. Shi, F.N., Trindade, T., Rocha, J., and Paz, F.A.A. (2008) Cryst. Growth Des, 8, 3917–3920. Xu, L., Choi, E.Y., and Kwon, Y.U. (2007) Inorg. Chem, 46, 10670–10680. Zhang, J., Chen, S.M., and Bu, X.H. (2008) Angew. Chem. Int. Ed, 47, 5434–5437. Zhang, J., Wu, T., Chen, S.M., Feng, P., and Bu, X.H. (2009) Angew. Chem. Int. Ed., 48, 3486–3490.

57. Lin, Z.J., Wragg, D.S., and Morris, R.E.

(2006) Chem. Commun, 2021–2023. 58. Xu, Y.P., Tian, Z.J., Wang, S.J., Hu, Y.,

59. 60.

61. 62.

63.

64.

65.

66.

67. 68. 69.

70.

71.

72.

73.

74.

Wang, L., Wang, B.C., Ma, Y.C., Hou, L., Yu, J.Y., and Lin, L.W. (2006) Angew. Chem. Int. Ed, 45, 3965–3970. Wragg, D.S. and Morris, R.E. Solid State Sci, in press. Cai, R., Sun, M.W., Chen, Z.W., Munoz, R., O’Neill, C., Beving, D.E., and Yan, Y.S. (2008) Angew. Chem. Int. Ed, 47, 525–528. Morris, R.E. (2008) Angew. Chem. Int. Ed, 47, 442–444. Xing, H.Z., Li, J.Y., Yan, W.F., Chen, P., Jin, Z., Yu, J.H., Dai, S., and Xu, R.R. (2008) Chem. Mater, 20, 4179–4181. Byrne, P.J., Wragg, D.S., Warren, J.E., and Morris, R.E. (2009) Dalton Trans, 795. Himeur, F., Wragg, D.S., Stein, I., and Morris, R.E. Solid State Sci., doi: 10.1016/j.solidstatesciences.2009.05.023. Lin, Z., Wragg, D.S., Warren, J.E., and Morris, R.E. (2007) J. Am. Chem. Soc, 129, 10334. Hulvey, Z., Wragg, D.S., Lin, Z., Morris, R.E., and Cheetham, A.K. (2009) Dalton Trans, 1131. Lin, Z., Slawin, A.M.Z., and Morris, R.E. (2007) J. Am. Chem. Soc, 129, 4880. Hanke, C.G. and Lyndon-Bell, R.M. (2003) J. Phys. Chem. B, 107, 10873. Amigues, E., Hardacre, C., Keane, G., Migaud, M., and O’Neill, M. (2006) Chem. Commun, 72. Morris, R.E., Burton, A., Bull, L.M., and Zones, S.I. (2004) Chem. Mater, 16, 2844–2851. Camblor, M.A., Villaescusa, L.A., and Diaz-Cabanas, M.J. (1999) Top. Catal, 9, 59–76. Zones, S.I., Darton, R.J., Morris, R., and Hwang, S.J. (2005) J. Phys. Chem. B, 109, 652–661. Villaescusa, L.A., Lightfoot, P., and Morris, R.E. (2002) Chem. Commun, 2220–2221. Bull, I., Villaescusa, L.A., Teat, S.J., Camblor, M.A., Wright, P.A., Lightfoot, P., and Morris, R.E. (2000) J. Am. Chem. Soc, 122, 7128–7129.

References 75. Villaescusa, L.A., Wheatley, P.S., Bull, I.,

Lightfoot, P., and Morris, R.E. (2001) J. Am. Chem. Soc, 123, 8797–8805. 76. Ma, H.J., Tian, Z.J., Xu, R.S., Wang, B.C., Wei, Y., Wang, L., Xu, Y.P., Zhang, W.P., and Lin, L.W. (2008) J. Am. Chem. Soc, 130, 8120.

77. Parnham, E.R. and Morris, R.E. (2006)

Chem. Mater, 18, 4882. 78. Parnham, E.R., Drylie, E.A., Wheatley,

P.S., Slawin, A.M.Z., and Morris, R.E. (2006) Angew. Chem. Int. Ed, 45, 4962.

105

107

4 Co-Templates in Synthesis of Zeolites Joaquin P´erez-Pariente, Raquel Garcı´a, Luis G´omez-Hortig¨uela, and Ana Bel´en Pinar

4.1 Introduction

Since the pioneering work by Barrer, the stabilizing role that guest molecules occluded inside zeolite cavities play in the crystallization of these materials has been widely recognized [1]. Apart from inorganic cations, whose inﬂuence in the zeolite crystallization is not discussed here, water and different types of organic molecules contribute to the stabilization of the otherwise intrinsically unstable low-density zeolite frameworks, at the expense of the more stable denser crystalline phases. In the particular case of organic molecules, the template concept was ﬁrst used by Aiello and Barrer [2], making reference to the mechanism that allows the trapping of tetramethylammonium (TMA) cations inside the gmelinite cages during the synthesis of zeolites offretite and omega. Since then, the concept of ‘‘templating’’ [3] has been extremely fruitful for the synthesis of a large variety of new zeolite structures [4, 5]. Although in most of the experimental approaches in this ﬁeld, just only one type of organic molecule is present in the synthesis gel as a structure-directing agent (SDA), there are also a number of structures that require the simultaneous presence of at least two different templates to crystallize. This chapter is devoted to the analysis of this strategy for the synthesis of zeolite-type materials, including both silica-based and aluminophosphate-based structures. Besides the structure-directing role of organic molecules, experimental evidence accumulated over the years shows the determining inﬂuence that inorganic guest species other than cations, such as water and ﬂuoride anions, exert in directing the crystallization of speciﬁc zeolitic structures, in combination with the organic molecules. In addition, this structure-directing concept could eventually be extended to the inorganic atoms of the framework other than Si, Al, or P that are incorporated in the microporous networks, such as germanium, for there are several Ge-containing materials which either do not have pure Si counterparts or crystallize more easily when Ge is present in the synthesis gel [6, 7]. However, we discuss the effect of the organic compounds on the synthesis of microporous

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

108

4 Co-Templates in Synthesis of Zeolites

materials in this chapter; inorganic species are taken into account only in those cases where they are chemically associated to the organic compound to promote zeolite crystallization.

4.2 Templating of Dual-Void Structures

Use of organic molecules in the synthesis of zeolites received a boost at the beginning of the 1980s, when many novel high-silica materials were claimed to crystallize from synthesis gels containing many different types of organic molecules as templates [8]. Most of these results were reported in patents, which also described some of the properties of the obtained zeolites. All these documents were comprehensively reviewed and analyzed in a work that still remains an invaluable source to study zeolite materials [9]. Some of these early works already reported the use of mixed-template systems, to which we refer to as co-templates, for the synthesis of high-silica zeolites, and they showed that the co-templating concept was already successfully applied at that time to crystallize zeolitic structures containing cages and channels of different size. What may be the ﬁrst example of this synthesis strategy was provided by Mobil researchers who, in 1980, reported the crystallization of zeolites ZSM-39 and ZSM-48 from gels containing TMA cations and propylamine [10–12]. The structure of ZSM-39 was later solved [13] and seems to be similar to that of dodecasil-3C (MTN structure type). It turned out to be composed of packing of two types of cages of different size, a smaller pentagonal dodecahedron and a larger hexadecahedron. Although no location of the organic molecules was found at that time, subsequent X-ray diffraction (XRD) studies of the as-made material synthesized from mixtures of pyridine and propylamine [14] conﬁrmed what was suspected when the structure was ﬁrst solved – the selective occupation of both cages by the two organic molecules as a function of their molecular size, the bulky pyridine molecule (or the TMA cation used in the earlier works) being occluded within the large cage, whereas the smaller propylamine molecule is accommodated inside the small one. Hence, crystallization of this material requires two different templates with different size to stabilize two different-sized cages. Synthesis of ZSM-39 was later optimized by using mixtures of TMABr (instead of TMACl) and ethylamine [15]. Similarly, ZSM-48 was obtained in the presence of TMA and n-propylamine by Chu in 1980 [16], although no experimental evidence of the occlusion of these organic molecules was provided. An improved synthesis of this structure involves the use of TMABr and octylamine [17]. The structure of this zeolite, ZSM-48, was initially explained using a model based on ferrierite sheets. An intergrowth of two of the structures resulting from the different conﬁgurations of the sheets was proposed as a model for the real disordered structure [18]. More recently, ZSM-48 has been described as a family of disordered materials assembled from silicate tubes made up of 10 T-atoms (where

4.2 Templating of Dual-Void Structures

T refers to tetrahedral atom sites). Connection occurs either via four-rings or via zig-zag crankshaft chains, forming a stacking of sheets of directly connected tubes. This building scheme gives rise to several ordered polytypes, two of which are equivalent to the models based on ferrierite sheets proposed earlier [19]. The ferrierite framework has a two-dimensional channel structure, formed by 10-ring channels parallel to the c-axis and 8-ring channels parallel to the b-axis. The intersection of the eight-ring channels with the smaller six-ring channels, which run along the c-axis, forms the ferrierite cavity, accessible through eight-ring windows. Then, the ferrierite structure is appropriate to probe a dual-templating concept, where the small cage and the 10-ring channel would be stabilized by suitable organic guests. An early example of the synthesis of ferrierite in the presence of two templates was reported in 1981 [20]. In this patent a synthetic ferrierite, FU-9, was prepared employing several combinations of TMA with different tertiary amines (Table 4.1). In the synthesis with TMA + trimethylamine (Si/Al = 10.5) the C/N ratio was found to be 3.6, which is between the C/N ratios of the free molecules, thus suggesting the incorporation of both SDAs in the zeolite. Replacement of triethylamine with tributylamine or triethanolamine resulted also in the crystallization of FU-9. The templating role of different organic molecules present in the synthesis gels leading to ferrierite crystals has been elucidated by XRD techniques. Fully connected Table 4.1

Synthesis of FER and MWW-type materials from co-template systems.

SDA1 FER-type materials Propylamine Triethanolamine Tributylamine Trimethylamine Propylamine bmp bmp bmp bmpm bmp bmpm Quinuclidine MWW-type materials TMAda+ TMAda+ TMAda+ TMAda+ Polycyclic amines

T (◦ )

SDA2

Phase

F−

TMA TMA TMA TMA Pyridine TMA Quinuclidine hydrochloride TEA TMA Quinuclidine hydrochloride TMA TMA

ZSM-48 FU-9 FU-9 FU-9 FER FER FER layered FER layered FER related FER + MWW FER + MWW MCM-65(CDO)

– – – – – – – – + – + 135, 150 + 135 + 135, 150 + 135 + 150 + 150 – 180

Dipropylamine Hexamethyleneimine Isobutylamine Piperidine Isobutylamine

ITQ-1 ITQ-1 SSZ-25 SSZ-25 SSZ-25

‘‘+’’ in the F− column indicates the presence of this anion in the gel.

– – – – –

150 150 170 170 170

References

[16] [20] [20] [20] [14] [21] [22] – – [22] – [23] [24] [24] [25] [25] [25]

109

110

4 Co-Templates in Synthesis of Zeolites

ferrierite zeolite was prepared with pyridine and propylamine, in HF/pyridine solvent [14]. Single crystal XRD studies showed that pyridine is located in both the ferrierite cavity and the 10-ring channel, while a small amount of propylamine is located exclusively in the channels. Propylamine thus seems to play a nucleating role in the synthesis of ferrierite, according to the small amount of this molecule found in the ﬁnal product and the decrease of the crystals size when increasing the amount of propylamine in the gel. Instead, ZSM-39 (MTN) was obtained in the absence of propylamine, which remarks the key role of these molecules in the crystallization of ferrierite under these synthesis conditions. This competitive phase is also obtained in the presence of both SDAs at higher temperatures or longer heating times. In this case, propylamine was found in the small cavity, and pyridine, which was too bulky to occupy the small cage, was located in the larger one. Furthermore, it has been shown that the simultaneous use of the bulky 1-benzyl-1-methylpyrrolidinium (bmp, Figure 4.1) and the smaller TMA cation leads to the crystallization of this zeolite structure [21]. When the size of the co-SDA increases to those of quinuclidine and tetraethylammonium (TEA) while the molecule remains the same, that is, bulky, the outcome of the synthesis is layered ferrierite materials. The structure of such phases is made up by the stacking of the individual ferrierite layers, with the organic molecules located in the interlayer region. As discussed later, for the MWW structure, there is a family of layered materials related to ferrierite, to which many different members belong, such as MCM-47, the borosilicate ERS-12, or the UZM series of materials [26–28]. In the synthesis using quinuclidine, the interlayer distance of these ferrierite layered phases decreased with the crystallization time, which was attributed to the exchange of bmp by quinuclidine as the crystallization proceeded [22]. When a bulkier organic molecule such as TEA was used as a co-SDA, related layered materials were obtained though, for this co-SDA, the interlayer distance did not vary with the crystallization time. Molecular mechanics calculations revealed a tendency of each of the SDAs used in these preparations to occupy speciﬁc sites within the ferrierite structure, evidencing a co-structure-directing effect and allowing to understand the experimental observations: the bulky bmp molecules accommodated along the 10-ring channels and the smaller co-SDA within the ferrierite cages (Figure 4.2). These calculations show that while the TMA cation has a strong preference for accommodation within the ferrierite cages, quinuclidine is too large to ﬁt properly in the ferrierite cavities and it would be better for it to be accommodated in a somewhat larger cavity. For TEA, its location within the ferrierite cavities would lead to an even more unstable situation, which explains that the layered FER structure obtained with TEA is not able to evolve to the fully condensed FER structure, in contrast to the one obtained with quinuclidine. Moreover, XRD studies show that the TMA cations are indeed located in the ferrierite cage, while bmp occupies the 10-ring channel (Pinar et al., manuscript in preparation). Replacement of the bulky SDA in these preparations, bmp, by a related chiral cation, 2-hydroxymethyl-1-benzyl-1-methylpyrrolidinium (bmpm, see Figure 4.1)

4.2 Templating of Dual-Void Structures

O

O

O

N

N

O (a)

O

N

N

N

N

(d)

O

O

N O O N O (b)

H3C

N

N O

O

H3 C N N

O

(c) N

CH3

N N CH 3

(e)

(f)

N

N

N (g)

(h)

(i)

N N

N OH (j)

(k)

Figure 4.1 Structures of some of the organic molecules used as co-structure-directing agents in the synthesis of zeolites: (a) Kryptoﬁx22, (b) Kryptoﬁx222, (c) Kryptoﬁx21, (d) 1,4,8,11-tetraazacyclotetradecane (cyclam), (e) 1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane (Tetramethylcyclam),(f)

(l) N, N, N-trimethyl-1-adamantanammonium (TMAda+ ), (g) hexamethyleneimine (HMI), (h) N, N-dimethyl-3-azoniabicyclo [4.2.1] nonane, (i) 1-benzyl-1-methylpyrrolidinium (bmp), (j) 2-hydroxymethyl-1-benzyl-1-methylpyrrolidinium (bmpm), (k) 4-methyl-2,3,6,7tetrahydro-1H,5H-pyrido [3.2.1-ij] quinolinium, and (l) quinuclidine.

was shown to direct the synthesis to ferrierite-related phases by using TMA as a co-SDA, evidencing again the high tendency of this cation to produce ferrierite. It is worth noting that in many of the above preparations that yielded ferrierite-related phases, the co-crystallization of a phase of the MWW family was observed at higher synthesis temperatures (150 ◦ C, see Table 4.1). Another material containing ferrierite layers, MCM-65, can also be synthesized by the combination of quinuclidine and TMA. The ferrierite layers of this precursor are stacked in such a manner that they render a new cage structure upon calcination. The structure of the calcined MCM-65, the CDO structure type, comprises a two-dimensional network of eight-ring channels [23]. Both SDAs used in its synthesis (quinuclidine and TMA) were found to be intact and rigidly held in the voids of the MCM-65 precursor.

111

112

4 Co-Templates in Synthesis of Zeolites

Assembly

bmp

TMA bmp

TMA

Figure 4.2 Scheme of the self-assembly of TMA-ﬁlled cavities around bmp molecules to give the ﬁnal ferrierite structure.

One of the zeolite structures frequently obtained in the presence of two SDAs is MWW, which is moreover often found to crystallize together with ferrierite in many preparations. The structure of this zeolite is composed of two independent channel systems accessible through 10-ring pores, one formed by large cages delimited by 10-ring apertures and the other one by two-dimensional sinusoidal ˚ of the cages. The channels that run perpendicular to the long dimension (18 A) three-dimensional MWW structure is usually obtained via a layered precursor MCM-22 (P) that forms the characteristic three-dimensional zeolite when calcined. There is a family of materials with a structure related to MWW, such as MCM-22, MCM-56, and MCM-49, where the differences between them are mainly due to the packing degree of the layers [29]; MCM-49 presents the same framework topology as calcined MCM-22, and MCM-56 maintains the layered structure on calcination. Related materials are the aluminosilicates SSZ-25 [30], the boron-containing analog ERB-1 [31], and the pure silica polymorph ITQ-1 [32]. Camblor et al. reported the synthesis of the pure silica zeolite ITQ-1 using N, N, N-trimethyl-1-adamantanammonium (TMAda+ , Figure 4.1) as the only SDA [32]. However the synthesis was hardly reproducible. They found that adding a second organic component to the system, hexamethyleneimine (HMI) or dipropylamine (DPA), allowed a faster and more reproducible synthesis [24]. They proposed that TMAda+ , since it is too large to ﬁt in the sinusoidal channel system, stabilizes the 12-ring cages, whereas the amine additives help in crystallizing ITQ-1 by ﬁlling the 10-ring sinusoidal channel system. Indeed, using a bulkier amine, diisobutylamine (DIBA), under the same conditions as HMI and DPA, did not result in the crystallization of ITQ-1. In addition, the amines seemed to exert some effect on the pH of crystallization, since the use of HMI as the hydrochloride salt yielded the zeolite MCM-35 (MTF structure type) instead of ITQ-1, where only the HMI molecules remained occluded within the cages of the structure [33].

4.3 Crystallization of Aluminophosphate-Type Materials

SSZ-25, a MWW-related material, was also synthesized in a two-component organic system, using the quaternary ammonium cation TMAda+ and different amines as pore ﬁllers. Zones et al. studied the crystallization of SSZ-25, replacing both components [25]. The system showed a high ﬂexibility and other adamantane derivatives, such as the free amine or the alcohol, could be used instead of TMAda+ at low concentrations to produce SSZ-25. In addition, under the same conditions, other large polycyclic hydrocarbon cations were also shown to yield SSZ-25, combined mostly with isobutylamine and piperidine. All these molecules possess a size that prevents them from ﬁtting in the 10-ring sinusoidal channels and, as commented above, they most likely occupy the large cages in the structure. It was shown that the adamantyl component can be decreased in the synthesis gel, which results in a larger uptake of the isobutylamine. However, a certain amount of this component is required to crystallize the zeolite, since in its absence, other zeolites are produced. Mixed-template systems involving simple and commonly used tetralkylammonium cations still have a large potential for discovering new zeolite structures that require a unique combination of those SDAs under suitable synthesis conditions. A few years ago, a research team of the UOP company reported the synthesis of two open-framework materials with Si/Al < 10, from TMA/TEA mixed systems: the zeolites UZM-4 and UZM-5 [34]. The ﬁrst one is a 12-ring channel structure with the BPH framework topology, initially synthesized as a beryllophosphate, and the second one is a novel 8-ring structure related to that of zeolite A. The gel rendering UZM-4 also contains lithium, and both organic molecules are incorporated into the structure, the TEA/TMA ratio being much higher in UZM-5. On some occasions, the mixed-template approach results in failure [35]. Zones et al. studied the structure-directing effect of several long symmetric diquaternary ammonium cations of varying chain length under different reaction conditions. One of the synthesis variables of these preparations was the use of a second SDA, the small TMA cation, together with a bulky diquaternary compound. However, in the presence of TMA cations, only clathrasil phases were obtained, probably templated by the TMA cation.

4.3 Crystallization of Aluminophosphate-Type Materials

The co-templating strategy has also been applied to the synthesis of different aluminophosphate structures. One of the ﬁrst examples was the crystallization of silicoaluminophosphate (SAPO)-37 from gels containing mixtures of TMA and tetrapropylammonium (TPA) hydroxides [36–38]. SAPO-37 possesses the faujasite structure, which provides a topological explanation for the role of each template: the TMA cations are located within the sodalite cages, while the bulkier TPA cations occupy the supercages. This material crystallizes in a relatively narrow range of TMA/TPA ratios in the synthesis gel. Unbalance of this ratio produces SAPO-20 (SOD structure-type) if an excess of TMA is used while, on the other side,

113

114

4 Co-Templates in Synthesis of Zeolites

predominance of TPA favors the crystallization of the large-pore SAPO-5, which possesses unidirectional 12-ring channels but no cavities. This example illustrates nicely what is commonly observed by using mixed-template systems for the crystallization of zeolite-type materials, namely, a delicate balance between both templates is required for successful synthesis, since both molecules cooperate to stabilize the dual-void structures in preference to the ones favored by each template separately. One of the zeolite structures frequently obtained with a variety of combinations of SDAs is LTA. This structure type is built up by the connection of double four-rings creating a framework with large α-cavities and smaller sodalite cages (Figure 4.3). The macrocycle, azaoxacryptand 4,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8.8.8] hexacosane (K222) (Figure 4.1), was shown to direct the formation of the aluminophosphate version of the LTA structure and its substituted derivatives when it was used as the only SDA [39]. In the presence of TMA cations and ﬂuoride anions, a rhombohedral variant of the LTA structure was obtained. The use of other Kryptoﬁx-n macrocycles led to the crystallization of the LTA topology in the presence of ﬂuoride and/or TMA cations. In all cases, the Kryptoﬁx-n molecules were most likely located within the α-cages of the structure, the ﬂuoride anions in the D4Rs and the TMA cations in the smaller sodalite cages [40], thus showing the cooperation of all the species in the construction of the cavities of the zeolite structure. A similar scenario was found with other combinations of SDAs. A copper–cyclam complex also produced LTA from aluminophosphate gels containing ﬂuoride and trimethylamine [41]. The Cu–cyclam complex was located within the α-cages, with trimethylamine in the sodalite cage and F− anions in the D4R units. Likewise, combination of diethanolamine, ﬂuoride anions, and TMA cations yielded the same framework structure, with each of them structure-directing one of the three polyhedra present [42]. Diethanolamine was located in the α-cage, with the nitrogen atom in the middle of the eight-ring of the cavity and the two ethanol

B

(a)

A

(b) Figure 4.3 (a) The structure of LTA, showing the smaller sodalite cages at the center of four large α-cages; (b) the SAV framework showing the bigger cages (labeled A) and the smaller cages (labeled B) of the structure; and (c) the structure of KFI showing the α-cages (A), as those in LTA, and the smaller MER cages (B).

A

(c)

B

4.3 Crystallization of Aluminophosphate-Type Materials

groups pointing to the contiguous sodalite cavities, while F− and TMA were located in the D4R unit and in the sodalite cage, respectively. Interestingly, eight water molecules were additionally located in the α-cage close to the six-rings, probably linked to the terminal OH groups of the amine through hydrogen bonds, and so they were also considered as templates. In this system, the replacement of the small diethanolamine molecule by the bulkier TPA cations led to the crystallization of materials having larger voids accessible through 12-ring apertures, possessing either FAU or AFR structures depending on speciﬁc synthesis conditions [43]. The aluminosilicate version of the LTA structure can also be prepared by using a mixture of templates. The cation 4-methyl-2,3,6,7-tetrahydro-1H,5H-pyrido [3.2.1-ij] quinolinium iodide (Figure 4.1) was able to direct the formation of this structure in germanosilicate preparations in ﬂuoride medium; this cation was shown to form self-assembled dimers within the α-cages of the structure. Interestingly, the addition of TMA to the synthesis gels, together with the quinolinium derivative, allowed for obtaining the pure silica version of the material, and both cations remained occluded in the structure [44]. The macrocycle tetramethylcyclam (1,4,8,11-tetramethyl-1,4,8,11-tetraazacyclotetradecane) was known to direct the synthesis of two metalloaluminophosphates, STA-6 (SAS structure type) for metals such as Mg, Mn, and Si, and STA-7 (SAV structure type) in preparations with Zn and Co [45]. Both are cage-like structures based on D6R as secondary building units, although the cavities are different in both materials. In particular, the framework of STA-7 possesses two three-dimensional channel systems of eight-ring openings and two cages of different size. The large cage of the SAV structure is stacked along the c-axis through planar eight-membered rings, creating one of the channel systems. The cage also possesses four elliptical eight-ring openings to the second channel system that is made up of smaller cages (Figure 4.3). Addition of TEA cations to the synthesis gels, which in their absence yield STA-6, directs the synthesis toward the formation of STA-[46]. Single crystal diffraction showed that the macrocycle resides in the large cages of the structure while the TEA cations are located in the smaller cages. Furthermore, together with TEA cations, tetramethylcyclam can be replaced with the related and less expensive cyclam (Figure 4.1) to direct the formation of STA-7, reducing the cost of production of this material. On the basis of the above studies with macrocycles, the co-templating approach was also used to prepare the aluminophosphate version of the aluminosilicate zeolite ZK-5 (KFI structure type) [46]. This structure possesses two types of cages: the large α-cage (the same found in the structure of LTA) and a smaller cage also found in the structure of zeolite merlinoite (MER). Computer modeling suggested the TEA cation as a good SDA for the MER cages. Therefore, the KFI structure could be obtained instead of LTA by using the strong structure-directing action of K222 for the α-cage and in the presence of TEA as a co-SDA. Crystal diffraction showed that TEA cations are present in the MER cages of the KFI structure, as was predicted by modeling [46].

115

116

4 Co-Templates in Synthesis of Zeolites

Application of the co-templating concept is by no means restricted to hydrothermal crystallization. The use of ionic liquids for the synthesis of zeolite-type materials, aluminophosphates in particular, provides opportunities for new advances in this area, as the organic cations forming part of the ionic liquid may themselves act as templates. The ﬁrst example of co-templating ionothermal synthesis of a new open-framework aluminophosphate denoted by JIS-1 has been recently reported by using the aromatic amine 1-methyl-imidazole (MIA) and the ionic liquid 1-methyl-3-ethylimidazolium bromide (EMIMBr) [47]. The structure consists of an anionic open framework with an Al/P ratio of 6/7 that possesses three-dimensional intersecting 10-ring, 10-ring, and 8-ring channels along the three crystallographic axes, where the protonated amine and cations of the ionic liquid are simultaneously found in the intersection of the channels.

4.4 Combined Use of Templating and Pore-Filling Agents

A usually high cost of SDAs and their difﬁculty to be removed from the microporous networks led to the search for new zeolite synthesis systems, in which the amount of the SDAs required to crystallize the zeolites is reduced. Here, we ﬁnd another interesting application for the use of combinations of organic molecules as SDAs in the synthesis of zeolites [48, 49]. Zones and Hwang proposed [48] a new zeolite synthesis system, in which a minor amount of the SDA is used to selectively specify the nucleation product, and then a larger amount of a cheaper, less selective, and smaller molecule is used to provide both pore-ﬁlling and basicity capacities in the zeolite synthesis as the crystal continues to grow. This new synthesis system provides a successful route toward the cost-effective making of a number of aluminosilicate materials and, indeed, new zeolite topologies have also been discovered. In addition to reducing the amount of the expensive SDAs required in the synthesis, these systems also accelerate crystallization rates. Interestingly, there is a reagent ﬂexibility of the pore-ﬁlling agent to be used. Furthermore, the smaller size of the pore-ﬁlling agent makes it easier to be removed from the zeolite network, which can be achieved by simple extraction. As an example, crystallization of SSZ-25 can be accomplished by adding minor amounts of N, N, N-trimethyladamantyl ammonium, which directs the formation of the zeolite, and major amounts of isobutyl amine, which are occluded during the crystallization as a pore-ﬁlling agent. In addition, a new intergrowth material (SSZ-47) related to the structural family NES/EUO/NON was obtained by using a large amount of isobutylamine together with a smaller amount of a group of bicycloorgano-cations [49], such as N, N-dimethyl-3-azoniabicyclo [4.2.1] nonane (Figure 4.1), providing evidence that this approach can also aid in the search for new zeolite materials. A particular use of these two-component systems is when degradable organic molecules are used as SDAs [50–52]. SDA molecules have to be removed from the zeolite frameworks before their use in adsorption and catalytic applications. Owing

4.5 Cooperative Structure-Directing Effects of Organic Molecules and Mineralizing Anions

to their usual large size, removal of these encapsulated species normally requires high temperature combustion that destroys them, and the associated energy released in combination with the formed water can be extremely detrimental to the zeolite structure. In order to avoid this, Lee et al. proposed [50–52] a new class of organic SDAs that have the potential to be degraded into fragments within the pores so that they can be readily extracted from the zeolite under mild conditions, thus avoiding the necessity of the high-temperature calcination process. Furthermore, the extraction of the molecular fragments would allow reuse (recycle) of the SDAs by re-assembly, which might decrease the cost of the zeolite synthesis, as usually SDA species are the most expensive components of the synthesis gel. One class of these degradable organic SDA molecules is ketal-containing species that are very stable at high pH but can be cleaved into ketones and diols at low pHs. These ketal molecules will be intact during the zeolite synthesis that typically occurs at high pH, and fragmented into pieces by lowering the pH (hydrolysis). Under these conditions, access of H2 O and H+ to the ketal group of the SDA is essential for the hydrolysis reaction to occur. If this access is prohibited, the SDAs will remain intact despite the chemical treatment. This is where the use of the second organic molecule, the pore-ﬁlling agent, plays its crucial role: the inclusion of a second organic molecule, smaller than the main SDA, facilitates its removal by extraction, and then the void space created is used by H+ and H2 O species to access the SDA for the cleavage reaction to occur. The degradable SDA would remain intact after the acidic treatment without the previous extraction of the pore-ﬁlling agent, which demonstrates that the extraction of these species, and thus the use of a two-component system, is essential for the successful application of degradable SDAs.

4.5 Cooperative Structure-Directing Effects of Organic Molecules and Mineralizing Anions

In some of the examples described above the ﬂuoride anions are found to be located in the structure, where they play an additional stabilizing role, showing preference for very small cages, such as the D4R in LTA structures. However, we pay attention in this section to a somewhat different situation, where mineralizing species such as F− and OH− anions interact with the organic guest species to provide the actual templating chemical entity. Such a cooperative structure-directing effect has been experimentally observed in the crystallization in ﬂuoride medium of the all-silica EUO structure, which contains 10-ring one-dimensional channels with side-pockets, by using a ﬂuorine-containing molecule as an SDA, o-ﬂuorobenzyl-benzyl-dimethylammonium [53]. A combination of XRD studies and computer simulation of the as-made material reveals the development of a strong interaction between the ﬂuoride anions occluded in the structure and the F-containing molecules located in the channel, in such a way that the actual arrangement of both species tends to maximize the electrostatic attraction between the positively charged N atom of the molecule and the F− anions while reducing

117

118

4 Co-Templates in Synthesis of Zeolites

Figure 4.4 Location of o-ﬂuorobenzylbenzyl-dimethylammonium in the EUO structures; ﬂuorinated rings are invariably located in the 10-ring channel (highlighted between dashed yellow lines). Fluoride anions are displayed as large pink balls, while ﬂuorine organic atoms are displayed as small

pink balls. The side-pockets where nonﬂuorinated aromatic rings are located are shown by dashed green lines. The repulsion between ﬂuoride and ortho-F located in the side-pockets that prevents this arrangement is shown by blue arrows.

the F− ···F(SDA) repulsion, leading to a well-deﬁned location of the ﬂuorinated aromatic rings in the channel rather than in the cavities (Figure 4.4). This determines the cooperative effect, between both chemical species in the stabilizing EUO crystals, that if no F is present in the SDA, the competing zeolite Beta is obtained, which also crystallizes if both the aromatic rings contain one F atom each. In this last case, strong F---F repulsion cannot be avoided when the ﬂuoride anions and the ﬂuorine-containing SDA are located inside EUO pores, and this makes the nucleation of this zeolite less favorable, leaving a chance for zeolite Beta to crystallize [54]. In the absence of ﬂuoride and low-valent heteroatoms, the necessity for compensating the positive charge of the organic SDA molecules implies the formation of negatively charged structural defects; however, in certain cases, it has been observed that the hydroxide anions are incorporated in the frameworks as charge-compensating species [55], leading to the effective incorporation of neutral SDA+ ···OH− adducts in the microporous framework. In a similar manner as previously described for o-ﬂuorobenzyl-benzyl-dimethylammonium in the synthesis of EUO, the presence of active groups (F atoms) in the organic SDA molecules, bis(o-ﬂuorobenzyl)-dimethylammonium, used for the synthesis of the AFI structure leads to the development of strong interactions between the organic molecule and other species present in the synthesis gels, such as the hydroxide anions, giving room to the formation of strongly bonded supramolecular organo-inorganic entities. These units are stabilized by electrostatic interactions between the hydroxide

4.6 Cooperative Structure-Directing Effect of Organic Molecules and Water

Figure 4.5 A model of the proposed orientation of hydroxide anions within the 12MR channels, bridging the ortho-ﬂuorinated organic SDA and the framework. Dashed green line indicates the H-bond and dashed blue line indicates the coordination bond with Al. Fluorine atom (blue) and pentacoordinated aluminum atom are displayed as balls.

anions and the positive charge localized around the N atom of the organic molecule, and especially by the formation of H-bonds between the neighboring F atoms (in the ortho position, i.e., spatially close to the ammonium group) and the hydroxide anion. Such a stable supramolecular arrangement is the actual SDA of the AFI structure, in which both species are ﬁnally incorporated in the framework, with the hydroxide anions also coordinatively bonded to Al framework ions, thus bridging the organic molecule and the framework (Figure 4.5). This mode of cooperative structure direction, where the organic molecules template the formation of the microporous structure in addition to contributing to the stabilization by developing nonbonded interactions, while the hydroxide anions act as charge-compensating units, carries the beneﬁt of incorporating a neutral entity, thus preventing the formation of negatively charged framework defects that destabilize the microporous structure.

4.6 Cooperative Structure-Directing Effect of Organic Molecules and Water

So far in this chapter, we have dealt only with co-templating effects where both co-templating agents were organic molecules and/or anions (ﬂuoride or hydroxide). However, there is also another component in the synthesis gels of zeolite-type materials that could, in principle, play a role in the structure direction of these materials, namely, the water molecules. Traditionally, the hydrothermal synthesis of microporous materials has involved the use of water as the solvent (although new synthesis trends propose the use of other solvents like alcohols or ionic liquids). Nevertheless, recent studies have demonstrated that water can play an important role in structure direction together with the usual organic molecules, apart from its

119

120

4 Co-Templates in Synthesis of Zeolites

role as a solvent [56–58]. This is especially true in the case of hydrophilic zeolite-type materials, which are susceptible to strong interactions with water molecules. High-silica zeolites are hydrophobic in nature, and therefore, no strong interaction and so no adsorption of water within these materials is usually found. However, this is not the case for AlPO-based frameworks, where the molecular-ionic nature of the network, which comprises Al3+ and PO4 3− units ionically bonded, and the strict alternation of Al and P ions provide these materials with a high hydrophilic character. Therefore, hydrophilic AlPO frameworks are more susceptible to strong interactions with water. A recent work by our group has evidenced a clear cooperation between water and organic molecules when directing the synthesis of the large-pore AlPO-5 aluminophosphate (AFI structure type) with different organic molecules. Water itself can template the formation of small secondary building units, like the 6-MR channels in the AFI structure, where a clear host–guest correlation between their shape and symmetry can be appreciated (Figure 4.6a). However, water molecules

(a)

(b)

(c)

Figure 4.6 Cooperative structure-directing effect of triethylamine and water molecules in the synthesis of the AFI structure. (a) Occlusion of guest species in the AFI framework. (b,c) Two views of the H-bonded chains of water around triethylamine molecules within the AFI channel.

4.6 Cooperative Structure-Directing Effect of Organic Molecules and Water

themselves cannot template the formation of large-pore topological items, such as the 12-ring channels, since the H-bonding network between the water molecules is not strong enough as to hold the large-pore architecture at the high synthesis temperatures usually required. In this regard, the larger organic molecules are inevitably required to template the synthesis of large-pore microporous materials. Nonetheless, water can play a cooperative role in directing the synthesis of hydrophilic microporous aluminophosphates, depending on the structure-directing efﬁciency and the hydrophilicity of the organic molecule employed as the main SDA. In cases where the interaction of the organic SDA molecules with the microporous framework is not very strong, water can compete for being occluded within the frameworks during crystallization, thus strongly contributing to the ﬁnal stabilization of the microporous structure by the development of H-bond interactions with the oxygen atoms of the framework. In this case, water molecules have a cooperative structure-directing effect on the organic molecules, where neither water nor the organic molecules themselves are the true SDAs, but a water–organic molecule aggregate in which both species cooperatively act as the SDA entity. Such a cooperativeness involves a synergetic effect of the two species: the large size of the organic SDA molecules provides the templating role of the large-pore architecture, allowing the development of nonbonded interactions with the framework, while the strong dipole of water molecules provides further strong interactions with the structure through the formation of H-bonds with the atoms of the oxygen framework, thus leading to a strong stabilization of the structure. Such a mode of structure direction could not be achieved by the occlusion of the isolated species: despite the high interaction of water molecules, they cannot hold the large-pore architecture of the microporous structures, while the weaker interaction of certain SDA organic molecules is not strong enough as to stabilize the structure to make the crystallization pathway viable. This cooperative effect is nicely illustrated by the crystallization of AlPO-5 in the presence of triethylamine [56] (Figure 4.6), where it can be clearly observed that water molecules arrange as H-bonded chains surrounding the triethylamine organic molecules, locating always close to the channel walls and thus developing strong H-bonds with the framework oxygen atoms. It seems that this cooperative water–organic structure-directing effect occurs for low-interacting organic molecules, where such low interaction needs to be compensated by the simultaneous occlusion of strongly interacting water molecules. Similar cooperative effects of water have been found for benzylpyrrolidine molecules when directing the synthesis of the AFI and the SAO structures, or methylamine in the synthesis of IST-1 and IST-2 [58]. Indeed, another nice example of this cooperative role of water was long time ago observed in the synthesis of Si and metal-substituted AlPO4 materials with the LTA structure, where water molecules are hydrogen bonded to the OH groups of diethanolamine, as has been previously mentioned in this chapter [42]. Instead, molecules that interact more strongly with the framework, like a benzylpyrrolidine analog with an attached methanol group, and (S)-N-benzyl-pyrrolidine-2-methanol, tend to direct the crystallization of the frameworks by themselves, preventing the simultaneous occlusion of water.

121

122

4 Co-Templates in Synthesis of Zeolites

4.7 Control of Crystal Size and Morphology

In the examples discussed above, the use of co-template systems severely alters the crystallization pathway of the synthesis gel to make the synthesis of zeolite structures feasible that would not otherwise be obtained in the absence of such speciﬁc combination of co-templating agents. However, there are also examples where such modiﬁcations of the nucleation and crystal growth are less severe, and although they do not lead to new structures, they nevertheless affect the crystal size and shape of the resulting zeolite, as well as its chemical composition, in a noticeable way. Crown ethers, 18-crown-6 and 15-crown-5, have been used for the synthesis of the cubic zeolite faujasite and its hexagonal analog EMT by using ethylene glycol and 1,3,5-trioxane as co-templates [59]. The reason for adding the latter two organic compounds is that they allow the crystallization of high-silica sodalite. As sodalite units are also present in both FAU and EMT structures, the association of these neutral organics with the crown ethers was supposed to provide good candidates to enhance further the Si/Al ratio of these two large-pore materials. Pure EMT and FAU zeolites were obtained from the large and the small crown ethers, respectively; in both cases, the crystal size of the obtained materials increased up to 3–5 µm, which was attributed to a decrease in the supersaturation of the system, since the aluminosilicate species were less soluble in the synthesis media containing the co-templates, evidencing a new effect of the presence of co-templates on the crystal size. However, despite some ethylene glycol was found in the EMT crystals, the Si/Al ratios of both materials were similar to those obtained by using the crown ethers alone, showing that in this case no control over the Si/Al ratio was provided by the use of the co-templates. Very large crystals of zeolite Y (Si/Al = 1.7) with diameters up to 245 µm were obtained by using bis(2-hydroxyethyl)dimethylammonium chloride (TCl) as a co-template together with triethanolamine. In this case, the amine is used as a chelating agent for the Al3+ ions in the gel in order to retard the nucleation of faujasite, thus leading to larger crystals. However, zeolite P was formed as an impurity [60]. Attempts to improve this synthetic procedure by adding seeds of zeolite Y did not completely eliminate the formation of zeolite P [61]. Nonetheless, these works provide a new example of the usefulness of using co-templates for inﬂuencing the crystal growth of the microporous materials. Finally, a dual-template approach has been proposed for the synthesis of nanosized edingtonite-type (EDI topology) zeolite from gels containing a mixture of TMA and copper amine complexes [Cu(NH3 )4 ]2+ [62]. Moreover, if only copper cations or ammonia are present, pure nanosized FAU zeolite is obtained. Both TMA and the copper complex are present in the crystals, and in some cases 82% of the channel intersections are ﬁlled by them, which points to a true co-structure-directing role of the molecules, apart from their effect on the crystal size.

4.9 Use of Co-Templates for Tailoring the Catalytic Activity of Microporous Materials

4.8 Membrane Systems

The control of the crystal size and shape provided by the co-templating zeolite synthesis strategy is particularly useful for the production of high-performing zeolite membranes. Mixed-template systems have also been employed to change the morphology of MFI crystals in order to improve the performance of membranes made up of this zeolite in separation processes [63]. The target here is to obtain zeolite crystals with b-axis as the preferential orientation, which are the most effective for that purpose [64]. Mixed systems containing N-ethyl-hexamethylenetetrammonium bromide as one template, and n-propylamine, n-butylamine, or ethylamine as the second template were used to crystallize pure silica MFI. It was found that n-propylamine is the most effective one in promoting anisotropic growth of MFI crystals. Both the amine and the ammonium amine, and the ammonium cation are found to be present in the zeolite crystals, playing a co-templating role in the formation and growth of the b-oriented crystals. SAPO-34 (CHA structure type) membranes can be conveniently used for the separation of CO2 /CH4 mixtures, and it has been reported that by using a combination of SDAs in the preparation of SAPO-34 membranes by seeding on porous stainless steel supports, higher ﬂuxes and selectivities are obtained [65]. TEAOH was used as the main template, and di-n-propylamine and cyclohexylamine as the secondary ones. The addition of both amines to the synthesis gel decreases the crystal size, but the former is more effective. In this case, cubic crystals with a narrow size distribution (0.7 ± 0.06) µm were obtained, resulting in more effective membranes for the separation of CO2 and CH4 : smaller crystals with narrow size distributions pack better than large crystals, which leads to a decrease in the size of the intercrystalline pores and so to higher CO2 /CH4 selectivities.

4.9 Use of Co-Templates for Tailoring the Catalytic Activity of Microporous Materials

A key issue in zeolite catalysis refers to the possibility of controlling the distribution of aluminum atoms in the framework, and hence that of their associated acid sites. The preference for different SDAs to be located within speciﬁc and different void architectures in a given zeolite structure gives unexpected perspectives for gaining an effective control over Al, for these molecules would drive aluminum toward different T-sites according to their location. This hypothesis has been recently experimentally proven in the case of zeolite ferrierite [66–68]. For this purpose, crystals of zeolite ferrierite (Si/Al = 15) were synthesized in ﬂuoride media in the absence of alkali cations but in presence of different SDAs, pyrrolidine, pyrrolidine + TMA, and TMA + bmp. X-ray reﬁnement of the crystals showed that pyrrolidine is occluded in both the cage and the 10-ring channels and, if pyrrolidine TMA is also present in the synthesis gel, this cation is located in some of the FER cages as well, but not in the 10-ring channel. If a TMA + bmp mixed-template system

123

124

4 Co-Templates in Synthesis of Zeolites

is used, TMA is found in the cages, while bmp is located in the 10-ring channel. It has been shown that the accessibility of a basic probe molecule such as pyridine to the active sites increases with the population of the ferrierite cages by TMA, with the sample containing only pyrrolidine exhibiting the lowest accessibility. Pyridine is too bulky to enter the cage through the eight-ring windows, so these results evidenced that the population of Br¨onsted acid sites within the cage and in positions accessible through the channel varies according to the combination of SDAs used in the synthesis. Catalytic activity of these materials in m-xylene isomerization and n-butene skeletal isomerization showed an increase with the acid sites accessibility. In line with the previous argument for ferrierite zeolite, a similar approach can be used for tailoring the catalytic activity of SAPO materials. In this case, the use of different co-templates is envisaged to control the incorporation mechanisms whereby heteroatoms are incorporated in the network of microporous materials, rather than to control their spatial distribution as in the previous case. Silicon can be incorporated in AlPO4 networks through different substitution mechanisms: SM2, which leads to the formation of isolated Si(OAl)4 , leading to one Br¨onsted site per Si ion with a low acid strength, or SM2 + SM3, giving room to the formation of stronger acid sites associated with silicon islands, but in a lower concentration. There is experimental evidence that different organic SDA molecules direct the incorporation of Si heteroatoms in AlPO4 networks through the different substitution mechanisms, thus altering the acidity and catalytic activity of the SAPO catalysts obtained. For instance, in SAPO-5, triethylamine favors the incorporation through SM2, leading to more abundant but weaker acid sites, while benzylpyrrolidine favors the insertion through the formation of Si islands, with a lower concentration of acid sites but of higher acidity. This different behavior of the SDA molecules is additive, that is, the use of combinations of the two molecules as SDAs leads to SAPO materials with Si distributions, and so with acid and catalytic properties related to the ratio of the two SDAs employed in the synthesis. Therefore, the catalytic activity of these materials in reactions that require different acid strengths can be easily tailored as needed by using combinations of organic molecules that direct the incorporation of heteroatoms through different replacement mechanisms in the desired ratio [69]. Other examples of the inﬂuence of mixed-template systems over the Si distribution in SAPO materials have been reported. The use of mixtures of diethylamine (DEA) and di-iso-propylamine in the synthesis of SAPO-11 led to a better dispersion of Si by favoring the substitution mechanism SM2 as compared with that occurring when both templates are used separately, and they are more active and selective to isomerization products in the hydroisomerization of n-tetradecane [70]. A similar effect has been found by adding a small amount of methylamine as a co-template in the synthesis of SAPO-11 and SAPO-34 (CHA structure) from di-n-propylamine (DPA) and TEAOH containing gels, respectively. The effect in this case is attributed to the fact that methylamine prevents silica polymerization in the gel, favoring the insertion of small Si oligomers in the framework, thus

4.10 Summary and Outlook

increasing the number of negative charges in the framework which are compensated by the incorporation of small amounts of methylammonium cations [71]. In the synthesis of the one-dimensional 12-ring channel SAPO-41, a mixture of two secondary amines with different chain lengths, DPA and DEA, led to materials with smaller crystal size and stronger acidity, where both templates were occluded inside the channels. In this case, the addition of DEA increases the average size of the Si islands. The obtained catalysts were also more active, yet less selective, in the hydroisomerization of n-octane [72]. As a conclusion, we have tried to highlight in this chapter the important inﬂuence that cooperative structure-directing effects between different species present in the synthesis gels, including those between different types of organic molecules and between these and mineralizing agents or water, exerts in molecular sieve science, ranging from the production of complex topological structures to the control of the crystalline growth as well as the tailoring of the catalytic activity of the materials. This synthesis strategy thus opens up new possibilities in the synthesis and applications of zeolite-type materials that would be otherwise unviable in traditional synthesis systems involving only one type of SDA.

4.10 Summary and Outlook

A comprehensive review of the zeolite materials obtained by the combined use of two different organic SDAs in the synthesis gel has revealed that this strategy is most suitable for the efﬁcient crystallization of structures built up by assembling cages and channels of different sizes and shapes. Following this approach, each template is generally accommodated in one speciﬁc site, dictated by geometrical correspondences between the guest template and the host void volume. While the above model on the action mechanism of dual-template systems can be recognized in many cases, progress in the characterization of actual crystals obtained by using either two or even only one template evidenced a more complex picture, yet even more challenging, in which inorganic species other than cations, namely, water molecules and anions such as hydroxyl and ﬂuoride, also cooperate with the organic guest molecules to stabilize the zeolite crystals. All these are but ingredients of a rich ‘‘chemical soup’’ that nurtures the nascent zeolite nuclei, which absorb from the medium whatever elements serve them in their continuous growth. In some cases, these small chemical entities serve to stabilize zeolite cages which are too small to be occupied by organic molecules, contributing in this way not in a small quantity to the overall stability of the structure, while in others they are actually involved in chemical interactions with the organic species to create supramolecular assemblies which behave as the actual SDAs for the nucleation of the corresponding zeolite structure. The additional stabilization energy provided by such chemical cooperativeness is often reﬂected in nucleation and crystal growth-depending properties of the resulting crystals, such as their size and/or shape, both being most relevant for catalysis and adsorption applications.

125

126

4 Co-Templates in Synthesis of Zeolites

The structure co-directing role of different templates, each of which would in principle be located in different zeolite cavities, would also open unexpected perspectives for inﬂuencing and eventually controlling the location of heteroatoms, such as aluminum, for example, in the framework. The assumption here is that each template would affect the location of the aluminum atoms in its vicinity in a nearly independent way; hence, the aluminum sitting could be modiﬁed by using different combinations of template molecules. As long as different templates would be occluded in different cavities, this strategy would result in crystals where the aluminum would also be unevenly distributed among different cavities. Moreover, if those cavities could accommodate more than one type of organic molecule, changes in the chemical nature of the corresponding template would further affect the location of aluminum, for it would be possible in this way to affect the template–aluminum interaction independently in each cavity. This strategy for inﬂuencing the aluminum’s location in the framework has already been proven to be valid for zeolite ferrierite, but it could be extended to many other structures as well, which contain cages and/or channels of different size. Taking all these considerations into account, it can be concluded that co-templating synthesis strategies entail speciﬁc characteristics that make them suitable for developing zeolite materials with new properties. Although this approach can be dated back to the earliest times of the discovery of high-silica zeolites and aluminophosphate materials, when it was used as a valuable tool for the crystallization of new structures, it has however been scarcely applied since then in the exploration of the synthesis of new zeolite topologies. Several interesting structures have been discovered in recent years by using different synthesis strategies, all of which involve however just one type of template. It would not be without interest to revisit such strategies by using co-SDAs that would be able to stabilize small cages or channels that, in combination with the void topologies of actual structures, would eventually render related low-density materials with higher complexity, or even totally new structures. This would however require a careful exploration of the synthesis parameters to overcome the crystallization of the competing cage-like structures that are usually formed by using the small templates alone. This work would beneﬁt by the use of advanced modeling techniques that can help in selecting the most appropriate template combinations. Control of heteroatom sitting would be feasible by a wise choice of SDAs, and this strategy can be extended to a range of zeolite-type materials, from zeolites to metal-containing aluminophosphates, and to heteroatoms ranging from aluminum in zeolites (or other trivalent elements or tetravalent ones like germanium, tin, or titanium) to silicon, cobalt, tin, zinc, or magnesium, to name a few, in aluminophosphates. The chemical phenomena encompassed under the heading of ‘‘co-templating’’ are adding new elements to the already complex ﬁeld of zeolite crystallization, where they contribute to reshape the classical scenario toward a more colorful one, one that would provide new resources for creative chemistry.

References

Acknowledgments

The authors acknowledge the Spanish Ministry of Science and Education for ﬁnancial support (project CTQ2006-06282). A. B. P. and L. G. H. are grateful to the Spanish Ministry of Science and Innovation for a predoctoral grant and a postdoctoral fellowship, respectively. R. G. acknowledges CSIC for the J.A.E. contract. References 1. Barrer, R.M. (1982) Hydrothermal Chem-

2. 3. 4. 5.

6.

7.

8. 9.

10. 11. 12.

13.

14.

15. 16.

istry of Zeolites, Academic Press, New York. Aiello, R. and Barrer, R.M. (1970) J. Chem. Soc. A, 1470. Flanigen, E.M. (1973) Adv. Chem. Ser, 121, 119. Davis, M.E. and Lobo, R.F. (1992) Chem. Mater, 4, 756. P´erez-Pariente, J. and G´omez-Hortig¨uela, L. (2008) Zeolites: From Model Materials to Industrial Catalysts, Chapter 3 Transworld Research Network, pp. 33–62. Corma, A., Navaro, M.T., Rey, F., and Valencia, S. (2001) Chem. Commun, 1486. Corma, A., D´ıaz-Caba˜ nas, M.J., Jorda, J.L., Martinez, C., and Moliner, M. (2006) Nature, 443, 842. Lok, B.M., Cannan, T.R., and Messina, C.A. (1983) Zeolites, 3, 282. Jacobs, P.A. and Martens, J.A. (1987) Synthesis of High-silica Aluminosilicate Zeolites, Elsevier, Amsterdam. Pelrine, B.P. (1981) US Patent 4,259, 306. Dwyer, F.G. and Jenkins E.E. (1981) US Patent 4,287,166. Casci, J.L. Lowe, B.M., and Whittam, T.V. (1981) UK Patent Application GB 2 077 709. Schlenker, J.L., Dwyer, F.G., Jenkins, E.E., Rohrbaugh, W.J., and Kokotailo, G.T. (1981) Nature, 294, 340. Weigel, S.J., Gabriel, J.C., Guti´errez-Puebla, E., Monge-Bravo, A., Henson, N.J., Bull, L.M., and Cheetham, A.K. (1996) J. Am. Chem. Soc, 118, 2427. Ref 9, recipe 3.b, p. 11. Chu, P. (1980) EPA 0023089.

17. Ref 9, p. 22. 18. Schlenker, J.L., Rohrbaugh, W.J., Chu,

19.

20. 21.

22.

23. 24.

25. 26.

27.

28.

29. 30. 31.

P., Valyocsik, E.W., and Kokotailo, G.T. (1985) Zeolites, 5, 355. Kirschhock, C.E.A., Liang, D., Van Tendeloo, G., F´ecant, A., Hastoye, G., Vanbutsele, G., Bats, N., Guillon, E., and Martens, J.A. (2009) Chem. Mater, 21, 371. Seddon, D. and Whittam, T.V. (1981) EPA 55,529. Pinar, A.B., G´omez-Hortig¨uela, L., and P´erez-Pariente, J. (2007) Chem. Mater, 19, 5617. Garc´ıa, R., G´omez-Hortig¨uela, L., D´ıaz, I., Sastre, E., and P´erez-Pariente, J. (2008) Chem. Mater, 20, 1099. Dorset, D.L. and Kennedy, G.J. (2004) J. Phys. Chem. B, 108, 15216. Camblor, M.A., Corma, A., D´ıaz-Caba˜ nas, M.J., and Baerlocher, C. (1998) J. Phys. Chem. B, 102, 44. Zones, S.I., Hwang, S.-J., and Davis, M.E. (2001) Chem. Eur. J, 7 (9), 1990. Burton, A., Accardi, R.J., Lobo, R.F., Falcioni, M., and Deem, M.W. (2000) Chem. Mater, 12, 2936. Millini, R., Carluccio, L.C., Carati, A., Bellussi, G., Perego, C., Cruciani, G., and Zanardi, S. (2004) Microporous Mesoporous Mater, 74, 59. Knight, L.M., Miller, M.A., Koster, S.C., Gatter, M.G., Benin, A.I., Willis, R.R., Lewis, G.J., and Broach, R.W. (2007) Stud. Surf. Sci. Catal, 170, 338. Roth, W.J. (2005) Stud. Surf. Sci. Catal, 158, 19. Zones, S.I. (1987) US Patent 4 665 110. Millini, R., Perego, G., Parker, W.O., Bellussi, G., and Carluccio, L. (1995) Microporous Mesoporous Mater, 4, 221.

127

128

4 Co-Templates in Synthesis of Zeolites 32. Camblor, M.A., Corell, C., Corma, A.,

33.

34. 35.

36.

37.

38.

39.

40.

41. 42.

43.

44.

45.

46.

47.

48. 49.

Diaz-Caba˜ nas, M.J., Nicolopoulos, S., Gonzalez-Calbet, J.M., and Vallet-Regi, M. (1996) Chem. Mater, 8, 2415. Barrett, P.A., Diaz-Caba˜ nas, M.J., and Camblor, M.A. (1999) Chem. Mater, 11 (10), 2919. Blackwell, C.S. et al. (2003) Angew. Chem. Int. Ed, 42, 1737. Jackowski, A., Zones, S.I., Hwang, S.-J., and Burton, A. (2009) J. Am. Chem. Soc, 131, 1092. Lok, P.M., Messina, C.A., Patton, R.L., Gajek, R.T., Cannan, T.R., and Flanigen, E.M. (1984) US Patent 4 440 871. Edwards, G.C., Gilson, P.J., and Mc Daniel, V. (1987) US Patent 4 681 864. de Saldarriaga, L.S., Saldarriaga, C., and Davis, M.E. (1987) J. Am. Chem. Soc, 109, 2686. Schreyeck, L., D’agosto, F., Stumbe, J., Caullet, P., and Mougenel, J.C. (1997) Chem. Commun, 1241. Paillaud, J.-L., Caullet, P., Schreyeck, L., and Marler, B. (2001) Microporous Mesoporous Mater, 42, 177. Wheatley, P.S. and Morris, R.E. (2002) J. Solid State Chem, 167, 267. Sierra, L., Deroche, C., Gies, H., and Guth, J.L. (1994) Microporous Mesoporous Mater, 3, 29. Sierra, L., Patarin, J., Deroche, C., Gies, H., and Guth, J.L. (1994) Stud. Surf. Sci. Catal, 84, 2237. Corma, A., Rey, F., Rius, J., Savater, M.J., and Valencia, S. (2004) Nature, 431, 287. Wright, P.A., Maple, M.J., Slawin, A.M.Z., Patinec, V., Aitken, R.A., Welsh, S., and Cox, P.A. (2000) J. Chem. Soc., Dalton Trans, 8, 1243. Castro, M., Garcia, R., Warrender, S.J., Slawin, A.M.Z., Wright, P.A., Cox, P.A., Fecant, A., Mellot-Draznieks, C., and Bats, N. (2007) Chem. Commun, 3470. Xing, H., Li, J., Yan, W., Chen, P., Jin, Z., You, J., Dai, S., and Xu, R. (2008) Chem. Mater, 20, 4179. Zones, S.I. and Hwang, S.-J. (2002) Chem. Mater, 14 (1), 313. Lee, G.S., Nakagawa, Y., and Zones, S.I. (2000) US Patent 6,156,290.

50. Lee, H., Zones, S.I., and Davis, M.E.

(2003) Nature, 425, 385. 51. Lee, H., Zones, S.I., and Davis, M.E.

(2005) J. Phys. Chem. B, 109, 2187. 52. Lee, H., Zones, S.I., and Davis, M.E.

53.

54.

55.

56.

57.

58.

59. 60.

61.

62.

63.

64.

65. 66.

67.

(2006) Microporous Mesoporous Mater, 88, 266. Arranz, M., Pe’rez-Pariente, J., Wright, P.A., Slawin, A.M.Z., Blasco, T., G´omez-Hortig¨uela, L., and Cora, F. (2005) Chem. Mater, 17, 4374. Arranz, M., Garc´ıa, R., and P´erez-Pariente, J. (2004) Stud. Surf. Sci. Catal, 154, 257. G´omez-Hortig¨uela, L., Cor`a, F., ´ M´arquez-Alvarez, C., and P´erez-Pariente, J. (2008) Chem. Mater, 20, 987. G´omez-Hortig¨uela, L., P´erez-Pariente, J., and Cor`a, F. (2009) Chem. Eur. J, 15, 1478. G´omez-Hortig¨uela, L., L´opez-Arbeloa, F., Cor`a, F., and P´erez-Pariente, J. (2008) J. Am. Chem. Soc, 130, 13274. Fernandes, A., Ribeiro, M.F., Borges, C., Lourenc¸o, J.P., Rocha, J., and Gabelica, Z. (2006) Microporous Mesoporous Mater, 90, 112. Chatelaine, T., Patarin, J., Soulard, M., and Guth, J.L. (1995) Zeolites, 15, 90. Ferchiche, S., Valcheva-Traykova, M., Vaughan, D.E.W., Warzywoda, J., and Sacco, A. Jr. (2001) J. Cryst. Growth, 222, 801. Berger, C., Gl¨aser, R., Rakoczy, R.A., and Weitkamp, J. (2005) Microporous Mesoporous Mater, 83, 333. Kecht, J., Mintova, S., and Bein, T. (2008) Microporous Mesoporous Mater, 116, 258. Yu, H., Wang, X.-Q., and Long, Y.-C. (2006) Microporous Mesoporous Mater, 95, 234. Lai, Z.P., Bonilla, G., D´ıaz, I., Nery, J.G., and Tsapatsis, M. (2003) Science, 300, 456. Carreon, M.A., Li, S., Falconer, J.L., and Noble, R.D. (2008) Adv. Mater, 20, 729. Pinar, A.B., P´erez-Pariente, J., and G´omez-Hortig¨uela, L. (2008) WO2008116958A1. ´ Pinar, A.B., M´arquez-Alvarez, C., Grande-Casas, M., and P´erez-Pariente, J. (2009) J. Catal, 263, 258.

References 68. M´arquez-Alvarez, C., Pinar, A.B.,

Garc´ıa, R., Grande-Casas, M., and P´erez-Pariente, J. (2009) Top. Catal, 52, 1281. ´ 69. G´omez-Hortig¨ uela, L., M´arquez-Alvarez, C., Grande-Casas, M., Garc´ıa, R., and P´erez-Pariente, J. (2009) Microporous Mesoporous Mater, 121, 129.

70. Liu, P., Rien, J., and Sun, Y. (2008)

Microporous Mesoporous Mater, 114, 365. 71. Fernandes, A., Ribeiro, F., Lourenc¸o,

J.P., and Gabelica, Z. (2008) Stud. Surf. Sci. Catal, 174A, 281. 72. Li, L. and Zhang, F. (2007) Stud. Surf. Sci. Catal, 170A, 397.

129

131

5 Morphological Synthesis of Zeolites Sang-Eon Park and Nanzhe Jiang

5.1 Introduction

Zeolites are crystalline inorganic materials with unique, characteristic micropores. These micropores in molecular scale, which have one to three dimensions, are due to interconnected cavities or channels. The types of assembly of cavities or channels control the size of micropores, orientation of microporous channels, and the morphology of zeolite particles. Undoubtedly, the morphology is originally related to the framework type and also closely related to the micropore size, crystal size, and shape, and directly affects the physicochemical properties of zeolites. Zeolites are extensively used in the ﬁelds of heterogeneous catalysis, separations, ion exchange, chemical separation, adsorption, host/guest chemistry, microelectronic devices, optics, and membranes [1–4]. Such applications of zeolites are strongly affected by the pore size, type of channels, and morphologies [5–9]. Morphological synthesis of zeolites is particularly important in catalytic applications where the particle shape can have a dramatic effect on the product distribution due to differences in rates of transport/diffusion and reaction. Recent research interests have also been directed toward the crystallography of single zeolite crystals by applying nanotechnology. Thus, there has been a great interest in developing synthetic approaches to control crystal size and morphology of zeolites [9–11]. Fine-tuning of zeolite crystal size and shape is usually accomplished by systematic variation of the composition of the precursor mixture (including the use of additives such as salts) [12]. But through the development of new technology and equipment, synthetic parameters such as temperature, pressure, and stirring rate or even gravity can be controlled. This chapter provides a brief introduction to the morphological synthesis of zeolites and related fabrication methods such as as microwave. Examples in the preparation of well-shaped single crystals, ﬁne control of common particles, and microwave driven fabrication are discussed.

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

132

5 Morphological Synthesis of Zeolites

5.2 Morphology of Large Zeolite Crystals

The synthesis of large zeolite single crystals is of a great interest for a large number of requirements, including single-crystal structure analysis; ﬁne-structure analysis; study of crystal growth mechanisms; study of adsorption and diffusion; and the determination of anisotropic electrical, magnetic, or optical properties [13–16]. For example, although the high surface areas of small zeolite particles might initially appear advantageous for heterogeneous catalysis, in the case of shape-selective catalysis by zeolites it is well established that a higher selectivity is offered by larger crystallites, because the internal surface area accounts for a much greater proportion of the total zeolite surface area [17]. Zeolites have distinct framework types that can be built in a periodic pattern by linking the basic building units (BBUs), called tetrahedrons. These BBUs can be continuously linked together to form more complex composite building units (CBUs), such as rings, which in turn will further lead to the next level, that is, cages (Figure 5.1). The type of building units or the way to connect them is the intrinsic factor deciding morphology, especially morphology of single crystals. And another important implication is to understand the crystallization of zeolites, that is, the ability to engineer crystal morphology and crystal size. 5.2.1 Large Crystals of Natural Zeolites

In nature, large zeolite crystals are often found in volcanogenic sedimentary rocks, which are believed to have formed via dissolution of volcanic glass [18]. Figure 5.2 and Table 5.1 show some well-known crystalline natural zeolites. Typical morphologies are the rhombic-shaped polyhedron or needle-shaped ﬁbrous zeolites of several centimeters in size. In nature, zeolites may grow under hydrothermal conditions in the earth’s crust, which are termed closed hydrologic conditions [19]. And natural zeolites usually occur in combined form with other silicate minerals 4-ring

5-ring

6-ring

Cancrinite cage

8-ring

Figure 5.1

5.6 Å

7.3 Å

10-ring

12-ring

a-Cage of zeolite A-LTA

Rings and cages frequently found in zeolite structures.

Sodalite cage

Supercage of zeolite X-FAU

5.2 Morphology of Large Zeolite Crystals

(a)

(b)

(d)

(c)

(e)

Figure 5.2 Some natural zeolite crystals: (a) ammonioleucite (Fujioka, Japan); (b) analcime (Quebec, Canada); (c) erionite (Oregon, USA); (d) natrolite (California, USA); and (e) mordenite (Washington, USA) (http://www.iza-online.org).

such as clays and dense forms of silica, and their chemical composition varies from one location to another. 5.2.2 Synthesis of Large Zeolite Crystals

Synthesis of large zeolite single crystals has been achieved by several excellent means [20]. A detailed study of the crystal structure, crystallization mechanism, and relation with morphology has become possible. Therefore, the synthesis of large, single zeolite crystals is highly desirable from the point of view of structural determination or discovering the crystallization mechanism. Although recent advances in X-ray powder diffractometry combined with sensitive spectroscopic methods such as multinuclear NMR methods have made structural determination from polycrystalline powders amenable (indeed most zeolites are structurally characterized by powder data) [20], diffraction data from a single crystal often remain the most accurate means of unambiguous structure elucidation, including the location of both the framework and extraframework atoms in a zeolite [21–28]. Hence it is possible to synthesize submicrometer-sized zeolite crystals [29–33] as well as very large crystals [34–37] by reproducible synthesis. The former relies on the syntheses where nucleation rates are very high, often involving monomeric silica and alumina sources, high alkalinity, and low temperatures. The latter relies on syntheses employing low-solubility silicon sources, ﬂuoride as the mineralizing

133

134

5 Morphological Synthesis of Zeolites Table 5.1

Zeolites

Material data for some nature zeolites (data from http://www.iza-online.org). Composition

Ammonioleucite |NH4 K)|[AlSi2 O6 ]

Framework type ANA

Analcime

|Na (H2 O)|[AlSi2 O6 ]

ANA

Erionite

Ca |K2 (Ca0.5 ,Na)8 (H2 O)30 |[Al10 Si26 O72 ]

ERI

Natrolite

Natrolite |Na2 (H2 O)2 | [Al2 Si3 O10 ]

NAT

Mordenite

|Na2 , Ca,K2 )4 (H2 O)28 | [Al8 Si40 O96 ]

MOR

CBUs

Morphology

6-2 or 6 or 4-[1,1] Tetragonal or 1-4-1 or 4 (SBUs) Isometric or pseudoisometric single crystals are trapezohedra in sizes ranging from millimeters to several centimeters d6r, can Hexagonal, 6/m2/ m2/m single crystals are hexagonal prisms terminated by a pinacoid with sizes under 3 mm, ﬁbrous, and wool-like nat Orthorhombic mm2 single crystals are pseudotetragonal prisms terminated by a pyramid sizes range from a few millimeters to several centimeters mor Orthorhombic mmm or mm2 single crystals are thin ﬁbers, 0.1–10 mm long

agent, and high temperatures. These methods mainly focus on lowering the speed of nucleation or crystallization steps extremely by (i) adding chemical reagents or (ii) applying novel synthetic conditions. Here we will show some typical examples. Various nucleation suppressing agents are applied to prepare large zeolite crystals with well-shaped morphology by extremely slow and controlled nucleation and crystallization steps. The most common LTA and FAU zeolites were synthesized into big crystals (Figure 5.3) by applying a nucleation suppression agent such as tertiary alkanolamine. The morphology of big crystals has the same shape symmetry as the framework type [38]. The unit cell and crystal structure would be understandable easily through such morphology of single crystals. The most interesting agent is the ﬂuoride anion. By applying the ﬂuoride anion as the mineralizing agent instead of hydroxide, large zeolite crystals with the framework types MFI (Figure 5.4), FER, MTT, MTN, and TON have been prepared.

5.2 Morphology of Large Zeolite Crystals (b)

(a)

50 µm

Cubic, Fd 3m

Figure 5.3

100 µm

Cubic, Pm3m

Large single crystals of (a) LTA and (b) FAU [38].

This method has allowed not only the formation of large zeolite crystals but also the growth of crystals with a high silica content and free of defects compared with those prepared by conventional routes [39, 40]. Even though F− has a tendency to form complexes with the initial reactant species such as silicon, these complexes slowly hydrolyze to release less ﬂuorinated silicon, which gradually supplies the nutrients for crystal growth, permitting large crystallites to be formed [41–43]. Under these conditions of low supersaturation, growth is favored at the expense of nucleation and thus a small number of large zeolite crystals are ultimately formed. The shape and size of large crystals are controllable by varying the chemical compositions (Figure 5.4b). Single crystals of Si-MFI (all-silica MFI) zeolite with different crystalline sizes ranging from 9 × 3 × 2 µm to 165 × 30 × 30 µm were obtained by adding benzene-1,2-diol as a complexing agent. Crystals synthesized in the presence of benzene-1,2-diol have a much larger size than those synthesized in its absence, and their size and shape were largely inﬂuenced by the content of benzene-1,2-diol in the reaction system (Figure 5.5). It is interesting to note that the aspect of ratio increased as a function of the content of benzene-1,2-diol, and the crystal size increased along the length rather than the width of the crystal [46]. Very interestingly, through a novel crystallization technique (the bulk-material dissolution (BMD) technique), giant zeolite crystals with size of several millimeters were synthesized by controlling the release and solubility of reactive solution

135

136

5 Morphological Synthesis of Zeolites

Crystal morphology

c

b a

a

x 0 0.5 2.5

400013

20KV

X500

60 µm

(a)

12SiO2 : xB2O3 : 4.5Na2O : 4.5 TPA2O : 2000H2O (b)

Figure 5.4 (a) (B, Al)-MFI (B-ZSM-5) [44]. (b) Modiﬁcation of silicalite morphology with addition of boric acid to conventional hydrothermal synthesis [45].

(a)

(b)

(c)

500001 20KV X2.00K 15.0um 200048 15KV X2.00K 15.0um 400013 20KV X500

68um

Figure 5.5 Crystal size of Si-MFI zeolite synthesized at 180 ◦ C with the molar composition: SiO2 -0.2TPABr-xR-0.5NaOH-30H2 O (R: benzene-1,2-diol) (a) x = 0; (b) x = 0.2; (c) x = 0.4 [46].

species in organothermal systems [47]. Si-MFI zeolite, Analcime (ANA), and JBM crystals with sizes of about 3 mm were successfully synthesized by using bulk materials (quartz tube) as the silica and alumina sources (Figure 5.6). Well-shaped giant crystals with different morphologies obviously prove that they originated from the zeolite framework type. Instead of above novel method, hydrothermal synthesis was intensively investigated for synthesizing single crystals of Al-MFI zeolite (ZSM-5) with varying chemical compositions and conditions to facilitate detailed structural studies of the MFI framework. Single crystals of Al-MFI zeolite were synthesized in systems containing Na+ -tetrapropylammonium (TPA), Li+ -TPA, and NH4 + -TPA. The samples consisted of a fully crystalline and pure zeolitic phase with good homogeneity with crystal sizes up to 420 µm [48]. In the alkaline-free NH4 + -TPA system, homogeneous and pure single crystals of Al-MFI zeolite were prepared of lengths up to 350 µm [49]. The crystal sizes and yields were found to depend on

5.2 Morphology of Large Zeolite Crystals

(a)

(b)

Orthorhombic, Pnma

137

(c)

Cubic, Ia3d

Orthorhombic, Pnma

Figure 5.6 (a) Giant crystals of Si-MFI zeolite; (b) Giant crystals of ANA zeolite; and (c) Giant crystals of JBW zeolite [47].

the water content of the starting reaction mixture and on the type of aluminum source. Large single crystals of Si-MFI zeolite have been synthesized with the choline cation, 1,4-diazabicyclo [2.2.2]octane, and tetramethylammonium (TMA) cations from an alkaline-free medium [50]. At high temperature (300 ◦ C) and high pressure (100 MPa), the structure-directing organic template of TPA+ is stabilized, and hydrolyzed tetraethylortosilicate (TEOS) grow to millimeter-sized big crystals of Si-MFI zeolite [51]. The elevated temperature and pressure favor the formation of crystals with improved quality. Prismatic Si-MFI zeolite crystals with a uniform size of about 0.7 × 0.2 × 0.2 mm have been obtained by heating a gel prepared from TMA-silicate solution, TPABr, and sodium hexaﬂuorosilicate at 250 ◦ C under a pressure of 80 MPa (Figure 5.7). The inﬂuence of synthesis conditions on the crystal size has been studied systematically by changing the temperature, pressure, and gel compositions. Under the speciﬁc conditions of 250 ◦ C and 80 MPa, a strong correlation between the crystal size and the F/Si mole ratio of the starting gel was found, which enabled the preparation of uniform crystals of Si-MFI zeolite with preset dimensions [51]. Generally, crystallization of large, single zeolite crystals can be achieved by controlling the nucleation and crystallization steps. But the various reagents applied would suppress the nucleation and slow down the crystal growth. At such low crystallization rates, it is difﬁcult to separate nucleation and crystallization clearly. So reagents such as F− , tetraethylammonium (TEA), or various amines and alcohols affect both processes. The synthesis conditions were varied by controlling

138

5 Morphological Synthesis of Zeolites (b)

(a)

0.5

1 mm

0

1

2

3 mm

1000 Max. size (µm)

0

800 600 400 200 0 0.0

(c)

0.5

1.0

1.5

F/Si mole ratio

Figure 5.7 (a) Optical micrograph of Si-MFI zeolite crystals synthesized at 300 ◦ C and 100 MPa; (b) Optical micrograph of Si-MFI zeolite crystals synthesized at 250 ◦ C and 80 MPa; (c) Correlation between Si-MFI zeolite crystal size and the reagent F/Si ratio adjusted by using Na2 SiF6 for syntheses at 250 ◦ C and 80 MPa [51].

pH of the synthetic precursor and/or by applying high pressure, temperature, and even gravity. Also, in space perfect single crystals of zeolites A and X have been synthesized [12]. All synthesized single crystals exactly reﬂect their own bulk symmetry from their unit cells in nanoscale.

5.3 Morphology Control of MFI Zeolite Particles (of Size Less than 100 µm)

Depending on the crystal morphology, pore openings of a particular channel system can be present at the crystal surface to different extents. As a consequence, the access to the intracrystalline volume may be facilitated or hindered. The crystal morphology and size deﬁne also the diffusion paths via the channels, which often have a great impact on their applications such as reaction kinetics. The diffusivity of guest molecules in zeolite crystals is closely related to the pore size, but they also depend on the crystal morphology. Especially as shape catalysts, the shape selectivity of zeolites is affected by their morphology (shape and size): for example, large crystals give higher selectivity but their long pathways decrease the efﬁciency.

5.3 Morphology Control of MFI Zeolite Particles (of Size Less than 100 µm)

So the selectivity and effectiveness need to be compromised by optimization of the crystal size and shape [37, 52]. Even though more than 130 different zeolites have been discovered so far, only a few are available as industrial catalysts. The crystal morphology of the materials possessing a three-dimensional channel system, for example, FAU- and LTA-type zeolites, is not expected to have a great impact on their properties. In contrast, the performance of materials with mono- or bidimensional channel systems might be strongly affected by the morphology of the crystals. Among them, we prefer to introduce morphological synthesis of zeolites having MFI-type framework. MFI-type zeolite possesses an anisotropic framework with two intersecting 10-ring channels. The straight channels are parallel to b axis and the zigzag channels with an estimated pore opening of 0.51 nm × 0.55 nm are parallel to the a axis (Figure 5.8). The b channels and a channels are interconnected with each other, so diffusion along the c direction is also possible. It has been known that MFI-type zeolites with various crystal shapes can be prepared, such as spherical, hexagonal twined disk, rodlike, and so on [53–68]. In the synthesis of Al-MFI, the various factors inﬂuencing the dimensions along each axis of the crystal have been investigated systematically [69]. The dependence of metal cations, structure-directing agents, chemical source, and compositions has been summarized by Singh and Dutta [70]. 5.3.1 Dependence of Structure-Directing Agents (SDAs)

The typical structure-directing agent (SDA) for MFI-type zeolite is the TPA cation. Instead of TPA, synthesis of MFI zeolite in the presence of dC6 (Figure 5.9) has been reported in several studies [72, 73]. The characteristic crystal shape of TPA-Si-MFI

0.54 × 0.56 nm

b 0.51 × 0.55 nm

a

[h0h]

c Figure 5.8

Pore structure of the AL-MFI [71].

139

140

5 Morphological Synthesis of Zeolites

c

c TPA

a (a)

b 5 µm dC7 N+

N+

(b)

1 µm dC6 N+

N+

(c)

5 µm

tC6 (d)

1 µm

N+

N+

N+

Figure 5.9 SEM images of Si-MFI: (a) pill- or cofﬁn-shaped crystals using TPA; (b) octagonal shaped crystals with twin intergrowths using dC7; (c) leaf-shaped crystals from dC6; (d) b-elongated leaf-shaped (or platelike) crystals from tC6 [74].

is hexagonal prismatic, more commonly referred to as a cofﬁn shaped, with the order of crystal dimensions Lc > La > Lb (where Li indicates crystal size along i axis). Tsapatsis et al. controlled the order of crystal dimensions to Lc > La = Lb with dimer of TPA (dC6, Figure 5.9) and to Lc > Lb > La with trimer tC6 [74]. The morphological changes for Si-MFI zeolites with amine additives and TPABr have been reported [75]. The TPABr-containing crystals are rather elongated (a × b × c = 80 × 40 × 20 µm3 ), whereas the crystals containing TPA and DPA were smaller (30 × 25 × 20 µm3 and 6 × 5 × 4 µm3 , respectively) and isometric in shape (Figure 5.10). Al-MFI crystals have been synthesized into cubic crystals in pyrrolidine-containing hydrous gels with uneven size and the diameter ranging from 0.5 to 4 µm [76]. By varying the TPABr content, Si-MFIs have been grown in a rodlike shape. Because fewer nuclei are formed at lower TPABr concentrations, the volume of the individual crystallites was inversely proportional to the initial TPABr concentration [77]. Si/TPA ratios have varied with values of 10, 24, and 48 [78]. With a ratio of 10, tablet-shaped crystals were formed with knobs at the top and bottom; for a ratio of 24, the crystals had a similar shape with sharp corners and were signiﬁcantly larger. The larger size was a reﬂection of lower TPA content and reduced rate of

5.3 Morphology Control of MFI Zeolite Particles (of Size Less than 100 µm)

(a)

(b)

25 µm

141

(c)

25 µm

Figure 5.10 SEM images of MFI-type zeolites prepared with structure directing agents: (a) TPABr; (b) tripropyl amine; and (c) dipropyl amine.

nucleation. With a Si/TPA ratio of 48, the size and shape remained the same as in 24, but there appeared to be a solid phase growing on the surface of the crystals. 5.3.2 Dependence on Alkali-Metal Cations

The morphology of AL-MFI was found to be dependent on the presence of alkali-metal ions [79]. Li and Na zeolites consisted of spheroidal 2–5 and 8−15 µm crystal aggregates of very small platelet-like units, respectively. K, Rb, and Cs zeolites consisted of twins of rounded (K, Rb) or sharp-edged crystals (Cs). (NH4 )-Al-MFI consisted of large lath-shaped, well-developed, and double-terminated single crystals. The (Li, Na)-, Na-, and (Na, K)-Al-MFI zeolites have spherical or egg-shaped polycrystallites [80], and similar morphology of the Al-MFI zeolites for Na and K have been observed [81]. Morphology of Al-MFI synthesized from Na, K-TPA depended on the relative ratio of the alkali-metal cations. With both Na and K cations present at a ratio of K/(K + Na) = 0.75, large crystal aggregates were obtained in the range of 5−10 µm [82, 83]. Here, some of the batch compositions were studied, that is, xNa2 O/8TPABr/ 100SiO2 /1000H2 O and xTPA2 O/(8 – 2x)TPABr/100SiO2 /1000H2 O, where x varies from 0.5 to 4.0. As the alkalinity of the reaction mixture was reduced from x = 4 to x = 0.5, the aspect ratio (length/width) of the crystals increased from 0.9 to 6.7. Both nucleation and crystallization occurred more rapidly in the presence of Na+ . Synthesis of Al-MFI in glycerol solvent has been reported, and the morphology of the crystals was found to be hexagonal columns [84]. Addition of Li2 O in the synthesis of zeolite TPA-Al-MFI with (NH4 )2 O/Al2 O3 = 38 produces unusually uniform, large, lath-shaped crystals of Al-MFI about 140 ± 10 µm in length [85].

1 µm

142

5 Morphological Synthesis of Zeolites

5.4 Morphological Synthesis by MW

Zeolites can be effectively and rapidly synthesized by using microwave heating techniques [86]. The advantages in the microwave synthesis of zeolites are homogeneous nucleation, fast synthesis by rapid heat-up time, selective activation of the reaction mixture by microwaves, phase selective synthesis by ﬁne-tuning of synthesis conditions, uniform particle, size-facile morphology control, fabrication of small crystallites and enhancement of crystallinity, and so on [87]. 5.4.1 Examples of MW Dependency

AFI-type molecular sieves Aluminophosphate ﬁve such as AlPO-5 and SAPO-5 have one-dimensional channels with micropores (0.73 nm) and they have been synthesized with various morphologies under microwave irradiation [88–91]. The morphologies were controlled by varying the reaction conditions and addition of extra components such as ﬂuoride and silica (Figure 5.11). Rodlike crystals (with aspect ratio of about 40) were obtained with the addition of ﬂuoride and increase of template and water concentrations. The platelike crystals (with an aspect ratio of about 0.2) were synthesized in an alkaline condition with the addition of an appropriate concentration of silica sol. In this case, it was supposed that silica might hinder the crystal growth in the c direction and the ﬂuoride ion might retard the nucleation rate. Most recently, Xu et al. controlled the morphology Si-MFI crystals (Figure 5.12) from a microwave-assisted solvothermal synthesis system in the presence of diols, that is, ethylene glycol (EG), diethylene glycol (DEG), triethylene glycol (TEG), and tetraethylene glycol (tEG) [92]. Under microwave radiation, the Si-MFI crystals with tunable sizes, shapes, and aspect ratios were crystallized.

(a)

(b)

5 µm

Acc.V Spot Magn Det WD 10.0 kV 3.0 5000× SE 10.1

5 µm

20 µm Acc.V Spot Magn Det WD 10.0 kV 3.0 1000× SE 4.9

Figure 5.11 SEM images of typical AFI molecular sieves: (a) platelike crystal and (b) rodlike crystal. White scale bar corresponds to 10 µm [87].

20 µm

5.4 Morphological Synthesis by MW

143

c

a

b

General

SEI

5.0kV ×18,000 1 µm WD 5.8mm

(a)

General

SEI

5.0kV ×19,000

1 µm WD 5.9mm

(b)

L

W

T General

SEI

5.0kV ×18,000

1 µm

WD 5.8mm

(c)

General

SEI

5.0kV

×18,000 1 µm WD 5.8mm

(d)

Figure 5.12 The SEM images of the Si-MFI crystals crystallized from the microwave-assisted solvothermal synthesis system in the presence of the diols: (a) EG, (b) DEG, (c) TEG, and (d) tEG. A schematic

identifying the crystal faces is shown on the right part of the ﬁgure. Gel composition (in molar ratio), SiO : TPAOH : EtOH : diols : H2 O = 1 : 0.357 : 4.0 : 7 : 21.55 [92].

5.4.2 Morphological Fabrication by MW

Fabrication of nanostructured zeolites has attracted much attention in order to (i) optimize zeolite performance (no pore blocking and zeolite diluting binding additives are present), easy handling, and attrition resistance; (ii) minimize diffusion limitations with the secondary larger pores; and (iii) apply to nonconventional applications, such as guest encapsulation, bioseparation, enzyme immobilization, and so on [33]. By applying various templates or nanotechniques, zeolites could be fabricated into membranes and ﬁlms, biomimic or hierarchical structures, and micro-/mesoporous materials [33]. Among them, fabrication of nanoporous materials using a chemical glue can be used for implementing nanoscopic or microscopic arrays of these materials. So far, there have been a few reports on the utilization of chemical glues such as inorganic glue [93], nano-glue [94], and organic covalent linkers [95]. Recently, we reported microwave fabrication of zeolites directly from the synthetic solution and proposed the incorporation of a transition metal as nano-glue [96].

144

5 Morphological Synthesis of Zeolites

In the microwave synthesis of Ti-MFI zeolite (TS-1), the surface titanol groups behave as inorganic glue and will stack Ti-MFI crystals to a ﬁbrous morphology. This technique could be expanded to the fabrication of zeolite ﬁlms and zeolite coatings [96]. Pure and metal-incorporated MFI crystals were synthesized by microwave heating. These samples will be denoted as M-MFI-MW, where M stands for the incorporated metal (Ti, Fe, Zr, and Sn) and MW for the microwave condition; metal-free MFI by the microwave synthesis will be denoted as Si-MFI-MW [96]. The microwave induces a dramatic change in the morphology depending on the composition. Si-MFI-MW and Ti-MFI-CH show the characteristic hockey-puck-like crystals of submicrometer sizes with well-developed, large (010) faces (Figure 5.13a and b). The microwave syntheses of metal (Ti, Sn) incorporated systems produced similar primary crystals, but in this case the crystals were all stacked on top of each other along their (010) direction (b axis) to form a wormlike or ﬁbrous morphology (Figure 5.13c and d). This stacking is sufﬁciently robust so as not to be destroyed by a sonication treatment for more than 1 h, indicating that this morphology is not a result of simple aggregation but of strong chemical bonding between the crystals. This ﬁbrous morphology persists as long as there is incorporated Ti, whose concentration is kept in the range of Si/Ti = 70−230. When the concentration of Ti is increased (Si/Ti ≤ 50), the product is composed of isolated ellipsoidal crystals (a)

(b)

1 µm

(c)

1 µm

(d)

2 µm

Figure 5.13 SEM images of (a) Si-MFI-MW, (b) Ti-MFI-CH (Si/Ti = 70), (c) Ti-MFI-MW(Si/Ti = 70), and (d) Sn-MFI-MW(Si/Sn = 70) [96].

5 µm

5.4 Morphological Synthesis by MW

(not shown). This is probably because the high concentration of Ti interferes with the crystal growth mechanism and the lack of ﬂat surfaces does not allow crystal stacking. The crystals with other incorporated metals (Fe, Zr, and Sn) also show the ﬁbrous morphology. High-resolution TEM image and ED patterns were used to observe the closely connected boundary between crystals (Figure 5.14). The connected parts have well-crystallized structures. We assume that these crystallized parts are mostly connected with the straight channels of each other except some edge parts, which allows the formation of mesopores. Although it is not clear how the incorporated metals induce the stacking of crystals under microwave conditions, it appears to be related to the magnitude of the local dipole moment of the M–O bonds that might be originated from the differences in the electronegativities. Electrically insulating materials absorb microwave energy through the oscillation of dipoles, and the magnitude of absorption increases with the increase of the dipole moment. The magnitude of a dipole moment is mainly determined by the difference in the electronegativities (χ) of the two bonded atoms. Therefore, the Ti–O bond (χ = 2.18 according to the Allred–Rochow scheme) [91] is a better microwave absorber than the Si–O bond (χ = 1.76). The Ti–O bonds on the surface are strongly activated by microwave absorption and can undergo condensation reactions to form Ti–O–Ti and/or Ti–O–Si bonds between crystals. The same explanation applies to the Fe-MFI-MW and Zr-MFI-MW zeolites because of the large χ values for Fe–O and Zr–O bonds. In the case of the Sn-MFI-MW, the χ value of the Sn–O bond (1.78) is rather small, close to that of Si–O bond. However, because Sn is large in size, the valence electron density of the Sn–O bond is shifted to the O side, making this bond more polar than estimated by the χ value alone, namely, the homopolar contribution to the dipole moment [96], and the above explanation of microwave absorption by polar bonds can be applied to the Sn-MFI case. Further work is needed to fully understand the role of the incorporated metals in the stacking of crystals under the microwave condition. 1 2 1

5 6

2

3 4 1 µm

6

Figure 5.14

5

4

3

HRTEM images and ED patterns of Ti-MFI-MW [87].

145

146

5 Morphological Synthesis of Zeolites

(a) 1 µm

(b)

(d)

14 (c)

12

1 µm

(b)

10.0 µm

10.0 µm

(c) 1 µm

(d)

Stacking layer

10 8

(a)

6 4 2

1 µm

0 360 10.0 µm

10.0 µm

480

600

720

Microwave power (W)

Figure 5.15 SEM images and average number of stacking layers of stacked Ti-MFI-MW synthesized under different microwave powers: (a) 360 W, (b) 480 W, (c) 600 W, and (d) 720 W [87].

Microwave power was controlled and varied for ﬁguring out the formation of nanostacked Ti-MFI zeolite [97]. FE-SEM images of Ti-MFI zeolites prepared at different powers are given in Figure 5.15. All the samples show stacked morphologies. The crystals are all stacked on top of each other along their (010) direction to form a wormlike or ﬁbrous morphology. And the average number of stacking layers increased from 7 to 13 as the microwave power was increased from 360 to 720 W. From the above discussion, we see that microwave can strongly affect the nano stacking process. This is because a higher power will give more condensation for the dehydration between hydroxyl groups on the crystal surface. 5.4.3 Formation Scheme of Stacked Morphology

In this study, the formation of the stacked morphology was observed more clearly through the SEM images of Ti-MFI depending on different microwave irradiation times [96]. At the ﬁrst stage of the microwave irradiation (Figure 5.16a–c), small zeolite seeds grew up to uniform crystals; after 40 min they started to form the stacked morphology and the number of stacks kept increasing till 60 min (Figure 5.16d–f). In the microwave synthesis, the silica precursor would be crystallized to uniform, hockey-puck-shaped morphology during the ﬁrst synthesis step. Under prolonged irradiation, those small crystals adhered together along the b orientation to form a stacked morphology. The surface Ti-OH groups seemed to be activated by microwaves and accelerate the condensation reaction between the OH groups on the surface of the crystals (Figure 5.17). MW irradiation induces a three-dimensional stacking of zeolite particles with opal-like morphology through bimetal incorporation [98]. Microwave synthesis of Al- and Ti-bimetal incorporated MFI zeolite ((Al, Ti)-MFI) gives both ﬁbrous and arrayed morphologies. Uniform zeolite crystals were stacked to form long lines and

5.4 Morphological Synthesis by MW

(a)

(b)

(c)

(d)

(e)

(f)

Figure 5.16 SEM of Ti-MFI with different microwave irradiation times: (a) 10 min, (b) 20 min, (c) 30 min, (d) 40 min, (e) 50 min, and (f) 60 min [87].

MW

165 °C 15 min

165 °C 30 min

Nano-sol of TS-1 Hockey puck shape

165 °C 90 min

165 °C 60 min

Ti

Si

Ti

Si

OH OH OH OH OH OH OH OH Nanoglue Si Ti Si Ti

Figure 5.17

Formation scheme of stacked Ti-MFI [87].

those lines were arrayed to form the three-dimensional opal structure (Figure 5.18). The submicrometer-sized pucklike crystals also were stacked face to face of the (010) plane (Figure 5.18c). Microwave synthesis of bimetal-incorporated systems produced the primary crystals, which were stacked on top of each other along their (010) direction to form long lines (Figure 5.18a–c). The void spaces in the nanoarrayed materials were observed through the SEM image of carbon replicas (Figure 5.18d). Xu and coworkers synthesized silicalite-1 (Si-MFI) crystals by applying microwave-assisted solvothermal heating (Figure 5.19). Even without the metal as a nano-glue, the zeolite particles could be stacked into ﬁbrous morphology

147

148

5 Morphological Synthesis of Zeolites

(a)

(b)

104674

15.0 kV

× 3.00 K 6.00 µm

104673

(c)

× 7.00 K 2.57 µm

15.0 kV

(d)

104672 15.0 kV

× 7.00 K 2.57 µm

602926

× 30.0 K 600 nm

15.0 kV

Figure 5.18 SEM images of nanoarrayed (Al, Ti)-MFI: (a) cross view, (b) top, (c) side, and (d) carbon replica.

(a)

General

(b)

SEI

5.0 kV

×7.000

1 µm

WD 7.9mm

(c)

General

General

SEI

5.0 kV

×18.000

1 µm

WD 6.0mm

(d)

SEI

5.0 kV

×30.000

100nm

WD 2.8mm

General

(f)

(e)

General

SEI

5.0 kV

×25.000

1 µm

WD 2.8mm

General

SEI

5.0 kV

(g)

SEI

5.0 kV

×33.000

100nm

WD 2.8mm

General

SEI

Figure 5.19 SEM images of Si-MFI crystals crystallized using different alcohol cosolvents under microwave radiation conditions: (a) ethylene glycol (x = 37), (b) methanol (x = 32.6), (c) ethanol (x = 24.3), (d) 1-propanol (x = 20.1), (e) isopropanol (x = 18.3), (f) n-butanol (x = 17.8), and (g) hexanol (x = 13.3). x is the dielectric constant [99].

5.0 kV

×33.000 100nm

WD 5.9mm

×19.000

1 µm

WD 7.9mm

5.5 Summary and Outlook

by controlling the dipolar cosolvents [99]. The low polarity (dielectric constant) of the cosolvent may favor the formation of abundant Si-OH groups in the precursor species and, on the surface of the nanocrystals formed at the early stage of the crystallization, might undergo further condensation to form self-stacked crystals due to the rapid crystallization under microwave conditions. The crystals are stacked on top of each other along their b direction to form a self-stacked morphology. The ‘‘ﬁber’’ cannot be destroyed even by long-time and strong ultrasonication, which indicates that strong chemical bonds might exist between the individual crystals.

5.5 Summary and Outlook

The representative techniques for the growth of large single crystals of zeolites are reviewed, especially for typical zeolites such as FAU, LTA, or MFI. To prepare perfect zeolite particles, we should understand the mechanism of nuclei formation and the crystal growth process properly in order to critically control the process. Morphology of large zeolite crystals reﬂects the expanded shape of their building units. For ﬁne-tuning the growth, the morphology of small zeolite particles was studied properly. Various additives, for example, alkali cations, alcohols, and amines, were investigated. Different alkali-metal cations result in distinct morphologies; a mixed cationic system will provide uniform, large, lath-shaped crystals (about 140 µm). The use of different structure-directing molecules changes the crystal morphology, as noted with the results in the case of various amines. Hydroxide ion content can alter the morphology signiﬁcantly. Even for a particular ion, such as TPA, the amount used can alter the morphology. The crystallites tend to be larger at lower template concentrations, presumably because of the formation of fewer nuclei. Morphologies of crystals from mixed solvents or nonaqueous media are distinct from comparable compositions in an aqueous medium. Microwaves can control the morphology of zeolites by utilizing the precursor solutions including lossy components such as metals and various organic and inorganic additives. They have played as the absorption sites of microwave energy, which were termed nano-glues, for the fabrication of oriented ﬁber morphology. Such preferred orientations were found to be helpful in the selective adsorption, transportation, or diffusion of longer molecules and could provide advanced shape-selective catalysis. Morphological synthesis of zeolite crystals is still an expanding area, and the preparation of high-quality zeolite crystals will continue to be of great importance. In order to maximize the performance of zeolite catalysts, it is important to understand the crystallographic organization within zeolite crystallites, particularly regarding access to the variously structured pores. Special attention is paid to the study of well-shaped zeolite crystals by utilizing in situ microspectroscopic techniques [100]. Unique insight into diffusion, intergrowth structure, and catalysis can be achieved by applying newly developed technology. It is the way to understand and design

149

150

5 Morphological Synthesis of Zeolites

appropriate zeolite catalysts for a speciﬁc reaction or study the speciﬁc reaction mechanism over zeolite catalysts [101]. Additionally, the catalytic performance of zeolites often depends on the morphology of primary zeolite particles, especially in the case of anisotropic zeolite crystals, where the required property is characteristic for a particular crystallographic direction of the zeolites. The crystal morphology and size deﬁne also the diffusion paths via the channels, which often have a great impact on the reaction kinetics [102]. Thus, the close control of zeolite crystal morphology is highly desirable for zeolite catalysis. Moreover, the morphological control of a given zeolite structure is desirable for the fabrication of zeolites into oriented membranes, sensor, or electric devices, especially to the hierarchical nanoporous structured materials. Finally, morphological control is one of the important keys to understand the mechanism of zeolite crystallization and their performance in various applications. Finding new and simple methods to control morphology is one of the challenges in the ﬁeld of academic study or industrial application of zoelites. Such studies make the future of morphological synthesis of zeolites crystals bright.

Acknowledgments

This work was supported by the Korea Science and Engineering Foundation grants (National Research Laboratory Program), BK21 and Nano Center for Fine Chemicals Fusion Technology.

References 1. Davis, M.E. (1991) Ind. Eng. Chem. 2. 3. 4. 5. 6.

7.

8.

Res., 30, 1675. Davis, M.E. (2002) Nature, 417, 813. Corma, A. (1995) Chem. Rev., 95, 559. Yu, J. and Xu, R. (2006) Chem. Soc. Rev., 35, 593. Csicsery, S.M. (1984) Zeolites, 4, 202. Weitkamp, J. and Puppe, L. (1999) Catalysis and Zeolites: Fundamentals and Applications, 1st edn, Springer, Berlin. Wojciechowski, B.W. and Corma, A. (1986) Catalytic Cracking: Catalysis, Chemistry, and Kinetics, Dekker, New York. Lai, Z., Bonilla, G., Diaz, I., Nery, J.G., Sujaoti, K., Amat, M.A., Kokkoli, E., Terasaki, O., Thompson, R.W., Tsapatsis, M., and Vlachos, D.G. (2003) Science, 300, 456.

9. Kuperman, A., Nadimi, S., Oliver, S.,

10. 11.

12.

13.

Ozin, G.A., Garces, J.M., and Olken, M.M. (1993) Nature, 365, 239. Feng, S. and Bein, T. (1994) Science, 265, 1839. Shi, F., Chen, X., Wang, L., Niu, J., Yu, J., Wang, Z., and Zhang, X. (2005) Chem. Mater., 17, 6177. (a) Singh, R. and Dutta, P.K. (2003) in Handbook of Zeolite Science and Technology (eds S.M.Auerbach, K.A. Carrado, and P.K. Dutta), Marcel Dekker, New York, p. 21; (b) Cundy, C.S. (2005) in Zeolites and Ordered Mesoporous Materials: Progress and Prospects, Studies in Surface Science and Catalysis, Vol. ˇ 157 (eds J. Cejka and H. van Bekkum), Elsevier, Prague, The 1st FEZA School on Zeolites, p. 101. (a) Beschmann, K., Kokotailo, G.T., and Reikert, L. (1988) Stud. Surf. Sci. Catal.,

References

14.

15.

16.

17.

18. 19.

20.

21. 22.

23.

24.

25.

26.

39, 355; (b) Muller, U. and Unger, K.K. (1988) Stud. Surf. Sci. Catal., 39, 101. Terasaki, O., Yamazaki, K., Thomas, J.M., Ohsuna, T., Watanabe, D., Saunders, J.V., and Barry, J.C. (1987) Nature, 330, 58. Cox, S.D., Gier, T.E., Stucky, G.D., and Bierleein, J. (1988) J. Am. Chem. Soc., 110, 2987. Qiu, S., Yu, J., Zhu, G., Terasaki, O., Nozue, Y., Pang, W., and Xu, R. (1998) Microporous Mesoporous Mater., 21, 245. (a) DiRenzo, F. (1998) Catal. Today, 41, 37; (b) Coker, E.N., Jansen, J.C., DiRenzo, F., Fajula, F., Martens, J.A., Jacobs, P.A., and Sacco, A. (2001) Microporous Mesoporous Mater., 46, 223. Mumpton, F.A. (1991) Proc. Natl. Acad. Sci. U.S.A., 96, 3463. Langella, A., Cappelletti, P., and de Gennaro, M. (2001) in Hydrologic Natural Zeolite Growth, Reviews in Mineralogy and Geology 45 (eds D.L.Bish and D.W. Wing), Geoscienceword p. 235. Lethbridge, Z.A.D., Williams, J.J., Walton, R.I., Evans, K.E., and Smith, C.W. (2005) Microporous Mesoporous Mater., 79, 339. St¨ocker, M. (1993) Acta Chem. Scand., 47, 935. Terskikh, V.V., Moudrakovski, I.L., Du, H.B., Ratcliffe, C.I., and Ripmeester, J.A. (2001) J. Am. Chem. Soc., 123, 10399. Binder, G., Scandella, L., Kritzenberger, J., Gobrecht, F., Koegler, J.H., and Prins, R. (1997) J. Phys. Chem. B, 101, 483. Megelski, S., Lieb, A., Pauchard, M., Drechsler, A., Glaus, S., Debus, C., Meixner, A.J., and Calzaferri, G. (2001) J. Phys. Chem. B, 105, 25. Jackson, K.T. and Howe, R.F. (1994) Zeolites and Microporous Crystals, Studies in Surface Science and Catalysis, Vol. 83 (eds H.T.Hattori and T. Yashima), Elsevier p. 187. Heink, W., K¨arger, J., Pfeifer, H., Salverda, P., Datema, K.P., and Nowak,

27.

28.

29.

30. 31.

32. 33.

34. 35.

36. 37. 38.

39.

40. 41. 42.

43.

44.

A. (1992) J. Chem. Soc., Faraday Trans., 88, 515. Snurr, R.Q., Hagen, A., Ernst, H., Schwarz, H.B., Ernst, S., Weitkamp, J., and K¨arger, J. (1996) J. Catal., 163, 130. Schwarz, H.B., Ernst, S., K¨arger, J., Knorr, B., Seiffert, G., Snurr, R.Q., Staudte, B., and Weitkamp, J. (1997) J. Catal., 167, 248. Karge, H.G. and Weitkamp, J. (1998) Zeolite Synthesis, Vol. 1, Springer-Verlag, Berlin. Cundy, C.S., Lowe, B.M., and Sinclair, D.M. (1990) J. Cryst. Growth, 100, 189. Persson, A.E., Schoeman, B.J., Sterte, J., and Ottesstedt, J.E. (1994) Zeolites, 14, 557. Mintova, S., Olson, N.H., Valtchev, V., and Bein, T. (1999) Science, 283, 958. Tosheva, L. and Valtchev, V.P. (2005) Chem. Mater., 17, 2494, and references therein. Charnell, J.F. (1971) J. Cryst. Growth, 8, 291. Sun, Y.Y., Song, T., Qiu, S., Pang, W., Shen, J., Jiang, D., and Yue, Y. (1995) Zeolites, 15, 745. Shimizu, S. and Hamada, H. (2001) Microporous Mesoporous Mater., 48, 39. DiRenzo, F. (1998) Catal. Today, 41, 37. Qiu, S., Yu, J., Zhu, G., Tarasaki, O., Nozue, Y., Pang, W., and Xu, R. (1998) Microporous Mesoporous Mater., 21, 245. Camblor, M.A., Villaescusa, L.A., and Diaz-Cabanas, M. (1999) Top. Catal, 9, 59. Axon, S.A. and Klinowski, J. (1992) Appl. Catal. A, 81, 27. Qiu, S., Pang, W., and Xu, R. (1997) Stud. Surf. Sci. Catal., 105, 301. Qui, S., Yu, J., Zhu, G., Terasaki, O., Nozue, Y., Pang, W., and Xu, R. (1998) Microporous Mesoporous Mater., 21, 245. Guth, J.L., Kessler, H., Higel, J.M., Lamblin, J.M., Patarin, J., Seive, A., Chezeau, J.M., and Wey, R. (1989) ACS Symp. Ser., 398, 176. Qiu, S., Pang, W., and Yao, S. (1989) Stud. Surf. Sci. Catal., 49, 133.

151

152

5 Morphological Synthesis of Zeolites 45. Jansen, J.C., Engelen, C.W.R., and van

46.

47. 48. 49. 50.

51. 52.

53. 54. 55.

56.

57. 58.

59.

60. 61.

62.

63.

64.

Beckkum, H. (1989) ACS Symp. Ser., 398, 257. Shao, C., Li, X., Qiu, S., Xiao, F.-S., and Terasaki, O. (2000) Microporous Mesoporous Mater., 39, 117. Shimizu, S. and Hamada, H. (1999) Angew. Chem. Int. Ed. Engl., 38, 2725. Kornatowski, J. (1988) Zeolites, 8, 77. Mueller, U. and Unger, K.K. (1988) Zeolites, 8, 154. Zhang, D., Qiu, S., and Pang, W. (1990) J. Chem. Soc. Chem. Commun., 1313. Wang, X. and Jacobson, A.J. (2001) Mat. Res. Soc. Symp., 658, GG8.1. Coker, E.N., Jansen, J.C., DiRenzo, F., Fajula, F., Martens, J.A., Jacobs, P.A., and Sacco, A. (2001) Microporous Mesoporous Mater., 46, 223. Ghamami, M. and Sand, L.B. (1983) Zeolites, 3, 155. Tuel, A. (1997) Stud. Surf. Sci. Catal., 105, 261. Iwasaki, A., Sano, T., and Kiyozumi, Y. (1998) Microporous Mesoporous Mater., 25, 119. Persson, A.E., Schoeman, B.J., Sterte, J., and Otterstedt, J.-E. (1994) Zeolites, 14, 557. Kalipcilar, H. and Culfaz, A. (2000) Cryst. Res. Technol., 35, 933. Cizmek, A., Subotica, B., Kralj, D., Babil-Ivancic, V., and Tonejc, A. (1997) Microporous Mesoporous Mater., 12, 267. Gao, F., Zhu, G., Li, X., Li, B., Terasaki, O., and Qiu, S. (2001) J. Phys. Chem. B, 105, 12704. Franklin, K.R. and Lowe, B.M. (1988) Zeolites, 8, 501. Burchart, E.V., Janse, J.C., van der Graaf, B., and van Bekkum, H. (1993) Zeolites, 13, 216. Beck, L.W. and Davis, M.E. (1998) Microporous Mesoporous Mater., 22, 107. de Moor, P.-P.E.A., Beelen, T.P.M., van Santen, R.A., Beck, L.W., and Davis, M.E. (2000) J. Phys. Chem. B, 104, 7600. Hussein, M.Z., Zainal, Z., and Masdan, S.A. (2001) Res. J. Chem. Environ., 5, 21.

65. Dwyer, J. and Zhao, J. (1992) J. Mater.

Chem., 2, 235. 66. Ke, J.-A. and Wang, I. (2001) Mater.

Chem. Phys., 68, 157. 67. Aiello, R., Crea, F., Nigro, E., Testa, F.,

68. 69.

70.

71. 72.

73.

74.

75.

76.

77. 78.

79. 80. 81. 82. 83. 84.

Mostowicz, R., Fonseca, A., and Nagy, J.B. (1999) Microporous Mesoporous Mater., 28, 241. Derouane, E.G. and Gabelica, Z. (1986) J. Solid State Chem., 64, 296. Ban, T., Mitaku, H., Suzuki, C., Matsuba, J., Ohya, Y., and Takahashi, Y. (2005) J. Cryst. Growth, 274, 594 –602. Auerbach, S.M., Carrado, K.A., and Dutta, P.K. (2003) Handbook of Zeolite Science and Technology, Marcel Dekker Inc., New York, p. 43. Lai, Z., Tsapatsis, M., and Nicolich, J.P. (2004) Adv. Funct. Mater., 14, 716. Beck, L.W. and Davis, M.E. (1998) Microporous Mesoporous Mater., 22, 107. de Moor, P., Beelen, T.P.M., van Santen, R.A., Beck, L.W., and Davis, M.E.J. (2000) Phys. Chem. B, 104, 7600. Bonilla, G., Diaz, I., Tsapatsis, M., Jeong, H.K., Lee, Y., and Vlachos, D.G. (2004) Chem. Mater., 16, 5697. Patarin, J., Soulard, M., Kessler, H., Guth, J.-L., and Baron, J. (1989) Zeolites, 9, 397. Suzuki, K., Kiyozumi, Y., Shin, S., Fujisawa, K., Watanabe, H., Saito, K., and Noguchi, K. (1986) Zeolites, 6, 290. Crea, F., Nastro, A., Nagy, J.B., and Aiello, R. (1988) Zeolites, 8, 262. Ahmed, S., El-Faer, M.Z., Abdillahi, M.M., Siddiqui, M.A.B., and Barri, S.A.I. (1996) Zeolites, 17, 373. Gabelica, Z., Blom, N., and Derouane, E.G. (1983) Appl. Catal., 5, 227. Crea, F., Aiello, R., Nastro, A., and Nagy, J.B. (1991) Zeolites, 11, 521. Aiello, R., Crea, F., Nastro, A., and Pellegrino, C. (1987) Zeolites, 7, 549. Erdem, A. and Sand, L.B. (1979) J. Catal., 60, 241. Lowe, B.M., Nee, J.R.D., and Casci, J.L. (1994) Zeolites, 14, 610. Kanno, N., Miyake, M., and Sato, M. (1994) Zeolites, 14, 625.

References 85. Nastro, A. and Sand, L.B. (1983) Zeo86.

87.

88. 89.

90. 91.

92.

93. Che, M., Masure, D., and Chaquin, P. lites, 3, 57. (1993) J. Phys. Chem., 97, 9022. (a) Cundy, C.S. (1998) Collect. Czech. 94. Morris, C.A., Anderson, M.L., Stroud, Chem. Commun., 63, 1699; (b) Cundy, R.M., Merzbacher, C.I., and Rolison, C.S., Plaisted, R.J., and Zhao, J.P. D.R. (1999) Science, 284, 622. 95. Kulak, A., Park, Y.S., Lee, Y.J., Chun, (1998) Chem. Commun., 1465. Park, S.-E. and Jiang, N. (2008) in Y.S., Ha, K., and Yoon, K.B. (2000) J. Zeolites: From Model Materials to Am. Chem. Soc., 122, 9308. ˇ 96. Hwang, Y.K., Chang, J.-S., Park, S.-E., Industrial Catalysis (eds J. Cejka, J. Kim, D.S., Kwon, Y.-U., Jhung, S.H., Perez-Pariente, and W.J. Roth), TransHwang, J.-S., and Park, M.-S. (2005) world Research Network, Kerala, p. 81. Carmona, J.G., Clemente, R.R., and Angew. Chem. Int. Ed. Engl., 44, 446. 97. Jin, H., Jiang, N., and Park, S.-E. Morales, J.G. (1997) Zeolites, 18, 340. (a) Yates, M.Z., Ott, K.C., Birnbaum, (2008) J. Phys. Chem. Solids, 69, 1136. 98. Choi, K.-M., David, R.B., Han, S.-C., E.R., and McCleeskey, T.M. (2002) and Park, S.-E. (2007) Solid State Angew. Chem. Int. Ed. Engl., 41, 476; Phenom., 119, 167. (b) Ganschow, M., Schulz-Ekloff, G., 99. Chen, X., Yan, W., Shen, W., Yu, J., Wark, M., Wendschuh-Josties, M., and Cao, X., and Xu, R. (2007) Microporous W¨ohrle, D. (2001) J. Mater. Chem., 11, Mesoporous Mater., 104, 296. 1823. Mintova, S., Mo, S., and Bein, T. (1998) 100. Roeffaers, M.B.J., Ameloot, R., Baruah, Chem. Mater., 10, 4030. M., Uji-I, H., Bulut, M., De Cremer, Jacobsen, C.J.H., Madsen, C., G., M¨uller, U., Jacobs, P.A., Hofkens, Houzvicka, J., Schmidt, I., and J., Sels, B.F., and De Vos, D.E. (2008) Carlsson, A. (2000) J. Am. Chem. J. Am. Chem. Soc., 130, 5763. 101. Schoonheydt, R.A. (2008) Angew. Chem. Soc., 122, 7116. Chen, X., Yan, W., Cao, X., Yu, J., and Int. Ed. Engl., 47, 9188. 102. Larlus, O. and Valtchev, V.P. (2004) Xu, R. (2009) Microporous Mesoporous Chem. Mater., 16, 3381. Mater., 119, 217.

153

155

6 Post-synthetic Treatment and Modiﬁcation of Zeolites Cong-Yan Chen and Stacey I. Zones

6.1 Introduction

Zeolites and other molecular sieves ﬁnd widespread applications in catalysis, adsorption, and ion exchange due to their unique shape selective properties and high internal surface areas [1, 2]. Their use in these applications depends largely on both the crystalline structures (e.g., 8-, 10-, 12-ring, or even larger pore openings and one-, two-, or three-dimensional channel systems) and framework compositions (e.g., Si/Al ratios and the associated acidity and hydrophobicity/hydrophilicity). According to the Structure Commission of the International Zeolite Association, 191 framework structure codes have been assigned to zeolites and zeotype materials as of July 2009 [3–5]. Various theoretical studies demonstrate that this number represents only a small fraction of the structures possible for zeolites and other molecular sieves [6–11]. The major challenge in tailoring and utilizing zeolites for speciﬁc applications remains the development of synthesis methods to produce desirable structures with desirable framework compositions. In principle, there are two routes to achieve this goal: (i) direct synthesis and (ii) post-synthetic treatment and modiﬁcation. In this chapter, we will ﬁrst brieﬂy discuss direct synthesis. Then the emphasis will be placed on two selected methods of the post-synthetic treatment and modiﬁcation of zeolites, namely, aluminum reinsertion into zeolite framework using aqueous Al(NO3 )3 solution under acidic conditions and preparation of pure-silica zeolites via hydrothermal treatment with acetic acid.

6.2 Direct Synthesis of Zeolites

The direct synthesis is the primary route of the synthesis of zeolites [12–28]. The major variables that have a predominant inﬂuence on the zeolite structure crystallized include the composition of synthesis mixture, synthesis temperature and time, as well as the preparation procedure of the synthesis mixture such as Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

156

6 Post-synthetic Treatment and Modiﬁcation of Zeolites

aging and seeding. Furthermore, the selection of the structure directing agents (SDAs) plays a critical role in the selective formation of zeolites, as reported in the above references. Depending on the nature of the zeolites involved and the chemistry of their formation, some zeolites can be synthesized in a broad spectrum of framework compositions, as exempliﬁed by IFR-type zeolites which are characterized by an undulating, one-dimensional, 12-ring channel system. IFR is the framework type code of the isostructural zeolites ITQ-4, SSZ-42, and MCM-58, according to the Structure Commission of the International Zeolite Association [3, 4]. As reported in the literature, IFR-type zeolites that can be synthesized via direct synthesis route include ITQ-4 (pure-silica synthesized via ﬂuoride route) [29, 30], Si-SSZ-42 (pure-silica synthesized via conventional hydroxide route) [31], B-SSZ-42 (borosilicate) [22, 32–34], Al-MCM-58 (aluminosilicate) [22, 34–37], Fe-MCM-58 (ferrosilicate) [37, 38], and Zn-SSZ-42 (zincosilicate) [31]. In contrast to IFR-type zeolites, the synthesis of some other structures succeeds only if certain heteroatom X (X = B, Ga, Ge, or Al, for example, or X = none of these elements for pure-silica zeolites) is present in the synthesis mixture and, in turn, incorporated into the framework. In many cases, certain zeolite structures containing speciﬁc heteroatoms can be synthesized only within a limited range of Si/X ratio or in the presence of certain speciﬁc structure directing agents. For example, borosilicate zeolite SSZ-33 (which contains intersecting 10-/12-/12-ring channels with CON topology [3, 4]) can be synthesized with N,N,N-trimethyl-8-tricyclo[5.2.1.02,6 ]decane ammonium cations as the structure directing agent [39, 40]. The direct synthesis of aluminosilicate SSZ-33 using this structure directing agent is so far not successful. On the other hand, aluminosilicate zeolite SSZ-26 (which has a very similar crystalline structure to that of SSZ-33) can be synthesized using N,N,N,N ,N ,N -hexamethyl[4.3.3.0]propellane-8,11-diammonium cations as the structure directing agent [39, 41]. However, this structure directing agent is more difﬁcult to synthesize and more expensive than N,N,Ntrimethyl-8-tricyclo[5.2.1.02,6 ]decane ammonium which is used for the synthesis of B-SSZ-33. Furthermore, Table 6.1 shows the aluminum, boron, and zinc effects on the synthesis of some zeolites when using different structure directing agents. Here we have different molar ratios of silicon to heteroatoms, Si/X, in the synthesis mixtures where X is Al, B, or Zn. These results present here some additional examples which demonstrate the complicated relationship between zeolite structures, framework compositions, and structure directing agents in the direct synthesis. In short, the direct synthesis route often does not lead to the formation of zeolites with the properties desirable for the target applications. As will be discussed in the next section, the post-synthetic treatment and modiﬁcation of zeolites constitute a powerful technique to reach this goal.

6.3 Post-synthetic Treatment and Modiﬁcation of Zeolites Aluminum, boron, and zinc effects on the synthesis of some zeolites when using different structure directing agents.

Table 6.1

Molar ratio in synthesis mixture SDA

N+ N+ (CH3)3

N+

N+ N+(CH3)3

N+(CH3)3

SiO2 (pure silica)

SiO2 /Al2 O3 <50

SiO2 /B2 O3 <30

SiO2 /ZnO <100

ZSM-12

Beta

Beta

VPI-8

ZSM-12

Beta

Beta

VPI-8

ZSM-12

Mordenite

Beta

Layered

SSZ-24

SSZ-25

SSZ-33

SSZ-31

Mordenite

SSZ-33

SSZ-31

SSZ-37

SSZ-33

–

VPI-8

N+

–

6.3 Post-synthetic Treatment and Modiﬁcation of Zeolites

In addition to the direct synthesis method, post-synthetic treatment often provides a more practical route to modify the zeolites to acquire desirable framework compositions and other properties. K¨uhl [42] and Szostak [43] have published two comprehensive review papers about the modiﬁcation of zeolites. These review papers cover many important aspects of post-synthetic treatment and modiﬁcation of zeolites such as ion exchange, preparation of metal-supported zeolites, dealumination, reinsertion of heteroatoms into zeolite framework, and other modiﬁcation methods. The readers are referred to these two review papers and the references therein for the details on the various post-synthetic treatment methods. In this chapter, we will review two newer methods of the post-synthetic treatment and modiﬁcation of zeolites for the lattice substitution, both of which are based on the same principle: the desired atoms such as Al or Si are inserted into lattice sites previously occupied by other T-atoms such as B. These two methods are given below: 1) Aluminum reinsertion into zeolite framework using aqueous Al(NO3 )3 solution under acidic conditions [44, 45]; 2) Synthesis of pure-silica zeolites via hydrothermal treatment with acetic acid [46].

157

158

6 Post-synthetic Treatment and Modiﬁcation of Zeolites

6.3.1 Aluminum Reinsertion into Zeolite Framework Using Aqueous Al(NO3 )3 Solution under Acidic Conditions

As discussed in Section 6.2, many borosilicate zeolites have been successfully synthesized via the direct synthesis route. However, borosilicate zeolites do not possess sufﬁciently high acidity required for many hydrocarbon reactions. The preparation of the catalytically more active aluminosilicate counterparts of these borosilicate zeolites relies, therefore, on the post-synthetic treatment. Various methods have been used to reinsert aluminum into the framework of zeolites [42, 43]. Two of these methods are succinctly described below: 1) Aqueous solution of sodium aluminate is used to insert aluminum into the tetrahedral vacancies and/or substitute aluminum for framework silicon [47]. It is noteworthy is that, due to the chemical nature of sodium aluminate, this method functions under basic conditions which leads to some dissolution of silicon species from the framework of zeolites. 2) Aluminum is inserted into the framework by treating zeolites with AlCl3 vapor in nitrogen at temperatures above 773 K [48–50]. This is basically a reversal of the dealumination reaction using SiCl4 vapor [51–55]. There is a newer post-synthetic treatment method to reinsert aluminum into the framework of zeolites, which operates with an aqueous solution of Al(NO3 )3 under acidic conditions [44, 45], in contrast to the method with sodium aluminate described above. At low pH, this technique minimizes the dissolution of silicon species from the framework of zeolites. This method will be reviewed next. 6.3.1.1 Experimental Procedures This post-synthetic technique consists of two elementary stages: (i) deboronation of borosilicate zeolites under acidic conditions and (ii) reinsertion of aluminum in the lattices of deboronated zeolites using an aqueous Al(NO3 )3 solution. Basically, there are two ways to perform the experiments [44, 45]:

1) Two-step method: These two steps are carried out sequentially. First, the calcined borosilicate zeolite is deboronated at room temperature in 0.01 M aqueous HCl solution for about 24 hours. The zeolite to HCl solution ratio of 1 : 40 by weight is usually used. After washing and air-drying, the deboronated zeolites are then combined with certain amount of aqueous Al(NO3 )3 solution (e.g., 1 g zeolite with 100 g of 1 M Al(NO3 )3 solution) and treated under various conditions (e.g., reﬂux for 100 hours). Typically, the temperature ranges between 363 and 443 K. The ﬁnal zeolite products are washed, ﬁltered, and air-dried. The resulting aluminosilicate zeolites are already in the H-form. 2) One-step method: Here the deboronation and aluminum reinsertion are conducted in the same single step. The experiments start directly with the parent borosilicate zeolites and Al(NO3 )3 solution. The deboronation is carried out by the Al(NO3 )3 solution which has a low pH (<3.5). Other experimental conditions are similar to those of the two-step method.

6.3 Post-synthetic Treatment and Modiﬁcation of Zeolites

Important experimental parameters for this aluminum reinsertion technique include Al(NO3 )3 /zeolite ratio, pH of the slurry, reaction time, and temperature (see above). 6.3.1.2 One-Step Method versus Two-Step Method The two-step method separates the deboronation and aluminum reinsertion steps, similar to many organic syntheses with a series of steps. It demonstrates how these two individual steps proceed and helps our scientiﬁc understanding of the chemistry involved. The one-step method has the advantage of combining these two steps together and providing a more efﬁcient way to reinsert aluminum into the framework of zeolites. Analytical results from X-ray diffraction (XRD), N2 /hydrocarbon adsorption, elemental analysis, 11 B magic angle spinning nuclear magnetic resonance (MAS NMR) with B-SSZ-33 and other zeolites reveal that the deboronated samples from the deboronation of two-step method are essentially boron-free, and their crystalline structures and micropore volumes remain unchanged [44, 45]. Likewise, the ﬁnal products of the one-step method have similar properties in terms of deboronation, namely, they are also boron-free. These results indicate that aqueous Al(NO3 )3 solutions in the one-step method have low enough pH (<3.5) to provide an acidic environment for efﬁcient deboronation, as does the aqueous HCl solution with the two-step method. Table 6.2 compares two Al-SSZ-33 zeolites prepared from B-SSZ-33 via two-step and one-step method, respectively [44, 45]. The molar (Si/B)bulk and (Si/Al)bulk ratios were measured via elemental analyses, while the molar (Si/Al)framework ratios for the framework compositions were determined via 27 Al MAS NMR. Table 6.2 Comparison of Al-SSZ-33 prepared from B-SSZ-33 via two-step and one-step methods.

Method Starting material: B-SSZ-33

Intermediate material: deboronated SSZ-33

Final product: Al-SSZ-33

Two-step

One-step

(Si/B)bulk

18.1

18.1

Micropore volume (milliliters per gram)

0.19

0.19

(Si/B)bulk

>200 (below detection limit)

–

Micropore volume (milliliters per gram)

0.18

–

(Si/Al)bulk

13.1

12.9

(Si/Al)framework Micropore volume (milliliters per gram)

16.6 0.21

16.3 0.21

159

160

6 Post-synthetic Treatment and Modiﬁcation of Zeolites 27

Al MAS NMR results showed that it is possible to exceed the boron lattice substitution in the resulting aluminum contents, but not all the aluminum is in the framework sites. These and additional 27 Al MAS NMR results (Chen, Zones, and Wilson, unpublished results) conﬁrm that most of aluminum in the resulting aluminosilicate zeolites is incorporated into the framework and these aluminum species are already placed into the tetrahedral positions upon the post-synthetic treatment described here. Initially we speculated that some ‘‘intermediate Al-species’’ were ﬁrst attached to the silanol nests (or defects) created by deboronation, and subsequently a calcination step at a high temperature (e.g., >623 K) might be needed to ﬁnally convert these ‘‘intermediate Al-species’’ into tetrahedral aluminum positioned in the framework of zeolites. But such ‘‘intermediate Al-species’’ were not observed in our experiments. Therefore, there is no need to add any additional steps such as calcination to reinsert aluminum into the framework of zeolites. The high micropore volumes of ∼0.2 ml/g (determined with cyclohexane adsorption at P/Po = 0.3 and room temperature) reveal that there is no pore blocking in the channels of the starting material B-SSZ-33, deboronated SSZ-33, and resulting Al-SSZ-33 samples. Supported by the results from XRD, N2 /hydrocarbon adsorption, and elemental analysis, both two-step and one-step methods essentially result in the same aluminosilicate products. Once the chemical roles of the Al(NO3 )3 and HCl solutions for deboronation are understood, from the practical point of view, the deboronation step with aqueous HCl solution can be omitted in practice and the one-step method is the preferred one to use due to its feature of simplicity.

6.3.1.3 Effects of the Ratio of Al(NO3 )3 to Zeolite In Section 6.3.1 we mentioned that the reinsertion of Al into frameworks can be accomplished with sodium aluminate under basic conditions [47]. It is also well known that framework aluminum of zeolites can be extracted via dealumination under acidic conditions [42, 43]. For example, we treated an Al-SSZ-33 sample ((Si/Al)bulk = 13.1) in a HCl solution (pH ∼0.8) under reﬂux for 32 hours. After this treatment in the acidic environment, the crystalline structure of the resulting product remained intact based on XRD and N2 adsorption results but its (Si/Al)bulk increased to 36.9. It reveals that the equilibrium does not favor aluminum reinsertion when we use aqueous solution of Al(NO3 )3 for aluminum reinsertion under acidic conditions. To shift the equilibrium toward aluminum reinsertion under acidic conditions, we have to increase the concentration of Al(NO3 )3 in the reaction slurry to force the conversion in an excess of Al cations. This trend and relationship between aluminum reinsertion and Al(NO3 )3 to zeolite ratio are demonstrated with the results shown in Table 6.3 [45]. When the weight ratio of Al(NO3 )3 solution to zeolite increases, the Si/Al ratio of the products decreases, indicating that more aluminum is reinserted. Although a signiﬁcant amount of Al(NO3 )3 is employed in this technique, it is noteworthy that the excess of Al(NO3 )3 can be always recycled and reused.

6.3 Post-synthetic Treatment and Modiﬁcation of Zeolites Preparation of Al-SSZ-33 from B-SSZ-33 at various ratios of Al(NO3 )3 solution to zeolite via two-step method.

Table 6.3

Zeolite

(Si/B)bulk

B-SSZ-33 Deboronated SSZ-33

18.1 >200 (below detection limit) – – – –

Al-SSZ-33 Al-SSZ-33 Al-SSZ-33 Al-SSZ-33 a S/Z

(Si/Al)bulk – – 24.7 20.1 17.0 13.1

Remarks Starting material Prepared via deboronation of B-SSZ-33 in 0.01 M HCl S/Z = 16 : 1a S/Z = 25 : 1 S/Z = 50 : 1 S/Z = 100 : 1

stands for the weight ratio of 1 M Al(NO3 )3 solution to deboronated SSZ-33.

6.3.1.4 Effects of pH, Time, Temperature, and Other Factors The solubility of aluminum salt becomes lower at higher pH, enhancing the precipitation of aluminum species. When using 1 M Al(NO3 )3 solution for aluminum reinsertion, the ﬁnal pH value of the zeolite/Al(NO3 )3 slurries is below 1. We extraneously raised the pH of the zeolite/Al(NO3 )3 slurries in two preparations by adding some ammonium acetate and water [44, 45]. Their ﬁnal pH was 3.34 and 4.10, whereas the resulting aluminosilicate products had a very low molar bulk Si/Al ratio of 5.5 and 7.5, respectively. These two samples with very high aluminum contents were, however, considerably less active than other Al-SSZ-33 materials prepared at lower pH. Apparently, the precipitation of aluminum species at high pH is the reason for it. Therefore, it is necessary to keep the pH of the zeolite/Al(NO3 )3 slurry below 3.5 to enhance the efﬁciency of aluminum incorporation. Aluminum reinsertion does not occur instantaneously. Shorter reaction time is accompanied by a less efﬁcient aluminum reinsertion. We usually run the experiments for at least 24 hours to reach the reaction equilibrium under the corresponding conditions [44, 45]. The reaction temperature stretches from about 363 to 443 K, which is also the typical temperature range for zeolite synthesis. A higher reaction temperature leads to a quicker reach for the equilibrium. So, the same synthesis devices can be used for both direct synthesis and post-synthetic treatment with Al(NO3 )3 . The experiments can be also easily carried out in heated glass ﬂasks under reﬂux. Although it is simple and convenient to carry out the reactions under static conditions (e.g., in a simple Teﬂon-lined autoclave), stirring, tumbling, or reﬂuxing is preferred in order to enhance the mass and heat transfer. 6.3.1.5 Applicable to Medium Pore Zeolite? Hydrated aluminum cations are too bulky to penetrate through the pores of medium pore zeolites. Therefore, this technique is not applicable to medium pore zeolites. This ﬁnding is supported by the following two parallel experiments conducted in our lab [44, 45]. B-ZSM-11 and Al-ZSM-11 zeolites were synthesized. B-ZSM-11 was then post-synthetically treated with Al(NO3 )3 solution in an attempt

161

162

6 Post-synthetic Treatment and Modiﬁcation of Zeolites

to convert it to its aluminosilicate counterpart. The resulting material is denoted here as B-(Al)-ZSM-11. When comparing B-(Al)-ZSM-11 with Al-ZSM-11 for n-hexane/3-methylpentane cracking, B-(Al)-ZSM-11 had a much lower activity than Al-ZSM-11. Results from elemental analyses indicated that B-(Al)-ZSM-11 took very little aluminum. Since both B-(Al)-ZSM-11 and Al-ZSM-11 exhibit similar XRD patterns characteristic of ZSM-11 and had the same micropore volume (0.17 ml/g) as determined by N2 adsorption, the inactivity of the B-(Al)-ZSM-11 is not due to any pore plugging. We conclude that the real reason is related to the mismatch between the medium pore zeolites (e.g., ZSM-11) with relatively small pore sizes and the bulky hydrated aluminum cations. In addition, based on this feature of medium pore zeolite in relation to the post-synthetic treatment with Al(NO3 )3 , we can use this technique as a tool, together with other physicochemical and catalytic methods, to distinguish medium pore zeolites from large or extra-large pore ones when we have a new zeolite with unknown structure. 6.3.2 Synthesis of Hydrophobic Zeolites by Hydrothermal Treatment with Acetic Acid

Hydrophobic pure-silica zeolites are useful materials for the separations of hydrophobic from hydrophilic compounds [56]. Defect-free or defect-deﬁcient pure-silica zeolites are particularly desirable for the characterization and determination of zeolite structures [57]. Various techniques have been developed to make such materials [42, 43]. Reactions with SiCl4 vapor [51–55] and hexaﬂuorosilicate [54, 58, 59] as well as steaming [60] are three well-known techniques for the dealumination of zeolites and preparation of pure-silica zeolites. As discussed in Section 6.3.1.2, boron in borosilicate zeolites can be easily hydrated and removed from the framework. Similar to the aluminum reinsertion discussed above, the silanol nests (also called vacancies or defects) created by deboronation can be repopulated with silicon. Jones et al. developed a new method to synthesize hydrophobic pure-silica zeolites via post-synthetic treatment of borosilicate zeolites with aqueous acetic acid [46]. With this single-step method, boron is expelled from borosilicate zeolites under acidic conditions and the defects created by the boron removal are subsequently healed with silicon dissolved from other parts of the crystal. By use of this procedure, highly crystalline, hydrophobic pure-silica CIT-1 and SSZ-33 (CON topology) zeolites were synthesized for the ﬁrst time. In the following sections, the features of this technique will be reviewed and some experimental results from our lab will be discussed. 6.3.2.1 Experimental Procedures The acetic acid treatment experiments are typically carried out between 373 and 458 K for six days. The calcined borosilicate zeolite is treated in Teﬂon-lined autoclaves in an oven with rotation at 60 rpm. In a typical experiment, 0.2 g of zeolite is added to a 45 ml autoclave ﬁlled with 25 g of water and 10 g of glacial acetic acid (pH ∼1.65). After six days of heating, the solid products are washed

6.3 Post-synthetic Treatment and Modiﬁcation of Zeolites

extensively with water and acetone and recovered by ﬁltration. So, the experiments are conducted in a way similar to the direct synthesis of zeolites. 6.3.2.2 Highly Crystalline Pure-Silica Zeolites Prepared via This Technique Jones et al. [46] reported successful synthesis of highly crystalline, hydrophobic pure-silica zeolites CIT-1 and SSZ-33 (both CON topology [3, 4]) as well as ERB-1 (MWW topology) and beta (*BEA topology, originally synthesized in ﬂuoride media) under appropriate conditions. They used XRD, 29 Si BD NMR, elemental analysis, as well as N2 and H2 O adsorption to investigate the parent zeolites and the products from acetic acid treatment. The acetic acid treated samples remain crystalline as revealed by the XRD and N2 adsorption results. 29 Si BD NMR results show impressive differences between the parent and acetic acid treated zeolites. Signiﬁcant sharpening of the Q4 peaks in the 29 Si BD NMR spectra allows crystallographically distinct T-sites to be distinguished. The enhanced hydrophobicity of the acetic acid treated zeolites is evidenced by the reduced amount of Q3 silicon species and the decreased H2 O adsorption capacity. This method is effective on zeolites CIT-1, SSZ-33, and ERB-1, all containing 10- and 12-ring channels. But it is noteworthy that the 12-ring pore openings of ERB-1 do not open to the exterior of the crystal. This method appears to be less efﬁcient on ZSM-5, likely suggesting that the small pore apertures of this zeolite could limit migration of soluble silica species. Jones et al. [46] also reported that some beta zeolite samples are either less sensitive (e.g., no signiﬁcant change in hydrophobicity) or oversensitive (e.g., loss of porosity or collapse of the crystalline structure) to this technique within the condition ranges of their investigation. Apparently, further investigations are needed to address these issues. 6.3.2.3 Effects of Type of Acid, pH, Temperature, and Other Factors Jones et al. [46] compared acetic acid with various aqueous mineral acids (HCl, HNO3 , and H2 SO4 ) for this technique and found that treatment with mineral acids results in materials with more signiﬁcant porosity losses than those from treatment with acetic acid. It is speculated that the mineral acids more readily solubilize silica than does the organic acid. Therefore, acetic acid is preferred for this technique. The treatment appears to be most effective near the isoelectric point of silica (pH 0–2). At these conditions, the dissolution of zeolite is slow enough to prevent signiﬁcant loss of microstructure while at the same time allowing sufﬁcient dissolution of silica to provide soluble silicon species for healing of the defects without an external silicon source. Temperature typically ranges between 373 and 458 K and is another important factor. For zeolites that contain no F− , such as CIT-1, SSZ-33, and ERB-1 (synthesized in hydroxide media), higher temperatures (433–458 K) give better results with limited loss in porosity due to structural degradation. In contrast, zeolites synthesized in the presence of F− require lower temperatures for effective healing of defects with silicon, likely due to structural degradation by traces of residual F− when at higher temperature.

163

164

6 Post-synthetic Treatment and Modiﬁcation of Zeolites

Jones et al. [46] also reported that additional aluminum or gallium source can be added to the acetic acid solution, and the silanol defects that are not ﬁlled with aluminum or gallium are repopulated with silicon during the treatment. As a result, it is possible to produce a more hydrophobic metallosilicate zeolite than the one containing internal silanol defects. 6.3.2.4 Experimental Results from Our Lab We prepared the following three SSZ-33 samples to investigate their physicochemical and catalytic properties [61]:

1)

A borosilicate SSZ-33 was synthesized using N,N,N-trimethyl-8-ammoniumtricycle-[5.2.1.02,6 ]decane hydroxide as structure directing agent [40]. It was calcined following the standard procedure to remove the occluded structure directing agent molecules. The calcined sample is denoted as B-SSZ-33. 2) A deboronated SSZ-33 was prepared by stirring B-SSZ-33 in an aqueous HCl solution (0.01 M) at room temperature and a zeolite-to-solution weight ratio of 1 : 40 for 24 hours. The deboronation creates SiOH nests in place of boron T-atoms and the resulting sample is denoted as SiOH-SSZ-33. 3) A pure-silica SSZ-33 was prepared by tumbling B-SSZ-33 in acetic acid (28.6 wt%) at 458 K and a zeolite-to-solution weight ratio of 1 : 25 for six days according to Jones et al. [46]. The resulting pure-silica zeolite is denoted as Si-SSZ-33. Each of B-SSZ-33, SiOH-SSZ-33, and Si-SSZ-33 was impregnated with 0.5 wt% Pt. The resulting Pt-loaded catalysts are denoted as Pt/B-SSZ-33, Pt/SiOH-SSZ-33, and Pt/Si-SSZ-33, respectively. These were investigated with XRD, elemental analysis, thermogravimetric analysis (TGA), N2 physisorption, H2 chemisorption if applicable, 11 B MAS NMR, 29 Si BD NMR, hydrocracking of n-octane, and other techniques. All samples (B-SSZ-33, SiOH-SSZ-33, and Si-SSZ-33 as well as Pt/B-SSZ-33, Pt/SiOH-SSZ-33, and Pt/Si-SSZ-33) possess XRD peaks characteristic of SSZ-33 and are highly crystalline. They all have a micropore volume of about 0.2 cc/g as determined by N2 adsorption. These results indicate that the SSZ-33 structure is well retained after deboronation with HCl, treatment with acetic acid, and Pt-loading. Results from elemental analysis and 11 B MAS NMR indicate that boron is essentially completely removed from the zeolite in SiOH-SSZ-33 and Si-SSZ-33. Figure 6.1 shows the 29 Si BD NMR spectra of B-SSZ-33, SiOH-SSZ-33, and Si-SSZ-33. Typical of borosilicate zeolites, the B-SSZ-33 sample shows a broad resonance of Q4 silicon T-atoms at about −110 ppm, indicating that the seven crystallographically distinct T-atoms of SSZ-33 are not distinguishable as a result of the distribution of chemical shifts caused by the presence of lattice boron atoms and framework defects. Consistent with this is a signal at −100 ppm from Q3 silicon atoms. SiOH-SSZ-33 exhibits an increased amount of Q3 silicon atoms that are created as SiOH nests in the framework upon the deboronation with HCl. The treatment with acetic acid expels boron from the zeolite and the defects created by

6.3 Post-synthetic Treatment and Modiﬁcation of Zeolites

165

Si-SSZ-33

SiOH-SSZ-33

B-SSZ-33 −80

−90

Figure 6.1

29 Si

−100

−110 ppm

−120

−130

−140

BD NMR spectra of B-SSZ-33, SiOH-SSZ-33, and Si-SSZ-33 [61].

deboronation are subsequently healed with silicon dissolved from other parts of the crystal [46]. With Si-SSZ-33, the Q3 silicon groups are reduced via acetic acid treatment, rendering a more deﬁned 29 Si BD NMR spectrum, which is indicative of fewer internal defects in Si-SSZ-33. TGA results also provide useful information on the hydrophobicity, dehydration, and dehydroxylation of B-SSZ-33, SiOH-SSZ-33, and Si-SSZ-33. As shown in Figure 6.2, these samples exhibit two distinct stages of weight loss: the ﬁrst weight loss (∼298–373 K) is due to dehydration and the second stage (above ∼573 K) is related to dehydroxylation. The lower weight loss of SiOH-SSZ-33 versus B-SSZ-33 between ∼298 and 373 K reﬂects some difference in hydrophobicity between B-SSZ-33 and SiOH-SSZ-33. When compared to B-SSZ-33, SiOH-SSZ-33 loses more weight above ∼573 K. This weight loss is attributed to dehydroxylation of the 100 Si-SSZ-33

SiOH-SSZ-33

85

600

400

Temperature

90

Pt/SiOH-SSZ-33

85

B-SSZ-33 0 (a)

80

20 40 60 80 100 120 140 160 180 Time (min)

600

400

Pt/B-SSZ-33 200

80

800

95 Weight (%)

Temperature

Temperature (K)

Weight (%)

95 90

Pt/Si-SSZ-33

800

200 0

(b)

Figure 6.2 TGA of (a) B-SSZ-33, SiOH-SSZ-33, and Si-SSZ-33 and (b) Pt/B-SSZ-33, Pt/SiOH-SSZ-33, and Pt/Si-SSZ-33. All samples were saturated with moisture in ambient atmosphere [61].

20 40 60 80 100 120 140 160 180 Time (min)

Temperature (K)

100

166

6 Post-synthetic Treatment and Modiﬁcation of Zeolites

silanol groups present in abundance in SiOH-SSZ-33. Si-SSZ-33 clearly exhibits a higher hydrophobicity (reduced dehydration) and lower content of structural defects (reduced dehydroxylation). The TGA proﬁles and weight losses of Pt/B-SSZ-33 and Pt/SiOH-SSZ-33 catalysts are essentially the same as those of their parent zeolites B-SSZ-33 and SiOH-SSZ-33. Pt/Si-SSZ-33 sample contains some more water than Si-SSZ-33, maybe due to the presence of Pt particle. The Pt dispersion in Pt/B-SSZ-33, Pt/SiOH-SSZ-33, and Pt/Si-SSZ-33 catalysts was determined via H2 chemisorption. As reported earlier [61], Pt/B-SSZ-33 has the best Pt dispersion, most likely due to the stronger Pt–B interaction versus Pt–SiOH interaction for Pt/SiOH-SSZ-33. The very low Pt dispersion of Pt/Si-SSZ-33 indicates that structural defects are needed to anchor and disperse Pt in zeolites for catalytic applications, although, by contrast, defect-free zeolite samples are always desirable for structure characterization and determination [57]. It is known that iso-alkanes (e.g., iso-butane) are preferably yielded by hydrocracking of normal alkanes (e.g., n-octane). With decreasing acid strength of the catalyst, the yield of n-butane from n-octane hydrocracking becomes more predominant. Sulﬁding minimizes the hydrogenolysis activity of the Pt-loaded catalysts. We measure the weak acid strength of the catalysts with the so-called Acidity Index [61, 62] which is deﬁned as the yield ratio of iso-butane to n-butane produced at 95% C5+ yield of n-octane (containing 20 ppm sulfur) hydrocracking over the sulﬁded catalysts. A lower value of the Acidity Index corresponds to a weaker acid strength. Pt/B-SSZ-33 and Pt/SiOH-SSZ-33 have an Acidity Index of 0.049 and 0.010, respectively. Pt/SiOH-SSZ-33 has a lower catalytic activity than Pt/B-SSZ-33, as evidenced by a difference in reaction temperature of 661 K for Pt/B-SSZ-33 versus 672 K for Pt/SiOH-SSZ-33 to reach 95% C5+ yield of n-octane hydrocracking. As a reference, Pt/mordenite has an Acidity Index above 3, as expected from the higher acid strength of this aluminosilicate zeolite. With its higher Pt dispersion, higher Acidity Index, and higher catalytic activity as compared to Pt/SiOH-SSZ-33, Pt/B-SSZ-33 appears to have a stronger ability of anchoring and dispersing Pt particles inside the channel system of SSZ-33 and exhibits a higher acid strength of the acid sites. Both factors are attributed to the framework boron sites. Because of the lack of acid function as well as Pt aggregation and its consequent low catalytic activity, the Acidity Index of Pt/Si-SSZ-33 is difﬁcult to be determined meaningfully. 6.4 Summary and Outlook

K¨uhl and Szostak have published two comprehensive review papers on the post-synthetic treatment and modiﬁcation of zeolites [42, 43]. These review papers cover many important aspects of post-synthetic treatment and modiﬁcation of zeolites such as ion exchange, preparation of metal-supported zeolites, dealumination, reinsertion of heteroatoms into zeolite framework, and other modiﬁcation methods. Complementary to these two papers, here we have reviewed two newer methods for post-synthetic lattice substitution. These two methods are (i) aluminum reinsertion into zeolite framework using aqueous Al(NO3 )3 solution under

References

acidic conditions and (ii) synthesis of pure-silica zeolites via hydrothermal treatment with acetic acid. Both methods use borosilicate zeolites as starting materials and operate on the basis of the same principle, namely, the desired atoms such as aluminum or silicon are inserted into lattice sites previously occupied by boron atoms. They are carried out under conditions typical of direct synthesis of zeolites. Aluminum reinsertion into zeolite framework using aqueous Al(NO3 )3 solution operates at pH below 3.5. It is known, however, that the reaction equilibrium favors dealumination over aluminum reinsertion under acidic conditions. Therefore, a high ratio of Al(NO3 )3 to zeolite in an excess of aluminum cations is required to force the equilibrium to shift toward aluminum reinsertion under acidic conditions. Although a signiﬁcant amount of Al(NO3 )3 is employed in this technique, it is noteworthy that the excess of Al(NO3 )3 can be always recycled and reused. In addition, the dissolution of silicon species from the framework is minimized under acidic conditions, overcoming the inherent disadvantage of the aluminum reinsertion method using sodium aluminate under basic conditions. Aluminum reinsertion method using aqueous Al(NO3 )3 solution is not applicable to medium pore zeolites because hydrated aluminum cations are too bulky to penetrate through the pores of medium pore zeolites. As demonstrated by converting B-SSZ-33, B-UTD-1, and many other borosilicate zeolites to their catalytically more active aluminosilicate counterparts, this method has proved to be a useful way to synthesize large or extra-large pore aluminosilicate zeolites which are difﬁcult or impossible to be prepared via direct synthesis. In the future, it will continue to serve as a practical and efﬁcient tool to synthesize these aluminosilicate zeolites before a more economic method of direct synthesis is found. Hydrothermal treatment of calcined borosilicate zeolites with aqueous acetic acid near the isoelectric point of silica (pH 0–2) effectively produces hydrophobic pure-silica zeolites with few structural defects. Here, the boron is expelled from the zeolites and the defects created by the deboronation are subsequently healed with silicon dissolved from other parts of the crystal. By use of this technique, highly crystalline, hydrophobic, pure-silica CIT-1 and SSZ-33 zeolites were synthesized for the ﬁrst time. It appears to be a powerful alternative to other post-synthetic treatment methods such as steaming as well as reactions with SiCl4 vapor and hexaﬂuorosilicate. This technique provides a new route to synthesize high-quality zeolites for the characterization and determination of zeolite structures. Acknowledgments

We thank Chevron Energy Technology Company for support of our zeolite research, especially Dr. C.R. Wilson and Dr G.L. Scheuerman. References 1. Weitkamp, J. and Puppe, L. (eds) (1999)

Catalysis and Zeolites – Fundamentals and Applications, Springer, p. 564.

ˇ J., van Bekkum, H., Corma, A., 2. Cejka, and Sch¨uth, F. (eds) (2007) Introduction to Zeolite Science and Practice, Studies in

167

168

6 Post-synthetic Treatment and Modiﬁcation of Zeolites

3. 4.

5.

6.

7.

8.

9. 10. 11.

12.

13.

14.

15.

16.

Surface Science and Catalysis, Vol. 168, 3rd revised edn, Elsevier, p. 1058. http://www.iza-structure.org/ (last accessed February 03, 2010). Baerlocher, Ch., McCusker, L.B., and Olson, D.H. (2007) Atlas of Zeolite Framework Types, 6th revised edn, Elsevier, p. 398. McCusker, L.B. and Baerlocher, Ch. (2007) in Introduction to Zeolite Science and Practice, Studies in Surface Science and Catalysis, Vol. 168, 3rd revised ˇ edn (eds J. Cejka, H. van Bekkum, A. Corma, and F. Sch¨uth), Elsevier, pp. 13–37. Treacy, M.M.J., Rao, S., and Rivin, I. (1993) in Proceedings of the 9th International Zeolite Conference, Montreal, 1992 (eds R. von Ballmoos, J.B. Higgins, and M.M.J. Treacy), Butterworth-Heinemann, pp. 381–388. Treacy, M.M.J., Randall, K.H., Rao, S., Perry, J.A., and Chadi, D.J. (1997) Z. Krist., 212, 768–791. Treacy, M.M.J., Randall, K.H., and Rao, S. (1999) in Proceedings of the 12th International Zeolite Conference, Baltimore, 1998 (eds M.M.J.Treacy, B.K. Marcus, M.E. Bisher, and J.B. Higgins), Materials Research Society, pp. 517–532. Falcioni, M. and Deem, M.W. (1999) J. Chem. Phys., 110, 1754–1766. Le Bail, A. (2005) J. Appl. Cryst., 38, 389–395. Foster, M.D. and Treacy, M.M.J. A Database of Hypothetical Zeolite Structures, http://www.hypotheticalzeolites.net (last accessed February 03, 2010). Zones, S.I. and Nakagawa, Y. (1994) Microporous Mesoporous Mater., 2, 543–555. Lobo, R.F., Zones, S.I., and Davis, M.E. (1995) J. Inclusion Phenom. Mol. Recognit. Chem., 21, 47–78. Zones, S.I. and Davis, M.E. (1996) Curr. Opin. Solid State Mater. Sci., 1, 107–117. Zones, S.I. and Maxwell, I.E. (1997) Curr. Opin. Solid State Mater. Sci., 2, 55–56. Davis, M.E. and Zones, S.I. (1997) in Synthesis of Porous Materials: Zeolites, Clays and Nanostructures, (eds M.L.

17.

18.

19. 20.

21.

22.

23. 24.

25.

26.

27.

28.

29.

Occelli and H. Kessler), Marcel Dekker, pp. 1–34. Nakagawa, Y., Lee, G.S., Harris, T.V., Yuen, L.T., and Zones, S.I. (1998) Microporous Mesoporous Mater., 22, 69–85. Zones, S.I., Nakagawa, Y., Lee, G.S., Chen, C.Y., and Yuen, L.T. (1998) Microporous Mesoporous Mater., 21, 199–211. Millini, R., Perego, G., and Bellussi, G. (1999) Top. Catal., 9, 13–34. Camblor, M.A., Villaescusa, L.A., and D´ıaz-Caba˜ nas, M.J. (1999) Top. Catal., 9, 59–76. Lee, G.S., Nakagawa, Y., Hwang, S.J., Davis, M.E., Wagner, P., Beck, L.W., and Zones, S.I. (2002) J. Am. Chem. Soc., 124, 7024–7034. Zones, S.I. and Hwang, S.J. (2003) Microporous Mesoporous Mater., 58, 263–277. Corma, A. and Davis, M.E. (2004) Chem. Phys. Chem., 5, 304–313. Corma, A. (2004) in Book of Abstracts of the 14th International Zeolite Conference, Cape Town, 2004 (eds E. van Steen, L.H. Callanan, and M. Claeys), pp. 25–41. Also available on CD. Burton, A.W., Zones, S.I., and Elomari, S.A. (2005) Curr. Opin. Colloid Interface Sci., 10, 211–219. Burton, A.W. and Zones, S.I. (2007) in Introduction to Zeolite Science and Practice, Studies in Surface Science and Catalysis, Vol. 168, 3rd revised edn (eds ˇ J. Cejka, H. van Bekkum, A. Corma, and F. Sch¨uth), Elsevier, pp. 137–179. Wilson, S.T. (2007) in From Zeolites to Porous MOF Materials, Proceedings of the 15th International Zeolite Conference, Beijing, 2007, Studies in Surface Science and Catalysis, Vol. 170 (eds R. Xu, Z. Gao, J. Chen, and W. Yan), Elsevier, pp. 3–18. Robson, H. and Lillerud, K.P. (2001) Veriﬁed Syntheses of Zeolitic Materials, 2nd revised edn, Elsevier on behalf of the Synthesis Commission of the International Zeolite Association, p. 266. Camblor, M.A., Corma, A., and Villaescusa, L.A. (1997) J. Chem. Soc., Chem. Commun., 749–750.

References 30. Barrett, P.A., Camblor, M.A., Corma, A.,

31.

32.

33.

34.

35. 36. 37.

38.

39.

40.

41.

42.

Jones, R.H., and Villaescusa, L.A. (1997) Chem. Mater., 9, 1713–1715. Chen, C.Y., Zones, S.I., Hwang, S.J., Burton, A.W., and Liang, A.J. (2007) in From Zeolites to Porous MOF Materials, Proceedings of the 15th International Zeolite Conference, Beijing, 2007, Studies in Surface Science and Catalysis, Vol. 170 (eds R. Xu, Z. Gao, J. Chen, and W. Yan), Elsevier, pp. 206–213. Chen, C.Y., Finger, L.W., Medrud, R.C., Crozier, P.A., Chan, I.Y., Harris, T.V., and Zones, S.I. (1997) J. Chem. Soc., Chem. Commun., 1775–1776. Chen, C.Y., Finger, L.W., Medrud, R.C., Kibby, C.L., Crozier, P.A., Chan, I.Y., Harris, T.V., Beck, L.W., and Zones, S.I. (1998) Chem. Eur. J., 4, 1312–1323. Chen, C.Y., Zones, S.I., Yuen, L.T., Harris, T.V., and Elomari, S.A. (1999) in Proceedings of the 12th International Zeolite Conference, Baltimore, 1998 (eds M.M.J. Treacy, B.K. Marcus, M.E. Bisher, and J.B. Higgins), Materials Research Society, pp. 1945–1952. Valyocsik, E.W. (1995) US Patent 5,441,721. Ernst, S., Hunger, M., and Weitkamp, J. (1997) Chem. Ing. Tech., 68, 77–79. Koˇsov´a, G., Ernst, S., Hartmann, M., ˇ and Cejka, J. (2005) Eur. J. Inorg. Chem., 1154–1161. Koˇsov´a, G., Ernst, S., Hartmann, M., ˇ and Cejka, J. (2004) in Book of Abstracts of the 14th International Zeolite Conference Cape Town, 2004, (eds E. van Steen, L.H. Callanan, and M. Claeys), pp. 362–363. Also available on CD. Lobo, R.F., Pan, M., Chan, I.Y., Medrud, R.C., Zones, S.I., Crozier, P.A., and Davis, M.E. (1994) J. Phys. Chem., 98, 12040–12052. Zones, S.I., Holtermann, D.L., Santilli, D.S., Jossens, L.W., and Kennedy, J.V. (1990) US Patent 4,963,337. Zones, S.I., Santilli, D.S., Ziemer, J.N., Holtermann, D.L., and Pecoraro, T.A. (1990) US Patent 4,910,006. K¨uhl, G.H. (1999) in Catalysis and Zeolites – Fundamentals and Applications (eds J. Weitkamp and L. Puppe), Springer, pp. 81–197.

43. Szostak, R. (2001) in Introduction to

44. 45.

46.

47.

48.

49.

50. 51.

52.

53. 54.

55.

56.

Zeolite Science and Practice, Studies in Surface Science and Catalysis, Vol. 137, 2nd completely revised and expanded edn (eds H. van Bekkum, P.A. Jacobs, E.M. Flanigen, and J.C. Jansen), Elsevier, pp. 261–297. Chen, C.Y. and Zones, S.I. (2002) US Patent 6,468,501. Chen, C.Y. and Zones, S.I. (2001) in Zeolites and Mesoporous Materials at the Dawn of the 21st Century, Proceedings of the 13th International Zeolite Conference, Montpellier, 2001, Studies in Surface Science and Catalysis, Vol. 135 (eds A. Galarneau, F. Di Renzo, F. Fajula, and J. Vedrine), Elsevier, pp. 211–218. Jones, C.W., Hwang, S.J., Okubo, T., and Davis, M.E. (2001) Chem. Mater., 13, 1041–1050. Sulikowski, B., Rakoczy, J., Hamdan, H., and Klinowski, J. (1987) J. Chem. Soc., Chem. Commun., 1542–1543. Jacobs, P.A., Tielen, M., Nagy, J.B., Debras, G., Derouane, E.G., and Gabelica, Z. (1983) in Proceedings of the 6th International Zeolite Conference, Reno, 1983 (eds D.H.Olson and A. Bisio), Butterworth, pp. 783–792. Anderson, M.W., Klinowski, J., and Liu, X. (1984) J. Chem. Soc., Chem. Commun., 1596–1597. Dessau, R.M. and Kerr, G.T. (1984) Zeolites, 4, 315–318. Beyer, H.K., Belenykaya, I. (1980) in Catalysis by Zeolites, Studies in Surface Science and Catalysis, Vol. 5 (eds B. Imelik et al.), Elsevier, pp. 203–210. Beyer, H.K., Belenykaya, I., Hange, F., Tielen, M., Grobet, P.J., and Jocobs, P.A. (1985) J. Chem. Soc., Faraday Trans. I, 81, 2889–2901. Grobet, P.J., Jacobs, P.A., and Beyer, H.K. (1986) Zeolites, 6, 47–50. Weitkamp, J., Sakuth, M., Chen, C.Y., and Ernst, S. (1989) J. Chem. Soc., Chem. Commun., 1908–1910. Li, H.X., Annen, M.J., Chen, C.Y., Arhancet, J.P., and Davis, M.E. (1991) J. Mater. Chem., 1, 79–85. Weitkamp, J., Kleinschmit, P., Kiss, A., and Breke, C.H. (1993) in Proceedings of the 9th International Zeolite Conference, Montreal, 1992, Vol. 2 (eds

169

170

6 Post-synthetic Treatment and Modiﬁcation of Zeolites

57.

58.

59. 60.

R. von Ballmoos, J.B. Higgins, and M.M.J. Treacy), ButterworthHeinemann, pp. 79–87. Fyfe, C.A., Gies, H., Kokotailo, G.T., Marler, B., and Cox, D.E. (1990) J. Phys. Chem., 94, 3718–3721. Skeels, G.W. and Breck, D.W. (1983) in Proceedings of the 6th International Zeolite Conference, Reno, 1983 (eds D.H.Olson and A. Bisio), Butterworth, pp. 87–96. Garralon, G., Fornes, V., and Corma, A. (1988) Zeolites, 8, 268–272. McDanial, C.V. and Maher, P.K. (1968) in Molecular Sieves (ed. R.M.Barrer), Society of Chemical Industry, p. 186.

61. Chen, C.Y., Zones, S.I., Hwang, S.J.,

and Bull, L.M. (2004) in Book of Abstracts of the 14th International Zeolite Conference, Cape Town, 2004 (eds E. van Steen, L.H. Callanan, and M. Claeys), pp. 1547–1554. Also available on CD. 62. Chen, C.Y., Rainis, A., and Zones, S.I. (1997) in Proceedings of Materials Research Society Symposium on ‘‘Advanced Catalytic Materials – 1996’’, vol. 454 (eds P.W. Lednor, M.J. Ledoux, D.A. Nagaki, and L.T. Thompson), Materials Research Society, pp. 205–215.

171

7 Structural Chemistry of Zeolites Paul A. Wright and Gordon M. Pearce 7.1 Introduction

Zeolites are tetrahedrally connected framework solids, based on silica, with intricate structures that possess channels and cages large enough to contain extra-framework cations and to permit the uptake and desorption of molecules varying from hydrogen to complex organics up to 1 nm in size. Their crystalline structure directly controls their properties and consequently their performance in applications such as ion exchange, separation, and catalysis, and is therefore of great interest to academics and technologists alike. A ‘‘ball and stick’’ representation of the most widely used zeolite A, with tetrahedral Al and Si atoms linked by O atoms and with chargebalancing Na+ cations, is given in Figure 7.1. Originally discovered as aluminosilicate minerals, synthetic zeolites with a range of compositions are now widely prepared and subsequently modiﬁed for a wide range of applications. Many excellent articles, reviews, and books describe the structures of these solids, often in well-illustrated texts [1–3] and online resources [4]. Here we start by summarizing their structural chemistry, beginning with the key features of the best known and most widely used zeolites A and Y. Besides covering the periodic structures of these solids, other features such as secondary mesoporosity are also important to their performance and are discussed. This is followed by a summary of the structural chemistry of some of the important zeolite types prepared using inorganic and simple organic cations prior to the 1990s. Over the last 20 years there has been a major international effort to prepare new zeolitic materials, in the search for improved adsorbents and catalysts. Much of this has focused on the exploration of the use of complex organic alkylammonium ions, synthesized speciﬁcally as potential ‘‘templates,’’ giving high-silica-content zeolites or pure silica polytypes. The diversity of structures has also been increased by the inclusion of elements other than Al for Si in the framework, which may be either aliovalent (2+ or 3+) or isovalent (4+), and the search for new materials is encouraged by the tantalizing arrays of hypothetical structures that have been shown to be energetically feasible [5, 6]. The remarkable products of this ongoing odyssey continue to show new structural features that are both intriguing and of practical importance or potential: increased crystallographic complexity leading to structures with novel Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

172

7 Structural Chemistry of Zeolites

c

b a

Figure 7.1 ‘‘Ball and stick’’ representation of part of the structure of the sodium form of zeolite A (Na-A), showing ordered Si and Al framework atoms (dark and light grey spheres) linked by O atoms (red spheres).

Extra-framework Na+ cations (orange spheres) occupy sites in the six-membered and eight-membered rings (where this refers to the number of tetrahedrally coordinated cations in the ring).

combinations of pore size, geometry and connectivity; coordination of framework cations above fourfold; chirality and mesoporosity; ordered defects; and layered zeolitic precursors. The second main section of this chapter therefore concentrates on some of the latest structural chemistry to be discovered and its signiﬁcance. The structures and structural chemistry described in this chapter have been established by a combination of diffraction methods. Synthetic zeolites rarely crystallize as single crystals of sufﬁcient size or quality for single crystal diffraction, so that the careful and often inspired analysis of powder diffraction proﬁles, often in combination with electron microscopy, has been essential. This is described elsewhere in this book. Similarly, a full picture of how these structures arise can only be achieved through an understanding of the conditions and mechanism of their crystallization. The ﬁrst steps in unraveling this process are underway and are also described in this volume.

7.2 Zeolite Structure Types Exempliﬁed by Those Based on the Sodalite Cage 7.2.1 Introduction

Strictly deﬁned, zeolites are aluminosilicates with tetrahedrally connected framework structures based on corner-sharing aluminate (AlO4 ) and silicate (SiO4 )

7.2 Zeolite Structure Types Exempliﬁed by Those Based on the Sodalite Cage

tetrahedra. Conceptually, they are based on pure silica frameworks with Si substituted by Al. This aliovalent (Al3+ ↔ Si4+ ) substitution imparts an overall negative charge to the framework. This is balanced by the presence of extra-framework charge-balancing cations within the pore space, which is also able to take in neutral atoms and molecules small enough to enter via the pore windows. A simpliﬁed empirical formula for an aluminosilicate zeolite is Mn+ x/n Alx Si1−x O2 · yX

(7.1)

where Mn+ represents inorganic or organic cations and X are included or adsorbed species. The building blocks of the aluminosilicate framework are the tetrahedra. Typi˚ with OTO angles (where cally, Al–O and Si–O bond distances are 1.73 and 1.61 A, T is the tetrahedral cation) close to the tetrahedral angle, 109.4◦ [7]. There is more variation in the TOT bond angles between tetrahedra, where the average angle is 145◦ but there is considerable spread. (A recent review of the crystallographic data on pure silica zeolites gives the following distribution: range 133.6–180◦ , mode 148◦ , mean 154(±9)◦ [8].) The variation in TOT angles is in large part responsible for the great structural diversity they exhibit. Values close to 180◦ are often reported, although this can commonly be attributed to fractional atomic coordinates of oxygen atoms averaged over different positions. A recent electron microscopic and diffraction study of SSZ-24 [9], a pure silica polymorph, sheds some light on the 180◦ TOT bond angle observed in this solid. In addition to the usual crystallographic repeat along the channel axis (parallel to the linear bond) an incommensurate structural repeat was observed, with a repeat unit of 2.63 × the c repeat. This indicates that there is a long-range modulation in the deviation of this TOT bond angle from 180◦ , with the O atom moving off its average position. Al–O–Al linkages are not observed in hydrothermally synthesized aluminosilicate zeolites, because of the unfavorable interaction of adjacent negative charges associated with aluminate tetrahedra – an observation expressed as L¨owenstein’s rule. A similar rule applies for Ga–O–Ga linkages in gallosilicates. At higher Al (or Ga) contents, L¨owenstein’s rule results in short-range ordering (conﬁrmed by solid-state NMR) and ultimately long-range ordering. When Si/Al approaches 1, in zeolite A, for example, there is strict alternation between Si and Al. Several elements other than Al are able to substitute for Si in tetrahedral positions in the framework. These include the divalent cations Be2+ and Zn2+ , other trivalent cations such as B3+ , Ga3+ , and Fe3+ , and tetravalent cations such as Ti4+ and Ge4+ . These can substitute either at low levels, where they may not affect the zeolite structure to form, or at higher concentrations, where Zn2+ , Be2+ , and Ga3+ , for example, tend to give rise to novel structures. This can be ascribed in part to their ˚ Zn2+ , 0.60 A; ˚ B3+ , 0.11 A; ˚ Al3+ , 0.39 A; ˚ different ionic radii [10] (Be2+ , 0.27 A; ˚ Fe3+ , 0.49 A; ˚ Ti4+ , 0.42 A; ˚ Ge4+ , 0.39 A) ˚ compared to Si4+ , 0.26 A, ˚ Ga3+ , 0.47 A; which in some cases favor particular structural units. The incorporation of Ge in silicate frameworks, for example, favors the formation of D4Rs [11]. Topological examination of the possible ways of assembling tetrahedra within all possible space group symmetries indicates that the number is practically limitless,

173

7 Structural Chemistry of Zeolites

18 16

Organic SDA Inorganic only

14 Structure types

12 10 8 6 4 2 0 67 68 – ′ 70 71– ′ 73 74– ′ 76 77– ′ 79 80– ′ 82 83– ′ 85 86– ′ 88 89– ′ 91 92– ′ 94 95– ′ 97 98– ′ 00 01– ′ 03 04– ′ 06 07– ′ 09

174

Date Figure 7.2 Chart showing the history of the ﬁrst preparations of synthetic silicates with tetrahedral open frameworks, as reported in the ‘‘Atlas of Zeolite Structures’’ [4]. A distinction is made between those

prepared by including only inorganic cations in extra-framework positions and those that include organic species as structure-directing agents, or SDAs.

even taking into account the constraints of energetic feasibility. Considering only silicates, 144 framework topologies have so far been observed, either as natural minerals or synthetic materials [4]. Figure 7.2 shows the trend in the ﬁrst observation of different synthetic zeolitic silicate structures, making a distinction between materials prepared with inorganic or organic cations acting as structure-directing agents (SDAs). This ﬁgure underlines the increasing importance of organic species, typically alkylammonium cations, controlling the synthesis of novel structures. Because these are typically bulky molecular cations, their inclusion in the pore space requires the silicate framework to have lower negative charge density than those formed in the presence of metal cations, so that in aluminosilicates, for example, these organically ‘‘templated’’ zeolites have high framework Si/Al ratios. A full description of all observed zeolite structure types is available on the web, courtesy of the Structure Commission of the International Zeolite Association [4]. Each observed framework topology is given a three letter code (e.g., FAU for faujasite, MFI for ZSM-5, etc.), and for the ﬁrst-observed type structure, full details (symmetry, atomic coordinates, secondary building units (SBUs), coordination sequences of tetrahedral nodes, pore connectivity, pore dimensions, etc.) are given. Typically, zeolites with windows deﬁned by 8-rings (made up of eight tetrahedrally coordinated cations and eight bridging O atoms) with pore sizes of around 4 A˚ are described as small-pore zeolites, those with 10-ring windows as medium-pore ˚ those with 12-ring windows as large-pore zeolites zeolites (typically about 5–5.5 A),

7.2 Zeolite Structure Types Exempliﬁed by Those Based on the Sodalite Cage

˚ and those with windows made up by larger rings (14-ring or 18-ring, for (7–8 A) example) are described as extra-large-pore solids. An account is also kept of the different chemical compositions of reported structures which have each topology type, along with references. This ‘‘Atlas of Zeolite Structures’’ includes all porous solids which possess fully four-connected nets, regardless of chemical composition, so that there are many phosphates, germanates, nitrides, and even sulﬁdes that fulﬁll these geometric requirements. This compilation, which is rigorously checked and kept up-to-date and contains many of the most relevant structural references, is therefore the essential source of information for zeolite structures. 7.2.2 The Framework: Secondary Building Units in Zeolite Structural Chemistry

Silicate (and aluminate, borate, gallate, germanate, titanate, etc.) tetrahedra are the primary building units of zeolites, but consideration of the observed framework types leads naturally to their description in terms of SBUs which are characteristic arrangements of tetrahedra. These are assembled, either on their own or in combination with others, to give the periodic structures. Whether these units exist independently, for example, in solution, remains an open question. Examples of SBUs are given in Figure 7.3, where only the topology of the linked tetrahedral

(a)

(b)

(c) Figure 7.3 Examples of secondary building units (SBUs) found in zeolite structures, represented by single lines joining the tetrahedral atom (T-atom) positions. (a) Rings; (b) double four-membered rings (D4Rs), D6Rs, and pentasil units; and (c, left to right) cancrinite, sodalite, paulingite, and α-cages.

175

176

7 Structural Chemistry of Zeolites

cations is given – in reality, these will be linked by O atoms. The SBUs include rings with different numbers of tetrahedral cations (three-rings, four-rings, etc.), double four-rings, D4Rs, (which contain two rings of four tetrahedral cations), D6Rs, and an array of polyhedral units and cages containing faces with even number of edges, including the cancrinite cage, the paulingite cage, the sodalite or β-cage and the α-cage. One way to represent these cages is in terms of their faces (or rather the number of edges on each face). In this way, a D4R is represented as [46 ], a D6R as [46 62 ], and a sodalite cage as [68 46 ]. Besides these polyhedral units, many high silica zeolites have SBUs containing ﬁve-rings (including the so-called pentasil zeolites ZSM-5 and ZSM-11). In these latter solids the characteristic pentasil unit is observed. Besides these polyhedral units, characteristic chains are observed in zeolite structures, such as the zig-zag, sawtooth, crankshaft, natrolite 4=1, and pentasil chains (Figure 7.4). The crankshaft chain is a key building unit in zeolites, in both single and double chains. Adjacent crankshaft chains can be related either by a mirror plane (giving the double crankshaft chain) or by a center of inversion (giving the narsarsukite chain, observed in zeolites but particularly common in aluminophosphates). The important natural zeolites of the natrolite family contain the 4=1 chain and zeolite ZSM-5 the pentasil chain. Among all the SBUs, the sodalite cage, or β-cage, is of great importance, because it is a key SBU in two of the most important zeolites, A and Y. We will therefore use the class of zeolites containing sodalite cages to illustrate some of the key features of zeolite structural chemistry.

(a)

(b)

(c)

(d)

Figure 7.4 Chains found in zeolite structures: (a) zig-zag, (b) sawtooth, (c) double crankshaft, (d) narsarsukite, (e) natrolite 4=1, and (f) pentasil chains. For clarity, in this and most of the following framework representations, single lines link the tetrahedral atom (T-atom) positions and O atoms are omitted.

(e)

(f)

7.2 Zeolite Structure Types Exempliﬁed by Those Based on the Sodalite Cage

7.2.3 Assembling Sodalite Cages: Sodalite, A, Faujasites X and Y, and EMC-2

The mineral sodalite is a so-called ‘‘feldspathoid,’’ the semi-open structure of which consists entirely of face-sharing sodalite cages (β-cages) (Figure 7.5a). The sodalite cages are connected through six-rings which permit the passage of water but no other molecules [4]. Nevertheless, the ability of the sodalite cage to host brightly colored sulﬁde radicals gives rise to important blue and purple pigments and semi-precious stones (ultramarine, Lapis lazuli, etc.) [3]. If the sodalite cages

(a)

(b)

(c)

(d) Figure 7.5 Arrangement of sodalite cages in (a) sodalite, (b) zeolite A, (c) faujasite (with supercage, right), (d) EMC-2 (with supercage, right).

177

178

7 Structural Chemistry of Zeolites

are linked through their four-rings via D4R units, the zeolite A framework results (topology type LTA, Figure 7.5b). Besides D4Rs and β-cages, the structure contains α-cages, which share eight-ring openings that can allow the uptake of small ˚ In the aluminum-rich zeolite A (Si/Al = 1) molecules with molecular sizes up to 4 A. many extra-framework cations are required for charge balance, and these are distributed between the β- and α-cages. Those in the α-cage can inﬂuence the pore size if they are close to the window. For Na–A the effective pore size is around 4 A˚ (‘‘zeolite 4A’’), whereas for the potassium form, K–A, the larger cations restrict the effective window size (‘‘zeolite 3A’’). If Na+ is exchanged by half the number of Ca2+ cations, the effective pore size is increased (‘‘zeolite 5A’’). Sodalite cages can also be assembled into frameworks by being linked through D6Rs on their six-ring faces (Figures 7.5c,d). Whereas there is only one way of arranging the cages for sodalite and zeolite A, there are different ways of linking layers of sodalite cages through D6Rs. The two end member variants of stacking of layers of sodalite cages are cubic zeolites with the FAU structure type [12], where FAU refers to the mineral form of this material, faujasite and its hexagonal polytype EMC-2 (structure type EMT) [4]. In faujasitic materials the Si/Al ratio can vary from 1.1 to inﬁnity (for X, Si/Al = 1.1–1.8, for Y, Si/Al = 1.8–∞). Disordered stacking sequences are also observed [13]. Which polytype crystallizes is determined by the cationic and other templating species included in the synthesis gel. Besides the D6R and β-cages, larger cavities, or supercages, are formed. 7.2.4 Faujasitic Zeolites X and Y as Typical Examples

The faujasitic zeolites X and Y provide excellent examples of many of the important types of zeolite structural chemistry. They can be prepared directly with Si/Al ratios from 1.1 to 5, and the Si/Al ratio can be increased by post-synthetic treatments, even to pure silica solids. In Al-containing X and Y the siting of extra-framework cations introduced during synthesis or subsequently by cation exchange is very important in the determination of properties, and has been extensively studied. Some of the sites observed in these studies are shown in Figure 7.6. These include sites in the D6Rs, in the sodalite cages, and in the supercages. If the deammoniation of the NH4 -exchanged Y is performed under carefully controlled conditions, the protonic form of the zeolite is prepared. Neutron diffraction studies of this solid have shown that the protons are located on bridging oxygen atoms, Si–OH–Al, giving strong Bronsted acid sites [14]. One of the most important uses of zeolite Y is as a solid acid catalyst for oil cracking. However, the need for both hydrothermal stability and acidity in working catalysts requires very high Si/Al framework ratios. This can be achieved by the process of ultrastabilization, in which the ammonium form is heated in an atmosphere containing steam. The result is deammoniation and the preparation of Bronsted acidic bridging protons on the framework. Under these conditions, Al is removed from the framework to give extra-framework aluminum in the pores and this leaves vacant tetrahedral sites. Simultaneously, Si can leave lattice sites and migrate to ﬁll

7.2 Zeolite Structure Types Exempliﬁed by Those Based on the Sodalite Cage

SI′ SIII

SI

SII

SII′

Figure 7.6 Cation sites (SI, SI , SII, SII , and SIII) observed in zeolites X and Y. (Reproduced with permission from [3]).

these vacant sites, resulting in a more hydrothermally stable framework with an increased Si/Al ratio and secondary mesopores [3]. A similar process is observed for some other zeolites (such as zeolite Rho), but for zeolites with very high aluminum contents there is insufﬁcient Si to ﬁll the vacancies and the structure collapses. Where secondary mesoporosity is observed, it improves access of molecules to the interior of the crystallites and has an important beneﬁcial effect in catalytic reactions. 7.2.5 Key Inorganic Cation-Only Zeolites Pre-1990

Once the pioneers of synthetic zeolite chemistry had shown that it was possible to prepare aluminosilicate zeolites such as A in the laboratory, the way was open for an exhaustive investigation of the use of inorganic cations as SDAs in these syntheses, varying cation ratios, Si/Al ratios, temperature, and time. Of the synthetic zeolites produced in this work, mordenite, zeolite L, chabazite and related zeolites, and ZK-5 and Rho are important examples [4]. Mordenite is typically prepared with a Si/Al ratio around 5–10 from Na+ -containing gels. It has a framework that contains both 4-rings and 5-rings, giving a system of one-dimensional 12-ring channels connected via ‘‘side-pockets’’ with staggered 8-ring openings to adjacent 12-ring channels (Figure 7.7a). Extra-framework cations are mainly located in these side-pockets. Mordenite is hydrothermally very stable and secondary mesoporosity can be introduced, improving transport between the large channels. These properties, together with

179

180

7 Structural Chemistry of Zeolites

(a)

(b) Figure 7.7 Framework topologies of (a) mordenite and (b) zeolite L, in each case viewed down the 12MR channel.

strong solid acidity when in the protonic form, have led to its application in petrochemical conversion processes. Zeolite L is a typical product of syntheses in the presence of K+ cations, with a moderate Si/Al ratio of about 3 (Figure 7.7b). The key structural elements of the framework are columns of cancrinite cages alternating with D6Rs that link to give a hexagonal zeolite with one-dimensional 12-ring channels. In between the 12-ring openings, the channels open out to a larger diameter. K+ cations adopt sites in the cancrinite cages (suggesting a strong templating role), in the intercage regions and in the large channels. The zeolite chabazite can also be prepared from K+ -containing gels, and in the + Li -exchanged form it has important properties in the noncryogenic separation of oxygen and nitrogen in air. Structurally, chabazite is one of a large family of synthetic and natural zeolites that can be thought to be built up entirely from six-rings, stacked in well-ordered sequences according to their position (A, B, or C) at (0,0), ( 23 , 13 ), or ( 13 , 23 ) in the xy plane of a hexagonal cell (Figure 7.8a). In chabazite, all six-rings are part of D6Rs: the stacking sequence being AABBCC. All D6Rs have the same orientation, and link to other D6Rs to give a structure that contains

7.2 Zeolite Structure Types Exempliﬁed by Those Based on the Sodalite Cage

(a)

(b)

(c) Figure 7.8 Arrangement of 6MR SBUs giving (a) chabazite, (b) erionite, and (c) offretite topologies, with associated cages and channel (right).

the chabazite cage. Each chabazite cage is linked to six others via eight-ring windows giving a highly porous, three-dimensionally connected eight-ring structure. Variation in the six-ring stacking sequence gives structures with cages and channels of different sizes: the cage and channel of erionite and offretite, respectively, two other zeolites of this family that are prepared with K+ and Na+ cations that have the stacking sequences AABAAC and AAB, are shown in Figure 7.8b,c. The zeolites ZK-5 and Rho are two cubic, small-pore zeolites with high pore volume, which contain α-cages (Figure 7.9). Both are prepared with alkali-metal cations, including K+ and Cs+ . In ZK-5, as in chabazite, the structure can be thought to be built entirely from D6Rs, giving rise to α- and pau cages that alternate along the crystallographic axes, giving a cage structure connected by planar eight-ring windows. In zeolite Rho the structure can be envisaged as α-cages linked via D8Rs, giving two interpenetrating pore volumes. Each pore volume is itself connected via eight-rings, but the two different pore systems are not connected.

181

182

7 Structural Chemistry of Zeolites

(a) Figure 7.9

(b) Arrangement of α-cages in the cubic zeolites (a) ZK-5 and (b) Rho.

There is no inaccessible pore space in the zeolite Rho structure and so it has a high pore volume. When in the Cs+ -form, it shows structural ﬂexibility upon dehydration, as the D8Rs distort to allow better coordination to the Cs+ cations [15]. 7.2.6 Structures Templated by Simple Alkylammonium Ions

Once the success of using alkali-metal cations as SDAs was apparent, it was not long before organic alkylammonium cations were investigated for the same purpose. In the period before the 1990s, these were mainly commercially available ions, especially tetraalkylammonium ions NR4 + (R = CH3 , C2 H5 , C3 H7 , and C4 H9 : TMA+ , TEA+ , TPA+ , and TBA+ , respectively). Using these, often in combination with alkali-metal cations, a new series of zeolites with medium to high Si/Al ratios was obtained [4]. Using TMA+ gives, in addition to the structures with the LTA topology with Si/Al > 1 (ZK-4), the important structures EAB, gmelinite, and ZSM-4 (which has the structure of the mineral mazzite), each of which contains cages. The cages in these structures (gmelinite cages are present in both gmelinite and ZSM-4) are shown in Figure 7.10. Among the most important new structure types prepared with TEA+ were zeolite Beta and ZSM-12. Although it was one of the ﬁrst zeolites to be prepared, the structure of zeolite Beta deﬁed solution until 1988 due to its disorder [16]. The structure is built up from layers with tetragonal symmetry that can be connected via stacking offsets of ±a/3 and ±b/3. In zeolite Beta, these offsets occur with a high degree of disorder, but this does not block the straight 12-ring channels parallel to the layers (shown in Figure 7.11), or the undulating and interconnected 12-ring channels that run perpendicular to these. The high silica content of Beta infers to it a high degree of stability, and by virtue of its 3D connectivity and ease of synthesis it is one of the most important zeolites used in catalysis. ZSM-12 is another zeolite that crystallizes in the presence of TEA+ . It has a 1D 12-ring channel system, the projection of which is similar to that of the straight channel in zeolite Beta.

7.2 Zeolite Structure Types Exempliﬁed by Those Based on the Sodalite Cage

(a) Figure 7.10

(a)

(b) Cages found in (a) EAB and (b) gmelinite (and ZSM-4 or mazzite).

(b)

Figure 7.11 The topology of zeolite Beta, viewed along one set of its straight 12MR channels (a). A single layer of the Beta structure is shown (b).

One of the major breakthroughs in zeolite chemistry came with the synthesis of ZSM-5 using TPA+ cations as SDAs. This solid possesses a three-dimensionally connected 10MR pore system consisting of intersecting straight and undulating medium-pore channels [17]. The pore structure, together with the high stability and strong acid strength that result from its high Si/Al ratio (reaching inﬁnity for its pure silica polymorph, silicalite-1), makes it a highly active and shape-selective catalyst, particularly in conversions of monoaromatics. The structure of ZSM-5 is built up from the pentasil units shown in Figure 7.12. These link to form chains, which in turn link to form sheets. The ZSM-5 structure results when these sheets are linked across a center of inversion, as shown in Figure 7.12, with the TPA+ cations located at the channel intersections. Straight 10-ring channels run parallel to the sheets, and are connected by undulating channels that lie in the plane parallel to them. Although there are only two sets of channels, they are connected so that any part of the pore space in a crystal is accessible to any other. If the same sheets are stacked so that they are related to adjacent sheets by mirror planes, a different structure, ZSM-11, is formed, also with a 3D 10-ring channel system. ZSM-11 is

183

184

7 Structural Chemistry of Zeolites

(a)

(b)

(d)

(c)

(e)

Figure 7.12 The ZSM-5 (MFI) structure can be built up successively: (a) the pentasil unit; (b) chains of pentasil units; (c) layers of these chains; and (d) layers linked across inversion centers. If the layers are linked across mirror planes, (e), ZSM-11 results.

prepared in the presence of TBA+ cations, among other SDAs. The observation that different structure types can often be formed from the same sheets, but with different stacking sequences, is a common one and the ‘‘Atlas’’ lists several families of this type [4]. In this ﬁrst period of zeolite structure synthesis in the presence of organic SDAs, initial efforts were made to use more complex alkylammonium ions. Casci used a series of diquaternary cations of the form [(H3 C)3 N(CH2 )n N(CH3 )3 ]2+ (where n = 3–10) and discovered the high silica 1D 10-ring zeolite EU-1 [18], which was templated by hexamethonium ions (n = 6). This was one of the ﬁrst examples of the use of more complex SDAs. 7.2.7 Lessons from Nature

From the elucidation of the structure of the natural mineral sodalite in 1930 to the present day, natural zeolites have provided an inspiration to zeolite chemists. Many

7.3 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features

of the commercially important zeolites have been observed as minerals, either before or after they have been synthesized in the laboratory, and frequently as crystals suitable for single-crystal diffraction. The structures of ferrierite, mordenite, faujasite, and mazzite, for example, were all determined from the minerals, whereas for mutinaite (MFI type), tschernichite (*BEA), and direnzoite (ECR-1, EON) the crystal structures were ﬁrst solved from synthetic samples before their natural counterparts were unearthed. For zeolite chemists, the observation of mineral zeolite structures with attractive features but no synthetic counterpart acts as a spur, because it indicates that their synthesis should be feasible, and furthermore that this should be possible without the use of organic templates, which are not thought to be involved under geological conditions. The mineral boggsite, for example, ﬁrst found and solved in 1990, has a 2D channel system in which 12-ring and 10-ring channel systems intersect. The possible advantages such a pore system could possess in terms of shape selectivity make it an attractive target. Subsequent synthetic studies have realized this 12-ring × 10-ring feature of pore systems in several novel solids. More recently, discovery of the aluminosilicate mineral direnzoite, an ordered intergrowth of mordenite and mazzite sheets [19], strongly suggests that this should be preparable without the complex organic that was initially used for its synthetic counterpart ECR-1 [20]. Recent reports suggest this to be the case [21]. Finally, some zeolite mineral structures remain without synthetic counterparts. Examples include boggsite, terranovaite (which has a 2D 10-ring channel system), and tsch¨ortnerite (TSC). TSC, which like CHA and KFI is composed entirely of D6Rs (Figure 7.13) [4], has a remarkable interconnected pore system containing β-cages, α-cages, and a larger supercage, and its synthesis remains an attractive target.

7.3 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features 7.3.1 Introduction

The discovery that mono, di, and triquaternary alkylammonium cations, some prepared speciﬁcally, could direct the crystallization of new zeolite structures prompted several research groups to prepare a wide variety of such species with novel geometries and to screen them as SDAs in a variety of gel compositions. This approach has since been extended by the use of phosphonium ions [22]. The groups of Zones (Chevron) and Corma (ITQ, Valencia) pioneered these studies and realized that expert synthetic organic chemistry was an integral part of them. Other academic and industrial research groups have also made important and sustained contributions in this area, including those of Davis (Caltech), Hong (Taejon), and Xou (Stockholm) and groups at ExxonMobil, UOP, Mulhouse and the Institut Francais du Petrole. Examples of their work are referenced in the following text

185

186

7 Structural Chemistry of Zeolites

(b)

(a)

(c) Figure 7.13 The tsch¨ortnerite mineral’s framework structure (a) contains the tsch¨ortnerite supercage (b), a combination of D6Rs, β-cages, α-cages, and the supercage (c).

alongside some of the structures they have produced. In parallel with this, strategies of using combinations of organic bases as potential co-template mixtures have also had important results in the synthesis of zeolites and related solids, particularly as shown via recent high-throughput studies by Blackwell et al. at UOP [23]. In addition to the use of designed potential SDAs, modiﬁcation of the inorganic components of the gel has been found to play a key role in phase selectivity. The work of Camblor at the ITQ on the introduction of ﬂuoride ions into low water content synthesis gels gave many new porous silica polymorphs of zeolites with large pore volumes, where the ﬂuoride ions have a dual role of assisting silicate condensation and crystallization and balancing the positive charge on the framework [24]. Early examples of the success of this model for new structures include the highly porous small-pore zeolite ITQ-3 (with 2D 8-ring pore system), the large-pore 1D channel structure ITQ-4 (there are now many 1D 12-ring channel structures known), and the 3D 12-ring ITQ-7 [25]. This approach has subsequently been used very productively by many researchers in the ﬁeld. The structural role of the ﬂuoride is discussed in Section 7.3.3. Changing the composition and elemental ratios of framework-forming cations has also been found to exert a large inﬂuence on the phase to form. Zones has investigated the effect of variation of Si/Al and Si/B ratios in the gel as additional

7.3 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features

parameters in the syntheses with new potential SDAs [26], and Corma (and the Mulhouse and Stockholm groups) found that the inclusion of germanium had a strong structure-directing inﬂuence, because of its propensity to favor the formation of D4Rs [11, 27, 28] . Furthermore, the addition of inorganic cations and variation of alkalinity (OH− /T) have been shown to have an important inﬂuence on the zeolite structure to form [29]. A combination of these innovative synthetic strategies has been responsible for the upsurge in reported novel synthetic zeolite types since 1990 (Figure 7.2). The need for state-of-the-art structural characterization, combined with specialist organic synthesis and high-throughput screening of potential SDAs and mixtures of SDAs under a very wide range of synthetic variables, means that the synthesis of ever more complex structures is a specialized enterprise. This approach has brought many signiﬁcant structural highlights in the last decade or so, including larger pore sizes, new pore connectivities, increased structural complexity, a broadening of the compositional range of known structure types, and chiral structures. In the second main part of this chapter we summarize some of the most important developments. 7.3.2 Novel Structures and Pore Geometries

One of the most obvious advances in new structural chemistry has been the preparation of zeolites with extra-large pores, larger than the 12-rings found in faujasites and Beta (Table 7.1). The ﬁrst 14-ring zeolite, UTD-1 [30] (University of Texas, Dallas), was prepared using the permethylated cobalticenium ion (Figure 7.14). Other 14-ring window pure silicate and germanosilicates were prepared at Caltech (CIT-5) [31], at Chevron, (SSZ-53 and SSZ-59) [32], and at Mulhouse (IM-12) [33], and a beryllosilicate (OSB-1) with 14-ring openings in the framework has also been synthesized [34]. Furthermore, two silicates with channels bounded by 18-rings have been reported, the gallosilicate ECR-34 [35] and most recently the germanosilicate ITQ-33 [36]. The dimensions of the pore openings as deﬁned crystallographically are given in Table 7.1. The large-pore nature of ECR-34 has been demonstrated by the adsorption of large hydrocarbons such as perﬂuorotri-n-butylamine. The gallium and germanium contents have important roles in directing these structures, for example, in favoring D4Rs in ITQ-33, and tend to reduce their overall hydrolytic stability. Nevertheless, these structures signpost the way to thermally regenerable extra-large-pore acid catalysts. Finally, a germanosilicate ITQ-37 with channels linked by highly noncircular 30-rings has recently been reported [37]. This is described in more detail in Section 7.3.5. Many new structure types with 3D pore connectivity have resulted from these structures, including those with connectivity in all dimensions via openings of 10-ring or greater (Table 7.2). These are of great interest as adsorbents and catalysts because of their enhanced molecular transport properties and their resistance to blocking in catalytic reactions. TNU-9 [38] and SSZ-74 [39], for example, add to the important class of zeolites with 3D 10-ring channel systems, previously

187

188

7 Structural Chemistry of Zeolites Table 7.1

Window sizes of extra-large pore silicates.

Zeolite (code)

Framework composition

UTD-1 (DON) CIT-5 (CFI) SSZ-53 (SFH) SSZ-59 (SFN) IM-12 (UTL) OSB-1 (OSO) ECR-34 (ETR) ITQ-33

SiO2

ITQ-37

Connectivity (MRs)

Window dimensions of ˚ the largest pore (A)

Reference (year)

14

(8.2 × 8.1)

[30] (1999)

SiO2

14

(7.5 × 7.2)

[31] (1998)

Si0.97 B0.03 O2

14

(8.7 × 6.4)

[32] (2003)

Si0.98 B0.02 O2

14

(8.5 × 6.2)

[32] (2003)

Si0.82 Ge0.18 O2

14 × 12

(9.5 × 7.1) × (8.5 × 5.5)

[33] (2004)

Si0.66 Be0.33 O2

14 × 8 × 8

[34] (2001)

Si0.75 Ga0.24 Al0.01 O2

18 × 8 × 8

(7.3 × 5.4) × (3.3 × 2.8) × (3.3 × 2.8) (10.1) × (6.0 × 2.5)

Si0.66 Al0.04 Ge0.30 O2

18 × 10 × 10

Si0.58 Ge0.42 O2

30 × 30 × 30

(12.2) × (6.1 × 4.3) × (6.1 × 4.3) (19.3 × 4.9) × (19.3 × 4.9) × (19.3 × 4.9)

[35] (2003) [36] (2006) [37] (2009)

exempliﬁed by ZSM-5 and ZSM-11. They are the most complex structures yet observed and are discussed further in Section 7.3.4, along with the 2D 10-ring zeolite IM-5 [40]. In addition, new 3D-connected 12-ring silicates ITQ-17 [25] and germanosilicates ITQ-17 [41], ITQ-21 [42], and ITQ-26 [43] add to the previously known structures with this connectivity, the faujasite and Beta structures (see, for example, Figure 7.15). ITQ-17 is related to the disordered Beta structures originally prepared, having the same framework layers, but these are stacked differently, in an ordered tetragonal arrangement and including D4Rs. In this structure, originally hypothesized and named Beta polymorph C (Beta C), the three perpendicular 12-ring channel systems intersect at the same place. Originally prepared as a germanate FOS-5 [44], this has more recently been prepared as the germanosilicate ITQ-17. ITQ-21 and -26 also possess 3D 12-ring channel systems, and here again the D4Rs typical of germanium-containing silicates are a crucial structural element. One of the most important novel classes of 3D-interconnected channel structures that has been prepared contains 12-ring channels intersecting with 10-ring channels. The zeolites CIT-1 [45], ITQ-24 [46], and MCM-68 [47] are examples of this type of framework, each prepared with a complex SDA. There is considerable interest in investigating possible novel shape-selective catalytic performances in this type of structure. Other novel solids with 3D connectivity include ITQ-33 (18 × 10 × 10). In addition to these new structures with 3D-connected porosity,

7.3 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features

Table 7.2 Recent (post-1990) tetrahedrally connected zeolite structures with 3D connectivity via at least 10MR openings, compared with ZSM-5, Y, and Beta zeolites.

Zeolite (code)

Framework composition

Space group

Pore system (MR)

Dimensions ˚ (A)

Reference (year)

Faujasite (FAU) Beta (*BEA)

Si1−x Alx O2 (x = 0–0.4) Si1−x Alx O2 (x = 0–0.05)

Fd3m

12 × 12 × 12

P4122

12 × 12 × 12

[12] (1958) [16] (1988)

ZSM-5 (MFI) ITQ-7 (ISV)

Si1−x Alx O2 (x = 0–0.05) SiO2

Pnma

10 × 10*

P42 /mmc

12 × 12 × 12

CIT-1 (CON)

Si0.96 B0.04 O2

C2/n

12 × 12 × 10

ITQ-17 (BEC)

Si0.64 Ge0.36 O2

P42 /mmc

12 × 12 × 12

ITQ-21 (no code) ITQ-24 (IWR)

Si0.66 Ge0.34 O2

Fm-3c

12 × 12 × 12

Si0.89 Ge0.09 Al0.06 O2

Cmmm

12 × 10 × 10

MCM-68 (MSE)

Si0.90 Al0.10 O2

P42 /mnm

12 × 10 × 10

ITQ-33 (no code) TNU-9 (TUN) SSZ-74 (-SVR) ITQ-26 (IWS) ITQ-37 (no code)

Si0.66 Al0.04 Ge0.30 O2

P6/mmm

18 × 10 × 10

Si0.95 Al0.05 O2

C2/m

102 × 10*

Si0.96 0.04 O2

Cc

10 × 10*

Si0.8 Ge0.2 O2

I4/mmm

12 × 12 × 12

Si0.58 Ge0.42 O2

P41 32 or P43 32

30 × 30 × 30

(7.4) × (7.4) × (7.4) (6.7 × 6.6) × (6.7 × 6.6) × (5.6 × 5.6) (5.1 × 5.5) × (5.3 × 5.6) (6.5 × 6.1) × (6.5 × 6.1) × (6.6 × 5.9) (7.0 × 6.4) × (7.0 × 5.9) × (5.1 × 4.5) (6.7 × 6.6) × (6.7 × 6.6) × (5.6 × 5.6) (7.5) × (7.5) × (7.5) (5.8 × 6.8) × (4.6 × 5.3) × (4.6 × 5.3) (6.8 × 6.4) × (5.8 × 5.2) × (5.2 × 5.2) (12.2) × (6.1 × 4.3) × (6.1 × 4.3) (5.6 × 5.5),(5.5 × 5.1) × (5.5 × 5.4) (5.9 × 5.5) × (5.6 × 5.6) (7.05) × (7.3 × 7.0) × (7.3 × 7.0) (19.3 × 4.9) × (19.3 × 4.9) × (19.3 × 4.9)

* indicates

indicates

that 3D connectivity is achieved via the intersection of two channel systems. a tetrahedral cation vacancy.

[17] (1978) [25] (1999) [45] (1995) [41] (2001) [42] (2002) [46] (2003) [47] (2006) [36] (2006) [38] (2006) [39] (2008) [43] (2008) [37] (2009)

189

190

7 Structural Chemistry of Zeolites

(a)

(b)

(c)

(d)

(e)

(g)

(f)

Figure 7.14 Projections down the extra-large pore channels of the 14MR zeolites (a) UTD-1, (b) CIT-5, (c) SSZ-53, (d) SSZ-59, and (e) IM-12 and of the 18MR zeolites (f) ECR-34, and (g) ITQ-33.

several new 2D-connected materials have been prepared and their structures solved (MCM-22, ITQ-3, ITQ-13, ITQ-22, SSZ-56, etc.) [4]. Besides possessing interesting gas adsorption and catalytic properties in their calcined forms, post-synthetic treatment is a possible route to secondary mesoporosity to increase the dimensionality of molecular transport. Finally, new structures that show chirality have been prepared and are described in Section 7.3.5: the chiral and mesoporous ITQ-37 is a remarkable example. Another major result that has been achieved by these synthetic studies is a widening of the available compositional range of zeolites with known structures. The pure silica version of zeolite A (ITQ-29) [48] has been prepared by using organic species (with low charge densities) in the syntheses, rather than Na+ cations. This silica shows much higher hydrolytic stabilities than zeolite A. Similar results have

7.3 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features

(a)

(b)

Figure 7.15 Projections down 12MR channels of (a) ITQ-17 (BEC) and (b) ITQ-26, both of which possess 3D-connected 12MR channel systems.

been achieved in UZM-4 (BPH) [23], which is a higher silica and more stable version (Si/Al > 1.5) of Linde Q (BPH, Si/Al = 1.1) [4]. In a similar way, templating studies have given silicate versions of structures initially prepared as germanates or aluminophosphates. Among the latter, SSZ-16, SSZ-24, SSZ-55, and SSZ-73 are the silica versions of AlPO4 -56, AlPO4 -5, AlPO4 -36, and STA-6, respectively [4]. It is likely that silicate versions of other aluminophosphates without existing zeolite analogs will also be obtained: the silicate versions of the larger pore, 3D-connected DAF-1(DFO) and STA-14(SAO) [4] would be of particular interest. 7.3.3 Expansion of the Coordination Sphere of Framework Atoms

Although zeolites are deﬁned by their tetrahedrally connected frameworks, there are some examples where their framework cations adopt higher coordination. The most important of these are when Si expands its coordination by bonding to ﬂuoride during synthesis and when framework Ti expands its coordination, for example, with water or upon uptake of hydrogen peroxide. Whereas the ﬁrst observation is important in understanding ﬂuoride synthesis [24], the second has important catalytic consequences [49]. The ﬂuoride ion behaves as an efﬁcient mineralizing agent in the synthesis of pure silica zeolites [24], where it catalyzes the hydrolysis of silica and enables the formation of silicate frameworks at pH values of 7–9, where no reaction would occur in its absence. A series of crystallographic and NMR studies have shown that in silicas prepared in ﬂuoride media, F− ions are present in the as-prepared solids, coordinated to lattice silicon atoms, where they raise the coordination number of Si to 5 (SiO4 F) [50]. The ﬂuoride ion is often found to occur within small cages, distributed over a number of different silicon sites with partial occupancy. For example, it is found in [46 ] cages (LTA, AST) and also in cages in nonasil [41 54 62 ], EU-1 [41 54 62 ], silicalite (MFI [41 52 62 ], ITQ-4 (IFR [42 64 ]) SSZ-23 [43 54 ], and so on. This is illustrated in Figure 7.16 for F− in the [41 54 62 ] cage of EU-1 [51]. In this

191

192

7 Structural Chemistry of Zeolites

Figure 7.16 Part of the as-prepared EU-1 structure prepared in ﬂuoride medium. The ﬂuoride ion is located within a small cage and is connected to one of the Si atoms, raising its coordination ﬁvefold. (Reproduced with permission from [50]).

way F− acts as an inorganic SDA. At the site where it coordinates to the silicon, the tetrahedrally arranged SiO4 group is distorted so that the three OTO angles that are closest to the F− ion are increased to minimize the O–F repulsion [8]. Once incorporated into the structure, the F− ion balances the positive charge of the alkylammonium ion template. Calcination results in simultaneous removal of both the organic cation and the F− , leaving a SiO2 framework with very few framework defects. Such materials are hydrophobic and as a direct result have potential applications in adsorption. Titanium can substitute for Si at low concentrations in pure silica zeolites, and adopts tetrahedral coordination once the as-prepared solid is calcined and dehydrated. Upon exposure to, for example, aqueous hydrogen peroxide [52], it expands its coordination, acting as a Lewis acid, and the local geometry is distorted. In this way, titanosilicate zeolites can act as important oxidation catalysts, especially when they activate hydrogen peroxide [49].

7.3 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features

7.3.4 The Current Limits of Structural Complexity in Zeolites

The ongoing synthetic efforts described above continue to give rise to structures with increasing complexity, both in terms of their crystallographic description and the diversity of their framework architecture. In this section, we discuss in detail the current limits of structural complexity, as deﬁned by the number of crystallographically distinct tetrahedral cation environments (T-sites) present in the repeat unit of the structure. In terms of crystallographic complexity, the structures of both zeolite A and zeolite Y are very simple, for although there are many tetrahedra within the unit cells of each framework, they are all related by the many symmetry elements present in the two structures, so there is only one unique position in each case. By contrast, the structure of ZSM-5 is far more complex and its framework structure is built from layers with 12 crystallographically distinct T-sites. As described in Section 7.3.1, each layer is linked to symmetrically equivalent layers by centers of symmetry. (In fact, high silica materials with this structure distort to lower symmetry at low temperatures so that they have 24 different sites, but all the sites retain their original coordination sequences and the distortion is a subtle one) [53, 54]. Recently two zeolite structures have been solved which have 24 crystallographically distinct sites, each with a different coordination sequence, TNU-9 [38] and IM-5 [40]. A third reported structure, SSZ-74 [39], is similarly complex, with 23 different T-sites and an ordered vacancy. None forms as single crystals and their complexity proved a major challenge to structure solution. Fortunately, in a tour-de-force of powder crystallography, by combining experimental X-ray diffraction (XRD) and phases determined from high-resolution electron microscopy with crystallographic computing algorithms, the groups of McCusker, Baerlocher, and Terasaki (for TNU-9) and McCusker, Baerlocher, and Zou (for IM-5 and SSZ-74) have successfully elucidated the structures of these materials. For TNU-9, high-resolution powder XRD and HRTEM have been successfully combined with the FOCUS program, designed to search trial electron density maps for tetrahedra. Sufﬁcient correct phase information was obtained from the electron micrographs to make this possible. For IM-5 and SSZ-74, similar types of experimental diffraction and imaging data have been combined with so-called enhanced charge ﬂipping crystallographic algorithms adapted from other application programs. These charge ﬂipping programs can be applied without applying structural constraints to obtain solutions (and so are generally applicable, regardless of framework connectivity). The structure of TNU-9 has a very similar projection (down [010]) to that observed for ZSM-5, but has two different sets of straight 10-ring channels, labeled A and B in Figure 7.17, rather than one. Perpendicular to [010], channels of type B are linked to each other via short 10-ring channels and to channels of type A via 10-ring windows. It is instructive to investigate how a structure as complex as TNU-9 is built from repeating units, and as a consequence to suggest how it may assemble from solution, in the presence of organic SDAs [55]. The TNU-9 framework can be built up from a single kind of chain, similar to what is seen in ZSM-5. These

193

194

7 Structural Chemistry of Zeolites

B

A

y x

(a)

z

4.9 Å

5.4 Å

y z x y

x

(b)

(c)

z

Figure 7.17 The complex topology of TNU-9 (a) contains two kinds of channel (b), as depicted in pink and blue, which show complex 3D connectivity. The framework itself is built from one kind of chain that links to form layers (c), which join to give the ﬁnal structure. (Reproduced with permission from [38] and [52]).

chains are connected via mirror planes to give sheets which are asymmetric (i.e., one side is different from the other). These sheets can only be linked to each other via their similar sides, so that there are two different intersheet regions. Modeling indicates that the organic SDA is able to interact favorably in different sites in both intersheet regions, so indicating a mechanism by which stacking of layers can be favored during synthesis [55]. The high silica zeolite IM-5 is, like ZSM-5 and TNU-9, highly thermally stable and shows interesting performance for hydrocarbon processing and selective NO reduction [40]. The structure has a 2D-connected 10-ring channel system. It is, like TNU-9, similar in projection to ZSM-5 but rather than complete 3D ˚ that contain channels that show connectivity there are slabs (thickness 25 A) connectivity in three dimensions, complete with complex channel intersections. Each slab is isolated from adjacent sheets by impermeable silicate sheets. The catalytic performance could only be properly explained once this connectivity was understood. The work also described how, because the charge ﬂipping structure solution methodology did not need to make use of symmetry, it determined the location of 288 silicon atoms and 576 oxygen atoms in the unit cell. This is very encouraging, as ever more complex structures are prepared, and suggests that an inability to grow single crystals should not limit our discovery of their intricate

7.3 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features

architectures. Both TNU-9 and IM-5 crystallize in the presence of the same organic SDA, bis-N-methylpyrrolidiniumbutane, which has also been shown to favor the crystallization of several other zeolites depending on solution pH, Si/Al ratio, additional cations, and so on [29]. This is a clear indication that the detailed composition of the hydrothermal reaction mixture plays a crucial role in supplying species for the growth of these complex solids. Finally, a third highly complex zeolite, SSZ-74, prepared using a template very similar to that used for TNU-9 and IM-5 (bis-N-methylpyrrolidiniumheptane) has recently been solved by the same combination of XRD, TEM, and application of the charge ﬂipping algorithm used successfully for IM-5. The structure possesses 23 different T-sites and an ordered tetrahedral vacancy (effectively a 24th T-site that is vacant). The structure contains an undulating 10-ring channel connected by straight 10-ring channels and leading to 3D connectivity limited by 10-rings similar to that for ZSM-5. The implications of the observed vacant site are discussed in more detail in Section 7.3.6. 7.3.5 Chirality and Mesoporosity

As described above, considerable progress has been made toward the synthesis of zeolite structures with different connectivities and larger pore sizes, at least up to the 12.2 A˚ circular pore openings observed in ITQ-33. One of the key remaining challenges is to produce batches of chiral zeolites which consist entirely of crystals of one enantiomer, so that these might ﬁnd application in enantioselective separation and catalysis. Very few silicate zeolites with potentially chiral porous structures are known, and the most important examples are described below. Zeolite beta, described in Section 7.3.3, typically exhibits stacking faults and does not crystallize as an ordered polymorph. Theoretically, though, there is a regular stacking sequence that would give a polymorph (Beta-A) that would be chiral, and efforts have been made to obtain this by the use of chiral templates. No fully ordered chiral polymorph A has yet been prepared, and one of the difﬁculties is to achieve chiral recognition over the long helical pitch of the channel that this polymorph would have, using either single molecules or molecules that could order. More recently, two zeolitic silicates have been prepared as mixtures of chiral crystals. The silicogermanate SU-32 is one of a family of fully tetrahedrally connected frameworks templated by the achiral ammonium ion (CH3 )2 CHNH3 + that are built up from chiral layers with 12-ring openings that consist of ‘‘4−1’’ repeating units of tetrahedra, which point up and down alternately (Figure 7.18) [56]. Relating adjacent layers of this type across inversion centers results in the achiral 12-ring structure SU-15, whereas stacking layers with a ±60◦ rotation between adjacent sheets (which enables fortuitous coincidence) results in chiral polymorphs of SU-32 (space groups P61 22 or P65 22). Remarkably, crystals form in a mixture of pure enantiomorphs, which is different from what is observed for zeolite Beta. The resulting structure comprises only D4Rs and 46 58 82 102 cavities that share 10-ring openings and make up helical channels which are either right

195

196

7 Structural Chemistry of Zeolites

or left handed, with pore openings 5 × 5.5 A˚ along the channels and intersected ˚ The by eight-ring channels that run parallel to the channel axis (4.7 × 3 A). challenge now is to prepare this topology in the form of a more stable silicate or aluminosilicate, and even to prepare only one of the two enantiomeric forms. The second type of zeolitic solid that crystallizes in chiral form is the silicogermanate ITQ-37 that crystallizes in the chiral space groups P41 32 and P43 32, with a structure related to the gyroidal (G) periodic minimal surface. This G-surface is exhibited by micelle-templated mesoporous amorphous silicas. In these mesoporous solids two pore systems of interconnected channels of opposite hand are separated by a silica wall located at the G-surface, whereas ITQ-37 can be thought of as having one of these pore systems empty and the other ﬁlled by a chiral zeolitic framework. The remarkable structure is beautifully illustrated by Sun et al. [37], and the cavity of this structure and the entrance window are illustrated in Figure 7.19. Besides possessing chiral channels, the structure has the lowest framework density observed for a zeolite (10.2 T/1000 A˚ 3 ) and cavity dimensions in the mesoporous ˚ As in other germanosilicate structures, D4Rs are important SBUs. regime (>20 A). In ITQ-37 these have one or two terminal hydroxy groups, and this interrupted nature of the framework is crucial in enabling the large cages to form. The generation of a material that is both intrinsically chiral and with mesoporous cavities is a signiﬁcant step forward, but before chiral zeolites can ﬁnd application,

(a)

a c b

(b) Figure 7.18 The chiral germanosilicate SU-32 is built from sheets of tetrahedra made up from 4−1 units alternatively pointing in opposite directions (a). These stack to give a chiral structure that possesses helices of cages (b). Adjacent cages are depicted in different colors.

7.3 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features

b a

(b)

(a)

(c)

Figure 7.19 (a) Part of the complex ITQ37 structure (as ‘‘ball and stick’’) viewed down [100] O atoms red, Si and Ge blue. The structure contains mesoporous cavities (b) accessible through strongly noncircular

windows (c). In (b) and (c) the germanosilicate tetrahedra, which can be four-, three-, or two-connected to other tetrahedra, are depicted in blue. Note the predominance of D4Rs in this structure.

they must be prepared as bulk samples of crystals of only one enantiomorph. This might be achieved by a combination of the use of chiral templates or surface-modifying agents coupled with chiral ampliﬁcation using seeding with chiral crystals of one enantiomorph only. 7.3.6 Ordered Vacancies and Growth Defects

Isolated structural defects in the frameworks of zeolites have long been postulated, for example, as intermediates in the process of ultrastabilization and generation of secondary mesoporosity in zeolite Y (Section 7.2.4). Elegant solid-state NMR studies have previously identiﬁed the inclusion of defects resulting from T-site vacancies as a method of charge balancing organic cations in pure silica polymorphs synthesized in alkaline media (in the absence of ﬂuoride). Measured compositions suggest that

197

198

7 Structural Chemistry of Zeolites

a vacancy includes two silanol groups and two siloxy groups pointing inwards and hydrogen bonded, resulting in a characteristic 1 H MASNMR signal observed at 10 ppm [57]. The recent observation of an ordered vacancy in the pure silica polymorph SSZ-74 (the pore structure of which was described in Section 7.3.4) is therefore of great interest. The crystallographic structure of the as-prepared form of this solid gives an atomistic picture of a vacancy and how it interacts with the organic cation (Figure 7.20). The vacant T-site is surrounded by four framework oxygen atoms forming a distorted tetrahedron. Two of these oxygen atoms make closer contacts with the nitrogen atom of the charged template and are taken to be siloxy groups (Si–O− ) leaving the other two as silanol groups, hydrogen atoms of which are involved in H-bonds with the siloxy oxygens. Although ordering of such defects is rare, it is likely that interactions of this type are a widespread mechanism of charge balancing of pure silica polymorphs templated by organocations when the option of the coordination of ﬂuoride to silicon is not available, so that the vacant T-site in SSZ-74 can be used as a model system. Upon calcination and removal of the organic, it is likely that at least some of these vacancies remain. If this is the case, the intriguing possibility of including additional functionality at this site (for example, a titanium atom) could enable site-selective catalysts to be prepared and permit crystallographic determination of the structure of these catalytically active sites. High resolution electron microscopy [58] of pure silica zeolite Beta has identiﬁed a second type of defect that also has relevance to the crystallization of high silica zeolites. The framework of Beta is made up of layers stacked with a displacement of one-third of the unit cell vector in either direction, as described in Section 7.2.6. In a recent study of zeolite Beta, TEM images show large-pore defects that can only be explained by the nucleation on a given layer of two domains that are stacked with opposite displacements (Figure 7.21). After three layers of growth, these become coincident again and the defect is healed. The implication of this observation is that zeolite Beta crystallizes by layer growth, and this can lead to additional nonperiodic porosity and the presence of extended defects. AFM studies reported elsewhere in this book by Anderson and Cubillas show that layer growth is a general mechanism and can be modeled atomistically. 7.3.7 Zeolites from Layered Precursors

As discussed above and elsewhere, recent microscopic studies have indicated that the growth of zeolites occurs by a layer-by-layer mechanism. A subset of zeolites has also been found which can be prepared via a two-step process that includes the post-synthetic condensation of layered silicate precursors prepared hydrothermally. The initial crystallization gives a layered silicate in which the silicate layers are terminated on each side by Si–OH groups of (SiO)3 SiOH, ‘‘Q3 ,’’ silicon atoms and separated from one another by organic SDAs. Upon calcination the organic material is removed and the silanol groups on adjacent layers condense (2 SiOH → SiOSi) to give a fully four-connected porous tetrahedral framework. Examples of this include the synthesis of the known zeolites ferrierite (FER) and MCM-22(MWW)

7.3 The Expanding Library of Zeolite Structures: Novel Structures, Novel Features

b a c

(a)

(b) Figure 7.20 Two views of the ordered framework vacancy in as-prepared SSZ-74. The silicate tetrahedra are shown in blue and the terminal hydroxyl group oxygen atoms of the SiOH/SiO units of the vacancy are

represented in orange. Note the proximity to the defect of the charged moieties of the bisalkylammonium template (C atoms as grey spheres) around the quaternary nitrogen atoms (blue).

from the precursor layered silicates PREFER and MCM-22(P), respectively [59, 60]. The latter conversion is represented schematically in Figure 7.22. In addition, the new structure types CDS-1 [61], RUB-24 [62], and RUB-41 [63] have been prepared from the layered silicate precursors PLS-1, RUB-18, and RUB-39, respectively. Such systems, and especially MCM-22(P) – MCM-22, have been chemically manipulated by careful treatment of the laminated phase in order to introduce catalytically active species and prepare pillared microporous solids with novel catalytic features. 7.3.8 Substitution of Framework Oxygen Atoms

As described above, zeolite frameworks show extensive substitution of silicon by other metal cations (divalent Be, Zn; trivalent B, Al, Ga, Fe; tetravalent Ge and Ti). Besides having a strong inﬂuence on the structure type that crystallizes, this also has a major effect on the stability and catalytic properties of the zeolite. Much less progress has been made in substituting framework oxygen, but there is a slow

199

200

7 Structural Chemistry of Zeolites

(a)

10 nm (b)

(c)

(a)

(b) (a) (b)

(c)

(d) (c) Figure 7.21 High-resolution electron microscopy of zeolite Beta along one set of its 12MR channels (a) reveals ‘‘double pore defects’’ that can be modeled (b) by layers stacked in two different orientations and coming back into coincidence after three

layers. This is consistent with a layer-by-layer growth model where domains with different stacking offsets form on the same layer and converge ((c), (i)–(iv)). (Reproduced with permission from [55]).

accumulation of evidence that partial substitution with retention of structure is possible. Interest has centered on the isoelectronic substitution of O with CH2 or NH groups. Initial attempts to include organic groups into the framework involved the inclusion of aminopropylsiloxanes as well as tetraethoxysilane into the synthesis gel, so that co-crystallization resulted in connectivity defects and pores lined with organic groups [64]. For the inclusion of methylene groups the most promising route is to use bis-(triethoxysilyl)methane, (EtO)3 Si–CH2 –Si(OEt)3 , as a silica precursor within the preparation, so that if complete incorporation into a tetrahedral lattice were achieved, one of the seven framework oxygen atoms would

7.4 Summary and Outlook

−n H2O

Figure 7.22 Scheme showing the condensation of the layered zeolitic precursor MCM-22(P) to the fully tetrahedrally connected zeolite MCM-22 upon heating. (Reproduced with permission from [3]).

be replaced by methylene groups. Yamamoto et al. [65] observed that analogs of zeolites A, Beta, and ZSM-5 can be prepared using this siloxane precursor, but that the siloxane is partially hydrolyzed to give a mixture of Si–CH3 and Si–OH as well as Si–CH2 –Si in the ﬁnal framework product. The inclusion of amine groups into framework positions has also been investigated, with the target of preparing stable shape-selective basic catalysts. The suggested route to the inclusion of NH groups that has received the most attention is via post-synthetic high temperature treatment with ammonia. The inclusion of NH groups into zeolites without loss of crystallinity has been demonstrated, and 29 Si MASNMR of the product solids gives signals at about −67 and −86 ppm that are attributed on the basis of theoretical calculations to Al–NH2 –Si–NH2 –Al and Al–NH2 –Si species (Si–NH–Si linkages are expected to be less favorable) [66]. If correct, this holds considerable potential in the preparation of solid base catalysts.

7.4 Summary and Outlook 7.4.1 Summary

From the foregoing discussion, it is clear that the structural features of the zeolites that are most important in applications of ion exchange, adsorption, and catalysis (such as zeolites A, X, Y, chabazite, mordenite, ZSM-5, and Beta) have been studied in great detail because of the direct link between structure and function. The crystallographic structures give time- and space-averaged atomic positions, whereas local features of framework and extra-framework cation disorder have been established by combining diffraction, spectroscopy (NMR, IR, etc.), and computational simulation. The structural stability and activity of zeolites have been established over a wide range of conditions, especially in situ under realistic working

201

202

7 Structural Chemistry of Zeolites

conditions for adsorption and catalysis. For example, as an early example of the study of the activation of a zeolitic catalyst, Ni2+ cation migration within Ni–Y has been followed in situ by diffraction and X-ray absorption spectroscopy [67]. In addition, the high temperature limits to structural stability have been probed by both diffraction and inelastic scattering methods, through which a model for the amorphization of zeolite structure has been established [68, 69]. Low-frequency phonon features in zeolites have been found to be responsible for destabilizing and collapsing the zeolite structure and for converting the resulting glass from a low-density to a high-density amorphous phase. In parallel to the increase in understanding of structural chemistry of important zeolite materials, many tens of new zeolite structures with increasing complexity have been prepared via novel synthetic routes, and their structures solved via advanced crystallographic methods. A simple investigation of the number of hypothetical tetrahedrally connected frameworks with levels of complexity up to those recently observed (24 topologically distinct sites in the unit cell) that could be energetically feasible soon indicates that the known structures are a very small subset of what should be achievable. Many of these new structures have arisen from syntheses that include heteroatoms (other than Al) in the framework: Ge has been of particular interest in this regard, giving striking structures at least partly due to its strong tendency to favor the formation of D4R SBUs. A smaller number of studies have shown that it is also possible to introduce C and N as framework constituents, in the place of linking O atoms. The range of known zeolite structural chemistry has therefore been widened greatly since 1990, through a combination of ingenious synthesis and structural analysis. However, this has taken place simultaneously with the development of microporous metal organic frameworks (MOFs) [70] and covalent organic frameworks (COFs) [71] and also mesoporous silicas [72]. These new families of porous solids have outstripped the zeolite ﬁeld in terms of their chemical and structural diversity, and there is no space to describe them in detail here, other than to state that mesoporous silicas, for example, can be prepared with narrow ˚ high porosities and stabilities, and distributions of pore sizes from 4 A˚ to >100 A, readily functionalizable surfaces, and that MOFs can be prepared, fully crystalline, that possess a much wider chemical range than zeolites, much higher surface areas, great ﬂexibility, inorganic and organic functionalities, and potentially in-built chirality. The outlook for research in porous solids has therefore never been more interesting: the key question here is what role will zeolite structures (compared to that of their more recently discovered relatives) have in future academic and industrial research? 7.4.2 Outlook

With great current interest in new types of porous solids, it should be stated that zeolites remain overwhelmingly the most important in applications. While this may be expected due to their being the ﬁrst materials discovered, synthesized, and

7.4 Summary and Outlook

investigated, it also derives from key structural and stability advantages they possess over mesoporous silicas and MOFs. These advantages are the relatively low cost of manufacture, at least for (alumino)silicates prepared without exotic templates, high cation exchange capacities (for ion exchange), the much stronger acidity of Bronsted acid sites they possess, a direct consequence of their crystalline structure (for acid catalysis), their ability to include Lewis acidic framework titanium sites (for peroxide activation and selective oxidation), and especially higher thermal and hydrothermal stability of high silica zeolites compared with either mesoporous silicas or MOFs (for most applications, especially catalysis). In addition, they have been shown to be biologically compatible [73]. As a result, developments in zeolite structural chemistry are likely to continue to be of great importance in order to improve their performance as functional materials, especially in their traditional uses (the other materials will ﬁnd their own ﬁelds of application). We also envisage developments in at least two broadly related areas of structural chemistry: New structure types and chemistries, and control of morphology and microstructure. Current methods in the discovery of novel zeolite structures combine the synthesis of novel organic templates with variation in the inorganic composition of the gel. Whereas the synthesis of organic compounds proceeds via well-established methodologies, inorganic syntheses under hydrothermal conditions proceed via poorly understood processes of aging, nucleation, and growth. Consequently, high-throughput techniques that explore a wide range of composition space and reaction conditions will play a role in these studies. There is also likely to be signiﬁcant progress in preparing zeolites by the simultaneous use of two or more organic SDAs in the same synthesis. This approach has already been successful for zeotypes with different cages in the structure (STA-7(SAV) and STA-14(KFI)) [74] and for channel and cage sites in ferrierite [75]. It is likely that structurally complex structures will have several potential template sites of different sizes and shapes (as does TNU-9 [38]) that would better suit different SDAs. The trend to preparing novel topologies by the inclusion of elements such as Ge or Ti will also continue to be productive, with the proviso that there is a balance between the higher Ge-content silicate materials having greater structural diversity but lower hydrothermal stability than silicate or aluminosilicate analogs. The novel structure types containing Al that result will generate potential solid acid catalysts and, if Ti is included, selective oxidation catalysts, in each case with new shape selectivities. In most of the examples of crystalline structure given in this chapter, it is assumed that the zeolite framework extends indeﬁnitely. In reality the particle will have a certain size and shape, and may contain microstructural features such as twin planes and stacking faults. Recent attempts to control these features have been made, in order to tailor the material for particular applications such as enhancing diffusion along one-dimensional channel systems for catalysis, aligning crystals for sensing technologies, or organizing crystal orientations in membranes and thin ﬁlms [76]. In terms of size, zeolite crystallites from typical hydrothermal preparations are typically of the order of microns, but there has been much recent interest in their preparation as nanoparticles [77] (as precursors for thin ﬁlm or membrane growth,

203

204

7 Structural Chemistry of Zeolites

for example) or as single crystals approaching millimeters in dimensions, which are large enough to enable measurements of anisotropic diffusion, mechanical properties, and so on [78]. The morphology of zeolite crystals, which is a direct consequence of the growth rates on different faces, is found to vary strongly with the gel composition and conditions of crystallization, including the SDA used. The work of the group of Tsapatsis in controlling the crystallographic orientation in silicalite membranes via choice of SDA is an elegant example [79]. Recent advances in understanding the mechanism of crystal growth [80] should ultimately enable morphological control and therefore the control of surfaces that are exposed. The nature and extent of the surface of zeolite crystals is important because it is that part of the crystal that interfaces with the external environment, both during the crystallization and, in the zeolite product, during its application as an adsorbent, ion exchanger, catalyst, or in a medical application. The increased resolution of surface microscopies, coupled with transmission electron microscopy, surface spectroscopies, and measurements of surface charge, are all likely to be important in research in understanding zeolite surface structures as a complement to their bulk crystallographic structure.

References 1. McCusker, L.B. and Baerlocher, Ch.

2.

3.

4.

5.

6.

7.

(2007) in Introduction to Zeolite Science and Practice, 3rd edn, Studies in Surface Science and Catalysis, Vol. 168 (eds J. ˇ Cejka, H. van Bekkum, A. Corma, and F. Schuth), Elsevier, pp. 13–37. Lobo, R.F. (2004) in Handbook of Zeolite Science and Technology(eds S.Auerbach, K. Carrado, and P. Dutta), Marcel Dekker, New York. Wright, P.A. (2007) Microporous Framework Solids, RSC Publishing, Cambridge. Database of Zeolite Structures, Structure Commission of the International Zeolite Association, http://www.izastructure.org/databases/ (last accessed February 03, 2010). Delgado-Friedrichs, O., Dress, A.W.M., Huson, D.H., Klinowski, J., and Mackay, A.L. (1999) Nature, 400, 644–647. Foster, M.D., Simperler, A., Bell, R.G., Delgado-Friedrichs, O., Paz, F.A.A., and Klinowski, J. (2004) Nat. Mater., 3, 234–238. Liebau, F. (1985) Structural Chemistry of Silicates: Structure, Bonding and Classiﬁcation, Springer-Verlag, Berlin, pp. 14–30.

8. Wragg, D.S., Morris, R.E., and

9.

10. 11.

12. 13.

14.

15.

16.

Burton, A.W. (2008) Chem. Mater., 20, 1561–1570. Liu, Z., Fujita, N., Terasaki, O., Ohsuna, T., Hiraga, K., Camblor, M.A., Diaz-Cabanas, M.-J., and Cheetham, A.K. (2002) Chem. Eur. J., 8, 4549–4556. Shannon, R.D. (1976) Acta Cryst., A32, 751–767. (a) Li, H. and Yaghi, O.M. (1998) J. Am. Chem. Soc., 120, 10569–10570; (b) O’Keeffe, M. and Yaghi, O.M. (1999) Chem. Eur. J., 5, 2796–2801. Baur, W.H. (1964) Am. Mineral., 49, 697–704. Newsam, J.M., Treacy, M.M.J., Vaughan, D.E.W., Strohmaier, K.G., and Mortier, W.J. (1989) Chem. Commun., 493–495. Czjzek, M., Jobic, H., Fitch, A.N., and Vogt, T. (1992) J. Phys. Chem., 96, 1535–1540. Parise, J.B., Gier, T.E., Corbin, D.R., and Cox, D.E. (1984) J. Phys. Chem., 88, 1635–1640. (a) Higgins, J.B., LaPierre, R.B., Schlenker, J.L., Rohrman, A.C., Wood, J.D., Kerr, G.T., and Rohrbaugh, W.J. (1988) Zeolites, 8, 446–452;

References

17.

18.

19. 20.

21.

22.

23.

24.

25.

26.

27.

28.

(b) Newsam, J.M., Treacy, M.M.J., Koetsier, W.T., and de Gruyter, C.B. (1988) Proc. R. Soc. Lond. A, 420, 375–405. Kokotailo, G.T., Lawton, S.L., Olson, D.H., and Meier, W.M. (1978) Nature, 272, 437–438. Casci, J.L. (1994)in Zeolites and Related Microporous Materials : State of the Art 1994, Studies in Surface Science and Catalysis, Vol. 84 (eds J.Weitkamp, H.G. Karge, H. Pfeifer, W. Holderich), Elsevier, pp. 133–140. Galli, E. and Gualtieri, A.F. (2008) Am. Mineral., 93, 95–102. Chen, C.S.H., Schlenker, J.L., and Wentzek, S.E. (1996) Zeolites, 17, 393–400. Song, J.W., Dai, L., Ji, Y.Y., and Xiao, F.S. (2006) Chem. Mater., 18, 2775–2777. Corma, A., Diaz-Cabanas, M.J., Jorda, J.L., Rey, F., Sastre, G., and Strohmaier, K.G. (2008) J. Am. Chem. Soc., 130, 16482–16483. Blackwell, C.S., Broach, R.W., Gatter, M.G., Holmgren, J.S., Jan, D.Y., Lewis, G.J., Mezza, B.J., Mezza, T.M., Miller, M.A., Moscoso, J.G., Patton, R.L., Rohde, L.M., Schoonover, M.W., Sinkler, W., Wilson, B.A., and Wilson, S.T. (2003) Angew. Chem. Int. Ed., 42, 1737–1740. Camblor, M.A., Villaescusa, L.A., and Diaz-Cabanas, M.-J. (1999) Top. Catal., 9, 59–76. Villaescusa, L.A., Barrett, P.A., and Camblor, M.A. (1999) Angew. Chem. Int. Ed., 38, 1997–2000. e.g. Jackowski, A., Zones, S.I., Hwang, S.-J., and Burton, A.W. (2009) J. Am. Chem. Soc., 131, 1092–1100. Corma, A. (2004)in Recent Advances in the Science and Technology of Zeolites and Related Materials, Proceedings of the 14th International Zeolite Conference, Studies in Surface Science and Catalysis, Vol. 154 (eds E. van Steen, M. Claeys, and L.H. Callanan), Elsevier, Amsterdam, pp. 25–40. Sastre, G., Vidal-Moya, J.A., Blasco, T., Rius, J., Jorda, J.L., Navarro, M.T., Rey, F., and Corma, A. (2002) Angew. Chem. Int. Ed., 41, 4722–4726.

29. Hong, S.B., Lear, E.G., Wright, P.A.,

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

43.

Zhou, W., Cox, P.A., Shin, C.-H., Park, J.-H., and Nam, I.-S. (2004) J. Am. Chem. Soc., 126, 5817–5826. Wessels, T., Baerlocher, Ch., McCusker, L.B., and Creyghton, E.J. (1999) J. Am. Chem. Soc., 121, 6242–6247. Yoshikawa, M., Wagner, P., Lovallo, M., Tsuji, K., Takewaki, T., Chen, C.Y., Beck, L.W., Jones, C., Tsapatsis, M., Zones, S.I., and Davis, M.E. (1998) J. Phys. Chem. B, 102, 7139–7147. Burton, A.W., Elomari, S., Chen, C.Y., Medrud, R.C., Chan, I.Y., Bull, L.M., Kibby, C., Harris, T.V., Zones, S.I., and Vittoratos, E.S. (2003) Chem. Eur. J., 9, 5737–5748. Paillaud, J.L., Harbuzaru, B., Patarin, J., and Bats, N. (2004) Science, 304, 990–992. Cheetham, A.K., Fjellvag, H., Gier, T.E., Kongshaug, K.O., Lillerud, K.P., and Stucky, G.D. (2001) Stud. Surf. Sci. Catal., 135, 158. Strohmaier, K.G. and Vaughan, D.E.W. (2003) J. Am. Chem. Soc., 125, 16035–16039. Corma, A., Diaz-Cabanas, M.J., Jorda, J.L., Martinez, C., and Moliner, M. (2006) Nature, 443, 842–845. Sun, J., Bonneau, C., Cantin, A., Corma, A., Diaz-Cabanas, M.J., Moliner, M., Zhang, D., Li, M., and Zou, X. (2009) Nature, 458, 1154–1157. Gramm, F., Baerlocher, C., McCusker, L.B., Warrender, S.J., Wright, P.A., Han, B., Hong, S.B., Liu, Z., Ohsuna, T., and Terasaki, O. (2006) Nature, 444, 79–81. Baerlocher, Ch., Xie, D., McCusker, L.B., Hwang, S.-J., Chan, I.Y., Ong, K., Burton, A.W., and Zones, S.I. (2008) Nat. Mater., 7, 631–635. Baerlocher, Ch., Gramm, F., Mass¨uger, L., McCusker, L.B., He, Z., H¨ovmuller, S., and Zou, X. (2007) Science, 315, 1113–1116. Corma, A., Navarro, M.T., Rey, F., Rius, J., and Valencia, S. (2001) Angew. Chem. Int. Ed., 40, 2277–2280. Corma, A., Diaz-Cabanas, M.J., Martinez-Triguero, J., Rey, F., and Rius, J. (2002) Nature, 418, 514–517. Dorset, D.L., Strohmaier, K.G., Kliewer, C.E., Corma, A., Diaz-Cabanas, M.J.,

205

206

7 Structural Chemistry of Zeolites

44.

45. 46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

57.

58.

Rey, F., and Gilmore, C.J. (2008) Chem. Mater., 20, 5325–5331. Conradsson, T., Dadachov, M.S., and Zou, X.D. (2000) Microporous Mesoporous Mater., 41, 183–191. Lobo, R.F. and Davis, M.E. (1995) J. Am. Chem. Soc., 117, 3764–3779. Castaneda, R., Corma, A., Fornes, V., Rey, F., and Rius, J. (2003) J. Am. Chem. Soc., 125, 7820–7821. Dorset, D.L., Weston, S.C., and Dhingra, S.S. (2006) J. Phys. Chem. B, 110, 2045–2050. Corma, A., Rey, F., Rius, J., Sabater, M.J., and Valencia, S. (2004) Nature, 431, 287–290. Perego, C., Carati, A., Ingallina, P., Mantegazza, M.A., and Bellussi, G. (2001) Appl. Catal. A, 221, 63–72. Koller, H., Wolker, A., Villaescusa, L.A., Diaz-Cabanas, M.J., Valencia, S., and Camblor, M.A. (1999) J. Am. Chem. Soc., 121, 3368–3376. Arranz, M., P´erez-Pariente, J., Wright, P.A., Slawin, A.M.Z., Blasco, T., G´omez-Hortiguela, L., and Cora` , F. (2005) Chem. Mater., 17, 4374–4385. Bonino, F., Damin, A., Ricchiardi, G., Ricci, M., Spano, G., D’Aloisio, R., Zecchina, A., Lamberti, C., Prestipino, C., and Bordiga, S. (2004) J. Phys. Chem. B, 108, 3573–3583. Fyfe, C.A., Gobbi, G.C., Klinowski, J., Thomas, J.M., and Ramdas, S. (1982) Nature, 296, 530–533. Fyfe, C.A., Kennedy, G.J., Kokotailo, G.T., Lyeria, J.R., and Fleming, W.W. (1985) J. Chem. Soc., Chem. Commun., 740–742. Hong, S.B., Min, H.K., Shin, C.-H., Cox, P.A., Warrender, S.J., and Wright, P.A. (2007) J. Am. Chem. Soc., 129, 10870–10885. Tang, L., Shi, L., Bonneau, C., Sun, J., Yue, H., Ojuva, A., Lee, B.-L., Kritikos, M., Bell, R.G., Bacsik, Z., Mink, J., and Zou, X. (2008) Nat. Mater., 7, 381–385. Koller, H., Lobo, R.F., Burkett, S.L., and Davis, M.E. (1995) J. Phys. Chem., 99, 12588–12596. Wright, P.A., Zhou, W., Perez-Pariente, J., and Arranz, M. (2005) J. Am. Chem. Soc., 127, 494–495.

59. Schreyeck, L., Caullet, P., Mougenel,

60.

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

J.C., Guth, J.L., and Marler, B. (1996) Microporous Mater., 6, 259–271. Leonowicz, M.E., Lawton, J.A., Lawton, S.L., and Rubin, M.K. (1994) Science, 264, 1910–1913. Ikeda, T., Akiyama, Y., Oumi, Y., Kawai, A., and Mizukami, F. (2004) Angew. Chem. Int. Ed., 43, 4892–4896. Marler, B., Str¨oter, N., and Gies, H. (2005) Microporous Mesoporous Mater., 83, 201–211. Wang, Y.X., Gies, H., Marler, B., and M¨uller, U. (2005) Chem. Mater., 17, 43–49. Tsuji, K., Jones, C.W., and Davis, M.E. (1999) Microporous Mesoporous Mater., 19, 339–349. Yamamoto, K., Nohara, Y., Domon, Y., Takahashi, Y., Sakata, Y., Pl`evert, J., and Tatsumi, T. (2005) Chem. Mater., 17, 3913–3930. Hammond, K.D., Dogan, F., Tompsett, G.A., Agarwal, V., Conner, W.C., Grey, C.P., and Auerbach, S.M. (2008) J. Am. Chem. Soc., 130, 14912–14913. Dooryhee, E., Catlow, C.R.A., Couves, J.W., Maddox, P.J., Thomas, J.M., Greaves, G.N., Steel, A.T., and Townsend, R.P. (1991) J. Phys. Chem., 95, 4514–4521. Greaves, G.N., Meneau, F., and Sankar, G. (2002) Nucl. Instr. Methods Phys. Res. B, 199, 98–105. Greaves, G.N., Meneau, F., Majerus, O., Jones, D.G., and Taylor, J. (2005) Science, 308, 1299–1302. F´erey, G. (2007)in Introduction to Zeolite Science and Practice, 3rd revised edn, Studies in Surface Science and Catalysis, ˇ Vol. 168 (eds J. Cejka, H. van Beckkum, A. Corma, F. Sch¨uth), Elsevier, pp. 327–374. El-Kaderi, H.M., Hunt, J.R., Mendoza-Cortes, J.L., Cˆot´e, A.P., Taylor, R.E., O’Keeffe, M., and Yaghi, O.M. (2007) Science, 316, 268–272. Zhao, D. and Wan, Y. (2007)in Introduction to Zeolite Science and Practice, 3rd revised edn, Studies in Surface Science ˇ and Catalysis, Vol. 168 (eds J. Cejka, H. van Beckkum, A. Corma, and F. Sch¨uth), Elsevier, pp. 241–300.

References 73. Schainberg, A.P.M., Ozyegin, L.S.,

Kursuoglu, P., Valerio, P., Goes, A.M., and Leite, M.F. (2005)in Bioceramics 17, Key Engineering Materials, Vol. 284–286 (eds P.Li, K. Zhang, and C. W. Colwell), Trans Tech Publications, pp. 561–564. 74. Castro, M., Garcia, R., Warrender, S.J., Wright, P.A., Cox, P.A., Fecant, A., Mellot-Draznieks, C., and Bats, N. (2007) Chem. Commun., 3470–3472. 75. Pinar, A.B., Gomez-Hortiguela, L., and P´erez-Pariente, J. (2007) Chem. Mater., 19, 5617–5626.

76. Drews, T.O. and Tsapatsis, M. (2005)

77. 78.

79.

80.

Curr. Opin. Colloid Interface Sci., 10, 233–238. Tosheva, L. and Valtchev, V.P. (2005) Chem. Mater., 17, 2494–2513. Lethbridge, Z., Williams, J.J., Walton, R.I., Evans, K.E., and Smith, C.W. (2005) Microporous Mesoporous Mater., 79, 339–352. Choi, J., Ghosh, S., Lai, Z.P., and Tsapatsis, M. (2006) Angew. Chem. Int. Ed., 45, 1154–1158. Brent, R. and Anderson, M.W. (2008) Angew. Chem. Int. Ed., 47, 5327–5330.

207

209

8 Vibrational Spectroscopy and Related In situ Studies of Catalytic Reactions Within Molecular Sieves Eli Stavitski and Bert M. Weckhuysen

8.1 Introduction

Scientists in the ﬁeld of catalysis are in a perpetual pursuit of optimization of existing processes in terms of activity/selectivity as well as development of new ones. In many cases, trial-and-error modus operandi prevails upon the rational approach, even with a level of sophistication and knowledge modern science and technology delivers. In order to improve on this situation, catalyst scientists require better insight into the key stages of the reaction process and, in particular, the catalyst’s modus operandi. Armed with a rigorous understanding of this, it could then be possible to prepare the archetypal designer catalyst with the desired superior performance for the reaction in question. However, such information can only be obtained reliably by monitoring a catalyst ‘‘in action.’’ In order to do this, it is essential to adapt catalytic reactors and/or spectroscopic/scattering techniques to study the processes in real time – an approach that gives rise to the ﬁeld of in situ or operando spectroscopy [1, 2]. A catalytic cycle typically consists of a sequence of reaction steps that describe the transformation of substrate molecules into a ﬁnal reaction product at a catalytically active site. Although scientists have been working for decades to decipher such events, only in a limited number of cases has a deeper understanding of the processes involved been achieved. If such information is to be obtained, one requires sufﬁcient detailed information about the catalyst material at each step in its life cycle: that is, synthesis → calcination → activation → reaction → deactivation → regeneration (where possible). Conventional characterization (hereby termed accordingly ex situ characterization) focuses on the study of the catalyst materials at these various stages, away from the reactor and is often performed under ambient conditions; that is, at room temperature and atmospheric pressure. Although this approach yields interesting information, it is incapable of providing direct insight into the processes occurring in the catalyst during the course of the reaction. Therefore, the catalyst scientist is forced to develop analytical tools that enable the continuous monitoring of the catalyst ‘‘in action.’’ However, we observe that at this stage a drawback of in situ methods is that the gas and/or liquid phase Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

210

8 Catalytic Reactions Within Molecular Sieves

surrounding the catalyst are probed simultaneously, including both the active surface and the inactive bulk of the material in question, making the interpretation prone to ambiguities. One can trace the origin of in situ spectroscopy in catalysis back to 1954 with two seminal papers from the Eischens group [3, 4]. In these reports, the interaction of carbon monoxide with Cu, Pt, Pd, and Ni supported on SiO2 and of ammonia molecules with cracking catalysts were studied with infrared (IR) spectroscopy. It is noteworthy that, from the modern standpoint, it can be argued to what extent the study in question can be considered as in situ as the conditions described are a far cry from those actually found in a catalytic process. Nevertheless, they represented an important step forward for the in situ approach as for the ﬁrst time they considered the importance of the dynamics of a catalyst surface in the presence of adsorbates. Indeed, it was most probably the ﬁrst spectroscopic-reaction cell designed for measuring IR spectra of heterogeneous catalysts. Since these pioneering studies, continuous progress toward the use of in situ spectroscopic techniques can be observed. This is illustrated in Figure 8.1, which shows the evolution of the number of in situ spectroscopy papers in the zeolite literature over the past decades. Of all the commonly employed in situ techniques, IR spectroscopy has perhaps the longest history and is most often used in the ﬁeld of zeolite research. In the early stage of the development, IR measurements were made using self-supported wafers (simple transmission/absorption measurements). However, since this time, IR has been developed further in order to obtain better quality spectra more quickly. Such improvements have been brought about by the development of more sensitive detection systems and improved sampling methods and enable the measurement of IR spectra under more relevant reaction conditions. A great 300

225

150

1963

75 1971

1979

1985 1991 1997

0

2003 2009

Figure 8.1 Estimated number of journal publications on in situ characterization of catalyst materials (based on the results of an ISI and Chemical Abstracts database search using the terms ‘‘in situ’’ and ‘‘zeolite’’).

8.2 Acidity Determination with IR Spectroscopy of Probe Molecules

improvement in the technique was developed in the early 1980s by the introduction of Fourier transform IR (FTIR) instruments, which allowed for short recording times (seconds to minutes) and high resolution (0.5–4 cm−1 ), adjustable to obtain the desired signal-to-noise ratio. Nowadays price-wise, the FT-IR spectrometer is comparatively cheap and, as such, is one of the workhorses in a typical academic and industrial heterogeneous catalysis laboratory. As most probably known to the reader, IR probes the transitions between vibrational states of the absorber over an energy range of 0.496 > E > 0.0496 eV (4000−400 cm−1 commonly known as medium/mid infrared or MIR) and 0.0496 > E > 1.24 × 10−3 eV (400−10 cm−1 known as far infrared or FIR). Thus, the technique is able to probe the chemical and geometric structure of (adsorbed) molecules (by causing changes in the molecular dipolar moment) and solids (changes in lattice vibrations and/or acoustic modes), thus covering the whole gamut of catalyst characterization. In general, there are three commonly employed modes by which IR spectroscopy is used to probe catalytic reactions under in situ conditions. First of all, there is transmission IR, which involves the preparation of the catalyst sample in the form of a self-supporting disk and is perhaps the most widely used. A second way of collecting IR data is by diffuse reﬂectance infrared Fourier transform (DRIFT) spectroscopy, which measures light scattering and absorption phenomena. It is particularly attractive because it does not employ complicated sample preparation approaches and therefore is often very close to realistic reaction conditions. On the other hand, the interpretation of DRIFT spectra is based on the phenomenological theory of Kubelka and Munk and some precautions have to be taken when comparing transmission IR and DRIFT with each other. The goal of this chapter is to give the reader some background of what currently is possible with the IR technique and related spectroscopic techniques for monitoring physicochemical phenomena during, for example, a catalytic reaction within molecular sieves. The ﬁrst part of the chapter is dedicated to use of IR to investigate acid–base properties of zeolites by making use of relevant probe molecules. The second part involves the study of zeolite synthesis processes, while the third part discusses two prototypical examples of reactions catalyzed by molecular sieves, namely the catalyst removal of NOx and the methanol-to-oleﬁn (MTO) process. A ﬁnal section is devoted to the use of IR microspectroscopy to shed insight into template decomposition phenomena, adsorption, and alignment of probe molecules and catalytic processes taking place within large zeolite crystals. The chapter ends with some concluding remarks.

8.2 Acidity Determination with IR Spectroscopy of Probe Molecules

Acid–base reactions within zeolite-based materials are perhaps the most technologically relevant class of heterogeneous catalytic processes. Similar to other solid acids, zeolites possess both Brønsted and Lewis acid sites, which are typically hydroxyl groups and coordinatively unsaturated cations, respectively.

211

212

8 Catalytic Reactions Within Molecular Sieves

Either Brønsted or Lewis acidity can dominate, depending on the chemical content, crystalline structure, postsynthetic treatments and, last but not the least, the state of hydroxylation under the reaction conditions. In order to draw connections between the acidic properties of zeolites and their catalytic properties, it is essential to obtain quantitative information on the number, nature, location, and strength of the acidic sites. In general, there are two theories that describe the acid–base properties of the solid materials, that is, Brønsted and Lewis theories. In the former theory, an elementary acid–base reaction is a proton transfer from an acid (AH) to the base (B): AH + B → A + BH. In this case, acid strength may be deﬁned as the tendency to give up a proton, which can be quantiﬁed by determining the state of equilibrium, for example, in aqueous solution. In the Lewis approach, any species possessing a vacant orbital can be considered as an acid. The acid–base reaction A + B → A – B is then the formation of a bond involving unshared electron pair of B and the vacant orbital of A. However, it is important to emphasize that the determination of acidity based on these concepts is strictly valid only for acidic molecules dissolved in homogeneous media (solvents). It is evident that acid groups located on surfaces of catalytic solids cannot be treated in the same way because the concept of homogeneous medium has lost any validity. This consideration is particularly valid for acid (and basic) groups located in the channels and cavities of microporous catalysts which, by deﬁnition, are highly inhomogeneous. IR has established itself as an essential tool to investigate intermolecular interactions, for example, hydrogen and coordination bondings. It has been shown that the most valuable information on the zeolite acidic properties can be obtained using adsorbed probe molecules interacting with acid sites. Perturbations introduced to the IR spectra of both the surface groups and probe molecules upon the interaction, such as intensity variations and frequency shifts, can be translated into the properties of relevant acid sites. The extensive criteria for a selection of suitable probe molecules have been formulated in great detail (e.g., [5–7]). In general, small and weakly interacting molecules are recommended for probing surface properties of acidic and basic solids. Other criteria include the following: (i) selectivity of interaction with speciﬁc surface site; (ii) low reactivity, even at catalytically relevant conditions; (iii) detectable magnitude of spectral response upon interaction; (iv) high and experimentally measurable extinction coefﬁcients; and (v), whenever possible, the reactant itself shall be used as a probe molecule. IR spectra of activated zeolites typically feature two or more major bands in a hydroxyl spectral region (in some cases, more groups of bands can be observed, as, for example, in the case of SSZ-33 zeolite [8]). The typical values for band positions are summarized in Table 8.1. Firstly, silanol groups give rise to bands centered at 3710–3760 cm−1 (Figure 8.2). External and internal silanols (four individual components) can be differentiated in the deconvoluted spectra [9]. In the case of large zeolite crystallites, the individual contribution can be even resolved in the spectra [10]. Secondly, the IR band at ∼3600 cm−1 is attributed to the O–H stretching mode of the Si(OH)Al bridged group. The latter is the most important chemical entity as these hydroxyl groups are typically strong Brønsted acid sites.

8.2 Acidity Determination with IR Spectroscopy of Probe Molecules Table 8.1

IR band positions for hydroxyl groups in various zeolite structuresa.

Zeolite

Silanol OH groups position (cm−1 )

Bridged OH group position (cm−1 )

References

SM-5 Mordenite

3747 3745 3740 3714 3746 3746 3710 (shoulder) 3740

3612 3605 (3612, 3585) 3640, 3550 3667 3602 (3609, 3601, 3587, 3565) 3602 (3565, 3592, 3610) 3610 (3631, 3617, 3600) 3616

[12] [13, 14] [15] [16] [17] [17] [18, 19] [20]

ZSM-2 Ferrierite UZ-4 SAPO-34 SZ-13

a Only a selection of zeolites, relevant to the scientiﬁc content of this chapter are included in this table. For a comprehensive set of data, the reader is referred to [22].

Simulated spectrum

Absorbance

0.1

Original spectrum

Nonacidic part

Acidic part

Stronger acidity

Weaker acidity 3800

3700

3600

Wavenumbers

3500

3400

(cm−1

)

Figure 8.2 The hydroxyl region of the IR spectrum of ultrastable zeolite Y and its deconvolution into individual components. (Copyright American Chemical Society 1993; Reproduced with permission from [11].)

This band may appear asymmetric due to heterogeneity of the sites located at different crystallographic positions within zeolite unit cell. In the cases when zeolite crystallites are defect rich, a broad band at 3460 cm−1 , assigned to hydrogen-bonded silanol groups located in the interior of the crystal, appears in the spectrum [10]. Several classes of probe molecules have been successfully used to characterize acidic properties of zeolites, namely, diatomic molecules (CO, N2 ),

213

8 Catalytic Reactions Within Molecular Sieves

nitrogen-containing organic bases (pyridine and its derivatives, nitriles), and hydrocarbons (ethene, benzene). CO forms hydrogen-bonded complexes with hydroxyl groups, which can be identiﬁed in the C–O and O–H spectral regions. Upon the H-band formation, O–H stretching band is redshifted and broadened, with the magnitude of shift related to the acidity of the site. At the same time, the C–O band is blueshifted compared to the gas-phase CO frequency. In the case of dinitrogen, the molecule looses its inversion center upon binding, making molecular vibration IR active. Small molecular diameter, chemical inertness, and weak basicity of N2 have made it a very useful probe. Pyridine, ﬁrst proposed as a reporter molecule in as early as 1963 [21], leads to a similar effect on the O–H bands, whereas ring vibrations in the region 1600−1400 cm−1 allow to identify pyridine bound to different surface sites (see Table 8.2 and Figure 8.3). An especially advantageous approach involves using two probes: one is weakly basic, such as CO to access the acidic strength of the sites and a strong base, for example, pyridine, to obtain information of their nature and location within zeolite framework. Several comprehensive reviews covering investigation of zeolite acidic properties with probe molecules were published in the [23]. Therefore, our aim is to highlight some main applications of the IR technique from research work of the last decade in this chapter.

Absorbance (a.u.)

214

(a)

(b)

(c) 1680

1640

1600

1560

Wavenumber (cm

1520 −1)

Figure 8.3 IR spectra taken after pyridine adsorption and subsequent evacuation at (a) 373; (b) 623; and (c) 773 K. (Copyright Royal Society of Chemistry 1996; Reproduced with permission from [24].)

1480

1440

8.2 Acidity Determination with IR Spectroscopy of Probe Molecules

When several acid sites differing in their nature and strength, coexist in a zeolite material, they can be distinguished using one or more molecular probes. For example, two types of Brønsted sites were identiﬁed in mordenite, that is, hydroxyl groups located in the main channels and in the side pockets [25, 26]. The former are more acidic, as evidenced by the redshift upon CO adsorption, and accessible to pyridine. Buzzoni et al. compared the acidity of several acidic zeolites using pyridine [27]. It was shown that for all the materials studied pyridine reacts quantitatively with the available acid sites to form pyridinium species. Formation of pyridine dimers was also detected. Unlike zeolites, featuring three-dimensional pore system (ZSM-5 and β), mordenite showed a pore blockage upon pyridine adsorption. Moreover, the acid sites in the small side pockets of the mordenite channels were also shown not to be accessible. This conclusion is conﬁrmed by another study where only the partial protonation of pyridine on mordenite was observed [11]. It was inferred that pyridine could only slightly enter the side pockets, making protonation improbable due to sterical hindrances. Pyridine derivatives are claimed to be advantageous over pyridine when one probes weak Brønsted acid sites due to their stronger basicity and weaker afﬁnity for Lewis sites, as steric hindrances induced by the methyl groups arise. In addition, several IR bands in the spectra of absorbed methylpyridines are sensitive to the acid site strength. For example, the 2,6-dimethylpyridine spectrum informs of the strength of the Brønsted acid sites not only by the position of the (NH) band (like for pyridinium species) but also more directly by the position of the ν 8a band. Sensible correlations have been established between the spectral shifts upon CO adsorption, the *(NH) as well as the ν 8a band position upon 2,6-DMP adsorption (Table 8.2) [28]. An interesting application of bulky pyridine derivatives is the assessment of acid sites with different accessibility. We illustrate this point with a few examples. One of the important aspects of zeolite catalysis is to overcome the transport limitations, which stems from the exclusively microporous character of these materials by Table 8.2

IR band assignments of different pyridine specie detected on solid acidsa.

Band position (cm−1 )

Assignment

Acid site

1397 1455 1490 1545 1576 1621 1635

PyH+ Py Py + PyH+ PyH+ Py Py PyH+

B L B+L B L L B

From [24]. = pyridine; PyH+ = pyridinium ion; B = Brønsted acid site; L = Lewis acid site.

a Py

215

216

8 Catalytic Reactions Within Molecular Sieves

100 80

85

68 80 60 58 60 32 30

33

40

20 10

20 11 3

6

4

DeAlSP1

0

DeAlSP2

CH3 HC3

CH3 HC3 CH3N HC3 CH3

SP DeAlSP

DeAlSP3 HC3

N

CH3

HC3

N

CH3

N

CO

Figure 8.4 Distribution of Brønsted acid sites of different accessibility over mordenite samples studied. (Copyright Elsevier 2007; Reproduced with permission from [28].)

introducing controlled mesoporosity. Such a treatment would unavoidably affect the acidic properties. A methodology for characterization and quantiﬁcation of the acidic sites of different accessibility has been developed and applied for a series of dealuminated mordenites. These materials were probed with IR spectroscopy of adsorbed alkylpyridines of increasing molecular dimensions, followed by the characterization of the nonaccessible sites using subsequently adsorbed CO [29], allowing for a step-by-step characterization of the nature and strength of sites with different accessibility. The results of this approach are shown in Figure 8.4. It was shown that upon dealumination of the small-port mordenite samples the acid sites in the side pockets become accessible for pyridine due to the partial destruction of side pockets upon dealumination. Subsequently, the formation of secondary mesoporosity systems due to further dealumination make the zeolite crystals completely accessible to relatively bulky molecules, such as lutidine, collidine, and 2,6-di-tert-butylpyridine. In another related study, desilicated ZSM-5 materials have been probed by applying CO and collidine (2,4,6-trimethylpyridine) [10]. CO adsorption at low temperatures showed no signiﬁcant alterations in Brønsted acidity after desilication. Collidine, which is too bulky to enter the micropore system of ZSM-5, was used to probe the surface of the mesopores, where Lewis acid sites can be generated from dislodged framework aluminum ions. By ﬁrst saturating the zeolite with collidine

8.2 Acidity Determination with IR Spectroscopy of Probe Molecules

and subsequently adsorbing CO, it was demonstrated that almost all Lewis sites coordinate collidine, whereas the Brønsted acidity was continuously protected in the micropore system. These observations allowed ruling out the higher Brønsted acidity as well as the synergism between two types of acid sites as an explanation for the enhanced catalytic activity of the desilicated material [10]. Thibault-Starzyk et al. determined an accessibility of the acid sites in hierarchical zeolites [30]. These materials have been extensively studied in the last decade, as they exhibit an improved catalytic performance compared to their microporous parents due to the integration of the catalytic properties of the native micropores and the facilitated transport via a mesopore network. It was shown that with the postsynthetic desilication up to 40% of the ZSM-5 acid sites could be made available for molecules as bulky as collidine. An accessibility index (ACI), which is deﬁned as the number of acid sites detected by adsorption of the probe molecules divided by the total amount of acid sites in the zeolite based on the measured aluminum content was shown to be a quantitative indicator for evaluating the effectiveness of synthetic strategies for hierarchical zeolites preparation. Another interesting route toward mesopore introduction in zeolites is partial dissolution in highly alkaline media, followed by recrystallization into mesostructure. The strength of the acid sites of recrystallized Beta zeolite was studied by IR spectroscopy of adsorbed CO [31]. Two types of Brønsted acid sites with different acid strength were identiﬁed in recrystallized samples. This was rationalized in terms of highly crystalline areas of zeolite bearing stronger sites, and less crystalline areas featuring weaker sites, due to the partial destruction of Si–O and Al–O bonds. Next, pyridine and 2,6-di-tert-butylpyridine were selected as probe molecules to assess the accessibility of Brønsted acid sites. The highest content of the sites accessible for pyridine was observed on parent zeolite, whereas the sample that recrystallizes in mild conditions has shown the highest amount of sites accessible for 2,6-di-tert-butylpyridine. More severe conditions of recrystallization lead to acid properties similar to those of mesoporous MCM-41. An inﬂuence of dealumination on acidic properties of zeolite Y was reported by Datka et al. [15]. Steaming was shown to lead to the appearance of two new IR bands, assigned to strongly acidic hydroxyls, which interact with extra-framework Al, and Al–OH groups. Two kinds of Al–OH groups can be differentiated, that is, one accessible and another inaccessible to benzene molecules. The dealumination was reverted by KOH treatment, leading to reincorporation of Al ions into the framework and reconstruction of the hydroxyls IR bands of the parent material. Pyridine sorption has shown that in both the realuminated and the nondealuminated samples a part of hydroxyl group is inaccessible to the bulky molecules. Acidic treatment of dealuminated HY was also used to reinsert Al into the zeolite structure: depending on the conditions used, a degree of realumination of 60% can be achieved, with almost all Lewis sites related to extra-framework Al converted into Brønsted ones [32]. Recently developed zeolite structures have also fallen into the scope of acidity measurements. For example, SUZ-4 zeolite, structurally related to ferrierite, has been investigated by Zholobenko et al. [17]. The ferrierite family has attracted attention due to the ability to catalyze the selective transformation of oleﬁns. Four

217

218

8 Catalytic Reactions Within Molecular Sieves

major types of hydroxyl groups of different nature could be identiﬁed in the OH region of ferrierite spectrum, which were assigned to (i) bridging OH groups located in 10-ring channels; (ii) large cages in eight-ring channels; (iii and iv) in eight- and six rings, respectively. It was suggested that extended rings at the intersections of eight- and six-ring channels are enriched with substituting aluminum, resulting in a high concentration of OH groups in the large cages in eight-ring channels. In the case of SUZ-4 zeolite, the deconvolution of the spectrum gave three IR bands. The intensity of the band, assigned to OH groups in six rings, decreased compared to that in ferrierite, indicating that the nonexchangeable potassium cations are concentrated in small cages or in double six rings. The hydroxyls in 10 rings, which constitute about 50% of Si(OH)Al groups, are accessible to both n-hexane and isobutane. The peak at 3592 cm−1 (40%) assigned to OH groups in eight-ring channels is affected with n-hexane but not with isobutane. A comparison of the OH band spectral shifts to n-hexane adsorption has suggested the following order of acidic strength in zeolites: H-ZSM-5 > H-SUZ-4 ≥ H-ferrierite. Another zeolite structure with relatively unexplored catalytic applications is ZSM-2, with pore openings of 0.74 nm. Covarrubias et al. synthesized and characterized nano-sized ZSM-2 particles [16]. From the pyridine adsorption experiments, both Brønsted and Lewis sites were identiﬁed. Steaming of this zeolite led to an increase in the number of Lewis sites at the expense of the B counterpart, possibly due to dealumination. Alongside zeolites, the acidic properties of silicoaluminophosphates have been investigated by the IR technique. For example, SAPO-34 cannot be probed with ˚ However, with ammonia as a probe pyridine due to small pore openings (4.4 A). molecule, both Lewis and Brønsted sites in SAPO-34 can be accessed [18]. The latter constitute most of the acid sites; they exhibit moderate acidity which, along with the shape selectivity, accounts for high selectivity of the catalyst toward light oleﬁn formation. When probing SAPO-34 with CO and ethene, the nature of the acid sites can be further clariﬁed [19]. Three distinct hydroxyl groups were identiﬁed: two major components attributed to different crystallographic positions and the third exhibiting acidic properties comparable with those of zeolites (330 cm−1 shift upon CO adsorption) related to species formed at the borders of silica islands or inside aluminosilicate domains. A related material, namely, high-silica chabazite H-SSZ-13 exhibits four types of hydroxyl groups with similar acidic strength but with different accessibilities for CO [20]. Overall, little difference has been found in acidity between this material and SAPO-34.

8.3 Zeolite Synthesis Processes

Zeolites are built from tetrahedral SiO4 and AlO4 units. The manner in which these basic structural units are connected deﬁnes the ﬁnal porous architecture of the channel networks, that is, the pore diameter and interconnection of microporous channels. Even subtle variations in these parameters can signiﬁcantly affect the

8.3 Zeolite Synthesis Processes

chemical reactivity and catalytic performance of the material. As the synthetic conditions directly determine the structure of the resulting zeolites, a number of studies have been directed toward obtaining a detailed understanding of the processes that occur during zeolite preparation. In more detail, nucleation and growth behavior has to be correlated with the changes in the nature of aluminum and silicon source, and synthesis duration, temperature, and pressure. Even though a considerable amount of research has been devoted to the investigation of underlying synthesis factors, a lack of basic insight into the process has led to a rather empirical approach in the attempts of designing new zeolite structures and modifying existing ones. In what follows, we brieﬂy discuss a variety of applications of in situ techniques to investigate zeolite synthesis processes. In one of the pioneering studies, Engelhardt and coworkers used 29 Si and 27 Al MAS NMR to investigate the nature of the intermediates during zeolite A synthesis [33]. The initial gel was found to consist of tetrahedral Si(OAl)4 and Al(OSi)4 , which form an amorphous network with alternating Al–Si ordering. This structure is converted into a highly crystalline zeolite with time. There was no direct correlation between the initial Si/Al ratio and the intermediate aluminosilicate gels composition. The variations have been explained in terms of different aluminum and silicon sources used [34]. This method has been extended to other systems, such as mordenite and sodalite. However, in all experiments of this kind, the solid and liquid phases were separated. Shi et al. demonstrated that the MAS MNR technique could be applied in situ to follow the speciation as a function of time [34]. The main distinct species observed during zeolite A synthesis were amorphous Al(OSi)4 , and the growing zeolite phase with small fractions of Si(OH)4 and silicalite species. Higher concentration of the gel led to faster crystallization due to a higher number of nuclei formed. The experimental ﬁndings were in favor of the crystallization model in which a growth proceeds through a fast deposition of the Si(OH)4 and Al(OH)4 − on the surface. Small-angle X-ray scattering (SAXS) measurements showed that before the zeolite A crystallization begins uniformly sized precursor particles of about 10 nm are formed [35]. The crystallization of SAPO-34 has been monitored by Vistad et al. using in situ X-ray diffraction and NMR [36]. Heating rates were shown to signiﬁcantly inﬂuence the course of the process, for example, slow heating produces the precursor phase, which is critical to obtain SAPO-34. The transformation of this layered phase into a chabazite structure occurs through the partial dissolution of the former. As no changes in the diffraction peak widths were observed, other rearrangement mechanisms were ruled out. NMR experiments (Figure 8.5) allowed the identiﬁcation of four steps in the SAPO-34 crystallization with increasing temperature, namely, dissolution of the initial gel, formation of the four-ring structures, formation of the layered AlPO4 -F prephase, and, ﬁnally, dissolution of the latter and condensation of the four-ring structures into a triclinic chabazite structure. Incorporation of silicon was proposed to be a limiting step of the process. The synthesis process of aluminophosphates (AlPOs) has also attracted significant attention. For example, the crystallization of the cobalt-containing AlPO4 -5 (CoAPO-5) has been investigated with a combined Raman/SAXS/UV–vis/X-ray

219

220

8 Catalytic Reactions Within Molecular Sieves ×6 30

20

10 0 (ppm)

−10

25 °C

−20

170 °C – 240 min

170 °C – 130 min

162 °C – 67 min 156 °C – 53 min 126 °C – 24 min 103 °C – 11 min 25 °C – 0 min 80

60

40

20

0 (ppm)

−20

−40

−60

−80

Figure 8.5 In situ 27 Al NMR spectra recorded during synthesis of triclinic SAPO-34. (Copyright American Chemical Society 2003; Reproduced with permission from [37].)

absorbtion setup [38]. Raman spectroscopy allowed the identiﬁcation of the instantaneous formation of the Al–O–P bonds in the gel with a wide distribution of particle sizes, as shown by SAXS [35]. Then, the one-dimensional chains with alternating Al–P arrangement are formed, most probably from four-membered units. The chains condense into larger 1D rodlike structures, which undergo rearrangement into 2D and, subsequently 3D network. In situ UV–vis data shows the gradual transformation of the octahedral coordination of the Co ions in the precursor gel into a tetrahedral one in the resulting material before the onset and during crystallization. These ﬁndings are corroborated by X-ray absorbtion data. Conformation of the structure-directing agent molecules in the synthesis of AlPO-5 and metal-subsituted APO-34 has been assessed with Raman [39]. A strong interaction between the template molecules and inorganic network containing transition metal ions leads to deformation of the template structure; this does not occur in case of the synthesis of a metal-free AlPO-based molecular sieve, as illustrated in Figure 8.6. Raman spectroscopy often presents challenges when applied for studying zeolite synthesis due to the strong ﬂuorescence of the reaction mixture. This prompts a use of, for example, a UV laser source, which allows avoiding the ﬂuorescence and increasing the sensitivity. Fan et al. have successfully applied UV Raman to study hydrothermal crystallization of zeolite X [40]. It was shown that amorphous solid phase initially dissolves to form monomer silicate species in the liquid phase, while amorphous aluminosilicate species composed of predominately four rings

8.4 Selection of Zeolite-Based Catalytic Reactions

221

C

C

C

C

+ N

C

C

C

C

C

(i) tg.tg conformation

(ii) tt.tt conformation

C C

C

+

N

C

C

C C

(i) (ii) (ii) (i)

e Tim

e Tim

600 (a)

620

640 660 680 Raman shift (cm−1)

700

600

720 (b)

620

640 660 680 Raman shift (cm−1)

Figure 8.6 Time-resolved in situ Raman spectra for synthesis gels with (a) no framework substitution and (b) 30% substituted Zn2+ . The structures of the two tetraethylammonium hydroxide template conformations are also indicated. (Copyright American Chemical Society 2006; Reproduced with permission from [39].)

are formed in the solid phase in the early stage of nucleation. These four rings are connected with each other via double six rings together with the monomer silicate species in the liquid phase to form the crystalline zeolite X framework. Another experimental obstacle is the development of a suitable reaction cell. Successful examples of hydrothermal vessels allowing for laser excitation have been reported [38], including variable focal point design allowing to probe the liquid and solid phase independently [40]. Furthermore, a setup for microwave-assisted zeolite synthesis adapted for simultaneous SAXS/WAXS and Raman measurements has been developed by Tompsett et al. [41]. 8.4 Selection of Zeolite-Based Catalytic Reactions 8.4.1 Catalytic Decomposition of Nitric Oxides

Ever-tightening control over nitrogen-containing emissions has prompted an extensive search for effective catalyst materials for nitric oxides abatement, often referred to as DeNOx . A number of oxide catalysts have been proposed for this process. Protonated and metal-containing zeolites were shown to be active in various DeNOx routes, such as direct decomposition and selective catalytic

700

720

222

8 Catalytic Reactions Within Molecular Sieves

reduction (SCR) by hydrocarbons and ammonia. In what follows, the insights into the mechanism of DeNOx on zeolites gained by using in situ spectroscopy are discussed. Copper-containing ZSM-5 was found to be active in NO decomposition. However, the nature of the active center is still under discussion despite the large characterization efforts. It was established that during pretreatment and under reaction conditions di- and monovalent copper ions, along with bi- and, possibly, polynuclear complexes, are present in the materials, On the basis of IR investigations of CuHZSM-5 and CuZSM-5 with NO and NO/O2 , CuI sites were suggested to be catalytically active [42]. In support of this proposal, photoluminescence measurements have demonstrated a correlation between the CuI concentration and the activity toward NO decomposition [43, 44]. On the basis of a combination of IR spectroscopy and the molecular modeling, mono-adducts of NO with CuI are proposed to be the key intermediates in the reaction, with the N–N bond formed in the interaction of Cu-coordinated and gas-phase NO molecules [45]. In contrast, Kucherov et al. reported on the strong interaction of NO with CuII centers based on EPR (electron paramagnetic resonance) spectroscopy [46]. This technique suggests three different types of paramagnetic CuII ions being formed upon CuZSM-5 dehydration [47]. Only one species react with adsorbed NO to form CuII –NO, while two other species are inactive. This is illustrated in Figure 8.7. CuI is proposed to react with two NO molecules to form the complex, which subsequently transforms into CuII (NO)O− . Adsorbed species are generated via the interaction of CuII O− with NO. Accordingly, a mechanism involving the aforementioned intermediates is proposed. Dehydrated Cu-ZSM-5

×3

g = 2.272

dpph

(a) Then 50 torr of NO adsorption for 10 min

×3

dpph

g = 2.314 200 G (b)

Figure 8.7 EPR spectra of (a) dehydrated CuZSM-5, (b) after 50 torr NO adsorption for 10 minutes. (Copyright Elsevier 2000; Reproduced with permission from [47].)

8.4 Selection of Zeolite-Based Catalytic Reactions

Mathisen et al. compared the redox behavior of copper ions in CuZSM-5 in the SCR of NOx by propene [48]. In the case of CuZSM-5 prepared by ion exchange, CuII can be reversibly reduced to CuI and backward by propene and NOx , respectively, as shown by XAS. No copper oxide species or metallic copper clusters were detected. On the basis of these ﬁndings, the authors postulate that the redox mechanism and not copper dimers are responsible for SCR activity. As to the CuAPO-5 material, acid sites and not framework copper ions are suggested to be catalytically active toward NOx reduction. Ganemi et al. tested highly siliceous copper ion-exchanged ZSM-5 in the direct decomposition of NO [49]. The optimal Si/Al ratio was established to allow for one ion exchange site per channel intersection; 200% Cu exchange was shown to yield best performance. Both unidentate and bidentate NO3 − were detected in the IR spectrum. In the latter, the NO3 species are bound to the bridged Cu2 + –O2− –Cu2 + sites, and the authors assume that these complexes block the sites active in NO decomposition. Another possible binuclear copper intermediate, that is, bis(µ-oxo)dicopper has been ﬁrst predicted on the basis of theoretical considerations [50] and then identiﬁed in overexchanged CuZSM-5 materials under reaction conditions by UV–vis and EXAFS (extended X-ray adsorption ﬁne structure) spectroscopies [51]. This complex takes a role of continuous O2 production and releases and maintains a sustainable catalytic cycle. Beyond binuclear copper complexes, chain-like copper oxide structures were identiﬁed in CuZSM-5 with exchange rates of 75–100% based on the results of EPR and UV–vis spectroscopy [52]. As these species can be easily oxidized and reduced, their involvement in the reaction is likely. Another zeolite-based catalyst active in catalytic reduction of NOx is FeZSM-5. High iron loadings up to Fe/Al = 1 can be achieved when sublimation of FeCl3 is used to introduce the metal into the zeolite [53]. Oxygen-bridged binuclear iron complexes were proposed to be key intermediates in the reaction [53]. EPR spectroscopy indicated that reactive iron is present as Fe3+ under oxidative catalytic conditions [54]. Different iron species exist in the catalysts under reaction conditions, that is, Fe3+ in octahedral and tetrahedral coordination and iron oxide clusters, all with different ox-red behavior [55]. Formation of NO2 /NO3 species coordinated to iron ions was demonstrated by IR as illustrated in Figure 8.8. These complexes are reduced by butane to yield the adsorbed cyanide and isocyanide groups, which decompose further to N2 and CO2 [56]. The importance of adsorbed NO2 has been underlined in a combined IR/kinetic study focusing on the NO-assisted N2 O decomposition [57]. A peroxide ion was proposed to be another active intermediate in the reaction based on the EPR and UV–Raman data [58]. A mechanism that involves the coordination of molecular oxygen as a peroxide ion with its further conversion into a bridged di-oxygen was proposed. In the case of the overexchanged FeZSM-5 catalyst materials, the existence of binuclear iron complexes has been demonstrated by EXAFS spectroscopy [59]. Distorted octahedral FeIII sites in these complexes are highly reactive, capable of breaking the N–O bond [59], as was shown in the study of SCR of NO with isobutene [60]. Treatment with isobutene alone results in a slight reduction of

223

8 Catalytic Reactions Within Molecular Sieves 3.5

Fen +(NO2) and Fen +(NO3) 1620

3.0

1602 1635 1577 NO+

2.5 Absorbance

224

Fe2+(NO) 1876

1549

298 K in NO and O2 2.0 323 K 373 K

1.5

423 K 473 K

1.0 523 K 573 K 623 K

0.5 673 K 2400

2200

2000

1800

1600

1400

Wavenumber (cm−1)

Figure 8.8 Infrared spectra of FeZSM-5 taken during a temperature ramp exposed to 5000 ppm NO and 1% O2 is passed over the catalyst after it had been exposed to this mixture for 20 minutes at room temperature. Nitrite and nitrate regions are indicated. (Copyright Springer 1999; Reproduced with permission from [56].)

iron. On the contrary, feeding the catalyst materials with NO and oxygen brings the average iron oxidation state up from 2.3 to 2.8. The number of Fe-neighboring atoms increases, indicating the partial reoxidation of binuclear sites or formation of N-coordinated species. In the SCR conditions, complete reoxidation of iron occurs, indicating that the reactivity originates from oxygen vacancies. Unfortunately, as EXAFS was unable to distinguish between light-scattering elements, the exact structure of the reactive complex could not be recovered. In another EXAFS study, the spectra of the catalyst before and after reaction with N2 O were found to be identical, suggesting that only a very small concentration of sites, which is not detectable by EXAFS, is responsible for the activity [61]. In situ M¨ossbauer spectroscopy also gives evidence toward formation of binuclear iron complexes. The spectra of the active species were shown to be very similar to those of the enzymatic Fe–Fe complex. These species comprise more than 60% of the total metal content. Reduced Fe2+ ions were shown to reversibly oxidize to Fe3+ by nitrous oxide, generating active α-oxygen species bringing about oxidation activity [62]. Interestingly, both iron atoms are capable of α-oxygen generation,

8.4 Selection of Zeolite-Based Catalytic Reactions

hence Dunbkov et al. argued that α-sites are, in fact, monatomic entities in a paired arrangement spectroscopically registered as dinuclear complexes. Unlike in the case of hydrocarbons, when ammonia is used as a reducing agent, ammonium nitrite was shown to be formed by IR spectroscopy [63]. At higher temperatures, nitrite decomposes to N2 and H2 O. The effect of water on the reaction rate was rationalized. At low temperatures, water competes with ammonia and NOx for adsorption sites, decreasing the nitrite formation. Furthermore, P´erez-Ram´ırez et al. reported on the SCR of N2 O by carbon monoxide [64]. According to the proposed mechanism, which is based on the combined data from UV–vis and EPR spectroscopy, CO removes oxygen species from the Fe–O–Fe complexes, thus, liberating then for N2 O adsorption. A correlation between the concentration of isolated Fe3+ and the N2 O conversion was established. Schwidder et al. assessed the role of Brønsted acidity in the NO SCR [65]. Several Fe-MFI catalysts with similar structure of Fe sites were found to exhibit a signiﬁcantly different performance in SCR of NO with isobutane and with NH3 . The dramatic differences observed have been attributed to their strongly different acidity properties, which have been characterized by IR of adsorbed pyridine. The variations indicate an essential role of Brønsted sites in this reaction. Zeolites exchanged with metals other than copper and iron, for example, nickel and cobalt were also shown to be active in the DeNOx process [66, 67]. Mihaylov et al. studied the process of NO adsorption on NiY and NiZSM-5 zeolites [68]. Only mononitrosyl species were shown to form in the case of NiY zeolite, whereas for NiZSM-5 dinitrosyls were also detected. Nitrites and nitrates formed on NiZSM-5 upon coadsorption of NO/O2 are highly reactive toward SCR by methane: this ﬁnding has been rationalized in terms of electrophilicity and coordinative saturation of Ni ions in zeolite environments. The lower coordination number of Ni2+ in ZSM-5 framework allows for simultaneous coordination of the nitrate and the reducing hydrocarbon molecule. 8.4.2 Methanol-to-Oleﬁn Conversion

Methanol is a valuable chemical, which can be made from the synthesis of gas and further converted into light alkenes and gasoline-range hydrocarbons. The zeolite-catalyzed conversion of methanol to hydrocarbons (MTH) is commonly referred to as MTO and MTG (methanol to gasoline) depending on the desired products. The MTH chemistry and its commercial potential have been known for decades. In 1986, a New Zeeland MTG facility was started up, but, due to a drop in crude oil prices, only the methanol synthesis step was left on stream. The Topsøe integrated gasoline synthesis (TIGAS), based on H-ZSM-5 as catalyst, was demonstrated on a pilot scale, but has never been scaled up. Later, more attention was paid to the MTO reaction, that is, polymer-grade ethene and propene from methanol using the UOP/Norsk hydro technology based on the silicoaluminophosphate H-SAPO-34.

225

226

8 Catalytic Reactions Within Molecular Sieves

In a simpliﬁed route of the MTO reaction, methanol is dehydrated to form dimethyl ether and water, followed by formation of alkenes. The long-standing question has been the speciﬁc mechanism behind this latter step, in which C–C bonds are formed from oxygenates (methanol/dimethyl ether). In the ﬁrst in situ IR study of MTO chemistry, ZSM-5 catalysts were exposed to methanol and dimethyl ether at elevated temperatures [69]. Methoxy groups formed upon protonation of organic molecules by zeolite hydroxyl groups and subsequent elimination of water were identiﬁed for both reactants. These species were suggested to be essential for the formation of the ﬁrst C–C bond. The reactivity of a methoxy species depends on the C–O bond strength, which is determined by the acid strength of the zeolite hydroxyl species. The authors noticed that the conclusions made are most probably relevant to the initial step of the reaction [69]. Comparison of various HZSM-5 zeolites containing different concentrations of framework and extra-framework aluminum has shown that several adsorbed species formed upon methanol adsorption, that is, methanol hydrogen-bonded to Brønsted acid sites, chemisorbed methanol as methoxy groups formed on Brønsted acid sites, silanol groups and extra-framework AlOH sites [70]. Dimethylether was found to form hydrogen bonds with silanol groups and Brønsted acid sites, and forms the same methoxy species generated from methanol at higher temperatures. Analysis of the IR spectra and the initial hydrocarbon products for different zeolite samples indicated that the Brønsted acid sites and the methoxy groups formed on them are the key to hydrocarbon formation. Extra-framework aluminum and the methoxy groups formed at silanol and AlOH were not found to play any direct role in the catalytic reaction. Solid-state MAS NMR has also proved to be an excellent tool to study reactive intermediates participating in the MTO reaction pathways. A number of ex situ studies, in which the catalyst samples were exposed to methanol, thermally treated, and subsequently quenched, exist. In the pioneering work by Anderson and Klinowski [71], the speciation of the aromatics formed in the zeolite channels was carried out. Interestingly, the distribution of the polymethylated benzenes was found to signiﬁcantly deviate from thermodynamic equilibrium. The 1,2,4,5-tetramethylbenzene is formed in larger quantities than the thermodynamically favored 1,2,3,5-tetramethylbenzene. The authors rationalized this ﬁnding in ˚ which ﬁts into the ZSM-5 terms of the smaller size of the former (6.1 vs 6.7 A), channel intersection dimensions. Other aromatic species formed included xylenes and smaller amounts of tri-, penta-, and hexamethylbenzenes. Direct observation of the induction period (when an equilibrium between methanol, dimethyl ether, and water was established) and the onset of hydrocarbon synthesis with the primary product being ethylene have been observed with in situ NMR [72]. These ﬁndings suggested ethylene to be the ‘‘ﬁrst’’ oleﬁn or, in other words, the ﬁrst C–C bond formed. The formation of signiﬁcant amounts of methane was explained by abstraction of the hydride by the active surface methoxy species. When multinuclear solid-state NMR was applied to study the adsorption of methanol on the H-ZSM-5 zeolite, formation of hydrogen-bonded neutral methanol molecules as well as partially protonated clusters with three methanol molecules

8.4 Selection of Zeolite-Based Catalytic Reactions

coordinated a bridged hydroxyl group [73]. These clusters with up to four methanol molecules could be observed also in zeolite Y [74]. However, in the case of H-β, silanol groups also participate in methanol absorption, forming weak H-bonds. Signiﬁcantly large adsorbate complexes with up to seven molecules per SiOHAl or SiOH group were shown to form in the latter case. It was established that the initial introduction of oleﬁns into the feed reduces the induction period and enhances the catalytic activity through the formation of the aromatic species [75]. Haw et al. investigated the interaction of ethene with ZSM-5 with MAS NMR using a pulse-quench catalytic reactor [76]. During the induction period, signals of cyclopentenyl carbenium ion appeared. These species are suggested to be key intermediates in the working MTO catalysts, as they act as reservoirs of cyclic dienes which are easier to methylate. Further NMR work, ﬁrst published in the patent literature [77, 78], has given direct insight into the chemical nature of the catalytic scaffold formed within the zeolite channels, in which the formation and breakage of carbon–carbon bonds takes place. This so-called hydrocarbon pool was found to contain benzenes, naphthalenes, and methylated derivatives thereof. Direct evidence toward the formation of methylated aromatics was reported by Song et al. for SAPO-34 [79]. The average number of methyl groups per aromatic ring was shown to reach a maximum of about 4, which are consumed during the reaction (Figure 8.9) [80]. No induction period was observed when the catalyst was pretreated with methanol pulse, allowing the aromatic pool to be formed. In a subsequent study, the naphthalene moiety methylated with up to four groups was found to be the major participants in the process [81]. Wang et al. have demonstrated in a continuous ﬂow experiment that surface methoxy groups are very active and capable of methylating toluene and cyclohexane, paving the way toward the hydrocarbon pool [82]. An experiment in which the feed of 13 C-enriched methanol was switched to 12 C-methanol has also substantiated the relevance of the alkylated aromatics [83]. Upon the switch, the 13 C NMR signal of alkyl group decreased, indicating that those species participate in alkylation and splitting-off reactions. A coupling of NMR to optical ﬁber-based UV–vis spectroscopy accomplished by Hunger et al. allowed the identiﬁcation of the formation of cyclic compounds and carbenium ions characterized by absorption at 300–400 nm at temperatures as low as 413 K [84]. The same combination was also used to study the coking and regeneration of H-SAPO-34 [85]. At temperatures up to 623 K, polyalkylaromatics and enylic carbenium ions were registered from the NMR and UV–vis spectra, respectively. On average, about 0.4 aromatic rings per chabazite cage are formed in the reaction with one to four methyl groups per aromatic ring, depending on the temperature. At 673 K, the formation of polymethylated anthracenes has become evident, which are thought to be responsible for the deactivation of the catalyst occurring in these conditions. Treatment in air at 773 K was found necessary to nearly completely remove both substituted aromatic and polyaromatic compounds. UV–vis and confocal ﬂuorescence microspectroscopy is another tool to assess not only the nature of the carbonaceous species leading to deactivation of MTO

227

228

8 Catalytic Reactions Within Molecular Sieves

130 134

20 25

Meave

60 min ∗

∗

1.9

12 min ∗

∗

2.7

6 min ∗

∗

3.0

3 min ∗

∗

3.7

1 min ∗

∗

4.8

0 min ∗

∗

5.6

250

200

150

100 ppm

50

0

Figure 8.9 13 C CP-MAS NMR spectra (75 MHz) showing the loss of methyl groups as a function of time from methylbenzenes trapped in the SAPO-34 cages at 673 K. For each case, a fresh catalyst bed was used

−50

to convert 0.1 ml. The average numbers of methyl groups per ring, Meave , are shown. (Copyright American Chemical Society 2001; Reproduced with permission from [81].)

catalysts but also their preferential location within a zeolite particle [86]. Mores et al. have demonstrated that in ZSM-5 coke initially forms in the near-surface area of the crystals and gradually diffuses into the particle (Figure 8.10). In the case of SAPO-34, the formation of aromatic coke precursors is limited to the near-surface region of SAPO-34 crystals, thereby creating diffusion limitations for the coke front moving toward the middle of the crystal during the MTO reaction. From the spectral and spatially resolved data, graphite-like coke deposited on the external crystal surface in the case of ZSM-5 and aromatic species formed inside the zeolite channels (for both catalysts) could be distinguished. IR spectroscopy was applied to investigate the inﬂuence of the acid site density on performance in the MTO reaction [87]. It was shown that within the series of mordenite samples with Si/Al ratio varying in the range 5–105 rapid accumulation of polyaromatic compounds occurs in the Al-enriched catalyst materials. The formation of compounds with three or four fused aromatic rings leads to the loss of the active alkylbenzene intermediates as well as to the blockage of mordenite channels. On the other hand, the sparse distribution of acid sites in more siliceous

8.4 Selection of Zeolite-Based Catalytic Reactions

0′

4′

7′

9′

10′ 12′

15′

17′

55′

60′

(a) (c)

(b) Figure 8.10 Fluorescent carbonaceous species formed in H-ZSM-5 crystal during MTO reaction with time-on-stream. Confocal slices are taken at laser excitation (a) 488 nm and (b) 561 nm and the schematic

representation of the slice where the confocal ﬂuorescence measurement has been performed (c). (Copyright Wiley-VCH 2008; Reproduced with permission from [85].)

samples suppresses the condensation of alkylbenzenes in their pores and lengthens the catalyst lifetime. In the work of Park et al., several zeolites with different pore topologies and acidities were compared, namely, SAPO-34, ZSM-5, H-β, mordenite, and HY [88]. All the zeolites were shown to exhibit initial high conversion of methanol; however, their deactivation rates varied signiﬁcantly, in more detail, activity of SAPO-34, ZSM-5, and H-β were maintained even after several hours on stream, whereas mordenite showed rapid deactivation. IR spectroscopy was used to follow the formation of the aromatic intermediates. Intensity of the band at 1465 cm−1 assigned to methylbenzenes was found to strongly correlate with the number of strong acid sites, leading to the conclusion that only these sites are relevant for the MTO process. The absorption band at around 1589 cm−1 due to polyaromatic compounds dominated in the case of HY, as shown in Figure 8.11. It is suggested that deactivation is related to the condensation of alkylbenzenes to large molecules in large cavities of these materials, and subsequent pore blockage. On the contrary, smaller cages of SAPO-34, ZSM-5, and H-β zeolites suppressed the formation of polyaromatics as evidenced from the IR spectra shown in Figure 8.11. As a result, the catalysts show a longer lifetime. On the basis of this spectroscopic and kinetic data, the authors deduced a relation between the pore and cavity geometry, product distribution, and deactivation behavior.

229

Absorbance

2 min 3600

2962 2863

3200

2400

2000

Wavenumber (cm−1)

2800

FAU 1337

1600

1200

Absorbance 90 min

2 min 3600

0.1

2925 2870

3200

2400

2000 Wavenumber (cm−1)

2800

CHA

1605

1589

1600

1200

Figure 8.11 In situ IR spectra of materials adsorbed and occluded on FAU (a) and SAPO-34 (b) zeolites during the MTO reaction: reaction temperature = 623 K. (Copyright Elsevier 2009; Reproduced with permission from [88].)

90 min

0.1

1461 1380

230

8 Catalytic Reactions Within Molecular Sieves

8.5 IR Microspectroscopy

8.5 IR Microspectroscopy

One of the emerging investigation methods in the ﬁeld of catalysis is microspectroscopy. Examples include UV–vis, Raman, X ray, ﬂuorescence, and magnetic resonance imaging (MRI) techniques. One of these methods is IR microscopy, which allows obtaining vibrational spectra from sample regions as small as small as a few micrometers. Large (up to hundreds of micrometers) zeolite crystals are commonly used for IR microspectroscopic studies on molecular diffusion, template decomposition, and catalytic reactivity. The usage of IR microspectroscopy in the ﬁeld of zeolites has to the best of our knowledge been pioneered by van Bekkum et al. to analyze element distribution in boron-ZSM-5. Later on, in a ﬁrst in situ study, the template decomposition in ZSM-5 crystals focused on the template decomposition process in ZSM-5 crystals [89]. Three steps were distinguished: (i) increasing the mobility of the tetrapropylammonium cations; (ii) partial oligomerization of the template fragments (both up to 613 K), and (iii) above 633 K, decomposition of the template via Hoffman elimination, with the formation of either dipropylamine or dipropylammonium ions, depending upon the aluminum concentration in the zeolite. Sch¨uth used polarized IR microscopy to reveal the molecular orientation of adsorbates in silicalite crystals [90]. It was shown that the IR spectrum of the adsorbed p-xylene strongly depends on the light polarization. Analysis of the IR bands and their polarization dependence indicated that the on-adsorbed p-xylene molecules align with the straight ZSM-5 channels. The method also elegantly revealed areas with different pore orientation. However, these ﬁndings were not interpreted, as the intergrowth architecture of ZSM-5 crystals has not been unambiguously known. Acidity of zeolites can also be probed in a spatially resolved manner using pyridine as a probe molecule [91]. Both OH and pyridine ring vibration spectral regions showed that acidic sites are more abundant in the center of the ZSM-5 crystal. Interestingly, no correlation of the acidic with Al concentration gradient measured with X-ray microprobe analysis has been observed (Figure 8.12). IR microspectroscopy has proven to be an excellent tool to follow the diffusion of organic molecules in zeolites, especially when combined with interference microscopy. In a ﬁrst study of this kind, an approach for the measurement of adsorbate concentrations in zeolites, based on IR [92, 93], was adopted. With an IR microscope, diffusion coefﬁcients of toluene were determined in individual ZSM-5 crystals varying in their degree of intergrowth [94]. Severely intergrown crystals showed diffusion coefﬁcients three orders of magnitude lower then perfect single crystals. It was proposed that this signiﬁcant effect is due to different channel orientation in single and twinned crystals. Chmelik et al. reported on the effect of surface modiﬁcation of MFI crystals [95]. In an attempt to disentangle the effect of surface and bulk diffusion barriers, the crystals were treated with methylated silanes, which block the pores and effectively increase the surface barrier. The rates of diffusion into and out of the crystals were

231

8 Catalytic Reactions Within Molecular Sieves 1.2

1.2

c

1 0.8

0.2

(a)

a

0.6 0.4

c

1 Absorbance

Absorbance

232

q = 0°

(b)

a

0.6 0.4 0.2

q = 90°

0 3300 3200 3100 3000 2900 2800 2700 Wavenumber (cm−1)

0.8

q = 0°

q = 90°

0 3300 3200 3100 3000 2900 2800 2700 Wavenumber (cm−1)

Figure 8.12 Spatially resolved polarized IR spectra of a p-xylene loaded silicalite I crystal. θ = 0◦ corresponds to polarization perpendicular to the crystal’s long axis. Crystal orientated with (010) parallel to beam axis. (a) Middle of crystal. (b) End of crystal. (Copyright American Chemical Society 1992; Reproduced with permission from [89].)

shown to be correlated well with the strength of the surface resistance, with the untreated samples exhibiting the fastest uptake/release rates and crystals exposed to tripropylchlorosilane revealing very slow or no uptake, which indicates a nearly complete pore blockage. In a recent report, IR microspectroscopy was applied to study alkane diffusion and adsorption in metal-organic frameworks (MOFs) [96]. It should be noted that the most prohibiting limitation of the conventional IR microscope is its spatial resolution, which cannot be improved beyond ∼20 µm with the conventional light source. One of the ways to overcome this problem is to use synchrotron light with brightness 100–1000 times higher than that of a globar light source. The improved characteristics of a synchrotron-based setup allow bringing the resolution to diffraction-limited values. Using this method, the ﬁrst in situ study of the zeolite-catalyzed styrene oligomerization was carried out, where carbocationic reaction intermediates were identiﬁed on the basis of the spatially resolved IR spectra [97]. Some typical IR spectra and the related 2D IR map during the styrene oligomerization reaction are illustrated in Figure 8.13. Moreover, using inherently polarized synchrotron light, the orientation of the product molecules in the zeolite channels was determined. More speciﬁcally, it was found that the carbocationic ﬂuorostyrene dimer was located in the straight channels of the ZSM-5 crystal.

8.6 Concluding Remarks and Look into the Future

It is now commonly accepted that in situ characterization studies of a catalyst material at work are crucial if one is to comprehend the key catalytic steps in the reaction mechanism [96]. If these steps are well understood, critical improvements

8.6 Concluding Remarks and Look into the Future

30

0.2

46 min

y (µm)

43 min 31 min 28 min 26 min

233

20 0.1 10 0 0

10

20

30

40

x (µm)

21 min 17 min

0 min 1600 1560 1520 1480 1440 1600 1560 1520 1480 1440 (a) Wavenumber

(cm−1

)

(b)

−1

Wavenumber (cm )

Figure 8.13 (a) Spatially resolved IR spectra of an individual H-ZSM-5 crystal (5 µm × 5 µm area) taken in situ during the 4-ﬂuorostyrene oligomerization reaction. A band at 1534 cm−1 (marked by arrow) is due to the carbocationic ﬂuorostyrene dimer. (b) Intensity of the IR band at 1534 cm−1

1600 1560 1520 1480 1440 Wavenumber (cm−1)

mapped over the crystal after reaction and IR spectra taken from the edge and the body of the crystal, demonstrating differences in the intensity ratio of the bands. (Copyright Wiley-VCH 2008; Reproduced with permission from [96].)

can be made to the catalytic process in order to make it more efﬁcient. Continued improvements in the spectroscopic instrumentation performance (i.e., better time/spatial resolution brought about by more powerful light sources and better detection systems), the embracing of new – often complementary – spectroscopic techniques, and the development of new spectroscopic-reaction setups means that there are more opportunities than ever to obtain fundamental insight into catalytic processes within molecular sieves, making these truly exciting times for the catalyst scientist. Vibrational spectroscopy takes a special as well as central position in the development of structure–activity relationships for heterogeneous catalysis as it is one of the few characterization methods, which provide detailed molecular insight in the adsorbed species on the catalyst surface, including reaction intermediates as well as deactivation products. As shown in this chapter, this has been especially fruitful in characterizing acid and redox active centers in molecular sieves. It would be highly welcomed if this vibrational information could be conﬁned to the nanoscale of molecular sieves in order to complement the detailed knowledge available from electron microscopy methods. A way to do so is to combine vibrational spectroscopy measurements with near-ﬁeld optical methods. This is the ﬁeld of scanning near ﬁeld optical microscopy (SNOM), including tip-enhanced Raman spectroscopy and scanning near ﬁeld infrared microscopy (SNIM). Although these methods are still in their infancy and appropriate in situ cells are not yet developed, mainly due to the instability of the

234

8 Catalytic Reactions Within Molecular Sieves

reﬂection tip, it would offer the possibility to relate spatial heterogeneities in catalytic solids to speciﬁc reactivity and deactivation patterns. With this knowledge in mind, advanced nanoscale structuring of catalytic solids could become within reach.

Acknowledgment

E.S. and B.M.W. acknowledge the Netherlands Organization for Scientiﬁc Research (NWO-CW) for ﬁnancial support (Veni, Vici and Top grants).

References 1. Banares, M.A. and Wachs, I.E. (2002) J. 2. 3.

4.

5. 6.

7. 8.

9.

10.

11. 12.

13.

Raman Spectrosc., 33, 359. Weckhuysen, B.M. (2003) Phys. Chem. Chem. Phys., 5, 4351. Eischens, R.P., Plisken, W.A., and Francis, S.A. (1954) J. Chem. Phys., 24, 1786. Eischens, R.P. and Pliskin, W.A. (1958) in Advances in Catalysis and Related Subjects, Vol. 10 (eds D.D. Eley, W.G. Frankenburg, and V.I. Komarewsky), Academic Press, New York, p. 2. Paukshtis, E.A. and Yurchenko, E.N. (1983) Russ. Chem. Rev., 52, 242. Lercher, J.A., Gr¨undling, C., and Eder-Mirth, G. (1996) Catal. Today, 27, 353. Knozinger, H. and Huber, S. (1998) J. Chem. Soc., Faraday Trans., 94, 2047. Gil, B., Zones, S.I., Hwang, S.J., ˇ Bejblova, M., and Cejka, J. (2008) J. Phys. Chem. C, 112, 2997. Hoffmann, P. and Lobo, J.A. (2007) Microporous Mesoporous Mater., 106, 122. Holm, M.S., Svelle, S., Joensen, F., Beato, P., Christensen, C.H., Bordiga, S., and Bjorgen, M. (2009) Appl. Catal. A: Gen., 356, 23. Makarova, M.A. and Dwyer, J. (1993) J. Phys. Chem., 97, 6337. Zecchina, A., Spoto, G., and Bordiga, S. (2005) Phys. Chem. Chem. Phys., 7, 1627. Bevilacqua, M., Alejandre, A.G., Resini, C., Casagrande, M., Ramirez, J., and Busca, G. (2002) Phys. Chem. Chem. Phys., 4, 4575.

14. Zholobenko, V.L., Makarova, M.A., and

15. 16.

17.

18.

19.

20.

21. 22.

23. 24.

25. 26. 27.

Dwyer, J. (1993) J. Phys. Chem., 97, 5962. Datka, J., Sulikowski, B., and Gil, B. (1996) J. Phys. Chem., 100, 11242. Covarrubias, C., Quijada, R., and Rojas, R. (2009) Microporous Mesoporous Mater., 117, 118. Zholobenko, V.L., Lukyanov, D.B., Dwyer, J., and Smith, W.J. (1998) J. Phys. Chem. B, 102, 2715. del Campo, A.E.S., Gayubo, A.G., Aguayo, A.T., Tarrio, A., and Bilbao, J. (1998) Ind. Eng. Chem. Res., 37, 2336. Martins, G.A.V., Berlier, G., Coluccia, S., Pastore, H.O., Superti, G.B., Gatti, G., and Marchese, L. (2007) J. Phys. Chem. C, 111, 330. Bordiga, S., Regli, L., Cocina, D., Lamberti, C., Bjorgen, M., and Lillerud, K.P. (2005) J. Phys. Chem. B, 109, 2779. Parry, E.R., (1963) J. Catal., 2, 371. Karge, H.G. and Geidel, E. (2004) in Molecular Sieves: Characterization I (eds H.G. Karge and J. Weitkamp), Springer, Berlin. Lavalley, J.C. (1996) Catal. Today, 27, 377. Barzetti, T., Selli, E., Moscotti, D., and Forni, L. (1996) J. Chem. Soc., Faraday Trans., 92, 1401. Maache, M., Janin, A., Lavalley, J.C., and Benazzi, E. (1995) Zeolites, 15, 507. Datka, J., Gil, B., and Kubacka, A. (1997) Zeolites, 18, 245. Buzzoni, R., Bordiga, S., Ricchiardi, G., Lamberti, C., Zecchina, A., and Bellussi, G. (1996) Langmuir, 12, 930.

References 28. Oliviero, L., Vimont, A., Lavalley, J.C.,

29.

30.

31.

32.

33. 34. 35.

36.

37.

38.

39.

40.

41.

42. 43.

Sarria, F.R., Gaillard, M., and Mauge, F. (2005) Phys. Chem. Chem. Phys., 7, 1861. Nesterenko, N., Thibault-Starzyk, F., Montouillout, V., Yuschenko, V., Fernandez, C., Gilson, J., Fajula, F., and Ivanova, I. (2004) Microporous Mesoporous Mater., 71, 157. Thibault-Starzyk, F., Stan, I., Abell´o, S., Bonilla, A., Thomas, K., Fernandez, C., Gilson, J., and P´erez-Ram´ırez, J. (2009) J. Catal., 264, 11. Ordomsky, V.V., Murzin, V.Y., Monakhova, Y.V., Zubavichus, Y.V., Knyazeva, E.E., Nesterenko, N.S., and Ivanova, I.I. (2007) Microporous Mesoporous Mater., 105, 101. Oumi, Y., Takahashi, J., Takeshima, K., Jon, H., and Sano, T. (2007) J. Porous Mater., 14, 19. Engelhardt, G., Fahlke, B., Magi, M., and Lippmaa, E. (1983) Zeolites, 3, 292. Engelhardt, G., Fahlke, B., M¨agi, M., and Lippmaa, E. (1985) Zeolites, 5, 49. Sankar, G., Okubo, T., Fan, W., and Meneau, F. (2007) Faraday Discuss., 136, 157. Vistad, O.B., Akporiaye, D.E., and Lillerud, K.P. (2001) J. Phys. Chem. B, 105, 12437. Vistad, O.B., Akporiaye, D.E., Taulelle, F., and Lillerud, K.P. (2003) Chem. Mater., 15, 1639. Grandjean, D., Beale, A.M., Petukhov, A.V., and Weckhuysen, B.M. (2005) J. Am. Chem. Soc., 127, 14454. O’Brien, M., Beale, A., Catlow, C., and Weckhuysen, B.M. (2006) J. Am. Chem. Soc., 128, 11744. Fan, F., Feng, Z., Li, G., Sun, K., Ying, P., and Li, C. (2008) Chem. Eur. J., 14, 5125. Tompsett, G., Panzarella, B., Conner, W., Yngvesson, K., Lu, F., Suib, S., Jones, K., and Bennett, S. (2006) Rev. Sci. Instrum., 77, 124101. Szanyi, J. and Paffett, M.T. (1996) J. Catal., 164, 232. Dedecek, J., Sobalik, Z., Tvaruzkova, Z., Kaucky, D., and Wichterlova, B. (1995) J. Phys. Chem., 99, 16327.

44. Wichterlova, B., Dedecek, J., and

45. 46.

47.

48.

49. 50.

51.

52.

53. 54.

55.

56.

57. 58.

59.

60.

61. 62.

Vondrova, A. (1995) J. Phys. Chem., 99, 1065. Pietrzyk, P., Gil, B., and Sojka, Z. (2007) Catal. Today, 126, 103. Kucherov, A.V., Gerlock, T.L., Jen, H.W., and Shelef, M. (1995) Zeolites, 15, 9. Park, S.K., Kurshev, V., Luan, Z.H., Lee, C.W., and Kevan, L. (2000) Microporous Mesoporous Mater., 38, 255. Mathisen, K., Nicholson, D.G., Beale, A.M., Sanchez-Sanchez, M., Sankar, G., Bras, W., and Nikitenko, S. (2007) J. Phys. Chem. C, 111, 3130. Ganemi, B., Bjornbom, E., and Paul, J. (1998) Appl. Catal. B: Environ., 17, 293. Goodman, B.R., Schneider, W.F., Hass, K.C., and Adams, J.B. (1998) Catal. Lett., 56, 183. Groothaert, M.H., van Bokhoven, J.A., Battiston, A.A., Weckhuysen, B.M., and Schoonheydt, R.A. (2003) J. Am. Chem. Soc., 125, 7629. Yashnik, S., Ismagilov, Z., and Anufrienko, V. (2005) Catal. Today, 110, 310. Chen, H. and Sachtler, W.M.H. (1998) Catal. Today, 42, 73. Kucherov, A., Montreuil, C., Kucherova, T., and Shelef, M. (1998) Catal. Lett., 56, 173. Schwidder, M., Grunert, W., Bentrup, U., and Bruckner, A. (2006) J. Catal., 239, 173. Lobree, L.J., Hwang, I.C., Reimer, J.A., and Bell, A.T. (1999) Catal. Lett., 63, 233. Pirngruber, G.D. and Pieterse, J.A.Z. (2006) J. Catal., 237, 237. Gao, Z.X., Kim, H.S., Sun, Q., Stair, P.C., and Sachtler, W.M.H. (2001) J. Phys. Chem. B, 105, 6186. Battiston, A.A., Bitter, J.H., and Koningsberger, D.C. (2000) Catal. Lett., 66, 75. Battiston, A.A., Bitter, J.H., and Koningsberger, D.C. (2003) J. Catal., 218, 163. Pirngruber, G.D., Roy, P.K., and Weiher, N. (2004) J. Phys. Chem. B, 108, 13746. Dubkov, K.A., Ovanesyan, N.S., Shteinman, A.A., Starokon, E.V., and Panov, G.I. (2002) J. Catal., 207, 341.

235

236

8 Catalytic Reactions Within Molecular Sieves 63. Sun, Q., Gao, Z.X., Wen, B., and

64. 65.

66. 67. 68.

69. 70.

71. 72.

73. 74. 75. 76.

77. 78. 79. 80.

81.

Sachtler, W.M.H. (2002) Catal. Lett., 78, 1. Perez-Ramirez, J., Kumar, M.S., and Bruckner, A. (2004) J. Catal., 223, 13. Schwidder, M., Kumar, M.S., Bentrup, U., Perez-Ramirez, J., Brueckner, A., and Gruenert, W. (2008) Microporous Mesoporous Mater., 111, 124. Li, Y.J. and Armor, J.N. (1993) Appl. Catal. B: Environ., 2, 239. Brosius, R. and Martens, J.A. (2004) Top. Catal, 28, 119. Mihaylov, M., Hadjiivanov, K., and Panayotov, D. (2004) Appl. Catal. B: Environ., 51, 33. Forester, T. and Howe, R. (1987) J. Am. Chem. Soc., 109, 5076. Campbell, S.M., Jiang, X.Z., and Howe, R.F. (1999) Microporous Mesoporous Mater., 29, 91. Anderson, M. and Klinowski, J. (1989) Nature, 339, 200. Munson, E.J., Kheir, A.A., Lazo, N.D., and Haw, J.F. (1992) J. Phys. Chem., 96, 7740. Hunger, M. and Horvath, T. (1996) J. Am. Chem. Soc., 118, 12302. Hunger, M. and Horvath, T. (1997) Catal. Lett., 49, 95. Dahl, I.M. and Kolboe, S. (1994) J. Catal., 149, 458. Haw, J.F., Nicholas, J.B., Song, W.G., Deng, F., Wang, Z.K., Xu, T., and Heneghan, C.S. (2000) J. Am. Chem. Soc., 122, 4763. Song, W.G., Fu, H., and Haw, J.F. (2001) J. Am. Chem. Soc., 123, 4749. Xu, T. and White, J.L. (2004) US Patent 6,734,330. Xu, T. and White, J.L. (2004) US Patent 6,743,747. Song, W.G., Haw, J.F., Nicholas, J.B., and Heneghan, C.S. (2000) J. Am. Chem. Soc., 122, 10726. Song, W.G., Fu, H., and Haw, J.F. (2001) J. Phys. Chem. B, 105, 12839.

82. Wang, W., Buchholz, A., Seiler, M., and

83. 84. 85.

86.

87.

88. 89. 90. 91. 92. 93. 94.

95.

96.

97.

Hunger, M. (2003) J. Am. Chem. Soc., 125, 15260. Seiler, M., Wang, W., Buchholz, A., and Hunger, M. (2003) Catal. Lett., 88, 187. Hunger, M. and Wang, W. (2004) Chem. Commun., 584. Jiang, Y., Huang, J., Marthala, V.R.R., Ooi, Y.S., Weitkamp, J., and Hunger, M. (2007) Microporous Mesoporous Mater., 105, 132. Mores, D., Stavitski, E., Kox, M.H.F., Kornatowski, J., Olsbye, U., and Weckhuysen, B.M. (2008) Chem. Eur. J., 14, 11320. Park, J.W., Kim, S.J., Seo, M., Kim, S.Y., Sugi, Y., and Seo, G. (2008) Appl. Catal. A: General, 349, 76. Park, J.W. and Seo, G. (2009) Appl. Catal. A: Gen., 356, 180. Nowotny, M., Lercher, J., and Kessler, H. (1991) Zeolites, 11, 454. Schuth, F. (1992) J. Phys. Chem., 96, 7493. Schuth, F. and Althoff, R. (1993) J. Catal., 143, 388. Niessen, W. and Karge, H.G. (1993) Microporous Mesoporous Mater., 1, 1. Karge, H.G. and Niessen, W. (1991) Catal. Today, 8, 451. Muller, G., Narbeshuber, T., Mirth, G., and Lercher, J.A. (1994) J. Phys. Chem., 98, 7436. Chmelik, C., Varmla, A., Heinke, L., Shah, D.B., Karger, J., Kremer, F., Wilczok, U., and Schmidt, W. (2007) Chem. Mater., 19, 6012. Chmelik, C., Kaerger, J., Wiebcke, M., Caro, J., van Baten, J.M., and Krishna, R. (2009) Microporous Mesoporous Mater., 117, 22. Stavritski, E., Kox, M.H.F., Swart, I., de Groot, F.M.F., and Weckhuysen, B.M. (2008) Angew. Chem. Int. Ed., 47, 3543.

237

9 Textural Characterization of Mesoporous Zeolites Lei Zhang, Adri N.C. van Laak, Petra E. de Jongh, and Krijn P. de Jong

9.1 Introduction

Zeolites and related zeolite-analog materials have found major industrial applications as catalysts and adsorbents, especially in reﬁnery and petrochemical processes where zeolite catalysts exhibit unique shape selectivity endowed by their crystalline yet highly porous structures with well-deﬁned channels of molecular dimensions [1–5]. However, as one side effect of their microporous structure, zeolite catalysts often suffer from restricted diffusion of guest species [6, 7]. Mass transport to and from the active sites inside the micropores (known as conﬁgurational diffusion) is much slower than that of molecular and Knudsen diffusion, which has led to lower catalyst utilization, and sometimes fast deactivation due to coke formation. Different approaches have been proposed to alleviate the diffusion limitation and enhance the accessibility of the internal sites [8–21]. One strategy is to synthesize novel zeolitic structures with a larger pore size. Various large-pore zeolites and zeolite analogs have been obtained, for example, VPI-5 [22], UTD-1 [23], SSZ-53 and SSZ-59 [24], ITQ-15 [25], ITQ-21 [26], ITQ-33 [27], and ITQ-37 [28]. However, despite the fact that a considerable amount of knowledge has been gained on the formation mechanisms of zeolites and numerous types of theoretical structures have been proposed for predicative synthesis, synthesis by design of novel zeolite structure is still a challenging topic, and most new structures are discovered by trial-and-error processes. Another strategy is to decrease the effective intracrystalline diffusion path length, which has given rise to the so-called hierarchical zeolites, that is, zeolites featuring multiscale porosity other than microporosity [12, 15]. This can be achieved by generating intracrystalline mesopores or, alternatively, by downsizing the zeolite crystals, thereby increasing the intercrystalline porosity. Among the typical examples are the well-known ultrastable Y (USY) family materials featuring mesopores [29, 30], which can be obtained by dealumination via steam treatments and are being widely used in the petrochemical industry. Over the past decades, a wealth of

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

238

9 Textural Characterization of Mesoporous Zeolites

methods have been practised and proven to be effective for generating additional meso- and macropores in zeolite materials. Intercrystalline pores can be tuned by assembling nanosized zeolites or by depositing them onto a porous support, for example, mesoporous silicas [31–33]. Methods for obtaining intracrystalline pores can be generally categorized into two types, that is, templating methods [10] and postsynthesis treatments [11, 21]. For the templating method, various materials have been employed as space ﬁllers during zeolite synthesis, including ‘‘soft templates,’’ for example, surfactant [34–36], polymer [37, 38], and starch [39], and ‘‘hard templates,’’ for example, carbon materials (carbon nanoﬁber [40], carbon black [41], carbon nanotube [42], carbon aerogel [43–45]), and nanosized CaCO3 [46]. Postsynthesis treatment methods normally involve demetallization, that is, selective leaching of the framework-constituting species, for example, dealumination by steaming and/or acid leaching [21, 47–59], desilication by base leaching [11, 60–65], and detitanation with hydroperoxide [66–69], thereby leaving voids inside the zeolite particles. Depending on the framework structure and composition of the zeolite, different methods show different features in generating additional porous structure with regard to pore size, pore shape, pore volume, tortuosity, and connectivity. These methods may be complementary to one another. Recently, a synthesis approach combining two methods (template + base leaching) has also emerged [70]. Owing to the presence of the hierarchical porous structures, improved diffusion and catalysis properties have generally been observed [59, 71–76]. Despite the rapid enrichment of the toolkit for generating additional porosity, detailed and explicit characterization of these porous structures has received less attention. Although a wealth of techniques are now available for assessment of the porous structure [77], with regards to zeolites gas physisorption (nitrogen and argon) and transmission electron microscopy (TEM) are still the prevailing methods, which provide information on pore size, pore volume, and limited information on pore shape and connectivity. Hierarchy does not mean simple combination of micropores and meso- or macropores, but rather how the pores at different levels are organized and interconnected in three dimensions [12, 15], as the latter often determines the performance of the materials in practical uses [78]. Other potential techniques, which have been specially developed for evaluating porous materials, especially mesoporous materials, have been largely ignored for the characterization of mesoporous zeolite materials. It is the intention of the current chapter to highlight these techniques as well as some novel techniques and bring them to the front stage of characterization of hierarchical zeolites. As shown below, unprecedented information about the porous structure of hierarchical zeolites has been obtained by these characterization methods, which complement the results from the conventional gas physisorption and TEM analyses. We ﬁrst review the various approaches for generating additional pores, then emphasize on the different techniques that have been used for characterization of these materials. The characterization methods are sectioned according to different techniques involved, although some of them actually rely on similar theories.

9.2 Methods for Generating Meso- and Macropores in Zeolites

9.2 Methods for Generating Meso- and Macropores in Zeolites

Recent years have seen the rapidly growing diversity of the protocols for generating additional pores in zeolite materials, as well as the wide applicability of these methods to various zeolite structures. A number of excellent reviews concerning the synthesis approaches have been published [8–21]. In this section, a short summary of the synthesis methods of hierarchical zeolites featuring meso- and/or macropores is given. 9.2.1 Postsynthesis Modiﬁcation

Postsynthesis modiﬁcation generally involves selective leaching of the framework species by treating the zeolites with steam, acid, base, or complexing agents, thereby leaving additional voids. This can be achieved by dealumination, desilication, or selective leaching of other constituting elements of the zeolite framework. 9.2.1.1 Dealumination Among the various postsynthesis treatment methods, dealumination by steaming and/or acid leaching is probably the most renowned one. Steaming is usually performed at a temperature above 500 ◦ C by contacting the ammonium or proton form of the zeolites with steam. During the treatment, Al–O–Si bonds undergo hydrolysis and aluminum is expelled from the framework, causing vacancy defects (silanol nest) and partial amorphization of the framework. Some less stable and mobile silicon species migrate and condense with silanols at other sites. Such a healing process results in the ﬁlling of some vacancies and large voids originating from expelled aluminum and mobile silicon species [79, 80], as depicted in Figure 9.1. In regions of high defect concentrations, spherical mesopores can coalescence into cylindrical pores. As amorphous debris is deposited on the mesopore surface or on the external surface of the treated zeolite crystals, which causes partial blockage of the micropores, subsequent mild acid washing is often necessary to remove these species. Diluted mineral acids (nitric acid and hydrochloric acid) or organic acid (for example, oxalic acid) can be used for this purpose. According to such a mechanism, the formation of mesopores is highly dependent on the Al concentration and the stability of Al sites against hydrolysis. Therefore, most work on steaming has been performed on zeolites with low pristine Si/Al ratios, for example, zeolite Y [29, 30, 48, 51, 80–88] and mordenite [82, 89–91]. Other examples include mazzite [92, 93], omega [50], ferrierite [94], and ZSM-5 [54, 95]. Dealumination can also be performed by only acid leaching with concentrated acid solutions. According to the same reasoning as indicated above, the effectiveness of this method also depends on the framework types and compositions of the zeolites used. In the case of mordenite, which has pseudo 1D channels, this method has been found to be especially effective and has been extensively investigated [59, 91, 96–99]. A correlation has been found between the extent of mesopore

239

240

9 Textural Characterization of Mesoporous Zeolites Al extraction

1

Si migration

Si 2

Si

3

Ultrastable Y

Figure 9.1 Schematic presentation of the formation of mesopores. The grid represents the zeolite framework, the black dots are framework aluminum atoms, the open circles are aluminum atoms extracted from the framework, and the dotted lines indicate the mesopores. (Adapted from [79].)

generation and the so-called symmetry index, that is, the ratio of the XRD (X-ray diffraction) peak intensities ([111] + [241]/[350]) [100]. The symmetry index has been proposed as an indicator of the extent of stacking faults inside the framework. Dealumination is assumed to take place preferentially at these stacking faults. The mesoporous mordenites thus obtained exhibit a 3D structure and have been applied in industrial processes for the transalkylation of diisopropyl benzene and benzene to yield cumene and (hydro)isomerization of alkanes.

9.2 Methods for Generating Meso- and Macropores in Zeolites

Dealumination leads, apart from mesopore generation, to an increase in the framework Si/Al ratios, thereby enhancing hydrophobicity and (hydro)thermal stability of the zeolites. One of the most well-known examples of mesoporous zeolites obtained by dealumination is the family of USY and related materials, which are invaluable industrial cracking catalysts. Meanwhile, the increased Si/Al ratios also lead to a decrease in the acid site density and consequently enhance the acidity of the acid sites, both exerting strong effects on the catalysis performance. Besides acids, other chemicals can also be used to withdraw aluminum from the framework; examples are SiCl4 [83, 101], (NH4 )2 SiF6 [83, 102], and EDTA (ethylenediaminetetraacetic acid) [57, 103]. However, caution should be taken to avoid severe framework collapse when the extraction of aluminum by EDTA or (NH4 )2 SiF6 is faster than the migration of silicon species during the treatment [104]. 9.2.1.2 Desilication As with aluminum, framework silicon can also be selectively extracted from the framework to generate mesopores. This has mostly been done by treating a zeolite with a base, for example, NaOH, KOH, LiOH, NH4 OH, and Na2 CO3 , or a speciﬁc acid like HF. While base treatment is usually used for the removal of amorphous gel impurities from zeolite crystals, its potential for generating mesopores has been long ignored. Dessau reported in 1992 that hollow crystals were obtained by reﬂuxing large ZSM-5 crystals in Na2 CO3 aqueous solution [60]. It was revealed that highly selective dissolution of the interior of the crystals had occurred, while the exterior surface remained relatively intact. This result hinted toward the inhibiting role of aluminum during the base treatment and gave direct evidence of aluminum zoning in large-crystal ZSM-5 synthesized in the presence of quaternary directing agents. Suboti´c et al. further investigated the role of aluminum during the base treatment of ZSM-5 [105, 106]. However, the evolution of the porous structure was not investigated in detail. Ogura et al. reported the ﬁrst explicit evidence of mesopore formation in ZSM-5 crystals by NaOH treatment [62]. Groen et al. reported the detailed investigation of base treatment conditions for optimizing the mesopore formation in a series of papers [11, 107–122]. They found that for ZSM-5 crystals there appears to be an optimal window of Si/Al ratio in the parent zeolite, that is, SiO2 /Al2 O3 = 50–100 (molar ratio) to achieve optimal mesoporosity with high mesopore surface areas up to 235 m2 g−1 , while still preserving the intrinsic crystalline and acidic properties (Figure 9.2) [119]. While ZSM-5 is still the most intensively investigated zeolite [39, 63–65, 123–128], desilication has recently also been applied to mordenite [74, 112], beta [113, 129], and ZSM-12 [75]. More recently, hollow nanoboxes of ZSM-5 and TS-1 were obtained by treating the zeolites with tetrapropylammonium hydroxide solution. A dissolution–recrystallization mechanism was proposed for the formation of such unique nanoporous structures [130, 131]. Similar to dealumination, desilication inevitably modulates the framework Si/Al ratios; however, in this case, it resulted in decreased Si/Al ratios. Moreover, some extraframework aluminum species are often observed after the base treatment [132]. Therefore, an additional acid treatment or ion-exchange step is needed to remove these species for opening the micropores and mesopores.

241

242

9 Textural Characterization of Mesoporous Zeolites

Al prevents Si extraction Si/Al ≤ 15

NaOH

Limited mesopore formation

Aluminum Silicon

Si/Al ∼ 25 – 50

NaOH

Si/Al ≥ 200

NaOH

Optimal Si/Al range Mesopores in the range 5 – 20 nm

Excessive Si dissolution Large mesoand macropores

Figure 9.2 Simpliﬁed schematic representation of the inﬂuence of the Al content on the mechanism of pore formation during the desilication treatment of MFI zeolites. (Reprinted with permission from [119].)

9.2.1.3 Detitanation For zeolite containing other metals in the framework, mesopores can also be generated by similar selective leaching method. Schmidt et al. reported that mesoporous ETS-10 titanosilicate can be obtained by postsynthesis treatment with hydrogen peroxide [67–69]. The materials exhibited increased external surface areas and improved performances in the Beckmann rearrangement of cyclohexanone oxime to ε-caprolactam. The various postsynthesis treatment methods mentioned above have proved effective for generating mesopores in zeolites. The intracrystalline diffusion pathlength is shortened by virtue of the mesopores. Improved diffusion [59, 74, 75] and catalytic properties [39, 67, 112–114, 123–125, 133] have been generally observed. However, it is still difﬁcult to generate the additional pores in a controllable way in terms of pore shape, pore volume, and connectivity. Moreover, it inevitably involves the change of framework composition, and partial amorphization of the crystalline structure. Therefore, in practical applications, it is sometimes difﬁcult to disentangle the contribution of improved textural properties from that of varied framework composition. Another effect that has received less attention is the change of morphology and surface properties of the zeolite crystals by the postsynthesis treatment [62, 97], which can exert strong effects on the adsorption properties.

9.2 Methods for Generating Meso- and Macropores in Zeolites

9.2.2 Templating Method

A more straightforward method for generating mesopores is the templating method, which has been extensively used for the synthesis of mesoporous materials [16, 134, 135]. Different from the postsynthesis treatment method, templates are employed during the zeolite crystallization and selectively removed afterward. Therefore, the framework composition can be predetermined on the basis of the synthesis gel. Various types of template materials have been employed. On the basis of the structure and rigidity, they can be roughly categorized into hard template and soft template [10, 15]. 9.2.2.1 Hard Template Hard templates here refer to materials with relatively rigid structure that are not supposed to deform during the zeolite synthesis. In this respect, carbon materials are ideal candidates for generating mesopores because of their inertness, rigidity, robustness, diversity, and easy removal by combustion [8, 12, 15, 16]. Carbon was initially used for making nanosized zeolites by the conﬁned growth of zeolite crystals between the spaces of carbon particles [136–138]. Afterward, it was revealed that, by controlling the growth of the zeolite crystals, the carbon templates can be entrapped inside the zeolite crystals [41]. Mesoporous crystals with intracrystalline mesopores can be synthesized by the subsequent removal of carbon via combustion (Figure 9.3). Carbon-templating routes have received wide attention and have developed rapidly during the last few years. Different types of carbon materials, for example, carbon black [40, 139–146], carbon nanoﬁber [40], carbon nanotube [42, 147], carbon Pores created by combustion of carbon particles

Carbon particles about 12 nm

O2

+ CO2

550 °C

About 1 µm zeolite crystal grown in pore system of carbon

Mesoporous zeolite single crystal

Figure 9.3 Schematic presentation of the growth of zeolite crystals around carbon particles. (Reprinted with permission from [41].)

243

244

9 Textural Characterization of Mesoporous Zeolites

aerogel [43–45, 148–150], carbon by sugar decomposition [151, 152], ordered mesoporous carbon [153, 154], and colloid-imprinted carbon (CIC) [155–158] have been used for the synthesis of mesoporous zeolites with various framework types, for example, MFI [40–45, 136–140, 142–144, 148–159], FAU [44, 138, 150], MEL [72, 140, 145, 159, 160], BEA [138, 159], MTW [140, 146, 161], CHA [159], LTA [138], and AFI [159]. It is noted that using ordered nanoporous CIC as the templates, Fan et al. have synthesized MFI single crystals with ordered imprinted mesoporosity [158]. Recently, using the same concept, nanosized CaCO3 [46] and resorcinol-formaldehyde aerogels [162, 163] have also been employed for synthesizing mesoporous zeolites. Similarly, zeolites featuring hierarchical structures can been synthesized via macrotemplating using polystyrene beads [164, 165], resin beads [166, 167], polyurethane foam [168], and even biological materials like bacteria [169], wood [170], and leaves [171, 172]. However, in most of the cases, the products are formed by coating of the templates with zeolite synthesis solution, and occur as polycrystalline ensembles of nanosized zeolites. Alternatively, such macrostructure can also be formed by assembling preformed colloidal zeolite precursor around the templates. 9.2.2.2 Soft Template The concept of soft template is adapted from the micelle templating synthesis of mesoporous materials [173, 174]. Ordered mesoporous materials have been synthesized by assembling zeolite ‘‘seed,’’ with micelle templates. The ‘‘seed’’ can be synthesized either from the precursor solution [175–191], or by decomposition of zeolites [192–195], the so-called top-down approach. Such materials, albeit mostly do not exhibit a discernible zeolite phase, have shown enhanced thermal and hydrothermal stability and enhanced catalytic activity compared to conventional mesoporous materials with amorphous pore walls [175, 176, 180–187, 189, 192–196]. In some diffusion-limited reactions, their catalytic properties are comparable and even surpass those of the microporous zeolite counterparts [175, 185, 186, 189, 193, 194, 197]. Attempts to generate mesopores using supramolecular micelles during zeolite synthesis met little success and, in most of the cases, ended up with a phase-segregated composite of zeolite crystals and mesoporous materials with amorphous pore walls [198–200]. This stressed the importance of modulating the interplay between the formation of the mesoporous structures and the pore wall crystallization. Recently, this problem has been solved by an elegant approach using novel templates of alkoxysilane-containing surfactant or polymer [34, 35, 38]. The presence of the siloxy group increased the interaction between the templates and the pore wall, and helped in the retention of the mesoporous structure during the pore wall crystallization (Figure 9.4). The resulting zeolites showed a highly crystalline structure and uniform mesoporosity. Moreover, the mesopore size can be controlled in a similar manner as for mesoporous materials, for example, MCM-41, by using surfactant molecules with different chain lengths or polymers with different molecular weights. In this sense, the mesoporous zeolite obtained by this approach is a synergetic product, rather than a combination between zeolites and mesoporous materials. Other templates, for example, starch [39, 201] and

9.2 Methods for Generating Meso- and Macropores in Zeolites

Si(OR)3

Si(OR)3

Si(OR)3 Zeolite Nucleation

Si(OR)3

Si(OR)3

Silylated polymer

Proto-Zeolite

Nucleated zeolite–polymer composite Crystal growth

Intracrystal polymer network formation

Figure 9.4 Conceptional approach to the synthesis of a zeolite with intracrystal mesopores using a silylated polymer as the template. (Reprinted with permission from [38].)

polydiallyldimethylammonium chloride [37], have also been used for the synthesis of mesoporous zeolites. Compared to the postsynthesis methods, the templating approach presents several advantages. First, mesopores can be generated without affecting the framework composition, thus making it possible to investigate separately the effects of textural and framework properties. Secondly, in principle, the pore volume, pore shape, and connectivity can be tuned in a more controllable way by choosing proper templates. However, the effects of these templates upon the zeolite composition and phase purity of the ﬁnal product should be taken into account, both during the crystallization process as well as the template removal process. 9.2.3 Other Methods

Mesoporous zeolites can also be synthesized in the absence of templates. As mentioned before, in addition to the intracrystalline pores, mesopores can also arise from the stacking of nanosized zeolites as intercrystalline pores. However, conventional colloidal or nanosized zeolites suffer from the difﬁculties of separation. A solution to this is to synthesize assemblies of nanosized zeolites by controlled nucleation and growth of the zeolite crystals [202, 203]. Alternatively, small zeolite particles can also be deposited onto mesoporous support [32, 33], thereby generating mesoporous composite materials. Gagae et al. [204]

245

246

9 Textural Characterization of Mesoporous Zeolites

and Stevens et al. [205] have obtained zeolite nanocrystals embedded in a mesomatrix by acidifying zeolite seed or precursor solutions. Recently, a new synthesis approach, which combined the sol–gel transformation and the zeolite crystallization, was reported by Wang and coworkers [206–209]. By steam-assisted crystallization of a previously aged precursor gel, mesoporous composite TUD-C with integrated ZSM-5 nanocrystals was obtained [207]. Tetrapropylammonium hydroxide, which was used as the template for the formation of the microporous structure of zeolites, is believed to act also as a scaffolding agent for the mesopore formation. Later on, using a similar concept, another composite material TUD-M was synthesized by reversing the preparation steps, that is, a ﬁrst hydrothermal treatment of the precursor gel followed by the sol–gel transformation [208]. Summarizing this section, we can see that numerous methods are available for generating meso- and macroporous zeolite materials. Each method features its own advantages over the others, and they are complimentary to each other. Another important issue concerning the different approaches for obtaining hierarchical zeolites, which is rarely addressed in the open literature, is the potential for large-scale production and the corresponding costs. For traditional processes, such as dealumination by steaming, engineers have created the means to scale-up from laboratory grams to the tons necessary in industrial processes. For desilication, the fast reaction time of 30 minutes could be problematic, resulting in inhomogeneously treated zeolites. The templating approaches have the disadvantage that they bring additional costs with them, not only in raw chemicals as, for instance, for the micelle route but also in the case of carbon tubes; an additional synthesis step for obtaining the template is also needed. Despite the fact that increasing the scale of a process generally leads to lower costs, upscaling and costs of zeolite production are aspects that should not be overlooked. Nevertheless, in order to arrive at a deeper understanding of the efﬁciency and working mechanisms of these methods, as well as the impacts of these pores on the zeolite performance in practical use, it is vital to have a clear picture of the textural properties and structures of the hierarchical zeolites. In the following part of this chapter, we elaborate on the different characterization techniques for evaluating the textural porosity, with emphasis on some novel techniques (electron tomography (ET), optical microscopy), as well as some well-known techniques that have received less attention (thermoporometry, mercury porosimetry, etc.).

9.3 Characterization of Textural Properties of Mesoporous Zeolites 9.3.1 Gas Physisorption

Among the various techniques available for porous structure analysis, gas physisorption is still the standard and mostly used technique [210]. This has been due to the well-developed theory, as well as the easy operation and wide availability

9.3 Characterization of Textural Properties of Mesoporous Zeolites

of the experimental equipment. This technique accurately determines the amount of gas adsorbed on a solid material at a certain temperature and pressure, and yields important information on the porous structure (pore volume, speciﬁc surface area, pore size distribution (PSD)) and the pore surface properties. Nitrogen and argon are the mostly used adsorbates. Usually isotherms are measured at liquid nitrogen temperature and pressures varying from vacuum to 1 bar. The presence of pores of different dimensions can be discerned from the shape of the isotherms. For microporous materials, like zeolites, Ar is advantageous over N2 because the presence of a quadrupolar moment in N2 can result in enhanced interaction with the heterogeneous surface of the zeolite framework, which results in difﬁculties in accessing the pore sizes and pore shapes [211]. As the interpretation of the recorded values usually relies on simpliﬁed models, the accuracy of the results is strongly dependent on the validity of the assumptions inherent to the model [210]. For example, in reports on zeolites, the BET (Brunauer–Emmett–Teller) speciﬁc surface area is often given [212]. However, since the micropore ﬁlling does not fulﬁll the conditions for multilayer adsorption, which is the basis of BET theory, the reported BET data do not represent a real physical surface area. In particular, for materials with multiscale pores, the interference of mesopores on the multilayer adsorption makes the interpretation more complicated. Nevertheless, for comparative studies, the BET surface area can be used as a value proportional to the volume adsorbed. For detailed description of gas physisorption analysis of zeolites, we refer to [213]. Here, we only discuss some speciﬁc phenomena related to zeolites featuring additional porosity. Figure 9.5 shows typical nitrogen physisorption results of zeolite NaY and the samples obtained after different postsynthesis treatments [214]. The descriptions of the samples and the corresponding textural parameters derived from the isotherms are listed in Table 9.1. For NaY, a type I isotherm was observed, which is typical for chemisorption on nonporous materials, or physisorption on materials with only microporosity. After postsynthesis modiﬁcation, type IV isotherms appeared, indicating the formation of mesopores. Moreover, different hysteresis loops were observed for samples obtained by different treatments. Hysteresis appearing in the multilayer range of physisorption isotherms is usually associated with capillary condensation in mesopores. Although the factors that affect the adsorption hysteresis loops are still not well understood, the shapes of hysteresis loops have often been identiﬁed with speciﬁc pore geometries. The steep hysteresis loop in sample HMVUSY (high-meso very ultra stable Y) indicates the presence of close-to-cylindrical mesopores with relatively uniform PSD, while the ﬂat and wide hysteresis loop in XVUSY (eXtra very ultrastable Y) is more complex, and can be attributed to the presence of slit pores or ink-bottle-shaped pores with small openings, or due to the effect of a pore network, which is discussed in more detail below. For materials with multiscale pores, especially for microporous materials like zeolites, the comparative analysis methods, like t plot [215], αs plot [213], and θ plot [216] methods, are generally used for the estimation of pore volume and speciﬁc surface area of pores of different dimensions by comparing the isotherms

247

9 Textural Characterization of Mesoporous Zeolites

0.8 HMVUSY + 0.15

0.7

XVUSY

0.6

NaY − 0.08

Vol ads (ml g−1)

USY

0.5 0.4 0.3 0.2 0

0.2

0.6

0.4

0.8

1

p /p 0

(a) 0.9 0.8

HMVUSY XVUSY USY NaY

0.7 dV /dlog d

248

0.6 0.5 0.4 0.3 0.2 0.1 0 1

10

(b)

100

Pore diameter (nm)

Figure 9.5 Nitrogen physisorption isotherms and the BJH pore size distribution curves calculated by the desorption branches for several Y zeolites (from top to bottom: HMVUSY, XVUSY, USY, and NaY).

For clarity, the isotherm of NaY has been shifted 0.08 cm3 g−1 downward and the isotherm of the HMVUSY has been shifted upward by 0.15 cm3 g−1 . (Reprinted with permission from [214].)

of samples under investigation with those of reference samples. For the t-plot analysis, the multilayer ﬁlm thickness of adsorbate (t-values) is determined on a reference nonporous solid with similar surface properties to the samples under investigation. Alternatively, standard reference plots can be used, for instance, the Harkins and Jura equation for silica and alumina substrates: t (nm) = 0.1

13.99 0.034 − log(p/p0 )

1/2 (9.1)

In a t-plot analysis, the volume adsorbed on the sample under investigation at different pressures is plotted against t and the corresponding statistical average

9.3 Characterization of Textural Properties of Mesoporous Zeolites Physical properties of NaY and the mesoporous Ys: USY, XVUSY, and HMVUSYa.

Table 9.1

Sample

NaY USY XVUSY HMVUSY

Si/Al bulk (at/at)

Si/Al XPS (at/at)b

a0 (nm)

% Yc

Vmicro (cm3 g−1 )d

Vmeso (cm3 g−1 )d

Sext (m2 g−1 )e

2.6 2.6 39.3 5.0

2.8 1.1 71.3 1.4

2.469 2.450 2.423 2.427

100 87 72 71

0.34 0.26 0.28 0.15

0.05 0.11 0.25 0.47

8 63 120 146

Results from [214]. (ultrastable Y) was obtained by steaming treatment, XVUSY (eXtra very ultrastable Y) was obtained by steaming twice and acid leaching, HMVUSY (high-meso very ultrastable Y) was obtained by hydrothermal treatment. b Si/Al ratios determined by XPS measurements. c Relative crystallinity. dV micro and Vmeso : micropore volume and mesopore volume determined by t-plot analysis. e Sum of external and mesopore surface area calculated from the t plot. a USY

layer thickness is calculated from the standard isotherm obtained with a nonporous reference solid. For a nonporous sample, a straight line through the origin is expected. Deviation in shape of the t plot from linearity indicates the presence of pores of a certain dimension. By this means, values like the micropore volume, mesopore volume, macropore volume, and mesopore and external surface areas can be calculated. In a similar manner to that of the t plot, the textural properties can also be evaluated from the αs plot, wherein αs is deﬁned by the ratios between the volume adsorbed in the sample under investigation and the volume adsorbed in the reference solid at a certain relative pressure [217]. For the assessment of microporosity, the thickness of the multilayer is irrelevant and it has been suggested to replace t by the ‘‘reduced’’ adsorption [210]. Nevertheless, both methods generally give consistent results for micropore assessment. Using the t-plot method, the effects of the postsynthesis treatment on the porous structure was investigated for the zeolite Y samples shown in Figure 9.1. From Table 9.1, one can see that upon steaming and acid-leaching treatments, the mesopore volume increases signiﬁcantly at the expense of their micropore volume. This has been attributed to the blocking of the micropores with extraframework species generated during the hydrothermal treatment and/or the partial amorphization of the framework. From the isotherms, the PSD curves can be derived. For mesoporous materials, the BJH (Barret–Joyner–Halenda) method [218], which is based on the Kelvin equation for the hemispherical meniscus, is still the most frequently used method. The adsorption process in mesopores often associates with capillary condensation. Preferentially, the desorption branch of the isotherms is used for PSD analysis, as it is closer to the thermodynamic equilibrium [219]. Figure 9.5b shows the PSD curves calculated from the desorption branches of the isotherms using the

249

250

9 Textural Characterization of Mesoporous Zeolites

BJH method. For NaY, as expected, essentially no mesopores were observed. For HMVUSY, a centered peak was observed around 10 nm, which correlated well with the steep hysteresis observed in the isotherms. Peaks around 3–4 nm in the PSD curves appeared for all the treated samples, which corresponded with the sudden closing of the hysteresis loop in the isotherms at a partial pressure of 0.4–0.5. Such a coincidence does not imply similarity among the porous structures of the treated samples. Peaks in this range have often been observed with mesoporous materials, especially zeolites with mesopores, and have been erroneously attributed to the presence of true pores with sizes of 3–4 nm. It can actually be related to the nature of the adsorptive rather than solely the nature of the adsorbent. This phenomenon is often referred to as the cavitation effect, and has been discussed by Neimark and coworkers [220, 221]. As depicted in Figure 9.6, for spherical pores with narrow necks smaller than 4 nm, cavitation of the pores always occurs at a partial pressure corresponding to an apparent pore size of around 4 nm [222]. For nitrogen physisorption, pores with sizes smaller than 4 nm show no hysteresis and exhibit reversible adsorption and desorption isotherms. This is due to the instability of the hemispherical meniscus during desorption in pores with a critical size of ∼4 nm, which is caused by the increased chemical potential of the pore walls provoking spontaneous nucleation of a bubble in the pore liquid. In ink-bottle pores as shown in Figure 9.6, the pore emptying during desorption is delayed because the meniscus in the necks is strongly curved and prevents evaporation. The tensile strength of the condensate in the cavities has a maximum limit, which corresponds to a partial pressure of 2 nm

Vads (cm3 STP g−1)

Vads (cm3 STP g−1)

Vads (cm3 STP g−1)

4 nm

Vads (cm3 STP g−1)

6 nm

10 nm

0.0 0.2 0.4 0.6 0.8 1.0

p /p0

p /p0

1

10

100

Pore diameter (nm)

1

6 10

100

Pore diameter (nm)

0.0 0.2 0.4 0.6 0.8 1.0

dV/dlog d (cm3 g−1)

p /p0

dV/dlog d (cm3 g−1)

dV/dlog d (cm3 g−1)

0.0 0.2 0.4 0.6 0.8 1.0

p /p0 dV/dlog d (cm3 g−1)

0.0 0.2 0.4 0.6 0.8 1.0

1

4 10

100

Pore diameter (nm)

Figure 9.6 N2 adsorption and desorption isotherms at 77 K (middle) and corresponding PSD (bottom) as derived from BJH model of 10-nm cavities with entrance sizes between 2 and 10 nm. (Reprinted with permission from [222].)

1

4 10

100

Pore diameter (nm)

9.3 Characterization of Textural Properties of Mesoporous Zeolites

0.4–0.5 for nitrogen at the boiling temperature [223, 224]. Below this limit, the condensed liquid no longer withstands the tension and evaporates, causing the so-called cavitation. Therefore, for hysteresis with steep desorption in the range of 0.4–0.5, it can be concluded that pores are present with cavity diameter larger than ∼4 nm, and neck size smaller than ∼4 nm. It is therefore necessary to have other evidences to better determine the size and shape of the mesopores. Similar effects can be expected for pore networks wherein larger pores are only accessible through smaller pores [225]. For discrimination between cavitation and capillary desorption, Ar adsorption can provide valuable evidence. In that case, cavitation occurs at a different partial pressure, for example, ∼0.35 at 77 K. In summary, although the peaks in the BJH desorption curves may be considered as ‘‘artifacts,’’ they actually, in many cases, point to the existence of mesoporous cavities in the zeolite crystals. In fact, this peak can be used to estimate the pore volume associated with these cavities in case independent information from ET or thermoporometry is available (see below). Recent results have shown that, using the BJH method, pore size is often underestimated [226, 227] because the BJH method does not take into account the inﬂuence of solid–ﬂuid interactions on capillary condensation [228]. For mesopores in zeolites, which are, in most of the cases, irregular, advanced calculation methods should be used. Recent developments of new methods, like BJH-BdB (Brockhoff–de Boer) [226], BdB-FHH (Frenkel–Halsey–Hill) [229], KJS (Kruk–Jaroniec–Sayari) [230], and NLDFT (Non-Local Density Functional Theory) [231], may provide more accurate estimation of the porous structure in mesoporous zeolites. Apart from the intracrystalline mesopores discussed above, mesopores can also arise from the intercrystalline spaces between nanosized zeolites. Intercrystalline pores often feature very similar hysteresis loops to those of intracrystalline pores. In that case, it is important to employ other characterization techniques for data interpretation. 9.3.2 Thermoporometry

Another method for assessing mesoporosity is thermoporometry [232]. Similar to gas physisorption techniques, it is also based on the inﬂuence of the surface of the solid sample on the phase transitions of the adjoining medium. Thermoporometry relies on the depression of the triple point of a liquid ﬁlling a porous material. The triple-point temperature of these systems depends on the solid–liquid and the liquid–gas interfaces. The depression of the triple-point temperature is generally due to the strong curvature of the solid–liquid interface present within the small pores. Thus, the shift in the equilibrium temperature of the liquid–solid transformation allows the determination of the PSD (for example, 2–60 nm with water as the adsorbate) and pore volume [233–241]. Additionally, the shape of the mesopores can be estimated from this technique by comparing the corresponding crystallization and fusion thermograms. A full thermodynamic description of this phenomenon was given by Brun et al. [232]. In principle, any liquid adsorbate of known thermodynamic properties can be employed for thermoporometry, given

251

252

9 Textural Characterization of Mesoporous Zeolites

that they exhibit suitable interaction with the walls of the porous materials. This hints at one important advantage of thermoporometry over other techniques: molecules relevant to the practical applications of the porous materials can be used as the adsorbates [234]. However, so far water and benzene are still the most investigated adsorbate molecules, as rather limited information is available for other adsorbates. Although thermoporometry has proven to be a powerful technique for the characterization of mesoporous materials, its application to mesoporous zeolites has been rarely reported. The only example, as far as we know, is reported by Janssen et al. on the characterization of mesoporous Y zeolites [242]. A series of zeolite Y samples were analyzed using differential scanning calorimeter (DSC) with water as the probe molecule. In a typical measurement, the sample is ﬁrst strongly cooled, and then the heat ﬂux during slow heating is recorded as a function of the temperature. The depression of the melting point of water, T(K), corresponds to the pore radius, r (nm), according to the modiﬁed Gibbs–Thomson relation derived by Brun et al. [232]: r = A − B/T

(9.2)

For heating scans, A = 0.68 nm and B = 32.33 nm K are used, while for cooling scans, A = 0.57 nm and B = 64.67 nm K are used. In the above equation, a correlation for a nonfreezing layer of 0.8 nm has already been included. Because the delayed nucleation, which is often encountered during cooling scans, can inﬂuence the pore size calculated from the freezing point depression, the melting point depression is more reproducible and most often used [235]. Figure 9.7 shows the DSC curves (heat ﬂow vs temperature) during heating (a) and cooling (b) of water, MCM-41 and different Y zeolites: NaY, USY, XVUSY, and HMVUSY (see Table 9.1 for the description of the sample codes). The textural parameters shown in Table 9.2 are determined by thermoporometry. The amount of water used was about three times the total pore volume determined with nitrogen physisorption. Before the heat ﬂows were measured during heating and cooling, the sample was ﬁrst cooled to −60 ◦ C in order to freeze all the water in the system. In the heating scan (−60 to −2 ◦ C), no peaks were observed for pure (bulk) water. For the other samples, a peak arose up to about −2 ◦ C due to the melting of the water in the intercrystalline spaces. Compared to NaY, additional peak appeared at about −10 ◦ C for USY, XVUSY, and HMVUSY, and −40 ◦ C for MCM-41, respectively, which indicated the presence of mesopores in these samples. For MCM-41, the pore size determined by the peak onset in the endotherm (−50 ◦ C) according to the Gibbs–Thomson relation is 2.7 nm, and that determined by the peak maximum is 3.0 nm, both of which are in good agreement with the value calculated by nitrogen sorption (2.8 nm). For the mesoporous zeolites, the peaks are considerably wider than that of MCM-41, indicating the wide PSD, which is also consistent with the nitrogen physisorption results (Table 9.2). After subtraction of the background, the pore volumes can be calculated by integrating the peaks using the heating rate and the temperature dependence of both the ice density and the enthalpy of fusion [232]. These results are also compiled in Table 9.2. More importantly, in the exotherms of mesoporous Y samples, two peaks can be observed. The peaks

9.3 Characterization of Textural Properties of Mesoporous Zeolites

253

Heating (endotherm)

−0.01

Heat flow (W g−1)

−0.01 −0.02 −0.012

−0.03

−0.014

−0.04

Water NaY − 0.01 XVUSY − 0.004 HMVUSY − 0.004 MCM41 − 0.006 USY

−0.016

−0.05 USY

−0.06

−60

−50

−40

−30

−20

−10

0

−0.018

T (°C)

(a)

Cooling (exotherm) 0.09

0.016

Heat flow (W g−1)

0.08 USY

0.07

0.015

0.06 0.05

0.014

0.04 0.03

0.013

0.02

MCM41 + 0.03 HMVUSY + 0.02 XVUSY + 0.008 NaY + 0.005 Water USY

0.01 0 (b)

−60

−50

−40

−30

−20

−10

0.012 0

T (°C)

Figure 9.7 DSC curves (heat ﬂow vs temperature) during warming (a) and cooling (b) of water, NaY, XVUSY, HMVUSY, and MCM-41 (left axis) and USY (right axis). For clarity, some of the curves have been shifted

upward or downward. The order of the samples in the legend (from top to bottom) is the order of the curves at the left side of the graphs. (Reprinted with permission from [242].)

at higher temperatures can be attributed to the heterogeneous nucleation in the cylindrical pores that extended to the external surface. The peaks around −40 ◦ C, due to the absence of a counterpart in the endotherms, have been attributed to homogeneous nucleation of supercooled water trapped inside cavities, which are only accessible through micropores from the exterior. This water is not in contact with any ice crystals that could serve as nucleation sites. Such reasoning is supported by the observation that the pore volume calculated from the endotherm from −35 to −2 ◦ C is smaller than that from −60 to −2 ◦ C (0.028–0.033 ml g−1 ). Therefore, in this way, the cylindrical pores and the cavities can be distinguished from thermoporometry. Thus, the relative contribution of the different pores to the total pore volumes, that is, cylindrical pores that extend to the external surface and

254

9 Textural Characterization of Mesoporous Zeolites

Table 9.2 Peak onsets, peak maximums, and peak areas of the heating and cooling DSC curves of water-containing USY, XVUSY, HMVUSY, and MCM-41.

Sample

Tonset (◦ C) ±2 ◦ C

Pore D (nm)a

Heating (endotherm) −34 3 −28 4 −32 3 −30 4 −25 4 −25 4 −50 2.7 Cooling (exotherm) USY −9.0g >15.5 −40.1 n.p.h g >34.3 XVUSY >−3.9 −38.2 n.p.h g HMVUSY >−7.0 >19.6 −38.2 n.p.h MCM-41 −38.5 4.5 USYe USYf XVUSYe XVUSYf HMVUSYe HMVUSYf MCM-41

Tmax (◦ C) Pore D (nm)a

Peak area Pore volume (cm3 g−1b ) (J g−1 )

−10.5 −10.4 −4.1 −4.1 −5.3 −5.3 −40.3

7.5 7.6 17.1 17.1 13.6 13.6 3.0

2.3 2.0 14.7 12.0 18.5 18.0 16.3

−16.5 −42.7 −5.9 −40.7 −11.8 −40.2 −42.5

9.0 n.p.h 23.1 n.p.h 12.1 n.p.h 4.2

n.d.i 0.3 n.d.i 2.1 n.d.i 0.6 15.9

0.033 0.028 0.18 0.14 0.24 0.23 0.62

Mesopore % Cavitiesd diameter (nm)c N2 physisorption 4−20 20 4−40

29

4−25

7

2.8

Results from [242]. Calculated with Eq. (9.2). b Calculated by correcting the peak area for the density of ice and the enthalpy of fusion. c Mesopore diameter determined by nitrogen physisorption using the desorption branches of the isotherms. d Proportion of the cavities volumes in the mesopore volumes. e Heating from −60 to −2 ◦ C. f Heating from −35 to −2 ◦ C after cooling from −2 to −35 ◦ C. g Not exactly determined due to overlap with the peak at −2 ◦ C. h n.p. = not possible to determine with Eq. (9.2) due to homogeneous nucleation. i n.d.= not determined due to overlap with the peak at −2 ◦ C. a

the cavities, which are only accessible through micropores, can be calculated. For USY, the relative volume fraction of cavities according to thermoporometry is 15% (1 – 0.028/0.033, Table 9.2), which is in good agreement with that determined by nitrogen physisorption (20%, Table 9.2). From this example, we can see that thermoporometry is a powerful technique for the characterization of mesoporous zeolites. Compared to gas physisorption, the sample preparation and measuring procedure in thermoporometry are much simpler and faster, and a fully automated apparatus including a sample changer for rapid product control is possible. However, thermoporometry is not yet a standardized method as is gas physisorption. For fragile samples, the possible deformation and even the collapse of the pore structure due to the expansion of liquid crystallization need to be taken into account. Some questions related to the data interpretation are still not well resolved, for example, the uncertainty about the thickness of the nonfreezable liquid layer [241], the enthalpy of fusion

9.3 Characterization of Textural Properties of Mesoporous Zeolites

at low temperature [242], and the different contributions of pore shape factor and the pore size to the origin and width of the hysteresis between endotherms and exotherms [233]. Moreover, extensive data exist only for water and benzene, which limits the application of this method [237]. Nevertheless, as shown above, with thermoporometry it is possible to discriminate the cavities inside the mesoporous zeolites, wherein homogeneous nucleation takes place, and in the mesopores that extend to the external surface in which heterogeneous nucleation takes place. Moreover, the relative volume portion of different pores can be easily determined. The full potential of thermoporometry for characterizations of mesoporous zeolites is still to be explored, especially by using adsorbates relevant to the practical applications of the zeolites [234]. 9.3.3 Mercury Porosimetry

Mercury porosimetry (MP) is a widely accepted method for the characterization of porous materials with macro- and mesopores [77]. It relies on the relationship between the pore radius r and the hydrostatic pressure P at which mercury can enter the pores. For a nonwetting liquid like mercury, a positive excess hydrostatic pressure P is needed, and P is varied inversely with r. With an increase in the pressure applied, mercury enters pores in decreasing order of size. According to this relation, the pore size can be calculated using the Washburn equation [243]: r = −2γ cosθ/P

(9.3)

wherein r is the pore radius of an equivalent cylindrical pore, γ is the surface tension of mercury, θ is the contact angle between the mercury and the material, and P is the applied pressure. Traditionally, a value of 484 mN m−1 is used for γ and 141◦ for θ. Experimentally, the pressure can be varied between 0.1 and 2000 bars, corresponding to the pore radii in the range of 75 µm to 3.5 nm. Moreover, in combination with nitrogen physisorption, it is possible to differentiate between the mesopores that extend to the external surface and the cavities that are only accessible through micropores. While nitrogen physisorption probes both the micro- and mesopores, MP only accesses pores larger than 3.5 nm. Although MP is a well-developed method for meso- and macropore analysis, the reported examples of application of MP to mesoporous zeolites are rather limited. Lohse et al. [244] and Janssen et al. [214] used MP to characterize mesoporous Y zeolites obtained by dealumination. It has been shown that MP gave consistent results compared to those of nitrogen physisorption for pores larger than 4 nm in terms of pore volume as well as PSD. More recently, MP has also been applied to mesoporous ZSM-5 crystals obtained by desilication [245]. Figure 9.8 shows the PSD curves of the nontreated sample (Z-200-nt) and the base-treated sample (Z-200-at) obtained by MP and nitrogen physisorption. It is noted that the peaks around 2 nm in the PSD curve of the nitrogen physisorption are not related to a speciﬁc pores, but rather due to the so-called ﬂuid-to-crystalline-like phase transition of the adsorbed phase in the MFI structure [246], which has

255

9 Textural Characterization of Mesoporous Zeolites 8

dV/dlog d (cm3 g−1)

256

6

4

2

0

1

10

100

1000

Pore diameter (nm)

Figure 9.8 Pore size distribution of the parent zeolite Z-200-nt (open symbols) and the base-treated zeolite Z-200-at (solid symbols) derived from (, ) N2 adsorption, and (, ) MP measurements. (Reprinted with permission from [245].)

been observed frequently. For pores larger than 100 nm, both nontreated and treated samples exhibited similar pore sizes, indicating that the morphology of the crystals did not change after the base treatment. In the mesopore region, both techniques gave comparable results. The pore volume in the range of 4–100 nm is 0.41 cm3 g−1 by nitrogen physisorption and 0.31 cm3 g−1 by MP. The lower pore volume from MP has been attributed to the elastic compression of the sample during the high-pressure measurements. MP gave a PSD centered at around 25 nm, while nitrogen physisorption gave a broader BJH PSD centered at 45 nm derived from the adsorption branch of the isotherm. The smaller size with a yet narrower PSD from MP is probably related to the pore-network effect. Because the pores are invaded in decreasing order of size in MP measurement, it implies that the sequential ﬁlling of the pores is dictated primarily by their mode of interconnection. Nevertheless, MP results provided evidence that the mesoporosity created by desilication is largely accessible from the external surface of the zeolite crystals, which is important for improved transport properties. In this context, MP provides a good estimation of the amount of the mesopores that extend to the external surface, which are expected to alleviate the diffusion limitation to a larger extent compared to the cavities inside the zeolite crystals. The above example also hints that caution should be taken when elastic or fragile samples are under investigation. 9.3.4 Electron Microscopy 9.3.4.1 SEM and TEM Electron microscopy (EM) is one of the most straightforward techniques for porous structure analysis, and has been used extensively, in many cases coupled with

9.3 Characterization of Textural Properties of Mesoporous Zeolites

nitrogen physisorption, to give direct information on the textural properties [247]. Although a number of working modes are available, scanning electron microscopy (SEM) and TEM are the most frequently used techniques for porous materials. SEM works by rastering a beam of electrons over the sample surface and determining the intensity of the backscattered electrons or secondary electrons at different positions. Therefore, surface properties like morphology and composition can be obtained. With respect to zeolites with additional meso- and macropores, enhanced surface roughness is expected. In contrast to SEM, TEM relies on transmitted and diffracted electrons. As the electron wave passes through the sample, the interaction between electrons and the sample results in a change in the amplitude and phase of the electron wave. The overall change depends on the thickness and composition of the samples. Regions with higher porosity appear as a lighter area in the bright ﬁeld TEM image due to the lower mass density. Therefore, a direct imaging of the pores can be achieved. Ogura et al. reported the formation of mesoporous ZSM-5 crystals by base leaching with 0.2 M NaOH aqueous solution [62]. After base leaching, nitrogen physisorption measurements showed slightly lowered micropore volume (0.17 increased to 0.13 cm3 g−1 ), yet considerably enhanced mesopore volume (0.07 increased to 0.28 cm3 g−1 ). SEM images showed some grooves and voids on the surface of the treated sample (Figure 9.9), indicative of the etching of the materials by base solution. Such morphology with enhanced surface roughness has been frequently observed for mesoporous zeolites obtained by postsynthesis treatments [97], as well as by templating methods [143, 151, 159]. The nonuniform distribution of aluminum in ZSM-5 crystals has been a matter of debate. By treating the ZSM-5 crystals with Si/Al ratio of 35–115 with base solution (0.5 M Na2 CO3 , 16–20 hours at reﬂux), Dessau et al. obtained a series of hollow ZSM-5 crystals [60]. Figure 9.10 shows the SEM image of a typical treated example. After the alkaline treatment, the Si/Al ratio decreased from 62 in the original crystals to 8 in the treated ones. This result indicated the preferential dissolution of silicon species over aluminum species under basic conditions. The tetrahedral aluminum centers with negative charges are supposed to be inert to the attack of hydroxide. The formation of hollow crystals provided unambiguous evidence for aluminum zoning in ZSM-5 crystals, with higher aluminum concentration close to the exterior of the crystals. Similar phenomena were observed later by Groen et al. treating large ZSM-5 crystals (∼20 µm, Si/Al ratio of 41) with 0.2 M NaOH solution [110]. With SEM–EDX (energy dispersive X-ray spectroscopy) analysis, spatially resolved elemental distribution was obtained in the crystals. As shown in Figure 9.11, a gradient of aluminum concentration can be observed in the parent crystals, while silicon is rather uniformly distributed. Point analysis over multiple crystals revealed that the aluminum concentration is 30 times higher at the edge than that in the center of the crystals. As a result of this, hollow crystals with inner macropores and relatively intact outer surface were obtained after the base treatment. Such a large void only moderately enhances the mesopore surface area, which was conﬁrmed by nitrogen sorption measurements (5 vs 30 m2 g−1 , parent vs treated sample).

257

258

9 Textural Characterization of Mesoporous Zeolites

(a)

(b)

500 nm

Figure 9.9 SEM images of (a) the parent and (b) the alkaline-treated ZSM-5 crystals. (Reprinted with permission from [62].)

0030

15KV

×15,000

1 µm

WD 8

Figure 9.10 SEM image of ZSM-5 crystal (Si/Al = 62) treated with Na2 CO3 solution. (Reprinted with permission from [60].)

9.3 Characterization of Textural Properties of Mesoporous Zeolites

Al

O

259

Si

20 µm Nontreated

Alkaline treated

(a)

(b)

Figure 9.11 (a) SEM micrograph of synthesized large ZSM-5 crystals and (b) SEM-EDX micrographs of the nontreated and alkaline-treated (0.2 M NaOH) ZSM-5 crystals (17 µm). Blue, red, and yellow colors represent aluminum, oxygen, and silicon concentrations, respectively. (Reprinted with permission from [11].)

Carbon templating has proven to be an effective method for generating well-deﬁned mesopores in zeolites. Using carbon black as the template, Christensen et al. obtained mesoporous ZSM-5 crystals [41, 143]. Figure 9.12 shows the SEM images of the conventional zeolite (synthesized without carbon) and the mesoporous zeolite crystals. Although the typical cofﬁn shape of MFI crystals was observed for both, the surface morphologies were very different with the mesoporous sample, showing a much rougher surface. The TEM image provided clear evidence of the highly mesoporous structure of the samples (Figure 9.12). (a)

(c)

(b)

2 µm

100 nm

Figure 9.12 (a, b) SEM images of the conventional (a) and mesoporous zeolite catalysts (b), (c) TEM of an isolated single ZSM-5 crystal and the diffraction pattern (inset) obtained from the same crystal. (Reprinted with permission from [41] and [143].)

260

9 Textural Characterization of Mesoporous Zeolites

From SEM and TEM analyses, the mesoporous crystals appeared to consist of agglomerated small crystals. However, electron diffraction measurement over the area covering the whole crystal indicated the single-crystal nature of the sample (inset of Figure 9.12c), which further supports the hypothesis that carbon particles were acting as the template and were entrapped during the synthesis. The advantages of the carbon-templating method are further illustrated by its wide applicability to other zeolites, as well as the diversity of available carbon sources. According to the synthesis mechanism using a hard template, under well-deﬁned synthesis conditions, the size, shape, and tortuosity of the mesopores show ﬁdelity with the morphology of the template materials. Consequently, the shape and size of the mesopores can be controlled by using proper template materials with a different morphology. This has been indeed observed for mesoporous silicalite-1 crystals with intracrystalline cylindrical pores synthesized using multiwall carbon nanotubes [42]. As shown in Figure 9.13, the zeolite crystals show single-crystal-like morphology with well-deﬁned facets, however, with extensive cylindrical pores throughout the crystals. Moreover, the cylinder pores occurred as imprinted patterns of the carbon nanotube, further conﬁrming the templating role of the carbon nanotubes. For in-depth mesostructure analysis as well as the study of the formation mechanism of the mesopores, the detailed structure analysis with high resolution, probably down to atomic scale, is appreciated. In this regard, high-resolution transmission electron microscopy (HRTEM) is one of the suitable techniques. However, many zeolite specimens with suitable size and thickness for (HR)TEM analysis, especially those of mesoporous zeolites, are sensitive to electron beam irradiation. To overcome this limitation, Sasaki et al. equipped the electron microscope with a

Zeolite single crystal (0.1 – 10 µm) Carbon nanotube (1 – 20 nm)

20 nm

(a)

(b)

50 nm

(c)

Figure 9.13 (a) Schematic illustration of the synthesis principle for crystallization of mesoporous zeolite single crystals. (b) and (c) TEM images of the multiwall carbon nanotube (b) and the mesoporous silicalite-1 crystal obtained using carbon nanotubes as the template (c). (Reprinted with permission from [42].)

9.3 Characterization of Textural Properties of Mesoporous Zeolites

high-sensitivity SSC (slow scan CCD) camera, which allowed the HRTEM measurement to be performed with only 1/50 of the electron dose of the usual conditions [88]. Figure 9.14 displays the HRTEM image for the steam-dealuminated Y zeolite observed along the [110] direction, showing the mesopore channels as white contrast. By detailed analysis of the micrographs, they observed that microtwins existed along the mesopores. It was therefore suggested that the formation of mesopores start from the preexisting twin planes, which would present less stable regions in the crystals. In this context, the density of the mesopore can be controlled by adjusting the twin plane density in the parent crystal. With combined SEM and HRTEM measurements on EMT/FAU intergrowth zeolites, Terasaki et al. also observed that dealumination occurred preferentially in regions with a higher level of stacking disorder [248]. 3D TEM TEM analysis has been one of the prevailing techniques and has provided a wealth of straightforward information on the structures of the mesopores in zeolites. The size and shape of the mesopores on individual particles can be

30 nm

20 nm

Figure 9.14 HRTEM image for the dealuminated zeolite Y observed along the [110] zone axis. The white contrast with columnar shape shows the mesopore. (Reprinted with permission from [88].)

20 nm

261

262

9 Textural Characterization of Mesoporous Zeolites

directly viewed with TEM, which complements the global information of ensembles obtained from other macroscopic techniques, like nitrogen physisorption. However, as conventional TEM image is a 2D projection of a 3D object, no unambiguous information on the shape, connectivity, location, and 3D orientation of mesopores can be derived. For example, in steamed Y zeolites, spherical and cylindrical pores have been frequently observed, while in 2D projection cylindrical pores may appear as spheres. Such discrepancy cannot be resolved by conventional TEM [249, 250]. Ultramicrotomy has been used for dealuminated Y zeolites to view the internal porous structure with ultrathin specimen (∼20–50 nm) [85, 87]. However, fracture of the samples occurred during the sectioning process, which complicated the results interpretation. All these have necessitated a real 3D imaging of the intact sample. Toward this, the stereo-TEM has been applied for mesoporous zeolites [48]. By tilting the specimen of a steamed Y crystal from 0 to 50◦ with 10◦ intervals, Sasaki et al. observed that round pores observed in one orientation showed up as cylindrical pores when viewed after tilting over 50◦ [88]. Stereo-TEM has also been employed for imaging of mesoporous ZSM-5 crystals obtained with carbon nanotubes [147]. A step further toward 3D imaging is the development of ET [251]. With ET, a series of 2D TEM images is recorded by tilting the specimen in a large angular range. These images are subsequently aligned with respect to a common origin and tilt axis, and used for the computation of the 3D image of the object under investigation. With state-of-the-art equipment, a typical series of 141 images is collected in the range of −70 to 70◦ with angular increment per 1◦ . For a proper alignment of these images, markers like metal nanoparticles (for example, 5-nm gold particles) are often used. Alignment is achieved by least-square ﬁtting of the positions of markers. According to the Projection–Slice theorem, the Fourier transform of a 2D projection is equal to a central slice through the 3D Fourier transform of the object [252]. By combining the series of the projections at different tilts and through an inverse Fourier transform process, the 3D morphology of the object can be obtained (Figure 9.15). Such a process, also known as reconstruction, provides the 2D images of the virtual slices of the object at different heights. For a detailed description of the basic principles and work ﬂow of ET, we refer to the book by Deans et al. [252]. The applications of ET to catalysis materials have recently been exploited and highlighted in a recent review [253]. With respect to zeolites, ET has been applied to image the mesoporous structure of zeolites obtained by postsynthesis treatment [110, 214, 254–256] as well as by the carbon-templating method [40]. The ﬁrst application of ET to zeolites, and also the ﬁrst to heterogeneous catalysts, was reported by Koster et al. [254, 256]. Figure 9.16 shows the conventional TEM image and the digital thin slice (0.6 nm) obtained by ET of the acid-leached H-mordenite crystal, which is among one of the most important zeolites with industrial applications. Black dots correspond to the gold particles as markers for alignment. The ET analysis showed the existence of mesopores inside the crystal with unprecedented clarity, while in conventional TEM image such information is obscured by the variant sample thickness and/or by superimposition or different pores in the conventional TEM images.

9.3 Characterization of Textural Properties of Mesoporous Zeolites

3D Object

3D FT of object

3D IFT

Z

Y

X Multiple tilts TEM

2D FT

2D Projection Figure 9.15 ET reconstruction scheme using the Projection–Slice theorem. FT denotes the Fourier transform and IFT denotes the inverse Fourier transform. By combining projections recorded at multiple tilts, the 3D FT

3D FT slice of the object is probed. An inverse Fourier transform then gives the reconstruction of the 3D object. (Reprinted with permission from [253].)

Another series of industrially important zeolites, that is, zeolite Y and the corresponding dealuminated counterparts have also been investigated with ET [255]. Figure 9.17 shows the 2D TEM and ET images of the NaY as well as the samples obtained after different treatments; that is, steaming (USY), seaming twice followed by acid leaching (XVUSY). As expected, no mesopores can be observed in the NaY sample. Compared to conventional TEM images, mesopores can be better resolved in the ET images. The diameters of the mesopores evaluated with ET are 3–20 nm for USY and 4–34 nm for XVUSY, which correlate well with the values estimated by nitrogen physisorption: 4–20 nm for USY and 4–40 nm for XVUSY, respectively. Moreover, ET images of a series of slices showed that, in USY and XVUSY, a signiﬁcant fraction of the mesopores are cavities, although there were some cylindrical pores extending to the external surface. The presence of both types of mesopores was also indicated by the nitrogen physisorption measurements. More importantly, ET images of the steamed sample (USY) showed some dark cavities and dark bands surrounding the crystals, which could not be clearly distinguished in the 2D TEM image. Combining the nitrogen physisorption and XPS analysis (Si/Al atomic ratio), such areas were attributed to the amorphous materials enriched with aluminum (Si/Al = 1.1 at/at from XPS). On the basis of these observations, a mechanism for the mesopore formation during steaming and acid leaching was proposed as depicted in Figure 9.17g, which supported earlier studies [79], but added key aspects of mesopore formation. Thus, the mesoporous system

263

264

9 Textural Characterization of Mesoporous Zeolites

V V V

(a)

V

V V

(b)

Figure 9.16 (a) Conventional TEM image indicating the mesopores in the crystallite (white spots, arrows) and several gold beads (black dots, 5 nm in diameter) on the grid for the alignment. (b) Digital slice (0.6 nm

thick) through the 3D reconstruction of the crystallite showing the mesopores inside the crystallite (arrows). Scale bar is 100 nm. (Reprinted with permission from [254].)

originated from cavities with amorphous materials deposited therein. During the second steaming and acid leaching, small cavities coalescenced to form larger cavities and cylindrical pores. The extraframework aluminum species, which were clearly visible in USY as dark areas, were removed in the subsequent treatment of XVUSY. By virtue of the detailed information provided by ET, a quantitative structural analysis on individual crystals is possible. As an example, PSD obtained by quantiﬁcation results from ET is shown in Figure 9.18, which agreed well with that from nitrogen physisorption for bulk analysis [257]. Thus, results from both methods complement each other, whereas unlike gas physisorption ET does not assume a speciﬁc model for the pore shape and provides direct information on the shape, size, and connectivity of the mesopores. ET has also been employed for the study of mesoporous zeolites formed by other methods. For small ZSM-5 crystals (400–700 nm in size), upon desilication by alkaline treatment, ET analysis showed rather uniform and interconnected

9.3 Characterization of Textural Properties of Mesoporous Zeolites

200 nm

(a)

200 nm

(b)

200 nm

200 nm

(d)

(c) 200 nm

200 nm

(f)

(e)

Steam

NaY

Steam 2

USY

Acid

(g)

XYUSY

Figure 9.17 2D-TEM images of NaY (a), USY (c), and XVUSY (e), 3D ET slices of NaY (b, 1.7 nm), USY (d, 1.7 nm), and XVUSY (f, 1.25 nm), and the model for the generation of mesopores in zeolite Y (g). (Reprinted with permission from [255].)

265

266

9 Textural Characterization of Mesoporous Zeolites

Pore volume (%)

50 USY XVYSY

40 30 20 10 0

0

5

10 15 Pore size (nm)

(a)

20

200 nm (b)

Figure 9.18 (a) Mesopore size distributions of USY and XVUSY obtained by image analysis of electron tomograms. (b) 3D representation of XVUSY showing the mesopores through the transparent surface of the particle. (Adapted from [257].)

mesopores developed in the interior of the crystals, while the outer surface remained relatively unaffected [110]. These results provide additional evidence of the aluminum zoning in the small ZSM-5 crystals, which play a directing role for the mesopore formation during the base treatment. As a last example, ET has also been applied to mesoporous zeolites obtained by the carbon-templating method [40]. For silicalite-1 samples synthesized with carbon nanoﬁbers and carbon black aggregates, cylindrical mesopores were formed that start at the external surface of the zeolite crystals. However, the tortuosity of the mesopores templated by the carbon black aggregates was much higher than the cylindrical pores obtained by the carbon nanoﬁbers. For larger zeolite crystals, a higher amount of carbon aggregates can be enclosed in the zeolites, leading to ink-bottle-type mesopores. From the above examples, one can see that, compared to conventional TEM, ET can yield unprecedented and unambiguous information, both qualitatively and quantitatively, on the pore shape, pore size, and connectivity inside individual crystals, which complement the results from other bulk analysis and provide new insights into the formation mechanism of mesopores and opportunities for optimizing the textural properties of zeolites. 9.3.5 NMR Techniques 9.3.5.1 129 Xe NMR Spectroscopy Nuclear magnetic resonance (NMR) spectroscopy is a powerful method enriched with a collection of techniques useful for structure characterization of zeolite materials. A wealth of information on the composition, structural defects, and

9.3 Characterization of Textural Properties of Mesoporous Zeolites

pore surface properties of conventional zeolites are available with 27 Al, 1 H, and 29 Si NMR spectroscopies. With respect to the porous structure of zeolites, 129 Xe NMR has proven to be an important and valuable technique [258–260]. 129 Xe is an inert, nonpolar, spherical atom with a large electron cloud, which is sensitive to its environment. Depending on the various interactions with its surrounding, a wide range of chemical shift has been observed for 129 Xe, based on which the so-called xenon NMR spectrometry has been developed. For xenon adsorbed in a porous solid, the observed 129 Xe chemical shift is the weighted average of different types of interactions on the NMR timescale [260]. For zeolites, the observed chemical shift of xenon is the sum of several terms corresponding to the various perturbations around it, including Xe–Xe collisions, electronic ﬁeld exerted by cations, interactions with the pore surface, and the dimension and geometry of the cages and channels. Generally, the smaller the pore size, the higher is the chemical shift observed with respect to that of the gas-phase xenon due to the strongly restricted diffusion of xenon. While the chemical shift of xenon reﬂects the pore dimensions, the xenon isotherm can provide information on the pore volumes. Although considerable knowledge of 129 Xe NMR has been accumulated on zeolite structure, compositions, cations siting, and so on, it has been rarely applied to mesoporous zeolites. For dealuminated Y [261], mordenite [261, 262], and CSZ-[263] zeolites, a decreased 129 Xe shift has been observed, which has been attributed to the increased mean free path caused by the formation of meso- and macropores. More recently, laser-hyperpolarized (HP) 129 Xe NMR 2D exchange spectroscopy (EXSY) has been applied by Liu et al. to investigate the mesoporous ZSM-5 zeolites synthesized by the starch-templating method [201]. The optical pumping by laser for the production of HP xenon facilitates its working at very low concentration of xenon (∼1%) under continuous ﬂow [264]. Therefore, the observed 129 Xe chemical shift would mainly reﬂect the interaction between xenon atoms and the solid surface. The HP 129 Xe 2D-EXSY NMR technique is a powerful tool to study the dynamic process of the adsorbed xenon in porous solid, and has shown potential to reveal textural connectivity in such hierarchical zeolites. Figure 9.19 shows the continuous-ﬂow HP 129 Xe NMR spectra of mesoporous ZSM-5 with a Si/Al atomic ratio of 50 at variable temperatures. The peaks at 0 ppm are due to xenon from the gas phase. From 293 to 153 K, there is one downﬁeld signal a, which was attributed to the xenon adsorbed in the micropores. The chemical shift of this signal increased with the decrease in the temperature due to the enhanced Xe-surface and Xe–Xe interactions at lower temperatures. At 153 K, a new upﬁeld signal b emerged, which indicates that there is another type of pores with a wide size distribution in the mesoporous ZSM-5, and the fast exchange of xenon among these pores exists above 153 K. Figure 9.19b shows the 2D EXSY spectrum of the same sample at 143 K with mixing time τmix = 1 ms. The 2D EXSY spectrum is generally obtained by monitoring frequencies before and after a so-called mixing time, τmix , during which spin exchange and/or molecular reorientation motions can occur. Changes in the NMR frequencies can be observed as the off-diagonal intensities in the 2D spectrum. The cross-peaks indicate an exchange of xenon atoms between the corresponding environments on the

267

268

9 Textural Characterization of Mesoporous Zeolites a b 138 K 143 K 153 K 163 K 173 K

100

213 K 233 K

ppm

193 K

200

253 K 273 K 293 K 300 (a)

200

100 ppm

150

t mix = 1 ms

250

0

250 (b)

200

150 ppm

100

Figure 9.19 (a) Laser-hyperpolarized 129 Xe NMR spectra of the Xe adsorbed in mesoporous ZSM-5 with an Si/Al ratio of 50 at a variant temperature from 293 to 138 K. (b) Laser-hyperpolarized 129 Xe 2D-EXSY NMR spectra of the same sample at 143 K with mixing times τmix = 1 ms. (Reprinted with permission [201].)

diagonal within the period of τmix . The cross-peaks observed in Figure 9.19b clearly demonstrated the exchange of xenon between the micropores and mesopores. No such cross-peaks can be observed for the mixing time shorter than 0.2 ms. The cross-peaks became stronger with longer τmix of 5 ms. Therefore, the exchange of xenon between these two types of porous domains takes place at a timescale of less than 1 ms. For a better understanding, the authors also prepared a sample by mixing conventional ZSM-5 with a commercial SiO2 with similar mesopore size to that of the mesoporous ZSM-5 sample. The mixture sample also exhibited comparable textural properties with those of the mesoporous ZSM-5 sample, for example, micropore volumes of 0.10 cm3 g−1 for both, mesopore volumes of 0.30 versus 0.34 cm3 g−1 , BET surface areas of 373 versus 350 m2 g−1 for the mixture and the mesoporous zeolite respectively. In contrast, under the same experimental conditions, 2D EXSY spectra showed no exchange between the conventional ZSM and the silica at τmix of 1 ms. The exchange signals were only observable with τmix > 5 ms. This indicates that the exchange of xenon in the mixture sample, which is actually interparticle diffusional exchange, occurred slower than that in the mesoporous ZSM-5 sample. Therefore, the microporous and mesoporous domains in the mesoporous zeolite were much closer and had a better connectivity than that in the physical mixture. This example showed that unprecedented information on the connectivity and communications between the micropores and mesopores can be obtained by the 2D EXSY 129 Xe NMR technique. The full potential of 129 Xe NMR for the characterization of mesoporous zeolites is still to be explored, especially in a quantitative way.

9.3 Characterization of Textural Properties of Mesoporous Zeolites

9.3.5.2 PFG NMR As the most important impact of generating textural mesopores in zeolites is to alleviate the diffusion limitation, a direct diffusion measurement of guest molecules, especially those molecules relevant to the practical applications of zeolites, is valuable. For various mesoporous zeolites, macroscopic diffusion measurements have indeed shown increased apparent diffusivities [71, 72, 74–76], which have been mostly attributed to the shortened diffusion path lengths. However, to gain a more insightful knowledge of the effects of the textual mesopores on the diffusion properties, a microscopic diffusion measurement is needed. In this context, pulsed ﬁeld gradient (PFG) NMR is a versatile technique and has been widely used for studying intracrystalline diffusion in porous materials, especially zeolite materials. The PFG NMR is based on the dependence of the Larmor frequency of spins on the amplitude of the applied ﬁeld [6]. Superimposing a large constant magnetic ﬁeld by a magnetic ﬁeld gradient pulse, one can label the positions of the spins by the Larmor frequency and hence the accumulated phases due to the rotation with the speciﬁc Larmor frequency in the local ﬁeld. If the molecules where the spins are located do not change their positions during the time of the diffusion measurements, the signal remains at the maximum level. However, if the molecular displacement occurs during the observation time interval, the measured signal decays to a certain extent. The attenuation of the NMR signal ( ) for the normal diffusion obtained with the 13-interval PFG NMR can be represented by the following equation [265]:

= exp(−4γ 2 δ 2 g 2 Dt)

(9.4)

wherein D is the diffusivity, t is the effective diffusion time, δ is the duration of the applied gradient pulses with the amplitude g, and γ is the gyromagnetic ratio. By measuring the attenuation of the signal as a function of the amplitude of the applied ﬁeld gradient (g) with all the rest of the parameters constant, one can determine the diffusivity D. Using PFG NMR, Kortunov et al. investigated the USY zeolite with n-octane and 1,3,5-triisopropylbenzene as the probe molecules [266]. USY was obtained by steaming treatment of ammonium-form Y zeolite (crystal size ∼3 µm). Nitrogen physisorption showed the slightly decreased micropore volume (from 0.28 to 0.24 cm3 g−1 ), yet increased mesopore volume (from 0.07 to 0.17 cm3 g−1 ) after the steaming treatment. Figure 9.20 shows the measured dependencies of the effective n-octane diffusivities on the root mean square displacement r 2 1/2 ( r 2 (t) = 6Dt) in the ammonium-form Y as well as the USY. However, no clear difference was observed for r 2 (t) 1/2 smaller than ∼1.5 µm (half of the crystal size), which revealed that the mesopores formed by steaming are essentially of no inﬂuence for the intracrystalline diffusion of n-octane. To further elucidate on this point, 1,3,5-triisopropylbenzene, whose size is larger than the micropore openings and hence not expected to enter the micropores, was used as the probe molecules. Again, similar diffusion behavior was observed between the parent and the steamed samples, with the intercrystalline diffusion dominating

269

9 Textural Characterization of Mesoporous Zeolites

Ammonium-ion exchanged Y USY 213 K 193 K 173 K

1 × 10−10

Deff (m2s−1)

270

10−11

10−12 0

1

2 〈

r 2〉1/2

Figure 9.20 Dependencies of the effective diffusivities on the mean square displacements measured for n-octane by PFG NMR (points) and those obtained by the dynamic Monte Carlo simulations (lines) for a cubic

3

(µm) lattice with a size of 2.3 × 2.3 × 2.3 µm3 . The boundaries of the simulation lattice were assumed to be impenetrable for diffusing molecules. (Reprinted with permission from [266].)

in both cases. As for USY, it has been well established with 3D TEM that a signiﬁcant part of the mesopores is actually cavities, which are only accessible via the micropores surrounding them. In line with this, the PFG NMR study showed that the intracrystalline diffusion is essentially unaffected by the mesopores in USY. The presence of the cavities induces only a moderate increase in the intracrystalline diffusivities, as the overall process is controlled by the slowest step, that is, the diffusion through the micropores. These results emphasize the importance of the shape and connectivity of the mesopores, in addition to the pore size and volume, when an alleviated diffusion limitation of zeolites is aimed at. NMR techniques combined with a wide range of applicable probe molecules can provide important information on the mesopores in zeolites, especially the shape and connectivity of the mesopores. It is important to note that, other potential NMR techniques, which have been well investigated for mesoporous materials, have not been exploited for mesoporous zeolites. Among them, NMR cryoporosimetry is an especially informative one, which relies on the dependencies of the phase-transition behavior of conﬁned guest species on the pore size of the solid matrix. For example, 1 H NMR spectra of water conﬁned in mesopores have been used to characterize ordered mesoporous silica MCM-41 [267]. A simple relation has been found between the freezing point depression and the pore size.

9.3 Characterization of Textural Properties of Mesoporous Zeolites

9.3.6 In situ Optical and Fluorescence Microscopy

In situ ﬂuorescence microscopy has been widely applied to cellular biology, and has recently found potential as an advanced technique for studying heterogeneous catalysts under working conditions due to its high spatiotemporal resolution and sensitivity [268, 269]. Nonuniform catalytic behaviors over large, single ZSM-5 and SAPO-34 crystals have been unraveled using a combined technique of in situ optical and ﬂuorescence microscopy [270–272]. The structures of the sub-building units of large CrAPO-5, SAPO-34, SAPO-5, and ZSM-5 crystals have also been suggested by imaging the zeolite crystals during the template decomposition process [273]. More recently, Kox et al. showed that the spatially resolved catalytic activity of mesoporous ZSM-5 crystals in the oligomerization of various styrene compounds can be accessed using such a combined measurement, which has provided spatially resolved direct evidence of the effect of mesopores during the catalytic events [133]. Figure 9.21 shows the SEM images of the parent ZSM-5 and the mesoporous ZSM-5 obtained by desilication. After desilication, the upper ‘‘lid’’ disappeared in

5 µm

5 µm

(a)

0.8

(b)

0.4

A

Mesoporous Parent 0

(c)

I

II

400 (d)

500 l (nm)

600

Figure 9.21 (a,b) SEM images of the parent (a) and the mesoporous (b) H-ZSM-5 crystals. (c) Microphotographs of the parent (I) and mesoporous (II) crystals after oligomerization with 4-methoxystyrene. (d) The corresponding optical absorption spectra. (Reprinted with permission from [133].)

700

271

272

9 Textural Characterization of Mesoporous Zeolites

the mesoporous crystal. After the addition of 4-methoxystyrene and subsequent heat treatment at 373 K, as revealed by the optical micrographs, the parent sample was mostly colored in the upper lid, whereas the mesoporous sample exhibited more uniform coloration (Figure 9.21). It is therefore possible that the ‘‘lid’’ unit has a different pore orientation with the rest parts of the crystals, and restricts the diffusion of guest molecules into the interior parts of the crystals. After desilication, a higher accessibility of the inner parts of the crystals can be expected. Also shown in Figure 9.21 are the optical absorption spectra. The bands at 595 and 650 nm can be attributed to the dimeric and trimeric (or higher) oligomers, respectively. Upon desilication, the formation of trimeric or higher carbocations is restricted. This can be due to the shortened diffusion length in the mesoporous samples. Consequently, the chance of successive reactions over other Brønsted acid sites is decreased. The same type of reaction was conducted under confocal ﬂuorescence microscopy for both crystals. In line with these observations, the ﬂuorescence snapshots also showed more uniform coloration over the mesoporous crystals (Figure 9.22). Furthermore, the confocal ﬂuorescence signal was more intense in the mesoporous sample, which can be rationalized by using the optical absorption spectra, that is, the optical absorption in the excitation wavelength (561 nm) directly corresponds to the ﬂuorescence intensity of the crystals. Moreover, based on these results, a novel building scheme was proposed for ZSM-5 crystals (Figure 9.22C). This example shows the impact of mesopores on the zeolite-catalyzed reactions directly. After the desilication treatment of ZSM-5 crystals, the microporous channels in large zeolite crystals are more accessible. Despite the fact that it is still difﬁcult to get detailed information on the porous structure of mesoporous zeolites from optical and ﬂuorescence microscopy due to the limited resolution (a few micrometers), in situ measurement under typical catalytic reaction conditions a

c (A)

b

a

b

c (B)

(C)

Figure 9.22 (A,B) Confocal ﬂuorescence images measured in the in situ spectroscopic cell after oligomerization of 4-methoxystyrene at 373 K on the parent (A) and mesoporous (B) crystals (excitation at 561 nm, detection at 580–640 nm,). Indices (a–c) correspond to the upper horizontal plane, middle

horizontal plane parallel to the upper plane, and vertical intermediate plane, respectively. Measurements were performed ≈5 minutes after exposure to the styrene compound. (C) Schematic representation of the ZSM-5 crystal showing its individual building blocks. (Reprinted with permission from [133].)

9.4 Summary and Outlook

using these techniques are expected to shed more insights into the practical applications of these mesoporous zeolites.

9.4 Summary and Outlook

The beneﬁts of generating hierarchical pores in zeolite crystals have received general recognition in academic ﬁelds as well as industrial catalytic processes. A number of methods for generating additional pores in zeolites are now available and are being developed for a wider application to zeolites with different structure and compositions. Meanwhile, novel methods aiming for a higher extent of controllability and versatility still continue to come up. Different methods proceed according to different mechanisms, and therefore result in different end products in terms of textural properties and compositions. Nevertheless, whichever method is of choice in practice, a clear picture of the textural property and its evolution is always a prerequisite for a deeper understanding of the formation mechanisms of the mesopores and the impact of these pores on the practical applications of the zeolites, as well as for further optimizing the textural properties toward improved performance of the zeolites. It is clear that, not only the pore size and pore volume, which have been frequently reported as indicator values, but also the shape, location, and connectivity of these mesopores are important factors that determine the performances of zeolite. We have summarized here the characterization methods that have been used for studying the textural properties of zeolites, more speciﬁcally mesoporous zeolites. While conventional methods like nitrogen physisorption and TEM have been well developed and are still prevailing, the limits of these methods are evident. Other characterization methods summarized in this review, which have shown potential in providing complementary information on shape and connectivity of the mesopores, should be further exploited for characterizing mesoporous zeolites. Among others, ET is a unique technique and it provides direct and detailed information on the textural properties in both a qualitative and a quantitative way. Other methods, like thermoporometry, mercury intrusion porosimetry, which may be easier to implement, however, have not been widely applied in mesoporous zeolites. One possible obstacle is the uncertainty in explaining the characterization results, which in many cases is caused by the complex nature of the mesopores with irregular shape and broad size distribution in most mesoporous zeolites, especially those obtained by postsynthesis treatments. Therefore, advances in the theories of these methods toward a more realistic model are highly anticipated. With respect to this, the knowledge gained from well-deﬁned mesoporous materials, like MCM-41, is expected to facilitate their applications to mesoporous zeolites. One may also note that there are still a number of potential techniques for porous structure analysis that have not been applied to mesoporous zeolite characterization, for example, quasi-equilibrium thermodesorption, spectroscopic ellipsometry, light transmission, and radiation scattering of neutrons and X rays, to mention just

273

274

9 Textural Characterization of Mesoporous Zeolites

a few. The progress of the characterization techniques, both experimentally and theoretically, will deﬁnitely beneﬁt and accelerate the ‘‘designed synthesis’’ of mesoporous zeolites.

Acknowledgments

We thank the ﬁnancial support from the National Research School Combination Catalysis (NRSCC) and the ACTS/ASPECT program from The Netherlands Organization for Scientiﬁc Research (NWO-CW).

References 1. Breck, D.W. (1974) Zeolite Molecular

2.

3.

4. 5.

6.

7. 8. 9.

10.

11.

12. 13.

Sieves, Structure, Chemistry and Uses, Wiley-Intersciences, New York. Barrer, R.M. (1978) Zeolites and Clay Minerals as Sorbents and Molecular Sieves, Academic, London. van Bekkum, H., Flanigen, E.M., Jansen, J. C. (eds) (1991) Introduction to Zeolite Science and Practice, Elsevier, Amsterdam. Corma, A. (1995) Chem. Rev., 95, 559–614. van Steen, E., Claeys, M., and Callanan, L.H. (eds) (2005) Recent Advances in the Science and Technology of Zeolites and Related Materials, vol. 154B, Elsevier, Amsterdam. K¨arger, J. and Ruthven, D.M. (1992) Diffusion in Zeolites and Other Microporous Solids, John Wiley & Sons, Inc., New York. K¨arger, J. and Freude, D. (2002) Chem. Eng. Technol., 25, 769–778. ˇ Cejka, J. and Mintova, S. (2007) Catal. Rev.-Sci. Eng., 49, 457–509. Drews, T.O. and Tsapatsis, M. (2005) Curr. Opin. Colloid Interface Sci., 10, 233–238. Egeblad, K., Christensen, C.H., Kustova, M., and Christensen, C.H. (2008) Chem. Mater., 20, 946–960. Groen, J.C., Moulijn, J.A., and P´erez-Ram´ırez, J. (2006) J. Mater. Chem., 16, 2121–2131. Hartmann, M. (2004) Angew. Chem. Int. Ed., 43, 5880–5882. Kirschhock, C.E.A., Kremer, S.P.B., Vermant, J., Van Tendeloo, G., Jacobs,

14.

15.

16. 17. 18.

19. 20. 21.

22.

23.

24.

25.

P.A., and Martens, J.A. (2005) Chem. Eur. J., 11, 4306–4313. Meynen, V., Cool, P., and Vansant, E.F. (2007) Microporous Mesoporous Mater., 104, 26–38. P´erez-Ram´ırez, J., Christensen, C.H., Egeblad, K., Christensen, C.H., and Groen, J.C. (2008) Chem. Soc. Rev., 37, 2530–2542. Sch¨uth, F. (2003) Angew. Chem. Int. Ed., 42, 3604–3622. Sch¨uth, F. (2005) Annu. Rev. Mater. Res., 35, 209–238. Tao, Y.S., Kanoh, H., Abrams, L., and Kaneko, K. (2006) Chem. Rev., 106, 896–910. Tosheva, L. and Valtchev, V.P. (2005) Chem. Mater., 17, 2494–2513. Tosheva, L. and Valtchev, V.P. (2005) C. R. Chim., 8, 475–484. van Donk, S., Janssen, A.H., Bitter, J.H., and de Jong, K.P. (2003) Catal. Rev.-Sci. Eng., 45, 297–319. Davis, M.E., Saldarriaga, C., Montes, C., Garces, J., and Crowder, C. (1988) Nature, 331, 698–699. Freyhardt, C.C., Tsapatsis, M., Lobo, R.F., Balkus, K.J., and Davis, M.E. (1996) Nature, 381, 295–298. Burton, A., Elomari, S., Chen, C.Y., Medrud, R.C., Chan, I.Y., Bull, L.M., Kibby, C., Harris, T.V., Zones, S.I., and Vittoratos, E.S. (2003) Chem. Eur. J., 9, 5737–5748. Corma, A., D´ıaz-Caba˜ nas, M.J., Rey, F., Nicolooulas, S., and Boulahya, K. (2004) Chem. Commun., 1356–1357.

References 26. Corma, A., D´ıaz-Caba˜ nas, M.,

27.

28.

29.

30.

31.

32. 33. 34.

35. 36. 37.

38. 39.

40.

41.

42.

Martinez-Triguero, J., Rey, F., and Rius, J. (2002) Nature, 418, 514–517. Corma, A., D´ıaz-Caba˜ nas, M.J., Jord´a, J.L., Mart´ınez, C., and Moliner, M. (2006) Nature, 443, 842–845. Sun, J.L., Bonneau, C., Cant´ın, A., Corma, A., D´ıaz-Caba˜ nas, M.J., Moliner, M., Zhang, D.L., Li, M.R., and Zou, X.D. (2009) Nature, 458, 1154–1157. Choifeng, C., Hall, J.B., Huggins, B.J., and Beyerlein, R.A. (1993) J. Catal., 140, 395–405. Beyerlein, R.A., ChoiFeng, C., Hall, J.B., Huggins, B.J., and Ray, G.J. (1997) Top. Catal., 4, 27–42. Mavrodinova, V., Popova, M., Valchev, V., Nickolov, R., and Minchev, C. (2005) J. Colloid Interface Sci., 286, 268–273. On, D.T. and Kaliaguine, S. (2002) Angew. Chem. Int. Ed., 41, 1036–1040. On, D.T. and Kaliaguine, S. (2003) J. Am. Chem. Soc., 125, 618–619. Choi, M., Cho, H.S., Srivastava, R., Venkatesan, C., Choi, D.H., and Ryoo, R. (2006) Nat. Mater., 5, 718–723. Choi, M., Srivastava, R., and Ryoo, R. (2006) Chem. Commun., 4380–4382. Srivastava, R., Choi, M., and Ryoo, R. (2006) Chem. Commun., 4489–4491. Xiao, F.S., Wang, L.F., Yin, C.Y., Lin, K.F., Di, Y., Li, J.X., Xu, R.R., Su, D.S., Schl¨ogl, R., Yokoi, T., and Tatsumi, T. (2006) Angew. Chem. Int. Ed., 45, 3090–3093. Wang, H. and Pinnavaia, T.J. (2006) Angew. Chem. Int. Ed., 45, 7603–7606. Mei, C.S., Wen, P.Y., Liu, Z.C., Liu, H.X., Wang, Y.D., Yang, W.M., Xie, Z.K., Hua, W.M., and Gao, Z. (2008) J. Catal., 258, 243–249. Janssen, A.H., Schmidt, I., Jacobsen, C.J.H., Koster, A.J., and de Jong, K.P. (2003) Microporous Mesoporous Mater., 65, 59–75. Jacobsen, C.J.H., Madsen, C., Houzvicka, J., Schmidt, I., and Carlsson, A. (2000) J. Am. Chem. Soc., 122, 7116–7117. Schmidt, I., Boisen, A., Gustavsson, ˚ E., Stahl, K., Pehrson, S., Dahl, S.,

43.

44.

45.

46.

47.

48.

49.

50.

51.

52.

53.

54.

55.

56.

Carlsson, A., and Jacobsen, C.J.H. (2001) Chem. Mater., 13, 4416–4418. Fang, Y.M., Hu, H.Q., and Chen, G.H. (2008) Microporous Mesoporous Mater., 113, 481–489. Tao, Y.S., Kanoh, H., and Kaneko, K. (2003) J. Phys. Chem. B, 107, 10974–10976. Tao, Y.S., Tanaka, H., Ohkubo, T., Kanoh, H., and Kaneko, K. (2003) Adsorpt. Sci. Technol., 21, 199–203. Zhu, H., Liu, Z., Wang, Y., Kong, D., Yuan, X., and Xie, Z. (2008) Chem. Mater., 20, 1134–1139. Meyers, B.L., Fleisch, T.H., Ray, G.J., Miller, J.T., and Hall, J.B. (1988) J. Catal., 110, 82–95. Cartlidge, S., Nissen, H.U., and Wessicken, R. (1989) Zeolites, 9, 346–349. Horikoshi, H., Kasahara, S., Fukushima, T., Itabashi, K., Okada, T., Terasaki, O., and Watanabe, D. (1989) Nippon Kagaku Kaishi, 398–404. Chauvin, B., Massiani, P., Dutartre, R., Figueras, F., Fajula, F., and Descourieres, T. (1990) Zeolites, 10, 174–182. Guisnet, M., Wang, Q.L., and Giannetto, G. (1990) Catal. Lett., 4, 299–302. Wang, Q.L., Giannetto, G., and Guisnet, M. (1991) J. Catal., 130, 471–482. Wang, Q.L., Giannetto, G., Torrealba, M., Perot, G., Kappenstein, C., and Guisnet, M. (1991) J. Catal., 130, 459–470. Zholobenko, V.L., Kustov, L.M., Kazansky, V.B., Loefﬂer, E., Lohse, U., and Oehlmann, G. (1991) Zeolites, 11, 132–134. Beyerlein, R.A., Choifeng, C., Hall, J.B., Huggins, B.J., and Ray, G.J. (1994) in Fluid Catalytic Cracking III – Materials and Processes, vol. 571 (eds M.L.Occelli and P. Oconnor), ACS, Washington, D.C., pp. 81–97. Nesterenko, N.S., Thibault-Starzyk, F., Montouillout, V., Yuschenko, V.V., Fernandez, C., Gilson, J.P., Fajula, F., and Ivanova, I. (2004) Microporous Mesoporous Mater., 71, 157–166.

275

276

9 Textural Characterization of Mesoporous Zeolites 57. Katada, N., Kageyama, Y., Takahara,

58.

59.

60.

61.

62.

63.

64.

65.

66.

67.

68.

69. 70.

71.

72.

K., Kanai, T., Begum, H.A., and Niwa, M. (2004) J. Mol. Catal. A: Chem., 211, 119–130. van Bokhoven, J.A., Tromp, M., Koningsberger, D.C., Miller, J.T., Pieterse, J.A.Z., Lercher, J.A., Williams, B.A., and Kung, H.H. (2001) J. Catal., 202, 129–140. van Donk, S., Broersma, A., Gijzeman, O.L.J., van Bokhoven, J.A., Bitter, J.H., and de Jong, K.P. (2001) J. Catal., 204, 272–280. Dessau, R.M., Valyocsik, E.W., and Goeke, N.H. (1992) Zeolites, 12, 776–779. Lietz, G., Schnabel, K.H., Peuker, C., Gross, T., Storek, W., and V¨olter, J. (1994) J. Catal., 148, 562–568. Ogura, M., Shinomiya, S.Y., Tateno, J., Nara, Y., Kikuchi, E., and Matsukata, H. (2000) Chem. Lett., 882–883. Ogura, M., Shinomiya, S.Y., Tateno, J., Nara, Y., Nomura, M., Kikuchi, E., and Matsukata, M. (2001) Appl. Catal. A: Gen., 219, 33–43. Suzuki, T. and Okuhara, T. (2001) Microporous Mesoporous Mater., 43, 83–89. Su, L.L., Liu, L., Zhuang, J.Q., Wang, H.X., Li, Y.G., Shen, W.J., Xu, Y.D., and Bao, X.H. (2003) Catal. Lett., 91, 155–167. Goa, Y., Yoshitake, H., Wu, P., and Tatsumi, T. (2004) Microporous Mesoporous Mater., 70, 93–101. Pavel, C.C., Palkovits, R., Sch¨uth, F., and Schmidt, W. (2008) J. Catal., 254, 84–90. Pavel, C.C., Park, S.H., Dreier, A., Tesche, B., and Schmidt, W. (2006) Chem. Mater., 18, 3813–3820. Pavel, C.C. and Schmidt, W. (2006) Chem. Commun., 882–884. Holm, M.S., Egeblad, K., Vennestrom, P.N.R., Hartmann, C.G., Kustova, M., and Christensen, C.H. (2008) Eur. J. Inorg. Chem., 5185–5189. Cavalcante, C.L., Silva, N.M., Souza-Aguiar, E.F., and Sobrinho, E.V. (2003) Adsorption, 9, 205–212. Christensen, C.H., Johannsen, K., T¨ornqvist, E., Schmidt, I., Topsøe, H.,

73.

74. 75.

76.

77.

78. 79. 80. 81.

82.

83.

84.

85. 86. 87. 88.

89. 90.

91. 92.

and Christensen, C.H. (2007) Catal. Today, 128, 117–122. Hoang, V.T., Huang, Q.L., Malekian, A., Eic, M., Do, T.O., and Kaliaguine, S. (2005) Adsorption, 11, 421–426. Li, X.F., Prins, R., and van Bokhoven, J.A. (2009) J. Catal., 262, 257–265. Wei, X.T. and Smirniotis, P.G. (2006) Microporous Mesoporous Mater., 97, 97–106. Xu, B., Bordiga, S., Prins, R., and van Bokhoven, J.A. (2007) Appl. Catal. A: Gen., 333, 245–253. Rouquerol, J., Avnir, D., Fairbridge, C.W., Everett, D.H., Haynes, J.H., Pernicone, N., Ramsay, J.D.F., Sing, K.S.W., and Unger, K.K. (1994) Pure Appl. Chem., 66, 1739–1758. Gheorghiu, S. and Coppens, M.O. (2004) AICHE J., 50, 812–820. Marcilly, C. (1986) Pet. Technol., 328, 12–18. Kerr, G.T. (1967) J. Phys. Chem., 71, 4155–4156. Morin, S., Gnep, N.S., and Guisnet, M. (1998) Appl. Catal. A: Gen., 168, 63–68. Coster, D., Blumenfeld, A.L., and Fripiat, J.J. (1994) J. Phys. Chem., 98, 6201–6211. Triantaﬁllidis, C.S., Vlessidis, A.G., and Evmiridis, N.P. (2000) Ind. Eng. Chem. Res., 39, 307–319. Bor´eave, A., Auroux, A., and Guimon, C. (1997) Microporous Mater., 11, 275–291. Patzelov´a, V. and Jaeger, N.I. (1987) Zeolites, 7, 240–242. Zukal, A., Patzelov´a, V., and Lohse, U. (1986) Zeolites, 6, 133–136. Lynch, J., Raatz, F., and Dufresne, P. (1987) Zeolites, 7, 333–340. Sasaki, Y., Suzuki, T., Takamura, Y., Saji, A., and Saka, H. (1998) J. Catal., 178, 94–100. Hong, Y., Gruver, V., and Fripiat, J.J. (1994) J. Catal., 150, 421–429. Meyers, B.L., Fleisch, T.H., Ray, G.J., Miller, J.T., and Hall, J.B. (1988) J. Catal., 110, 82–95. Lee, K.H. and Ha, B.H. (1998) Microporous Mesoporous Mater., 23, 211–219. Dutartre, R., de M´enorval, L.C., Di Renzo, F., McQueen, D., Fajula,

References

93.

94.

95.

96.

97.

98.

99.

100.

101.

102.

103.

104. 105.

106.

107.

F., and Schulz, P. (1996) Mesoporous Mater., 6, 311–320. McQueen, D., Chiche, B.H., Fajula, F., Auroux, A., Guimon, C., Fitoussi, F., and Schulz, P. (1996) J. Catal., 161, 587–596. Pellet, R.J., Casey, D.G., Huang, H.M., Kessler, R.V., Kuhlman, E.J., Oyoung, C.L., Sawicki, R.A., and Ugolini, J.R. (1995) J. Catal., 157, 423–435. Rozwadowski, M., Komatowski, J., Wloch, J., Erdmann, K., and Golembiewski, R. (2002) Appl. Surf. Sci., 191, 352–361. Tromp, M., van Bokhoven, J.A., Oostenbrink, M.T.G., Bitter, J.H., de Jong, K.P., and Koningsberger, D.C. (2000) J. Catal., 190, 209–214. Giudici, R., Kouwenhoven, H.W., and Prins, R. (2000) Appl. Catal. A: Gen., 203, 101–110. van Donk, S., Bitter, J.H., Verberckmoes, A., Versluijs-Helder, M., Broersma, A., and de Jong, K.P. (2005) Angew. Chem. Int. Ed., 44, 1360–1363. Olken, M.M. and Garc´es, J.M. (1992) in Proceedings of 9th International Zeolite Conference, vol. 2, Butterworth-Heinemann, Boston, p. 559. Lee, G.-S.J., Garc´es, J.M., Meima, G.R., and van der Aalst, M.J.M. (1989) US Patent 5243116. Sulikowski, B., Borb´ely, G., Beyer, H.K., Karge, H.G., and Mishin, I.W. (1989) J. Phys. Chem., 93, 3240–3243. Le van mao, R., Vo, N.T.C., Sjiariel, B., Lee, L., and Denes, G. (1992) J. Mater. Chem., 2, 595–599. Datka, J., Kolidziejski, W., Klinowski, J., and Sulikowski, B. (1993) Catal. Lett., 19, 159–165. Kerr, G.T., Chester, A.W., and Olson, D.H. (1994) Catal. Lett., 25, 401–402. ˇ zmek, A., Suboti´c, B., Smit, ˇ Ciˇ I., Tonejc, A., Aiello, R., Crea, F., and Nastro, A. (1997) Microporous Mater., 8, 159–169. ˇ zmek, A., Suboti´c, B., Aiello, R., Ciˇ Crea, F., Nastro, A., and Tuoto, C. (1995) Microporous Mater., 4, 159–168. Groen, J.C., Peffer, L.A.A., Moulijn, J.A., and P´erez-Ram´ırez, J. (2005) Chem. Eur. J., 11, 4983–4994.

108. Groen, J.C., Moulijn, J.A., and

109.

110.

111.

112.

113.

114.

115.

116.

117.

118.

119.

120.

121.

122.

P´erez-Ram´ırez, J. (2007) Ind. Eng. Chem. Res., 46, 4193–4201. Groen, J.C., Maldonado, L., Berrier, E., Bruckner, A., Moulijn, J.A., and P´erez-Ram´ırez, J. (2006) J. Phys. Chem. B, 110, 20369–20378. Groen, J.C., Bach, T., Ziese, U., Paulaime-Van Donk, A.M., de Jong, K.P., Moulijn, J.A., and P´erez-Ram´ırez, J. (2005) J. Am. Chem. Soc., 127, 10792–10793. Groen, J.C., Moulijn, J.A., and P´erez-Ram´ırez, J. (2005) Microporous Mesoporous Mater., 87, 153–161. Groen, J.C., Sano, T., Moulijn, J.A., and P´erez-Ram´ırez, J. (2007) J. Catal., 251, 21–27. Groen, J.C., Abello, S., Villaescusa, L.A., and P´erez-Ram´ırez, J. (2008) Microporous Mesoporous Mater., 114, 93–102. Groen, J.C., Bruckner, A., Berrier, E., Maldonado, L., Moulijn, J.A., and P´erez-Ram´ırez, J. (2006) J. Catal., 243, 212–216. Groen, J.C., P´erez-Ram´ırez, J., and Peffer, L.A.A. (2002) Chem. Lett., 94–95. Groen, J.C., Zhu, W.D., Brouwer, S., Huynink, S.J., Kapteijn, F., Moulijn, J.A., and P´erez-Ram´ırez, J. (2007) J. Am. Chem. Soc., 129, 355–360. Groen, J.C., Caicedo-Realpe, R., Abello, S., and P´erez-Ram´ırez, J. (2009) Mater. Lett., 63, 1037–1040. Groen, J.C., Hamminga, G.M., Moulijn, J.A., and P´erez-Ram´ırez, J. (2007) Phys. Chem. Chem. Phys., 9, 4822–4830. Groen, J.C., Jansen, J.C., Moulijn, J.A., and P´erez-Ram´ırez, J. (2004) J. Phys. Chem. B, 108, 13062–13065. Groen, J.C., Peffer, L.A.A., Moulijn, J.A., and P´erez-Ram´ırez, J. (2004) Microporous Mesoporous Mater., 69, 29–34. Groen, J.C., Peffer, L.A.A., Moulijn, J.A., and P´erez-Ram´ırez, J. (2004) Colloids Surf. A, 241, 53–58. Groen, J.C., Peffer, L.A.A., Moulijn, J.A., and P´erez-Ram´ırez, J. (2005) in Nanoporous Materials, IV vol. 156 (eds

277

278

9 Textural Characterization of Mesoporous Zeolites

123.

124.

125.

126.

127. 128.

129.

130.

131. 132.

133.

134. 135. 136.

137. 138.

139.

A.Sayari and M. Jaroniec), Elsevier, Amsterdam, pp. 401–408. Gopalakrishnan, S., Zampieri, A., and Schwieger, W. (2008) J. Catal., 260, 193–197. Bjorgen, M., Joensen, F., Holm, M.S., Olsbye, U., Lillerud, K.P., and Svelle, S. (2008) Appl. Catal. A: Gen., 345, 43–50. Jin, L.J., Zhou, X.J., Hu, H.Q., and Ma, B. (2008) Catal. Commun., 10, 336–340. Choi, D.H., Park, J.W., Kim, J.H., Sugi, Y., and Seo, G. (2006) Polym. Degrad. Stab., 91, 2860–2866. Zhao, L., Shen, B.J., Gao, F.S., and Xu, C.M. (2008) J. Catal., 258, 228–234. Song, Y.Q., Zhu, X.X., Song, Y., Wang, Q.X., and Xu, L.Y. (2006) Appl. Catal. A: Gen., 302, 69–77. P´erez-Ram´ırez, J., Abello, S., Bonilla, A., and Groen, J.C. (2009) Adv. Funct. Mater., 19, 164–172. Wang, Y.R., Lin, M., and Tuel, A. (2007) Microporous Mesoporous Mater., 102, 80–85. Wang, Y.R. and Tuel, A. (2008) Microporous Mesoporous Mater., 113, 286–295. Holm, M.S., Svelle, S., Joensen, F., Beato, P., Christensen, C.H., Bordiga, S., and Bjørgen, M. (2009) Appl. Catal. A: Gen., 356, 23–30. Kox, M.H.F., Stavitski, E., Groen, J.C., P´erez-Ram´ırez, J., Kapteijn, F., and Weckhuysen, B.M. (2008) Chem. Eur. J., 14, 1718–1725. Ciesla, U. and Sch¨uth, F. (1999) Microporous Mesoporous Mater., 27, 131–149. He, X. and Antonelli, D. (2002) Angew. Chem. Int. Ed., 41, 214–229. Jacobsen, C.J.H., Madsen, C., Janssens, T.V.W., Jakobsen, H.J., and Skibsted, J. (2000) Microporous Mesoporous Mater., 39, 393–401. Madsen, C. and Jacobsen, C.J.H. (1999) Chem. Commun., 673–674. Schmidt, I., Madsen, C., and Jacobsen, C.J.H. (2000) Inorg. Chem., 39, 2279–2283. Chou, Y.H., Cundy, C.S., Garforth, A.A., and Zholobenko, V.L. (2006) Microporous Mesoporous Mater., 89, 78–87.

140. Kustova, M.Y., Rasmussen, S.B.,

141.

142.

143.

144.

145.

146.

147.

148.

149.

150.

151.

152.

153. 154. 155.

156.

Kustov, A.L., and Christensen, C.H. (2006) Appl. Catal. B: Environ., 67, 60–67. Kustov, A.L., Hansen, T.W., Kustova, M., and Christensen, C.H. (2007) Appl. Catal. B: Environ., 76, 311–319. Schmidt, I., Krogh, A., Wienberg, K., Carlsson, A., Brorson, M., and Jacobsen, C.J.H. (2000) Chem. Commun., 2157–2158. Christensen, C.H., Johannsen, K., Schmidt, I., and Christensen, C.H. (2003) J. Am. Chem. Soc., 125, 13370–13371. Christensen, C.H., Schmidt, I., Carlsson, A., Johannsen, K., and Herbst, K. (2005) J. Am. Chem. Soc., 127, 8098–8102. Kustova, M.Y., Hasselriis, P., and Christensen, C.H. (2004) Catal. Lett., 96, 205–211. Wei, X.T. and Smirniotis, P.G. (2006) Microporous Mesoporous Mater., 89, 170–178. Boisen, A., Schmidt, I., Carlsson, A., Dahl, S., Brorson, M., and Jacobsen, C.J.H. (2003) Chem. Commun., 958–959. Tao, Y.S., Kanoh, H., and Kaneko, K. (2003) J. Am. Chem. Soc., 125, 6044–6045. Tao, Y., Hattori, Y., Matumoto, A., Kanoh, H., and Kaneko, K. (2005) J. Phys. Chem. B, 109, 194–199. Tao, Y.S., Kanoh, H., Hanzawa, Y., and Kaneko, K. (2004) Colloids Surf. A, 241, 75–80. Zhu, K., Egeblad, K., and Christensen, C.H. (2007) Eur. J. Inorg. Chem., 3955–3960. Kustova, M., Egeblad, K., Zhu, K., and Christensen, C.H. (2007) Chem. Mater., 19, 2915–2917. Fang, Y.M. and Hu, H.Q. (2006) J. Am. Chem. Soc., 128, 10636–10637. Fang, Y.M. and Hu, H.Q. (2007) Catal. Commun., 8, 817–820. Li, H.C., Sakamoto, Y., Liu, Z., Ohsuna, T., Terasaki, O., Thommes, M., and Che, S.N. (2007) Microporous Mesoporous Mater., 106, 174–179. Kim, S.S., Shah, J., and Pinnavaia, T.J. (2003) Chem. Mater., 15, 1664–1668.

References 157. Yoo, W.C., Kumar, S., Wang, Z.Y.,

158.

159.

160.

161.

162.

163. 164.

165.

166.

167. 168.

169.

170.

171.

172.

Ergang, N.S., Fan, W., Karanikolos, G.N., McCormick, A.V., Penn, R.L., Tsapatsis, M., and Stein, A. (2008) Angew. Chem. Int. Ed., 47, 9096–9099. Fan, W., Snyder, M.A., Kumar, S., Lee, P.S., Yoo, W.C., McCormick, A.V., Penn, R.L., Stein, A., and Tsapatsis, M. (2008) Nat. Mater., 7, 984–991. Egeblad, K., Kustova, M., Klitgaard, S.K., Zhu, K.K., and Christensen, C.H. (2007) Microporous Mesoporous Mater., 101, 214–223. Christensen, C.H., Schmidt, I., and Christensen, C.H. (2004) Catal. Commun., 5, 543–546. Kustova, M.Y., Kustov, A., Christiansen, S.E., Leth, K.T., Rasmussen, S.B., and Christensen, C.H. (2006) Catal. Commun., 7, 705–708. Li, W.C., Lu, A.H., Palkovits, R., Schmidt, W., Spliethoff, B., and Sch¨uth, F. (2005) J. Am. Chem. Soc., 127, 12595–12600. Tao, Y.S., Kanoh, H., and Kaneko, K. (2005) Langmuir, 21, 504–507. Holland, B.T., Abrams, L., and Stein, A. (1999) J. Am. Chem. Soc., 121, 4308–4309. Rhodes, K.H., Davis, S.A., Caruso, F., Zhang, B.J., and Mann, S. (2000) Chem. Mater., 12, 2832–2834. Naydenov, V., Tosheva, L., and Sterte, J. (2003) Microporous Mesoporous Mater., 66, 321–329. Naydenov, V., Tosheva, L., and Sterte, J. (2002) Chem. Mater., 14, 4881–4885. Lee, Y.J., Lee, J.S., Park, Y.S., and Yoon, K.B. (2001) Adv. Mater., 13, 1259–1263. Zhang, B.J., Davis, S.A., Mendelson, N.H., and Mann, S. (2000) Chem. Commun., 781–782. Dong, A.G., Wang, Y.J., Tang, Y., Ren, N., Zhang, Y.H., Yue, J.H., and Gao, Z. (2002) Adv. Mater., 14, 926–929. Valtchev, V., Smaihi, M., Faust, A.C., and Vidal, L. (2003) Angew. Chem. Int. Ed., 42, 2782–2785. Valtchev, V.P., Smaihi, M., Faust, A.C., and Vidal, L. (2004) Chem. Mater., 16, 1350–1355.

173. Beck, J.S., Vartuli, J.C., Roth, W.J.,

174.

175.

176.

177. 178. 179. 180.

181.

182.

183.

184.

185.

186.

187.

188.

189.

Leonowicz, M.E., Kresge, C.T., Schmitt, K.D., Chu, C.T.W., Olson, D.H., and Sheppard, E.W. (1992) J. Am. Chem. Soc., 114, 10834–10843. Zhao, D.Y., Huo, Q.S., Feng, J.L., Chmelka, B.F., and Stucky, G.D. (1998) J. Am. Chem. Soc., 120, 6024–6036. Bagshaw, S.A., Baxter, N.I., Brew, D.R.M., Hosie, C.F., Nie, Y.T., Jaenicke, S., and Khuan, C.G. (2006) J. Mater. Chem., 16, 2235–2244. Bagshaw, S.A., Jaenicke, S., and Khuan, C.G. (2003) Catal. Commun., 4, 140–146. Liu, Y. and Pinnavaia, T.J. (2002) Chem. Mater., 14, 3–5. Liu, Y. and Pinnavaia, T.J. (2002) J. Mater. Chem., 12, 3179–3190. Liu, Y. and Pinnavaia, T.J. (2004) J. Mater. Chem., 14, 1099–1103. Liu, Y., Zhang, W.Z., and Pinnavaia, T.J. (2000) J. Am. Chem. Soc., 122, 8791–8792. Liu, Y., Zhang, W.Z., and Pinnavaia, T.J. (2001) Angew. Chem. Int. Ed., 40, 1255–1258. Ooi, Y.S., Zakaria, R., Mohamed, A.R., and Bhatia, S. (2004) Appl. Catal. A: Gen., 274, 15–23. Zhang, Z.T., Han, Y., Zhu, L., Wang, R.W., Yu, Y., Qiu, S.L., Zhao, D.Y., and Xiao, F.S. (2001) Angew. Chem. Int. Ed., 40, 1258–1262. Han, Y., Xiao, F.S., Wu, S., Sun, Y.Y., Meng, X.J., Li, D.S., Lin, S., Deng, F., and Ai, X.J. (2001) J. Phys. Chem. B, 105, 7963–7966. Han, Y., Wu, S., Sun, Y.Y., Li, D.S., and Xiao, F.S. (2002) Chem. Mater., 14, 1144–1148. Xiao, F.S., Han, Y., Yu, Y., Meng, X.J., Yang, M., and Wu, S. (2002) J. Am. Chem. Soc., 124, 888–889. Sun, Y.Y., Han, Y., Yuan, L.N., Ma, S.Q., Jiang, D.Z., and Xiao, F.S. (2003) J. Phys. Chem. B, 107, 1853–1857. Di, Y., Yu, Y., Sun, Y.Y., Yang, X.Y., Lin, S., Zhang, M.Y., Li, S.G., and Xiao, F.S. (2003) Microporous Mesoporous Mater., 62, 221–228. Lin, K.F., Sun, Z.H., Sen, L., Jiang, D.Z., and Xiao, F.S. (2004) Microporous Mesoporous Mater., 72, 193–201.

279

280

9 Textural Characterization of Mesoporous Zeolites 190. Xia, Y.D. and Mokaya, R. (2004) 191.

192.

193.

194.

195.

196. 197.

198.

199.

200.

201.

202.

203. 204.

205.

J. Mater. Chem., 14, 863–870. ´ ´ Agundez, J., D´ıaz, I., M´arquez-Alvarez, C., P´erez-Pariente, J., and Sastre, E. (2003) Chem. Commun., 150–151. Wang, S., Dou, T., Li, Y.P., Zhang, Y., Li, X.F., and Yan, Z.C. (2005) Catal. Commun., 6, 87–91. Ivanova, I.I., Kuznetsov, A.S., Yuschenko, V.V., and Knyazeva, E.E. (2004) Pure Appl. Chem., 76, 1647–1658. Goto, Y., Fukushima, Y., Ratu, P., Imada, Y., Kubota, Y., Sugi, Y., Ogura, M., and Matsukata, M. (2002) J. Porous Mater., 9, 43–48. Wang, H., Liu, Y., and Pinnavaia, T.J. (2006) J. Phys. Chem. B, 110, 4524–4526. Liu, Y. and Pinnavaia, T.J. (2004) J. Mater. Chem., 14, 3416–3420. Han, Y., Li, N., Zhao, L., Li, D.F., Xu, X.Z., Wu, S., Di, Y., Li, C.J., Zou, Y.C., Yu, Y., and Xiao, F.S. (2003) J. Phys. Chem. B, 107, 7551–7556. Prokesova-Fojokova, P., Mintova, S., ˇ Cejka, J., Zilkova, N., and Zukal, A. (2006) Microporous Mesoporous Mater., 92, 154–160. Karlsson, A., St¨ocker, M., and Schmidt, R. (1999) Microporous Mesoporous Mater., 27, 181–192. Huang, L.M., Guo, W.P., Deng, P., Xue, Z.Y., and Li, Q.Z. (2000) J. Phys. Chem. B, 104, 2817–2823. Liu, Y., Zhang, W.P., Liu, Z.C., Xu, S.T., Wang, Y.D., Xie, Z.K., Han, X.W., and Bao, X.H. (2008) J. Phys. Chem. C, 112, 15375–15381. Majano, G., Mintova, S., Ovsitser, O., Mihailova, B., and Bein, T. (2005) Microporous Mesoporous Mater., 80, 227–235. Fang, Y.M., Hu, H.Q., and Chen, G.H. (2008) Chem. Mater., 20, 1670–1672. Gagea, B.C., Liang, D., van Tendeloo, G., Martens, J.A., and Jacobs, P.A. (2006) Stud. Surf. Sci. Catal., 162, 259–266. Stevens, W.J.J., Meynen, V., Bruijn, E., Lebedev, O.I., van Tendeloo, G., Cool, P., and Vansant, E.F. (2008) Microporous Mesoporous Mater., 110, 77–85.

206. Wang, J., Groen, J.C., and Coppens,

207.

208.

209.

210.

211.

212.

213.

214.

215. 216.

217. 218.

219. 220. 221. 222.

223. 224.

225.

M.O. (2008) J. Phys. Chem. C, 112, 19336–19345. Wang, J., Groen, J.C., Yue, W., Zhou, W., and Coppens, M.O. (2007) Chem. Commun., 4653–4655. Wang, J., Groen, J.C., Yue, W., Zhou, W., and Coppens, M.O. (2008) J. Mater. Chem., 18, 468–474. Wang, J., Yue, W.B., Zhou, W.Z., and Coppens, M.O. (2009) Microporous Mesoporous Mater., 120, 19–28. Sing, K.S.W., Everett, D.H., Haul, R.A.W., Moscou, L., Pierotti, R.A., Rouquerol, J., and Siemieniewska, T. (1985) Pure Appl. Chem., 57, 603–619. Storck, S., Bretinger, H., and Maier, W.F. (1998) Appl. Catal. A: Gen., 174, 137–146. Brunauer, S., Emmett, P.H., and Teller, E. (1938) J. Am. Chem. Soc., 60, 309–319. Gregg, S.J. and Sing, K.S.W. (1982) Adsorption, Surface Area and Porosity, 2nd edn, Academic Press, London. Janssen, A.H., Koster, A.J., and de Jong, K.P. (2002) J. Phys. Chem. B, 106, 11905–11909. Lippens, B.C., Linsen, B.G., and de Boer, J.H. (1964) J. Catal., 3, 32–37. Jaroniec, M., Madey, R., Choma, J., McEnaney, B., and Mays, T.J. (1989) Carbon, 27, 77–83. Jaroniec, M., Kruk, M., and Olivier, J.P. (1999) Langmuir, 15, 5410–5413. Barret, E.P., Joyner, L.G., and Halenda, P.H. (1951) J. Am. Chem. Soc., 73, 373–380. Ball, P.C. and Evan, R. (1984) Langmuir, 5, 714–723. Ravikovitch, P.I. and Neimark, A.V. (2002) Langmuir, 18, 9830–9837. Ravikovitch, P.I. and Neimark, A.V. (2002) Langmuir, 18, 1550–1560. Groen, J.C. (2007) Mesoporus Zeolites Obtained by Desilication, PhD thesis, Technische Universiteit Delft. Kadlec, O. and Dubinin, M.M. (1969) J. Colloid Interface Sci., 31, 479–489. Burgess, C.G.V. and Everett, D.H. (1970) J. Colloid Interface Sci., 33, 611–614. Seaton, N.A. (1991) Chem. Eng. Sci., 46, 1895–1909.

References 226. Ojeda, M.L., Esparza, J.M., Campero,

227.

228.

229.

230. 231. 232.

233. 234.

235.

236.

237.

238.

239. 240.

241.

242.

243. 244.

A., Cordero, S., Kornhauser, I., and Rojas, F. (2003) Phys. Chem. Chem. Phys., 5, 1859–1866. Ravikovitch, P.I., Wei, D., Chueh, W.T., Haller, G.L., and Neimark, A.V. (1997) J. Phys. Chem. B, 101, 3671–3679. Ravikovitch, P.I. and Neimark, A.V. (2001) J. Phys. Chem. B, 105, 6817–6823. Lukens, W.W., Schmidt-Winkel, P., Zhao, D.Y., Feng, J.L., and Stucky, G.D. (1999) Langmuir, 15, 5403–5409. Kruk, M., Jaroniec, M., and Sayari, A. (1997) Langmuir, 13, 6267–6273. Neimark, A.V. (1995) Langmuir, 11, 4183–4184. Brun, M., Lallemand, A., Quinson, J.-F., and Eyraud, C. (1977) Thermochim. Acta, 21, 59–88. Denoyel, R. and Pellenq, R.J.M. (2002) Langmuir, 18, 2710–2716. Robens, E., Benzler, B., and Unger, K.K. (1999) J. Therm. Anal. Calorim., 56, 323–330. Cuperus, F.P., Bargeman, D., and Smolders, C.A. (1992) J. Membr. Sci., 66, 45–53. Iza, M., Woerly, S., Danumah, C., Kaliaguine, S., and Bousmina, M. (2000) Polymer, 41, 5885–5893. Ferguson, H.F., Frurip, D.J., Pastor, A.J., Peerey, L.M., and Whiting, L.F. (2000) Thermochim. Acta, 363, 1–21. Jallut, C., Lenoir, J., Bardot, C., and Eyraud, C. (1992) J. Membr. Sci., 68, 271–282. Faivre, C., Bellet, D., and Dolino, G. (1999) Eur. Phys. J. B, 7, 19–36. Ishikiriyama, K., Todoki, M., and Motomura, K. (1995) J. Colloid Interface Sci., 171, 92–102. Endo, A., Yamamoto, T., Inagi, Y., Iwakabe, K., and Ohmori, T. (2008) J. Phys. Chem. C, 112, 9034–9039. Janssen, A.H., Talsma, H., van Steenbergen, M.J., and de Jong, K.P. (2004) Langmuir, 20, 41–45. Le´on y Le´on, C.A. (1998) Adv. Colloid Interface Sci., 76-77, 341–372. Lohse, U. and Mildebrath, M. (1981) Z. Anorg. Allg. Chem., 476, 126–135.

245. Groen, J.C., Brouwer, S., Peffer, L.A.A.,

246.

247.

248.

249.

250. 251. 252.

253.

254.

255.

256.

257.

258.

259.

260. 261.

262.

and P´erez-Ram´ırez, J. (2006) Part. Part. Syst. Char., 23, 101–106. Groen, J.C., Peffer, L.A.A., and P´erez-Ram´ırez, J. (2003) Microporous Mesoporous Mater., 60, 1–17. Gai, P.L. and Buyes, E.D. (2003) Electron Microscopy in Heterogeneous Catalysts, Institute of Physics, Bristol. Ohsuna, T., Terasaki, O., Watanabe, D., Anderson, M.W., and Carr, S.W. (1994) Chem. Mater., 6, 2201–2204. Ersen, O., Hirlimann, C., Drillon, M., Werckmann, J., Tihay, F., Pham-Huu, C., Cruciﬁx, C., and Schultz, P. (2007) Solid State Sci., 9, 1088–1098. de Jong, K.P. and Koster, A.J. (2002) ChemPhysChem, 3, 776–780. Midgley, P.A. and Weyland, M. (2003) Ultramicroscopy, 96, 413–431. Deans, S.R. (1983) The Radon Transform and Some of its Applications, John Wiley & Sons, Inc., New York. Friedrich, H., de Jongh, P.E., Verkleij, A.J., and de Jong, K.P. (2009) Chem. Rev., 109, 1613–1629. Koster, A.J., Ziese, U., Verkleij, A.J., Janssen, A.H., and de Jong, K.P. (2000) J. Phys. Chem. B, 104, 9368–9370. Janssen, A.H., Koster, A.J., and de Jong, K.P. (2001) Angew. Chem. Int. Ed., 40, 1102–1104. Koster, A.J., Ziese, U., Verkleij, A.J., de Graaf, J., Gues, J.W., and de Jong, K.P. (2000) Stud. Surf. Sci. Catal., 130, 329–334. Ziese, U., Gommes, C.J., Blacher, S., Janssen, A.H., Koster, A.J., and de Jong, K.P. (2005) Stud. Surf. Sci. Catal., 158, 633–638. Springuel-Huet, M.A., Bonardet, J.L., G´ed´eon, A., and Fraissard, J. (1999) Magn. Reson. Chem., 37, S1–S13. Dybowski, C., Bansal, N., and Duncan, T.M. (1991) Annu. Rev. Phys. Chem., 42, 433–464. Fraissard, J. and Ito, T. (1988) Zeolites, 8, 350–361. Kneller, J.M., Pietrass, T., Ott, K.C., and Labouriau, A. (2003) Microporous Mesoporous Mater., 62, 121–131. Springuel-Huet, M.A. and Fraissard, J.P. (1992) Zeolites, 12, 841–845.

281

282

9 Textural Characterization of Mesoporous Zeolites (2007) Catal. Today, 126, 44–53.

263. Cotterman, R.L., Hickson, D.A.,

264.

265.

266.

267.

268.

Cartlidge, S., Dybowski, C., Tsiao, C., and Venero, A.F. (1991) Zeolites, 11, 27–34. Raftery, D., MacNamara, E., Fisher, G., Rice, C.V., and Smith, J. (1997) J. Am. Chem. Soc., 119, 8746–8747. Cotts, R.M., Hoch, M.J.R., Sun, T., and Markert, J.T. (1989) J. Magn. Reson., 83, 252–266. Kortunov, P., Vasenkov, S., K¨arger, J., Valiullin, R., Gottschalk, P., El´ıa, M.F., Perez, M., St¨ocker, M., Drescher, B., McElhiney, G., Berger, C., Gl¨aser, R., and Weitkamp, J. (2005) J. Am. Chem. Soc., 127, 13055–13059. Schmidt, R., Hansen, E.W., St¨ocker, M., Akporiaye, D., and Ellestad, O.H. (1995) J. Am. Chem. Soc., 117, 4049–4056. Roeffaers, M.B.J., Hofkens, J., de Cremer, G., de Schryver, F.C., Jacobs, P.A., de Vos, D.E., and Sels, B.F.

269. Roeffaers, M.B.J., Sels, B.F., Uji-i, H.,

270.

271.

272.

273.

de Schryver, F.C., Jacobs, P.A., and de Vos, D.E. (2006) Nature, 439, 572–575. Mores, D., Stavitski, E., Kox, M.H.F., Kornatowski, J., Olsbye, U., and Weckhuysen, B.M. (2008) Chem. Eur. J., 14, 11320–11327. Stavitski, E., Kox, M.H.F., and Weckhuysen, B.M. (2007) Chem. Eur. J., 13, 7057–7065. Kox, M.H.F., Stavitski, E., and Weckhuysen, B.M. (2007) Angew. Chem. Int. Ed., 46, 3652–3655. Karwacki, L., Stavitski, E., Kox, M.H.F., Kornatowski, J., and Weckhuysen, B.M. (2007) Angew. Chem. Int. Ed., 46, 7228–7231.

283

10 Aluminum in Zeolites: Where is it and What is its Structure? Jeroen A. van Bokhoven and Nadiya Danilina

10.1 Introduction

Aluminum in a zeolite framework position introduces the negative charge in the framework that requires a cation to balance it. This gives zeolites much of their properties, such as cation exchange capacity and the ability to catalyze reactions. The number of active sites often scales with the number of aluminum atoms in the framework. The zeolite hydrophobicity and hydrophilicity also strongly depend on the aluminum content, which affect the adsorption of polar and apolar molecules into the zeolite [1, 2]. Thus, besides structure, the silicon to aluminum ratio determines the zeolite performance. The silicon to aluminum ratio can be accurately determined by using a combination of characterization methods, such as 29 Si and 27 Al magic-angle spinning nuclear magnetic resonance (MAS NMR) and elemental analysis. From the 27 Al MAS NMR spectra the framework and extra-framework nature of aluminum can be also established. The aluminum coordination strongly varies with the pretreatment condition to which the zeolite was exposed. For example, exposing an acidic zeolite to moisture causes partial dealumination [3, 4]. Such structural changes are to a large extent reversible and strongly affect the catalytic performance of the zeolite. Steaming, a treatment that is used at a very large scale in industry to improve zeolite performance, causes dealumination, formation of new aluminum species, creation of Lewis acidity, and the formation of secondary mesopores. This treatment succeeds in improving the zeolite performance in terms of activity and stability in catalytic cracking, alkylation, acylation, and so on [5]. An additional factor that affects the zeolite properties is the distribution of aluminum. This can be regarded in two manners, which affect the zeolite performance in a different way. The ﬁrst is the zoning of aluminum over the individual crystals and the second is the distribution over the crystallographic T sites. A T site is a crystallographic position, which is occupied by a tetrahedrally coordinated atom, usually silicon and aluminum. The number and degeneracy of T sites and the spatial aluminum distribution in a zeolite vary with zeolite structure type [6, 7].

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

284

10 Aluminum in Zeolites: Where is it and What is its Structure?

For most of the zeolite structure types, the aluminum distribution remains unidentiﬁed. Recent progress in the above-mentioned ﬁelds will be described using selected examples of the literature.

10.2 Structure of Aluminum Species in Zeolites

Framework aluminum is tetrahedrally coordinated and 27 Al MAS NMR under hydrated conditions typically represents this species as a narrow resonance at a chemical shift around 60 ppm. The exact chemical shift depends on the local structure of the aluminum and varies with the averaged Al–O–Si angles according to δiso = −0.5θ + 132

(10.1)

This empirical formula has been successfully used to interpret 27 Al MAS NMR data of various zeolites and in some cases, aluminums that occupy different crystallographic T sites were resolved, which is described further below. The aluminum coordination strongly varies with the exact pretreatment condition. High temperature steam-activation or treatment of the zeolite in the presence of moisture results in partial collapse of the zeolite framework, which in addition to the tetrahedrally coordinated framework aluminum leads to various other aluminum species. These can be three- and ﬁve-coordinated and (distorted) tetrahedrally and octahedrally coordinated. Such species can be identiﬁed by different characterization tools, such as NMR [8–10], X-ray photoelectron spectroscopy (XPS) [11], and X-ray absorption spectroscopy (XAS) [12–14]. Because of its high resolution and widespread availability of magnets, NMR is the most used method. High ﬁeld magnets and the use of advanced pulse schemes, such as multiple quantum-NMR, provide high-resolution data that yield the aluminum coordination unambiguously. Various schemes to provide a quantitative analysis have been described [15, 16]. Zeolite samples are generally hydrated before performing an NMR measurement, because NMR spectra of hydrated zeolites identify the framework aluminum as the narrow resonance around 60 ppm. Non-hydrated or dehydrated samples show very broad and often overlapping resonances, because aluminum is a quadrupolar nucleus, which is sensitive to its three-dimensional surrounding. Resonances can even become broadened beyond detection; the corresponding aluminum species are called NMR-invisible. However, recent technology and the correct use of experimental conditions should in general enable a quantitative detection of all aluminum in a sample [15]. Thus, the aluminum structure under dehydrated conditions has been successfully determined by 27 Al MAS NMR. The framework and extra-framework aluminum of non-hydrated zeolites were determined of a steamed zeolite Y of silicon to aluminum ratio of 2.7 using a spin-echo pulse scheme [16]. Various aluminum species were detected: Tetrahedrally coordinated framework aluminums that were charge balanced by protons and/or sodium, respectively extra-framework aluminum, present in form of extra-framework aluminum cations and octahedrally

10.2 Structure of Aluminum Species in Zeolites

coordinated aluminum in neutral extra-framework aluminum oxide clusters. The kind and amount of species signiﬁcantly differed from those that are typically detected in steamed zeolites that were measured after exposure to moisture. In such sample, in addition to the typical tetrahedrally coordinated framework aluminum, a large amount of distorted tetrahedrally coordinated aluminum was detected, two types of octahedrally coordinated aluminum, which differed in their symmetry, and a small fraction of ﬁve-coordinated aluminum [17]. Moisture clearly plays a dominant role in affecting the aluminum coordination. In situ XAS at the Al K edge has identiﬁed a strong variation of the aluminum coordination as function of temperature [18]. The fraction of octahedrally coordinated aluminum, which is present in steamed zeolite Y (USY) when measured at room temperature in the presence of moisture, decreased upon increasing the temperature [3, 19]. About 50% of the octahedral aluminum lowered its coordination to tetrahedral aluminum upon heating to 400 ◦ C. Heating a zeolite to a temperature in excess of 400 ◦ C causes only small changes in the zeolite structure. An under-coordinated, most probably threefold coordinated aluminum species were detected even in the presence of small quantities of moisture [20]. Thus, when performing a catalytic reaction at high temperature, the structure that is typically measured by 27 Al MAS NMR is not representative of the structure of the catalyst under catalytically relevant conditions. This might be one of the reasons, why the role of extra-framework aluminum species on catalytic performance has remained poorly understood for so long. 10.2.1 Reversible versus Irreversible Structural Changes

A reversibility of the aluminum coordination in zeolites has been shown for many zeolites [3, 21–23]. The charge-compensating cation in zeolite beta was shown to strongly affect the aluminum coordination. Proton-exchanged beta showed a large fraction of up to 25% of octahedrally coordinated aluminum, which was quantitatively reverted to tetrahedral coordination after ion exchange with alkali cations and also after treatment with ammonia [21]. In situ Al K edge XAS [18] and 27 Al MAS NMR [21–23] showed that part of the framework tetrahedrally coordinated aluminum in proton charge-balanced zeolites is converted into an octahedral coordination at room temperature. This octahedrally coordinated aluminum is unstable and reverts into the tetrahedral coordination by heating above 100 ◦ C [18]. However, the presence of base-like ammonia is required to quantitatively restore the framework structure [21, 22]. Because of the ability of the aluminum atom to reversibly change its coordination, the octahedrally coordinated aluminum has been called framework-associated aluminum. The amount of framework-associated aluminum that forms in a zeolite depends on the silicon to aluminum ratio and the zeolite framework type [3]. The higher the aluminum content, the more octahedral aluminum forms. Among the different zeolites, the aluminum atoms in zeolite beta show a strong tendency of adopting the octahedral coordination.

285

286

10 Aluminum in Zeolites: Where is it and What is its Structure?

10.2.2 Cautionary Note

Various characterization methods require different pretreatments of the zeolite samples. Infrared spectroscopy is measured after heating a zeolite at high temperature to remove moisture from the sample, 27 Al MAS NMR is measured after exposure to moisture to obtain high-resolution spectra, XPS is measured in vacuum, and so on. As a result, the structure that is probed differs. 10.2.3 Development of Activity and Changing Aluminum Coordination

The formation of framework-associated octahedrally coordinated aluminum is related to a loss of catalytic performance [24, 25]. The development of the aluminum coordination was compared to the high-temperature propane cracking activity. Zeolite Y with silicon to aluminum ratio of 2.6 was treated according to Scheme 10.1. Four samples were formed, which were characterized by NMR and infrared spectroscopy, X-ray diffraction, and nitrogen physisorption (Figures10.1–10.3 and Table 10.1). NH4 Y contained only tetrahedrally coordinated aluminum as indicated by a narrow resonance around 60 pp m in the 27 Al MAS NMR spectrum. However, exposure to moisture after the high-temperature treatment caused the formation of octahedrally coordinated aluminum indicated by the resonance at about 0 ppm (Figure 10.1). The transformation from framework tetrahedral to octahedral coordination occurs already at room temperature [18, 19, 23]. Removal of ammonia at high temperature and measuring an infrared spectrum shows the typical stretching hydroxyl groups in the supercage and sodalite cage at about 3640 and 3550 cm−1 , respectively (Figure 10.2). Also the XRD pattern showed loss of crystallinity indicated by the increased width of the diffraction peaks (Figure 10.3). Treatment with ammonia at 150 ◦ C before measuring infrared, 27 Al MAS NMR, and XRD yielded virtually identical spectra and patterns as the parent material. This IR after calcination 27

NH4Y

550 °C

Moisture HY

Al MAS NMR under wet conditions

150 °C with NH3

H(NH3)Y

Propane cracking activity Scheme 10.1 Treatment scheme of zeolite Y before kinetic measurement and characterization.

550 °C

No moisture HY ′

10.2 Structure of Aluminum Species in Zeolites

H(NH3)Y HY NH4Y 100

80

Figure 10.1

60

27 Al

40 ppm

20

−20

0

MAS NMR spectra of zeolite Y, treated according to Scheme 10.1.

H H

O

O

S

A

S

H(NH3)Y HY NH4Y 3800

3600

3400

Wavenumber (cm−1) Figure 10.2

FTIR spectra of zeolite Y, treated according to Scheme 10.1.

shows that the original structure can be completely recovered and that complete re-insertion of now-tetrahedral aluminum into the framework occurs. However, heating HY prior to treatment with ammonia causes additional structural damage that cannot be simply repaired by treatment under ammonia (Scheme 10.2). The catalytic activity of the parent NH4 Y and ammonia-treated sample H(NH3 )Y were virtually identical and much higher than that of the HY. That the catalytic activity in HY was much lower can be understood from the infrared spectrum, which shows a very low number of Brønsted acid sites (Figure 10.2). The heating of HY during the pretreatment of the infrared measurement, respectively, catalytic reaction causes a signiﬁcant structural damage. A pretreatment of HY at high temperature (550 ◦ C) before measuring nitrogen physisorption also showed a micropore volume and

287

288

10 Aluminum in Zeolites: Where is it and What is its Structure?

H(NH3)Y

HY

NH4Y

10

20

30

40

2q Figure 10.3

XRD patterns of zeolite Y, treated according to Scheme 10.1.

Table 10.1 Results of nitrogen physisorption for zeolite Y, treated according to Scheme 10.1.

Sample

NH4 Y HY H(NH3 )Y HY

BET surface area (m2g–1 )

Micropore volume (cm3g–1 )

810 450 770 230

0.32 0.18 0.31 0.09

BET surface area of 0.09 cm3 g−1 and 230 m2 g−1 compared to 0.32 cm3 g−1 and 810 m2 g−1 for NH4 Y (Table 10.1). Apparently, the defects that were formed after the exposure of the sample to moisture are starting points for the further structural collapse, which occurs upon heating. Unlike the initial formation of octahedrally coordinated aluminum, these structural changes are much more severe and irreversible. Scheme 10.2 (adapted from [26]) summarizes the structural changes that occur after the various treatments. The role of steaming on generating highly active and selective catalysts has been debated in the literature for over decades [13, 27–29]. Because of the limitations in structure determination under catalytically relevant conditions in the past, structure–performance relations do not exist for many zeolite-catalyzed reaction systems. The complexity of chemical reactions over zeolites makes it likely that different reactions are affected differently by steaming. The extent of inﬂuence

10.3 Where is the Aluminum in Zeolite Crystals? NH4+ O

TO Si TO

OSi Al

O Si

OT

−NH3 450 °C

OSi

H O

TO

OSi

Si

Al

H2O

Si

O Si

OT

OSi

TO

(b) +NH3

Scheme 10.2

OH

OT H2O OH H2O (c)

Complete reversal to (a)

Species (c) High T

H2O Al

RT TO

(a)

150 °C

O

TO

289

Irreversible structural collapse, the more so under moisture

Development of zeolite structure under various conditions.

of extra-framework species to activate the reactants [30], of mesopores to alleviate diffusion limitation [31], and of isolation of the Brønsted acid sites [32] will depend on the kind of reaction that is performed as well as on the reaction conditions.

10.3 Where is the Aluminum in Zeolite Crystals?

Zeolite performance is affected by many parameters, mainly by zeolite structure type, silicon to aluminum ratio, and the distribution of aluminum. In a single crystal of zeolite of a particular structure type, the aluminum may not be homogeneously distributed throughout the crystal, a phenomenon that is called zoning. This spatial aluminum distribution should be differentiated from the aluminum and silicon distribution over the crystallographic T sites within the zeolite structure [33–35]. Although recent progress has been made in this ﬁeld, the assignment of aluminum and silicon distributions to their occupancy of the crystallographic T sites remains one of the largest challenges in the structure determination of zeolites. 10.3.1 Aluminum Zoning

The physical properties and catalytic behavior of aluminosilicates are closely related to the aluminum content and distribution in their frameworks and over the crystal. It is well known that aluminum is often non-uniformly distributed in different zeolites [36, 37]. This aluminum zoning plays an important role in the design of catalytically active sites and their accessibility. The occurrence and kind of aluminum zoning in zeolites depend on many factors, such as zeolite framework [38, 39], silicon to aluminum ratio [40], crystal size [39], and synthesis parameters [35, 41–43]. The aluminum distribution on the surface and in the bulk of ZSM-5 has been most investigated, sometimes giving contradictory results. A heterogeneous aluminum distribution in large crystals of ZSM-5 was described by electron

290

10 Aluminum in Zeolites: Where is it and What is its Structure?

microprobe analysis [36, 39]. The crystals were synthesized in the presence of tetrapropylammonium ions and were about 40 µm in size and had silicon to aluminum ratio between 50 and 100. The analysis of a volume of about 2 µm3 of the sample grinded with a diamond paste to remove the outer layers of the crystals showed a silicon-rich core and an aluminum-rich rim. The aluminum concentration in the rim was up to 10 times higher than in the core of the crystal. The zoning was symmetric and the surface aluminum concentration was fairly constant. The non-uniform aluminum distribution was explained by the crystallization mechanism of ZSM-5. Initially, an aluminum-rich gel is formed at the bottom of the reactor and the MFI phase nucleates from the silica-rich solution. At a later stage, aluminum, released from the gel, is progressively incorporated into the ZSM-5 framework. A similar behavior was reported for smaller ZSM-5 crystals of about 1 µm [37, 44]. The spatial aluminum distribution was found to be dependent on the synthesis route [41, 43]. By using different templates, the spatial distribution of aluminum in ZSM-5 crystals can thus be controlled. Tetrapropylammonium ions as template lead to strongly zoned proﬁles with the aluminum enriched in the outer shell of the crystals. In contrast, using 1,6-hexanediol as template or using a completely inorganic reaction mixture leads to homogeneous aluminum proﬁles. An explanation of the difference between the templates was given in terms of the interaction with the template molecule. Tetrapropylammonium ions in the presence of alkali ions interact preferentially with the silicate species, thus leading to incorporation of these species in the growing crystals. In a tetrapropylammonium ion-free system, the sodium ions, which interact strongly with aluminate species, are assumed to play the templating role. Nevertheless, there are studies conﬁrming the homogeneous aluminum distribution across the tetrapropylammonium-ZSM-5 crystals of varying size [38, 42] or even a presence of a siliceous outer surface and an aluminum-rich interior [45]. In these studies surface-sensitive techniques, such as Auger electron spectroscopy and electron microprobe analysis, were used to obtain the depth proﬁles. It was reported that different Si/Na, Si/tetrapropylammonium ion, and Si/H2 O ratios in the synthesis gel can lead to reverse silicon to aluminum ratio proﬁles across a crystal [41]. An aluminum-rich core was obtained, when low Si/Na and Si/tetrapropylammonium ion ratios and larger amount of water were used in the synthesis. The inﬂuence of the nature and the amount of alkali cations introduced into the synthesis gel on the aluminum distribution were also studied [46]. It was found that the counter-ion, such as lithium, sodium, potassium, rubidium, and cesium, does not inﬂuence the surface aluminum concentration and its distribution. In these studies, crystals of different size (0.4−100 µm) were compared. The surface analysis was performed using X-ray photoelectron spectroscopy and energy-dispersive X-ray spectroscopy, while the bulk composition was analyzed by proton-induced gamma-ray emission. The penetration depth for these techniques is about 10 nm, 1 µm, and 10 µm, respectively [41]. Only a few studies deal with aluminum zoning throughout the crystals of zeolites other than ZSM-5. Zeolite beta, synthesized using tetrapropylammonium hydroxide, showed an aluminum-rich core and silicon-rich shell, a behavior that

10.3 Where is the Aluminum in Zeolite Crystals?

is reverse compared to what is generally observed in ZSM-5 crystals [47]. The same, however, was observed for Na-X crystals [48]. Based on a theoretical study, aluminum-enrichment on the surface of the crystals was reported for zeolite Y [49]. On the other hand, applying Auger electron spectroscopy [38], secondary ion mass spectrometry [7], and fast atom bombardment mass spectrometry [50], it was shown that the aluminum distribution in zeolites Y, A, X, and ZSM-5 was homogeneous. It is obvious that the choice of the right technique is crucial to characterize the aluminum zoning. A reliable method may be scanning electron microscopy with energy-dispersive X-ray mapping [30, 32]. Because the penetration depth of the electron beam is about 1 µm, the crystal should be cut or the outer layers removed to insure the analysis is performed throughout the whole volume and not restricted only to the surface. Rubbing off surface layers or polishing the crystal result in a rough surface or it may break or crush crystals, which will affect the signal. A smooth cut of the crystal can be obtained by using a focused ion beam, Figure 10.4 shows a large crystal of ZSM-5, which was cut with a focused ion beam and its aluminum, silicon, and oxygen distribution using the energy-dispersive X-ray mapping (not published results). This crystal was synthesized in the presence of tetrapropylammonium ions as template. Most of the aluminum is concentrated in the 2−3 µm broad rim of the crystal at the expense of silicon. Desilication of the crystals and subsequent scanning electron microscopy analysis were also used to visualize aluminum zoning [41, 51]. Base-treatment might change the elemental distribution and/or affect the defect concentration. Figure 10.5 shows the scanning electron micrograph of a base-treated ZSM-5 crystal, of which the original was shown in Figure 10.4. Their morphology, hollow cofﬁns, conﬁrmed that the crystals had a silicon-rich core, which was leached out, and an aluminum-rich rim. 4 µm

4 µm

(a)

(b)

4 µm

4 µm

(c)

(d)

Figure 10.4 SE micrograph of the FIB cut crystal (a) and EDX (energy dispersive X-ray spectroscopy) maps, representing aluminum (b), silicon (c), and oxygen (d) distribution.

291

292

10 Aluminum in Zeolites: Where is it and What is its Structure?

Figure 10.5

SEM images of large crystals of NaOH-treated ZSM-5.

10.3.2 Aluminum Distribution Over the Crystallographic T Sites

The chemical shift in 27 Al MAS NMR is determined by the local geometry of the aluminum atom. Equation (10.1) gave the empirical formula that relates the averaged T-O-T angle to chemical shift. In certain cases, the framework aluminum atoms are partially resolved in 27 Al MAS NMR [52]. Figure 10.6 shows an example of zeolite beta. Two partially overlapping resonances, referred to as Td1 and Td2, represent the tetrahedrally coordinated framework aluminum. The different chemical shift of Td1 and Td2 originates from the different local structure of the aluminum atoms that occupy different T sites, notably the averaged T-O-T

Td2 Td1

80

60

40

20

0

ppm Figure 10.6

27

Al MAS NMR of zeolite beta.

−20

10.3 Where is the Aluminum in Zeolite Crystals?

293

Td2 80 70

215 110 56

Relative intensity

Si/Al

27

60

40

20 ppm

0

50 Td1

40 30 20

10.5 80 (a)

60

10

9 −20

0 (b)

9 10.5 27 56 110 215 Si/Al ratio

Figure 10.7 (a) 27 Al MAS NMR of zeolite beta of different silicon to aluminum ratios and (b) relative intensities of Td1 (blue) and Td2 (magenta).

angle (Eq. (10.1)) [22]. Zeolite beta can be synthesized with different silicon to aluminum ratio. Quantiﬁcation of the 27 Al MAS NMR spectra of these indicated a change in the relative ratio of intensity of the two contributions (Figure 10.7a,b). The higher the silicon to aluminum ratio was, the higher the relative intensity of Td2. This shows that depending on the aluminum content of the framework, the aluminum has the tendency to occupy different crystallographic sites. Similar results were obtained for ZSM-5 [53], mordenite [54], and NH4 -Y [55]. In a previous section, the instability of framework aluminum in proton-exchanged zeolites was described. After converting the zeolite beta to the proton form, a large fraction of octahedrally coordinated aluminum appeared (Figure 10.7a). Interestingly, only one of tetrahedral resonances decreased intensity, viz. Td2. A more harsh treatment even removed most of the framework aluminum that was associated with Td2, while the intensity of the resonance Td1 only slightly changed. This indicates that the stability of framework aluminum depends on the speciﬁc aluminum occupancy of T sites. Moreover, the higher the silicon to aluminum ratio is, the lower the fraction of octahedrally coordinated aluminum (Figure 10.7a), in all cases, originating mainly from Td2. The fewer aluminum atoms occupy the framework, the more stable they are to moisture as described above. Obviously, a zeolite with very high silicon to aluminum ratio is hydrophobic, which prevents water to adsorb in the pores, which thus protects the framework aluminum from being exposed to water. More recently, 27 Al MQ MAS NMR on ZSM-5 samples that were synthesized using different methods identiﬁed that the aluminum distribution depends on the synthesis parameters [53]. Using mixed quantum mechanics/molecular dynamics calculations a more precise correlation between chemical shift and occupancy of T sites was proposed [56]. Throughout the set of samples, 10 resonances

294

10 Aluminum in Zeolites: Where is it and What is its Structure?

were identiﬁed, the assignment of which was tentative for only 4 resonances. Thus, a much more precise assignment of aluminum onto crystallographic sites could be performed. This represents a signiﬁcant improvement in interpreting 27 Al MAS NMR spectra compared to the previous studies. The effect of using different organic molecules as structure-directing agents on the distribution of aluminum over the T sites was recently investigated for ferrierite [57, 58]. Different ferrierite samples were synthesized in the absence of sodium cations by using 1-benzyl-1-methylpyrrolidinium, tetramethylammonium cations, and pyrrolidine. The distribution of the acid sites of the zeolite was associated to the aluminum distribution. The acidity of the materials was investigated by FTIR spectroscopy and pyridine adsorption, and furthermore, correlated to their activity in m-xylene and n-butene isomerization. The distribution of bridging hydroxyl groups that were located in the ferrierite cage accessible through windows of eight-membered rings or in the 10-membered ring channel varied with the speciﬁc combination of templates used in the synthesis. The dependence of aluminum distribution over the crystallographic T sites in ZSM-5 on the silicon to aluminum ratio and the source of silicon and aluminum was shown by using cobalt(II) ions as probes for the so-called ‘‘Al pairs,’’ [Al-O-(Si-O)1,2 -Al] [34, 35, 59]. Co2+ ions in dehydrated zeolites are suggested to be coordinated only to framework oxygen atoms and balanced by two framework AlO2 − tetrahedrals. The concentration of Co2+ in Co-exchanged zeolites corresponds to the concentration of the ‘‘Al pairs’’ in the zeolite framework. The distribution of Co2+ ions at cationic sites can be estimated from UV–Vis spectra. The non-random aluminum distribution was also found for mordenite [54, 60, 61], zeolite beta [60], zeolite omega [62], faujasite [55], ferrierite [57, 58], and merlinoite [63]. NMR spectroscopy was mainly used to determine the aluminum and silicon distribution. CD3 CN and benzene adsorption measurements and DFT quantum chemical calculations showed that aluminum is preferentially sitting in the T site of mordenite, located at the bottom of the side pockets [60, 61]. By ﬁtting the results of the numerical simulated annealing calculations to the experimental data obtained by 27 Al and 29 Si MAS NMR, it was suggested that aluminum preferentially occupies the tetrahedral site in the six-membered rings of zeolite omega [62]. Although the majority suggests the aluminum distribution in the zeolite framework to be non-random, a few groups concluded a random aluminum distribution based on the 27 Al MQ, 29 Si, and 1 H MAS NMR experiments [64–67]. The above-mentioned studies were able to partially resolve the aluminum distribution and further progress is expected to be made. Recently, an alternative method to determine the silicon and aluminum distribution over the T sites has been proposed [33]. The method is based on the formation of an X-ray standing wave, which is generated in a single crystal of a zeolite. When an X-ray beam is diffracted, the incoming X-ray and the diffracted one interfere with each other. They thus generate a standing wave, which is parallel to the d-spacing. Thus, when looking perpendicular to the d-planes, there are certain positions in the unit cell, which experience X-ray intensity by positive interference and areas that have no or less intensity by negative interference. Figure 10.8 shows a schematic representation of a standing wave along a unit cell. The yellow planes represent the

10.3 Where is the Aluminum in Zeolite Crystals?

Figure 10.8

Schematic representation of an X-ray standing wave in the scolecite structure.

maximal intensity of the standing wave that is generated by interference between the initial and diffracted X-ray of the (040) reﬂection in zeolite scolecite, which is a naturally occurring zeolite that has two crystallographic T sites. Now, if an element is positioned where X-ray intensity is high, thus, at the position of one of the yellow planes in Figure 10.6, this atom will be excited and emits ﬂuorescence radiation, whose energy is typical of its Z number. By detecting the ﬂuorescence radiation in an energy-dispersive manner, aluminum and silicon atoms can be distinguished. Changing the energy of the incoming X-ray will vary the interference pattern and move the intensity perpendicularly to the d-planes. Alternatively, varying the incoming angle of the X-ray to the d-plane will also cause the standing wave to propagate through the unit cell. In such an experiment, the occupancy of the crystallographic sites by aluminum or silicon can be unambiguously determined by simulation of the intensity modulation using classical scattering theory. This synchrotron experiment has been explored on zeolite scolecite. It was unambiguously illustrated that aluminum preferentially occupies a single T site. Simulations showed that there are no fundamental reasons, why such measurement could not be performed on zeolite crystals with a structure as complex as that of ZSM-5 if it has silicon to aluminum ratio of about 25. The most important limitation is the required minimal crystal size, which depends on the state of the art equipment at the synchrotron beam line. Important additional information about the local structure surrounding aluminum sites has come from simulations [68–72]. The simulated annealing, the procedure of standard Monte Carlo ﬁnite temperature techniques, based on the statistical mechanics and combinatorial optimization [73], was used to study the aluminum distribution in zeolite omega [62]. The aluminum partitioning ratio between the two crystallographically unequivalent sites in the framework of zeolite

295

296

10 Aluminum in Zeolites: Where is it and What is its Structure?

omega was obtained from 27 Al MAS NMR spectra and ﬁtted by trial and error method to the numerically calculated values. It has been concluded that aluminum is non-randomly distributed in the framework, which is in a good agreement with the experimental results described above. The possible aluminum and proton sitting in the zeolite ITQ-7 was proposed, combining force ﬁeld atomistic simulations and FTIR experiments [74]. It was assumed that aluminum distribution is controlled by the energetics during the synthesis process and that the interaction between the structure-directing agent and the zeolite framework has to be involved in the model. Based on this assumption, variation of the structure-directing agent can be employed to control the distribution of framework negative charge in zeolites. Similar results were obtained for FAU and EMT zeolite structures using a combination of 29 Si MAS NMR and model generation by computer algorithm [6]. These ﬁndings were recently fortiﬁed by the experimental results [57, 58]. Recently, an alternative theoretical approach to determine the aluminum distribution in the zeolite framework was presented [68]. Experimentally accessible properties, such as adsorption and diffusion, are crucially dependent on the aluminum distribution and associated cation distribution and can be directly correlated to those. Molecular simulations have been performed to obtain the adsorption and diffusion properties of alkanes in a variety of zeolitic structures by varying the position of the aluminum atoms. Henry coefﬁcients of linear alkanes were computed for MOR, FER, and TON zeolite structures at a ﬁxed silicon to aluminum ratio by varying the aluminum distribution. Direct comparison with the experimentally obtained coefﬁcients allowed conclusions about preferential aluminum sitting. Nevertheless, theoretical approaches using ab initio calculations [75] or Monte Carlo simulations [76–79] face severe limitations. Small clusters used for ab initio computations cannot be used for description of the spatial distribution of aluminum. Monte Carlo simulations are based on the assumption that aluminum distribution in zeolites is random. Moreover, theoretical studies cannot reﬂect the potential dynamics effect occurring during zeolite synthesis. The overall conclusion is that the aluminum distribution in the zeolite framework is neither random, nor controlled by a simple rule, but that it depends on the conditions of the zeolite synthesis. The aluminum distribution is not dominated by their relative stability in the various sites, but by the conditions of the synthesis. Such prospect opens interesting possibilities to tune the catalytic performance in terms of activity and selectivity.

10.4 Summary and Outlook

Aluminum in zeolites is essential for their performance. This makes its analysis indispensable in terms of its structure, its position within a crystal, and its occupancy of crystallographic T sites. The structure of a zeolite is strongly dependent on the pretreatment as well as on the conditions of measurement. The zeolite

10.4 Summary and Outlook

structure is especially sensitive to moisture. At room temperature, part of the tetrahedrally coordinated aluminum in a proton-exchanged zeolite can be converted into octahedral aluminum, the amount of which depends on zeolite structure type and silicon to aluminum ratio. This change in coordination is completely reversible after treatment with a base, such as ammonia. Zeolite activation by steam causes large structural changes. The amount and kind of aluminum species depend on the extent of hydration and temperature. Hydration increases the amount of high-coordinated aluminum. Dehydration and/or increasing the temperature convert part of the octahedral coordination into a tetrahedral one. Moreover, at temperature higher than 400 ◦ C, three-coordinated aluminum forms. These defect sites and the framework-associated octahedrally coordinated aluminum make the zeolite less stable and more sensitive to further heating. This sensitivity of aluminum species to different pretreatment conditions makes it difﬁcult to understand their inﬂuence on catalytic performance. To determine structure–selectivity relationship for aluminum species in catalytic processes, in situ characterization techniques are required. Aluminum zoning has been widely reported. Its extent mainly depends on the synthesis conditions. Using tetrapropylammonium ions in the synthesis of ZSM-5 produces a silicon-rich core and aluminum-rich outer rim. Other templates might provide different distributions, however. This enables control of the distribution in terms of zoning of the aluminum. For the characterization of aluminum zoning, the choice of the right technique is crucial. Large, but incomplete, progress has been obtained in determining the aluminum and silicon occupancy of the crystallographic T sites. This distribution is not dominated by the thermodynamic stability of the T site occupying either aluminum or silicon, but it is determined by the synthesis conditions. NMR, X-ray scattering (X-ray standing waves), and computer simulation techniques are emerging that promises obtaining further insight. The prospect of being able to control the aluminum zoning as well as its occupancy over the crystallographic T sites opens exciting possibilities to synthesize zeolite material with unique properties and unprecedented activity and high selectivity. The structure of catalytically active sites, notably those of Lewis acid sites, and the location of aluminum in the crystalline framework remain two of the most pressing questions in structure-analysis in the ﬁeld of zeolites. The physical properties and catalytic behavior of aluminosilicates depend on the aluminum content, the structure of aluminum species, and aluminum distribution in the zeolite framework and over the crystal. The aluminum content can be unambiguously obtained from the elemental analysis, but it is still challenging to identify and characterize the structure of the aluminum species, especially under catalytically relevant conditions. Moreover, to analyze the aluminum zoning and foremost to clearly assign the aluminums to the speciﬁc T sites remains challenging. In the ﬁeld of NMR, there is a constant development of advanced techniques, such as two-dimensional multiple-quantum magic-angle spinning NMR, 27 Al magic-angle spinning/multiple-quantum MAS and 27 Al– 14 N transfer of population

297

298

10 Aluminum in Zeolites: Where is it and What is its Structure?

in double-resonance NMR [80], and 2D 27 Al →29 Si rotor-assisted population transfer and Carr-Purcell-Meiboom-Gill heteronuclear correlation NMR [81], which promise to being able to further clarify the structure of aluminum. As the aluminum coordination strongly varies with the pretreatment conditions, it is very important to characterize it in situ. In situ Al K edge (XANES) X-ray absorption near edge structure [20] and in situ NMR [82] have opened up this possibility. It is likely that the structure of aluminum in zeolites under pretreatment and reaction conditions will be unambiguously determined in the near future. Thus, the structure of the catalytically active site and its interaction with reactants and intermediates under reaction conditions will be accessible. The structure of Lewis acid sites requires particular attention and a combination of methods that enable to determine the Lewis acid character in terms of amount and strength of the aluminum species, such as infrared and NMR spectroscopy, are particularly promising. The knowledge of the aluminum distribution over the crystallographic T sites and thus the location of the catalytically active sites would allow exactly correlating the catalytic performance of aluminosilicates in terms of activity, selectivity, and stability. Numerous techniques have been applied or developed to characterize the aluminum distribution over the crystallographic T sites in a zeolite framework. The various methods based on NMR, X-ray standing waves, UV–Vis, theory, and so on, are likely to further develop and a combined approach is most promising.

Acknowledgment

We acknowledge the Swiss National Science Foundation for ﬁnancial support.

References 1. Barthomeuf, D. (1980) Stud. Surf. Sci. 2.

3.

4.

5. 6.

7.

Catal., 5, 55. Llewellyn, P.L., Grillet, Y., and Rouquerol, J. (1994) Langmuir, 10, 570. Omegna, A., van Bokhoven, J.A., and Prins, R. (2003) J. Phys. Chem. B, 107, 8854. Xu, B., Rotunno, F., Bordiga, S., Prins, R., and van Bokhoven, J.A. (2006) J. Catal., 241, 66. Szostak, R. (1991) Stud. Surf. Sci. Catal., 58, 153. Feijen, E.J.P., Lievens, J.L., Martens, J.A., Grobet, P.J., and Jacobs, P.A. (1996) J. Phys. Chem., 100, 4970. Dwyer, J., Fitch, F.R., Machado, F., Qin, G., Smyth, S.M., and Vickerman, J.C. (1981) J. Chem. Soc., Chem. Commun., 422.

8. Pfeifer, H., Freude, D., and Hunger, M.

(1985) Zeolites, 5, 274. 9. Anderson, M.W. and Klinowski, J.

(1989) Nature, 339, 200. 10. Ramdas, S. and Klinowski, J. (1984)

Nature, 308, 521. 11. Collignon, F., Jacobs, P.A., Grobet,

P., and Poncelet, G. (2001) J. Phys. Chem. B, 105, 6812. 12. Joyner, R.W., Smith, A.D., Stockenhuber, M., and van den Berg, M.W.E. (2004) Phys. Chem. Chem. Phys., 6, 5435. 13. van Bokhoven, J.A., Kunkeler, P.J., van Bekkum, H., and Koningsberger, D.C. (2002) J. Catal., 211, 540. 14. van Bokhoven, J.A., Sambe, H., Ramaker, D.E., and Koningsberger, D.C. (1999) J. Phys. Chem. B, 103, 7557.

References 15. Kraus, H., M¨ uller, M., Prins, R.,

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30. 31.

and Kentgens, A.P.M. (1998) J. Phys. Chem. B, 102, 3862. Jiao, J., Kanellopoulos, J., Wang, W., Ray, S.S., Foerster, H., Freude, D., and Hunger, M. (2005) Phys. Chem. Chem. Phys., 7, 3221. van Bokhoven, J.A., Roest, A.L., Koningsberger, D.C., Miller, J.T., Nachtegaal, G.H., and Kentgens, A.P.M. (2000) J. Phys. Chem. B, 104, 6743. van Bokhoven, J.A., van der Eerden, A.M.J., and Koningsberger, D.C. (2002) Stud. Surf. Sci. Catal., 142, 1885. Omegna, A., Prins, R., and van Bokhoven, J.A. (2005) J. Phys.Chem. B, 109, 9280. van Bokhoven, J.A., van der Eerden, A.M.J., and Koningsberger, D.C. (2003) J. Am. Chem. Soc., 125, 7435. Bourgeatlami, E., Massiani, P., Direnzo, F., Espiau, P., Fajula, F., and Courieres, T.D. (1991) Appl. Catal., 721, 139. van Bokhoven, J.A., Koningsberger, D.C., Kunkeler, P., van Bekkum, H., and Kentgens, A.P.M. (2000) J. Am. Chem. Soc., 122, 12842. Wouters, B.H., Chen, T.H., and Grobet, P.J. (1998) J. Am. Chem. Soc., 120, 11419. Katada, N., Kanai, T., and Niwa, M. (2004) Microporous Mesoporous Matter., 75, 61. Xu, B., Sievers, C., Hong, S.B., Prins, R., and van Bokhoven, J.A. (2006) J. Catal., 244, 163. Xu, B. (2007) Structure-performance relationships in solid-acid aluminosilicates, Doctoral thesis, ETH, 16922. Kung, H.H., Williams, B.A., Babitz, S.M., Miller, J.T., Haag, W.O., and Snurr, R.Q. (2000) Top. Catal, 10, 59. Beyerlein, R.A., Choi-Feng, C., Hall, J.B., Huggins, B.J., and Ray, G.J. (1997) Top. Catal, 4, 27. Zholobenko, V.L., Kustov, L.M., Kazansky, V.B., Loefﬂer, E., Lohse, U., and Oehlmann, G. (1991) Zeolites, 11, 132. Lukyanov, D.B. (1991) Zeolites, 11, 325. Williams, B.A., Babitz, S.M., Miller, J.T., Snurr, R.Q., and Kung, H.H. (1999) Appl. Catal. A, 177, 161.

32. Xu, B., Bordiga, S., Prins, R., and van

33.

34.

35. 36. 37.

38. 39. 40.

41. 42. 43.

44. 45.

46.

47. 48. 49. 50.

51.

Bokhoven, J.A. (2007) Appl. Catal. A, 333, 245. van Bokhoven, J.A., Lee, T.-L., Drakopoulos, M., Lamberti, C., Thiess, S., and Zegenhagen, J. (2008) Nat. Mater., 7, 551. Dedecek, J., Kaucky, D., and Wichterlova, B. (2001) Chem. Commun., 970. ˇ Gabova, V., Dedecek, J., and Cejka, J. (2003) Chem. Commun., 1196. von Ballmoos, R. and Meier, W.M. (1981) Nature, 289, 782. Derouane, E.G., Gilson, J.P., Gabelica, Z., Mousty-Desbuquoit, C., and Verbist, J. (1981) J. Catal., 71, 447. Suib, S.L. and Stucky, G.D. (1980) J. Catal., 65, 174. Chao, K.-J. and Chern, J.-Y. (1988) Zeolites, 8, 82. Sklenak, S., Dedecek, J., Lo, C., Wichterlova, B., Gabova, V., Sierka, M., and Sauer, J. (2009) Phys. Chem. Chem. Phys., 11, 1237. Debras, G., Gourgue, A., and Nagy, J.B. (1985) Zeolites, 5, 369. Lin, J.-C. and Chao, K.-J. (1986) J. Chem. Soc., Faraday Trans. 1, 82, 2645. Althoff, R., Schulz-Dobrick, B., Sch¨uth, F., and Unger, K. (1993) Microporous Mater., 1, 207. Dessau, R.M., Valyocsik, E.W., and Goeke, N.H. (1992) Zeolites, 12, 776. Hughes, A.E., Wilshier, K.G., Sexton, B.A., and Smart, P. (1983) J. Catal., 80, 221. Nagy, J.B., Bodart, P., Collette, H., El Hage-Al Asswad, J., and Gabelica, Z. (1988) Zeolites, 8, 209. Perez-Pariente, J., Martens, J.A., and Jacobs, P.A. (1986) Appl. Catal., 31, 35. Weeks, T.J. and Passoja, D.E. (1977) Clays Clay Miner., 25, 211. Corma, A., Melo, F.V., and Rawlence, D.J. (1990) Zeolites, 10, 690. Dwyer, J., Fitch, F.R., Qin, G., and Vickerman, J.C. (1982) J. Phys. Chem., 86, 4574. Groen, J.C., Bach, T., Ziese, U., Paulaime-van Donk, A.M., de Jong, K.P., Moulijn, J.A., and Perez-Ramirez, J. (2005) J. Am. Chem. Soc., 127, 10792.

299

300

10 Aluminum in Zeolites: Where is it and What is its Structure? 52. Abraham, A., Lee, S.H., Shin, C.H.,

53.

54. 55. 56.

57.

58.

59.

60. 61.

62.

63.

64.

65.

66.

Hong, S.B., Prins, R., and van Bokhoven, J.A. (2004) Phys. Chem. Chem. Phys., 6, 3031. Han, O.H., Kim, C.S., and Hong, S.B. (2002) Angew. Chem. Int. Ed. Engl., 41, 469. Lu, B., Kanai, T., Oumi, Y., and Sano, T.J. (2007) Porous Mater., 14, 89. Koranyj, T.I. and Nagy, J.B. (2007) J. Phys. Chem. C, 111, 2520. Sklenak, S., Dˇedeˇcek, J., Li, C., Wichterlov´a, B., G´abov´a, V., Sierka, M., and Sauer, J. (2007) Angew. Chem. Int. Ed. Engl., 46, 7286. Pinar, A.B., Marquez-Alvarez, C., Grande-Casas, M., and Perez-Pariente, J. (2009) J. Catal., 263, 258. Marquez-Alvarez, C., Pinar, A.B., Garcia, R., Grande-Casas, M., and Perez-Pariente, J. (2009) Top. Catal, 52, 1281. Dedecek, J., Kaucky, D., Wichterlova, B., and Gonsiorova, O. (2002) Phys. Chem. Chem. Phys., 4, 5406. Koranyj, T.I. and Nagy, J.B. (2005) J. Phys. Chem. B, 109, 15791. Bodart, P., Nagy, J.B., Debras, G., Gabelica, Z., and Jacobs, P.A. (1986) J. Phys. Chem., 90, 5183. Li, B., Sun, P., Jin, Q., Wang, J., and Ding, D. (1999) J. Mol. Catal. A, 148, 189. Kennedy, G.J., Afeworki, M., and Hong, S.B. (2002) Microporous Mesoporous Mater., 52, 55. Smith, J.V. (1971) in Molecular Sieve Zeolites-1 (eds E.M.Flaningen and L.B. Sand), American Chemical Society, Washington, DC, p. 171. Sarv, P., Fernandez, C., Amoureux, J.-P., and Keskinen, K. (1996) J. Phys. Chem., 100, 19223. Jarman, R.H., Jacobson, A.J., and Melchior, M.T. (1984) J. Phys. Chem., 88, 5748.

67. Raatz, F., Roussel, J.C., Cantiani, R.,

68.

69.

70. 71. 72.

73. 74. 75. 76. 77. 78. 79. 80.

81.

82.

Ferre, G., and Nagy, J.B. (1987) in Innovation in Zeolite Materials Science (eds P.J.Gorbet, W. Morties, W.J. Vansant and E.F. Schultz-Ekloff), Elsevier Science Publishers B.V., Amsterdam, p. 301. Garcia-Perez, E., Dubbeldam, D., Liu, B., Smit, B., and Calero, S. (2007) Angew. Chem. Int. Ed. Engl., 46, 276. (a) Ehresmann, J.O., Wang, W., Herreros, B., Luigi, D.P., Venkatraman, T.N., Song, W.G., Nicholas, J.B., and Haw, J.F. (2002) J. Am. Chem. Soc., 124, 10868; (b) Brandle, M., Sauer, J., Dovesi, R., and Harrison, N.M. (1998) J. Chem. Phys., 109, 10379. Eichler, U., Brandle, M., and Sauer, J. (1997) J. Phys. Chem. B, 101, 10035. Beerdsen, E., Smit, B., and Calero, S. (2002) J. Phys. Chem. B, 106, 10659. Beerdsen, E., Dubbledam, D., Smit, B., Vlugt, T.J.H., and Calero, S. (2003) J. Phys. Chem. B, 107, 12088. Kirkpatrick, S., Gelatt, C.D., and Vecchi, M.P. Jr. (1983) Science, 220, 671. Sastre, G., Fornes, V., and Corma, A. (2002) J. Phys. Chem. B, 106, 701. Derouane, E.G. and Fripiat, J.G. (1985) Zeolites, 5, 165. Takaishi, T. and Kato, M. (1995) Zeolites, 15, 689. Takaishi, T., Kato, M., and Itabashi, K. (1995) Zeolites, 15, 21. Feng, X. and Hall, W.K. (1997) Catal. Lett., 46, 11. Rice, M.J., Chakraborty, A.K., and Bell, A.T. (1999) J. Catal., 186, 222. Abraham, A., Prins, R., van Bokhoven, J.A., van Eck, E.R.H., and Kentgens, A.P.M. (2009) Solid State Nucl. Magn. Reson., 35, 61. Kennedy, G., Wiench, J.W., and Pruski, M. (2008) Solid State Nucl. Magn. Reson., 33, 76. Hunger, M. (2004) Catal. Today, 97, 3.

301

11 Theoretical Chemistry of Zeolite Reactivity Evgeny A. Pidko and Rutger A. van Santen

11.1 Introduction

Nowadays, advances in chemical theory rely to a signiﬁcant extent on computations. Novel theoretical concepts arising directly from experiments usually require further investigations using computational modeling. This is necessary to develop a conclusive molecular-level picture of the observed phenomenon. As a result, computational methods are nowadays widely and extensively applied in the chemical, physical, biomedical, and engineering sciences. They are used in assisting the interpretation of experimental data and increasingly in predicting the properties and behavior of matter at the atomic level. Computational simulation techniques can be divided into two very broad categories. The ﬁrst is based on the use of interatomic potentials (force ﬁelds). These methods are usually empirical and do not consider explicitly any electron in the system. The system of interest is described with functions (normally analytical), which expresses its energy as a function of nuclear coordinates. These are then used to calculate structures and energies by means of minimization methods, to calculate ensemble averages using Monte Carlo simulations, or to model dynamic processes via molecular dynamics simulations using classical Newton’s law of motion. The current capabilities and challenges for the corresponding computational methods have been recently reviewed in excellent papers by Woodley and Catlow [1] and by Smit and Maesen [2, 3]. The second category includes quantum chemical methods based on the calculation of the electronic structure of the system. Such methods are particularly important for processes that depend on bond breaking or making, which include, of course, catalytic reactions. Hartree Fock (HF), Density Functional Theory (DFT), and post-HF ab initio approaches have been used in modeling zeolites, although DFT methods have predominated in recent applications. This chapter focuses on illustrating the power and capabilities of modern quantum chemical techniques, and aims at highlighting the key areas and challenges Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

302

11 Theoretical Chemistry of Zeolite Reactivity

in theoretical chemistry of zeolites and microporous materials. The text is organized as follows. First, we brieﬂy discuss the methodological aspects of quantum chemical modeling of zeolites (Section 11.2). This will involve a brief review of the capabilities and limitations of the traditional electronic structure methods and structural models used for studying microporous materials and their reactivity. Section 11.3 will illustrate the current advances in quantum chemical calculations of zeolite-catalyzed transformations of hydrocarbons. The approaches toward improvement of DFT with respect to a better description of weak van der Waals interactions will be discussed. In the next Section 11.4, the methanol-to-oleﬁn (MTO), which is an industrially important process will be used to illustrate the power of computational methods in revealing the molecular mechanism of such complex catalytic reaction. Sections 11.5 and 11.6 will show how novel reactivity and structural concepts can be derived from theoretical calculations to rationalize the experimental observations. In these sections, we will discuss recent chemical concepts of the conﬁnement-induced reactivity of microporous materials (Section 11.5) and nonlocalized charge compensation of extra-framework cations in zeolites (Section 11.6). The latter section also illustrates recent progress in revealing local structure of zeolites from ab initio calculations. The text is then concluded with an outlook on future challenges of computational chemistry of zeolites.

11.2 Methodology

Quantum chemical electronic structure methods have a long history of applications to investigation of structural and reactivity properties of zeolites. In general, the goal of quantum chemical methods is to predict the structure, energy, and properties of a system containing many particles. The energy of the system is expressed as a direct function of the exact position of all atoms and forces that act upon the electrons and the nuclei of each atom. The exact quantum mechanical calculation of electronic structure is possible only for a very limited number of systems and thus, a number of simplifying approximations resulting in different quantum chemical methods are required to solve it for larger systems. Below, we present a simple overview of the important limitations of various quantum chemical methods as applied to the zeolite chemistry. More detail and in-depth discussion on the electronic structure calculations can be found in a number of very good references [4–9]. Electronic structure methods can be categorized as ab initio wavefunctions-based methods, ab initio density functional or semiempirical methods. All of them can be applied for the solution of different problems in zeolite chemistry. Wavefunction methods start with the HF solution and have well-prescribed methods that can be used to increase its accuracy. One of the deﬁciencies of the HF theory is that it does not treat dynamic electron correlation, which refers to the fact that the motion of electrons is correlated so as to avoid one another. The neglect of this effect can cause very serious errors in the calculated energies, geometries, vibrational, and other molecular properties.

11.2 Methodology

11.2.1 Ab initio Methods

There are numerous so-called post-HF methods for treating correlated motion between electrons. One of the most widely used approaches is based on the deﬁnition of the correlation energy as a perturbation. In other words, the conﬁgurational interactions are treated as small perturbations to the Hamiltonian. Using this expansion the HF energy is equal to the sum of the zero and ﬁrst order terms, whereas the correlation energy appears only as a second order term. The second order Møller–Plesset perturbation theory (MP2) typically recovers 80–90% of the correlation energy, while MP4 provides a reliably accurate solution to most system. Another approach to treat the electron correlation is the conﬁguration interactions method (CI). The general solution strategy for it is to construct a trial wavefunction that is comprised of a linear combination of the ground-state wavefunctions and excited-state wavefunctions. The trial wavefunction can include the exchange of one, two, or three electrons from the valence band into unoccupied orbitals; these are known as conﬁguration interaction singles (CIS), conﬁguration interaction doubles (CID), and CI triples and allow for single, single/double, and single/double/triple excitations. Coupled cluster (CC) methods differ from perturbation theory in that they include speciﬁc corrections to the wavefunction for a particular type to an inﬁnite order. CC theory therefore must be truncated. The lowest level of truncation is usually at double excitations (CCSD) since the single excitations do not extend the HF solution. CC theory can improve the accuracy for thermochemical calculations to within 1 kcal mol−1 . Despite using these methods a very high accuracy can be achieved, almost all of the post-HF methods are prohibitive for calculation of reliable models of zeolites due to very high computational costs. Only the computationally cheapest post-HF methods can be currently used for studying zeolites. Thus, the application of the post-HF methods in zeolite chemistry is limited to the MP2 method, which still can be applied to calculations of only rather small zeolite models. One notes however that when the resolution of identity (RI) approximation is used [10], the resulting RI–MP2 method can in principle be used for calculations of systems containing more than hundred of atoms. 11.2.2 DFT Methods

A more attractive method is DFT. DFT is ‘‘ab initio’’ in the sense that it is derived from the ﬁrst principles and does not usually require adjustable parameters. These methods formally scale with increase in the number of basis functions (electrons) as N3 , and thus, permit more realistic models compared to the higher-level post-HF methods, which usually scale as N5 for MP2 and up to N7 for MP4 and CCSD(T). On the other hand, the theoretical accuracy of DFT is not as high as the higher-level ab initio wavefunction methods.

303

304

11 Theoretical Chemistry of Zeolite Reactivity

The DFT is attributed to the work of Hohenberg and Kohn [11], who formally proved that the ground-state energy for a system is a unique functional of its electron density. Kohn and Sham [12] extended the theory to practice by showing how the energy can be portioned into kinetic energy of the motion of the electrons, potential energy of the nuclear–electron attraction, electron–electron repulsion, which involves with Coulomb as well as self interactions, and exchange correlation, which covers all other electron–electron interactions. The energy of an N-particle system can then be written as E[ρ] = T[ρ] + U[ρ] + Exc [ρ]

(11.1)

Kohn and Sham demonstrated that the N-particle system can be written as a set of n-electron problems (similar to the molecular orbitals in wavefunction methods) that could be solved self-consistently [12]. Although the DFT is in principle an exact approach, unfortunately, because the exact expression for the electron density functional as well as for the exchange correlation energy are not known, lots of assumptions and approximations are usually made. The most basic one is the local density approximation (LDA). It assumes that the exchange–correlation per electron is equivalent to that in a homogeneous electron gas, which has the same electron density at a speciﬁc point r. The LDA is obviously an oversimpliﬁcation of the actual density distribution and usually leads to overestimation of calculated bond and binding energies. The nonlocal gradient corrections to LDA functional improve the description of electron density. In this case, the correlation and exchange energies are functionals of both the density and its gradient. The gradient corrections take on various different functionals such as B88 [13], PW91 [14], PBE [15], for example. However, the accuracy of those is typically less than what can be expected from high-level ab initio methods. One notes that the HF theory provides a more exact match of the exchange energy for single determinant systems. Thus, numerous hybrid functionals have been developed, where the exchange functional is a linear combination of the exact HF exchange and the correlation (and exchange) calculated from pure DFT. The geometry and energetics calculated within this approach (B3LYP and B3PW[16], MPW1PW91 [17], PBE0 functional [18], etc.) are usually in a good agreement with the experimental results and with those obtained using the post-HF methods. However, hybrid functionals still fail to describe chemical effects mainly based on electron–electron correlation such as dispersion and other weak interactions [19, 20]. 11.2.3 Basis Sets

As it was mentioned above, the energy in the DFT methods is formally a function of the electron density. However, in practice the density of the system ρ(r) is written as a sum of squares of the Kohn–Sham orbitals: (11.2) ρ(r) = |ψi (r)|2

11.2 Methodology

This leads to another approximation usually done both in DFT and wavefunction-based methods. It consists in representation of each molecular orbital by a speciﬁc orthonormal basis set. The true electronic structure of a system can be in principle mathematically represented by an inﬁnite number of basis functions. However, due to computational limitations, these functions are truncated in practice and described by a ﬁnite number of basis sets resulting in some potential loss in accuracy. A wide range of different basis sets currently exists and the choice of a certain one strongly depends on the solution method used, the type of the problem considered, and the accuracy required in each particular case. These functions can take on one of several forms. The most commonly used approaches use either a linear combination of local atomic orbitals, usually represented by Gaussian-type functions (GTO), or a linear combination of plane-waves (PWs) as basis sets. GTO basis sets are widely used in calculations of molecular systems, which also include cluster models of zeolites discussed in more detail in Section 11.2.4 of this chapter. GTO basis sets are implemented in various available quantum chemical softwares (Gamess-UK [21], Gaussian 03 [22], Tubomole [23], etc.). Since many years, they have been used in implementation of HF- and post-HF methods, as well as in DFT. PWs are more popular in simulations of solids (e.g., zeolite crystals). This is mainly due to the fact that their application to periodic systems is straightforward. The corresponding calculations are faster both for computing energies and gradients as compared to the approaches employing GTO basis sets. As a result, the PW approach is widely used in computer programs such as CASTEP [24], CPMD [25], and VASP [26], and so on, which have found a wide range of applications in studying various periodic systems usually by means of ‘‘pure’’ DFT (without exact HF exchange). In addition to the PW codes developed to model condensed phase systems, the CRYSTAL 06 [27] program that utilizes GTO basis sets can be used for studying both periodic and molecular systems within the same formalism irrespectively of the dimensionality of the system. It is important to note that in general, when a sufﬁciently large number of basis functions are employed, the computational results obtained using either GTO or PW basis sets are essentially the same [28]. When the PW basis set expansion of the electronic wavefunction is used, the number of PW components needed to correctly describe the behavior of the wavefunction near the nucleus is prohibitively large. To address this problem, within the PW approach the core electrons are described using the pseudopotential approximation. In this case, it is assumed that the core electrons do not signiﬁcantly inﬂuence the electronic structure and properties of atoms and therefore, the ionic potential that arises from the nuclear charge and frozen core electron density is replaced by an effective pseudopotential. Although within the GTO approach, core electrons can be treated explicitly, the pseudopotential approximation can also be employed to reduce number of basis functions in calculations without dramatic loss of accuracy. This is in particular useful for the description of heavy atoms.

305

306

11 Theoretical Chemistry of Zeolite Reactivity

11.2.4 Zeolite Models

The approximations done in order to compute energies by different quantum chemical methods as well as the use of ﬁnite basis set for the description of molecular orbitals are not the only factors leading to limited accuracy. When modeling zeolites, one can seldom take into account all of the atoms of the system (Figure 11.1a,b). Typically, a limited subset of the atoms of the zeolite is used to construct an atomistic model. The size of the model describing the reaction environment can be critical for obtaining the reliable results. Indeed, although the coordination of small molecules such as CO, CH4 , H2 O to isolated cations can be studied very accurately using even CC theory, obviously the results thus obtained can hardly represent adsorption to the exchangeable cations in zeolites. This is an example of a competition of ‘‘model’’ versus ‘‘method’’ accuracy. The current progress in computational chemistry also made it possible to use rather efﬁciently periodic boundary conditions in DFT calculations of solids. This

SIII

SII

(a)

(b)

M2+

(d)

Fd 3m 576 atoms 192T atoms (c)

144 atoms 48T atoms

M2+

M2+

16T cluster M+ M2+

(e)

6T cluster

(f)

3T embedded cluster

Figure 11.1 Structural models of zeolite with faujasite topology (a) crystallographic unit cell with the Fd3m symmetry (b), a smaller rhombohedral lower-symmetry unit cell (c), cluster models containing 16 (d) and 6 (e) T-atoms representing, respectively,

(g)

12T embedded cluster

local structure of two and one SII faujasite sites with exchangeable cations stabilized at them, parts (f) and (g) show 3T and 12T clusters representing cation sites SIII and SII , respectively, embedded into the rhombohedral faujasite unit cell.

11.3 Activation of Hydrocarbons in Zeolites: The Role of Dispersion Interactions

allows theoretical DFT studies of structure and properties of some zeolites usually with relatively small unit cells using a real crystal structure as a model (Figure 11.1c). Such periodic DFT calculations of zeolites are, however, mostly limited to the use of LDA and GGA density functionals. Another method for the zeolite modeling is a so-called cluster approach. Here, only a part of the zeolite framework, containing ﬁnite number of atoms, is considered, whereas the inﬂuence of the rest atoms of the zeolite lattice is neglected (Figure 11.1d,e). A minimum requirement to the zeolite model is that it involves the reactive site or the adsorption site together with its environment. In this case, the model is built of several TO4 units to mimic the local structure of a part of the zeolite. Although this approach results in some loss of ‘‘model’’ accuracy, it can be very useful for the analysis of different local properties of zeolites such as elementary reaction steps, adsorption, and so on. In addition, in the case of cluster modeling the higher-level ab initio methods as well as hybrid density functionals can be successfully used. Hybrid quantum chemical embedding schemes form a wide range of popular methods for computational modeling of zeolites. They allow the combination of two or more computational techniques in one calculation and make it possible to investigate the chemistry of such systems like zeolites with high precision. The region of the system where the chemical process takes place (similar to that used for cluster modeling) is treated with an appropriately accurate method, while the remainder of the system is treated at a lower level of theory (Figure 11.1f,g). The main difﬁculty comes in linking two different regions together. The link region is usually deﬁned in order to provide an adequate transfer of information between the inner and outer regions. The energy for this system is then calculated as Ehybrid (System) = Ehigh (Model) + Elow (System) − Elow (Model)

(11.3)

where Ehigh (Model) refers to the energy calculated at higher level for the inner core region only. Elow (System) – Elow (Model) refers to the difference in energy between the full system and the core region both calculated at the low level of theory. Although usually quantum chemical methods (QM) are used for the description of the core region, while the rest atoms of the system are treated by molecular mechanics, it can be any lower level method which is faster than that used for the core. Therefore, using cluster embedding one can rather accurately investigate the local chemistry of zeolites and also take into account the longer range effects.

11.3 Activation of Hydrocarbons in Zeolites: The Role of Dispersion Interactions

One of the main challenges in modeling of chemical reactions in zeolites is to predict accurately energies of adsorption and reaction proﬁles for hydrocarbon transformations in microporous materials. The current method of choice for modeling reactivity of such complex catalytic materials as zeolites is DFT. However,

307

308

11 Theoretical Chemistry of Zeolite Reactivity

commonly used density functionals fail to describe correctly the long-range dispersion interactions [20]. The dominant interactions between the hydrocarbon species and the zeolite walls correspond to weak van der Waals interaction of dispersive nature, which therefore cannot be correctly computed within the conventional DFT. This may result not only in inaccurate computed energetics of chemical reactions but also in wrong prediction of stability or reactivity trends for systems, where the impact of dispersive interactions on the total stabilization energy of the reaction intermediates and transition states is not uniform along the reaction coordinates. Note that dispersion is an intermolecular correlation effect. As it has been mentioned above, the simplest electronic structure method that explicitly describes electron correlation is MP2 theory. However, MP2 calculations for periodic systems are presently feasible only for very small unit cells containing only few atoms in combination with small basis sets. Recently, an embedding scheme to introduce local corrections at post-HF level to DFT calculations on a periodic zeolite model has been proposed [29, 30]. This approach allows an accurate modeling of structural and electrostatic properties of the zeolite reaction environment by using the periodic DFT calculations. The reﬁnement for the self-interaction effects and van der Waals interactions between the adsorbed reactants and the zeolite walls is achieved by applying resolution of identity implementation of the MP2 method (RI–MP2) to a cluster model representing the essential part of the framework that is embedded into the periodic model of the zeolite. The thus designed MP2:DFT approach is suited for studying reactions between small or medium-sized substrate molecules and very large chemical systems as zeolite crystals and allows quantitative computing reaction energy proﬁles for transformations of hydrocarbons in microporous matrices with near chemical accuracy. To illustrate this, let us consider interaction of isobutene with Brønsted acid site of a zeolite (Scheme 11.1) as a prototype of Brønsted acid-catalyzed transformations of hydrocarbons. This reaction is not only interesting from the practical point of view due to its relation to the skeletal isomerization of butenes [31]. It also attracts attention of many theoretical groups because of the fundamental question of whether it is possible to form and stabilize the tert-butyl carbenium ion in zeolite microporous matrix as a reaction intermediate [32]. There are several studies that report DFT calculations on protonation of isobutene employing rather small cluster models to mimic local surrounding Brønsted acid site in a zeolite (see e.g., [33, 34]). A very strong dependency of the relative stabilities of the protonated products on the level of computations and more importantly on the size of the cluster model was found. The only minima on the potential energy surface obtained within the cluster modeling approach corresponds to the covalently bound alkoxides, while the carbenium ions are present as very short-lived transition states. On the other hand, when the long-range interactions with the zeolite framework and its structural details are explicitly included in the computation either within the embedded cluster approach with a very large part of the zeolite lattice as a low-level model [35] or by using periodic DFT [36, 37], a local minimum on the

11.3 Activation of Hydrocarbons in Zeolites: The Role of Dispersion Interactions

CH3 H3C O

Si CH3 H2C

CH3 +

Si

H O

Si Al

O

Si

H O

CH3 Al

O

Si

p-complex (1)

Al

O

(2) Si

t-Bu carbenium ion CH3 H3C C CH 3

CH3 H2C

CH3

Si

O

Al

O

Si

(3)

t-butoxide CH3 C H3C H CH2 Si

O

Al

O

(4) Si

i- butoxide Scheme 11.1

Protonation of isobutene on a zeolitic Brønsted acid site.

potential energy surface corresponding to the tert-butyl cation can be located for various zeolite topologies. Indeed, applying DFT under periodic boundary conditions to the realistic system containing isobutene adsorbed in ferrierite a rather different picture was observed [37] as compared to the situation when a small cluster model was used to mimic the zeolite active site [34]. Only the π-complex of butane (1) with the Brønsted acid site of the zeolite was found to be more stable than the isolated alkene separated from the zeolite [37]. The stabilization however was rather minor (PBE/PW, FERpbc , Table 11.1). The DFT-computed adsorption energies did not exceed a few kilojoules per mole that is much less than what would be expected for such a system. The existence of the local minimum on the potential energy surface corresponding to the tert-butyl carbocation (2) was reported. Its stability was shown to be at least comparable to that of covalently bound tertiary butoxide (3). Inclusion of zero-point vibrations and ﬁnite temperature effects further stabilized the carbenium ions relative to the covalently bound alkoxides. It was concluded that already at 120 K formation of tert-butyl cation in H-ferrierite becomes thermodynamically more favorable than formation of the covalently bound species [37]. However, this theoretical prediction lacks the experimental support, because simple carbenium ions have never been observed by either NMR or infrared spectroscopy upon oleﬁn adsorption to hydrogen forms of zeolites [38]. This inconsistency may not be ascribed to any deﬁciency of the zeolite model used in the computational studies, and therefore, must be due to the inaccuracies of the computational method (DFT) used either in respect to description of the self-interaction effect or dispersive interactions. Tuma and Sauer [30] computed the relative stabilities of the possible products of interaction of isobutene with H-ferrierite by means of the MP2:DFT hybrid method. A cluster model containing 16T atoms at the intersection of 8-membered

309

b Values

−13 57 10 10

−28 – −35 −54

−61 8 −67 −59

in parenthesis are taken from [37]. in parenthesis are corrected for BSSE.

a Values

(1) (2) (3) (4)

PBE/CBS, 16T [39]

B3LYP/ DZ, 3T [34]

M06–L/ CBS, 16T [39]

−63 41 −67 −67

MP2/CBS, 16T [39]

−49 – −62 −145

B3LYP: MM, FERpbc [40] −79 – −67 −94

MP2//B3 LYP: MM FERpbc [40] −16 (−10)a 8 (36)a 19 (17)a −3 (5)a

PBE/PW, FERpbc [30]

−92 −67 −78 −94

PBE+D, FERpbc [41]

−77 (−44)b −13 (−8)b −66 (20)b −80 (−27)b

MP2:DFT FERpbc [30]

−78 −21 −48 −73

Best estimate FERpbc [30]

Calculated reaction energies ( E, kJ mol−1 ) for the formation of the π complex of isobutene, of the tert-butyl carbenium ion, and of the tert-butoxide and iso-butoxide in acidic zeolites.

Table 11.1

310

11 Theoretical Chemistry of Zeolite Reactivity

11.3 Activation of Hydrocarbons in Zeolites: The Role of Dispersion Interactions

and 10-membered channels of H-FER including the Brønsted acid site was deﬁned for the MP2 level within the full periodic model that is in turn was described within DFT. The MP2 calculations on cluster models were performed using local basis set constructed from Gaussian functions. To avoid possible errors associated with the use of the limited-sized localized basis sets, the computational results were corrected for basis set superposition error (BSSE) and extrapolated to the complete basis set (CBS) limit. Furthermore, the results obtained within the embedded cluster approach were extrapolated to inﬁnite cluster size (i.e., to the periodic limit). To access the reliability of the chosen theoretical methods for hydrocarbon reactions in zeolites, comparison with the results of CCSD(T) calculations was performed. It was concluded that the MP2 method allows chemically accurate description of the system considered. Indeed, it was clearly shown that the post-HF corrections to the reaction energy proﬁles obtained by pure DFT (PBE exchange–correlation functional) are substantial (Table 11.1). More importantly, they are not uniform for different structures formed within the zeolite pores. When dispersion is included at the MP2 level, the adsorption energy of isobutene changes from −16 kJ mol−1 to the realistic value of −78 kJ mol−1 . Stabilization of the covalently bound alkoxides due to van der Waals interaction with the zeolite walls is even larger (best estimate, Table 11.1). Surprisingly, it was found that the impact of dispersion interactions on the stabilities of the protonated species is the lowest for the tert-butyl carbenium ion. The corresponding reaction energy is lowered only by 30 kJ mol−1 . As a result, the carbenium ion structure was shown to be the least stable species among the structures considered, whereas periodic PBE calculations predict this species to be only 15 kJ mol−1 less stable than the iso-butoxide species. When dispersion is included this energy gap becomes three times larger and reaches 52 kJ mol−1 [30]. Unfortunately, despite all the efforts made to reach high computational accuracy, the reported MP2:DFT results were not corrected for the ﬁnite temperature effects. Therefore, no deﬁnite conclusion can be made on the relative stabilities of the species formed upon isobutene protonation in ferrierite. Nevertheless, this large energy difference suggests that although there is a chance that at higher temperatures the equilibrium will shift toward the formation of tert-butyl carbenium ion, at ambient temperatures in line with the experimental observations the formation of covalently bound alkoxide species is preferred. Very recently, the applications of the MP2:DFT approach to computational studies of reactivity of zeolites have been extended to the investigation of methylation of small alkenes with methanol over zeolite HZSM-5 [42]. Besides the highly sophisticated theoretical methods employed in this study, to the best of our knowledge this is the ﬁrst ab initio periodic study of reactivity of microporous materials with such a complex structure as MFI. Similarly to the above-considered protonation of oleﬁns, the zeolite framework has been represented by a periodically repeated MFI unit cell, while the interactions due to the conﬁnement of the reacting species in the microporous space and the energetics of their transformations at the Brønsted acid site have been reﬁned by applying the MP2 correction to a cluster model embedded in the periodic structure (Figure 11.2).

311

312

11 Theoretical Chemistry of Zeolite Reactivity

m

O

CH3m

C4H8

Hm

Hz Al

c a

38T42H:MFIPBC Figure 11.2 Transition state structure for t-2-butene methylation with methanol over HZSM-5 zeolite. (a) Depicts the corresponding periodic MFI model shown along the straight channel. Highlighted atoms correspond to the largest 38T embedded cluster (enlarged in (b), boundary

38T42H H atoms are omitted for clarity) treated at the MP2 level of theory in [42] (created with permission using the supplementary materials provided with the [42], http://pubs.acs.org/doi/suppl/10.1021/ ja807695p).

The reaction chosen for the computational study is of high relevance to the industrially important MTO process. Reaction rates and activation barriers for the methylation of small alkenes over HZSM-5 are directly available from experimental studies [43, 44]. Thus, this reaction and the respective experimental data can be used to compare performance, accuracy, and predicting power of the currently widely used pure DFT methods and of the more advanced quantum chemical techniques (e.g., DFT+D and MP2 methods). A simpliﬁed schematic energy diagram [42] for this reaction is depicted in Figure 11.3. Several assumptions must be made at this step to compare the experimental and the computational results. The experimental kinetic studies indicate that oleﬁn methylation is a ﬁrst order reaction with respect to alkene concentration and zero order with respect to methanol concentration. The resulting experimental barriers [43, 44] represent apparent activation energies with respect to the state, in which methanol is adsorbed at the Brønsted acid site of a zeolite and alkene is in the gas-phase (Figure 11.3a). Secondly, the methylation reaction is assumed to take place via an associative one-step mechanism rather than a two-step consecutive process involving the formation of a methoxy group covalently bound to the zeolite walls. Although previous theoretical studies performed using a small 4T cluster model [45] have contributed signiﬁcantly to the molecular-level understanding of the mechanistic details of this catalytic process, the thus computed apparent activation barriers were signiﬁcantly overestimated and could not even reproduce the experimentally observed trends in their dependency on the alkene chain length (B3LYP4T and PBE4T in Figure 11.3b). The former effect is mainly caused by the well-known drawback of the cluster modeling approach that is mainly due to the lack of the electrostatic stabilization of the polar transition states by the zeolite

11.3 Activation of Hydrocarbons in Zeolites: The Role of Dispersion Interactions

313

200

Enthalpy

CH3OH (g) R = (g)

CH3OH (ads) R = (g)

Apparent activation energy H2O (g) CH3R = (g)

Eads CH3OH (ads) R = (ads)

(a)

H2O (ads) CH3R = (ads) Reaction coordinate

Apparent activation energy, kJ mol−1

Transition state 180 160 140 120 100 80 60 40 20 0 (b)

Ethene

Propene

B3LYP4T PBE4T PBE+D MP216T:DFTPBC

t-2-butene

PBEPBC DFT−EXP PBEPBC+∆Eads Best estimate Experimental

Figure 11.3 Simpliﬁed reaction energy diagram for an alkene methylation with methanol over acidic zeolite (a) and the respective apparent activation barriers computed using various methods (b) [42].

lattice [46–48]. Indeed, the apparent activation barriers calculated using a periodically repeated MFI unit cell decrease substantially (PBEPBC in Figure 11.3b). For ethane the calculated barrier is only 15 kJ mol−1 higher than that obtained from the experiment. Interestingly, this value corresponds almost exactly to the difference between the experimental and DFT-computed adsorption energies of ethylene on HZSM-5. Therefore, the absence of the hydrocarbon chain-length dependency of the apparent activation barriers calculated within DFT is primarily associated with its poor description of the dispersion effects. Indeed, the implication of the local post-HF correction within the MP2:DFT approach signiﬁcantly improves the qualitative picture, although the thus obtained results still deviate by 8–20 kJ mol−1 from the experimental values. This mismatch is further reduced after the extrapolation of the high-level correction results to the periodic and CBS limits (‘‘best estimate’’ in Figure 11.3b), while the subsequent corrections for ZPE and ﬁnite temperature effects allow reproduction of the experimental apparent enthalpy barriers with nearly chemical accuracy (deviation between 0 and 13 kJ mol−1 ). One notes that these deviations are contributed by uncertainties in both the computational and the experimental results. It has been convincingly shown [42] that the errors associated with various ﬁtting and modeling procedures used in the theoretical study are of the same order as the uncertainties in the energetics derived from the experimental data. This means that the MP2:DFT method by Sauer and coworkers [29, 30, 42]

314

11 Theoretical Chemistry of Zeolite Reactivity

allows to compute energy parameters for the reactions in zeolites that quantitatively agree with the experimental data. Nevertheless, although the proposed DFT:MP2 scheme allows the very accurate calculations of adsorption and reaction energies in microporous space, the associated computations are still too demanding to be used for comprehensive studies and for an in-depth theoretical analysis of various factors that inﬂuence the selectivity and reactivity patterns of the zeolite catalyst. The authors state in [42] that ‘‘the hybrid DFT:MP2 method is computationally expensive and not suited for routine studies on many systems.’’ Thus, there is still a strong desire for a robust computational tool aspiring to provide with reliable predictions for hydrocarbon transformations in zeolites that must combine efﬁciency and chemical accuracy of DFT methods along with the proper account for van der Waals dispersive interactions. This is reﬂected by the fact that the improvement of DFT toward a better description of nonbonding interactions is currently an active research area in theoretical chemistry. The most pragmatic solution for this problem is to involve in the calculations force ﬁelds based on the empirically ﬁtted interatomic potential. The state-of-the-art examples of those show extremely good results of quantitative quality for the prediction of structural properties of microporous materials and for the description of the processes that are mainly inﬂuenced by the nonbonding interactions [1–3]. The computational simplicity of the force ﬁeld approach allows simulations of even dynamical properties of chemical systems composed of more than 106 atoms at time scales up to nanoseconds. However, again due to the simplistic form of the interatomic potentials, they cannot be directly used to describe processes associated with bond breaking and making, that is, chemical reactivity. Thus, there are numerous approaches that in one way or another make use of the empirically derived nonbonding interatomic potentials combined with the electronic structure calculations to amend the results of DFT toward better description of van der Waals interactions. One can see from the results presented in Figure 11.3 that the periodic DFT calculations (DFTPBE ) can be improved substantially by adding the contribution from van der Waals interactions at the initial state obtained as a difference between the underestimated DFT-predicted adsorption energies and those obtained from DFT−EXP ). An associated computational procedure has been the experiment (Eads proposed by Demuth et al. [49] and Vos et al. [48]. It involves the correction of the periodic DFT results for van der Waals interactions using an add-on empirical 6–12 Lennard–Jones potential (Eq. (11.4)) acting between the atoms of the conﬁned hydrocarbon molecule and of the microporous matrix. The correction in this case is applied for the ﬁxed DFT optimized structures. Aij Bij − 6 (11.4) EvdW (rin ) = rij12 rij A similar m ethod that provides the possibility to optimize structures with inclusion of van der Waals interactions is the density functional theory plus damped dispersion (DFT+D) approach [50]. This scheme consists in adding a semiempirical

11.3 Activation of Hydrocarbons in Zeolites: The Role of Dispersion Interactions

term E(D) to the DFT energy E(DFT) resulting in the dispersion-corrected energy E(DFT+D). E(D) in this case is expressed as a sum over pairwise interatomic interactions computed using a force-ﬁeld-like potential truncated after the ﬁrst term (Eq. 11.5). cij E(D) = −s6 fD (rij ) (11.5) rij6 where cij are the dispersion coefﬁcients, the damping function fD (rij ) removes contributions for short-range interactions, while the global scaling parameter s6 depends on the particular choice of the exchange–correlation functional. The DFT+D approach has been parameterized for many atoms and a wide variety of functionals and can be used in a combination with popular quantum chemical programs [41]. When applied to chemical processes in microporous materials, this approach has been shown to provide realistic adsorption energies for hydrocarbons in all-silica zeolites [41]. Although the DFT+D signiﬁcantly improves the pure DFT results for the reaction energies (Table 11.1) and activation barriers (Figure 11.3b) for the conversion of hydrocarbons over acidic zeolites, these still signiﬁcantly deviate from the higher-level ab initio MP2:DFT or experimental result. The DFT+D apparent activation energies are systematically underestimated by ∼20–30 kJ mol−1 (Figure 11.3b), while the qualitative trend in the hydrocarbon chain dependency is perfectly predicted. On the other hand, the thus computed relative stabilities of the products of protonation of isobutene differ from the best estimate values from MP2:DFT substantially (Table 11.1). All of the above-considered computational techniques involve rather computationally demanding periodic DFT calculations of a large zeolite unit cells as the base for the geometry optimization of the structures of intermediates and transition states. Although the results thus obtained do not suffer from the artiﬁcial effects associated with the model accuracy, these methods may be unfeasible for such tasks as thorough computational screening of the catalytic performance of zeolite-based catalysts. In this case, the use of a hybrid QM:force ﬁeld (QM:MM) approach may help to reduce the associated computational requirements. This method may be viewed as a ‘‘lower-level’’ analog of the MP2:DFT approach. In this case, the ab initio part (usually treated by DFT methods) describing the bond rearrangement at the zeolite active site is intentionally limited to a small part of the zeolite, while the van der Waals and electrostatic interactions with the remaining zeolite lattice are described using a computationally less demanding force ﬁeld methods. This methodology allows fast and rather accurate calculation of the heats of adsorption and reaction energies of various hydrocarbons in zeolites [40, 51, 52]. However, when using the conventional DFT as the ‘‘high-level’’ method, the correct description of the dispersion contribution to the adsorption energy of longer-chain hydrocarbons requires the use of very small, usually containing only 3T atoms, cluster model [40]. The energetics can be substantially improved by correcting the DFT results by single-point MP2 calculations. The thus computed energetics (MP2//B3LYP : force ﬁeld) of isobutene protonation in H-FER zeolite agree reasonably well with those obtained using the MP2:DFT scheme (Table 11.1).

315

316

11 Theoretical Chemistry of Zeolite Reactivity

The performance of DFT itself may also be substantially improved by parameterization of the exchange–correlation functionals. Zhao and Truhlar have recently reported a family of meta-GGA functionals (M05 [53], M06 [54], and related functionals) in which the performance in describing nonbonding interactions (Table 11.1) as well as in predicting reaction energies and activation barriers is signiﬁcantly improved compared to the conventionally used GGA and hybrid functionals. The hybrid methods involving a combination of such density functionals and well-parameterized force ﬁelds are anticipated to be very efﬁcient and accurate for the investigations of zeolite-catalyzed reactions [39]. Nevertheless, the simpliﬁcations involved in the above methods, such as the assumption of pairwise additivity of van der Waals interactions, the presence of empirically ﬁtted parameters both in the force ﬁelds and in the parameterized density functionals can lead to unreliable results for systems dissimilar to the training set. The recently proposed nonlocal van der Waals density functional (vdW-DF) [55] is derived completely from ﬁrst principles. It describes dispersion in a general and seamless fashion, and predicts correctly its asymptotic behavior. Self-consistent implementations of this method both with PW [56] and Gaussian basis sets have been reported [57]. Until now, the vdW-DF method has been successfully applied to weakly bound molecular complexes, polymer crystals, and molecules adsorbed on surfaces (see e.g., [56, 57] and references therein). However, to the best of our knowledge, its applicability to modeling chemical reactions within zeolite pores has not been investigated yet. Summarizing, there is a strong desire for the efﬁcient and accurate computational tool for studying chemical reactivity of zeolites that is able to correctly predict effects due to nonbonding interactions in the microporous matrix. Most of the currently available computational techniques involve numerous approximations and often contain empirically ﬁtted parameters. In this respect, the hybrid MP2:DFT method by Tuma and Sauer [29, 30, 37] is useful to generate reliable datasets for various chemical processes in zeolite, on which parameterization of force ﬁelds, QM:MM methods, as well as assessment of the performance of various exchange–correlation functionals and various dispersion correction schemes can be based. The correct description of weak nonbonding interactions within the intermediates and transition states involved in catalytic conversions of hydrocarbons in zeolites is important not only for the fundamental understanding of these processes but also for the generating reliable microkinetic models able to predict the reactivity and selectivity patterns of microporous catalysts in various reactions.

11.4 Molecular-Level Understanding of Complex Catalytic Reactions: MTO Process

Molecular-level determination of the reaction mechanism of complex catalytic transformations in zeolites based solely on experimental studies is usually an extremely challenging task. Theoretical methods based on quantum chemical calculations are in contrary ideally suited for revealing the molecular mechanism

11.4 Molecular-Level Understanding of Complex Catalytic Reactions: MTO Process

and for identifying the elementary reaction steps of such processes. This section illustrates the recent advances in understanding the molecular-level picture of the industrially important MTO process from quantum chemical calculations. The MTO process catalyzed by acidic zeolites has been subject of extensive experimental studies driven by the possibility of converting almost any carbon-containing feedstock (i.e., natural gas, coal, biomass) into a crucial petrochemical feedstock like ethylene and propylene. The actual reaction mechanism of this process has been a topic of intense debates for the last 30 years [58, 59]. Initially, the research was focused on the formation of the ﬁrst C−C bond via the combination of two or more methanol molecules to produce alkene and water [58, 59]. Such a ‘‘direct’’ mechanism involved only methanol and C1 derivatives. An alternative mechanism has been suggested by Dahl and Kolboe [60] that assumed the formation of some ‘‘hydrocarbon pool’’ species that is continually adding and splitting reactants and products. Recently, both the experimental results by Haw et al. [61, 62] and quantum chemical studies by Lesthaeghe et al. [63, 64] provide evidence for the preference of the latter mechanism. Indeed, Lesthage et al. [63, 64] screened practically all of the possible direct C−C coupling reaction routes over a zeolitic Brønsted acid site modeled using a small 5T cluster at the B3LYP/6-31G(d) level of theory. The combination of the thus obtained results in complete pathways and calculation of barrier heights as well as the rate coefﬁcients clearly showed that there is no successful pathway leading to the formation of ethylene or any intermediate containing a C−C bond from only methanol. These results are in line with the experimental observations of the very low activity of methanol and DME (dimethyl ether) over HZSM-5 zeolite in the absence of organic impurities acting as a hydrocarbon pool species [61, 62]. It was concluded that the failure of the direct C−C coupling mechanism is mainly due to the low stability of the ylide intermediates and the highly activated nature of the concerted C−C bond formation and C−H bond breaking involved in these mechanisms. Both of these effects were attributed to the low basicity of the framework zeolite oxygens that cannot efﬁciently stabilize the respective species. The more likely pathway involves organic reaction centers trapped in the zeolite pores which act as cocatalysts. In particular, experimental studies have proved formation of various cyclic resonance-stabilized carbenium ions (Scheme 11.2) upon the MTO process in the microporous space. Stable dimethylcyclopentenyl (5) and pentamethylbenzenium (6) cations in HZSM-5 [59, 65] and hexamethylbenzenium (7) and heptamethylbenzenium (8) cations in zeolite HBEA [66, 67] have been detected by various spectroscopic methods. Obviously, the catalytic activity is therefore inﬂuenced both by the nature of the hydrocarbon pool species and by the

(5)

Scheme 11.2

(6)

(7)

(8)

Stable cyclic carbocations experimentally detected in zeolites.

317

11 Theoretical Chemistry of Zeolite Reactivity

318

topology of the zeolite framework that determines the preferred pathway for the transformation of these bulky species. An attempt to separate these effects was done by using quantum chemical calculations [68]. The geminal methylation of various methylbenzenes with methanol over a zeolitic Brønsted acid site (Figure 11.4a) was modeled with a 5T cluster model. Note that such a small cluster may be viewed as a general representation of any aluminosilicate. It does not mimic structural features of any particular zeolite and therefore completely neglects all of the effects due to the sterics and electrostatics of the microporous space. Although the activation energies computed were substantially overestimated, the obtained results indicate the increasing reactivity of larger polymethylbenzenes (Figure 11.4b). The effect of the zeolite topology has been included by extending the calculations to larger clusters containing 44T and 46T atoms, where the catalytically active site and the reactants were still modeled at DFT level using an embedded 5T cluster, while the remaining part of the cluster model was treated at HF level [68]. It has been concluded that the structural features of the zeolite framework play a major role in the reaction kinetics. The reaction rates for the geminal methylation of hexamethylbenzene follow the order: CHA >> MFI > BEA (Figure 11.4c). These striking differences in reactivity can be attributed to the molecular recognition features of the zeolite cages (Figure 11.4d–f). Indeed, the size and the shape of the chabazite cage were shown to be ideal for this reaction step. The larger pores of BEA + CH3OH

+ H2 O

H-zeolite

(a)

(b)

200 180 160 140 120 100 80 60 40 20 0

195 170

165

162 115 104

0

Toluene Durene HMB

Relative energy (∆E, kJ mol−1)

Relative energy (∆E, kJ mol−1)

(d) 180 160 140 120 100 80 60 40 20 0 −20

MFI

BFA MFI 144 CHA 126

(e) 61

55 CHA

29 0

(c)

Figure 11.4 Methylation of polymethylbenzenes in acidic zeolites (a) and the computed activation barriers and reaction enthalpies depending on the nature of the organic molecule (b) and on the zeolite topology (c). The molecular recognition effects are illustrated with the schematic

−8 (f)

representation of the resulting carbenium ions in the conﬁned space and with the structures of the transition states for methylation of HMB in zeolites BEA (d), MFI (e), and CHA (f). (Adapted from [68, 69] with the help of the authors.)

11.4 Molecular-Level Understanding of Complex Catalytic Reactions: MTO Process

beta zeolite (BEA) are not able to provide an effective electrostatic stabilization of the organic species. On the other hand, the small pores of HZSM-5 zeolite (MFI) restrict the formation of large carbocations, suggesting thus that the hydrocarbon pool in this case involves less bulky species. Combining the effects of such a transition state shape selectivity of MFI framework and those caused by the nature of the hydrocarbon pool species, the highest reactivity was predicted for the less sterically demanding 1,2,4-trimethylbenzene and pentamethylbenzene. This is in line with the experimental observations of Svelle et al. [70] on the higher reactivity of the smaller methylbenzenes in HZSM-5 zeolite. Finally, combining the theoretical and experimental results a complete catalytic cycle for the conversion of methanol to oleﬁns (MTO) in HZSM-5 zeolite (Figure 11.5) was presented by McCann et al. [71]. For each elementary reaction step, activation barriers and reaction rates were computed. This complete route convincingly explains the formation of the experimentally detected cations (5) and (6) all the way from toluene, multiple scrambling of labeled carbons from the methanol feed into the hydrocarbon pool species, as well as the substantial selectivity to isobutene during the MTO process in HZSM-5. Very recently, the mechanism of MTO reaction over the hexamethylbenzene species encapsulated in the chabazite cages of the aluminophosphate HSAPO-34 (CHA topology) has been thoroughly investigated by means of periodic DFT calculations [72]. The results obtained indicate that the catalytic reactivity of this system is to a signiﬁcant extent controlled by the intrinsic acidity of the zeolite catalysts. A plausible role of water is suggested, which can act as the promoter for the H+ shift between the conﬁned organic species and the zeolite framework to facilitate the elimination of alkenes. The computed apparent activation barrier of the rate-determining steps for the production of ethene was 230 kJ mol−1 and that for the production of propene was 206 kJ mol−1 . The authors thus concluded that the MTO process over the hexamethylbenzene/HSAPO-34 catalysts is selective toward propene. The reported activation energies seem to be signiﬁcantly overestimated. This can be due to either the actual choice of the hydrocarbon pool species or the deﬁciencies of the computational methodology used in this study. Note that the hydrocarbon pool mechanism of the MTO process in zeolites involves a tight conﬁnement of the hydrocarbon molecules in the microporous space. Their transformations during the catalytic process substantially alter their conformation, size, and length of the side-chains, and so on. Thus, the impact of intermolecular van der Waals interactions between the encapsulated organic molecules and the zeolite walls in different reaction intermediates and transition states on their stability may signiﬁcantly vary along the reaction coordinate. This in turn may have a great inﬂuence not only on the exact computed values of enthalpies and activation barriers of the elementary reaction steps but also on the qualitative trends in the energetics of the MTO catalytic process. These effects have been completely neglected in the computational studies discussed above and therefore the inﬂuence of dispersion on various elementary reaction steps of the MTO process in zeolites still has to be investigated.

319

Z-

ZZ-

3.92E+11 Z(24.3)

8T:46T 2-layer ONIOM cluster (B3LYP/6-31G(d):HF/6-31G(d) //(B3LYP/6-31G(d):MNDO)

Z-

C4Z8

Z-

Z-

Z-H

1.18E+2 (120.1)

1.71E+11 (14.6)

1.60E+0 (154.9)

H 3.05E+ 12(4.5)

Geminal methylation & deprotonation

1.11E+6 8.10E+8 Z-H (75.1) Z(54.7) Z-H CH3OH H 2O H2O

M2

9.47E+ 12(4.4) Z-H H2O

1.00E−1 s−1 < k < 1.00E+2 s−1 (slow) 1.00E+2 s−1 < k < 1.00E+5 s−1 1.00E+5 s−1 < k < 1.00E+8 s−1 (fast)

Rate coefficients at 673 K

9.55E+8 3.66E+8 Z-H Z(54.8) (47.8) CH3OH H 2O

H 1.26E+0 (151.9)

M3 Geminal methylation & deprotonation

Z-H

H 2O

CH3OH

Z-H

CH3OH H2O

M2 (slow)

Geminal methylation

H2O

1.30E−1 (162.5)

Geminal methylation & deprotonation H

Z-H 1.85E+7 Z- 7.57E+7 Z-H CH3OH (42.5) H2O (68.9) H2O

M1

H2O

M3 (slow)

M1 (slow)

CH3OH

M4

Z-H

1.59E+7 Z-H Z(62.0) CH3OH H 2O

M4

I3

CH3OH

Complete cycle

I1 (fast)

I2 (fast)

3.21E+8 Z-H (58.3)

Level of theory

Z-

4.93E+8 (51.3)

Ring contraction

ZC4Z8

1.21E+06 (94.5)

Z-

1.02E+6 (87.0)

Isobutene elimination

8.95E−3 (203.4)

1.00E+6 (94.2)

I2

1.58E+3 (126.2) Z-

6.05E+2 (112.6)

Ring expansion & deprotonation HH 4.37E+5 1.89E+3 (111.2) (95.2)

Figure 11.5 Full catalytic cycle for carbon-atom scrambling and isobutene formation from methanol through a combined methylbenzene/cyclopentenyl cation pool mechanism in HZSM-5. Calculated rate constants at 673 K are given in per second and reaction barriers at 0 K (in brackets) are given in kilojoules per mole. (Adapted from [71] with the help of the authors. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.)

I1

I3

320

11 Theoretical Chemistry of Zeolite Reactivity

11.5 Molecular Recognition and Conﬁnement-Driven Reactivity

Summarizing, the computational work performed until now has greatly improved our fundamental understanding of the elementary reaction steps involved to produce light alkenes from aromatic reaction centers during the MTO process over acidic zeolites. However, the thorough investigation of the stability and reactivity of other possible hydrocarbon pool species conﬁned within the pores of zeolite is needed to rationalize the experimentally observed selectivity patterns. The extension of the theoretical studies to other zeolite catalysts with different topologies and framework composition is also very important, because the resulting different geometrical constraints and different acidic properties will probably lead to different major catalytic cycles. Furthermore, the mechanism of the initial formation of the hydrocarbon pool cocatalysts in the zeolite pores, and related the question of how the ﬁrst C–C bond is formed have not been addressed yet. Rationalization of these issues is necessary for the creation of a complete molecular-level picture of the MTO process over acidic zeolites. The resulting fundamental understanding of various factors underlying the reactivity of zeolites in the MTO process will ultimately create a possibility to rationally design a catalyst with speciﬁc local spatial environment and framework composition toward the optimal catalytic performance and desired product selectivity.

11.5 Molecular Recognition and Conﬁnement-Driven Reactivity

The chemical reactivity of zeolites mainly arises from the presence of various ‘‘defect’’ sites in the ideal full-silica framework. Such defects can be either framework sites formed upon the isomorphous substitution of silicon with other ions (Sn4+ , Ti4+ , etc.) in the zeolite lattice or extra-framework cationic species that compensate for the negative charge on the framework due to the presence of lower-valent ions in the framework (Al3+ , Ga3+ , etc.). Depending on the type and the properties of such sites the chemistry of the resulting catalyst can vary dramatically. Besides the intrinsic chemical properties, steric effects in the form of shape selectivity are well known to be important for the reactivity of the zeolite-based materials [3, 73, 74]. Depending on the size of the zeolite cavities and channels only molecules below a certain size or of a speciﬁc shape are allowed to reach the active site or to leave the zeolite matrix. In addition, steric constraints imposed by the zeolite structure on a particular reaction intermediate or a transition state can deﬁne formation of certain product molecules. The effects of the zeolite topology on the relative stabilities of different reaction intermediates formed in the microporous space have been illustrated in the previous Sections 11.3 and 11.4. It should be noted that in all these cases, it was assumed that the chemical reaction is controlled by the zeolite topology and the intrinsic properties of the active species, whereas their intrazeolite arrangement is usually neglected. For the case of low-silica zeolites the latter factor can become crucial. The high density of the exchangeable ions in these materials can allow formation of multiple noncovalent interactions with the adsorbed molecules resulting in their

321

322

11 Theoretical Chemistry of Zeolite Reactivity

speciﬁc orientation and activation in the zeolite cage. This resembles the way enzymes activate substrates [75]. They use molecular recognition features to orient conﬁned molecules at the active site resulting in their speciﬁc chemical activation. Similar effects have been also reported for numerous examples in coordination chemistry and homogeneous catalysis [76]. This section illustrates the importance of molecular recognition features in zeolites for the adsorption properties and reactivity. The term molecular recognition is usually applied when selective binding and formation of speciﬁc activated complexes is a result of multiple attractive noncovalent interactions. Therefore, one expects the associated effects to have a signiﬁcant impact on the adsorption properties of microporous materials. Recently, adsorption of various probe molecules such as CO [77–79], CH3 CN [80], CO2 [81], and N2 O4 [82, 83] in zeolites has been investigated by periodic DFT calculations. The obtained results indicate that the vast majority of adsorption complexes in cation-exchanged Al-rich zeolites cannot be described in terms of the interactions with a speciﬁc type of single cation sites. Instead, they usually involve numerous intermolecular interactions between the conﬁned molecules and the exchangeable cations in the zeolite. Even for such a small molecule as CO, formation of the linearly bridged CO adsorption complexes were detected in Na- and K-containing ferrierites with Si/Al ratio of 8 using a combined computational and spectroscopic approach [78]. Interaction with dual adsorption sites was found to be slightly more favorable (∼5–10 kJ mol−1 ) as compared to coordination to a single exchangeable cation. Furthermore, very recently evidence for the formation of similar complexes in high-silica Na-ZSM-5 and K-ZSM-5 has been reported [79]. Systematically, lower CO adsorption energies for the potassium-containing zeolites have been reported. This is in line with the expected lower Lewis acidity of the larger alkaline cations. The probability of the speciﬁc interaction of CO with multiple cation sites was shown to depend on the Al content and hence on the density of the exchangeable cations in the microporous matrix, on their ionic radius and on the zeolite topology. Similarly, a preference for the formation of bridging adsorption complexes of CO2 in Na-FER zeolite with high concentration of the exchangeable cations was demonstrated by the combination of variable-temperature IR spectroscopy and periodic DFT calculations [81]. Nevertheless, it was concluded that the maximum loading of CO2 apparently does not depend on the concentration of Na+ in the ferrierite zeolite. Molecular adsorption of N2 O4 on Na-, K-, and Rb-containing zeolites Y and X was also investigated by means of periodic DFT calculations [82, 83]. Bonding within the adsorption complex formed upon adsorption of a nonpolar N2 O4 on alkali-exchanged faujasite is very weak and corresponds to the induced polarization of the adsorbed species by the extra-framework cations. Strength of such interactions correlates with the size of the alkali ion. The smaller the cation is, the stronger its polarizing ability and Lewis acidity are. As a result, similar to the cases of CO and H2 adsorption on alkali-exchanged zeolites [79, 84], one expects a decrease of the N2 O4 adsorption energies simultaneously with the increase of the ionic radius

11.5 Molecular Recognition and Conﬁnement-Driven Reactivity Adsorption energies (kJ mol−1 ) of N2 O4 molecules at SII and SIII cationsa in the cage of alkaline-exchanged zeolites X [82] and Y [83].

Table 11.2

Na+ K+ Rb+

SII /X

SIII /X

SII a /Y

SII b /Y

SIII /Y

−10 −10 −21

−11 −15 −26

−11 −11 −18

−11 −10 −16

−33 −16 −29

a The S and S II III sites of the faujasite lattice are illustrated in Figure 11.1a.

of the zeolitic cations. The computational results show, however, the opposite trend (Table 11.2). This phenomenon can be best rationalized with the example of lower-silica modiﬁcation of faujasite that is zeolite X [82]. A high density of the extra-framework cations in the cage of zeolite X allows formation of multiple interactions between the adsorbed N2 O4 molecule and the exchangeable ions (Figure 11.6a–c). The size of the intrazeolite cations strongly inﬂuences the probability of such multicentered binding. Indeed, in the case of NaX zeolite, the smaller sodium ions neighboring the primary adsorption site are located too far to form interatomic contacts of a notable strength with the adsorbed molecule and are able to only slightly polarize it. On the other hand, when the ionic radius of the cations increases numerous additional adsorbate–adsorbent interactions can be formed. Whereas, in the cases of NaX and KX (Figure 11.6a,b, respectively), one can distinguish the primary and the secondary interactions in the corresponding adsorption complexes, the interatomic distances for the respective contacts between the N2 O4 conﬁned in the cage of RbX with the accessible extra-framework Rb+ cations become very similar (Figure 11.6c). In the latter case, not only the exchangeable cations nearest to the main adsorption site but almost all Rb+ in the faujasite supercage interact with the adsorbed N2 O4 . As a result, despite the expected lower polarizing ability of the softer cations and hence the weaker individual interatomic contacts formed with them upon adsorption, the overall interaction energy increases along with the increase of the ionic radius of the exchangeable cations. Besides the higher probability for the formation of the multiple-centered adsorption complexes, the unexpected enhancement of the interaction energy with the larger exchangeable cations can be also contributed by the differences in the effective shielding of those when stabilized at cation sites of zeolites. Indeed, larger alkaline or alkali-earth cations (K+ , Rb+ , Ca2+ , etc.) at the SII sites (six-membered rings, Figure 11.1a) of faujasite are only slightly shielded by the surrounding framework oxygens. As a result, the effective polarization of the adsorbed molecules in their ﬁeld increases as compared to the cases of smaller exchangeable cations (Li+ , Mg2+ , etc.), which are strongly shielded at the SII positions of faujasite. This has

323

11 Theoretical Chemistry of Zeolite Reactivity

324

Molecular adsorption KX

RbX SIII

3.565

SII

72

24

SII

3.0

2.

80

2

SII

3.763

S

4.29 III 5.15 8 47 4.

SII

3.709

4.660 5.455 4 19 5.

3.296

2.795

SII

4.369

3.150

4.088

4 53 4.

36

3 5.

3.1

NaX

SIII

SIII

(b)

(a)

3.532 SII

SIII

(c)

Na

K

Rb

Disproportionation

RbX

KX

NaX

SIII 29 0 . NOd3− 3 3.53

2.

NOd3−

2

(d)

(e)

2.854

2

SII

2.029 d+

SIII

NO

(f)

94

9 2.

2.211

SIII

NO

2.164

2.3

2.190 2.425 d+ 57

2.23

SIII

3.054 2. 95 0

2.48

8

SII

3.054

NOd3− SII

83

3

SIII

2.948 1.999 d+

NO

SII N O Al Si

Figure 11.6 Molecular adsorption of N2 O4 (a–c) and forδ+ ion pair (d–f) due to N2 O4 dismation of NOδ− 3 · · · NO proportionation in alkali-exchanged zeolites X. (Adapted from [85].)

been recently demonstrated by a combination of IR spectroscopy and cluster ab initio calculations of adsorption of light alkanes on alkali-earth exchanged zeolite Y [86, 87]. Furthermore, it has been shown that the differences in the effective shielding of the exchangeable cations in zeolites can indirectly affect the preferred conformation in the resulting adsorption complexes via favoring additional secondary van der Waals interactions between the adsorbed molecules and the framework oxygens of the cation site. In the case of N2 O4 molecular adsorption, the combination of both effects due to the effective shielding and due to the formation of multiple interaction with the exchangeable cations results in the partial restoration of the expected trends in the adsorption energies when the primary adsorbate–adsorbent interaction takes place with the weakly shielded alkaline cation stabilized at the SIII site (four-membered ring, Figure 11.1a) of zeolite Y (Table 11.2). These effects are much more pronounced when polar species are formed in the cation-exchanged zeolite matrices. Disproportionation of N2 O4 in alkali-exchanged

11.5 Molecular Recognition and Conﬁnement-Driven Reactivity δ+ faujasites results in the formation of rather polar NOδ− species. Inter3 · · · NO action of the involved charged fragments with the corresponding counterion of the zeolite matrix facilitates the cleavage of the contact ion pair. Periodic DFT calculations [82, 83] show the dominant role of the exchangeable cations for the reactivity of alkaline zeolite in the N2 O4 disproportionation reaction. Stability of the reaction products was found to correlate very well with the size, intrazeolite arrangement, and mobility of the exchangeable cations. It was shown that in the case of Na-containing faujasites, the smaller size of the sodium cations and their limited mobility do not allow formation of the appropriate conﬁguration of the active site to stabilize the negatively charged NOδ− 3 (Figure 11.6d). In contrast, the larger and mobile Rb+ cations shape a perfect stabilizing environment for the nitro group (Figure 11.6f). Thus, it has been concluded that the molecular recognition features of the microporous matrix facilitate the charge separation upon the N2 O4 disproportionation in a fashion of a polar solvent. The cooperative effect of the extra-framework cations, their intrazeolite arrangement and mobility induced by adsorption, as well as the steric properties of the zeolite cage are crucial for the reactivity of the cation-exchanged zeolites in this reaction. Recently a concept of conﬁnement-induced reactivity, which is interrelated with the above-discussed molecular recognition features of the intrazeolite void space, has been put forward to rationalize the photo-catalytic activity of rather inert alkali-earth containing faujasites in oxidation of alkenes [85, 88]. The chemical reaction in this case is initiated by a visible-light induced electron transfer between the adsorbed oleﬁn and dioxigen. Experimental studies indicated a signiﬁcant decrease of the energy of such an electron excitation when the reagents are loaded into the zeolite matrix containing alkaline- or alkali-earth cations [89–91]. This effect has been initially attributed to the stabilization of the hydrocarbon·O2 charge-transfer state by the interaction of the strong electrostatic ﬁeld in the zeolite cage with the large dipole moment generated upon the excitation. In contrary, a theoretical analysis of the initial step of photochemical activation of 2,3-dimethyl-2-butene (DMB) and O2 coadsorbed in a 16T cluster model (Figure 11.1d) representing a part of Ca-, Mg-, and Sr-exchanged supercage of zeolite Y indicate that the effect of the electrostatic ﬁeld of the zeolite cavity is only minor for the reactivity of these microporous materials [88]. It has been proposed that the role of the zeolite in the photo-oxidation of alkenes with molecular oxygen is the complexation of the reagents to the extra-framework cations in a pre-transition state conﬁguration. This results in conﬁnement of these molecules with a speciﬁc orientation. The resulting formation of a π−π intermolecular complex (Figure 11.7) substantially increases the probability of the photo-initiation of the oxidation reaction. The relative orientation and the distance between the adsorbed reagents, which depend strongly on the size of the exchangeable cations in the adsorption site, are crucial for the chemical reactivity of these systems. It has been concluded that a high density, speciﬁc location, and size of the exchangeable cations in the zeolite result in molecular recognition of the adsorbed species and their chemical activation. Thus, molecular recognition features of the intrazeolite space are crucial both for the adsorption properties and for the chemical reactivity of the microporous

325

326

11 Theoretical Chemistry of Zeolite Reactivity

hn

HOMO

LUMO

Figure 11.7 Frontier molecular orbitals of the DMB · O2 adsorption complex in CaY involved in the intermolecular electron excitation. (Adapted from [85].)

materials. The presence of multiple interaction sites in the zeolite cage can control the preferred conformations of the adsorbed molecules and their chemical transformations. Further theoretical studies are needed to understand better the impact and the role of the multiple interaction sites in the conventional zeolite-catalyzed reactions. Indeed, the computational studies discussed so far suggest that there is a possibility of interaction between the distant active species and bulkier reaction intermediates formed in the zeolite channels. We anticipate that the reactivity and selectivity trends predicted by computational studies may alter substantially when a realistic chemical composition of the zeolite catalysts is considered.

11.6 Structural Properties of Zeolites: Framework Al Distribution and Structure and Charge Compensation of Extra-framework Cations

Catalytically active species, that is, protons, metal cations or more complex cationic aggregates, compensate for the negative charge of the microporous aluminosilicate framework. The crystallographic position of the corresponding negatively charged aluminum-occupied oxygen tertrahedra ([AlO4 ]− ) governs to a large extent the location of the active sites that in turn may have a signiﬁcant inﬂuence on the catalytic activity and selectivity [92, 93]. Thus, understanding the exact Al sitting and the factors which govern its distribution in zeolites is strongly desired. The related question is how the Al distribution inﬂuences structural and chemical properties of conﬁned cationic species as well as their location in the zeolite matrix. An in-depth discussion of the problem of Al sitting in zeolites is provided in Chapter 14 of this book. Therefore, in this section we limit ourselves to only a brief illustration of the recent progress made in revealing the location of framework aluminum in high-silica zeolites by combined spectroscopic and quantum chemical studies, while the main focus will be on the recent theoretical concept of the structures of extra-framework cations. Recently, a tool based on the complementary use of the high resolution of 27 Al 3Q MAS (magic-angle spinning) NMR spectroscopy and DFT calculations

11.6 Structural Properties of Zeolites

has been proposed for studying the local geometry of framework [AlO4 ]− units and for the identiﬁcation of the Al sitting in high-silica zeolites [94–96]. The respective experimental technique allows identiﬁcation and quantiﬁcation of the 27 Al resonances corresponding to individual T-sites in the zeolite framework. High-silica zeolite models with different framework Al distribution were optimized using a hybrid DFT:force ﬁeld method where the higher-level method is applied to cluster models containing the Al atom surrounded by at least ﬁve coordination shells. The remaining part of the zeolite was then described by a force ﬁeld method. A bare charged framework of ZSM-5 zeolite (MFI topology) that includes neither cations nor water molecules was used as a relevant model. Each model contained only one type of Al substitution. After the structure determination, NMR shielding tensors were calculated for the atoms of the optimized clusters using the gauge-independent atomic orbital (GIAO) methods [97]. The resulting calculated isotropic 27 Al NMR shifts were then used to assign and rationalize the experimental observations. The results thus obtained allowed to assign the observed 27 Al resonances to the particular T-sites in the MFI framework [94–96]. Although a trend has been detected for smaller 27 Al isotropic chemical shifts with increasing average T–O–T angle [95] that had been previously proposed as a tool for the identiﬁcation of the Al sitting in zeolites [98], the corresponding correlation was shown to be not suitable for the assignment purposes. A very important conclusion was made that the local geometry of framework [AlO4 ]− tetrahedra cannot be deduced directly from the experimental NMR data, whereas such information can only be obtained from the theoretical calculations [95]. Concerning the relative location of the anionic [AlO4 ]− sites in the framework, the very recent study by Dˇedeˇcek et al. [96] has shown that the inﬂuence of the second Al site in the next nearest or in the next-next-nearest framework position on the 27 Al isotropic shift is not uniform. Thus, it has been concluded that this combined NMR and DFT approach is only suitable for determining the Al sitting in zeolites with only very low local density of aluminum. In other words, the concentration of the Al−O-(SiO)n −Al (n = 1 or 2) sequences in the framework must be negligible. Revealing the Al distribution and the rules, which govern Al sitting, in high-silica zeolites is very important to understand the molecular structure and chemical reactivity of cationic species conﬁned in the microporous matrix. For univalent cations, a well-accepted model is localization of the positively charged ions in the vicinity of the negatively charged [AlO4 ]− framework tetrahedral units [99]. For cations with a higher charge, this model requires close proximity of the aluminum substitutions in the zeolite framework. This requirement may not always be met, especially for high-silica zeolites. Indeed, even closely located [AlO4 ]− tetrahedra may not necessarily face the same zeolitic ring or even channel preventing thus a direct charge compensation of multivalent cations (see e.g., [96]). An alternative model involves indirect compensation of the multiple-charge cations by distantly placed negative charges of the zeolite lattice. This concept has been initially put forward to account for the high reactivity of Zn-modiﬁed ZSM-5

327

328

11 Theoretical Chemistry of Zeolite Reactivity

in alkane activation [100, 101]. In this case, part of the exchangeable Zn2+ cations is located in the vicinity of one framework anionic [AlO4 ]− site, while the other negative site required for the overall charge neutrality is located at a longer distance, where it does not directly interact with the extra-framework positive charge. The existence of such species has been supported by spectroscopic [100–103] and theoretical studies [104, 105]. However, a ﬁrm theoretical evidence for a structural model, in which the positions of multivalent cations in zeolite are not dominated by the direct interaction between the mononuclear M2+ cation and the framework charge, has not yet been presented. A useful structural model includes then the presence of extra-framework oxygen-containing anions that coordinate to the metal (M), resulting in the formation of multinuclear cationic complexes with a formal charge of +1 such as [M3+ = O2− ]+ , [M2+ −OH− ]+ , and so on For example, formation of the isolated gallyl GaO+ ions was proposed to be responsible for the experimentally observed enhancement of the dehydrogenation catalytic activity of ZSM-5 modiﬁed with Ga+ upon the stoichiometric treatment with N2 O [106]. However, a comprehensive computational analysis of possible reaction paths for the light alkane dehydrogenation indicated that the isolated GaO+ ions cannot be responsible for the catalytic activity [107]. An alternative interpretation is the formation of multiple-charged extra-framework oligomeric (GaO)n n+ cations. Stability and reactivity of such species in high-silica mordenite have been studied by periodic DFT calculations [108, 109]. It has been shown that isolated gallyl ions tend to oligomerize resulting in formation of oxygen-bridged Ga3+ pairs. The stability of the resulting cationic complexes does not require proximate Al substitution in the framework (Figure 11.8). The theoretical calculations indicate that the oligomers with a higher degree of aggregation can be in principle formed in oxidized Ga-exchanged zeolites [109].

[(GaO+)2] [AIO4] [AIO4]

[(Ga2O2)2+]SP

[(Ga2O2)2+]

[AIO4]

[AIO4]

GaO+

[AIO4]

GaO+

Ga2O22+

[AIO4] Ga2O22+ [SiO4]0

Ga O Al Si

−50

−117

Figure 11.8 Structures of (GaO)2 2+ isomers in a high-silica (Si/Al = 23) mordenite model. The numbers under the structures correspond to the DFT-computed reaction energies (E in kilojoules per mole) for the stoichiometric oxidation of two exchangeable Ga+ sites with N2 O toward the respective cations [109].

[SiO4]0

−166

11.6 Structural Properties of Zeolites

329

It has been concluded that the formation of the favorable coordination environment of the metal centers via interaction with basic oxygen anions dominates over direct charge compensation and leads to clustering of the extra-framework species. The presence of multiply charged bi- or oligonuclear metal oxide species in zeolites does not require the immediate proximity of an equivalent number of negative framework charges. Nonlocalized charge compensation is expected to be a common feature of high-silica zeolites modiﬁed with metals ions. The corresponding theoretical concept is argued to be useful to develop new structural models for the intrazeolitic active sites involving multiple metal centers. Finally, implication of the concept of nonlocalized charge compensation allowed considering formation of other possible multinuclear Ga-containing intrazeolite species in high-silica zeolite matrix. The performance of different cationic oxygenand sulfur-containing Ga clusters in light alkane dehydrogenation has been analyzed by means of periodic DFT calculations [110]. From the results thus obtained a structure–reactivity relationship of a remarkable predictive power has been derived for their catalytic performance. It has been shown that the computed activation free energies as well as the free energy changes of the important elementary steps of the catalytic ethane dehydrogenation cycle scale linearly with the values of free energy change of the active site regeneration (Figure 11.9). The later parameter has been chosen as the reactivity descriptor because it reﬂects strength of the associated active Lewis acid–base pairs. The relationship presented points to the optimum composition and structure of the intrazeolite Ga cluster – [H-Ga(O)(OH)Ga]2+ – for the catalytic dehydrogenation of light alkanes [110]. Similar active species have been previously proposed to account for the enhanced dehydrogenation activity of Ga-modiﬁed ZSM-5 zeolite upon water cofeeding [109, 111]. Nevertheless, so far the preference for the nonlocalized charge compensation in zeolites has been convincingly shown only for Ga-containing species stabilized

(a)

Gibbs free energy of H2 recombination 0 (∆G823K, kJ mol−1) ◦

350

#

DG 823K, kJ mol−1

0 ∆G 823K , kJ mol−1

Strength of Lewis acid-base pair 200 4+ [Ga4O4] 150 [Ga2S2]2+ 100 [GaS2HGaH]2+ 50 [Ga2O2]2+ 0 [Ga4O4]4+ −50 [GaO2HGaH]2+ 2+ [Ga S ] 2 2 −100 C–H cleavage R 2 =0.989 C2H4 desorption −150 H2 recombination −200 −200 −150 −100 −50 0 50 100 150

300 250 200 150 100

S* Ga Ga* S S* O* HGa Ga* HGa Ga* O* S O H Ga Ga* H O

(b)

Figure 11.9 (a) Gibbs free energies (G823 K ) of elementary reaction steps of C2 H6 dehydrogenation and (b) activation Gibbs free energies (G#823 K ) of the C−H activation and of the H2 recombination step plotted against the re◦ spective values of G823 K of H2 recombination [110].

O

R 2 =0.986 R 2 =0.953

C–H cleavage H recombination

2 50 −200 −150 −100 −50

Ga* O O O* GaGaGa

0

50

100

Gibbs free energy of H2 recombination 0 (∆G823K, kJ mol−1)

150

330

11 Theoretical Chemistry of Zeolite Reactivity

in a single zeolite topology. Although, these results have been anticipated to be valid also for other metal ions in different microporous matrices, the expansion of theoretical studies to other metal-containing zeolites with different topologies and framework composition is needed to generalize this theoretical structural concept. Further theoretical efforts in understanding the fundamental factors that govern Al sitting and their relative distribution in zeolite framework are strongly desired to create a resolved molecular-level picture of the structural properties of these microporous materials.

11.7 Summary and Outlook

Computational modeling is becoming one of the key contributors to the zeolite science. Theoretical methods play a pivotal role in assisting the interpretation of the experimental data, revealing the structural and chemical properties of microporous materials, and in developing the molecular-level understanding of mechanistic aspects of catalytic reactions in conﬁned space. Obviously, it was impossible to review all of the computational methods and areas of their application in zeolite sciences on pages available here. In this chapter we attempted to illustrate the current capabilities and limitations of promising quantum chemical methods as applied to zeolite sciences. The power of quantum chemical techniques in rationalizing the complex chemical processes in microporous matrices and in developing novel useful chemical and structural concepts is highlighted. There are two major future challenges remaining in the computational chemistry of zeolites. Thanks to the great development of theoretical methodologies and the rapid growth in hardware performance we are now able to model rather accurately various aspects of the chemical processes taking place in the zeolite void space. This allows us to unravel molecular details of many known processes and to understand the fundamental factors that determine and control the chemical reactivity of microporous catalysts. The next step is the development of ab initio-based computational approaches that will serve as a tool for the prediction of chemical reactivity of zeolite catalysts. This however is not a trivial task taking into account the high complexity of the associated chemical processes, many aspects of which are yet not well understood. The second challenge is to develop novel hierarchical approaches integrating various levels of theory into one multiscale simulation to cover the disparate length and time scales and allow a comprehensive theoretical kinetic description of a working catalyst. The ab initio electronic structure calculations discussed in this chapter provide important molecular-level information about the details of the elementary reactions involved in the catalytic processes. Combining the results of such calculations with the statistical simulations (e.g., kinetic Monte Carlo) to account for the interplay between all elementary processes involved in the catalytic

References

cycle at the mesoscopic scale, and with the macroscopic theories, which describe the effects of heat and mass transfer, would ultimately result in the complete description of the zeolite catalysis.

References 1. Woodley, S.M. and Catlow, R. (2008) 2. 3. 4.

5. 6.

7.

8.

9.

10.

11. 12. 13. 14.

15. 16. 17. 18.

Nat. Mater., 7, 937. Smit, B. and Maesen, T.L.M. (2008) Chem. Rev., 108, 4125. Smit, B. and Maesen, T.L.M. (2008) Nature, 451, 671. Jensen, F. (1999) Introduction to Computational Chemistry, Wiley-Interscience, New York. Levine, I.N. (1983) Quantum Chemistry, Allyn and Bacon, Boston. Leach, A.R. (1996) Molecular Modeling: Principles and Applications, Pearson Education, Harlow. Foresman, J.B. and Frish, A. (1996) Exploring Chemistry with Electronic Structure, 2nd edn, Gaussian, Pittsburg. Parr, R.G. and Yang, W. (1989) Density Functional Theory of Atoms in Molecules, Oxford University Press, New York. Young, D.C. (2001) Computational Chemistry: A Practical Guide for Applying Techniques to Real-World Problems, Wiley-Interscience, New York. Feyereisen, M., Fitzgerald, G., and Komornicki, A. (1993) Chem. Phys. Lett., 208, 359. Hohenberg, P. and Kohn, W. (1964) Phys. Rev., 136, B864. Kohn, W. and Sham, L. (1965) Phys. Rev., 140, A1133. Becke, A.D. (1988) Phys. Rev. A, 38, 3098. Perdew, J.P., Chevary, J.A., Vosko, S.H., Jackson, K.A., Pederson, M.R., Singh, D.J., and Fiolhais, C. (1992) Phys. Rev. B, 46, 6671. Perdew, J.P., Burke, K., and Ernzerhof, M. (1996) Phys. Rev. Lett., 77, 3865. Becke, A.D. (1993) J. Chem. Phys., 98, 5648. Adamo, C. and Barone, V. (1998) J. Chem. Phys., 108, 664. Adamo, C. and Barone, V. (1999) J. Chem. Phys., 110, 6158.

19. Johnson, E.R. and DiLabio, G.A. (2006)

Chem. Phys. Lett., 419, 333. 20. Zhao, Y. and Truhlar, D.G. (2005) J.

Chem. Theory Comput., 1, 415. 21. Guest, M.F., Bush, I.J., van Dam,

H.J.J., Sherwood, P., Thomas, J.M.H., van Lenthe, J.H., Havenith, R.W.A., and Kendrick, J. (2005) Mol. Phys., 103, 719. 22. Frisch, M.J., Trucks, G.W., Schlegel, H.B., Scuseria, G.E., Robb, M.A., Cheeseman, J.R., Montgomery, J.A. Jr., Vreven, T., Kudin, K.N., Burant, J.C., Millam, J.M., Iyengar, S.S., Tomasi, J., Barone, V., Mennucci, B., Cossi, M., Scalmani, G., Rega, N., Petersson, G.A., Nakatsuji, H., Hada, M., Ehara, M., Toyota, K., Fukuda, R., Hasegawa, J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Klene, M., Li, X., Knox, J.E., Hratchian, H.P., Cross, J.B., Bakken, V., Adamo, C., Jaramillo, J., Gomperts, R., Stratmann, R.E., Yazyev, O., Austin, A.J., Cammi, R., Pomelli, C., Ochterski, J.W., Ayala, P.Y., Morokuma, K., Voth, G.A., Salvador, P., Dannenberg, J.J., Zakrzewski, V.G., Dapprich, S., Daniels, A.D., Strain, M.C., Farkas, O., Malick, D.K., Rabuck, A.D., Raghavachari, K., Foresman, J.B., Ortiz, J.V., Cui, Q., Baboul, A.G., Clifford, S., Cioslowski, J., Stefanov, B.B., Liu, G., Liashenko, A., Piskorz, P., Komaromi, I., Martin, R.L., Fox, D.J., Keith, T., Al-Laham, M.A., Peng, C.Y., Nanayakkara, A., Challacombe, M., Gill, P.M.W., Johnson, B., Chen, W., Wong, M.W., Gonzalez, C., and Pople, J.A. (2003) Gaussian 03, Revision B.05, Gaussian, Inc., Pittsburgh. 23. Ahlrichs, R. (1989) Chem. Phys. Lett., 162, 165. 24. Segall, M.D., Lindan, P.J.D., Probert, M.J., Pickard, C.J., Hasnip, P.J., Clark, S.J., and Payne, M.C. (2002) J. Phys. Condens. Matter., 14, 2717.

331

332

11 Theoretical Chemistry of Zeolite Reactivity 25. (a) Marx, D. and Hutter, J. (2000) Mod-

26.

27.

28.

29. 30. 31.

32. 33. 34. 35. 36.

37. 38.

39. 40.

41.

ern Methods and Algorithms of Quantum Chemistry, NIC, FZ J¨ulich, p. 301; (b) Andreoni, W. and Curioni, A. (2000) Parallel Comput., 26, 819. (a) Kresse, G. and Hafner, J. (1994) Phys. Rev. B, 49, 14251; (b) Kresse, G., and Furthm¨uller, J. (1996) Comput. Mater. Sci., 6, 15; (c) Kresse, G. and Furthm¨uller, J. (1996) Phys. Rev. B, 54, 11169. Dovesi, R., Saunders, V.R., Roetti, C., Orlando, R., Zicovich-Wilson, C.M., Pascale, F., Civalleri, B., Doll, K., Harrison, N.M., Bush, I.J., D’Arco, Ph., and Llunell, M. (2006) CRYSTAL2006 User’s Manual, Universita di Torino, Torino, (2008) http://www.crystal.unito.it. Tosoni, S., Tuma, C., Sauer, J., Civalleri, B., and Ugliengo, P. (2007) J. Chem. Phys., 127, 154102. Tuma, C. and Sauer, J. (2004) Chem. Phys. Lett., 387, 388. Tuma, C. and Sauer, J. (2006) Phys. Chem. Chem. Phys., 8, 3955. de M´enorval, B., Ayrault, P., Gnep, N.S., and Guisnet, M. (2005) J. Catal., 230, 38. Kato, T. and Reed, C.A. (2004) Angew. Chem. Int. Ed. Engl., 43, 2907. Boronat, M. and Corma, A. (2008) Appl. Catal. A Gen., 336, 2. Correa, R.J. and Mota, C.J.A. (2002) Phys. Chem. Chem. Phys., 4, 375. Boronat, M., Viruela, P.M., and Corma, A. (2004) J. Am. Chem. Soc., 126, 3300. Rozanska, X., van Santen, R.A., Demuth, T., Jutschka, F., and Hafner, J. (2003) J. Phys. Chem. B, 107, 1309. Tuma, C. and Sauer, J. (2005) Angew. Chem. Int. Ed. Engl., 44, 4769. Haw, J.F., Nicholas, J.B., Xu, T., Beck, L.W., and Ferguson, D.B. (1996) Acc. Chem. Res., 29, 259. Zhao, Y. and Truhlar, D.G. (2008) J. Phys. Chem. C, 112, 6860. de Moor, B.A., Reyniers, M.-F., Sierka, M., Sauer, J., and Marin, G.B. (2008) J. Phys. Chem. C, 112, 11796. Kerber, T., Sierka, M., and Sauer, J. (2008) J. Comp. Chem., 29, 2088.

42. Svelle, S., Tuma, C., Rozanska, X.,

43. 44.

45.

46.

47.

48.

49.

50.

51. 52.

53.

54. 55.

56.

57. 58. 59.

Kerber, T., and Sauer, J. (2009) J. Am. Chem. Soc., 131, 816. Svelle, S., Rønning, P.O., and Kolboe, S. (2004) J. Catal., 224, 115. Svelle, S., Rønning, P.O., Olsbye, U., and Kolboe, S. (2005) J. Catal., 234, 385. Svelle, S., Arstad, B., Kolboe, S., and Swang, O. (2003) J. Phys. Chem. B, 107, 9281. Rozanska, X., Saintigny, X., van Santen, R.A., and Hutschka, F. (2001) J. Catal., 202, 141. Rozanska, X., van Santen, R.A., Hutschka, F., and Hafner, J. (2001) J. Am. Chem. Soc., 123, 7655. Vos, A.M., Rozanska, X., Schoonheydt, R.A., van Santen, R.A., Hutschka, F., and Hafner, J. (2001) J. Am. Chem. Soc., 123, 2799. Demuth, T., Benco, L., Hafner, J., Toulhoat, H., and Hutschka, F. (2001) J. Chem. Phys., 114, 3703. (a) Grimme, S. (2004) J. Comput. Chem., 25, 1463; (b) Grimme, S. (2006) J. Comput. Chem., 27, 1787. Joshi, Y.V. and Thomson, K.T. (2008) J. Phys. Chem. C, 112, 12825. de Moor, B.A., Reyniers, M.-F., Sierka, M., Sauer, J., and Marin, G.B. (2009) Phys. Chem. Chem. Phys., 11, 2939. Zhao, Y., Schultz, N.E., and Truhlar, D.G. (2005) J. Chem. Phys., 123, 161103. Zhao, Y. and Truhlar, D.G. (2008) Theor. Chem. Acc., 120, 215. (a) Dion, M., Rydberg, H., Schr¨oder, E., Langreth, D.C., and Lundqvist, B.I. (2004) Phys. Rev. Lett., 92, 246401; (b) Dion, M., Rydberg, H., Schr¨oder, E., Langreth, D.C., and Lundqvist, B.I. (2005) Phys. Rev. Lett., 95, 109902(E). Thonhauser, T., Cooper, V.R., Li, S., Puzder, A., Hyldgaard, P., and Langreth, D.C. (2007) Phys. Rev. B, 76, 125112. Vydrov, Q.A., Wu, Q., and van Voorhis, T. (2008) J. Chem. Phys., 129, 014106. Stocker, M. (1999) Microporous Mesoporous Mater., 29, 3. Haw, J.F., Song, W.G., Marcos, D.M., and Micholas, J.B. (2003) Acc. Chem. Res., 36, 314.

References 60. (a) Dahl, I.M. and Kolboe, S. (1993)

61.

62.

63.

64.

65.

66.

67.

68.

69.

70.

71.

72.

73. 74. 75. 76.

Catal. Lett., 20, 329; (b) Dahl, I.M. and Kolboe, S. (1994) J. Catal., 149, 458. Song, W.G., Marcus, D.M., Fu, H., Ehresmann, J.O., and Haw, J.F. (2002) J. Am. Chem. Soc., 124, 3844. Marcus, D.M., McLachlan, K.A., Wildman, M.A., Ohresmann, J.O., Kletnieks, P.W., and Haw, J.F. (2006) Angew. Chem. Int. Ed. Engl., 45, 3133. Lesthaeghe, D., van Speybroeck, V., Marin, G.B., and Waroquier, M. (2006) Angew. Chem. Int. Ed. Engl., 45, 1714. Lesthaeghe, D., van Speybroeck, V., Marin, G.B., and Waroquier, M. (2007) Ind. Eng. Chem. Res., 46, 8832. Sassi, A., Wildman, M.A., Ahn, H.J., Prasad, P., Nicholas, J.B., and Haw, J.F. (2002) J. Phys. Chem. B, 106, 2294. Bjorgen, M., Bonino, F., Kolboe, S., Lillerud, K.-P., Zecchina, A., and Bordiga, S. (2003) J. Am. Chem. Soc., 125, 15863. Song, W.G., Nicholas, J.B., Sassi, A., and Haw, J.F. (2002) Catal. Lett., 81, 49. Lesthaeghe, D., van Speybroeck, V., Marin, G.B., and Waroquier, M. (2007) Angew. Chem. Int. Ed. Engl., 46, 1311. Lesthaeghe, D., van Speybroeck, V., Marin, G.B., and Waroquier, M. (2007) Stud. Surf. Sci. Catal., 170, 1668. Svelle, S., Joensen, F., Nerlov, J., Olsbye, U., Lillerud, K.-P., Kolboe, S., and Bjorgen, M. (2006) J. Am. Chem. Soc., 128, 14770. McCann, D.M., Lesthaeghe, D., Kletnieks, P.W., Guenther, D.R., Hayman, M.J., Van Speybroeck, V., Waroquier, M., and Haw, J.F. (2008) Angew. Chem. Int. Ed. Engl., 47, 5179. Wang, C.-M., Wang, Y.-D., Xie, Z.-K., and Liu, Z.-P. (2009) J. Phys. Chem. C, 113, 4584. Corma, A. (2003) J. Catal., 216, 298. Degnan, T.F. Jr. (2003) J. Catal., 216, 32. Ringe, D. and Petsko, G.A. (2008) Science, 320, 1428. Das, S., Brudvig, G.W., and Crabtree, R.H. (2008) Chem. Commun., 413.

77. OteroAre´an, C., Rodr´ıgue Delgado,

78.

79.

80.

81.

82.

83.

84.

85.

86. 87.

88. 89. 90. 91. 92.

93. 94.

M., L´opez Bauc¸a` , C., Vrbka, L., and Nachtigall, P. (2007) Phys. Chem. Chem. Phys., 9, 4657. Garrone, E., Bul´anek, R., Frolich, K., Otero Are´an, C., Rodr´ıgues Delgado, M., Turnes Palomino, G., Nachtigallov´a, D., and Nachtigall, P. (2006) J. Phys. Chem. B, 110, 22542. Otero Are´an, C., Rodr´ıgues Delgado, M., Frolich, K., Bul´anek, R., Pulido, A., Fiol Bibiloni, G., and Nachtigall, P. (2008) J. Phys. Chem. C, 112, 4658. Nachtigallov´a, D., Virbka, L., Budsk´y, O., and Nachtigall, P. (2008) Phys. Chem. Chem. Phys., 10, 4189. Pulido, A., Nachtigall, P., Zukal, A., ˇ Dom´ınguez, I., and Cejka, J. (2009) J. Phys. Chem. C, 113, 2928. Pidko, E.A., Mignon, P., Geerlings, P., Schoonheydt, R.A., and van Santen, R.A. (2008) J. Phys. Chem. C, 112, 5510. Mignon, P., Pidko, E.A., van Santen, R.A., Geerlings, P., and Schoonheydt, R.A. (2008) Chem. Eur. J., 14, 5168. Otero Are´an, C., Nachtigallov´a, D., Nachtigall, P., Garrone, E., and Rodr´ıgues Delgado, M. (2007) Phys. Chem. Chem. Phys., 9, 1421. Pidko, E.A. and van Santen, R.A. (2010) Int. J. Quantum Chem., 110, 210. Pidko, E.A. and van Santen, R.A. (2006) ChemPhysChem, 7, 1657. Pidko, E.A., Xu, J., Mojet, B.L., Lefferts, L., Subbotina, I.R., Kazansky, V.B., and van Santen, R.A. (2006) J. Phys. Chem. B, 110, 22618. Pidko, E.A. and van Santen, R.A. (2006) J. Phys. Chem. B, 110, 2963. Frei, H. (2006) Science, 313, 209. Vasenkov, S. and Frei, H. (1997) J. Phys. Chem. B, 101, 4539. Blatter, F., Sun, H., Vasenkov, S., and Frei, H. (1998) Catal. Today, 41, 297. Bhan, A., Allian, A.D., Sunley, G.J., Law, D.J., and Iglesia, E. (2007) J. Am. Chem. Soc., 129, 4919. Bhan, A. and Iglesia, E. (2008) Acc. Chem. Res., 41, 559. Sklenak, S., Dˇedeˇcek, J., Li, C., Wichterlov´a, B., G´abov´a, V., Sierka,

333

334

11 Theoretical Chemistry of Zeolite Reactivity

95.

96.

97. 98. 99.

100. 101. 102.

M., and Sauer, J. (2007) Angew. Chem. Int. Ed. Engl., 46, 7286. Sklenak, S., Dˇedeˇcek, J., Li, C., Wichterlov´a, B., G´abov´a, V., Sierka, M., and Sauer, J. (2009) Phys. Chem. Chem. Phys., 11, 1237. Dˇedeˇcek, J., Sklenak, S., Li, C., Wichterlov´a, B., G´abov´a, V., Brus, J., Sierka, M., and Sauer, J. (2009) J. Phys. Chem., 113, 1447. Wolinski, K., Hinton, J.H., and Pulay, P. (1990) J. Am. Chem. Soc., 112, 8251. Lippmaa, E., Samson, A., and Magi, M. (1986) J. Am. Chem. Soc., 108, 1730. Centi, G., Wichterlova, B., and Bell, A.T. (eds) (2001) Catalysis by Unique Metal Ion Structures in Solid Matrices, NATO Science Series, Kluwer Acadamic, Dordrecht. Kazansky, V.B. and Serykh, A.I. (2004) Phys. Chem. Chem. Phys., 6, 3760. Kazansky, V. and Serykh, A. (2004) Microporous Mesoporous Mater., 70, 151. Kazansky, V.B., Serykh, A.I., and Pidko, E.A. (2004) J. Catal., 225, 369.

103. Kazansky, V.B. and Pidko, E.A. (2005)

J. Phys. Chem. B, 109, 2103. 104. Zhidomirov, G.M., Shubin, A.A.,

105. 106.

107.

108.

109.

110. 111.

Kazansky, V.B., and van Santen, R.A. (2005) Theor. Chem. Acc., 114, 90. Pidko, E.A. and van Santen, R.A. (2007) J. Phys. Chem. C, 111, 2643. Rane, N., Overweg, A.R., Kazansky, V.B., van Santen, R.A., and Hensen, E.J.M. (2006) J. Catal., 239, 478. Pidko, E.A., Hensen, E.J.M., and van Santen, R.A. (2007) J. Phys. Chem. B, 111, 13068. Pidko, E.A., Hensen, E.J.M., Zhidomirov, G.M., and van Santen, R.A. (2008) J. Catal., 255, 139. Pidko, E.A., van Santen, R.A., and Hensen, E.J.M. (2009) Phys. Chem. Chem. Phys., 11, 2893. Pidko, E.A. and van Santen, R.A. (2009) J. Phys.Chem. B, 113, 4246. Hensen, E.J.M., Pidko, E.A., Rane, N., and van Santen, R.A. (2007) Angew. Chem. Int. Ed. Engl., 46, 7273.

335

12 Modeling of Transport and Accessibility in Zeolites Sof´ıa Calero Diaz

12.1 Introduction

Modeling plays an important role in the ﬁeld of zeolites and related porous materials. The use of molecular simulations allows the prediction of adsorption and diffusion coefﬁcients in these materials and also provides important information about the processes taking place inside the porous structures at the molecular level [1–10]. Hence, molecular modeling is a very good complement to experimental work in the understanding of the molecular behavior inside the pores. Despite the great interest and the applicability of zeolites, there are still many facets of the molecular mechanisms for a given reaction inside their pores that are poorly understood. These mechanisms are important for (i) the way the zeolite adsorbs, diffuses, and concentrates the adsorbates near the speciﬁc active sites; (ii) the interactions between the zeolite and the adsorbates and the effect on the electronic properties of the system; (iii) the chemical conversion at the active site; and (iv) the way the zeolite disperses the ﬁnal product. Detailed knowledge on the molecular mechanisms involved will eventually lead to an increase in the reaction efﬁciency. This chapter focuses on molecular modeling of transport and accessibility in zeolites. It describes how simulations have contributed to a better understanding of these materials and provides a summary of the state of the art as well as of current challenges. The chapter is organized as follows. First, common models and potentials are brieﬂy described. This is followed by a general overview on current simulation methods to compute adsorption, diffusion, free energies, surface areas, and pore volumes. The chapter continues with some examples on applications of molecular modeling to processes of interest from the industrial and environmental point of view. Finally, the chapter closes with a summary and some remarks on future challenges.

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

336

12 Modeling of Transport and Accessibility in Zeolites

12.2 Molecular Models

To perform molecular simulations in zeolites, adequate models for all the atoms and molecules involved in the system are needed together with inter- and intramolecular potentials to describe the interactions between them. This section deals with a discussion on the most common models and potentials used in the literature for the zeolite framework, the nonframework cations, and the guest molecules. 12.2.1 Modeling Zeolites and Nonframework Cations

The zeolite framework is usually built from silicon, aluminum, and oxygen with the crystallographic positions of these atoms taken from the dehydrated structures [11]. Zeolites with a Si/Al ratio higher than 1 can be obtained from random substitution of aluminum by silicon, either ignoring [12, 13] or taking into account distribution rules [14–18]. The aluminum atoms can also be assigned to energetic and entropic preferential positions [19–24], or using theoretical approaches that identiﬁes experimentally accessible properties dependent on the aluminum distribution and associated cation distribution [25, 26]. Substitution of silicon for aluminum generates a negative net charge in the zeolite framework that needs to be compensated by either nonframework protons or cations in order to make the zeolite charge neutral. Some models explicitly distinguishes silicon from aluminum assigning different charges not only to those atoms but also to the oxygen atoms bridging two silicon atoms, and the oxygen atoms bridging one silicon and one aluminum atom [16, 27]. The charge distribution on the oxygen framework is often considered static in such a way that the polarization of oxygen by nearby extra framework cations is neglected. The extra framework cations can either remain ﬁxed [28, 29] or move freely and adjust their position depending on their interactions with the system [16, 27, 30]. The latter requires potentials to predict the distribution of cations in the bare or loaded zeolite. The cation motions have to be sampled using displacements and random insertions that bypass energy barriers. A force ﬁeld is described as a set of functions needed to deﬁne the interactions in a molecular system. A wide variety of force ﬁelds that can be applied to zeolites exist. Among them, we ﬁnd universal force ﬁeld (UFF) [31], Discover (CFF) [32], MM2 [33], MM3 [34, 35], MM4 [36, 37], Dreiding [38], SHARP [39], VALBON [40], AMBER [41], CHARMM [42], OPLS [43], Tripos [44], ECEPP/2 [45], GROMOS [46], MMFF [47], Burchart [48], BKS [48], and specialty force ﬁelds for morphology predictions [49] or for computing adsorption [50]. In one approach, force ﬁelds are designed to be generic, providing a broad coverage of the periodic table, including inorganic compounds, metals, and transition metals. The diagonal terms in the force-constant matrix or these force ﬁelds are usually deﬁned using simple functional forms. Owing to the generality of parameterization, these force ﬁelds are normally expected to yield reasonable predictions of molecular structures. However, emphasis was given to improving the accuracy in predicting molecular

12.2 Molecular Models

properties while maintaining a fair broad coverage of the periodic table. To achieve this goal, the force ﬁelds require complicated functional forms [32, 34–37, 47], and the parameters are derived by ﬁtting to experimental data or to ab initio data. Surprisingly, these general force ﬁelds give very poor results for specialized systems as adsorption and diffusion in zeolites. It is for this reason that new force ﬁelds have been optimized for pure silica zeolites [51–53] and also for those with nonframework sodium, calcium, and protons [16, 54–56]. The new force ﬁeld parameters provide quantitative good predictions for adsorption and molecular transport in these systems [16, 51–58]. Most molecular simulation studies in zeolites are performed using the Kiselev-type potentials, where the zeolite atoms are held rigid at the crystallographic positions [59]. However, some authors have also investigated the effect of ﬂexibility, using a variety of potentials for the framework atoms [60–62] and testing the accuracy and viability by comparing the computed adsorption [63, 64], diffusion [62, 65, 66], IR spectra [60, 61], or structural parameters [67, 68] with experimental data. 12.2.2 Modeling Guest Molecules

To model guest molecules, rigid or ﬂexible models can be used. For simple molecules such as carbon dioxide, nitrogen, hydrogen, oxygen, or even water, rigid models with multipoles or polarization seem a good representation [30, 56, 57, 69–78]. Complex molecules such as hydrocarbons normally require ﬂexible models. A variety of ﬂexible models have been addressed in literature spanning from the simplicity and the efﬁciency of the united-atom models to the complexity and the accuracy of the full-atom models [38, 79–86]. These models typically include partial charges for all atoms, expressions for describing bond-bending, bond-stretching, and torsion motions, and Lennard-Jones or Buckingham potential parameters that are often obtained from the ﬁtting to ab initio [47, 87, 88] or to experimental vapor–liquid equilibrium data [53, 80, 89]. This is illustrated in Figure 12.1 with a comparison of the experimental vapor–liquid equilibrium curve (liquid branch) for ethylene [90], propylene [90], carbon dioxide [91], and argon [92], with computed data using available models of literature [53, 56, 64, 69, 93, 94]. The interactions between the guest molecules and the zeolite and nonframework cations must be reproduced with efﬁcient and accurate potentials. Although some authors have opted for more sophisticated models [4], a simple and computational efﬁcient option is the Kiselev-type model [59]. This model is based on Lennard-Jones potentials for the van der Waals interactions and on partial charges on all atoms of the system for the coulombic interactions that can be neglected for nonpolar guest molecules. The interactions of the guest molecules with the zeolite are dominated by the dispersive forces between the guest and the oxygen atoms of the structure [59], so the van der Waals interactions of guest molecules with the Si or Al atoms are often neglected. The development of transferable potentials provides accurate representation of the interactions of the experimental system that is being simulated remains

337

12 Modeling of Transport and Accessibility in Zeolites 400 Propylene

360 320 Temperature (K)

338

Carbon dioxide 280 240 200 160

Ethylene

Argon

120 80 0.2

0.4

0.6

0.8

1.0

1.2

1.4

Density (g ml−1)

Figure 12.1 Liquid branch of the vapor–liquid equilibrium curves for ethylene, propylene, carbon dioxide, and argon. Experimental data are shown as solid [90–92]. Simulated data using available models of

literature are shown as open triangles [93], ﬁlled triangles [53], open squares [93], ﬁlled squares [53], open circles [69], ﬁlled circles [56], down open triangles [94], and down ﬁlled triangles [64].

a challenge. Accurate parameters can predict topology-speciﬁc adsorption and transport properties, and this can be an effective tool to resolve discrepancies among experimental datasets. Most efforts to improve the models are focused on the optimization of the Lennard-Jones parameters [16, 52–55], and Dubbeldam et al. reported a method to obtain accurate sets of parameters for united atom models, based on the ﬁtting to experimental isotherms with inﬂection points [51]. Recent work has also highlighted that the partial charges assigned to the zeolite atoms are essential to reproduce experimental values of systems containing polar molecules [56, 57, 70, 95]. As an example, Desbiens et al. computed water adsorption in MFI using a host–guest Kiselev-type potential and a TIP4P model for water [96]. They found that this zeolite exhibits hydrophobic character only if the partial charge of the silicon atoms is kept below 1.7 a.u. The agreement with the experimental adsorption data drastically improved when the charge on the silicon atom was varied from 1.4 to 1.2. a.u. (Figure 12.2).

12.3 Simulation Methods

This section summarizes the speciﬁc methods to compute adsorption, free energies, surface areas, pore volumes, and diffusion in zeolites, and gives a short introduction of the most common simulation techniques. A more detailed description of these methods can be found in [97, 98]. In contrast to catalysis, modeling of adsorption and diffusion in zeolites is generally based on classical mechanics because quantum chemical calculation is still too expensive to compute statistical mechanical properties. In addition, adsorption in zeolites is dominated by dispersive

12.3 Simulation Methods

60

Loading (molecules / uc)

50 40 30 20 10 0

0

50

100

150

200

250

300

350

Pressure (MPa) Figure 12.2 Experimental and simulated liquid-phase adsorption isotherms of water in MFI at 300 K. Experimental data are shown in solid and dashed lines. Simulations were performed using the TIP4P water model with

a silicon partial charge of 1.4 a.u. (circles) and 1.2 a.u. (squares). The dotted lines are a guide to the eye. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission from Desbiens et al. [96].

interactions that are notoriously difﬁcult to treat quantum mechanically. Dispersion using MP2 limits the amount of atoms so severely that in most cases one cannot afford to use it to compute binding energies of a molecule with the entire periodic unit cell of a zeolite. These properties are accessible with plane-wave codes (e.g., VASP) using GGA- and LDA-type basis sets, but one can argue about the validity of these results for the computation of binding energies of small adsorbates. This chapter focuses on classical force ﬁeld–based methods for adsorption and diffusion, leaving catalysis and quantum approaches [88, 99, 100] beyond the scope of this work. 12.3.1 Computing Adsorption

Adsorption isotherms, Henry coefﬁcients, and isosteric heats of adsorption are important parameters for gas separation, catalysis, capture, and storage applications in zeolites that can be computed using molecular simulations. The simulations yield absolute values, which have to be converted to excess properties before comparing with experimental data [101, 102]. Adsorption isotherms are usually obtained using Monte Carlo simulations in the grand-canonical ensemble (GCMC). In this ensemble, the temperature, the volume, and the chemical potential of the adsorbed molecules are imposed. During the simulation, the molecules are exchanged with a reservoir at the same temperature

339

340

12 Modeling of Transport and Accessibility in Zeolites

and chemical potential. Hence, the number of molecules ﬂuctuates and during the simulation, the average number of adsorbate molecules is computed. The chemical potential is directly related to fugacity, which is computed from the pressure using an equation of state [23, 97]. Recently, transition matrix Monte Carlo has been proposed as an efﬁcient alternative for GCMC simulations [103, 104]. Henry coefﬁcients are directly related to the excess free energy (or excess chemical potential) of the adsorbed molecules. The free energy of a molecule cannot be directly computed using Monte Carlo or molecular dynamics (MDs) simulations techniques. At low-to-intermediate loading, special simulation techniques such as the Widom test particle method can be applied. Details on this and other available techniques used to compute free energies can be found in the literature [97]. In the limit of zero coverage, the heat of adsorption Qst can be obtained from the derivative of the Henry constant from the Clausius–Clapeyron equation (see, e.g., [105]), but in practice, it is more efﬁcient to obtain it directly from the energies of the system (see, e.g., [106]). Hence, both the Henry coefﬁcient and the isosteric heat of adsorption are usually computed from the total energy of the system in the canonical ensemble with a ﬁxed number of molecules (N), a given volume (V), and temperature (T). This requires two independent simulations to provide: (i) the energy of the single molecule inside the zeolite and (ii) the energy of the ideal gas situation [81, 106, 107]. This method of energy differences is very efﬁcient for pure silica structures, but it is unsuited to compute the isosteric heats of adsorption in structures with nonframework cations. When cations move around the energies of cation–cation, cation-framework energies are very large and the ﬁnal result consists of the subtraction of two very large energy values. The standard error is difﬁcult to reduce in such a system. In 2008, Vlugt et al. resolved that impediment by proposing a new method making use of as much cancellation in energy terms as possible. The method provides signiﬁcantly better and reliable results [108]. Figure 12.3 shows the isosteric heat of adsorption computed at 493 K for several n-alkanes in a LTA5A zeolite containing 32 sodium and 32 calcium cations per unit cell using the three methods. All force ﬁelds and models used for this comparison were taken from the literature [16, 52, 55]. The Monte Carlo simulations consist of random attempted moves that sample a given ensemble. Those moves are molecular translations and rotations for the NVT ensemble and additional molecular insertion and deletion for GCMC, and they are accepted or rejected with adequate criteria for the sampling of the corresponding ensemble probability function [97]. The NVT ensemble keeps ﬁxed the number of particles (N), the volume (V), and the temperature (T) of the system. The conventional Monte Carlo method is very efﬁcient for small molecules, but inefﬁcient for the insertion of big molecules such as long-chain hydrocarbons in the system. The conﬁgurational-bias Monte Carlo (CBMC) technique was developed to make possible the insertion for molecules in moderately dense liquids, initially for lattice models [109] and subsequently extended to continuous models [110–112]. It improves the conformational sampling of the molecules and increases the efﬁciency of the chain insertions, required for the calculations of the adsorption isotherms,

12.3 Simulation Methods

80

Q st (kJ mol−1)

70 60 50 40 30 20

1

2

3

4 5 6 7 Number of carbon atoms

Figure 12.3 Isosteric heat of adsorption at zero coverage for n-alkanes in zeolite LTA5A. The heats of adsorption were computed at 493 K using the method based on the Henry constant from the Clausius-Clapeyron equation [105] (), the method of energy

8

9

differences [81, 107] (), and the new method reported by Vlugt et al. [108] (×). The error bars are smaller than the symbol size, except for the method based on energy differences, for which the error bars are included.

the free energy, and the Henry coefﬁcients by many orders of magnitude. In a CBMC simulation, instead of inserting a molecule randomly in the host system, the molecule is gradually inserted atom by atom, avoiding overlap with the zeolite. The rules that accept or reject the grown molecule are chosen in such a way that they exactly remove the bias caused by the growing scheme [97]. Over the last few years, several improvements of the CBMC scheme have been proposed [113–115]. In the last years, there have been many simulations studies of adsorption in zeolites. Detailed overviews on adsorption studies using molecular simulations were compiled in 2001 by Fuchs and Cheettham [4], and more recently by Smit and Maesen [10]. A few examples on simulation of adsorption in zeolites are also highlighted in Section 12.3 of this chapter. 12.3.2 Computing Free Energy Barriers

The free energy methods used to obtain the Henry coefﬁcients can also be applied to compute the free energy as a function of the position in the zeolite and hence to obtain information about the free energy barrier that the molecules have to cross if they hop from one position to another. To compute the free energy as a function of the position in the zeolite, it is necessary to relate a position in the zeolite channel or cage to a reaction coordinate q. The computed free energy at a given position q contains both the potential energy and the entropy contribution that is directly related to the probability of the molecule to be found at q. A good example of the amount of information that can be provided using these methods

341

342

12 Modeling of Transport and Accessibility in Zeolites

7 Free energy barrier bF

6

Infinite dilution 1–6 molecules/cage 7 molecules/cage

5 4 3 2 1

Y

0

X

q A' q A

q" Reaction coordinate q

q B q B'

(a)

Free energy barrier bF

5

Infinite dilution 1–6 molecules/cage 7 molecules/cage

4 3 2 1

Z 0

X

qA

q"

qI

q"

qB

Reaction coordinate q

(b) Figure 12.4 Topology of the ERI-type zeolite showing the xy-direction (a) and the z-direction (b). Free energy proﬁles at 600 K for ethane in ERI at inﬁnite dilution and 1, 2, 3, 4, 5, 6, and 7 molecules per unit cell. The free energy proﬁles were obtained in the xy-plane with qA the center of a cage and qB

the center of a neighboring cage (c), and in the z-direction across a cage with qA the top of the cage, qI the middle of the cage, and qB the bottom of the cage (d). Reprinted with permission from Dubbeldam et al. [119]. Copyright  by the National Academy of Sciences.

can be taken from the studies of Dubbeldam et al. in ERI-type zeolites [116–119]. Figure 12.4 (taken from [119]) shows the computed free energies for ethane at several loadings in ERI-type zeolite. The maximum of the free energy is at q = 0 position, which corresponds to the dividing window q∗ , and the minimum values, which correspond to the values deep inside the cages (indicated as cages a and b in the ﬁgure). The free barriers obtained indicate that in the xy direction the hopping takes place on the hexagonal lattice, and each hop in the z direction is preceded by a hop in the xy-direction [119]. The analysis was performed as a function of loading. A reversal of the main direction of diffusion was discovered that by examining the free energy, proﬁles could be related to the hopping interactions and preferences of individual molecules.

12.3 Simulation Methods

(a)

(b)

(c)

(d)

Figure 12.5

Structure of a periodic unit cell of ITW (a), IWS (b), LTA (c), and IHW (d).

12.3.3 Computing Volume-Rendered Pictures, Zeolite Surface Areas, and Zeolite Pore Volumes

Volume-rendered (as in Figure 12.4) and isocontour pictures, surface areas, and pore volumes are extremely helpful to characterize zeolites. Figure 12.5 shows the isocontour pictures of the ITW, IWS, LTA, and IHW zeolites. To obtain these pictures, the zeolite unit cell is divided into a grid. Second, the free energy of a test particle is computed for all of the grid points. To obtain the energy landscape, the chosen energy value of the three-dimensional dataset is visualized and the zeolite framework is made transparent to avoid overlap. Usually, a high energy value is used to generate the pores and the framework walls. The surface area and the pore volume can be obtained indirectly from experimental isotherms using the BET theory and the saturation loading, respectively. Those parameters can also be computed from Monte Carlo simulations [120]. The surface area can be easily computed by rolling spherical test particles over the

343

344

12 Modeling of Transport and Accessibility in Zeolites

surface of the zeolite framework. If the spherical test particles are randomly shot into the zeolite, one can keep track of the fractions that do not overlap with the structure and compute the geometric the pore volume. These methods can be applied to both rigid and ﬂexible frameworks and also to pure silica zeolites and for those containing extra framework cations. Since these methods depend on the test particle used, to properly compute the pore volume and the surface area, the procedure must be repeated for a variety of test particles of arbitrary size, including point atoms. Table 12.1 compiles the computed pore volumes obtained for several zeolites using the described method. 12.3.4 Computing Diffusion

The way the molecules diffuse in the zeolite pores strongly inﬂuences the behavior of the structure during adsorption, separation, and catalytic processes. In recent years, simulations of diffusion of adsorbed molecules in zeolites have spurred important advances that can be attributed to both the development of efﬁcient algorithms and the massive increase of computer power. From a simulation point of view, computing diffusion coefﬁcients is challenging, and several reviews on this topic have been published recently [7–9, 121, 122]. The ﬁrst simulation studies for diffusion in conﬁned systems focused on self-diffusivity [123] calculations for a single component using equilibrium MD simulations. With growing computer power, the simulation studies have gradually shifted toward computing self-diffusivities for mixtures [124–126] as well as transport diffusivities [127–134] that are the relevant diffusion coefﬁcients for use in technological applications. Transport diffusivities can be obtained from three equivalent approaches that correspond to the Fick, the Onsager, and the Maxwell Stefan formulations [129, 135, 136]. The ﬁrst attempts to compute transport diffusivities in zeolites were achieved for methane in MFI [130] and LTA [131] using nonequilibrium MD methods. Most recently, the increase of computer power made it possible to compute transport diffusivities from equilibrium MD simulations [132–134, 137, 138], and also to extend those calculations for obtaining diffusivities over loading, providing insights into the mechanisms that control the molecular trafﬁc along the zeolite pores [139, 140]. Molecular simulations also play an important role in the understanding of multicomponent diffusion in zeolites by the development of theoretical models to predict multicomponent diffusion from single-component data [136, 141, 142]. This knowledge is essential for industrial applications, and experimental measurements for multicomponent adsorption and diffusion are difﬁcult with little available data in literature. A good example of the usefulness of molecular simulations in this context was reported by Skoulidas and Sholl [143]. Using MD simulations, they analyzed the effect of molecular loading and pore topology of the self- and transport diffusivities for a variety of molecules in four zeolites (MFI, MTW, ISV, and ITE). On the basis of this, Beerdsen et al. were able to classify the pore topologies by matching the free energy of the molecules with the zeolite structure [144].

12.3 Simulation Methods Table 12.1

Zeolite pore volumes obtained from molecular simulations.

Zeolite

Pore volume (cm3 g–1 )

Zeolite

Pore volume (cm3 g–1 )

Zeolite

Pore volume (cm3 g–1 )

ABW AEL AET AFG AFI AFO AFR AFS AFT AFX AFY AHT ANA APC APD AST ATN ATO ATT ATV AWW BEA (pol. A) BEA (pol. B) BEC BOG BPH CAN CAS CHA CHI CLO DDR DFO

0.0211 0.0902 0.1424 0.1287 0.1620 0.0875 0.2386 0.2897 0.1823 0.2199 0.3300 0.0459 0.0150 0.0331 0.0031 0.2274 0.0938 0.0810 0.1170 0.0279 0.1801 0.2763 0.2721 0.3289 0.2407 0.3235 0.1278 0.0155 0.2425 0.0069 0.3279 0.1400 0.2915

DOH DON EAB ECR EDI EMT ERI EUO FAU FAU (NaX) FAU (NaY) FER GME IHW ISV ITE ITW IWS JBW KFI LEV LIO LOS LOV LTA (ITQ-29) LTA4A LTA5A LTL LTN MAZ MEI MEL MEP

0.1665 0.1661 0.1945 0.2275 0.0535 0.3423 0.2227 0.1458 0.3680 0.3268 0.3050 0.1469 0.2384 0.1310 0.2863 0.2271 0.0957 0.2443 0.0338 0.2327 0.2191 0.1811 0.2060 0.0323 0.2854 0.2568 0.2605 0.1685 0.1894 0.1718 0.2912 0.1546 0.1357

MER MFI MFS MON MOR MTN MTT MWW MTW NAT OFF PAU RHO RON SAS SAT SBE SFF SFG SGT SOD SOF STF STT STW THO TON TSC VFI VSV WEN – –

0.1133 0.1642 0.1321 0.0326 0.1501 0.1704 0.0733 0.2033 0.1109 0.0025 0.2238 0.1620 0.2517 0.0930 0.2575 0.1779 0.3408 0.2036 0.1390 0.1782 0.1314 0.1451 0.2017 0.1916 0.2045 0.0185 0.0913 0.3702 0.2967 0.0544 0.0735 – –

Sodalite cages and nonaccessible pockets were blocked during the simulation.

The facility to analyze the molecular motion in the zeolite at a molecular level made molecular simulations a powerful tool to shed light on difﬁcult problems such as single-ﬁle [145–147] and resonant diffusion [148], window [116, 117] and levitation [9] effects, or molecular trafﬁc control [139, 149–151] and molecular path control [119, 152]. The development of special simulation techniques to compute diffusion in those processes that take place in a timescale inaccessible to MD is a general problem that has attracted great attention in the last years. Hence,

345

346

12 Modeling of Transport and Accessibility in Zeolites

approaches based on a sequence of rare events such as Bennet-Chandler [153], Ruiz-Montero [154], or transition path sampling [155] and those more recent such as transition interface sampling [156], temperature-accelerated dynamics [157], or long timescale kinetics Monte Carlo [157] can all be considered as rare event sampling (RES) methods. Earlier RES studies in zeolites [2] date back to the 1970s and they have been increasing in complexity over the years. In the 1990s, RES methods were applied to compute molecular diffusion at inﬁnite dilution in MFI [158–160] and faujasite [161–163]. Those studies were later extended to other complex topologies such as LTA, LTL, ERI, and CHA not only for inﬁnite dilution [119, 152, 164–166] but also for low, medium, and high loadings [119, 152, 166–171]. 12.4 Molecular Modeling Applied to Processes Involving Zeolites

As shown in previous sections, molecular simulation is becoming a powerful tool for gaining an insight into the adsorption and diffusion processes taking place in the zeolite pores at a molecular level. Since zeolites are in widespread use in industrial and environmental applications, this section is devoted to emphasize some examples of the active role that molecular modeling is playing in technological processes as well as in green chemistry. 12.4.1 Applications in Technological Processes

The adsorption, diffusion, separation, and catalytic properties of a given zeolite depend on multiple factors such as the topology and the surface properties of the structure (hydrophobicity, hydrophilicity), the shape and size of its pores, the Si/Al ratio, the number, location, and type of nonframework cations it contains, and its level of hydration. Molecular simulations are being applied to the study of these properties and factors aiming to predict the zeolite materials features and their performance in industrial processes. 12.4.1.1 Molecular Modeling of Conﬁned Water in Zeolites A current challenge is to provide information at a molecular level of the complex behavior of water when it is conﬁned into the zeolites pores. This is the ﬁrst step toward understanding the role of water in technological processes such as those involving ion exchange to reduce water hardness, for water/alcohol separations, and for water removal. Molecular modeling can be used to provide knowledge on the behavior of water conﬁned in a given zeolite and also on the way water inﬂuences the properties of the structure. The effect of conﬁned water on the properties of the zeolites depends not only on the level of hydration but also on the topology and chemical composition of the structure. Pure silica zeolites are highly hydrophobic, whereas the isomorphic substitution of silicon with aluminum and the subsequent neutralization of the framework with cations make the structure hydrophilic [172].

12.4 Molecular Modeling Applied to Processes Involving Zeolites

Recent simulation studies on hydrophilic zeolites have been applied to analyze the effect that the hydration level exerts on the framework stability and on the positions and mobility of the framework cations, as well as to understand the underlying mechanisms involved in the coordination of the water to the nonframework cations and to the zeolite frame. Simulation studies on hydrophobic structures are mostly focused on the structural and dynamical characteristics of water conﬁned in the zeolite and on the search of optimal molecular sieves for gas separation in the presence of water [57, 70, 76, 95, 173–176]. Numerous studies of water in hydrophobic zeolites (mainly MFI) have been performed over the years concluding that water–water interactions prevail over water–zeolite interactions. Therefore, the topology of the zeolite greatly inﬂuences the water properties as it shapes the molecular aggregates inside the pores. This is clearly illustrated in Figure 12.6 that shows the snapshots that Puibasset and Pellenq [177] reported

Y Z

X

Figure 12.6 Snapshots of water adsorption in MFI at 300 K for the relative pressures P/P0 = 0.0886, 0.1477, 0.2045, and 0.3409. For shake of clarity, the zeolite framework was removed by the authors. Figure from [177].

347

348

12 Modeling of Transport and Accessibility in Zeolites

for water conﬁgurations in MFI at four relative pressures. For the lower pressure, the adsorbed molecules of water are isolated in small clusters containing a few molecules. An increase of pressure leads isolated molecules to get together forming clusters that merge into larger ones until it creates a single (inﬁnite) cluster at the higher pressure. Detailed information on simulation studies of water conﬁned in hydrophobic and hydrophilic zeolites can be found in the review of Bougeard and Smirnov [178] as well as in very recent publications [57, 177, 179, 180]. 12.4.1.2 Molecular Modeling of Hydrocarbons in Zeolites Zeolites are used for a variety of industrial applications. Among them, they are widely used in the petrochemical industry for isomerization and cracking of hydrocarbons [181–183]. Zeolites become catalytically active by substitution of aluminum for silicon into the framework. Each substitution creates a negative charge in the framework that is compensated by a counterion or a proton. The location of these counterions inﬂuences the adsorption and the catalytic properties for these materials [25, 26]. In addition, zeolites act as shape-selective catalysts with a high selectivity and reaction yield. The shape and size of their pores control both the way reactants enter the zeolite to the catalytically active sites and the way reaction products leave the zeolite. Since the catalytic and structural properties of zeolites depend on the diffusivity of the adsorbed molecules, many simulation studies are focused on adsorption and diffusion of hydrocarbons [10]. As described in Section 12.2, these studies are being performed using advanced simulation techniques that from day to day increase in both speed and efﬁciency. A large number of groups have computed adsorption and diffusion of alkanes in zeolites. These works as well as current developments on molecular modeling applied to shape-selective catalysis are compiled in a recent review of Smit and Maesen [10]. Many examples of hydrocarbon molecules adsorbed and converted by the zeolites can be found in the literature [10]. Among them, the so-called window effect is an interesting example where molecular modeling is helpful to shed light on the microscopic mechanisms involved in heterogeneous catalysis. Most hydrocracking processes result in a product distribution with a single maximum. However, in 1968, Chen et al. found that ERI-type zeolites yield a bimodal product distribution [184]. This window effect was attributed to the relative diffusion rates of the alkanes in the zeolite, though most recent measurements failed to reproduce these diffusion values [185, 186]. Molecular simulation calculations corroborated, in 2003, the existence of the controversial window effect, showing that diffusion rates can increase by orders of magnitude when the alkane chain length increases, so that the shape of the alkane is no longer commensurate with that of a zeolite cage [116, 118]. The ﬁrst simulation studies were performed in structures consisting of channels (OFF) and in structures consisting of cages separated by small windows but differing in the size and shape of the cages as well as in the orientation of the windows with respect to the cages (ERI, CHA, and LTA). In contrast to the channel-like structures, the latter showed a nonmonotonic periodic dependence of the Henry coefﬁcients and heats of adsorption. In addition, when a molecule is incommensurate with the cage structure, the diffusion rate increases by orders of

12.4 Molecular Modeling Applied to Processes Involving Zeolites

magnitude. The simulation studies showed that the maximum length for a given hydrocarbon to ﬁt in a single cage was of 13, 11, and 23 carbon atoms for ERI, CHA, and LTA, respectively. These values are directly related to the local minima in the Henry coefﬁcients and heats of adsorption and to the local maxima in the diffusion coefﬁcients [118]. The unusually low adsorption obtained for chains of alkanes similar or larger than the zeolite cage provided an alternative cracking mechanism that affords prediction of selectivity as a function of cage size and opens the possibility of length-selective cracking, where the obtained distribution is controlled by choosing zeolites with cages of suitable size [116–118]. 12.4.1.3 Molecular Modeling of Separation of Mixtures in Zeolites Most industrial applications of adsorption involve mixtures. However, compared with the amount of literature about experimental studies on single-component adsorption in zeolites, fewer experimental studies have been published on mixtures [187–189]. This is due to the experimental difﬁculties to accurately determine the molecular composition of the adsorbed phase [190]. When experimental data are unavailable, molecular modeling becomes a very useful tool to predict the adsorption of mixtures in the zeolite as well as to ﬁnd out the separation mechanisms that take place inside the pores. Molecular separations of mixtures in zeolites are based not only on the adsorption selectivity but also on the difference between the diffusivities of the components of the mixture. The selectivity for a given zeolite can be further optimized by changing the type and the amount of nonframework cations [15, 23, 55, 58, 191]. The ﬁrst simulation studies on this topic were reported by Beerdsen et al. showing that an increase in the number of sodium cations in MFI- and MOR-type structures leads to an increase in the selectivity for adsorbing linear (MFI) and branched alkanes (MOR), respectively [23]. Further simulation studies in MFI showed that the adsorption of hydrocarbons increases with decreasing atomic weight of the nonframework cation [15]. Zeolite-based separation processes involve both mixture adsorption and diffusion that are strongly interrelated. To illustrate the idea, we have computed free energy proﬁles of propane and propylene in ITW at 500 K (Figure 12.7). Differences in the free energy barrier indicate that propane is virtually excluded from the zeolite, whereas propylene is adsorbed, supporting experimental ﬁndings. Free energies were computed using available well-tested models, force ﬁelds, and methods [52, 53, 116]. Simulation studies on zeolite-based separations are mostly applied to analyze the zeolite efﬁciency on the separation of mixtures of hydrocarbons, alcohol–water solutions, and natural gas puriﬁcation. This section sums up current work performed on hydrocarbons and alcohol–water mixtures, whereas studies on natural gas puriﬁcation are addressed in Section 12.4.2.2. The separation of mixtures of hydrocarbons is an important activity in several technological and petrochemical processes. For a given separation task, CBMC simulations allow the efﬁcient screening of zeolite topologies based on selective adsorption and foster the development of novel separation processes exploiting entropy and enthalpy effects [192–196]. Hence, for mixtures of linear alkanes,

349

12 Modeling of Transport and Accessibility in Zeolites

12 10 Free energy (k BT )

350

8 6 4 2 0 −0.4

−0.2

0.0

0.2

0.4

Reaction coordinate q axis c Figure 12.7 Free energy proﬁles of propane (solid line) and propylene (dashed line) in ITW-type zeolite at 500 K.

smaller molecules are preferentially adsorbed at high loadings, because the few empty voids are easily ﬁlled by them (size entropy effect). For mixtures of isomers formed by linear and branched alkanes, selectivity is in favor of those that pack more efﬁciently within the zeolite channel (conﬁgurational and length entropy effects). For a detailed overview on molecular simulation studies of mixtures of hydrocarbons in zeolites, the reader is referred to the recent review of Smit and Maesen [10]. An example where a mixture separation can be mainly attributed to enthalpic effects is reported by Ban et al. [197]. The authors computed adsorption isotherms for mixtures of benzene and propylene in several zeolites ﬁnding very high selectivities for benzene in structures formed by channels such as MOR, BEA, and MWW. The selectivity for benzene decreases in zeolites with large cavities such as NaY and it is even lower in MFI, formed by intersecting channels, with a reversal of the selectivity in favor of propylene for 373 K. The selectivity behavior can be explained based on enthalpic effects; in MOR, both propylene and benzene ﬁt in the main channel. However, benzene is preferentially adsorbed as its heat of adsorption is much larger. Similar features are found for BEA and MWW that almost exclusively adsorb benzene, independent of temperature. NaY shows larger cavities where the heat of adsorption of benzene decreases. This allows the adsorption of some molecules of propylene decreasing the adsorption selectivity for benzene. In MFI, benzene shows similar behaviour than in NaY. It is preferentially adsorbed at the intersections whereas propylene – that in absence of benzene can adsorb at both the intersections and the channels – is adsorbed in the channels (Figure 12.8). The location of benzene at the intersections is a key factor in the catalysts performance because of ‘‘trafﬁc junction’’ effects [198]. Hansen et al. presented a study using a combination of quantum chemistry, MC and MD simulations to

12.4 Molecular Modeling Applied to Processes Involving Zeolites

(a)

(b)

Figure 12.8 Snapshots of an equimolar mixture benzene/propylene adsorbed in MFI at low (a) and high (b) loading. Reprinted with permission from Ban et al. [197]. Copyright 2009 American Chemical Society.

predict performance for reaction of benzene with propylene using MFI catalysts [199]. These simulations are helpful in interpreting published experimental data. Zeolites are also widely used for solvent dehydration, since their polar nature make them optimal structures to separate water from organic compounds. The technological interest in ﬁnding zeolite membranes with high permselectivity of water to alcohols is leading researchers to study the processes involved on the water–alcohols separations using molecular simulations [71, 200, 201]. Hence, the water–alcohol adsorption and diffusion behavior in zeolites has been recently studied. Lu et al. [201] used GCMC simulations for the water–alcohol mixtures in MFI, MOR, CFI, and DON, and Yang et al. [71] analyzed the diffusion behavior of the mixture in MFI with equilibrium MD techniques. In spite of the improvements achieved on this topic, more efﬁcient and realistic models still need to be developed as the overall agreement with equilibrium experimental data is only qualitative. Detailed information on the most recent models as well as on previous available models can be found in [202]. 12.4.2 Applications in Green Chemistry

Green chemistry can be deﬁned as the use of chemistry to reduce or eliminate hazardous substances. Among a number of examples, zeolites may play a useful role carbon dioxide capture and natural gas puriﬁcation. This section explains the role that molecular simulations are currently playing in understanding the mechanisms involved in these processes at a molecular level. 12.4.2.1 Carbon Dioxide Capture The prediction of carbon dioxide adsorption on porous materials is becoming an imperative task not only for the need to develop cheap carbon dioxide capture

351

12 Modeling of Transport and Accessibility in Zeolites 70 LTA4A (Exp) [167] LTA4A (Sim) [28] LTA4A (Sim) [29] LTA4A (Sim) [33] NaY (Exp) [33] NaY (Sim) [29] NaY (Sim) [33]

60 Loading (molecules / uc)

352

50 40 30 20 10 0

1

10

100

Pressure (kPa)

Figure 12.9 Experimental [56, 207] and computed (using available force ﬁelds [28, 29, 56]) adsorption isotherms of carbon dioxide in LTA4A (with 96 aluminum atoms and sodium cations per unit cell) and in NaY (with 54 aluminum atoms and sodium cations per unit cell) at 298 K.

technologies but also for the improvement of gas separation processes such as natural gas puriﬁcation. Recent studies have found that carbon dioxide is strongly adsorbed by a variety of zeolites and that the adsorption efﬁciency strongly depends on the zeolite type and composition [56, 203–205]. For instance, the zeolite basicity and electric ﬁeld are essential factors on carbon dioxide adsorption that can be either induced or controlled by the nature and density of the cations within the zeolite pores [206]. At this point, molecular simulation can be extremely useful to systematically analyze the performance of the structures as a function of its size and shape, the type of the pore, the aluminum composition, and the nature of the nonframework cations. However, this task is seriously restricted by the lack of good and transferable force ﬁelds. Most of the works reporting molecular studies on carbon dioxide in zeolites are for all-silica structures and the few works reporting simulations in zeolites containing aluminum atoms compensate the net negative charge of the framework with sodium nonframework cations [28–30, 56]. Although force ﬁeld development for nonframework cations other than sodium is still an open task, Garcia-Sanchez et al. [56] developed a general force ﬁeld that outperforms previous force ﬁelds as it is more accurate, transferable between zeolite structures, and applicable to all Si/Al ratios. The adsorption isotherms computed with available force ﬁelds [28, 29, 56] and compared with experimental data [56, 207] are shown in Figure 12.9. 12.4.2.2 Natural Gas Puriﬁcation The natural gas composition is around 95% methane, traces of heavier gaseous hydrocarbons such as ethane and propane, and other light gasses such as carbon

12.5 Summary and Outlook

dioxide and nitrogen. The presence of carbon dioxide in natural gas reduces the combustion power efﬁciency and contributes to greenhouse gas emissions. Therefore, the production of cheap and clean fuel from natural gas is based on efﬁcient puriﬁcation processes. Zeolites are suitable materials for storage, separation, and puriﬁcation of natural gas for their thermal and mechanical stability, high carbon dioxide adsorption capacities, high rates of transport, and high selectivities. However, the mechanisms to identity the zeolite properties and performances on natural gas puriﬁcation are not well deﬁned yet and molecular simulations are currently being applied to make up for this deﬁciency. Several simulation studies have been reported on the adsorption and diffusion of the natural gas components. They have investigated the idea of analyzing the inﬂuence of several factors such as temperature, pressure, chemical composition, zeolite topology, and pore size on the adsorption [204, 208–214] and transport [124, 203, 205, 215–218] of these mixtures. For example, Krishna et al. reported molecular simulations for the self-diffusion of pure and equimolar mixture of carbon dioxide and methane in DDR, CHA, and MFI structures [205], and in a follow-up study, by using MC and MD simulations, screened 12 zeolite topologies looking for the optimal structure to separate the mixture, thereby ﬁnding the highest permeation selectivities for DDR and CHA zeolites [214]. Furthermore, they observed that the separation selectivity on these two structures is particularly enhanced by what they called the segregated nature of the mixture during adsorption. The segregation effect results in a preferential adsorption of carbon dioxide at the windows regions, hindering the diffusion of methane between cages and improving in this way, the selectivity [218]. These studies may lead to future strategies to improve natural gas separation processes.

12.5 Summary and Outlook

Molecular modeling is already a powerful tool to accurately predict transport and accessibility in zeolites. However, efﬁcient methods and good force ﬁelds capable of reproducing ideal experimental conditions for all zeolites are essential for this purpose. A number of simulation studies have been reported to exploit practical issues in zeolites. MC simulations predict adsorption isotherm, Henry coefﬁcients, isosteric heats of adsorption, cation distribution, preferable adsorption sites, and siting of molecules within the pores that are in good agreement with experiment. For diffusion, there is also good agreement with experimental data available, though most of the simulation work in diffusion focuses on self-diffusivities and methods to compute the corrected diffusivities for slow-diffusion systems are still scarce. Much progress has been made on the development of fast and efﬁcient methods and accurate and transferable force ﬁelds. However, signiﬁcant challenges remain in these areas. Future work is also expected to focus on the development of accurate potential models, particularly for different type of nonframework cations and for

353

354

12 Modeling of Transport and Accessibility in Zeolites

complex adsorbates. Finally, another on-going challenge is to combine methods based on both quantum and classical mechanics to the study of chemisorption and catalysis. The development of methods to adequately integrate diffusion, adsorption, and reaction kinetics will lead to the prediction of properties of new high-performance materials with increasing efﬁciency and speed. The microscopic information obtained from simulations provides the underlying knowledge from a molecular point of view that may guide to the development of more efﬁcient processes, to ﬁne-tune zeolites for a particular application, and also to steer the experimental effort in successful directions.

Acknowledgments

This work is supported by the Spanish ‘‘Ministerio de Educaci´on y Ciencia (MEC)’’ (CTQ2007-63229) and Junta de Andaluc´ıa (P07-FQM-02595). The author thanks J. M. Castillo, E. Garcia-Perez, A. Garcia-Sanchez, J. J. Gutierrez-Sevillano, and A. Martin-Calvo for their help with ﬁgures and especially for their joint work to provide the computed values listed in Table 12.1. The author is also thankful to T. J. H. Vlugt and S. Ban for Figure 12.8 and to D. Dubbeldam, T. J. H. Vlugt, J. A. Anta, and R. Krishna for a critical reading of the manuscript.

References 1. Karger, J. and Ruthven, D.M. (1992)

2.

3. 4. 5.

6.

Diffusion in Zeolites, John Wiley & Sons, Ltd, New York. Theodorou, D.N., Snurr, R.Q., and Bell, A.T. (1996)Molecular dynamics and diffusion in microporous materials,in Comprehensive Supramolecular Chemistry: Solid-state Supramolecular Chemistry–Two- and Three-dimensional Inorganic Networks, vol. 7, Pergamon, Oxford, p. 507–548. Keil, F.J., Krishna, R., and Coppens, M.O. (2000) Rev. Chem. Eng., 16, 71. Fuchs, A.H.and Cheetham, A.K. (2001) J. Phys. Chem. B, 105, 7375. Ramanan, H.and Auerbach, S.M. (2006)in Conference of the NATO-Advanced-Study-Institute on Fluid Transport in Nanoporous Materials (eds W.C.Conner and J. Fraissard), La Colle sur Loup, France, p. 93. Auerbach, S.M. (2006)in Conference of the NATO-Advanced-Study-Institute on Fluid Transport in Nanoporous Materials (eds W.C.Conner and J. Fraissard), La Colle sur Loup, France, p. 535.

7. Dubbeldam, D.and Snurr, R.Q.

8.

9. 10. 11.

12.

13. 14. 15.

(2006)3rd International Conference on Foundations of Molecular Modeling and Simulation (FOMMS), Blaine, p. 305. Jobic, H.and Theodorou, D.N. (2007) Microporous Mesoporous Mater., 102, 21. Yashonath, S.and Ghorai, P.K. (2008) J. Phys. Chem. B, 112, 665. Smit, B.and Maesen, T.L.M. (2008) Chem. Rev., 108, 4125. Baerlocher, C., Meier, W.M., and Olson, D.H. (2007) Atlas of Zeolite Structure Types, 6th revised edn, Elsevier, London. Auerbach, S.M., Bull, L.M., Henson, N.J., Metiu, H.I., and Cheetham, A.K. (1996) J. Phys. Chem., 100, 5923. Buttefey, S., Boutin, A., and Fuchs, A.H. (2002) Mol. Simul., 28, 1049. Dempsey, E., Kuehl, G.H., and Olson, D.J. (1969) Phys. Chem., 73, 387. Beerdsen, E., Dubbeldam, D., Smit, B., Vlugt, T.J.H., and Calero, S. (2003) J. Phys. Chem. B, 107, 12088.

References 16. Calero, S., Dubbeldam, D., Krishna, R.,

17. 18.

19. 20. 21.

22.

23. 24.

25.

26.

27. 28. 29. 30. 31.

32. 33. 34. 35.

Smit, B., Vlugt, T.J.H., Denayer, J.F.M., Martens, J.A., and Maesen, T.L.M. (2004) J. Am. Chem. Soc., 126, 11377. L¨owenstein, W. (1954) Am. Mineral., 39, 92. Garcia-Perez, E., Torrens, I.M., Lago, S., Dubbeldam, D., Vlugt, T.J.H., Maesen, T.L.M., Smit, B., Krishna, R., and Calero, S. (2005) Appl. Surf. Sci., 252, 716. Sastre, G., Fornes, V., and Corma, A. (2002) J. Phys. Chem. B, 106, 701. Sastre, G., Fornes, V., and Corma, A. (2002) J. Phys. Chem. B, 106, 701. Sklenak, S., Dedecek, J., Li, C., Wichterlova, B., Gabova, V., Sierka, M., and Sauer, J. (2009) Phys. Chem. Chem. Phys., 11, 1237. Macedonia, M.D., Moore, D.D., and Maginn, E.J. (2000) Langmuir, 16, 3823. Beerdsen, E., Smit, B., and Calero, S. (2002) J. Phys. Chem. B, 106, 10659. Mellot-Draznieks, C., Buttefey, S., Boutin, A., and Fuchs, A.H. (2001) Chem. Commun., 2200. Garcia-Perez, E., Dubbeldam, D., Liu, B., Smit, B., and Calero, S. (2007) Angew. Chem. Int. Ed., 46, 276. Liu, B., Garcia-Perez, E., Dubbeldam, D., Smit, B., and Calero, S. (2007) J. Phys. Chem. C, 111, 10419. Jaramillo, E.and Auerbach, S.M. (1999) J. Phys. Chem. B, 103, 9589. Jaramillo, E.and Chandross, M. (2004) J. Phys. Chem. B, 108, 20155. Maurin, L.P.and Bell, R.G. (2005) J. Phys. Chem. B, 109, 16084. Akten, E.D., Siriwardane, R., and Sholl, D.S. (2003) Energy Fuels, 17, 977. Rappe, A.K., Casewit, C.J., Colwell, K.S., Goddard, W.A., and Skiff, W.M. (1992) J. Am. Chem. Soc., 114, 10024. Hagler, A.T.and Ewig, C.S. (1994) Comput. Phys. Commun., 84, 131. Allinger, N.L. (1977) J. Am. Chem. Soc., 99, 8127. Allinger, N.L., Yuh, Y.H., and Lii, J.H. (1989) J. Am. Chem. Soc., 111, 8551. Allinger, N.L., Li, F.B., Yan, L.Q., and Tai, J.C. (1990) J. Comput. Chem., 11, 868.

36. Allinger, N.L., Chen, K.S., and Lii, J.H.

(1996) J. Comput. Chem., 17, 642. 37. Allinger, N.L. (2002) Abstr. Pap. Am.

Chem. Soc., 223, 97. 38. Mayo, S.L., Olafson, B.D., and

39. 40.

41.

42.

43. 44.

45.

46.

47. 48.

49. 50.

51.

52.

53.

54.

Goddard, W.A. (1990) J. Phys. Chem., 94, 8897. Allured, V.S., Kelly, C.M., and Landis, C.R. (1991) J. Am. Chem. Soc., 113, 1. Root, D.M., Landis, C.R., and Cleveland, T. (1993) J. Am. Chem. Soc., 115, 4201. Weiner, S.J., Kollman, P.A., Case, D.A., Singh, U.C., Ghio, C., Alagona, G., Profeta, S., and Weiner, P. (1984) J. Am. Chem. Soc., 106, 765. Brooks, B.R., Bruccoleri, R.E., Olafson, B.D., States, D.J., Swaminathan, S., and Karplus, M. (1983) J. Comput. Chem., 4, 187. Jorgensen, W.L.and Tiradorives, J. (1988) J. Am. Chem. Soc., 110, 1657. Clark, M., Cramer, R.D., and Vanopdenbosch, N. (1989) J. Comput. Chem., 10, 982. Momany, F.A., McGuire, R.F., Burgess, A.W., and Scheraga, H.A. (1975) J. Phys. Chem., 79, 2361. Hermans, J., Berendsen, H.J.C., Vangunsteren, W.F., and Postma, J.P.M. (1984) Biopolymers, 23, 1513. Halgren, T.A. (1992) J. Am. Chem. Soc., 114, 7827. Burchart, E.D., Jansen, J.C., and Vanbekkum, H. (1989) Zeolites, 9, 432. Lifson, S., Hagler, A.T., and Dauber, P. (1979) J. Am. Chem. Soc., 101, 5111. Momany, F.A., Carruthe., L.M., McGuire, R.F., and Scheraga, H.A. (1974) J. Phys. Chem., 78, 1595. Dubbeldam, D., Calero, S., Vlugt, T.J.H., Krishna, R., Maesen, T.L.M., Beerdsen, E., and Smit, B. (2004) Phys. Rev. Lett., 93, 8. Dubbeldam, D., Calero, S., Vlugt, T.J.H., Krishna, R., Maesen, T.L.M., and Smit, B. (2004) J. Phys. Chem. B, 108, 12301. Liu, B., Smit, B., Rey, F., Valencia, S., and Calero, S. (2008) J. Phys. Chem. C, 112, 2492. Calero, S., Lobato, M.D., Garcia-Perez, E., Mejias, J.A., Lago, S., Vlugt,

355

356

12 Modeling of Transport and Accessibility in Zeolites

55.

56.

57.

58.

59.

60.

61.

62.

63. 64.

65.

66. 67. 68. 69. 70.

71.

T.J.H., Maesen, T.L.M., Smit, B., and Dubbeldam, D. (2006) J. Phys. Chem. B, 110, 5838. Garcia-Perez, E., Dubbeldam, D., Maesen, T.L.M., and Calero, S. (2006) J. Phys. Chem. B, 110, 23968. Garcia-Sanchez, A., Ania, C.O., Parra, J.B., Dubbeldam, D., Vlugt, T.J.H., Krishna, R., and Calero, S. (2009) J. Phys. Chem. C, 113, 8814–8820. Castillo, J.M., Dubbeldam, D., Vlugt, T.J.H., Smit, B., and Calero, S. (2009) Mol. Simul., 35, 1067–1076. Garcia-Sanchez, A., Garcia-Perez, E., Dubbeldam, D., Krishna, R., and Calero, S. (2007) Adsorption Sci. Technol., 25, 417. Bezus, A.G., Kiselev, A.V., Lopatkin, A.A., and Du, P.Q. (1978) J. Chem. Soc. Faraday Trans. 2, 74, 367. Demontis, P., Suffritti, G.B., Quartieri, S., Fois, E.S., and Gamba, A. (1988) J. Phys. Chem., 92, 867. Nicholas, J.B., Hopﬁnger, A.J., Trouw, F.R., and Iton, L.E. (1991) J. Am. Chem. Soc., 113, 4792. Leroy, F., Rousseau, B., and Fuchs, A.H. (2004) Phys. Chem. Chem. Phys., 6, 775. Vlugt, T.J.H.and Schenk, M. (2002) J. Phys. Chem. B, 106, 12757. Garcia-Perez, E., Parra, J.B., Ania, C.O., Dubbeldam, D., Vlugt, T.J.H., Castillo, J.M., Merkling, P.J., and Calero, S. (2008) J. Phys. Chem. C, 112, 9976. Zimmermann, N.E.R., Jakobtorweihen, S., Beerdsen, E., Smit, B., and Keil, F.J. (2007) J. Phys. Chem. C, 111, 17370. Bouyermaouen, A.and Bellemans, A. (1998) J. Chem. Phys., 108, 2170. Hill, J.R.and Sauer, J. (1994) J. Phys. Chem., 98, 1238. Hill, J.R.and Sauer, J. (1995) J. Phys. Chem., 99, 9536. Harris, J.G.and Yung, K.H. (1995) J. Phys. Chem., 99 (31), 12021. Di Lella, A., Desbiens, N., Boutin, A., Demachy, I., Ungerer, P., Bellat, J.P., and Fuchs, A.H. (2006) Phys. Chem. Chem. Phys., 8, 5396. Yang, J.Z., Chen, Y., Zhu, A.M., Liu, Q.L., and Wu, J.Y. (2008) J. Memb. Sci., 318, 327.

72. Guillot, B.and Guissani, Y. (2001) J.

Chem. Phys., 114, 6720. 73. Pellenq, R.J.M., Roussel, T., and

74. 75.

76.

77.

78.

79.

80. 81. 82. 83.

84.

85.

86.

87.

88. 89. 90.

Puibasset, J. (2008) J. Int. Adsorpt. Soc., 14, 733–742. Darkrim, F.and Levesque, D. (1998) J. Chem. Phys., 109, 4981. Garberoglio, G., Skoulidas, A.I., and Johnson, J.K. (2005) J. Phys. Chem. B, 109, 13094. Trzpit, M., Soulard, M., Patarin, J., Desbiens, N., Cailliez, F., Boutin, A., Demachy, I., and Fuchs, A.H. (2007) Langmuir, 23, 10131. Cailliez, F., Stirnemann, G., Boutin, A., Demachy, I., and Fuchs, A.H. (2008) J. Phys. Chem. C, 112, 10435. Wender, A., Barreau, A., Lefebvre, C., Di Lella, A., Boutin, A., Ungerer, P., and Fuchs, A.H. (2006) Adsorpt. Sci. Technol., 24, 713. Nath, S.K., Escobedo, F.A., and de Pablo, J.J. (1998) J. Chem. Phys., 108, 9905. Martin, M.G.and Siepmann, J.I. (1998) J. Phys. Chem. B, 102, 2569. Smit, B.and Siepmann, J.I. (1994) J. Phys. Chem., 98, 8442. Jorgensen, W.L.and Swenson, C.J. (1985) J. Am. Chem. Soc., 107, 1489. Vlugt, T.J.H., Zhu, W., Kapteijn, F., Moulijn, J.A., Smit, B., and Krishna, R. (1998) J. Am. Chem. Soc., 120, 5599. Lubna, N., Kamath, G., Potoff, J.J., Rai, N., and Siepmann, J.I. (2005) J. Phys. Chem. B, 109, 24100. Zeng, Y.P., Ju, S.G., Xing, W.H., and Chen, C.L. (2007) Sep. Purif. Technol., 55, 82. Zeng, Y.P., Ju, S.G., Xing, W.H., and Chen, C.L. (2007) Ind. Eng. Chem. Res., 46, 242. Maple, J.R., Hwang, M.J., Stockﬁsch, T.P., and Hagler, A.T. (1994) Isr. J. Chem., 34, 195. Zhao, Y.and Truhlar, D.G. (2008) J. Phys. Chem. C, 112, 6860. Martin, M.G.and Siepmann, J.I. (1999) J. Phys. Chem. B, 103, 4508. Smith, B.D.and Srivastava, R. (1986) Thermodynamic Data for Pure Compounds: Part A, Hydrocarbons and Ketones, Elsevier, Amsterdam.

References 91. Stoll, J., Vrabec, J., and Hasse, H.

92. 93. 94. 95.

96.

97.

98.

99.

100. 101. 102. 103. 104. 105. 106. 107.

108.

109. 110. 111.

(2003)GVC/DECHEMA Annual Meeting, Mannhein, Germany, p. 891. Vrabec, J., Stoll, J., and Hasse, H. (2001) J. Phys. Chem. B, 105, 12126. Wick, C., Martin, M., and Siepmann, J.I. (2000) J. Phys. Chem. B, 104, 8008. Skoulidas, A.I.and Sholl, D.S. (2002) J. Phys. Chem. B, 106, 5058. Desbiens, N., Boutin, A., and Demachy, I. (2005) J. Phys. Chem. B, 109, 24071. Desbiens, N., Demachy, I., Fuchs, A.H., Kirsch-Rodeschini, H., Soulard, M., and Patarin, J. (2005) Angew. Chem. Int. Ed., 44, 5310. Frenkel, D.and Smit, B. (2002) Understanding Molecular Simulations: From Algorithms to Applications, 2nd edn, Academic Press, San Diego. Allen, M.P.and Tildesley, D.J. (1987) Computer Simulations of Liquids, Clarendon Press, Oxford. Bussai, C., Fritzsche, S., Haberlandt, R., and Hannongbua, S. (2004) J. Phys. Chem. B, 108, 13347. Schroder, K.P.and Sauer, J. (1996) J. Phys. Chem., 100, 11043. Talu, O.and Myers, A.L. (2001) AIChE J., 47, 1160. Myers, A.L.and Monson, P.A. (2002) Langmuir, 18, 10261. Chen, H.B.and Sholl, D.S. (2006) Langmuir, 22, 709. Shen, V.K.and Errington, J.R. (2005) J. Chem. Phys., 122, 8. Karavias, F.and Myers, A.L. (1991) Langmuir, 7, 3118. Vlugt, T.J.H., Krishna, R., and Smit, B. (1999) J. Phys. Chem. B, 103, 1102. Woods, G.B., Panagiotopoulos, A.Z., and Rowlinson, J.S. (1988) Mol. Phys., 63, 49. Vlugt, T.J.H., Garcia-Perez, E., Dubbeldam, D., Ban, S., and Calero, S. (2008) J. Chem. Theory Comput., 4, 1107. Harris, J.and Rice, S.A. (1988) J. Chem. Phys., 89 (9), 5898. Depablo, J.J., Laso, M., and Suter, U.W. (1992) J. Chem. Phys., 96, 6157. Siepmann, J.I.and Frenkel, D. (1992) Mol. Phys., 75, 59.

112. Frenkel, D., Mooij, G., and Smit, B.

113.

114.

115. 116.

117.

118. 119.

120. 121. 122. 123. 124. 125. 126.

127. 128.

129. 130.

131.

132.

(1992) J. Phys. Condens. Matter, 4, 3053. Consta, S., Vlugt, T.J.H., Hoeth, J.W., Smit, B., and Frenkel, D. (1999) Mol. Phys., 97, 1243. Consta, S., Wilding, N.B., Frenkel, D., and Alexandrowicz, Z. (1999) J. Chem. Phys., 110, 3220. Houdayer, J. (2002) J. Chem. Phys., 116, 1783. Dubbeldam, D., Calero, S., Maesen, T.L.M., and Smit, B. (2003) Phys. Rev. Lett., 90, 245901. Dubbeldam, D., Calero, S., Maesen, T.L.M., and Smit, B. (2003) Angew. Chem. Int. Ed., 42, 3624. Dubbeldam, D.and Smit, B. (2003) J. Phys. Chem. B, 107, 12138. Dubbeldam, D., Beerdsen, E., Calero, S., and Smit, B. (2005) Proc. Natl. Acad. Sci. U.S.A., 102, 12317. Walton, K.S.and Snurr, R.Q. (2007) J. Am. Chem. Soc., 129, 8552. Auerbach, S.M. (2000) Int. Rev. Phys. Chem., 19, 155. Sholl, D.S. (2006) Acc. Chem. Res., 39, 403. Demontis, P.and Suffritti, G.B. (1997) Chem. Rev., 97, 2845. Snurr, R.Q.and Karger, J. (1997) J. Phys. Chem. B, 101, 6469. Gergidis, L.N.and Theodorou, D.N. (1999) J. Phys. Chem. B, 103, 3380. Jost, S., Bar, N.K., Fritzsche, S., Haberlandt, R., and Karger, J. (1998) J. Phys. Chem. B, 102, 6375. Krishna, R.and van Baten, J.M. (2008) Chem. Eng. Sci., 63, 3120. Krishna, R.and van Baten, J.M. (2008) Microporous Mesoporous Mater., 109, 91. Krishna, R.and van Baten, J.M. (2009) Chem. Eng. Sci., 64, 870. Maginn, E.J., Bell, A.T., and Theodorou, D.N. (1993) J. Phys. Chem., 97, 4173. Fritzsche, S., Haberlandt, R., and Karger, J. (1995) Z. Phys. Chem.: Int. J. Res. Phys. Chem. Chem. Phys., 189, 211. Arya, G., Chang, H.C., and Maginn, E.J. (2001) J. Chem. Phys., 115, 8112.

357

358

12 Modeling of Transport and Accessibility in Zeolites 133. Sholl, D.S. (2000) Ind. Eng. Chem. Res., 134.

135. 136. 137.

138. 139. 140.

141. 142. 143. 144. 145. 146.

147.

148. 149. 150. 151. 152.

153. 154. 155.

39, 3737. Hoogenboom, J.P., Tepper, H.L., van der Vegt, N.F.A., and Briels, W.J. (2000) J. Chem. Phys., 113, 6875. Krishna, R.and Baur, R. (2003) Sep. Purif. Technol., 33, 213. Krishna, R.and van Baten, J.M. (2005) J. Phys. Chem. B, 109, 6386. Sastre, G., Catlow, C.R.A., Chica, A., and Corma, A. (2000) J. Phys. Chem. B, 104, 416. Sastre, G., Catlow, C.R.A., and Corma, A. (1999) J. Phys. Chem. B, 103, 5187. Derouane, E.G.and Gabelica, Z. (1980) J. Catal., 65, 486. Krishna, R., van Baten, J.M., and Dubbeldam, D. (2004) J. Phys. Chem. B, 108, 14820. Krishna, R.and van Baten, J.M. (2006) Chem. Phys. Lett., 420, 545. Skoulidas, A.I., Sholl, D.S., and Krishna, R. (2003) Langmuir, 19, 7977. Skoulidas, A.I.and Sholl, D.S. (2003) J. Phys. Chem. A, 107, 10132. Beerdsen, E., Dubbeldam, D., and Smit, B. (2006) Phys. Rev. Lett., 96, 4. Hahn, K., Karger, J., and Kukla, V. (1996) Phys. Rev. Lett., 76, 2762. Kukla, V., Kornatowski, J., Demuth, D., Gimus, I., Pfeifer, H., Rees, L.V.C., Schunk, S., Unger, K.K., and Karger, J. (1996) Science, 272, 702. Gupta, V., Nivarthi, S.S., Keffer, D., McCormick, A.V., and Davis, H.T. (1996) Science, 274, 164. Tsekov, R.and Evstatieva, E. (2005) Adv. Colloid Interface Sci., 114, 159. Brauer, P., Brzank, A., and Karger, J. (2003) J. Phys. Chem. B, 107, 1821. Clark, L.A., Ye, G.T., and Snurr, R.Q. (2000) Phys. Rev. Lett., 84, 2893. Harish, R., Karevski, D., and Schutz, G.M. (2008) J. Catal., 253, 191. Dubbeldam, D., Beerdsen, E., Calero, S., and Smit, B. (2006) J. Phys. Chem. B, 110, 3164. Chandler, D. (1978) J. Chem. Phys., 68, 2959. RuizMontero, M.J., Frenkel, D., and Brey, J.J. (1997) Mol. Phys., 90, 925. Dellago, C., Bolhuis, P.G., and Geissler, P.L. (2002) Adv. Chem. Phys., 123, 1.

156. van Erp, T.S., Moroni, D., and Bolhuis,

P.G. (2003) J. Chem. Phys., 118, 7762. 157. Sorensen, M.R.and Voter, A.F. (2000) J.

Chem. Phys., 112, 9599. 158. June, R.L., Bell, A.T., and Theodorou,

D.N. (1991) J. Phys. Chem., 95, 8866. 159. Forester, T.R.and Smith, W. (1997) J.

Chem. Soc., Faraday Trans., 93, 3249. 160. Maginn, E.J., Bell, A.T., and

161. 162.

163.

164.

165.

166.

167. 168. 169. 170. 171.

172.

173.

174.

175. 176.

Theodorou, D.N. (1996) J. Phys. Chem., 100, 7155. Jousse, F.and Auerbach, S.M. (1997) J. Chem. Phys., 107, 9629. Mosell, T., Schrimpf, G., and Brickmann, J. (1997) J. Phys. Chem. B, 101, 9476. Mosell, T., Schrimpf, G., and Brickmann, J. (1997) J. Phys. Chem. B, 101, 9485. Ghorai, P.K.R., Yashonath, S., and Lynden-Bell, R.M. (2002) Mol. Phys., 100, 641. Schuring, A., Auerbach, S.M., Fritzsche, S., and Haberlandt, R. (2002) J. Chem. Phys., 116, 10890. Dubbeldam, D., Beerdsen, E., Vlugt, T.J.H., and Smit, B. (2005) J. Chem. Phys., 122, 22. Nagumo, R., Takaba, H., and Nakao, S.I. (2008) J. Phys. Chem. C, 112, 2805. Gupta, A.and Snurr, R.Q. (2005) J. Phys. Chem. B, 109, 1822. Tunca, C.and Ford, D.M. (1999) J. Chem. Phys., 111, 2751. Tunca, C.and Ford, D.M. (2003) Chem. Eng. Sci., 58, 3373. Beerdsen, E., Smit, B., and Dubbeldam, D. (2004) Phys. Rev. Lett., 93, 24. Bowen, T.C., Noble, R.D., and Falconer, J.L. (2004) J. Memb. Sci., 245, 1. Channon, Y.M., Catlow, C.R.A., Gorman, A.M., and Jackson, R.A. (1998) J. Phys. Chem. B, 102, 4045. Demontis, P., Stara, G., and Suffritti, G.B. (2003) J. Phys. Chem. B, 107, 4426. Fleys, M.and Thompson, R.W. (2005) J. Chem. Theory Comput., 1, 453. Fleys, M., Thompson, R.W., and MacDonald, J.C. (2004) J. Phys. Chem. B, 108, 12197.

References 177. Puibasset, J.and Pellenq, R.J.M. (2008) 178. 179.

180.

181.

182.

183. 184. 185. 186.

187.

188.

189.

190.

191.

192. 193.

194. 195.

J. Phys. Chem. B, 112, 6390. Bougeard, D.and Smirnov, K.S. (2007) Phys. Chem. Chem. Phys., 9, 226. Ockwig, N.W., Cygan, R.T., Criscenti, L.J., and Nenoff, T.M. (2008) Phys. Chem. Chem. Phys., 10, 800. Demontis, P., Gulin-Gonzalez, J., Jobic, H., Masia, M., Sale, R., and Suffritti, G.B. (2008) Acs Nano, 2, 1603. Maesen, T.L.M., Krishna, R., van Baten, J.M., Smit, B., Calero, S., and Sanchez, J.M.C. (2008) J. Catal., 256, 95. Maesen, T.L.M., Beerdsen, E., Calero, S., Dubbeldam, D., and Smit, B. (2006) J. Catal., 237, 278. Maesen, T.L.M., Calero, S., Schenk, M., and Smit, B. (2004) J. Catal., 221, 241. Chen, N.Y., Lucki, S.J., and Mower, E.B. (1969) J. Catal., 13, 329. Magalhaes, F.D., Laurence, R.L., and Conner, W.C. (1996) AIChE J., 42, 68. Cavalcante, C.L., Eic, M., Ruthven, D.M., and Occelli, M.L. (1995) Zeolites, 15, 293. Talu, O., Li, J.M., and Myers, A.L. (1995) Adsorpt. J. Int. Adsorpt. Soc., 1, 103. Denayer, J.F., Ocakoglu, A.R., De Jonckheere, B.A., Martens, J.A., Thybaut, J.W., Marin, G.B., and Baron, G.V. (2003) Int. J. Chem. Reactor Eng., 1, A36. Denayer, J.F.M., Ocakoglu, R.A., Huybrechts, W., Dejonckheere, B., Jacobs, P., Calero, S., Krishna, R., Smit, B., Baron, G.V., and Martens, J.A. (2003) J. Catal., 220, 66. Ruthven, D.M. (1984) Principles of Adsorption and Adsorption Processes, Wiley-Interscience, New York. Lachet, V., Boutin, A., Tavitian, B., and Fuchs, A.H. (1999) Langmuir, 15, 8678. Krishna, R., Smit, B., and Vlugt, T.J.H. (1998) J. Phys. Chem. A, 102, 7727. Calero, S., Smit, B., and Krishna, R. (2001) Phys. Chem. Chem. Phys., 3, 4390. Krishna, R., Calero, S., and Smit, B. (2002) Chem. Eng. J., 88, 81. Krishna, R., Smit, B., and Calero, S. (2002) Chem. Soc. Rev., 31, 185.

196. Calero, S., Smit, B., and Krishna, R.

(2001) J. Catal., 202, 395. 197. Ban, S., Van Laak, A., De Jongh, P.E.,

198. 199.

200.

201.

202.

203.

204.

205.

206.

207.

208.

209. 210. 211. 212.

213.

Van der Eerden, J., and Vlugt, T.J.H. (2007) J. Phys. Chem. C, 111, 17241. Krishna, R.and van Baten, J.M. (2008) Chem. Eng. J., 140, 614. Hansen, N., Krishna, R., van Baten, J.M., Bell, A.T., and Kew, F.J. (2009) J. Phys. Chem. C, 113, 235. Kuhn, J., Castillo, J.M., Gascon, J., Calero, S., Dubbeldam, D., Vlugt, T.J.H., Kapteijn, F., and Gross, J. (2009) J. Phys. Chem. C, 113, 14290–14301. Lu, L.H., Shao, Q., Huang, L.L., and Lu, X.H. (2007)11th International Conference on Properties and Phase Equilibria for Product and Process Design, Crete, Greece, p. 191. Rutkai, G., Csanyi, E., and Kristof, T. (2008) Microporous Mesoporous Mater., 114, 455. Krishna, R., van Baten, J.M., Garcia-Perez, E., and Calero, S. (2007) Ind. Eng. Chem. Res., 46, 2974. Garcia-Perez, E., Parra, J.B., Ania, C.O., Garcia-Sanchez, A., Van Baten, J.M., Krishna, R., Dubbeldam, D., and Calero, S. (2007) Adsorpt. J. Int. Adsorpt. Soc., 13, 469. Krishna, R., van Baten, J.M., Garcia-Perez, E., and Calero, S. (2006) Chem. Phys. Lett., 429, 219. Bonenfant, D., Kharoune, M., Niquette, P., Mimeault, M., and Hausler, R. (2008) Sci. Technol. Adv. Mater., 9. Ahn, H., Moon, J.H., Hyun, S.H., and Lee, C.H. (2004) Adsorpt. J. Int. Adsorpt. Soc., 10, 111. Goj, A., Sholl, D.S., Akten, E.D., and Kohen, D. (2002) J. Phys. Chem. B, 106, 8367. Yue, X.P.and Yang, X.N. (2006) Langmuir, 22, 3138. Heuchel, M., Snurr, R.Q., and Buss, E. (1997) Langmuir, 13, 6795. Jia, Y.X., Wang, M., Wu, L.Y., and Gao, C.J. (2007) Sep. Sci. Technol., 42, 3681. Babarao, R., Hu, Z.Q., Jiang, J.W., Chempath, S., and Sandler, S.I. (2007) Langmuir, 23, 659. Ghouﬁ, A., Gaberova, L., Rouquerol, J., Vincent, D., Llewellyn, P.L., and

359

360

12 Modeling of Transport and Accessibility in Zeolites Maurin, G. (2009) Microporous Mesoporous Mater., 119, 117. 214. Krishna, R.and van Baten, J.M. (2007) Chem. Eng. J., 133, 121. 215. Sanborn, M.J.and Snurr, R.Q. (2000) Sep. Purif. Technol., 20, 1. 216. Papadopoulos, G.K., Jobic, H., and Theodorou, D.N. (2004) J. Phys. Chem. B, 108, 12748.

217. Babarao, R.and Jiang, J.W. (2008) Lang-

muir, 24, 5474. 218. Krishna, R.and van Baten, J.M. (2008)

Sep. Purif. Technol., 61, 414.

361

13 Diffusion in Zeolites – Impact on Catalysis Johan van den Bergh, Jorge Gascon, and Freek Kapteijn

13.1 Introduction

Zeolite catalysts and adsorbents are widely accepted in industry. In 1948, commercial adsorbents based on synthetic aluminosilicates zeolite A and X were available [1]. Zeolite Y as FCC catalyst was commercially available in [2]. Aluminum-containing zeolites are inherently catalytically active in several ways. The isomorphic substituted aluminum atom within the zeolite framework has a negative charge that is compensated by a counterion. When the counterion is a proton, a Brønsted acid site is created. Moreover, framework oxygen atoms can give rise to weak Lewis base activity. Noble- metal ions can be introduced by an ion exchange with the cations after synthesis. Incorporation of metals like Ti, V, Fe, and Cr in the framework can provide the zeolite with activity for redox reactions. A well-known example of the latter type is titanium silicalite-1 (TS-1): a redox molecular sieve catalyst [3]. Not only the catalytic activity makes zeolites particularly interesting but also the location of the active sites within the well-deﬁned geometry of a zeolite. Because of the geometrical constraints of the zeolite, the selectivity of a chemical reaction can be increased by three mechanisms: reactant selectivity, product selectivity, and transition state selectivity. In the case of reactant selectivity, bulky components in the feed do not enter the zeolite and will have no opportunity to react. When several products are formed within the zeolite, but only some are able to leave the zeolite, or some leave the zeolite more rapidly, the process is named product selectivity. When the geometrical constraints of the active site within the zeolite prohibit the formation of products or transition states leading to certain products, transition state selectivity is obtained [4, 5]. The importance of diffusion in (zeolite) catalysts arises from the fact that the catalytically active site needs to be reached by the reactants, and products need to move away from this site. As depicted in Figure 13.1 for a packed bed conﬁguration, reactants need to move from the bulk to the active site. As commonly applied in catalysis, the macroporous (dpore > 50 nm) particles at the bed level consist of pelleted smaller particles (crystals in the case of zeolites) with Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 1. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

362

Packed bed (m)

13 Diffusion in Zeolites – Impact on Catalysis Pelleted particles /crystals (mm)

Macro and meso pores in pellets

Liquid or gas film Individual crystals layer around pellets (µm)

Interparticle space

Zeolite pores and channels (0.4 – 1.0 nm)

Zeolite micro pores

Figure 13.1 The reactant pathway from bulk to active site encompasses diffusion processes on a broad length scale in common catalytic processes. (After Ruthven [6].)

meso- (2 nm < dpore < 50 nm) or even micropores (dpore < 2 nm). External resistances, which can stand in the way of optimal catalysis, are ﬁlm layers around the particles that can introduce external heat and mass transfer limitations. Additionally, macropores in the pelleted particles and meso- and micropores in the small particles or crystals can induce internal (diffusional) mass transport limitations. Furthermore, internal heat transfer limitations can occur due to poor heat removal from the (pelleted) particles. Most transport limitations and associated catalyst effectiveness are based on solving the reaction–diffusion problem in catalyst particles. Transformation of the governing differential equation yields a dimensionless parameter, the Thiele modulus, expressing the ratio of the kinetics and diffusion rates for the particle. This reaction–diffusion problem and the Thiele concept are textbook material [7] and are not discussed in this chapter. Because of the small pore size of zeolite catalysts, diffusion limitations may quickly arise. This chapter focuses on internal diffusion (limitations) in zeolite catalyst particles that considerably deviates from diffusion in meso- and macroporous materials. The impact of these deviations with respect to the applicability of the Thiele concept to estimate the catalyst effectiveness is evaluated. Furthermore, results of recently introduced spatially resolved measurement techniques are discussed in relation to the performance of a zeolite catalyst at the subcrystal level. Finally, the state-of-the-art approaches to overcome, prevent, or utilize diffusion limitations are discussed.

13.2 Diffusion and Reaction in Zeolites: Basic Concepts

Diffusion in the gas phase or in relative large pores (>100 nm [8]) is dominated by intermolecular collisions, and the ﬂux of component i can be described by

13.2 Diffusion and Reaction in Zeolites: Basic Concepts

the Maxwell–Stefan (MS) approach [9], in which forces acting on molecules (in diffusional processes, the gradient in thermodynamic potential) are balanced by the friction between molecules and, in case of porous materials, with a solid. In the latter case, this was coined the ‘‘dusty gas model.’’ This model was extended to zeolites by Krishna [10]. The often used Fick’s law is a simpliﬁcation of the generalized MS equations for thermodynamically ideal systems [9]. In the case of a porous material, a correction needs to be made to account for the porosity (ε) and tortuosity (τ ) of the material, leading to an ‘‘effective’’ diffusivity. ε Ni = − Di ∇Ci = −Deff i ∇Ci τ

(13.1)

In porous materials, when the mean free path of a molecule is in the order of or larger than the pore diameter (∼10–100 nm [8]), molecule-wall collisions start to dominate and the diffusivity can be described by the Knudsen diffusion mechanism. A ﬂux in such small pores can be presented as ε Ni = − Dkn, i ∇Ci , τ

Dkn, i

d0 = 3

8RT πMi

(13.2)

In the case of zeolites, the pores approach molecular dimensions (∼0.3–0.74 nm) and, consequently, mass transport through zeolite pores is determined by the interaction of the molecules with the zeolite pore wall (see also Chapter 24 on modeling transport in zeolites). Now molecules are adsorbed on the zeolite, lose their gaseous nature, and transport is often referred to as surface or zeolitic diffusion [8]. The ﬂux can now also be represented in a Fickian way; the concentration (qi ) represents the adsorbed amount on the zeolite or loading. A common unit for the loading is moles per kilogram; therefore, the zeolite density (ρ) is added to arrive at consistent dimensions: Ni = ρDi ∇qi

(13.3)

The tortuosity and porosity presented in Eq. (13.1) are not speciﬁed in Eq. (13.3); these are an inherent property of the diffusivity. Each zeolite has its own speciﬁc pore network with its own tortuosity and porosity: some zeolites are characterized by channels and intersections (e.g., MFI); others resemble systems of cages connected by windows (e.g., faujasite (FAU) and Linde type A (LTA)). Moreover, the pore network can be one-, two-, or three-dimensional with different pore sizes or connectivity in different directions, leading to diffusion anisotropy. Since zeolites are crystalline materials, this may imply that some crystal facets are better accessible than others, signiﬁcantly inﬂuencing local transport and potentially giving rise to concentration proﬁles inside the zeolite.

363

364

13 Diffusion in Zeolites – Impact on Catalysis

13.2.1 Importance of Adsorption

The adsorbed phase (qi ) in Eq. (13.3) is related to the gas phase fugacity through an adsorption isotherm, of which the classical example is the Langmuir isotherm: qi =

qsat i Ki fi 1 + Ki fi

(13.4)

An important difference between gas phase and adsorbed phase diffusion is the concentration level, being much higher in the case of adsorbed phase diffusion. When the gradient in chemical potential is taken as the fundamental driving force for diffusion [8, 9], a correction needs to be made to Eq. (13.3). Now, a so-called thermodynamic correction factor (i ) is introduced; the diffusivity is referred to as corrected or Maxwell–Stefan (MS) diffusivity:

-

-

Ni = −ρDi ∇ln fi = −ρDi i ∇qi ,

i =

d ln fi d ln qi

(13.5)

For a single-site Langmuir isotherm, the thermodynamic correction factor is given by i =

d ln fi 1 = , d ln qi 1 − θi

θi =

qi qsat i

(13.6)

In the limit of low loading, the thermodynamic correction factor approaches 1 and the MS and Fickian diffusivity are equal. Although the MS diffusivity appears to be physically more correct, the Fickian diffusivity remains very important since this diffusivity can be directly assessed in diffusion measurements. 13.2.2 Self-Diffusivity

In the previous text, only transport diffusivities are considered: diffusion induced by a concentration gradient. However, in the absence of concentration gradients, molecules also move, this feature is called Brownian motion or self-diffusion. This self- or tracer diffusivity (Dself ) is commonly deﬁned for a 3D space by the following equation, in which the mean square displacement ( r 2 (t) ) is proportional with time (t):

r 2 (t) = 6Dself t

(13.7)

Depending on the measurement technique, a self- or transport diffusivity (or both) can be obtained, whereas the transport diffusivity is required for design. Therefore, relating the two is often desired, and a direct relation is recently proposed and is discussed in more detail in the next section.

13.2 Diffusion and Reaction in Zeolites: Basic Concepts

13.2.3 Mixture Diffusion

Diffusion in mixtures appears to be best treated by the MS approach to mass transport [8, 9, 11]. Because of the relatively high concentrations adsorbed in the zeolite, interactions between molecules can play a signiﬁcant role in terms of ‘‘speeding up’’ or ‘‘slowing down’’ other components. In the MS approach besides the interaction (or ‘‘friction’’) of the individual molecule with the zeolite, also the interaction between the different diffusing molecules is accounted for and balanced with the driving force for mass transport: ρθi ∇ ln f =

n qj Ni − qi Nj j=1

+

Ni ; qsat i Di

(13.8) - i = 1, 2 . . . n Within this approach, the estimation of -Di j can be difﬁcult; however, a reasonable

-

sat qsat j qi Dij

estimation can be made through a logarithmic (‘‘Vignes’’) interpolation [9, 12] based on the single-component exchange diffusivities and a correction factor F for the conﬁnement of the molecules in the narrow zeolite pores [13]. For a single-component system of tagged and untagged species, the saturation capacities are equal, and one can show [13] that the single-component exchange coefﬁcient is related to the self-diffusivity and MS diffusivity as 1 1 1 = + DSelf ,i Di Dii

-

-

-

-

θi θi +θj

Dij = F · Dii

-

θi θi +θj

Djj

(13.9)

(13.10)

For mesoporous systems, the factor F equals 1, whereas for the microporous zeolites, <1 holds. This factor is fairly constant for a zeolite and depends on the pore size (see [13]). It is evident that in the case of mixture diffusion, an accurate estimation of the individual component loading in the zeolite and the driving force is required to satisfactory model in such a system. For zeolitic systems, the ideal adsorbed solution theory (IAST) [14] provides an acceptable mixture prediction based on the single-component isotherms [8, 15], but when adsorption heterogeneity becomes manifest, IAST tends to fail [16–18]. At signiﬁcant loading, the molecular interaction can play an important role, strongly inﬂuencing the reactant and product concentration proﬁles. When the loading is relatively low, the cross-correlation effects can often be ignored [19]. 13.2.4 Diffusion Measurement Techniques

For molecular gaseous and Knudsen diffusion, a reasonable estimation of the diffusivity can be made: in the case of Knudsen diffusion based on Eq. (13.2)

365

366

13 Diffusion in Zeolites – Impact on Catalysis

and in the case of gaseous diffusion by relations like that of Fuller et al. [20]. Although some generalizations on the diffusivities in zeolites can be made based on their class (e.g., cage-like) and pore size [21], accurate predictions from simple methods are not feasible. Therefore, signiﬁcant effort has been devoted to the development of new experimental techniques [22] and computer simulations (e.g., grand canonical Monte Carlo (GCMC) and molecular dynamics (MDs) [5, 23] to quantify adsorption and diffusion properties in zeolites. In the last decades, with improving experimental techniques and methodologies, many phenomena characteristic for zeolites have been clariﬁed, but also many new peculiarities of diffusion in zeolites have been discovered. The latter underlines the difﬁculty to capture diffusion in zeolites in a ‘‘simple’’ and general engineering approach, which is desired for process design. Excellent reviews on diffusion measurement techniques are available (e.g., [22, 24, 25]) and it is not our objective to repeat this in this chapter. Here, the focus is on the new, state-of-the-art time- and space-resolved techniques [26–28] that have provided exciting new insights and direct evidence for several phenomena in zeolite mass transport, for instance, surface barriers, diffusion anisotropy, and catalyst activity at the subcrystal level. Although these state-of-the-art techniques have clear added value, this means that in no way other techniques have lost their value [29]. A remark needs to be made on the controversy of the difference in diffusivities obtained by macroscopic techniques (e.g., transient adsorption uptake, membrane experiments, or zero length chromatography) and microscopic techniques (e.g., quasi elastic neutron scattering (QENS) and pulsed ﬁeld gradient nuclear magnetic resonance (PFG-NMR)). The measured diffusivities for the same system may differ in orders of magnitude [22, 30]; diffusivities measured by macroscopic techniques are sometimes found to be signiﬁcantly lower. These differences can be explained by the length scale of the measurements and the fact that, consequently, in macrotechniques, also surface and internal barriers are part of the investigated domain, whereas in microtechniques, the measured domain can be much smaller and probes only the internal zeolite pore structure. The consequence of these ﬁndings is twofold: diffusivities measured by macroscopic techniques are generally not representative for the true intracrystalline diffusivity and diffusivities measured by microscopic techniques are a priori not representative for a real system that may comprise internal and surface barriers. Therefore, interpretation and application of diffusivity data should be done with caution. 13.2.5 Relating Diffusion and Catalysis

In the classical reaction engineering approach, to account for internal diffusion limitations, the concept of catalyst effectiveness factor is introduced, representing the ratio between the observed reaction rate versus the expected (intrinsic) reaction

13.2 Diffusion and Reaction in Zeolites: Basic Concepts

rate at bulk concentration and temperature: robserved η= rintrinsic

(13.11)

On solving the mass balance of a ﬁrst-order irreversible reaction in a slab-like and spherical geometry, respectively, two simple relations to describe the catalyst effectiveness are obtained: 3 1 1 tan h(φ) and η = − (13.12) η= φ φ tan h(φ) φ which becomes a function only of the Thiele modulus [31] that can be calculated from the characteristic length for diffusion (L, ratio between volume and external surface area) of the particle, the intrinsic reaction rate constant based on the volume of the particle (k), and the effective diffusivity (Deff ). The Thiele modulus can be considered as the square root of the ratio of the characteristic time for diffusion and of reaction: k kL2 τdiff V (13.13) = = with L = φ=L Deff Deff τreaction A For cylindrical and spherical geometry, similar relations are obtained as Eq. (13.12) and, thus, a powerful, simple approach is found to describe the catalyst effectiveness in an accurate way. A textbook example of a successful application of the effectiveness factor and the Thiele modulus to model diffusion limitations is given by Post et al. [32]. They were able to model the catalyst effectiveness of cobalt Fischer Tropsch catalysts with different pore and particles sizes accurately within this concept. In the case of a zeolite catalyst, diffusion is sometimes described based on the adsorbed phase concentration, whereas the reaction rate is based on the bulk gas phase concentration or pressure. In this case, an adsorption constant (K) is added to the Thiele modulus to maintain its dimensionless character [33]: k (13.14) φ=L KDeff This explains in part the large differences in diffusivity values for zeolites and meso/macroporous materials. Post et al. have also provided an experimental veriﬁcation of the Thiele concept for the conversion of 2,2-dimethyl butane over a H-ZSM5 catalyst. At reaction temperatures ranging from 673 to 803 K, the measured effectiveness factor closely follows the predicted effectiveness calculated from the Thiele modulus [25, 33]. Figure 13.2 shows how the concentration proﬁle across a zeolite crystal at different values of the Thiele modulus and the dependence of the effectiveness factor on the Thiele modulus for the speciﬁc case of a slab-like geometry (Eq. (13.12)). Full utilization of the catalyst particle (η → 1) only takes place at very low values of the Thiele modulus (φ → 0). Contrarily, φ = 10 renders η = 0.1, meaning that only 10% of the catalyst volume is effectively used in the reaction. Transport limitations not only affect activity, but may also impact selectivity and stability

367

13 Diffusion in Zeolites – Impact on Catalysis

1

h

368

0.1 0.1

1 f

Figure 13.2 Concentration proﬁles across a zeolite crystal (sphere geometry) at different values of the Thiele modulus φ (from left to right: φ = 0.1, 1, 3, and 10) and the dependence of the effectiveness factor on the Thiele modulus. Low Thiele moduli lead to full catalyst utilization (φ → 0, η → 1),

10

whereas high Thiele moduli render a poorly utilized catalyst (φ → ∞, η → 1/φ). The reactant concentration across a zeolite crystal is exhausted (c/cs = 0) close to the surface at φ = 10, while being practically uniform and very similar to the surface concentration (c/cs = 1) at φ = 0.1. (After [34].)

[7, 34]. Especially in case of consecutive reactions, diffusion limitations should be avoided if the intermediate product is desired. To what extent the Thiele concept is applicable to zeolites is discussed in more detail in Section 13.3.4. 13.3 Diffusion in Zeolites: Potential Issues

Once the scenario has been set, now some special issues will be treated that can typically occur in zeolites, what the consequences are with respect to catalysis and, when possible, how to account for these effects in modeling studies. 13.3.1 Concentration Dependence of Diffusion

In contrast to gas phase reactions in meso- and macroporous materials, in the case of zeolites, relative high concentrations are common practice and the diffusivity is found to be (strongly) dependent on the occupancy of the zeolite. K¨arger et al. [35] have identiﬁed ﬁve types of loading dependencies for self-diffusivities. Figure 13.3 shows some typical transport diffusivity behaviors as a function of the loading. This ﬁgure captures in principle the dependencies originally postulated by K¨arger et al.,

13.3 Diffusion in Zeolites: Potential Issues

1.5 IV I

Ð/Ð(0)

1.0

II

0.5 III 0.0 0.0

0.5

1.0

q/q sat Figure 13.3 Common dependencies of the Maxwell–Stefan diffusivity on the loading on the zeolite, normalized to that at zero loading.

and accounting for the observation that when approaching saturation loading, the diffusivity appears to become very low in general. It goes without saying that not every diffusivity – concentration dependency ﬁts one of the four proﬁles. There is a broad spectrum of variation. Type I can be described by Eq. (13.5), which represents a loading-independent diffusivity, except at very high occupancies. However, it is frequently observed that the diffusivity decreases with loading and approaches very small values when the zeolite becomes saturated. This can be interpreted as a reduction of free volume available for diffusion and can be incorporated in the diffusivity modeling with the addition of a term that corrects this. The diffusivity follows in this case the so-called ‘‘strong conﬁnement’’ scenario, which is represented by type II in Figure 13.3 and can be described by

-Di = -Di(0)(1 − θi)

(13.15)

It is noteworthy to mention that in the case the diffusivity obeys the strong conﬁnement scenario (Eq. (13.15)) and the isotherm can be described by a single-site Langmuir isotherm, the thermodynamic correction factor (Eq. (13.6)) and conﬁnement term cancel and Fickian-type diffusion is found (Eq. (13.3)). Most zeolite–host systems can be treated well with Eq. (13.5) either with or without the strong conﬁnement correction. The diffusivities of many light gases in cage-like zeolites ﬁrst show a strong increase in diffusivity with increasing loading before eventually going to very small diffusivity values (type III in Figure 13.3) when approaching saturation loading. Although much evidence of this phenomena comes from computer simulations (e.g., MD) [21, 36], clear experimental evidence of strongly increasing diffusivities with loading is also present [22, 37, 38]. Two possible explanations are given in literature. The ﬁrst explanation relates this phenomenon to intermolecular repulsion effects [39, 40]: when the loading increases, the repulsion effects increase and, consequently, the activation barrier for diffusion is reduced

369

370

13 Diffusion in Zeolites – Impact on Catalysis

leading to an increased diffusivity [41]. A second, alternative explanation is found in the heterogeneity of adsorption on the zeolite [42, 43]. In the case of diffusion of a gas phase in porous medium, the gas is present as a continuum and can be treated as such. But, in the case of a zeolite, the adsorbed phase can be very well segregated. A well-known example is the adsorption of alkanes in zeolite MFI. Adsorption of butane and hexane yields a clear two-step isotherm, where the two steps can be related to adsorption in the channels and intersections of MFI, respectively [44, 45]. In a similar manner, segregated adsorption is found between adsorption sites in the cage and in the connecting window for strongly conﬁned molecules in small-pore zeolites consisting of cages connected by windows [17]. Under the assumption that molecules at the window determine mass transport in the zeolite pores, the observed loading dependency could be explained [42]. The last type of diffusivity behavior, type IV from Figure 13.3, is, for example, found for linear hydrocarbons in zeolite MFI [46] and seems to be related to the presence of two distinct adsorption sites in this zeolite system, that is, when the ﬁrst of the two sites becomes fully occupied, the diffusivity is strongly reduced. 13.3.2 Single-File Diffusion

Another very interesting phenomenon is single-ﬁle diffusion. Recently, this topic has been thoroughly reviewed by K¨arger [47]. This type of diffusion is characterized by the fact that molecules are not able to pass each other and maintain their initial sequence in time, which can be easily envisaged in 1D zeolite pores. Such a system is strongly correlated and the diffusion process is strikingly different from ordinary diffusion.

The mean square displacement ( r 2 (t) ) is proportional to the mobility factor (F) and the square root in time at sufﬁciently long observation times and inﬁnite length [48]: √ 2

(13.16) r (t) = 2F t It is already intuitively clear that, in this type of diffusion, molecules can severely hinder each other and, therefore, the diffusivity will be low compared with ordinary self-diffusion. This is quantiﬁed in Eq. (13.16) where the mean square displacement increases with the square root in time and not with time, as found for ‘‘normal’’ self-diffusion (Eq. (13.7)). In the simplest case, with equally spaced adsorption sites, the mobility factor is given by [48] F = λ2

1−θ 1 √ θ 2πτ

(13.17)

which is a function of the mean residence time at an adsorption site (τ ) and the jump distance (λ). However, a real zeolite catalyst has a ﬁnite size and open ends and it has been found that the mean square displacement of such a system eventually becomes proportional with time, related by an effective diffusion

13.3 Diffusion in Zeolites: Potential Issues

coefﬁcient dependent on the characteristic length of the channel (Eq. (13.18)) [49]. This type of diffusion is called center of mass diffusion and can be seen as a type of ordinary diffusion, although much slower. 2

1−θ λ r (t) = 2D t = 2Deff t θ L

(13.18)

Which model applies best is dependent on the length of the channel. Experimental evidence of single-ﬁle diffusion was measured by microscopic methods. One of the ﬁrst direct experimental evidence was provided by Gupta et al. [50] who measured the diffusivity of ethane on large AlPO4 -5 (AFI topology, 1D pore system) crystals by PFG-NMR. The crystals were large enough to measure the dependency of the mean square displacement with the square root of time. Also with QENS, direct evidence has been claimed. Jobic et al. [51] could describe their QENS spectra of cyclopropane in AlPO4 -5 and methane in ZSM-48 (a disordered 1D pore system) at higher loadings. Single-ﬁle diffusion affects the concentration proﬁles inside the catalyst and, consequently, the catalyst effectiveness. K¨arger et al. [48] introduced a characteristic diffusion time to extend its use to single-ﬁle systems: φ = 3kτintra (13.19) where the intracrystalline diffusion time (τintra ) for single-ﬁle diffusion is given by [48]: τintra =

L2 12D

(13.20)

With insertion of an appropriate description of the diffusivity (Eq. (13.18) in the Thiele modulus, a qualitative description of a model system of a ﬁrst-order irreversible reaction in a single-ﬁle diffusion regime was obtained. It must be noted that compared with ordinary diffusion at equal diffusivities, a much higher Thiele modulus is obtained (compare Eqs. (13.13) and (13.20)), leading to more pronounced diffusion limitations in the case of single-ﬁle diffusion. Although the existence of single-ﬁle diffusion has been demonstrated for several ideal cases, structural crystal defects and intracrystalline barriers can diminish the impact of single-ﬁle diffusion. For ethane diffusion in AlPO4 -5, Gupta et al. [50] found single-ﬁle diffusion using PFG-NMR, but Jobic et al. [51] found normal diffusion by QENS measurements. As reason for this difference, a difference in crystal batches was suggested [51]. Moreover, crystallites used in catalysis are often small and their internal transport may very well be described by the center of mass diffusion, a form of normal diffusion (Eq. (13.18)). Furthermore, the adsorbate concentrations in the zeolite at the relevant catalytic conditions are typically low (high temperature), a regime where single-ﬁle diffusion will not be very pronounced, when present. Evidence for single-ﬁle diffusion under catalytic conditions is only indirectly obtained. Zeolite mordenite is a well-known catalyst with a 1D pore structure, for which single-ﬁle effects have been claimed in catalysis (e.g., [52, 53]).

371

372

13 Diffusion in Zeolites – Impact on Catalysis

As a ﬁnal remark, besides negative impacts on catalysis, there have also been attempts to beneﬁt from single-ﬁle diffusion. There appears to be a theoretical basis to exploit single-ﬁle diffusion to enhance the reactivity under certain conditions by this type of molecular trafﬁc control [54]. Selectivity can be enhanced enormously by zeolite coatings on porous materials [55] as well as on zeolite catalysts [56]. 13.3.3 Surface Barriers

Surface barriers in zeolite systems have been discussed and used to explain diffusivity results from the late [57]. Recent advances in measuring techniques (Section 13.4), particularly interference microscopy, have provided overwhelming direct evidence for the presence of surface barriers for several zeolite-adsorbate systems (e.g., [26, 58]). The commonly applied assumption that the adsorbed concentration at the boundary of the crystal is in equilibrium with the gas phase concentration is violated in this case. Figure 13.4 shows clearly that in the case of i-butane in silicalite-1 (MFI), no surface barriers are present (Figure 13.4b) and in the case of methanol in ferrierite, (FER) this is clearly the case (Figure 13.4a): the concentration at the boundary of the crystal is not in equilibrium with the gas phase during the uptake. The exact nature of the surface barrier is not completely clear. Part of the explanation stems from the idea that molecules suffer from restrictions when entering the pores of the zeolite. From this point of view, molecules do not enter the pores of the zeolite directly but a weak surface adsorption step is involved in the adsorption process [60] as already proposed by Barrer [61]. The inﬂuence of this type of transport barrier has been reduced by Reitmeier et al. [62] by coating an amorphous silica layer (∼1–1.5 nm) on top of ZSM-5 crystals. With this coating, the local adsorption at the pore entrance and hence the uptake rate into the zeolite was increased.

(∆

≈∆

t = 110 s

∆y

c/

t = 1100 s

)/∆

t = 30 s

y

t = 750 s ∂c / ∂t ≈∆c /∆t

Before adsorption

t = 400 s

0

10

20

y (µm)

30

40

Equilibrium

0.8 t = 50 s

0.6 0.4 t = 10 s

t = 30 s

0.2 t=0s

t=0s

0.00

(a)

1.0 Crel. during uptake

2

∂y

c

t = 230 s

t = 1650 s

c/

0.05

t = 2250 s 2

0.10

Equilibrium after 77 min

t = 3750 s ∂

0.15

Gas phase outside of the crystal

Crystal 0.20

0.0 0

50 (b)

Figure 13.4 Transient intracrystalline concentration proﬁles of methanol in FER (a) and i-butane in silicalite-1 (b) measured by interference microscopy. Figures taken from [26] and [59], respectively (Copyright Wiley-VCH Verlag Gmbh & co. KGaA. Reproduced with permission).

5

10

15

x (µm)

20

25

30

35

13.3 Diffusion in Zeolites: Potential Issues

This is only one side of the story because also during desorption, surface barriers have been observed [63], indicating that it is not just a matter of entering the crystal. Tzoulaki et al. [64] have shown that traces of water form a surface barrier for isobutane entering silicalite-1, indicating that also other adsorbates can introduce additional surface barriers. Although many of the zeolite–guest systems investigated by interference microscopy have been shown to possess surface barriers, it is not clear when barrier effects are important or not. Crystals that exhibit surface barriers may be treated to remove these [59]. Two concepts have been introduced to describe the barrier effects: surface sticking probability and a surface permeability [65]. The surface permeability describes the transport of a molecule from the surface into the zeolite pore, and the sticking coefﬁcient represents the chance that a molecule that hits the surface also sticks to the surface. The surface permeability (α) relates the ﬂux through the surface to the concentration difference over the barrier, and can be considered as a mass transfer coefﬁcient: Nisurf = αi (Ci,eq − Ci,surf )

(13.21)

Using this approach, an estimation of the surface permeability can be made using time-resolved concentration proﬁles generated by Interference Microscopy (IM). It is found that, analogously to the diffusivity, the surface permeability is dependent on the concentration [66], and accurate predictions cannot be made up to now. The sticking coefﬁcient can vary by orders of magnitude. It can be very close to 1, as observed for n-butane in silicalite-[67], or be several orders lower, as found by Jentys et al. [68] for benzene, toluene, and xylene on ZSM-5. Clear guidelines for which systems low sticking coefﬁcients can be expected are still lacking. But the sticking probability is clearly a phenomenon that can impose limitations on the mass transport into the zeolite. Whether surface barriers become apparent is clearly dependent on the relation with the intracrystalline diffusivity. An analogy with the Biot mass number (Bi) for porous catalyst particles can be drawn. It represents the ratio of the resistance against effective diffusion inside a porous catalyst pellet and that of transport through the ﬁctitious ﬂuid ﬁlm surrounding the particle. Following a similar approach for a zeolite particle, the ratio of the characteristic times of intracrystalline diffusivity and surface permeability can be introduced [69] as τintra Lα = Bizeolite = (13.22) τsurf D So, in relation to reaction, an extra resistance is found besides the internal diffusion resistance: the surface barrier. It is evident that an optimal catalytic process meets the following criterion (provided the parameters are all based on identical concentrations): τintra , τsurf τrxn 1 L2 L , D α k

(13.23)

373

374

13 Diffusion in Zeolites – Impact on Catalysis

Depending on the nature of the barrier, reduction seems possible in certain cases. As previously noted, Reitmeijer et al. [62] succeeded by coating a thin top layer of amorphous silica on top of the crystals. Tzoulaki et al. [59] obtained surface barrier free silicalite-1 crystallites by an alkaline treatment, removing a layer of amorphous silica that was the suspected surface barrier. These two examples form a nice contrast, illustrating the need for further elucidation. The inﬂuence of the surface barrier on the catalyst effectiveness will be similar as an external mass transfer limitation: a reduced surface concentration that reduces the catalyst effectiveness. This effect of external barriers can be easily extended to intracrystalline barriers. Catalysis on the subcrystal level will be discussed in more detail in Section 13.3.4 and Chapter 8, where the existence and inﬂuence of intracrystalline barriers are nicely demonstrated. 13.3.4 The Thiele Concept: A Useful Approach in Zeolite Catalysis?

The Thiele concept, as described in Section 10.2.6, is very useful and popular because of its conceptual simplicity. In view of the peculiarities of zeolite systems presented in the previous sections, in which they clearly distinguish themselves from meso- and macroporous catalysts, one may wonder if this concept is also useful for zeolite catalysis. The important peculiarities are the loading dependency of diffusivity, single-ﬁle diffusion, surface barriers, and mixture diffusion. The concentration dependence of the diffusivity (Section 13.3.1) can lead to poor effectiveness predictions when a constant diffusivity is applied. Ruthven [70] accounted for the loading dependency in the Thiele modulus and derived some analytical solutions for some types of loading dependency. However, analytical expressions for each type of loading dependency do not seem to be very practical or feasible. One may wonder, however, if the effect of loading dependency prohibits the use of the Thiele concept. When the diffusivity is measured under practical conditions, an average diffusivity is obtained that may represent the system reasonably well. More importantly, when the loading in the zeolite is low, which is very common for catalytic conversion at elevated temperatures, these loading effects are less signiﬁcant and a constant diffusivity sufﬁces. Regarding single-ﬁle diffusion, it has been shown that with a modiﬁed Thiele modulus, a qualitative description of the catalyst effectiveness could be given (Section 13.3.2). Surface barriers can be treated as an external mass transport resistance. The surface permeability and sticking coefﬁcient can, however, only be determined by a limited number of techniques. Macroscopic techniques do not seem to be able to discriminate the surface barrier and internal diffusion [26]. A diffusivity determined by macroscopic techniques will then be an apparent quantity, but can be used for engineering purposes to estimate the zeolite’s performance. In Section 13.2.3, it was pointed out that diffusion in mixtures can be signiﬁcantly different from that in pure components. In the Thiele approach, only the reactant

13.4 Pore Structure, Diffusion, and Activity at the Subcrystal Level

single-component diffusivity is used, neglecting interactions with other adsorbates and counterdiffusion effects in small pores. Competitive adsorption and diffusion effects can negatively impact the catalyst effectiveness [15]. A detailed investigation of these effects was carried out recently by Hansen et al. [71] who analyzed the diffusion limitations in the alkylation of benzene over H-ZSM-5, combining different simulation tools. The usual approach to determine effectiveness factors for reactions in porous media, assuming a constant effective diffusivity, may lead to substantial deviations. A possible solution may be to determine the effective diffusivity at conditions relevant for practice. As a general conclusion, it appears that in many cases, the Thiele concept can be applied well for zeolite catalysts and remains a very valuable tool in the interpretation of reaction data as follows from several studies [72–75] that have successfully applied this approach. This approach should be applied with caution, however, for systems with signiﬁcant concentration levels in the zeolite where deviations can be expected.

13.4 Pore Structure, Diffusion, and Activity at the Subcrystal Level

One of the main challenges still remaining in the ﬁeld of molecular transport in zeolites is to match diffusivities measured by different techniques with catalytic performance data. The problem is twofold; ﬁrst, order-of-magnitude differences between the values of diffusivities determined by different techniques (macrovs. microscopic) are frequently reported for the same guest–host systems. As a consequence, it is extremely difﬁcult to directly incorporate diffusion data into catalysis studies: no more trends can be extracted from pure diffusion experiments, while the quantitative values relevant for catalysis are still not clear [76]. The second problem arises from the differences between different batches of the same zeolite. In the early 1970s, the sentence ‘‘your ZSM-5 is not mine’’ was coined. On the basis of these recent studies, this quote can be extended to the rest of the zeolite topologies: the time-ago believed single-crystal zeolites have been shown to be composed of several intergrowth building blocks. The interfaces of these subunits constitute diffusion boundaries due to a potential mismatch in the alignment of the microporous network or a different pore orientation. This mismatch will be different for each crystal and batch of zeolite. This fact has very important consequences for catalysis, since certain regions of the zeolite crystals may be inaccessible for reactant molecules [28]. During the past years, excellent papers have been published in the literature aiming at unraveling the orientation of the complex zeolite channel networks (of mainly, MFI, AFI, and chabasite (CHA)) by identifying the different building blocks [77–79]. Several techniques like AFM, optical and ﬂuorescence microscopy, electron backscatter diffraction/focused ion beam, FTIR, and Raman have been applied (see Chapter 8 for more details). In addition, the catalytic behavior of such crystals have been studied by spatial and time-resolved techniques that complement

375

376

13 Diffusion in Zeolites – Impact on Catalysis

the structural studies and demonstrate the nonuniform catalytic behavior of such zeolitic crystals [28, 80–85]. These studies indicate a trend that the larger the zeolite crystal, the less perfect its pore structure. The ﬁrst indications for building units involved in crystal growth of ZSM-5 were already published in the early 1990s [86]. By combining transmission electron microscopy (TEM) and scanning electron microscopy (SEM). Hay et al. reported that most frequently adjoining ZSM-5 crystals are rotated by 90◦ around a common c axis with an intergrowth nucleating from small areas on (010) faces of growing crystals. On (100) faces of large crystals, ramps were also observed in association with impurities. Consistent with these observations, one of the ﬁrst models was proposed for crystal growth. Along with the development of optical techniques, their application in the study of zeolite crystals has growth during the last decade. On pioneering work, Koricik et al. applied light microscopy to investigate sorption and mass transport phenomena in zeolites together with peculiarities of crystal morphology via coloring of zeolite crystals (MFI) using iodine indicators [87, 88]. Some years later, the group of K¨arger started the use of interference microscopy [89] in combination with FTIR [58] to elucidate macroscopic adsorbate distributions and crystal intergrowth. In this approach, the high spatial resolution of interference microscopy is complemented by the ability of FTIR spectroscopy to pinpoint adsorbates by their characteristic IR bands. For the ﬁrst time, two-dimensional concentration proﬁles of an unprecedented quality were reported, showing an inhomogeneous distribution of adsorbate [58]. These inhomogeneous proﬁles were attributed to regular intergrowth effects in CrAPO-5 (AFI structure). The space and time-resolved study of the detemplation process has been shown as a powerful technique for determining the intergrowth structure on the basis of in situ mapping of the template-removal process in individual zeolite crystals. When the formation of light-absorbing and -emitting species during the heating process is monitored by a combination of optical and confocal ﬂuorescence microscopy, as the accessibility of the porous network in the subunits varies, the individual building blocks can be readily visualized by monitoring the template-removal process in time. This concept has been successfully applied to four different zeolite crystals: CrAPO-5 (AFI structure), SAPO-34 (CHA structure), SAPO-5 (AFI structure), and ZSM-5 (MFI structure) [78]; the proposed intergrowth structures are depicted in Figures 13.5 and 13.6. Because of its high industrial relevance, zeolite ZSM-5 has been widely studied in order to elucidate if its cofﬁn-like crystals are the product of two or three interpenetrating crystals (two- and three-component models, Figure 13.6). The two-component model can be regarded as two interpenetrating crystals rotated by 90◦ around the common c axis [87, 90–92]. This model consists of two central and four side pyramidal subunits. In the so-called three-component model, the crystallographic axes maintain the same orientation across the entire crystal [93]. From the catalysis engineering point of view, there are important consequences arising from the accessibility of the pores, since according to the two-component model and due to the changed orientation within the pyramidal components, the

13.4 Pore Structure, Diffusion, and Activity at the Subcrystal Level

(a)

(b)

(c)

(d)

Figure 13.5 Normal and ‘‘exploded’’ representation of the intergrowth structures of different zeolite crystals as proposed by [78]: (a) CrAPO-5 (front subunits are not shown), (b) SAPO-34, (c) SAPO-5 (front subunits are not shown), and (d) ZSM-5.

c a

c b

I

a

c

VI b

V

b a

a

b

c

III

IV

c a (a)

c b

II

a

b

(b)

Figure 13.6 ‘‘Exploded’’ representation of (a) two-component and (b) three-component models of cofﬁn-shaped ZSM-5 crystals. Orientations of crystal axes in the individual subunits are given. Subunits in the

two-component models are denoted I–VI (see text). The straight pores align with the b crystal axis, whereas the zigzag pores extend along the a crystal axis. (Adapted from [77].)

377

378

13 Diffusion in Zeolites – Impact on Catalysis

sinusoidal channels appear not only at the (100) faces of the crystallite but also at its hexagonal (010) facets, implying that there is hardly any access to the straight channels from the outer crystal surface. Experimental results reported by different groups [77, 79] clearly indicate that differences are due to the zeolite batches, that are intrinsically different, conﬁrming the quote ‘‘your zeolite is not mine.’’ In view of the unlike pore orientation of the studied samples, probably together with differences in Al distribution along the crystals, differences in the catalytic as well as in the diffusion behavior are expected. In contrast to many surface science spectroscopic techniques, ﬂuorescence microscopy is capable of studying diffusion and catalysis in zeolite pores of crystals with high 3D spatial and temporal resolution [94]. If ﬂuorogenic probes that are smoothly transformed into ﬂuorescent molecules on chemical transformation are chosen, reaction and diffusion can be followed in a spatial and time-resolved manner [27, 28, 81, 82, 84, 85, 94, 95]. The ﬁrst examples of application of such techniques dealt with studying the acid-catalyzed oligomerization of furfuryl alcohol on HZSM-5 [84] and H-MOR [94]. Furfuryl alcohol oligomerization starts with alkylation of one furfuryl alcohol molecule by another in an electrophilic aromatic substitution (EAS). After some subsequent acid-catalyzed reaction steps, a family of ﬂuorescent compounds is formed. Especially interesting are the results obtained for H-MOR, ﬂuorescence time lapse measurements (Figure 13.7) clearly show the evolution of catalytic activity from two opposing crystal faces, whereas no light is emitted from the rest of the outer zeolite surface. Transmission images evidenced that the reactive crystal faces are the (001) planes. As the reaction carries on, ﬂuorescence propagates along the (001) direction, corresponding to the 12-ring channels in mordenite. As furfuryl alcohol is too large to access the eight-ring pores, no ﬂuorescence develops from the other faces of the crystal; this result represents an excellent picture of diffusion

Intensity (cps)

2500

20 µm (a)

Figure 13.7 Reactive zones inside a mordenite crystal during dehydration of 1,3-diphenyl-1,3-propanediol. For conditions, see Section 13.6. (a) False color ﬂuorescence image, (b) transmission image, and (c) line proﬁle of the ﬂuorescence intensity along the line indicated in (a). (Adapted from [94].)

1500 1000 500 0

20 µm (b)

2000

(c)

0 5 10 15 Line position (µm)

13.5 Improving Transport through Zeolite Crystals

limitations associated to the 1D pore system of MOR. When the same reaction was studied in a 3D pore system like ZSM-5, after initiation of the reaction, ﬂuorescence gradually spreads into the crystal starting from the (100) and (101) planes [84]. Other liquid-phase reactions like the acid-catalyzed oligomerization of styrene derivatives inside the pores of ZSM-5 crystals were studied by using optical and ﬂuorescence microspectroscopy and application of polarized light [28, 81, 83], yielding similar insight in diffusion barriers and pore orientations in the crystal. All these results conﬁrm the spatially nonuniform catalytic behavior of the studied zeolite samples, with speciﬁc parts of the crystals being hardly accessible to reactant molecules and pore orientations that deviate from that expected on the basis of the crystal orientation.

13.5 Improving Transport through Zeolite Crystals

Limitations due to restricted access, slow transport, and diffusion boundaries provoke a low catalyst utilization. In many cases, zeolites are victims and executioners. When size selectivity is the key of a process, large zeolite crystals are preferred in order to decrease external surface reaction contributions; the methylation of toluene on ZSM-5 [56] is a clear example, where operation under strong diffusion limitation conditions enhances the overall selectivity of the process. This fact brings the accessibility problem to the extreme and limits industrial plants to operate far below their full potential, although it can also be utilized to enhance the selectivity of the process, as shown by Van Vu et al. [56], who coated H-ZSM-5 crystals with various Si-to-Al with polycrystalline silicalite-1 layers. When applied to the alkylation of toluene with methanol, the silicalite coating signiﬁcantly enhanced para-selectivity up to 99.9% under all reaction conditions. The enhanced para-selectivity may originate from diffusion resistance through the inactive silicalite layer on the H-ZSM-5, resulting in increased diffusion length. It has become clear from Section 13.3.4 that the classical approach to determine effectiveness factors for reactions in porous media can be inaccurate when applied to zeolites. However, at least from a qualitative point of view, their use may help gaining insight into the performance and the limitations of zeolite-catalyzed reactions, and utilized for the design of zeotypes hampered less by diffusion limitations. According to Eq. (13.13), if a small Thiele modulus is needed, two different strategies can be followed: shortening the diffusion length L and/or enhancing the effective diffusivity Deff in the zeolite pores. The latter strategy has led to the development of ordered mesoporous materials (OMMs) [96], where diffusion is governed by Knudsen or bulk regimes. This approach is valid for processes where bulky molecules are involved that would exceed the size of the pores and cages of the zeolites or when size selectivity is not a priority. However, OMMs still suffer from a poor thermal stability due to their, in general, thin walls. Furthermore, the performance of their active sites is usually far below that of zeolites due to the amorphous character of their walls [96–98].

379

380

13 Diffusion in Zeolites – Impact on Catalysis

Ultra large zeolites

Delaminated zeolites

Unimodal pore systems

Nanosized zeolites Zeolite composites Mesoporous zeolites

Hierarchical pore systems

Figure 13.8 Different zeotypes with enhanced transport characteristics. Ultralarge-pore zeolites (usually 12MR) present an increased effective diffusivity, whereas delaminated, nanosized, composites, and mesoporous zeolites present shorter diffusion path lengths.

Parallel to the development of OMMs, a great effort has been devoted to enhance diffusion in zeolites while maintaining the other intrinsic properties of the material (Figure 13.8). The synthesis of new structures with large and ultralarge pores [99–103], the modiﬁcation of the textural properties of known frameworks by creating mesopores via synthetic [104, 105] or postsynthetic approaches (mainly acid (dealumination) or basic (desilication) leaching) [106–111], the synthesis of small zeolite crystals with a more convenient external to internal surface ratio [112], the synthesis of micromesoporous composites [113] by using mixed templated systems [114, 115] or by recrystallization [116–119] approaches, and the delamination [120–126] of crystalline-layered structures are the most often followed approaches [127]. Ultralarge-pore zeolites present substantially wider micropores than regular zeolite structures, enhancing the effective diffusivity, whereas in the other approaches, the diffusion path length for reactants and products is shorter: nanosized zeolites have, in addition to nanosized zeolitic pores, intercrystalline pores or voids; zeolite composites are composed of zeolite crystals supported on a material that is typically mesoporous or macroporous. Mesoporous zeolite crystals exhibit intracrystalline mesopores, and delaminated materials are formed by single layers organized in a ‘‘house of cards’’-like structure, where reaction can be considered to take place at the pore mouth, in fact one could no longer distinguish a pore. Thus, ultralarge-pore and delaminated zeolites possess unimodal systems, whereas the other materials are characterized by featuring hierarchical pore systems, since they combine the intracrystalline micropores with larger pores that can be either intercrystalline or intracrystalline. Because of this optimization, the diffusional resistance in these cases is mainly determined by the transport in the larger pores over longer distances. Ultralarge-pore zeolites like SSZ-53 [102] and delaminated zeolites like ITQ-2 [123] have been successfully applied to the hydrocracking of bulky molecules under mild conditions, showing an outstanding performance due an improved transport of molecules and a higher acidity. The partial conversion of a mesoporous material TUD-1 into BEA or Y-type zeolite and application in alkylation or hydrocracking indicated an effective diffusion improvement by up to 15 times [128].

13.5 Improving Transport through Zeolite Crystals

Some destructive methods like acid leaching or steaming (dealumination) turned out to be less efﬁcient to improve transport than expected. In the case of dealuminated ultrastable Y (USY) pellets (FCC catalyst), the rate of molecular exchange between catalyst particles and their surroundings is primary determined by the intraparticle diffusivity, that is, at the reaction temperature, diffusion is controlled by the macropores and not by the micro- or mesopores [129]. Moreover, when dealuminated crystals are used instead of catalyst pellets, the mesopores do not form an interconnected network, and the diffusion of guest molecules through the crystals via only mesopores is not possible [130]. Another method, desilication [131] seems to be much more effective, yielding to a greatly improved physical transport in the zeolite crystals as revealed by transient uptake experiments of neopentane in ZSM-5 [108] crystals and diffusion studies of n-heptane, 1,3-dimethylcyclohexane, n-undecane in mesoporous ZSM-12 [132] and diffusion and adsorption studies of cumene in mesopore structured ZSM-5 [133]. Up to 3 orders of magnitude-enhanced rates of diffusion were concluded in the hierarchical systems as compared with their purely microporous precursors due to improved accessibility and a distinct shortening of the micropores. Moreover, Brønsted acidity seems to be preserved, in contrast to dealumination. Catalytic testing of various mesoporous zeolites has shown the effectiveness of the desilication approach in the liquid-phase degradation of HDPE, cumene cracking, and methanol to gasoline on desilicated ZSM-5 [34]. A recent in situ microspectroscopic study on the oligomerization of styrene derivatives revealed an enhanced accessibility of the micropores in the hierarchical ZSM-5 zeolites obtained by desilication [81], but still a nonuniform catalytic behavior was discovered, due to a nonuniform distribution of the aluminum over the zeolite crystal, which seems to be an intrinsic phenomenon for this material. Templated mesoporous zeolites and zeolite nanoparticles deposited on different supports have been widely applied in catalysis. The activation energy of the vapor-phase benzene alkylation with ethylene to ethylbenzene was found to be higher for a carbon-templated ZSM-5 than that of the purely microporous zeolite (77 vs 59 kJ mol−1 ); this fact was attributed to the alleviated diffusion limitation in the case of the mesoporous crystals. Hierarchical mesoporous BEA zeolite templated with a mixture of organic ammonium salts and cationic polymers showed a higher activity in the alkylation of benzene with propan-2-ol than a microporous BEA sample with the same Si/Al ratio [134]. Catalytic test reactions on the oxidation of 1-naphthol over TS-1, Ti-coated MCF, and MCF materials coated with (TS-1) nanoparticles revealed increased 1-naphthol conversion and activity for the TS-1-coated MCF materials compared with the TS-1 zeolite due to the presence of mesopores. Moreover, an increased selectivity, hydrothermal stability, and the absence of titania leaching were observed for the TS-1-coated MCF materials in contrast to the Ti-coated MCF materials because titania was embedded in the zeolitic framework present in the TS-1 nanoparticles [135]. Catalytic tests on MAS-7 and MTS-9 (mesoporous materials made up of zeolite beta and TS-1 precursor particles, respectively) in the cracking and hydroxylation (with H2 O2 ) of different small and bulky molecules (cumene, phenol, trimethylphosphine (TMP), etc.)

381

382

13 Diffusion in Zeolites – Impact on Catalysis

showed high activity. The acylation of different amino derivatives with fatty acids is carried out smoothly and under green conditions when using UL-MFI type (mesoporous ZSM-5) as catalyst [136].

13.6 Concluding Remarks and Future Outlook

Zeolites are commonly applied versatile catalysts, but due to their small pores, they are also frequently associated with diffusion limitations. Because of their small pore size, mass transport is characterized by adsorbed phase diffusion (zeolitic diffusion). Since each zeolite has its speciﬁc pore connectivity and structure, much effort has been put in measuring and understanding the mass transport phenomena in these systems. Recently, new space and time-resolved measurement techniques, like interference, IR, and ﬂuorescence microscopy, have been introduced and provided a wealth of new insights in zeolite diffusion and catalysis on the (sub)crystal level. In contrast to gaseous molecular and Knudsen diffusion, zeolitic diffusion involves very narrow conﬁnements and high concentrations, leading to a loading-dependent diffusivity. Additionally, at high concentrations, signiﬁcant competitive adsorption and hindrance (‘‘friction,’’ ‘‘exchange’’) effects between guest molecules may occur. When the pore structure of a zeolite is 1D and molecules are unable to pass each other, single-ﬁle diffusion is the governing transport mechanism, which is much slower than ordinary diffusion. These phenomena, however, are not very prominent at low loadings, which are often in the case for catalysis at elevated temperatures, but dominant in liquid-phase reactions. Particularly, the microscopic techniques have revealed the existence of surface and internal barriers for several zeolite crystals. This was suggested decades ago but not until recently, it has been demonstrated so clearly for a variety of zeolites. Indications when zeolite–guest system barriers can be expected are not generalized, but their occurrence clearly increases with crystal size, having a twofold detrimental effect on the catalyst effectiveness, due to the combined lower effective diffusivity and the larger particle size. These new experimental techniques are able to unravel the orientation of the complex zeolite channel networks of ‘‘single crystals,’’ deviating from that expected on the basis of their geometry. Often they are composed of subunits with their own pore orientation, explaining important differences between different batches of the same zeolite topology. Since these techniques allow monitoring in time and space, local concentrations inside crystals can be used well for quantiﬁcation of surface barriers and diffusivities inside zeolite particles. So, 250 years after the word zeolite was coined and more than 50 years after the ﬁrst industrial application of such materials was developed, still new insights are obtained at the subcrystal level, impacting the relation ‘‘pore structure–diffusion–activity.’’

References

A well-established method to account for intraparticle diffusion in porous catalysts and quantify its impact on catalyst performance is the Thiele approach. The application has been very successful in catalysis and seems to be extendable to zeolite catalysts quite well at low concentrations. However, at higher concentrations, the loading dependency of diffusion, competitive adsorption effects, and strong hindrance (‘‘friction’’) can introduce severe deviations in the description of the catalyst effectiveness. Shortening the diffusion distance in the zeolite crystal is the best solution to utilize the intrinsic properties of the zeolite to the fullest, and many synthetic and posttreatment approaches are being explored to realize this. The incorporation of the nanosized zeolite structures in meso- and macroscopic bodies may alleviate the diffusional resistance in the zeolite, but those in the catalyst particle may remain, for which diffusion measurements remain essential. The evolution toward zeolitic materials with improved transport is a well-established ﬁeld of research, and the proofs of principle have been given. Zeolites are still materials with further design possibilities. In relation to diffusion, a careful analysis of the characteristic times of the various phenomena in a catalyst particle (reaction, diffusion in zeolite-, micro-, meso-, and macropores, barrier transport) may guide the way to compose the optimal hierarchical structure of a catalyst particle for use in practice. This implies careful experimentation and interpretation, including diffusion, on all these aspects, in combination with molecular modeling.

References 1. Corma, A. (2003) J. Catal., 216, 2.

3.

4. 5. 6.

7.

8.

298–312. Plank, C.J., Hawthorne, W.P., and Rosinski, E.J. (1964) Ind. Eng. Chem. Prod. Res. Dev., 3, 165–169. Taramasso, M., Perego, G., and Notari, B. (1983) Preparation of porous crystalline synthetic material comprised of silicon and titanium oxides. US Patent 4,410,501. Song, C. (2002) Cattech, 6, 64–77. Smit, B. and Maesen, T.L.M. (2008) Chem. Rev., 108, 4125–4184. Ruthven, D.M. (1984) Principles of Adsorption and Adsorption Processes, John Wiley & Sons, Inc., New York. Froment, G.F. and Bischoff, K.B. (1990) Chemical Reactor Analysis and Design, John Wiley & Sons, Inc., New York. Kapteijn, F., Zhu, W., Moulijn, J.A., and Gardner, T.Q. (2005) Zeolite membranes: modeling and application,

9. 10. 11.

12.

13.

14. 15. 16.

in Structured Catalysts and Reactors, Chapter 20 (eds A. Cybulski and J.A. Moulijn), Taylor & Francis Group, Boca Raton, pp. 701–747. Krishna, R. and Wesselingh, J.A. (1997) Chem. Eng. Sci., 52, 861–911. Krishna, R. (1993) Gas Sep. Purif., 7, 91–104. Kapteijn, F., Moulijn, J.A., and Krishna, R. (2000) Chem. Eng. Sci., 55, 2923–2930. van de Graaf, J.M., Kapteijn, F., and Moulijn, J.A. (1999) AIChE J., 45, 497–511. Krishna, R. and van Baten, J.M. (2009) Chem. Eng. Sci., 64, 870–882. doi: 10.1016/j.ces.2009.03.047. Myers, A.L. and Prausnitz, J.M. (1965) AIChE J., 11, 121–127. Baur, R. and Krishna, R. (2005) Catal. Today, 105, 173–179. Murthi, M. and Snurr, R.Q. (2004) Langmuir, 20, 2489–2497.

383

384

13 Diffusion in Zeolites – Impact on Catalysis 17. Krishna, R. and van Baten, J.M. (2008) 18. 19. 20.

21.

22.

23. 24. 25.

26.

27.

28.

29.

30.

31. 32.

33.

34.

Sep. Purif. Technol., 60, 315–320. Krishna, R. and van Baten, J.M. (2007) Chem. Phys. Lett., 446, 344–349. Habgood, H.W. (1958) Can. J. Chem. Rev. Can. Chim., 36, 1384–1397. Fueller, W.N., Schettler, P.D., and Giddings, J.C. (1966) Ind. Eng. Chem. Res., 58, 19–53. Krishna, R. and van Baten, J.M. (2008) Microporous Mesoporous Mater., 109, 91–108. K¨arger, J. and Ruthven, D.M. (1992) Diffusion in Zeolites, John Wiley & Sons, Inc. Coppens, M.O., Keil, F.J., and Krishna, R. (2000) Rev. Chem. Eng., 16, 71–197. Karger, J. (2003) Adsorpt. J. Int. Adsorpt. Soc., 9, 29–35. Ruthven, D.M. (2008) in Introduction to Zeolite Science and Practice (eds J. ˇ Cejka, H. van Bekkum, A. Corma, and F. Schueth), Elsevier, pp. 737–786. Karger, J., Kortunov, P., Vasenkov, S., Heinke, L., Shah, D.R., Rakoczy, R.A., Traa, Y., and Weitkamp, J. (2006) Angew. Chem. Int. Ed., 45, 7846–7849. Roeffaers, M.B.J., Sels, B.F., Uji-I, H., De Schryver, F.C., Jacobs, P.A., De Vos, D.E., and Hofkens, J. (2006) Nature, 439, 572–575. Kox, M.H.F., Stavitski, E., and Weckhuysen, B.M. (2007) Angew. Chem. Int. Ed., 46, 3652–3655. Post, M.F.M. et al. (1991) in Introduction to Zeolite Science and Practice (eds H. van Bekkum, E.M. Flanigen, and J.C. Jansen), Elsevier, Amsterdam, pp. 391–443. Jobic, H., Schmidt, W., Krause, C.B., and Karger, J. (2006) Microporous Mesoporous Mater., 90, 299–306. Thiele, E.W. (1939) Ind. Eng. Chem., 31, 916–920. Post, M.F.M., Vanthoog, A.C., Minderhoud, J.K., and Sie, S.T. (1989) AIChE J., 35, 1107–1114. Post, M.F.M., van Amstel, J., and Kouwenhoven, H.W. (1984) in Proceedings 6th International Zeolite Conference, Reno, 1983 (eds D. Olson and A. Bisio), Butterworth, Guildford. Perez-Ramirez, J., Christensen, C.H., Egeblad, K., Christensen, C.H., and

35. 36. 37. 38.

39.

40. 41. 42.

43. 44.

45.

46. 47. 48.

49. 50.

51.

52.

53.

54.

Groen, J.C. (2008) Chem. Soc. Rev., 37, 2530–2542. Karger, J. and Pfeifer, H. (1987) Zeolites, 7, 90–107. Skoulidas, A.I. and Sholl, D.S. (2003) J. Phys. Chem. A, 107, 10132–10141. Xiao, J.R. and Wei, J. (1992) Chem. Eng. Sci., 47, 1143–1159. Pantatosaki, E., Papadopoulos, G.K., Jobic, H., and Theodorou, D.N. (2008) J. Phys. Chem. B, 112, 11708–11715. Krishna, R., van Baten, J.M., Garcia-Perez, E., and Calero, S. (2007) Ind. Eng. Chem. Res., 46, 2974–2986. Xiao, J.R. and Wei, J. (1992) Chem. Eng. Sci., 47, 1123–1141. Beerdsen, E. and Smit, B. (2006) J. Phys. Chem. B, 110, 14529–14530. van den Bergh, J., Ban, S., Vlugt, T.J.H., and Kapteijn, F. (2009) J. Phys. Chem. C. 17840–17850. Coppens, M.O. and Iyengar, V. (2005) Nanotechnology, 16, S442–S448. Vlugt, T.J.H., Krishna, R., and Smit, B. (1999) J. Phys. Chem. B, 103, 1102–1118. Vlugt, T.J.H., Zhu, W., Kapteijn, F., Moulijn, J.A., Smit, B., and Krishna, R. (1998) J. Am. Chem. Soc., 120, 5599–5600. Krishna, R. and van Baten, J.M. (2005) Chem. Phys. Lett., 407, 159–165. Kaerger, J. (2009) Mol. Sieves, 7, 329–366. Karger, J., Petzold, M., Pfeifer, H., Ernst, S., and Weitkamp, J. (1992) J. Catal., 136, 283–299. Hahn, K. and Karger, J. (1998) J. Phys. Chem. B, 102, 5766–5771. Gupta, V., Nivarthi, S.S., Mccormick, A.V., and Ted Davis, H. (1995) Chem. Phys. Lett., 247, 596–600. Jobic, H., Hahn, K., Karger, J., Bee, M., Tuel, A., Noack, M., Girnus, I., and Kearley, G.J. (1997) J. Phys. Chem. B, 101, 5834–5841. de Gauw, F.J.M.M., van Grondelle, J., and van Santen, R.A. (2001) J. Catal., 204, 53–63. Lei, G.D., Carvill, B.T., and Sachtler, W.M.H. (1996) Appl. Catal. A: Gen., 142, 347–359. Neugebauer, N., Braeuer, P., and Kaerger, J. (2000) J. Catal., 194, 1–3.

References 55. Nishiyama, N., Ichioka, K., Park, D.H.,

56.

57.

58.

59.

60.

61. 62.

63.

64.

65. 66.

67.

68.

69.

70.

Egashira, Y., Ueyama, K., Gora, L., Zhu, W.D., Kapteijn, F., and Moulijn, J.A. (2004) Ind. Eng. Chem. Res., 43, 1211–1215. van Vu, D., Miyamoto, M., Nishiyama, N., Egashira, Y., and Ueyama, K. (2006) J. Catal., 243, 389–394. Kocirik, M., Struve, P., Fiedler, K., and Buelow, M. (1988) J. Chem. Soc., Faraday Trans. 1: Phys. Chem. Condens. Phases, 84, 3001–3013. Lehmann, E., Chmelik, C., Scheidt, H., Vasenkov, S., Staudte, B., Karger, J., Kremer, F., Zadrozna, G., and Kornatowski, J. (2002) J. Am. Chem. Soc., 124, 8690–8692. Tzoulaki, D., Heinke, L., Schmidt, W., Wilczok, U., and Karger, J. (2008) Angew. Chem. Int. Ed., 47, 3954–3957. Reitmeier, S.J., Mukti, R.R., Jentys, A., and Lercher, J.A. (2008) J. Phys. Chem. C, 112, 2538–2544. Barrer, R.M. (1990) J. Chem. Soc., Faraday Trans., 86, 1123–1130. Reitmeier, S.L., Gobin, O.C., Jentys, A., and Lercher, J.A. (2009) Angew. Chem. Int. Ed., 48, 533–538. Kortunov, P., Heinke, L., Vasenkov, S., Chmelik, C., Shah, D.B., Karger, J., Rakoczy, R.A., Traa, Y., and Weitkamp, J. (2006) J. Phys. Chem. B, 110, 23821–23828. Tzoulaki, D., Schmidt, W., Wilczok, U., and Kaerger, J. (2008) Microporous Mesoporous Mater., 110, 72–76. Karge, H.G. and Karger, J. (2009) Mol. Sieves, 7, 135–206. Heinke, L., Kortunov, P., Tzoulaki, D., and Karger, J. (2007) Phys. Rev. Lett., 99, 228301–228304. Simon, J.M., Bellat, J.P., Vasenkov, S., and Karger, J. (2005) J. Phys. Chem. B, 109, 13523–13528. Jentys, A., Mukti, R.R., and Lercher, J.A. (2006) J. Phys. Chem. B, 110, 17691–17693. Heinke, L., Kortunov, P., Tzoulaki, D., and Karger, J. (2007) Adsorpt. J. Int. Adsorpt. Soc., 13, 215–223. Ruthven, D.M. (1972) J. Catal., 25, 259–264.

71. Hansen, N., Krishna, R., van Baten,

72.

73.

74.

75. 76.

77.

78.

79.

80.

81.

82.

83.

84.

85.

J.M., Bell, A.T., and Keil, F.J. (2009) J. Phys. Chem. C, 113, 235–246. Christensen, C.H., Johannsen, K., Toernqvist, E., Schmidt, I., and Topsoe, H. (2007) Catal. Today, 128, 117–122. Haag, W.O., Lago, R.M., and Weisz, P.B. (1981) Faraday Discuss., 72, 317–330. Al-Sabawi, M., Atias, J.A., and de Lasa, H. (2008) Ind. Eng. Chem. Res., 47, 7631–7641. Wang, G. and Coppens, M.O. (2008) Ind. Eng. Chem. Res., 47, 3847–3855. Wloch, J. and Kornatowski, J. (2008) Microporous Mesoporous Mater., 108, 303–310. Stavitski, E., Drury, M.R., de Winter, D.A.M., Kox, M.H.F., and Weckhuysen, B.M. (2008) Angew. Chem. Int. Ed., 47, 5637–5640. Karwacki, L., Stavitski, E., Kox, M.H.F., Kornatowski, J., and Weckhuysen, B.M. (2007) Angew. Chem. Int. Ed., 46, 7228–7231. Roeffaers, M.B.J., Ameloot, R., Baruah, M., Uji-I, H., Bulut, M., De Cremer, G., Muller, U., Jacobs, P.A., Hofkens, J., Sels, B.F., and De Vos, D.E. (2008) J. Am. Chem. Soc., 130, 5763–5772. Stavitski, E., Kox, M.H.F., Swart, I., de Groot, F.M.F., and Weckhuysen, B.M. (2008) Angew. Chem. Int. Ed., 47, 3543–3547. Kox, M.H.F., Stavitski, E., Groen, J.C., Perez-Ramirez, J., Kapteijn, F., and Weckhuysen, B.M. (2008) Chem. Eur. J., 14, 1718–1725. Roeffaers, M.B.J., Ameloot, R., Bons, A.J., Mortier, W., De Cremer, G., de Kloe, R., Hofkens, J., De Vos, D.E., and Sels, B.F. (2008) J. Am. Chem. Soc., 130, 13516–1351. Stavitski, E., Kox, M.H.F., and Weckhuysen, B.M. (2007) Chem. Eur. J., 13, 7057–7065. Roeffaers, M.B.J., Sels, B.F., Uji-I, H., Blanpain, B., L’hoest, P., Jacobs, P.A., De Schryver, F.C., Hofkens, J., and De Vos, D.E. (2007) Angew. Chem. Int. Ed., 46, 1706–1709. Roeffaers, M.B.J., Sels, B.F., Loos, D., Kohl, C., Mullen, K., Jacobs, P.A.,

385

386

13 Diffusion in Zeolites – Impact on Catalysis

86. 87.

88. 89.

90.

91.

92.

93.

94.

95.

96.

97.

98. 99.

100.

Hofkens, J., and De Vos, D.E. (2005) Chemphyschem, 6, 2295–2299. Hay, D.G., Jaeger, H., and Wilshier, K.G. (1990) Zeolites, 10, 571–576. ` Kocirik, M., Kornatowski, J., MasarIk, V., Novak, P., Ziknov·, A., and Maixner, J. (1998) Microporous Mesoporous Mater., 23, 295–308. Geus, E.R., Jansen, J.C., and van Bekkum, H. (1994) Zeolites, 14, 82–88. Geier, O., Vasenkov, S., Lehmann, E., Karger, J., Schemmert, U., Rakoczy, R.A., and Weitkamp, J. (2001) J. Phys. Chem. B, 105, 10217–10222. Price, G.D., Pluth, J.J., Smith, J.V., Bennett, J.M., and Patton, R.L. (1982) J. Am. Chem. Soc., 104, 5971–5977. Weidenthaler, C., Fischer, R.X., Shannon, R.D., and Medenbach, O. (1994) J. Phys. Chem., 98, 12687–12694. Weidenthaler, C., Fischer, R.X., and Shannon, R.D. (1994) Zeolites and Related Microporous Materials: State of the Art 1994, Elsevier Amsterdam, New York. vol. 84, pp. 551–558. Agger, J.R., Hanif, N., Cundy, C.S., Wade, A.P., Dennison, S., Rawlinson, P.A., and Anderson, M.W. (2003) J. Am. Chem. Soc., 125, 830–839. Roeffaers, M.B.J., Hofkens, J., De Cremer, G., De Schryver, F.C., Jacobs, P.A., De Vos, D.E., and Sels, B.F. (2007) Catal. Today, 126, 44–53. Mores, D., Stavitski, E., Kox, M.H.F., Kornatowski, J., Olsbye, U., and Weckhuysen, B.M. (2008) Chem. Eur. J., 14, 11320–11327. Meynen, V., Cool, P., and Vansant, E.F. (2007) Microporous Mesoporous Mater., 104, 26–38. Ciesla, U. and Schueth, F. (1999) Microporous Mesoporous Mater., 27, 131–149. Taguchi, A. and Schueth, F. (2005) Microporous Mesoporous Mater., 77, 1–45. Lobo, R.F., Tsapatsis, M., Freyhardt, C.C., Khodabandeh, S., Wagner, P., Chen, C.Y., Balkus, K.J., Zones, S.I., and Davis, M.E. (1997) J. Am. Chem. Soc., 119, 8474–8484. Barrett, P.A., Diaz-Cabanas, M.J., Camblor, M.A., and Jones, R.H. (1998)

101.

102. 103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113. 114.

115.

J. Chem. Soc., Faraday Trans., 94, 2475–2481. Wessels, T., Baerlocher, C., McCusker, L.B., and Creyghton, E.J. (1999) J. Am. Chem. Soc., 121, 6242–6247. Tontisirin, S. and Ernst, S. (2007) Angew. Chem. Int. Ed., 46, 7304–7306. Shvets, O.V., Kasian, N.V., and Ilyin, V.G. (2008) Adsorpt. Sci. Technol., 26, 29–35. Egeblad, K., Kustova, M., Klitgaard, S.K., Zhu, K., and Christensen, C.H. (2007) Microporous Mesoporous Mater., 101, 214–223. Zhu, K., Egeblad, K., and Christensen, C.H. (2007) Eur. J. Inorg. Chem., 25, 3955–3960. Chauvin, B., Boulet, M., Massiani, P., Fajula, F., Figueras, F., and Descourieres, T. (1990) J. Catal., 126, 532–545. Chauvin, B., Massiani, P., Dutartre, R., Figueras, F., Fajula, F., and Descourieres, T. (1990) Zeolites, 10, 174–182. Groen, J.C., Zhu, W.D., Brouwer, S., Huynink, S.J., Kapteijn, F., Moulijn, J.A., and Perez-Ramirez, J. (2007) J. Am. Chem. Soc., 129, 355–360. Groen, J.C., Abello, S., Villaescusa, L.A., and Perez-Ramirez, J. (2008) Microporous Mesoporous Mater., 114, 93–102. Perez-Ramirez, J., Abello, S., Villaescusa, L.A., and Bonilla, A. (2008) Angew. Chem. Int. Ed., 47, 7913–7917. Perez-Ramirez, J., Abello, S., Bonilla, A., and Groen, J.C. (2009) Adv. Funct. Mater., 19, 164–172. Wang, X., Qi, G., and Li, G. (2007) Method for Preparing Nano Zeolite Catalyst and its Use in Methylbenzene and Trimethyl Benzene Transalkylation Reaction, CN1850337-A. ˇ Cejka, J. and Mintova, S. (2007) Catal. Rev., 49, 457–509. Wang, J., Groen, J.C., Yue, W., Zhou, W., and Coppens, M.O. (2008) J. Mater. Chem., 18, 468–474. Wang, J., Groen, J.C., Yue, W., Zhou, W., and Coppens, M.O. (2007) Chem. Commun., 4653–4655.

References 116. Liu, Y., Zhang, W.Z., and Pinnavaia,

117.

118.

119.

120.

121. 122.

123.

124.

125.

126.

127.

T.J. (2000) J. Am. Chem. Soc., 122, 8791–8792. Zhang, Z.T., Han, Y., Xiao, F.S., Qiu, S.L., Zhu, L., Wang, R.W., Yu, Y., Zhang, Z., Zou, B.S., Wang, Y.Q., Sun, H.P., Zhao, D.Y., and Wei, Y. (2001) J. Am. Chem. Soc., 123, 5014–5021. Zhang, Z.T., Han, Y., Zhu, L., Wang, R.W., Yu, Y., Qiu, S.L., Zhao, D.Y., and Xiao, F.S. (2001) Angew. Chem. Int. Ed., 40, 1258–1258. Mazaj, M., Stevens, W.J.J., Logar, N.Z., Ristic, A., Tusar, N.N., Arcon, I., Daneu, N., Meynen, V., Cool, P., Vansant, E.F., and Kaucic, V. (2009) Microporous Mesoporous Mater., 117, 458–465. Corma, A., Diaz, U., Domine, M.E., and Fornes, V. (2000) J. Am. Chem. Soc., 122, 2804–2809. Corma, A., Fornes, V., and Diaz, U. (2001) Chem. Commun., 2642–2643. Corma, A., Fornes, V., ` MartInez-Triguero, J., and Pergher, S.B. (1999) J. Catal., 186, 57–63. Corma, A., Martinez, A., and Martinez-Soria, V. (2001) J. Catal., 200, 259–269. Galletero, M.S., Corma, A., Ferrer, B., Fornes, V., and Garcia, H. (2003) J. Phys. Chem. B, 107, 1135–1141. Nguyen, C.T., Kim, D.P., and Hong, S.B. (2008) J. Polym. Sci. A: Polym. Chem., 46, 725–732. Wu, P., Nuntasri, D., Ruan, J.F., Liu, Y.M., He, M.Y., Fan, W.B., Terasaki, O., and Tatsumi, T. (2004) J. Phys. Chem. B, 108, 19126–19131. Shan, Z., Jansen, J.C., Yeh, Y.T., Koegler, J.H., and Maschmeyer, T. (2003) Catalyst containing microporous

128.

129.

130.

131.

132.

133. 134.

135.

136.

zeolite in mesoporous support and method for making same. International WO 03/045548 A1. Waller, P., Shan, Z.P., Marchese, L., Tartaglione, G., Zhou, W.Z., Jansen, J.C., and Maschmeyer, T. (2004) Chem. Eur. J., 10, 4970–4976. Kortunov, P., Vasenkov, S., Karger, J., Elia, M.F., Perez, M., Stocker, M., Papadopoulos, G.K., Theodorou, D., Drescher, B., McElhiney, G., Bernauer, B., Krystl, V., Kocirik, M., Zikanova, A., Jirglova, H., Berger, C., Glaser, R., Weitkamp, J., and Hansen, E.W. (2005) Chem. Mater., 17, 2466–2474. Kortunov, P., Vasenkov, S., Karger, J., Valiullin, R., Gottschalk, P., Elia, M.F., Perez, M., Stocker, M., Drescher, B., McElhiney, G., Berger, C., Glaser, R., and Weitkamp, J. (2005) J. Am. Chem. Soc., 127, 13055–13059. ` Groen, J.C., Caicedo-Realpe, R., Abello, S., and P´erez-Ram´ırez, J. (2009) Mater. Lett., 63, 1037–1040. Wei, X. and Smirniotis, P.G. (2006) Microporous Mesoporous Mater., 97, 97–106. Zhao, L., Shen, B., Gao, J., and Xu, C. (2008) J. Catal., 258, 228–234. Xiao, F.S., Wang, L.F., Yin, C.Y., Lin, K.F., Di, Y., Li, J.X., Xu, R.R., Su, D.S., Schlogl, R., Yokoi, T., and Tatsumi, T. (2006) Angew. Chem. Int. Ed., 45, 3090–3093. Trong-On, D., Ungureanu, A., and Kaliaguine, S. (2003) Phys. Chem. Chem. Phys., 5, 3534–3538. Musteata, M., Musteata, V., Dinu, A., Florea, M., Hoang, V.T., Trong-On, D., Kaliaguine, S., and Parvulescu, V.I. (2007) Pure Appl. Chem., 79, 2059–2068.

387

389

14 Special Applications of Zeolites Vı´ctor Sebasti´an, Clara Casado, and Joaquı´n Coronas

14.1 Introduction

Zeolites are crystalline, hydrated aluminosilicates having microporous, regular structures. The zeolite micropores are of molecular size, which give them adsorption, catalytic [1], and ion-exchange properties [2] of paramount importance in the chemical industrial ﬁeld. Moreover, interest is growing on the study of new zeolite applications related to process intensiﬁcation [3], green chemistry [4], hybrid materials [5], medicine [6, 7], animal food uses [8], optical- and electrical-based applications [9], multifunctional fabrics [10], and nanotechnology [11]. Furthermore, the concept of zeolite can be extended to the so-called porous tailored materials. These would include oxide molecular sieves, porous coordination solids, porous carbons, sol–gel-derived oxides, and porous heteropolyanion salts [12]. In consequence, the term zeolitic material is frequently used in a very broad sense. This chapter attempts to introduce what could be denominated special applications of zeolites related to coatings and membranes, host–guest interactions, medical and veterinary applications, and others such as racemic separations, magnetic zeolites, and hydrogen storage. 14.2 Zeolite Membranes

Zeolites are versatile materials for the preparation of coatings, selective membranes for molecular separations, and integrating reaction and separation in a solo device; this includes microreactor and -sensor applications. Although there are some pervaporation zeolite membranes commercially available and competitive against other kinds of membranes [13], in general, the reproducibility of zeolite membranes is very difﬁcult [14]: MFI-type zeolite membranes prepared in different laboratories not only present different values of permeance and selectivity but also different qualitative behavior. The reasons for all these discrepancies are not only related to the various thicknesses, the presence of an uncertain number of intercrystalline Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

390

14 Special Applications of Zeolites

defects, and the crystallographic orientation of the membrane but also to the inﬂuence of the porosity and chemical composition of the support on which the zeolite is synthesized [15]. Quite often, the zeolite is synthesized inside the support pores [16], while frequently the support is attacked [17] by the precursor gel of the zeolite affecting both its chemical composition and transport properties. A zeolite membrane prepared inside the support pores could be mechanically stronger than a continuous membrane synthesized on the top of a given support, whose advantage is that microstructure could be easily handled to obtain different crystallographical orientations [18, 19]. This is important since the orientation of a zeolite membrane exerts an important inﬂuence on its performance, as demonstrated in the p-xylene separation from m- or o-xylene [18, 20]. Zeolite membranes are often addressed from the point of view of the formation of a single coating, as in corrosion protection coatings [21], antimicrobial coatings [22], and zeolite-coated catalyst particles [23]. Furthermore, the synthesis of zeolites using ionic liquids, with low vapor pressure and with the double role of solvent and structure-directing agent, has been used to obtain new zeolitic phases and may have application in the manufacture of zeolite coatings at atmospheric pressure [24], suggesting the possibility of large-scale coating at relatively soft conditions without the necessity of conventional autoclaves. Recent innovative approaches to the preparation of zeolite membranes deal with the use of microwave radiation [25, 26], continuous [27] or semicontinuous synthesis systems [28], methods involving separation of nutrients [29], and low-temperature activation conditions [26, 30], among others. The synthesis and application of zeolites as membranes for gas separation, pervaporation, and membrane reactors have been reviewed in depth [31–38]; hence, this chapter focuses on the new applications mentioned above. 14.2.1 Membrane Reactors and Microreactors

One of the challenges for the processes limited by thermodynamic equilibrium is how to achieve high yields of the target products, simultaneously avoiding side reactions. Zeolite membrane reactors have shown a potential application for a wide range of reactions (dehydrogenation, partial oxidation, isomerization or esteriﬁcation, among others) as they combine the reaction and separation processes into one unit. They enable the enhancement of conversion by equilibrium displacement or by selectively removing reaction rate inhibitors [38]. Most applications of zeolite membrane reactors have concentrated on the dehydrogenation [39, 40], synthesis of methyl-tert-butyl ether (MTBE) [41], and esteriﬁcation [42] with MFI, MOR, and LTA-type zeolite membrane reactors. For instance, the dehydrogenation of ethylbenzene to styrene using a silicalite membrane reached an ethylbenzene conversion of 74.8% and a styrene selectivity of 97% at 610 ◦ C, compared with the ﬁxed-bed reactor values of 67.5 and 93%, respectively [40]. In addition, interesting results can be found in the esteriﬁcation of acetic acid with ethanol in a continuous tubular mordenite membrane reactor, obtaining conversions of about 90% that

14.2 Zeolite Membranes

were maintained for ﬁve days with an H2 O/EtOH separation factor as high as 192, and showing a great resistance to the acidic reaction medium [42]. The combination of the concepts of membrane reactor and process miniaturization allows for new routes for chemical syntheses that are more efﬁcient, cleaner, and safer. See in Figure 14.1 the comparison of a conventional membrane reactor (a tube constituting a multitubular reactor) and a microreactor system. Microreactors permit high heat and mass transfer rates, safer intrinsic operation, and easier scale-up by numbering up than conventional reaction systems. The incorporation of zeolites in microreactors as functional elements including catalysts [43] and membranes [44] has given rise to improvements in terms of selectivity and activity enhancement (see Table 14.1). However, it is highly necessary to solve serious problems regarding the adhesion and stability of the zeolite catalyst, as well as the pressure drop along the reactor and the control of the layer thickness. The growth of zeolite ﬁlms directly on the surface of microreactor channels seems to be a suitable strategy to overcome most of the above problems [45]. A clear example of the advantages of this new generation of reactors is the Knoevenagel condensation reaction where the selective removal of the by-product water through a ZSM-5 membrane microreactor led to a 25% improvement in the conversion [46].

5 µm

1 cm (a)

1 cm 100 µm

(b) Figure 14.1

(a) MFI-type zeolite membrane reactor. (b) MFI-type zeolite microreactor.

391

392

14 Special Applications of Zeolites Table 14.1

Performance and application of zeolite microreactors.

Zeolite

Reaction

Role of zeolite

Advantages of microreactor

CsNaX [43]

Knoevenagel condensation

Dehydration and catalyst

Fast mass transfer Narrow residence time distribution Contact enhancement

Selective oxidation of CO (SELOX) Ammoxidation of ethylene

Pt support and nanospace reactor Co support and nanospace reactor

Pt-ZSM-5 [47] Co-Beta zeolite [48]

Overcome the temperature gradients and pressure drops Efﬁcient heat transfer

14.2.2 Zeolite-Based Gas Sensors

A chemical sensor measures the concentration of a ﬁxed chemical compound and transforms this magnitude into an electrical signal, which allows the detection of the presence of this compound in a very short time. Zeolites can either be the essential components of chemical sensors or be used to increase their selectivity and sensitivity, the recent emphasis being on the role of zeolite as a sensitive element [49]. In fact, as shown in Table 14.2, the zeolite can be a functional element, with ion-conducting, adsorption, or catalyst effect. In other cases, the zeolite is considered as an auxiliary element playing the role of a ﬁlter, template, or host [49]. Among the ﬁrst gas-phase devices that were improved by zeolites were the quartz crystal microbalances (QCMs). In these piezoactive devices, the adsorption of a certain specimen produces a mass increase, which is translated into a change of resonance frequency. In a pioneering work, silicalite-1 crystals were coupled to the Table 14.2

Zeolite-based gas sensors. Note that ALPO4 is a pseudozeolite.

Sensor type

Principle of actuation

Property measured

Example of zeolite

Role of zeolite

Quartz crystal microbalances, surface acoustic waves, microcantilevers Capacitors

Piezoelectricity

Resonance frequency

Silicalite-1 [50–52], zeolite A [53, 54], AlPO4 –18 [54]

Adsorption

Dielectricity

Complex impedance Electrical resistance

Zeolite Y [55], ZSM-5 [56] Ferrierite [57], zeolite Y [57], silicalite-1 [58], zeolite A [58, 59]

Ionic conduction

Reactive semiconductor gas sensors

Chemisorption catalysis

Filtering

14.2 Zeolite Membranes

surface of a QCM for sensing ethanol [51]: the regular micropores of the zeolite were found to effectively control molecular access to the device, thus increasing its sensitivity and selectivity. Also, QCM sensors can be modiﬁed by other zeolitic material such as AlPO4 -18 and zeolite A for gas-sensing applications [54]. Surface acoustic wave sensors also exploit the mass change due to the adsorption of a given zeolite (silicalite-1) [52]. This change, similar to QCM sensors, is transformed into a frequency shift. Another class of sensors based on the same principle of piezoelectricity are microcantilevers working in a resonating mode. When some component is selectively adsorbed on the zeolite attached to the end of one of these microcantilevers, a speciﬁc change in the vibration frequency occurred [50, 60]. On the other hand, the fact that the adsorption of certain molecules modiﬁes the dielectric constant of zeolites has been applied to zeolite-coated interdigitated capacitors for sensing gases and butane at concentrations down to 10 ppm have been detected with a PtNaY-covered capacitor that showed no response to CO and H2 [55]. The selectivity of these sensors to hydrocarbons can be increased by introducing a chromium oxide ﬁlm between the gold-interdigitated electrodes and the conducting zeolite ﬁlm [56]. The response of the sensors mentioned above is based on physical changes. The working principle of reactive semiconductor gas sensors relies on the change of conductivity that takes place after exposure to a certain reducing atmosphere. The sensors can be modiﬁed with zeolites to improve their performance by using layers of previously synthesized zeolites as adsorbent barriers to eliminate interfering molecules [57], and by coating the surface of a semiconductor (Pd-doped SnO2 ) gas sensor with a zeolite ﬁlm. To this end, either MFI- or LTA-type zeolites were hydrothermally grown onto the SnO2 sensing surface, acting as barriers for interfering species [58]. The presence of the zeolite ﬁlm strongly reduces, and even suppresses (with the LTA-type zeolite ﬁlm), the sensor response to H2 , CH4 , or propane, while maintaining the sensitivity to CO and ethanol [61], as illustrated in Figure 14.2. In these sensors, together with the chosen zeolite, the quality of CO, EtOH H2, CH4, propane Zeolite layer Pd–SnO2 Alumina substrate Pt heater

Measurer

Figure 14.2 Functional scheme of a reactive semiconductor gas sensor modiﬁed with zeolite.

393

14 Special Applications of Zeolites

the zeolite layer itself is of tremendous importance, which in turn depends on the preparation conditions of the SnO2 [61]; it should be kept in mind that the zeolite is directly synthesized on top of the sensing layer. An improving effect in terms of fast sensor response was achieved through microdropping from a zeolite suspension previously obtained [59], that is, without the need of an in situ hydrothermal synthesis stage, risking the sensor integrity. 14.2.3 Mixed-Matrix Membranes

The above-mentioned difﬁculty of the fabrication of pure inorganic membranes has increased the attention to alternatives in the latest reviews on zeolite membranes [38]. Mixed-matrix membranes (MMMs) are prepared by the incorporation of zeolite particles in a polymer matrix, with the aim of combining the processability of polymers with the molecular-sieving effect of zeolites [38]. The recent developments in both small zeolite particles and glassy polyimides have given scope for many different studies in key gas mixtures, such as air [62–67], H2 /CH4 [68, 69], natural gas [64, 70–72], and most recently CO2 /N2 [65, 71, 73]. The goal is the achievement of gas separation synergetic effects on permeability and selectivity to overcome the so-called ‘‘upper bound’’ for polymeric membranes [74]. This ‘‘upper bound’’ is represented by the relationship (Eq. (14.1)) Pi = k α nij

(14.1)

where i stands for the most permeable gas component, α ij is the separation factor of the gas pair, and k, is referred to as the front factor and n, the slope in the log–log plot of this relationship, which is represented in Figure 14.3 for several gas pairs with updated data on pure polymeric membranes [75]. The permeability–selectivity 103 Scope for MMM

O2 / N2 Separation factor

394

102 CO2 /CH4 CO2 /N2 101

Polymer membranes

H2 /CH4 100 10−13

10−8

10−3

102

107

Permeability (barrer)

Figure 14.3 Robeson’s upper bound [75] for the separation by polymeric membranes of typical gas mixtures.

14.2 Zeolite Membranes

trade-off for pure polymer membranes would be under the corresponding upper bound and scope for MMM development over this line. The permeability (P) of a gas through a membrane is proportional to the solubility (S) and diffusivity of the gas in the membrane (P = D·S). Adding inorganic nanoﬁllers into a polymer may affect gas separation in two ways: (i) the interaction between the polymer chains and nanoﬁllers may disrupt the polymer packing and increase the voids, and thus the gas diffusion and (ii) the functional groups on the surface of the inorganic ﬁller may interact with polar gases such as CO2 , improving the penetrant solubility in the membrane [76]. The effect of adding inorganic ﬁllers or molecular sieves into polymer membranes is generally described by Maxwell’s equation to model gas permeability across MMMs [70, 77], as the behavior of gases varies as a function of the spatial relationship of gas in the polymer matrix, as follows: Pd + 2Pc − 2d (Pc − Pd ) (14.2) Peff = Pc Pd + 2Pc + d (Pc − Pd ) where Peff , Pd , and Pc are the effective phase, dispersed phase, and continuous phase permeabilities, respectively, and d is the volume fraction of the dispersed phase or ﬁller. The problem of Maxwell’s equation is that it neglects the interaction between nanoﬁllers and polymer chains, nanoﬁllers and penetrants, so that poor adhesion between dispersed and continuous phases results in nonideal deviations from Eq. (14.2). When there are nonselective interfacial voids larger than penetrating molecules, which reduce the apparent selectivity and increase permeability, this is called sieve-in-a-cage morphology [68, 77]. Even if polymer–zeolite adhesion is good, a reduction in free volume may occur near the sieve surface, and this is called matrix rigidiﬁcation [78]. A model considering the diffusion of a component within a membrane containing a zeolite phase has been considered to be a Maxwell–Stefan model [79], which has been widely applied in the prediction of diffusive transport across inorganic membranes and materials. Several procedures have been proposed for improving the adhesion of zeolite and polymer phases in an MMM. The ﬁrst is the choice of polymer. If the compatibility between the inorganic ﬁller and a rubbery polymer is good due to the high mobility of the polymeric chains, the permselectivity reached is usually not very much higher than for the pure polymer [71]. Most researchers have thus focused on rigid polymers such as polyimides [69, 80, 81] and followed several strategies to modify the zeolite sieve surface to improve the adhesion and reduce excessive rigidiﬁcation of the matrix. Use of aminosilanes as coupling agents has also been explored [70] for incorporating zeolite A in a polyimide matrix. While this decreases the voids between polymer and zeolite, the permeability is also often dramatically reduced. For this purpose, block copolymers composed of a rigid polyimide part providing selectivity and a ﬂexible PDMS part have been proposed to embed zeolite L into a gas separation membrane [81], without the use of coupling agents. Terpenic resin can enhance chain ﬂexibility of glassy polyetheretherketone before incorporating zeolite A particles [65]. The aspect ratio of ﬁller particles may affect the dispersability and adhesion [71], and Maxwell’s equation has been successfully modiﬁed to predict

395

396

14 Special Applications of Zeolites

the gas separation performance through platelet-shaped molecular-sieving particles [82]. Another approach is the use of mesoporous zeolite ZSM-5 nanoparticles in a polyimide matrix [83], where the good contact is attributed to the ability of the polymer chains to penetrate into the mesopores, while the micropores provide shape and size selectivity, a H2 /CH4 separation factor of 168 being reported. Small pore Nu-6(2)-polysulfone hybrid membranes have also been obtained with a good enough contact to avoid any by-passing gas effect in the H2 /CH4 separation [68].

14.3 Host–Guest Interactions

With regard to zeolites, host–guest interactions describe complexes that are composed of a certain zeolite structure (host) and a molecule or ion (guest) held together by means of links usually differing from those of full covalent bonds (hydrogen bonding, van der Waals force, π-stacking, electrostatic interactions). The regular microporous structure of the zeolite, besides the conﬁnement to which the guest is submitted (in fact, the guest is immobilized in the zeolite framework), guarantees an improvement of the chemical and physical stability of the guest, whereas the porosity of the host keeps open the access of the guest to the surrounding ambient. This is crucial not only in photonic applications [84] but also in heterogeneous catalysis, where the term zeozyme has been often used to describe a situation in which a metal-encapsulated complex imitates the role of the active site, while the zeolite mantle replaces the surrounding protein moiety of metalloenzymes [85]. In addition, the practically unlimited combination of pore size and chemical composition associated with zeolites and related materials adds to these host–guest hybrids the possibility of tailoring their performance as drug delivery systems, catalysts, chromophores, and storage devices, among others. Another important issue deals with the ability of reaching supramolecular organization or patterning of guest molecules in nanochannels. In particular, the parallel channels of zeolite L are decisive to obtain zeolite crystals containing oriented ﬂuorophores mimicking the antenna system of green plants [86] for light-harvesting applications. Figure 14.4 is a scheme related to host–guest interactions and deals with the general aspects related to their synthesis, advantages, and proposed applications that are treated below. A pioneering strategy used to encapsulate different kinds of active components consisted in the entrapment of organic compounds into sol–gel matrixes [87, 88]. At relatively low temperature (<373–423 K), in which the desired guest is not thermally damaged, the sol–gel process has been used for the preparation of porous gel solids (SiO2 , Al2 O3 , TiO2 , etc.) doped with organic molecules. In the sol–gel process, the desired compounds to be encapsulated can be dissolved together with the precursors for the preparation of a colloidal suspension, which upon gelation would entrap (encapsulate) the guest molecules, such as rhodamine [89] or different enantiomer pairs [90]. The usual precursors are inorganic salts and organic compounds such as alkoxides, which are also the most widely employed in

HO

Si O OH OH Si HO OH HO HO O Si OH O CH3 Si N O HO OH OH CH3 OH Si OH HO OH O

OH

Sensors, lasers Optical switches Catalysis Drug delivery Storage devices Light harvesting Functionalization

Thermal stability Mechanical stability Tunability Optical transparency Light (UV) protection

Milling Solid-state reaction Ion exchange Sorption Grafting, anchoring Ship in the bottle Coinclusion

Metals Chromophores Metal complexes Drugs Additives

Guests

Figure 14.4 Examples of porous hosts (nonordered and ordered mesoporous silicas and zeolites) and host–guest interactions (related to the preparation of the hybrids and their advantages and possible applications).

Applications

FAU-type zeolite

Advantages

Host-guest interactions

Ordered mesoporous matrix (MCM-41-type)

Type of host

Preparation

Methylene blue in a nonordered porous solid

Si O HO HO Si OHHO OH Si O N HO OH H3C N+ S O− HO Si CH3 OH Si HO O HO O HO OH Si O Si Si HO HO OH HO HO

HO

O

HO HO O Si

14.3 Host–Guest Interactions 397

398

14 Special Applications of Zeolites

the sol–gel approach. The resulting materials are not ordered unless, for instance, a cationic surfactant is introduced in the synthesis sol, following a methodology parallel to that employed to obtain M41S-type ordered mesoporous silicas [91]. Host–guest composites based on zeolites and ordered related materials with pores in the range of micropores and even mesopores present advantages over nonordered silica-based porous materials owing to their well-deﬁned, uniform-sized systems of cages and channels. In this case, the zeolite–guest composite can be produced through many different successful strategies. Co and Ru complexes have been entrapped in the supercages of zeolite Y after producing CoY zeolite from NaY zeolite by ion exchange (ion exchange itself is also an encapsulation method where the guest is an organic or organometallic cation [92]) and then being mixed at 473 K in an excess of bipyridine ligand [93]. This procedure can be included in the so-called ship-in-the-bottle method. The method consists of the synthesis of the guest inside the zeolite microporosity from the corresponding direct precursors. These are small enough to diffuse inside the zeolite host and to generate the desired bulky species unable to go out. For instance, Co(salen) complexes are encapsulated by a similar two-step procedure: Co ion exchange followed by contact with salen ligand (C16 H16 N2 O2 ) [94]. The heterogeneous catalyst obtained was recycled at least ﬁve times without a signiﬁcant loss of activity in the oxidative carbonylation of aniline. Entrapping can be also the encapsulation during the synthesis of the zeolite, as for the inclusion of methylene blue (C16 H18 N3 S+ ) during the hydrothermal synthesis of NaY zeolite [95]. Light-emitting boron nitride (BN) nanoparticles-encapsulated-ZSM-5 was obtained by mechanical mixing of ZSM-5 with B2 O3 , followed by thermal treatment at 973 K [96]. If the size of the guest molecule is smaller than the free diameter of the zeolitic channels, gas or liquid sorption of the guest is possible in evacuated zeolites, as in the case of nile red (C20 H18 N2 O2 ), with a length of 1 nm and an approximate width of 0.6 nm, in NaY zeolite [97]. On the other hand, NaX zeolite exhibits a high degree of ﬂexibility, and at 453 K the inclusion of bulky molecules (with minimum diameters of 0.90–0.95 nm) was possible, even though the guest diameters exceeded the nominal 0.74 nm pore opening of the FAU-type zeolite [98]. Solid-state sorption is also a plausible option of encaging and NaX-encapsulated thioindigo (C16 H8 O2 S2 ) was prepared by solid-state reaction at 573 K between zeolite and dye [99]. Because this molecule (0.6 nm × 1.3 nm approximate dimensions) melts at 623 K, adsorption from the solid state took place, X-ray diffraction measurements revealing the location of the dye molecules in the FAU-type framework. Finally, covalent grafting or anchoring to the pore walls of the host for increasing the bonding strength with the guest is useful not only for mesoporous molecular sieves but also for zeolites [100]. Compiling the applications mentioned above, host–guest interactions have use in catalysis to heterogenize homogeneous catalysts [94], optochemical sensors [97], lasers [101], and other light-related systems such as optical switches. In the latter case, the gas permeation of an azobenzene-adsorbed FAU-type zeolite membrane changes by photoinduced switching, and the N2 /CO2 and CH4 /CO2 separation factors are higher at trans- than at cis-switching [39]. Besides, drug delivery (which is discussed later, perhaps, more related to the use of ordered mesoporous materials

14.4 Medical and Veterinary Applications

[102]), and what could be deﬁned as functionalization (for instance, the use of Ag– and Zn– exchanged zeolites as antimicrobial-ﬁnished textile products [103]), are ﬁelds of potential development attributed to host–guest interactions.

14.4 Medical and Veterinary Applications

The multiple uses of both natural and synthetic zeolitic materials are based on their adsorptive, ion-exchange, molecular-sieving, lack of toxicity, and catalytic properties. Over the last few years, many applications for zeolites have also emerged in the veterinary and medical ﬁelds [6, 104]. The principles of these new applications are in the established synergistic host–guest interaction framework already introduced. 14.4.1 Medical Applications

Considering the particular structure of zeolites, that is, the channels and cavities linked together, the variety of pore sizes and shapes [105], and the fact that the basic structure is biologically neutral, zeolites have a wide range of complementary applications from biochemistry, agro industry, detergents, and soil improvements to the nuclear industry [106]. Zeolites can contain water as a part of their structure; after this has been removed by heating, the basic framework is left intact and other solutions can be put through the structure, and thus the zeolite acts as a delivery system for the new ﬂuid. This capacity of selective adsorption and subsequent controlled release of different ions, which can be enhanced or modiﬁed by the presence of surfactants on the external surface [104], led to the consideration of zeolites as drug delivery systems. For instance, in gastrointestinal applications, zeolite X crystals [107] or ZSM-5 microspheres [108] are stable even at the low pH levels of the stomach, for releasing of ketoprofen and biotin/antibody, respectively, whereas tetracycline antacid, one of the ﬁrst broad-spectrum drugs able to control the pH in the stomach, is stored and released from CAN-type zeolite [109, 110]. In addition, antitumoral drug control release was proposed using zeolite CuX [111], and recently magnetic drug delivery, where the zeolite is the shell protecting a magnetic core, has also shown potential in the loading and release of doxorubicin [112]. Given the importance of NO as a biological agent in cardiovascular, nervous, and immune systems, the delivery of exogenous NO is an attractive therapy for a number of ailments. The range and diversity of its effects make the development of materials that can store signiﬁcant quantities of NO and deliver it in a controlled way to speciﬁc sites in the body a major area of research [113]. Many systems have been explored, for instance, co-exchanged zeolite A has been studied for the storage and delivery of NO in cardiovascular centers for the prevention of thrombosis on artiﬁcial surfaces [114]; this co-exchanged zeolite is able to store NO for days in dry

399

400

14 Special Applications of Zeolites

air, which can then be removed in the presence of moist gas to speciﬁc sites in the body. Pressure equalization steps have been seen to enhance recovery of component in pressure swing adsorption (PSA) processes ranging from a small-scale unit for medical oxygen using Ag-exchanged zeolite [115] for the production of O2 -rich air for patients with respiratory problems, to a large-scale hydrogen production unit [116]. The efﬁciency of zeolites to adsorb gases such as CO2 is of interest in anesthesia [117]. Diagnostic magnetic resonance imaging (MRI) [118] is another application of zeolites in medicine. MRI relies on administering contrast agents to patients to improve the diagnostic value of the resonance imaging. Contrast agents contain high spin metals that bind water molecules, thus yielding proton spin relaxation times faster than those with free water. However, these contrast agents cannot be administered directly owing to their toxicity. Zeolites are biocompatible nontoxic materials, where these agents can be lodged and delivered at the appropriate rate. Gd3+ is considered a good paramagnetic relaxation agent because of its large magnetic moment and nanosecond spin time. Therefore, it has been chosen by several authors as the exchangeable cation with some zeolites to provide Gd-zeolite nanocomposites with potential in radionuclide imaging and anticancer therapies [119]. Eu3+ and Tb3+ have been used to dope zeolite Y [120] to improve the linearity of the photostimulated luminescence intensity to the X-ray irradiation dose. As discussed above, the use of zeolites in sensors that detect small amounts of particular compounds in body ﬂuids is also important in medical applications [121–123]. Protein immobilization onto the surface of zeolites plays a role in the preparation of biocompatible materials for surgical implants: the adsorption of cytochrome c, one of the most characterized proteins, on FAU-, BEA-, and MFI-type zeolites has been studied [124]. Of late, zeolites have been presented as a means of controlling hemorrhage by adsorbing free water in blood and promoting clot formation, avoiding direct contact with the site of the injury, as advantageously as other biomaterials such as chitosan [125]. Ca-exchanged aluminophosphates are also reported to clot blood with 50% less energy release than conventional materials [7], reducing the burning of tissue and the subsequent pain. 14.4.2 Veterinary Applications

Mycotoxins are secondary metabolites coming from a diverse range of fungi, which are known to produce injurious effects on the health of animals and humans, such as carcinogenicity and mutagenicity. One of the most recent approaches to the prevention of mycotoxicosis in livestock is the addition, in the diet, of the nonnutritional adsorbents that bind mycotoxins in the gastrointestinal tract and capable of reducing their bioavailability [126, 127]. The best known mycotoxins are aﬂatoxins, which are most predominant in feedstock [128]. The adsorptive properties of zeolites and their low cost and availability in most countries make them attractive. In this context, the addition of natural zeolites and clays and bentonite to animal feeds is an easy way of protecting livestock from mycotoxins [8, 127]. The adsorption efﬁciency depends on the physicochemical structure of the

14.5 Other Applications

Animal feed

Zeolite adsorbent

Animal (e.g., chicken)

Toxin free animal for consumption

Toxins adsorbed in zeolite

Waste disposal Figure 14.5

Zeolites in the animal production cycle.

adsorbent, including dose, total charge and charge distribution, the size of pores and speciﬁc surface area, as well as physicochemical properties of the mycotoxin. In vitro experiments of mycotoxin binding by bentonite [127] and clinoptilolite [8] have been conducted with promising yet preliminary results for the prevention of aﬂatoxicosis in poultry. A method for screening the aﬂatoxin metabolites in Vietnamese pig urine as biomarker to mycotoxin exposure has been recently developed using several zeolites [128]. Other applications are supplementing diets with Ca- and P-binding zeolite A to manipulate dry cow rations in order to inﬂuence blood calcium levels in nonpregnant dairy cows so that parturient hypocalcemia (milk fever) is prevented [129]. The cation-exchange properties of zeolites have been evaluated for reducing ammonia emissions in the manure from excreta [130]. The role of the zeolite in the animal production chain is schematically shown in Figure 14.5. Of course, in vivo experiments and information on the toxin-binding andreleasing mechanisms still need to be investigated.

14.5 Other Applications

In addition to zeolite applications related to membranes and host–guest interactions, other emerging applications dealing with racemic separations, magnetic zeolites hydrogen storage, and solar adsorption are considered below. 14.5.1 Racemic Separations

Because of the importance of chirality in biological processes, research efforts have been targeted at the synthesis of chiral zeolites and related porous solids [131–134]. Besides the wide range of synthetic zeolitic materials obtained to date, only zeolite beta polymorph A [135] and microporous titanosilicate ETS-10[136]

401

402

14 Special Applications of Zeolites

are chiral. Nevertheless, these materials are very difﬁcult to obtain in their pure form. Attempts to synthesize chiral zeolites relying on chiral surfactants, which should organize silicate–surfactant assemblies into chiral conformations, have been doomed to failure. This is mainly because calcination at high temperature destroys the chiral conformation of such assemblies, resulting in either achiral or racemic mixtures of zeolites [137]. On the other hand, small changes in the nature of the catalyst utilized in chirotechnology applications can lead to loss of enantioselectivity [137, 138]. Therefore, zeolite catalysts with a chiral modiﬁcation are being used for that approach because of their well-deﬁned crystalline structure. For example, a chiral modiﬁcation of zeolite Y by dithiane oxide can lead to a catalyst able to carry out an enantioselective dehydration of racemic butan-2-ol [139]. This chiral modiﬁcation signiﬁcantly enhances the rate of dehydration of butan-2-ol to butenes in an enantiomerically discriminating pathway, since one enantiomer reacts preferentially to the other (70% conversion of S-Butan-2-ol vs. less than 2% for R-Butan-2-ol), although both are present within the micropores under same the reaction conditions. 14.5.2 Magnetic Zeolites

Zeolites are materials whose periodical spaces of molecular size are well arranged to have attracting forces to conﬁne electrons. Different types of cationic clusters can be generated in the porosity of the zeolite framework by adsorption of alkali metals into dehydrated zeolites. Many body effects, such as spontaneous magnetization or electronic conduction, have been observed if the clusters with s-electronic orbitals are aligned with an appropriate interaction: ferromagnetism in K clusters in LTA [140] and antiferromagnetism in Na clusters in SOD [141]. These phenomena are due to the framework of zeolites involving not only well-separated cages, but also windows with relatively small radius that connect cages to each other. The electrons of the adsorbed atoms can interact not only within one cage but also through the windows [142]. Magnetically modiﬁed zeolites offer the advantage of easy separation and recovery of powdered zeolites after catalytic runs or after environmental remediation by means of an applied magnetic ﬁeld [143]. There are mainly two methods to obtain magnetically modiﬁed zeolites, the ﬁrst refers to ion exchange followed by the reduction of the cations to their metallic state. The second method involves the modiﬁcation of zeolites with colloidal magnetite particles. On the other hand, the adsorption of paramagnetic molecules (e.g., NO) in zeolites is directly related not only to the magnetic ﬁeld applied but also to the porosity [144]. This result could support the application of ferromagnetic zeolite membranes for air separation. As ferromagnetic materials can be magnetized, a permeoselective ferromagnetic membrane based on speciﬁc interactions with paramagnetic O2 and diamagnetic N2 molecules [145] could lead to a very efﬁcient system for air separation.

14.5 Other Applications

14.5.3 Hydrogen Storage

The development of materials that are able to uptake, store, and release H2 on demand is crucial for the development of the so-called hydrogen economy [146, 147]. Safe and high-energy density storage of hydrogen is a key requirement for using hydrogen as a fuel. Sorption processes (physisorption and chemisorption) typically require highly porous materials in which the surface area available for hydrogen sorption is maximized, thus allowing the easy uptake and release of hydrogen to and from the material. In such a way, zeolites are potential hydrogen storage materials [148, 149]. The diameter of the cages and the channels of zeolites can be controlled by ion exchange, modifying the size and the valence state of the exchangeable cations. These tunable modiﬁcations have a direct effect on the chemical nature of the possible binding sites, the void space, and the ease of access of H2 to the zeolite structure [150]. Zeolites exhibit diverse behavior with respect to hydrogen adsorption [151, 152], depending on both the framework structure and the nature of the compensation cations. In addition, cations may act as binding sites for hydrogen molecules. The highest gravimetric storage capacity of 2.19 wt% was obtained for zeolite CaX, with a volumetric storage density of up to 31.0 kg of H2 per m3 (143 H2 molecules per unit cell) [151]. However, this value is below the H2 storage capacity obtained by other materials (metal hydride, MOF, and activated carbon, among others) (see Figure 14.6). A zeolite layer can be made impermeable to hydrogen at 293 K, even at hydrogen pressure as high as 6.6 MPa, by adsorbing methanol in the zeolite layer [156]. 8

77 K

H2 gravimetric capacity (%)

7

77 K

6

333 K

5 4 3 77 K

2 1 0 Zeolite

Metal hydride

MOF

Carbon material

H2 storage materials Figure 14.6 H2 gravimetric capacity of different materials: zeolite (CaX [151]), metal hydride (sodium alanate [153]), MOF (manganese benzenetristetrazolate [154]), and carbon material (chemically activated carbon [155]).

403

404

14 Special Applications of Zeolites

At higher temperatures, methanol desorbs and H2 ﬂux increases. The hydrogen permeance could be controlled by the rate of methanol on the feed side. This could be an example of storing H2 at high pressures in small pellets coated with a thin zeolite layer, which would allow high pressure inside the pellet and low pressure outside. Finally, other application of zeolites is solar adsorption for ice-making and air-conditioning systems [157, 158]. Solar refrigeration is an important use of solar energy because the supply of solar energy and the demand for cooling are greatest during the same season. The solar-powdered solid adsorption refrigeration system has been presented as a promising environmentally friendly and low-cost technology, and several inorganic material/cooling ﬂuid pairs have been studied. This is based on the fact that heat adsorption is stronger than condensation to liquid phase, thus transport of adsorbate from the liquid phase to the adsorbent occurs in the form of vapor, causing the temperature of the liquid to be lowered while that of the adsorbent rises. Thus, an important factor of the optimization of the refrigeration process is based on the thermodynamic equilibrium of the adsorbent/adsorbate working pairs. In the case of zeolites, the zeolite/water pair has shown some refrigerating effect, and it is one of the most widely studied pairs [158].

14.6 Summary and Outlook

Although signiﬁcant progress has been achieved in the last few years in the synthesis of zeolite membranes and materials, and in their applications, these still face some problems regarding reproducibility. Consequently, the preparation of membranes has been improved by several methods. In particular, microwave irradiation can considerably reduce synthesis time, which results in changes in morphology, crystallographic orientation, and composition compared with conventional heating. Other new approaches to prepare zeolite membranes involve continuous and semicontinuous synthesis systems, the method of separation of nutrients, low-temperature activation conditions, and the use of ionic liquids. With regard to industrial applications, at present, there are only a few pilot plants operating with zeolite membranes, mainly in the dewatering of alcohols. It seems to be only a question of time before a pilot plant using zeolite membranes for gas separation becomes feasible. Zeolite membranes can also be used in membrane reactors, where reaction and separation are combined, thereby enhancing the conversion by equilibrium displacement or selectively removing reaction rate inhibitors. From the point of view of process intensiﬁcation, microreactor technology is a new concept of the chemical reactor. Zeolites have been known as useful catalysts in a large variety of reactions, and zeolite microreactors may provide more efﬁcient, cleaner, and safer routes for chemical syntheses than conventional systems. However, one of the challenges that still remains is the development of a microreactor with a

14.6 Summary and Outlook

homogeneously distributed, mechanically stable, reproducible catalytic load that is sufﬁcient to achieve the desired yield under a given set of reaction conditions. Physical sensors are based on the change of response, be it resonance frequency, acoustic wave, or any vibration frequency originating from mass change produced by adsorption of a certain specimen on the sensor surface. Similarly, the working principle of reactive semiconductor gas sensors lies on the change of conductivity taking place after exposure to a certain reducing atmosphere. In any case, the interest in zeolites as sensors relies in their use as the sensitive component or as an auxiliary element with the primary role of a ﬁlter. Many challenges still remain concerning miniaturization, high throughput application to solve complex problems, modeling, and cost. Zeolites possess a useful structure for a wide range of host–guest interactions, guaranteeing an improvement of both the chemical and physical stability of the guest, while the porosity of zeolites keeps open the access of the guest to the surrounding ambient. This, together with the practically unlimited combination of pore size and chemical composition, makes zeolites a valuable material for drug delivery, storing, magnetic, photonic, and heterogeneous catalysis applications. Chirality is an important ﬁeld in biological and drug synthesis and puriﬁcation processes; therefore, research efforts have been targeted at the synthesis of chiral zeolites. As attempts to synthesize pure chiral zeolites have not been successful enough, a new approach based on the use of chiral surfactants and structure-directing agents is being applied to produce enantioselective materials. In the last few decades, the development of MMMs from both pure zeolite and polymer membranes has resulted in improved gas separation, membrane performance, and processability. Zeolites have been introduced as nanoﬁllers in polymer matrixes. Zeolites with several shapes and structures, as also different polymers, have been used for the key gas separations currently in vogue: O2 /N2 , H2 /CH4 , CO2 /CH4 , and recently, CO2 /N2 . Some results have shown promise in overcoming the maximum permselectivity upperbound performance for polymer membranes, but requires more research on the adhesion of nanoparticles and polymer chains to make stable, durable, and defect-free MMM. Accurate, speciﬁc mass transport models predicting the performance across MMM are also needed. Given the particular structure of zeolites, that is, the channels and cavities linked together, the variety of pore sizes and shapes and the fact that the basic structure is biologically neutral, zeolites have long attracted attention in medical applications such as drug delivery systems for gastrointestinal applications, antitumoral drug control release, and magnetic drug delivery. Therefore, zeolite application research and development seems to have become a hot topic for the coming years. Nevertheless, the real applicability of many of these forwarding applications has to be ﬁrst tested in vivo. With regard to veterinary applications, the contamination of foodstuffs and feedstuffs by mycotoxins is a worldwide problem and seems to be unavoidable, even though a considerable number of ﬁeld and preservation studies have been carried out to prevent mycotoxins. Several studies have focused on zeolites as a means of adsorbing mycotoxins from chicken, pig, or cow feed and thus removing

405

406

14 Special Applications of Zeolites

them from the food chain, with some in vitro studies showed promising results. In vivo experiments are yet to be addressed. Despite exhibiting hydrogen adsorption, because structural cations act as binding sites for hydrogen molecules, the capacity of zeolites as gravimetric storage systems is below that obtained by other materials (metal hydride, MOF, and activated carbon, among others). However, it is an interesting fact that a zeolite layer with methanol adsorbed can tune the ﬂux of H2 as a function of temperature.

References 1. Corma, A., Rey, F., Rius, J., Sabater,

2.

3. 4.

5.

6.

7.

8. 9.

10. 11. 12.

13.

M.J., and Valencia, S. (2004) Nature, 431, 287–290. Kuznicki, S.M., Bell, V.A., Hillhouse, H.V., Jacubinas, R.M., Braunbarth, C.M., Toby, B.H., and Tsapatsis, M. (2001) Nature, 412, 720–724. Stankiewicz, A. (2003) Chem. Eng. Process., 42, 137–144. Anastas, P.T., Kirchhoff, M.M., and Williamson, T.C. (2001) Appl. Catal. A: Gen., 221, 3–13. Choi, S., Coronas, J., Jordan, E., Oh, W., Nair, S., Okamoto, F., Shantz, D.F., and Tsapatsis, M. (2008) Angew. Chem.Int. Ed., 47, 552–517. Danilczuk, M., Dlugopolska, K., Ruman, T., and Pogocki, D. (2008) Mini Rev. Med. Chem., 8, 1407–1417. Galownia, J., Martin, J., and Davis, M.E. (2006) Microporous Mesoporous Mater., 92, 61–63. Oguz, H. and Kurtoglu, V. (2000) Br. Poult. Sci., 41, 512–517. Schwenn, H.J., Wark, M., Schulz-Ekloff, G., Wiggerss, H., and Simon, U. (1997) Colloid Polym. Sci., 275, 91–95. Grancaric, A.M., Markovic, L., and Tarbuk, A. (2007) Tekstil, 56, 533–542. Tsapatsis, M. (2002) AIChE J., 48, 654–660. Barton, T.J., Ull, L.M.B., Klemperer, W.G., Loy, D.A., McEnaney, B., Misono, M., Monson, P.A., Pez, G., Scherer, G.W., Vartuli, J.C., and Yaghi, O.M. (1999) Chem. Mater., 11, 2633–2656. Urtiaga, A.M., Gorri, E.D., G´omez, P., ˜ ez, R., and Ortiz, I. Casado, C., Ib´an (2007) Drying Technol., 25, 1819–1828.

14. Navajas, A., Mallada, R., Tellez, C.,

15.

16.

17.

18.

19.

20.

21.

22.

23.

24. 25.

Coronas, J., Menendez, M., and Santamaria, J. (2007) J. Membr. Sci., 299, 166–173. Bernal, M.P., Coronas, J., Menendez, M., and Santamar´ıa, J. (2003) Microporous Mesoporous Mater., 60, 99–110. Miachon, S., Landrivon, E., Aouine, M., Sun, Y., Kumakiri, I., Li, Y., Prokopova, O.P., Guilhaurne, N., Giroir-Fendler, A., Mozzanega, H., and Dalmon, J.A. (2006) J. Memb. Sci., 281, 228–238. Geus, E.R., Denexter, M.J., and Vanbekkum, H. (1992) J. Chem. Soc., Faraday Trans. 1, 88, 3101–3109. Lai, Z.P., Bonilla, G., Diaz, I., Nery, J.G., Sujaoti, K., Amat, M.A., Kokkoli, E., Terasaki, O., Thompson, R.W., Tsapatsis, M., and Vlachos, D.G. (2003) Science, 300, 456–460. Choi, J., Ghosh, S., Lai, Z.P., and Tsapatsis, M. (2006) Angew. Chem. Int. Ed., 45, 1154–1158. Lai, Z.P., Tsapatsis, M., and Nicolich, J.R. (2004) Adv. Funct. Mat., 14, 716–729. Cheng, X.L., Wang, Z.B., and Yan, Y.S. (2001) Electrochem. Solid State, 4, B23–B26. McDonnell, A.M.P., Beving, D., Wang, A.J., Chen, W., and Yan, Y.S. (2005) Adv. Funct. Mat., 15, 336–340. Nishiyama, N., Ichioka, K., Park, D.H., Egashira, Y., Ueyama, K., Gora, L., Zhu, W.D., Kapteijn, F., and Moulijn, J.A. (2004) Ind. Eng. Chem. Res., 43, 1211–1215. Morris, R.E. (2008) Angew. Chem. Int. Ed. Engl., 47, 442–444. Motuzas, J., Heng, S., Lau, P., Yeung, K.L., Beresnevicius, Z.J., and Julbe, A.

References

26.

27.

28.

29.

30.

31. 32. 33. 34. 35.

36.

37.

38.

39.

40. 41.

42.

(2007) Microporous Mesoporous Mater., 99, 197–205. Motuzas, J., Julbe, A., Noble, R.D., van der Lee, A., and Beresnevicius, Z.J. (2006) Microporous Mesoporous Mater., 92, 259–269. Pera-Titus, M., Bausach, M., Llorens, J., and Cunill, F. (2008) Sep. Puri. Technol., 59, 141–150. Pina, M.P., Arruebo, M., Felipe, A., Fleta, F., Bernal, M.P., Coronas, J., Menendez, M., and Santamaria, J. (2004) J. Membr. Sci., 244, 141–150. Mateo, E., Lahoz, R., de la Fuente, G.F., Paniagua, A., Coronas, J., and Santamaria, J. (2007) Chem. Mater., 19, 594–599. Heng, S., Lau, P.P.S., Yeung, K.L., Djafer, M., and Schrotter, J.C. (2004) J. Membr. Sci., 243, 69–78. Coronas, J. and Santamaria, J. (1999) Sep. Purif. Methods, 28, 127–177. Bein, T. (1996) Chem. Mater., 8, 1636–1653. Coronas, J. and Santamar´ıa, J. (2004) Top. Catal., 29, 29–44. Tavolaro, A. and Drioli, E. (1999) Adv. Mater., 11, 975–996. Caro, J., Noack, M., Kolsch, P., and Schafer, R. (2000) Microporous Mesoporous Mater., 38, 3–24. Bowen, T.C., Noble, R.D., and Falconer, J.L. (2004) J. Membr. Sci., 245, 1–33. McLeary, E.E., Jansen, J.C., and Kapteijn, F. (2006) Microporous Mesoporous Mater., 90, 198–220. Caro, J. and Noack, M. (2008) Microporous Mesoporous Mater., 115, 215–233. Weh, K., Noack, M., Hoffmann, K., Schroder, K.P., and Caro, J. (2002) Microporous Mesoporous Mater., 54, 15–26. Kong, C., Lu, J., Yang, J., and Wang, J. (2007) J. Membr. Sci., 306, 29–35. Salom´on, M.A., Coronas, J., Men´endez, M., and Santamar´ıa, J. (2000) Appl. Catal. Gen., 200, 201–210. ´ Mallada, R., de la Iglesia, O., Men´endez, M., and Coronas, J. (2007) Chem. Eng. J., 131, 35–39.

43. Cai, H., Zhang, X., Liu, H., and

44. 45.

46.

47.

48.

49.

50.

51. 52. 53. 54.

55.

56.

57. 58.

59.

Yeung, K.L. (2007) Chinese J. Catal., 28, 758–760. Chau, J.L.H., Leung, A.Y.L., and Yeung, K.L. (2003) Lab Chip, 3, 53–55. Sebasti´an, V., de la Iglesia, O., Mallada, R., Casado, L., Kolb, G., Hessel, V., and Santamar´ıa, J. (2008) Microporous Mesoporous Mater., 115, 147–155. Lai, S.M., Martin-Aranda, R., and Yeung, K.L. (2003) Chem. Commun., 218–219. de la Iglesia, O., Sebasti´an, V., Mallada, R., Nikolaidis, G., Coronas, J., Kolb, G., Zapf, R., Hessel, V., and Santamar´ıa, J. (2007) Catal. Today, 125, 2–10. Rebrov, E.V., Seijger, G.B.F., Calis, H.P.A., de Croon, M.H.J.M., van den Bleek, C.M., and Schouten, J.C. (2001) Appl. Catal. Gen., 206, 125–143. Sahner, K., Hagen, G., Sch¨onauer, D., Reiß, S., and Moos, R. (2008) Solid State Ionics, 179, 2416–2423. Zhou, J., Li, P., Zhang, S., Long, Y.C., Zhou, F., Huang, Y.P., Yang, P.Y., and Bao, M.H. (2003) Sens. Actuators, B Chem., 94, 337–342. Yan, Y.G. and Bein, T. (1992) Chem. Mater., 4, 975–977. Yan, Y.G. and Bein, T. (1992) J. Phys. Chem., 96, 9387–9393. Mintova, S., Mo, S.Y., and Bein, T. (2001) Chem. Mater., 13, 901–905. Vilaseca, M., Yague, C., Coronas, J., and Santamaria, J. (2006) Sens. Actuators, B Chem., 117, 143–150. Kurzweil, P., Maunz, W., and Plog, C. (1994) 5th International Meeting on Chemical Sensors, Rome. Hagen, G., Dubbe, A., Rettig, F., Jerger, A., Birkhofer, T., Muller, R., Plog, C., and Moos, R. (2006) Sens. Actuators, B Chem., 119, 441–448. Fukui, K. and Nishida, S. (1997) Sens. Actuators, B Chem., 45, 101–106. Vilaseca, M., Coronas, J., Cirera, A., Cornet, A., Morante, J.R., and Santamaria, J. (2002) 5th International Conference on Catalysis in Membrane Reactors (ICCMR-2002), Dalian, Peoples R China. Vilaseca, M., Coronas, J., Cirera, A., Cornet, A., Morante, J.R., and

407

408

14 Special Applications of Zeolites

60.

61.

62. 63. 64. 65.

66. 67.

68.

69.

70. 71. 72.

73.

74. 75. 76.

77. 78.

79.

Santamaria, J. (2008) Sens. Actuators, B Chem., 133, 435–441. Wakayama, T., Kobayashi, T., Iwata, N., Tanifuji, N., Matsuda, Y., and Yamada, S. (2006) Sens. Actuators, A Phys., 126, 159–164. Vilaseca, M., Coronas, J., Cirera, A., Cornet, A., Morante, J.R., and Santamaria, J. (2007) Sens. Actuators, B Chem., 124, 99–110. Moore, T.T. and Koros, K.J. (2005) J. Mol. Struc., 739, 87–98. Moore, T.T. and Koros, W.J. (2008) Ind. Eng. Chem. Res., 47, 591–598. Funk, C.V. and Lloyd, D.R. (2008) J. Membr. Sci., 313, 224–231. Clarizia, G., Algieri, C., Regina, A., and Drioli, E. (2008) Microporous Mesoporous Mater., 115, 67–74. Wang, H., Holmberg, B.A., and Yan, Y. (2002) J. Mater. Chem., 12, 3640–3643. Widjojo, N., Chung, T.-S., and Kulprathipanja, S. (2008) J. Membr. Sci., 325, 326–335. Gorgojo, P., Uriel, S., T´ellez, C., and Coronas, J. (2008) Microporous Mesoporous Mater., 115, 85–92. Zhang, Y., Musselman, I.H., Balkus, K.J.Jr., and Ferraris, J.P. (2008) J. Membr. Sci., 325, 28–39. Husain, S. and Koros, W.J. (2007) J. Membr. Sci., 288, 195–207. Jha, P. and Way, J.D. (2008) J. Membr. Sci., 324, 151–161. Perez, E.V., Balkus, K.J.Jr, Ferraris, J.P., and Musselman, I.H. (2009) J. Membr. Sci., 328, 165–173. Pechar, T.W., Tsapatsis, M., Marand, E., and Davis, R. (2002) Desalination, 146, 3–9. Robeson, L.M. (1991) J. Membr. Sci., 62, 165–185. Robeson, L.M. (2008) J. Membr. Sci., 320, 390–400. Cong, H., Radosz, M., Towler, B.F., and Shen, Y. (2007) Sep. Purif. Methods, 55, 281–291. Moore, T.T. and Koros, K.J. (2005) J. Mol. Struct., 739, 87–98. Hillock, A.M.W., Miller, S.J., and Koros, W.J. (2008) J. Membr. Sci., 314, 193–199. Sheffel, J.A. and Tsapatsis, M. (2009) J. Membr. Sci., 326, 595–607.

80. Mahajan, R., Burns, R., Schaeffer, M.,

81.

82. 83.

84.

85.

86.

87. 88.

89. 90.

91.

92. 93.

94.

95.

96.

and Koros, W.J. (2002) J. Appl. Polym. Sci., 86, 881–890. Pechar, T.W., Kim, S., Vaughan, B., Marand, E., Tsapatsis, M., Jeong, H.K., and Cornelius, C.J. (2006) J. Membr. Sci., 277, 195–202. Guseva, O. and Gusev, A.A. (2008) J. Membr. Sci., 325, 125–129. Zhang, L., Gilbert, K.E., Baldwin, R.M., and Way, J.D. (2004) Chem. Eng. Commun., 191, 665–681. Schulz-Ekloff, G., Wohrle, D., van Duffel, B., and Schoonheydt, R.A. (2002) Microporous Mesoporous Mater., 51, 91–138. Bennur, T.H., Srinivas, D., and Ratnasamy, P. (2001) Microporous Mesoporous Mater., 48, 111–118. Calzaferri, G. and Lutkouskaya, K. (2008) Photochem. Photobiol. Sci., 7, 879–910. Levy, D. (1997) Chem. Mater., 9, 2666–2670. Pagliaro, M., Ciriminna, R., and Palmisano, G. (2007) Chem. Soc. Rev., 36, 932–940. Avnir, D., Levy, D., and Reisfeld, R. (1984) J. Phys. Chem., 88, 5956–5959. Fireman-Shoresh, S., Avnir, D., and Marx, S. (2003) Chem. Mater., 15, 3607–3613. Beck, J.S., Artuli, J.C.V., Roth, W.J., Leonowicz, M.E., Kresge, C.T., Schmitt, K.D., Chu, C.T.W., Olson, D.H., Sheppard, E.W., McCullen, S.B., Higgins, J.B., and Schlenker, J.L. (1992) J. Am. Chem. Soc., 114, 10834–10843. Salavati-Niasari, M. (2008) Polyhedron, 27, 3207–3214. Briot, E., Bedioui, F., and Balkus, K.J. (1998) J. Electroanal. Chem., 454, 83–89. Li, G., Chen, L., Bao, J., Li, T., and Mei, F. (2008) Appl. Catal. A: Gen., 346, 134–139. Hoppe, R., Schulz-ekloff, G., Wohrle, D., Kirschhock, C., Fuess, H., and Uyytterhoeven, L. (1995) Adv. Mater., 7, 61–64. Li, X.T., Shao, C.L., Qiu, S.L., Xiao, F.S., Zheng, W.T., Ying, P.L., and

References

97.

98.

99.

100.

101.

102. 103.

104.

105. 106.

107.

108.

109.

Terasaki, O. (2000) Microporous Mesoporous Mater., 40, 263–269. Pellejero, I., Urbiztondo, M., Izquierdo, D., Irusta, S., Salinas, I., and Pina, M.P. (2006) Conference on Advanced Membrane Technol III Membrane Engineering for Process Intensiﬁcation, Cetraro. Coker, E.N., Roelofsen, D.P., Barrer, R.M., Jansen, J.C., and van Bekkum, H. (1998) Microporous Mesoporous Mater., 22, 261–268. Hoppe, R., Schulz-Ekloff, G., Woehrle, D., Kirschhock, C., and Fuess, H. (1994) Langmuir, 10, 1517–1523. Comes, M., Marcos, M.D., Martinez-Manez, R., Millan, M.C., Ros-Lis, J.V., Sancenon, F., Soto, J., and Villaescusa, L.A. (2006) Chem.-Eur. J., 12, 2162–2170. Vietze, U., Krauss, O., Laeri, F., Ihlein, G., Schuth, F., Limburg, B., and Abraham, M. (1998) Phys. Rev. Lett., 81, 4628–4631. Wang, S. (2009) Microporous Mesoporous Mater., 117, 1–9. Takai, K., Ohtsuka, T., Senda, Y., Nakao, M., Yamamoto, K., Matsuoka, J., and Hirai, Y. (2002) Microbiol. Immunol., 46, 75–81. Rivera, A., Far´ıas, T., Ruiz-Salvador, A.R., and M´enorval, L.Cd. (2003) Microporous Mesoporous Mater., 61, 249–259. Davis, M.E. (2002) Nature, 417, 813–829. Pavelic, K. and Hadzija, M. (2003) Medical Applications of Zeolites in Handbook of Zeolite Science and Technology (eds S.M. Auerbach, K.M. Carrado, and P.K. Dutta), Marcel Dekker, New York, pp. 1141–1173. Rimoli, M.G., Rabaioli, M.R., Melisi, D., Curcio, A., Mondello, S., Mirabelli, R., and Abignente, E. (2007) J. Biomed. Mater. Res. A., 87A, 156–164. Wong, L.W. (2006) Molecular delivery system based on the nanoporous zeolite microstructures, Doctoral thesis, Hong Kong University of Science and Technology. Linares, C.F. and Brikgi, M. (2006) Microporous Mesoporous Mater., 96, 141–148.

110. Linares, C.F., Solano, S., and Infante,

111.

112.

113. 114.

115.

116.

117.

118.

119.

120.

121. 122.

123. 124. 125.

126.

G. (2004) Microporous Mesoporous Mater., 74, 105–110. Uglea, C.V., Albu, I., Vatajanu, A., Croitoru, M., Antoniu, S., Panaitescu, L., and Ottenbrite, R.M. (1994) J. Biomater. Sci. Polym. Ed., 6, 633–637. Arruebo, M., Fern´andez-Pacheco, R., Irusta, S., Arbiol, J., Ibarra, M.R., and Santamar´ıa, J. (2006) Nanotechnology, 17, 4057–4064. Miller, M.R. and Megson, I.L. (2007) Br. J. Pharmacol., 151, 305–321. Wheatley, P.S., Butler, A.R., Crane, M.S., Fox, S., Xiao, B., Rossi, A.G., Megson, I.L., and Morris, R.E. (2006) J. Am. Chem. Soc., 128, 502–509. Santos, J.C., Cruz, P., Regala, T., Magalhaes, F.D., and Mendes, A. (2007) Ind. Eng. Chem. Res., 46, 591–599. Shin, H.-S., Kim, D.-H., Koo, K.-K., and Lee, T.-S. (2000) Adsorption, 6, 233–240. Juniot, A., Seltzer, S., Louvier, N., Milesi-Defrance, N., and Cros-Terraux, N. (1999) Ann. Fr. Anesth. R´eanim., 18, 319–331. Cleveland, Z.I. and Meersmann, T. (2007) Magn. Reson. Chem., 45, S12–S23. Chaudhary, R. (2008) Next Generation Larthanide–based Contrast Agents for Applications in MRI, Multimodel Imaging, and Anti-cancer Therapies, MSc, University of Toronto. Chen, W., Wang, S., Westcott, S.L., and Zhang, J. (2005) J. Appl. Phys., 97, 083506-1–8. Gillett, S.L. (1996) Nanotechnology, 7, 177–182. Huang, H., Zhou, J., Chen, S., and Huang, Y. (2004) Sens. Actuators B, Chem., 101, 316–321. Lin, L. and Guthrie, J.T. (2006) J. Membr. Sci., 278, 173–180. Tavolaro, P., Tavolaro, A., and Martino, G. (2009) Colloid Surf. B, 70, 98–107. Devlin, J.J., Kircher, S., Kozen, B.G., Littlejohn, L.F., and Johnson, A.S. (2009) J. Emerg. Med., 1–9. doi: 10.1016/j.jemermed.2009.02.017. Shariatmadari, F. (2008) World. Poultry Sci. J., 64, 76–84.

409

410

14 Special Applications of Zeolites 127. Kabak, B., Dobson, A.D.W., and

128.

129.

130.

131.

132. 133.

134. 135. 136.

137. 138.

139.

140. 141.

142.

143.

Var, I. (2006) Crit. Rev. Food Sci., 46, 593–619. Thieu, N.Q. (2008) Mycotoxins in Vietnamese Pig Feeds, Doctoral thesis, Swedish University of Agricultural Sciences. Pallesen, A., Pallesen, F., Jorgensen, R.J., and Thilsing, T. (2008) Vet. J., 175, 234–239. Ndegwa, P.M., Hristov, A.N., Arogo, J., and Shefﬁeld, R.E. (2008) Biosyst. Eng., 100, 453–469. Takagi, Y., Komatsu, T., and Kitabata, Y. (2008) Microporous Mesoporous Mater., 109, 567–576. Treacy, M.M.J. and Newsam, J.M. (1988) Nature, 332, 249–251. Burton, A.W., Elomari, S., Chan, I., Pradhan, A., and Kibby, C. (2005) J. Phys. Chem. B, 109, 20266–20275. Dartt, C.B. and Davis, M.E. (1994) Catal. Today, 19, 151–186. Lobo, R.F. and Davis, M.E. (1992) Chem. Mater., 4, 756–768. Anderson, M.W., Terasaki, O., Ohsuna, T., Philippou, A., Mackay, S.P., Ferreira, A., Rocha, J., and Lidin, S. (1994) Nature, 367, 347–351. Kesanli, B. and Lin, W.B. (2003) Coord. Chem. Rev., 246, 305–326. Feast, S., Raﬁq, M., Siddiqui, H., Wells, R.P.K., Willock, D.J., King, F., Rochester, C.H., Bethell, D., Page, P.C.B., and Hutchings, G.J. (1997) J. Catal., 167, 533–542. Wells, R.P.K., Tynjala, P., Bailie, J.E., Willock, D.J., Watson, G.W., King, F., Rochester, C.H., Bethell, D., Page, P.C.B., and Hutchings, G.J. (1999) Appl. Catal. Gen., 182, 75–84. Nozue, Y., Kodaira, T., and Goto, T. (1992) Phys. Rev. Lett., 68, 3789–3792. Srdanov, V.I., Stucky, G.D., Lippmaa, E., and Engelhardt, G. (1998) Phys. Rev. Lett., 80, 2449–2452. Igarashi, M., Kodaira, T., Shimizu, T., Goto, A., and Hashi, K. (2006) J. Phys. Chem. Solids, 67, 1063–1066. Bourlinos, A.B., Zboril, R., and Petridis, D. (2003) Microporous Mesoporous Mater., 58, 155–162.

144. Ozeki, S., Uchiyama, H., and Kaneko,

145. 146. 147. 148.

149. 150. 151.

152.

153.

154.

155.

156.

157.

158.

K. (1992) J. Colloid Interface Sci., 154, 303–304. Reich, S. and Cabasso, I. (1989) Nature, 338, 330–332. Marb´an, G. and Vald´es-Sol´ıs, T. (2007) Int. J. Hydrogen Energy, 32, 1625–1637. McDowall, W. and Eames, M. (2007) Int. J. Hydrogen Energy, 32, 4611–4626. Regli, L., Zecchina, A., Vitillo, J.G., Cocina, D., Spoto, G., Lamberti, C., Lillerud, K.P., Olsbye, U., and Bordiga, S. (2005) Phys. Chem. Chem. Phys., 7, 3197–3203. Kayiran, S.B. and Darkrim, F.L. (2002) Surf. Interface Anal., 34, 100–104. Fraenkel, D. (1981) J. Chem. Soc., Faraday Trans. 1, 77, 2029–2039. Langmi, H.W., Book, D., Walton, A., Johnson, S.R., Al-Mamouri, M.M., Speight, J.D., Edwards, P.P., Harris, I.R., and Anderson, P.A. (2005) J. Alloys Compd., 404, 637–642. Langmi, H.W., Walton, A., Al-Mamouri, M.M., Johnson, S.R., Book, D., Speight, J.D., Edwards, P.P., Gameson, I., Anderson, P.A., and Harris, I.R. (2003) J. Alloys Compd., 356, 710–715. Graham, D.D., Culnane, L.F., Sulic, M., Jensen, C.M., and Robertson, I.M. (2007) J. Alloys Compd., 446, 255–259. Forster, P.M., Eckert, J., Heikem, B.D., Parise, J.B., Yoon, J.W., Jhung, S.H., Chang, J.S., and Cheetham, A.K. (2006) J. Am. Chem. Soc., 128, 16846–16850. Jorda-Beneyto, M., Suarez-Garcia, F., Lozano-Castello, D., Cazorla-Amoros, D., and Linares-Solano, A. (2007) Carbon, 45, 293–303. Yu, M., Wyss, J.C., Noble, R.D., and Falconer, J.L. (2008) Microporous Mesoporous Mater., 111, 24–31. Dieng, A.O. and Wang, R.Z. (2001) Renewable Sustainable Energy Rev., 5, 313–342. Sumathy, K., Yeung, K.H., and Yong, L. (2003) Prog. Energy Combust., 29, 301–327.

411

15 Organization of Zeolite Microcrystals Kyung Byung Yoon

15.1 Introduction

The ability to rationally assemble complex structures from modular components of various sizes is an essential part of life and key to the success for the materials chemistry in the new millennium [1, 2]. During the last century, chemists and materials scientists have acquired enough ability to organize atoms and small molecules, which are equal to or smaller than 1 nm, into complex structures. In the twenty-ﬁrst century, chemists have begun including nano building blocks (1–10 nm) into their pools of building blocks. However, micrometer-sized building blocks have not been considered as a novel class of building blocks despite the fact that the ability to organize them will make chemistry and materials science ﬂourish even further. Using zeolite microcrystals as model microbuilding blocks, methods of organizing microcrystals into two- (2D) and three-dimensionally (3D) organized structures have been developed [3–23]. The use of zeolite microcrystals as microbuilding blocks has many additional advantages since each zeolite microcrystal has millions of regularly spaced nanochannels and billions of nanopores, which can be ﬁlled with a variety of functional molecules or nanoparticles. Accordingly, the 2D and 3D organized arrays or supracrystals of zeolite microcrystals with controlled orientations of the component crystals can be applied in various ﬁelds such as molecular separation, materials science, nonlinear optics, electrochemistry, energy and light harvesting. This chapter thus covers the new area of ‘‘organization of zeolite microcrystals’’ and application of the organized entities.

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

In chemistry and nature, atoms, molecules, enzymes, proteins, DNAs, RNAs, and nanoparticles usually self-assemble into larger organized entities. Like these Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

412

15 Organization of Zeolite Microcrystals

relatively much smaller building blocks, it has been shown that microcrystals can also be readily organized into organized entities such as monolayers, multilayers, and other functional materials. This section will show various examples of organization of zeolite microcrystals into various organized entities by self-assembly. 15.2.1 Monolayer Assembly on Solid Substrates

Monolayer assembly of zeolite microcrystals on substrates has been pursued by several researchers as a means to modify electrode surfaces [24–26] and to grow continuous zeolite ﬁlms by secondary growth [27, 28]. However, the coverage, close packing, and degree of uniform orientation of the attached crystals were not satisfactory due to the use of non-uniform crystals and the crystals that do not have wide and ﬂat facets. Indeed, coverage, close packing, and degree of uniform orientation increase dramatically when the zeolite crystals are uniform in size and when they have wide and ﬂat facets. Furthermore, it has been shown that the microcrystals can be attached to various substrate surfaces through various types of linkages. During the monolayer assembly, the type of the linkage and the method of attachment sensitively affect the rate of the attachment, coverage of the crystals on substrates, the degree of close packing (DCP) between the crystals, and the binding strength of the crystals onto the substrates. There are also many interesting features associated with the monolayer assembly of microcrystals on various substrates [3–23, 29, 30]. 15.2.1.1 Types of Linkages Several types of linkages have been developed and each type has its own characteristic features. Covalent Linkages Various covalent linkages have been produced by reacting alkyl amine and epoxide (Figure 15.1a–c) [3], alkyl amine and C60 (Figure 15.1d) [4], alkyl amine and aldehyde (Figure 15.1e) [5], alkyl amine and alkyl halide (Figure 15.1f ) [6], alklyl amine and alkyl isocyanate (Figure 15.1g) [7], zeolite surface hydroxide and alkyl halide (Figure 15.1h) [6], and zeolite surface hydroxide and alkyl isocyanate (Figure 15.1i) [7]. As for the alkyl amines, polyethylene imine (PEI, Figure 15.1b)) and dendritic polyamine (Figure 15.1c) have also been used [8]. These polymeric amines give rise to a stronger binding between the zeolite microcrystals and the substrates as will be described in detail in the later part of this chapter. Among various covalent linkages [4–9] the ether linkage formation through the reaction between the zeolite surface hydroxyl groups and the halopropyl groups tethered to the substrate surface (Figure 15.1h) [6] has been most widely used due to its simplicity. The above functional groups have mostly been attached to the solid surfaces by silylation [3–6, 8–20] although urethanation [7] can also be used instead. The characterization of covalent linkages between zeolite microcrystals and substrates is difﬁcult. However, X-ray reﬂectivity (a technique to analyze the thickness, roughness, and the density of thin ﬁlms supported on substrates) has

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

Zeolite

Zeolite

O NH2 NH2

NH2

NH2 NH2 O

O

Zeolite

Zeolite

NH2

NH2

O

NH2 Zeolite

413

H2N NH2 H2N NH2 H2N NH2 H 2N NH2 NH2 H 2N H2N NH2

CHO Zeolite

O NH2

NH2

NH2

CHO NH2

X

Glass

Glass

Glass

Glass

Glass

Glass

(a)

(b)

(c)

(d)

(e)

(f) Zeolite

Zeolite N

Zeolite

Zeolite

NCO NH2

Zeolite

OH

OH

X

NCO

Zeolite

O

N

H H N

H N

N N

Glass

Glass

Glass

Glass

Glass

Glass

(g)

(h)

(i)

(j)

(k)

(l)

NH2 = CH2CH2CH2NH2 O

= CH2CH2CH2OCH2(CHCH2O)

NCO = CH2CH2CH2NCO Figure 15.1

X = CH2CH2CH2X (X = Cl, Br, I) = CH2CH2CH2N(CH3)3+ = CH2CH2CH2CO2−

NH2 NH2 NH 2 NH2 NH2 NH2

N

= PEI = PDDA+ = PSS−

Various types of molecular linkage.

shown that it can analyze the differences of the linkages in terms of number density, thickness, and roughness of each linker [29]. Ionic Linkages Sodium butyrate (Na+ Bu− ) and trimethylpropyl ammonium iodide (TMPA+ I− ) groups tethered on the solid surfaces readily form ionic linkages between zeolite microcrystals and solid substrates (Figure 15.1j) [10]. The former is produced by converting the substrate-bound 3-cyanopropyl groups to butyric

O

414

15 Organization of Zeolite Microcrystals

CH2 CH

CH2

n

CH2

N+

n

Cl−

SO3−Na+ Na+PSS− (a) Figure 15.2

PDDA+Cl− (b)

Structures of Na+ PSS− and PDDA+ Cl− .

acid groups followed by conversion of the butyric acid to sodium butyrate. The latter is produced by permethylation of zeolite-bound aminopropyl groups to quaternary ammonium salts by treating them with methyl iodide. Upon mixing the Na+ Bu− -tethering substrates and TMPA+ I− -tethering zeolite microcrystals in ethanol, the zeolite microcrystals form monolayers on substrates. Polyelectrolytes such as poly(sodium 4-styrenesulfonate) (Na+ PSS− , Figure 15.2a) and poly(diallyldimethylammonium chloride) (PDDA+ Cl− , Figure 15.2b) can also be used as the intermediate layer between the zeolite microcrystals and substrates. Alternating multilayers of PSS− and PDDA+ can also be applied between the charged solid surfaces (Figure 15.1k). Since aminopropyl groups become positively charged in acidic media due to protonation, monolayers of zeolite microcrystals on substrates can also be formed through aminoproply/PSS− /aminopropyl linkages in acidic media. However, this method produces monolayers of zeolite microcrystals with unsatisfactory coverage and DCP [10]. Hydrogen Bonding Although H-bonding is much weaker than covalent and ionic bonding, zeolite microcrystals can still be attached to substrates through H-bonding. Thymine-tethering zeolite crystals (2.5 µm) readily form closely packed monolayers on adenine-tethering glass plates in an aqueous solution at room temperature (Figure 15.1l) [11]. Most interestingly, at the annealing temperature (55 ◦ C), at which bond breaking and bond reforming between the surface-tethered complementary DNA bases become very rapid, the assembly rate and DCP increased signiﬁcantly, due to faster surface migration of the crystals. Physical Adsorption Monolayers of zeolite microcrystals can also be formed on substrates by physical adsorption of the microparticles onto substrates. The methods include slow evaporation of the solution [31], dip-coating [28, 32], and convective assembly (vertical deposition) [33]. Polystyrene can also be applied for stronger adhesion [27]. Although physical adsorption is not a choice of method for monolayer assembly of zeolite microcrystals on substrates, it has been regarded as an important process during growth of zeolite ﬁlms on substrates [34].

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

(a)

RO

(b)

RS

(c)

SO

(d)

SWS

Figure 15.3 Four different types of reaction set-up that have been studied: reﬂux only (RO), reﬂux and stirring (RS), sonication only (SO), and sonication with stacking between the glass plates (SWS) [30].

15.2.1.2 Types of Substrates The substrate that has been most widely tested is glass [3, 4, 7–11, 13, 16–21, 23]. In addition to glass, glass ﬁber [8], silica [6], alumina [6], large zeolite [6], vegetable ﬁbers (cotton, linen, and hemp) [14], artiﬁcial ﬁbers (nylon and polyester), conducting substrates (Pt, Au, and ITO glass) [9], plastic (polycarbonate) [35], mesoporous silica supported on porous alumina [36, 37], and Si wafer [38] have been tested. In fact, any solid materials regardless of the shape and size can be used as substrates. 15.2.1.3 Methods The four methods that have been extensively tested are reﬂux only (RO), reﬂux and stirring (RS), sonication only (SO), and sonication with stacking (SWS) between the glass plates (Figure 15.3) [3–23, 29, 30]. For RO and RS methods, Teﬂon supports having four underlying supporting legs have been used (Figure 15.3a and b). The substrates coated with functional groups are normally placed vertically, one by one, into each groove engraved on the support. The Teﬂon support is subsequently immersed into the cylindrical reactor charged with a solution (usually toluene) dispersed with zeolite microcrystals. The solution is reﬂuxed without (RO) and with stirring (RS) with the help of a magnetic stirring bar placed under the four-legged-Teﬂon support. For SO and SWS methods, comb-shaped Teﬂon supports have been used (Figure 15.3c and d). The comb-shaped Teﬂon plates have been used as such with no underlying supporting legs so that the thin substrates inserted into the gaps can have direct contacts with the bottom of the round-bottomed ﬂasks to allow transmission of ultrasound with the highest possible power to the glass plates. For the SO method, substrate plates are inserted, piece by piece, into each gap of the comb-shaped support (Figure 15.3c). For the SWS method, a surface functionalized glass substrate (FG) is interposed between two clean bare glass (BG) plates, and three sets of BG/FG/BG stack are inserted, stack by stack, into each gap of the comb-shaped Teﬂon support (Figure 15.3d). Sonication of the reactors is usually carried out using an ultrasonic bath equipped with two ultrasound generators (28 kHz, 95 W each) attached to the bath from the bottom. Sonication always induces the temperature increase in the water inside the sonic bath and the increase

415

416

15 Organization of Zeolite Microcrystals

10 s

20 s

30 s

40 s

50 s

60 s

Figure 15.4 Digital camera images of the FG plates removed from the corresponding BG/FG/BG stacks that were kept in the toluene solution of silicalite-1 microcrystals ([C]F = 0.90 pM) for various periods (as

indicated) under the SWS condition showing that the attachment of the silicalite-1 microcrystals starts from the bottom and then propagate to the top with time [30]. (FG = 3-chloropropyl-coated glass)

in the temperature gives rise to the increase in the attachment rate in the case of the covalent linkage formation. Therefore, the temperature of the water inside the sonic bath should be controlled to be the same using a temperature-controlled immersion cooler, to eliminate the effect of temperature on the observed results. Interestingly, during the SWS method, the zeolite microcrystals ﬂow from the bottom to the top between the gap of FG and BG plates as shown in Figure 15.4. Furthermore, SWS method has many advantages over other methods in terms of coverage and DCP, as will be discussed in detail in the later section of this chapter. Regardless of the methods, there are always physically or very weakly adhered microcrystals on the chemically attached monolayers. Therefore, to remove the weakly adhered microcrystals from the chemically attached monolayers, the freshly prepared zeolite monolayer-attached substrates are usually mildly sonicated for 30 seconds in pure toluene using a sonic bath equipped with weaker ultrasound generators before analyzing the monolayers with scanning electron microscopy (SEM) and before conducting the measurements of the attached amounts on a microbalance. 15.2.1.4 Characteristic Points to Monitor the Quality of the Monolayers Coverage and Coverage-t Plot Coverage (in percentage) is a quantitative value, which is deﬁned as Eq. 15.1,

Coverage = (WA /WA,max ) × 100

(15.1)

where WA and WA,max represent the attached amount and maximum attached amount, respectively, on a substrate (on both sides). The plot of coverage with respect to time (coverage-time) for each monolayer gives more detailed information on the rate and mechanism of the monolayer assembly process.

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

Degree of Close Packing (DCP) DCP represents the degree of microparticles placed adjacent to each other. At the moment, there are no direct methods to endow quantitative values to DCP. Comparison of the SEM images gives an idea about the relative DCP. %R-t Plots At the present stage, there are no methods to measure the absolute binding strengths between the attached individual microcrystals and the substrates. Therefore, the relative binding strengths are estimated indirectly by comparing the percentages of the residual amounts (%R) of the attached zeolite microcrystals with respect to the initially attached amount after conducting sonication-induced detachment of the microcrystals from the substrates for a certain period of time in a fresh solvent [8]. The plot of the %R values with respect to time (%R-t plot) gives more information about the distribution of binding strength between the attached microcrystals and the substrates. The comparison of the %R-t plots of the monolayers allows the comparison of the relative binding strengths of the monolayers and the distribution of the binding strengths between the attached microcrystals and the substrates. The section ‘Surface Migration’ in Section 15.2.1.5 shows typical examples. For this, the sonication-induced detachment of the monolayers should be conducted under the same condition, such as the same ultrasonic bath, the same amount of water, the same position of the monolayer within the ultrasonic bath, and the same temperature. 15.2.1.5 Four Key Processes Occurring during Monolayer Assembly During the monolayer assembly, four key processes occur as illustrated in Figure 15.5 [30]. They are dispersion of microcrystals into the solution (dispersion), attachment of microcrystals onto 3-chloropropyl-tethering glass (CP-G) (surface attachment), migration of the randomly attached crystals leading to lateral close packing between the microcrystals (surface migration), and the replacement of the surface-bound microcrystals with those dispersed in solution (replacement). Dispersion The dispersion of zeolite microcrystals is important because the zeolite microcrystals sediment quickly [30]. If they once sediment, then they lose the chance to react with the substrates. To maintain dispersion of the microcrystals, the solution should be agitated by stirring, shaking, or ultrasonic irradiation. The dispersed microcrystals automatically have translational energies, with which the Replacement

Dispersion

Surface migration Linkage destructive Linkage non destructive

Surface attachment Figure 15.5 The four major processes that occur during the monolayer assembly of zeolite microcrystals in solution [30].

417

418

15 Organization of Zeolite Microcrystals

microcrystals can collide onto the substrates to undergo attachment through bond formation between the functional groups on the surfaces of zeolite microcrystals and substrates. The collision between the particles dispersed in the solution and those attached to the substrates help surface-bound microcrystals undergo surface migration, and replace the surface-bound microcrystals with those in the solution. As described in the later part of this chapter (Section 15.2.1.6), the type of agitation, in other words, the strength of agitation plays a critical role for the monolayer assembly process. The concentration of the zeolite microcrystals simply based on its added amount is deﬁned as the formal concentration, [C]F . Without dispersion of the microcrystals, unlike homogeneous solutions, [C]F cannot be assumed to be equal to the concentration of the zeolite microcrystals dispersed in the solution. Therefore, the concentration of the actually dispersed microcrystals [C]D should be deﬁned. The dispersed concentration [C]D is related to [C]F by multiplying a proportionality factor, fmethod , which is a function of the method, namely RO, RS, SO, and SWS (Eq. (15.2)). [C]D,method = fmethod [C]F

(15.2)

The estimated fRO value was 0.25 < fRO ≤ 0.5. The estimated fRS , fSO , and fSWS values were 1 [30]. Surface Attachment For the attachment of the microcrystals onto the substrate surface, the functional groups of the zeolite microcrystals and the substrates should ﬁrst encounter and undergo linkage formation reactions. For covalent linkages to occur, the two functional groups should collide at certain speeds higher than the threshold speeds and at certain angles. Thermal energies also facilitate the linkage formation [3–9, 17, 21]. Since the formation of ionic linkages is omni-directional (formed regardless of the direction of approach of the functional groups) and the oppositely charged centers attract each other regardless of the distance between the charged centers, the formation of ionic linkages is much easier than the formation of covalent linkages [10]. Accordingly, the number of ionic linkages is much larger than that of covalent linkages, giving rise to stronger binding strength between the crystals and substrates. Surface Migration For the surface migration to occur, the attached zeolite microcrystals should break the molecular linkages (denoted as linkage breakage) that were formed between a small number of the functional groups of the zeolite and the substrate, and subsequently form new molecular linkages (denoted as new linkage formation) between the unused, fresh functional groups tethered to the surfaces of zeolite microcrystals and those existing on the new area of the substrate [4, 11, 18]. Thus, for the surface migration to occur, the attached microcrystals should be very strongly agitated. This is the reason why the surface migration occurs very efﬁciently under the SO and SWS conditions. The reason for the formation of small numbers of linkages between the surfaces of the microcrystals and substrates is because the surface roughness values are larger than the lengths of the molecules tethered to the surfaces of both the

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

Zeolite

A

A

A

A B B B

A B B

Glass

Figure 15.6 Schematic illustration depicting the limited number of linkage between the zeolite-bound functional groups (a) and the glass-bound functional groups (b) due to the unevenness of both surfaces [8].

microcrystals and substrates as illustrated in Figure 15.6 [8, 29]. In the case of covalent linkages, after extensive surface migration, the number of surface-tethered functional groups, and eventually the number of linkages between the microcrystals and substrates, decreases, leading to the decrease in the binding strength (or %R) as will be discussed in detail in the next section [18]. The surface migration is classiﬁed into two types: surface destructive and surface nondestructive migrations [39, 40]. The former process occurs when the linkage is covalent while the latter process occurs when the linkage is ionic or hydrogen bonding. Consequently, the nature of linkage and the method sensitively affect the proﬁles of the coverage-time and %R-t plots. For instance, in the case of covalent linkage, coverage and the binding strength gradually decrease with time under the condition of SO and SWS (see below) in which the surface migration takes place very rapidly. On the contrary, in the case of hydrogen bonding and ionic bonding, the number of linkages and hence the binding strength do not change with time even under the SO and SWS conditions [40]. The above dramatic difference is well demonstrated by the coverage-t and %R-t plots of the monolayers of silicalite crystals (size: 1.3 × 0.6 × 1.8 µm3 ) prepared on glass plates with three different types of linkages as schematically illustrated in Figure 15.7. The linkages, are (Figure 15.7a) the covalent linkage achieved by reaction between bare silicalite crystals (Z) and CP-G plates, (Figure 15.7b) the direct ionic linkage achieved by reaction between (Na+ Bu− )-tethering silicalite (Z – ) crystals and (TMPA+ I− )-tethering glass (G+ ) plates (size: 1.8 × 1.8 cm2 ), and (Figure 15.7c) (Na+ PSS− ) and (PDDA+ Cl− )-mediated ionic bonding between Z− and G+ , respectively. The monolayers are denoted by (a), (b), and (c), respectively, in the ﬁgure. The method was called SWS. Figure 15.8a and b show the proﬁles of the coverage with respect to the reaction time for the periods of 0–8 minutes and 0–5 hours, respectively. Consistent with the fact that the binding strength increases in the order A < B < C (see below) [10, 40], the reason why the proﬁles are different in Figure 15.8a is because their surface migration rates are different and the rate of surface migration decreases as binding strength increases. This example clearly demonstrates that the rate of monolayer assembly increases as the rate of surface migration increases and as the binding strength decreases.

419

420

15 Organization of Zeolite Microcrystals

Na+ −

Surface migration

Na+ −

Na+ −

Zeolite

Na+ −

OH

OH

Z/CP-G

Na+ −

−

−

−

+

+

+

+

Zeolite

OH

Zeolite

(a)

Na+ −

−

+ −

+ −

+

−

+ −

−

−

−

−

−

−

+

+

+

+

+

+

(b)

Z−/G+

(c)

Z−/PDDA+/PSS−/G+

Figure 15.7 Schematic illustration of two different types of reaction, bond breakage (a) and charge shift ((b) and (c)) that occur during the surface migration of zeolite microcrystals on glass [40].

During the reaction period of 5 hours, the coverage of curve A gradually decreases from 100 to 64% (Figure 15.8b). In contrast, those of curves B and C decrease only to 96 and 97%, respectively. This shows that the initial coverages of curves B and C essentially remain unaltered with reaction time. In other words, the number of ionic centers on Z− , G+ , and PDDA+ /PSS− /G+ remains intact even after undergoing surface migration for extended periods of time. Consistent with the above inference, while the %R-t plot of curve A gradually decreases from ∼70 to ∼30% with time, those of curves B and C reach ∼94% during the initial period and then remain constant at the initial value during the tested 5 hours (Figure 15.8c). This result shows that the initial binding strengths of curve B and C (the initial numbers of ionic linkages) remain constant despite the fact the crystals underwent surface migration (and the replacement of the attached crystals with those in solution as will be described in the next section) for extended periods of time. Conversely, the above results verify that the covalently attached microcrystals indeed migrate on the substrate surface by repeatedly undergoing the cycle of linkage breakage and new linkage formation. Thus, the ionically attached microcrystals migrate on the surface by ‘‘charge shift’’ (Figure 15.7b and c). Therefore, the surface migration of the crystals can also be classiﬁed into ‘‘linkage-destructive’’ and ‘‘linkage-nondestructive’’ depending on the type of linkage – covalent or ionic.

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

100 A

B

Coverage (%)

80

C

60 40 20 0 0

1

2

(a)

3 4 5 6 Reaction time (min)

7

8

C

100

B Coverage (%)

80 60

A

40 20 0 0

1

(b)

2 3 Reaction time (h)

100

4

5

B C

80

%R

60 40 A

20 0 0 (c)

1

2 3 Reaction time (h)

Figure 15.8 The coverage-t plots of the monolayer assembly of the zeolite microcrystals on glass plates with three different types of linkages shown in Figure 15.7 (as labeled) during the period of 0–8 minutes

4

5

(a) and 0–5 hours (b), respectively, and the %R-t plot of the monolayers (c) assembled on glass plates with the three different types of linkage as labeled [40].

421

422

15 Organization of Zeolite Microcrystals

Replacement Replacement of the surface-attached zeolite microcrystals with those dispersed in solution is readily observed during continuation of the monolayer assembly procedure with the substrates already fully covered with zeolite microcrystals with the same zeolites but incorporating ﬂuorescent molecules [18]. Thus, the glass plates fully covered with bare zeolite microcrystals (BZ-G) were ﬁrst prepared through the covalent linkage formation between the glass-bound 3-choloropropyl groups and the surface hydroxyl groups of zeolite microparticles, and they were subsequently allowed to undergo further monolayer assembly in the solution dispersed with ﬂuorescent zeolite microcrystals (FL Z) under the conditions of RO, RS, SO, and SWS for a desired period of time (t). The resulting glass plates are denoted as (FL Z/BZ-G)method -t. Note that the ﬂuorescent zeolite is denoted as FL Z to differentiate it from the functional group-coated zeolite (FZ) (see Section 15.2.1.3). Figure 15.9 shows the luminescence images of (FL Z/BZ-G)RO -6 hours, FL ( Z/BZ-G)RO -24 hours, (FL Z/BZ-G)RS -6 hours, (FL Z/BZ-G)RS -24 hours, (FL Z/ BZ-G)SO -30 seconds, (FL Z/BZ-G)SO -300 seconds, (FL Z/BZ-G)SWS -30 seconds, and 0h

6h

24 h

(a)

(b)

(c)

(d

(e)

(f)

RO

RS

0s

30 s

300 s

(g)

(h)

(i)

(j)

(k)

(l)

SO

SWS

Figure 15.9 The luminescence images of BZ/G (a), (FL Z/BZ-G)RO -6h (b), (FL Z/BZ-G)RO -24h (c), BZ/G (d), (FL Z/BZ-G)RS -6h (e), (FL Z/BZ-G)RS -24h (f), BZ/G (g), (FL Z/BZ-G)SO -30s (h), (FL Z/BZ-G)SO -300s (i), BZ/G (j), (FL Z/BZ-G)SWS -30s (k), and (FL Z/BZ-G)SWS -300s (l), respectively [30].

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

1.0

B 7%

Normalized I F

R 0.8

R

A 4%

0.6 550

600

650

0.4

B 33%

0.2

A 10%

0.0 550 (a)

600

1.0 R Normalized I F

650

700

550

600

(b)

Wavelength (nm)

650

700

Wavelength (nm)

B 106% R

0.8 B 76%

0.6 0.4 0.2

24%

20%

A

A

0.0 550 (c)

600 650 Wavelength (nm)

700

550 (d)

600 650 Wavelength (nm)

700

Figure 15.10 Luminescence spectra. (a) (FL Z/BZ-G)RO -6h (A), (FL Z/BZ-G)RO -24h (B), and FL Z-G (R). (b) (FL Z/BZ-G)RS -6h (A), (FL Z/BZ-G)RS -24h (B), and FL Z-G (R). (c) (FL Z/BZ-G)SO -30s (A), (FL Z/BZ-G)SO -300s (B), and FL Z-G (R). (d) (FL Z/BZ-G)SWS -30s (A), (FL Z/BZ-G)SWS -300s (B), and FL Z-G (R) [30].

(FL Z/BZ-G)SWS -300 seconds, respectively. Figure 15.10 shows the ﬂuorescence spectra of the resulting glass plates and the reference glass plate, which is fully covered with FL Z (FL Z-G). Based on the relative ﬂuorescence intensities of the FL Z-treated BZ-G plates with respect to the reference FL Z-G, the degrees of replacement were measured, which were 4% (RO, 6 hours), 7% (RO, 24 hours), 10% (RS, 6 hours), 33% (RS, 24 hours), 20% (SWS, 0.5 minutes), 76% (SWS, 5 minutes), 24% (SO, 0.5 minutes), and 106% (SO, 5 minutes), respectively. Thus, the replacement readily occurs regardless of the methods and the degree of replacement increases in the order RO < RS <<< SWS < SO. Interestingly, it was also revealed that the replacement rate is signiﬁcantly higher than the detachment rate under the SO condition, which suggests that the rate of collision-induced detachment of the surface-bound microcrystals with those in solution is signiﬁcantly higher than that of the self-detachment [18]. Thus, replacement, and hence the collision of surface-bound microcrystals with those in solution, are also one of the key processes that occur during the monolayer assembly of microcrystals on substrates and it signiﬁcantly affects the processes of the monolayer assembly and the qualities of the produced monolayers.

423

15 Organization of Zeolite Microcrystals

15.2.1.6 Effect of Method on Rate, DCP, Coverage, and Binding Strength The method of assembly very sensitively affects the above factors [18]. For instance, as the coverage-t plots (Figure 15.11) during the initial 10 minutes period show, the rate of monolayer assembly of zeolite microcrystals on glass through covalent linkage formation between the glass-bound 3-choloropropyl and the surface hydroxyl groups of zeolite microparticles increases in the order RO < RS << SO < SWS. The coverage and DCP also increase in the same order as the SEM images (insets) show. Thus, the method very sensitively affects the rate, coverage, and DCP. As noted from Figure 15.11, the formal concentration [C]F also sensitively affects the monolayer assembly rate in the cases of RO and SWS. In the case of RO, the strong dependence of the rate on [C]F arises due to the low concentration of the dispersed microparticles [C]D due to low fRO (see Section 15.2.1.5 ‘Dispersion’). In the case of SWS, since [C]F = [C]D , the observed dependence of the rate on [C]F indicates the rate of intercalation of the microparticles into the gap between BG and FZ also depend on [C]D . The sonication induced dramatic increases in the rate, coverage, and DCP in the cases of SO and SWS occurs due to a large increase in the rate of migration of the substrate-bound crystals on the surface (surface migration) caused by sonication-induced powerful agitation of the attached crystals by ultrasound, giving

Coverage (%)

100

RS

RO

80 60 40 20 0 0

(a)

2

4

6

8

10

(b) 0

2

4

6

8

10

100 Coverage (%)

424

80 60 40 20 SO

0 (c)

0

10 µm

SWS 2

4

6

8

10

Reaction time (min)

Figure 15.11 Coverage-t proﬁles for the attachment of zeolite (silicalite-1) microcrystals onto CP-G during the period of 10 minutes obtained under the conditions of RO (a), RS (b), SO (c), and SWS (d), respectively, for [C]F = 0.45 (empty circle), 0.90 (ﬁlled square), and 1.80 pM (empty

(d) 0

2

4

6

8

10

Reaction time (min)

triangle), respectively. The inset shown in each panel shows the SEM image of each silicalite-1-coated plate obtained after 2 minutes under each condition with [C]F = 0.90 pM. The scale bar in each SEM image represents 10 µm [30].

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

80

80

60

60

40

40

20

20 0

0 0

3

6

100

80

80

60

60

40

40

20

20

0

9 12 15 18 21 24

(b)

0 0

3

6

100

80

80

60

60

40

40

20

20

0 (c)

%R

SWS

100

Coverage (%)

%R

SO

0 0

1

2

3

9 12 15 18 21 24

4

5

6

Reaction time (h) Figure 15.12 Coverage-t (open circle) plots for the monolayer assembly of silicalite-1 crystals on 3-chloropropyl coated glass plate and the corresponding %R–t (ﬁlled circles) plots for the silicalite-1-coated glass plates

100

100

80

80

60

60

40

40

20

20

0 (d)

0 0

1

2 3 4 5 Reaction time (h)

produced under the conditions of RO (a), RS (b), SO (c), and SWS (d), respectively, for the period shown in the x axis with [C]F = 0.90 pM [30].

rise to large increases in the rates of bond breaking and bond reforming between the microcrystals and substrates. Thus, strong agitation of the surface-bound microcrystals is crucial to speed up the monolayer assembly process and to increase the coverage and DCP. Moreover, in the case of SWS, the fast inﬂux (inﬂow) of zeolite microparticles from the bottom to the interior of the glass stacks, which subsequently leads to the fast ﬂow of the particles from the bottom to the top within the glass stack, contributes signiﬁcantly to the high coverage, rate, and DCP. The uniform upward movement of particles within the glass stacks can be named as a vectorial surface migration, giving rise to very tight lateral close packing between the crystals [18, 40]. The coverage-t and %R-t plots (Figure 15.12) of the same monolayer assembly taken during the longer periods of time (24 hours for RO and RS and 6 hours for SO and SWS) clearly show the method-dependent variation of the coverage and %R, which also reﬂects the binding strength. Under the conditions of RO and RS, during 24 hours, the coverage progressively increases to 100% and %R progressively increases to 43%, respectively, indicating that both the number of attached crystals and the number of the molecular linkages between each crystal and glass plate progressively increase during the period (Figure 15.12a and b). Under the SO and SWS conditions, the coverage and %R

6

Coverage (%)

(a)

100

Coverage (%)

100

%R

RS

100

Coverage (%)

%R

RO

425

15 Organization of Zeolite Microcrystals

rapidly increase to 100 and ∼60% during the initial assembly period of 5 minutes (Figure 15.12c and d) and maintain the similar coverage and %R during the next 1 hour. However, they decrease gradually to ∼80 and 20–30%, respectively, during the course of 6 hours, indicating the progressive decrease in the number of molecular linkages between the crystal and the substrate as a result of the extensive undergoing of the linkage-destructive surface migration and replacement of the substrate-bound crystals with those dispersed in solution. This phenomenon is the same with the one described in Section 15.2.1.5 ‘Surface Migration’ (Figure 15.8b and c) As noted, the %R-increasing period is much longer under the conditions of RO and RS conditions than under the SWS and SO conditions. This suggests that the repetition rate of the cycle of the bond breaking and bond reforming is much faster under the sonication conditions than under the RO and RS conditions. The overall order of the repetition rates is RO ≤ RS < SWS < SO. 15.2.1.7 Factors Affecting Binding Strengths The two most important factors that affect binding strengths are the number of the linkage and the strength of each linkage. In general, the number of linkage between each zeolite crystal and the substrate is higher with ionic than with covalent linkage due to the fact that the formation of ionic bonding is omni-directional, works well regardless of the distance between the positive and negative centers, and does not require kinetic energies for the bonding formation. Regardless of the type of linkage, the use of polymeric linkages gives rise to large increases in the number of linkage between the crystal and the substrate because the ﬂexible polymeric linkages readily adjust themselves into the valleys of the rough surfaces of both zeolite microcrystals and substrates as illustrated in Figure 15.13. For example, in the case of the monolayer assembly of 3-aminopropyl-coated zeolite (AP-Z) microcrystals to the terminal epoxide-coated glass (EP-G) substrates, the %R 100 EP/PEI/EP

X X A A

X A

A X

A X

60

EP

X A

PEI

80

A X A X

AP

EP/DPA/EP

%R

426

X A

40

X A

EP

X A

A

A A X

A X

A X

DPA AP

20 AP/EP 0

X

0

10

20 30 40 Sonication time (min)

50

60

Figure 15.13 %R–t plots of the monolayers of zeolite-A crystals assembled on glass plates through AP/EP (bottom), EP/dendritic polyamine (DPA)/EP (middle), and EP/PEI/EP (top) linkages [8].

A

X X A A

X X A A

EP AP

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

decreases to 5% when they are directly connected, decreases to 45% when dendritic polyamine (see Figure 15.1c) is applied between the terminal epoxide-coated zeolite (EP-Z) and EP-G, and decreases to 75% when PEI (see Figure 15.1b) is applied between the EP-Z and EP-G. Although small, the slower rate of monolayer assembly of Z− onto G+ in the presence of PDDA+ /PSS− in Figure 15.8a under the SWS condition also demonstrates the same phenomenon that polymer induces large increases in the number of linkage between the crystal and the substrate. As described in the previous Section 15.2.1.6, the method of monolayer assembly also sensitively affects the binding strength. For example, the binding strength between the crystals and substrates increases in the order RO ≈ RS << SO < SWS during the period of 1 hour (Figure 15.12). As noted, in the cases of the monolayers that were assembled on glass using SO and SWS methods, the binding strength decreases as a result of repeatedly undergoing the linkage-destructive surface migration and replacement of the substrate-bound crystals with those dispersed in solution. Interestingly, the binding strength between the crystals and the substrates increases dramatically upon cross-linking the adjacent crystals while leaving the number of the linkage between each crystal and the substrate constant (Figure 15.14) [17]. Thus, upon cross-linking the adjacent AP-tethering microcrystals (Figure 15.14a) with terephthaldicarboxyaldehyde (TPDA) (Figure 15.14b), %R increases signiﬁcantly (Figure 15.14c). Regardless of the type of linkage and method, the zeolite microcrystals become almost ﬁrmly attached to the substrates when the monolayers assembled on glass and other metal oxides substrates are calcined at high temperatures under the ﬂow of air. 15.2.1.8 Driving Forces for Uniform Orientation and Close Packing Silicalite-1 or MFI type crystals can be attached to substrates in two different orientations, a and b orientations, that is, a and b axes pointing normal to the substrate, respectively. b orientation is most dominant (> 99%), and a orientation is found only during the initial stages. The reasons for the degrees of uniform orientation being high are as follows. First, the attached crystals are continuously replaced by those dispersed in solution. Second, the binding strength between each microcrystal and substrate increases with increasing the contact area. Consequently, the probability of a substrate-bound microcrystal to be replaced by those in solution decreases when they are attached with the largest-area face. Thus, the largest-area face determines the ﬁnal orientation of the attached crystals. This phenomenon seems to be responsible also for formation of b-oriented continuous ﬁlms on substrates during silicalite-1 ﬁlm growth [34]. Zeolite-L can be synthesized in two different morphologies, cylindrical and hexagonal column [41]. The cylindrical zeolite-L crystals can be assembled into vertically oriented monolayers (Figure 15.15a), while hexagonal columnar zeolite-L crystals can be assembled into horizontally oriented monolayers (Figure 15.15b) [19]. The former occurs because the crystals have ﬂat faces only at the bases, and the latter occurs because the areas of the side planes are signiﬁcantly larger than those of basal planes. Assembly method also affects the degree of uniform orientation.

427

15 Organization of Zeolite Microcrystals

Cross-linked

Noncross-linked

Glass =

(a)

NH2 (AP)

= OHC

(b)

CHO (TPDA)

100 Cross-linked 80

60 %R

428

40 Noncross-linked

20

0 0 (c)

5

10

15

20

25

30

Sonication time (min)

Figure 15.14 Illustrations of noncross-linking (a) and cross-linking (b) between the adjacent microcrystals, and the %R-t plots of the monolayers assembled on glass plates with two different bonding modes, noncross-linking (bottom) and cross-linking (top) [17].

For instance, the degree of vertical orientation of cylindrical zeolite-L crystals with the aspect ratio (the ratio of the length with respect to the diameter of the bottom face) >2 increases in the order RS < SO and SWS. When the aspect ratio is less than 1, SWS also readily produces monolayers of c-oriented crystals [42]. In relation to the above, electric ﬁeld–driven alignment of long ZSM-5 crystals was demonstrated [43]. The physical method is limited to those crystals having net sizable intrinsic dipole moments. Growth of vertically oriented aluminophosphate molecular sieves using anodized alumina discs was also demonstrated [44]. However, the degrees of coverage, close packing, and uniform orientation cannot compete with those shown in Figure 15.15.

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

10 µm (a)

20 µm (b) Figure 15.15 Monolayers of closely packed and vertically aligned cylindrical zeolite-L crystals (a) and horizontally aligned hexagonal columnar zeolite-L crystals (b) assembled on a 3-chloropropyltethering glass plate. Insets show cross sections; scale bar 2 µm [19].

15.2.2 Patterned Monolayer Assembly on Substrates

Micropatterned monolayers of zeolite microcrystals have the potential to be applied as combinatorial catalysts and low-dielectric packing materials for integrated circuits [45, 46]. The ﬁrst step is preparation of micropatterned monolayers of functional groups on substrates using microcontact printing [21] or photopatterning [13, 20]. A typical SEM image of a micropatterned monolayer of ZSM-5 crystals using microcontact printing is shown in Figure 15.16a. The related micropatterned microporous and mesoporous silica ﬁlms were prepared by several research groups [47–49]. Photochemical degradation of organic linker groups tethered to glass surfaces is a highly versatile and effective way for preparing glass plates patterned with organic linker groups [13]. Thus, when a 3-halopropyl-tethering glass (XP-G) plate was

429

430

15 Organization of Zeolite Microcrystals

(a)

(b)

(c)

Figure 15.16

30 µm

6 µm

100 µm

10 µm

50 µm

SEM images of micropatterned monolayers of zeolite microcrystals [13, 21].

exposed to UV light under a photomask, the UV-exposed halopropyl groups were selectively degraded. When the patterned XP-G plates are used as the substrates, 3-halopropyl-tethering zeolite (XP-Z) crystals formed monolayers only on the UV-exposed areas (Figure 15.16b), while the bare crystals formed monolayers only on the unexposed areas (Figure 15.16c). Patterned continuous silicalite-1 ﬁlms are obtained by immersing photo-patterned glass plates into the corresponding synthesis gel [13]. A weaker attractive force between the colloidal seed crystals and gold was also utilized to produce patterned continuous silicalite-1 ﬁlms on silicon wafers [38]. Photoinduced

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

decarboxylation of glass-bound silver butyrate (C3 H7 -CO2 − Ag+ ) is an efﬁcient way for micropatterned monolayer assembly of zeolite microcrystals on substrates through ionic linkages [20]. 15.2.3 Multilayer Assembly on Substrates

After assembling monolayers, the same method can be repeated for multilayer assembly. For this, strong linkages are necessary because the multilayers tend to be detached from the substrates as the number of the layers increases. Accordingly, ionic linkage combined with several layers of polyelectrolytes can be used for multilayer assembly to strengthen the binding strength [10]. Indeed, the assembly of pentalayers of zeolite microcrystals on glass was demonstrated using PSS− /PDDA+ /PSS− as the repeating linkage units (Figure 15.17). Since the zeolite microcrystals are not very uniform in size and shape, the surface of the monolayer becomes rough. The surface roughness also increases as the number of the layer increases. Accordingly, as the number of layer increases, the degree of uniform orientation and DCP of the top layer decreases due to the increase in the roughness of the under layer. 15.2.4 Organization into 2D Arrays on Water

In addition to the mono and multilayer assembly of zeolite microcrystals in speciﬁc crystal orientations on solid substrates through well-deﬁned chemical linkages [4–23, 30, 39, 40], a method to organize microbuilding blocks in the form of monolayers on water through interdigitation of the surface-tethered hydrophobic hydrocarbon (HC) chains was explored. In fact, such a strategy has been adopted by several groups [50–55] as a means for organizing nanocrystals. Whitesides and his coworkers have extensively studied self-assembly of meso-sized (10 µm to 5 mm) hydrophobic building blocks in water [1]. The HC-coated cubic zeolite microcrystals readily form closely packed monolayers at the air–water interface through interdigitation of surface-tethered HCs, and on glass plates after transferring onto glass plates by dip coating [56]. Interestingly, while the mode of networking was face-to-face (FTF) contact when the water contact angles of HCs were lower 77◦ (Figure 15.18a), it changed to edge-to-edge (ETE) contact mode when the water contact angles of HCs were higher than 102◦ (Figure 15.18b). When the water contact angles of HCs were intermediate (82◦ ), both modes appeared in the monolayers, with about equal populations (Figure 15.18c). The ETE mode of close packing occurs as a means to decrease the repulsion between the hydrophobic chains on the zeolite surface and the underlying water by decreasing the surface area of a zeolite microcrystal that are contacting with water. The resulting monolayers of cubic zeolite microcrystals with their threefold axes oriented perpendicular to substrates would be useful for application of the zeolite monolayers for advanced materials.

431

432

15 Organization of Zeolite Microcrystals

I−

I−

Zeolite

Zeolite

Zeolite 2 µm

50 µm

(a)

Zeolite

1 µm Zeolite (b) CH2 CH

CH2

n

= PSS−

CH

= SO3−Na+

PDDA+

Figure 15.17 SEM images of a pentalayer of silicalite crystals assembled on a glass plate through ionic linkages with PSS− /PDDA+ /PSS− as the intermediate linkers: top view (a) and cross-section (b) [10].

N+ Cl−

n

I−

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

5 µm

(a)

(b)

(c)

25 µm

25 µm

Figure 15.18 (a) SEM image showing the monolayer of cubic zeolite-A microcrystals (1.7 µm) supported on glass with the FTF mode of contacting between the neighboring crystals with HC = n-octyl or n-dodecyl (θ ≤ 77◦ ). (b) SEM image showing a monolayer of cubic zeolite-A microcrystals supported

2.5 µm

2.5 µm

on glass with mostly ETE mode with HC = n-octadecyl or n-heptadecaﬂuorodecyl (θ ≥ 102◦ ). (c) SEM image showing a monolayer of cubic zeolite-A microcrystals supported on glass with both FTF and ETE in about equal populations with HC = methyl n-undecanoate (θ = 82◦ ) [56].

Thus, cubic microcrystals can be organized in the form of closely packed monolayers with the threefold (C3 ) axis of each crystal orienting perpendicular to the plane of the support through the ETE contacting mode. Furthermore, interdigitation between the hydrophobic chains along the six edges is enough to interlink or to organize cubic microcrystals not only on water but also on ﬂat glass surfaces with each crystal standing with a vertex pointing down even in the absence of supporting buoyant force exerted from the underlying water layer. It is also worthwhile to note that while the cubes inside the closely packed monolayers of crystals in the ETE contacting mode have six contacts with the six neighbors, those

433

434

15 Organization of Zeolite Microcrystals

Toluene

−

−

−

− −

− − −

−

−

−

−

−

−

−

−

−

− − − − − −

−

−

−

−

−

−

− −

−

− − − − − −

Water (a)

No SDS

(b)

4 mmol

NaA

(c)

>10 mmol

SDS

Figure 15.19 Illustrations of a water droplet in toluene acting as a template to attract hydrophilic zeolite crystals (a) and the aligned anionic surfactant molecules (SDS) acting as nanotools for the alignment of zeolite-A

crystals at the water-toluene interface at an intermediate concentration of SDS (b), and a deformed structure of a water pool caused by high concentrations of SDS (c) [22].

at the boundary have even three contacts with other crystals. This indicates that the van der Waals force arising from interdigitation between very short hydrophobic chains is much stronger than one would imagine. This phenomenon is likened to the one that geckos with relatively large weights can readily adhere to any smooth surfaces by using a large number (∼500 000) of 30 − 130 µm-long keratinous hairs or setae through van der Waals interaction [57]. 15.2.5 Organization into Surface-Aligned Zeolite Microballs

Using the water droplets dispersed in toluene as templates zeolite nanocrystals (150 nm) can be organized into self-supporting 1–20 µm sized zeolite microballs [22] (Figure 15.19a). The water droplets are prepared by adding small amount of water into the toluene predispersed with zeolite nanocrystals followed by vigorous sonication. Strong bonding develops between the bare zeolite nanocrystals due to formation of direct siloxyl linkages between the nanocrystals during sonication by undergoing dehydration reaction between the surface hydroxyl groups. In pure water, the microballs of randomly oriented nanocrystals are formed (Figure 15.19a). Sodium dodecylsulfate (SDS) induces the zeolite nanocrystals in the two outermost surface layers of microballs uniformly aligned (Figures 15.19b and 15.20a and b). The surface alignment occurs because dodecylsulfate ions self-assemble at the water-toluene interface with the negatively charged polar heads pointing to water droplets (Figure 15.19b). Thus, to minimize the electrostatic repulsion with the negative heads, zeolite nanocrystals uniformly align also to minimize surface area. At higher concentrations of SDS, the zeolite nanocrystals assembled into anisotropic (nonspherical) structures (Figures 15.19c and 15.20c) due to the decrease in the surface energy. Perforated zeolite microballs are produced during the initial periods of sonication (Figure 15.20d). The shaded spots in Figure 15.20a represent the internal voids covered with thin layers (usually

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

3 µm

0.8 µm

10 µm

1 µm

2 µm

(a)

435

(b)

5 µm

1 µm ×20k

(c) Figure 15.20 SEM images of a microball composed of octahedral zeolite-X crystals (size = 150 nm) and its uniformly aligned surface (a), an egg-shaped ball composed of cubic zeolite-A crystals (size = 150 nm) and

10 µm

2 µm

(d) its uniformly aligned surface (b), a highly deformed ball with aligned surface (c), and a perforated ball showing the 3D network of voids (d) [22].

monolayers) of zeolite. The 3D networks of the perforated spherulites can be seen from Figure 15.20d (right). The highly perforated microballs of zeolite nanocrystals can be utilized as highly effective catalysts and adsorbents. The microballs also show the anatomies of emulsions. 15.2.6 Self-Assembly of Substrate-Tethering Zeolite Crystals with Proteins

d-glucose-tethered zeolite microcrystals and α-glucosidase (Figure 15.21a) and d-biotin-tethered zeolite microcrystals and avidin (Figure 15.21b) readily self-assemble into thin (2–20 µm) and long (>1 cm) ﬁbrils (Figure 15.21c) upon stirring them in aqueous buffer solutions where the protein activities are highest [12, 15]. The morphology of the ﬁbrils is sensitively governed by the protein/zeolite ratio. In the case of d-biotin-tethering zeolite microcrystals and avidin, discrete

1 µm 2 µm

436

15 Organization of Zeolite Microcrystals

OH OH O O

OH OH

O OH OH

O

OH OH

O OH OH

O

O C

O C

O C

O

O

O

O HN

OH OH

O NH

HN

S

S O C O

(CH2)11 (CH2)11 (CH2)11 O Si O

(a)

O

Si O

Zeolite

O

Si O O

NH

O C CH3

O

(CH2)11 (CH2)10 (CH2)11 O

(b)

Si O O

Si

O

O

Si O O

Zeolite

100 µm (c) 2 µm

× 7.0 k

5 µm

(d)

Figure 15.21 SEM images showing the ﬁbrous zeolite-A-α-D-glucosidase composite material (a) and a discrete cluster of biotin-tethering zeolite-A crystals (b) [12, 15].

15.2 Organization of Zeolite Microcrystals into Functional Materials by Self-Assembly

clusters of zeolite crystals with sizes of 5–10 µm (Figure 15.21d) are produced when the protein to zeolite ratio is 0.1. This is an unprecedented phenomenon that a protein and its substrate-tethering inorganic microcrystals self-assemble into structured aggregates. Although the mechanism is unclear, in particular as to why the two components grow into axial symmetry, this method is a way to organize zeolite microcrystals with protein. 15.2.7 In Situ Self-Organization of Zeolite Crystals into Arrays during Synthesis

The methods to assemble monolayers of zeolite microcrystals on various substrates have been well explored. However, as already described in Section 15.2.1.8, the orientations of the crystals on the substrates are determined by the phenomenon that the largest-area faces attach to the substrate. Accordingly, vertical orientations of other axes on the substrates have not been possible. On the contrary, it was revealed that zeolite crystals can be self-assembled into aligned monolayers during their synthesis. Furthermore, it was discovered that all their three axes can be uniformly aligned with their orientations parallel to the surface normal. Such an in situ self-organization method was derived from the facile transformation of polyurethane sponges into self-supporting silicalite-1 foams [58, 59]. Based on this ﬁnding, uniformly aligned polyurethane ﬁlms were used as the templates for orientation-controlled synthesis of 2D and 3D arrays of silicalite-1 crystals, where the crystal orientation is controlled by varying the nature of polyurethane [23]. For instance, the uniformly aligned polyurethane ﬁlms are produced by conducting the cycle of the alternative introduction of PDI (Figure 15.22a) and BDO (Figure 15.22b) or PDI and TBE (Figure 15.22c) onto glass substrates for 500 times. The produced uniformly aligned polyurethane ﬁlms are denoted as (PDI/BDO)500 /G and [(PDI/TBE)500 /G], respectively. When the hydrothermal reaction of silicalite-1 is carried out in the presence of (PDI/BDO)500 /G, closely packed 2D arrays of c-oriented silicalite-1 crystals are produced on the glass substrates (Figure 15.22d). There are some areas covered with second layers of c-oriented crystals (Figure 15.22d), indicating that production of even 3D arrays of uniformly aligned silicalite-1 crystals is also possible by optimization of the condition. When [(PDI/TBE)500 /G] are used, closely packed 2D arrays of a-oriented silicalite-1 crystals are produced on the glass substrates (Figure 15.22e). Supramolecular organization of the hydrolyzed organic products and the reactive silicon species seems to be responsible for the above phenomenon. The organic species in synthesis gels (such as TPA+ ) have been known as structure directors for creation of nanopores in certain shapes, sizes, and networks in zeolites. The in situ self-organization phenomenon indicates that the organic species can also serve as ‘‘orientation directors’’ for the produced crystals. In relation to this, the growth of randomly scattered zincophosphate and AlPO4 -5 crystals in certain orientations on alkyl phosphate-coated gold has been carried out [60].

437

438

15 Organization of Zeolite Microcrystals

O

NCO HO

OH

NCO PDI

BDO

(a)

(b)

O

C

O(CH2)2OH

C

O(CH2)2OH TBE

(c)

500 nm

(d)

300 nm

(e)

WD17.4 mm

200 nm

(f)

Figure 15.22 Monolayer (a) and double layer (b) of closely packed, c-oriented silicalite-1 crystals grown on (PDI/BDO)500 /G, and a monolayer of closely packed a-oriented silicalite-1 crystals grown on (PDI/TBE)500 /G (c) [23].

15.3 Monolayer Assembly of Zeolite Microcrystals by Dry Manual Assembly

The previous methods for monolayer assembly of microcrystals have depended on self-assembly of the crystals during their assembling processes in solution (denoted as ‘‘wet self-assembly’’). In fact, wet self-assembly has also been the method of choice for the monolayer assembly of nanometer-scale building blocks [61–65], while the manual attachment of dry building blocks with hands (denoted as ‘‘dry manual assembly’’) on adhesive-coated substrates has been the method for the monolayer assembly of millimeter-to-centimeter building blocks on substrates (ﬂoors and walls). Therefore, the method for monolayer attachment of building blocks on substrates has to switch from the wet self-assembly to dry manual assembly at some stage as the size of the building block increases. It was elucidated that, for the monolayer assembly, the upper size limit by wet self-assembly is ∼3 µm, the lower size limit for the dry manual assembly is ∼0.5 µm, and dry manual assembly is superior to self-assembly in the overlapping region (∼0.5–3 µm) with

15.3 Monolayer Assembly of Zeolite Microcrystals by Dry Manual Assembly

Figure 15.23 Photographic images showing the process of monolayer assembly of silicalite-1 microcrystals on glass through ionic bonding by forced manual assembly (rubbing) [19].

respect to rate, DCP, uniform orientation of the assembled microcrystals, substrate area, and ecological considerations. For the monolayer assembly of microcrystals on substrates by dry manual assembly, ionic bonding and hydrogen bonding were most effective. In the case of hydrogen bonding, the combination of zeolite-bound surface hydroxyl groups/PEI/the hydroxyl groups on the surface of BG substrates works very effectively. Chitosan can also be used as the intermediate layer instead of PEI [35]. Hydrogen bonding between the surface hydroxyl groups was also used instead of polymer linkers [38]. Practically, zeolite powders are simply rubbed for 10–20 seconds onto substrates (18 × 18 mm2 ) with a ﬁnger (Figure 15.23). To avoid contamination of the glass and microcrystals with moisture and salt from the ﬁnger, a soft latex glove can be used. Instead of ﬁngers, the soft plates made with PDMS work very well for large-area substrates. The coverage and DCP of the monolayers of silicalite-1 microcrystals are similar to those prepared by SWS with the ionic bonding between TMPA+ and Bu− as the ionic linkage (Figure 15.24). However, dry manual assembly is incomparably simpler than sonication and does not require solvents, reactors, or other equipment (such as a sonication bath), and that the required period for monolayer assembly was shorter: 10–20 seconds in the case of rubbing versus 180 seconds in the case of sonication. Furthermore, there were no physically adsorbed second-layer crystals on the monolayers when prepared by rubbing, because the weakly adhered microcrystals are removed from the chemically attached crystals during rubbing. Interestingly, in the monolayers of silicalite-1 crystals prepared by SWS, there are many crystals that carry a 90 ◦ -intergrown parasitic crystal [66, 67] on the (010) plane of the mother crystal (see the inset and the circled crystals in Figure 15.24b), whereas the monolayers prepared by rubbing do not have such crystals as all the parasitic components were dislodged from the mother crystals during monolayer assembly (see the inset and the circled crystals in Figure 15.3a). This phenomenon further shows that the binding strength between the parasitic and mother crystals at the interface is weaker than the binding strength between the microcrystal and substrate. Although the silicalite-1 crystal morphology can be controlled by structure -directing agents [68], it is usually difﬁcult to prepare batches that do not have the

439

440

15 Organization of Zeolite Microcrystals

10 µm (a)

10 µm (b)

(c)

(d)

Figure 15.24 SEM images of the monolayers of TMPA+ I− -coated silicalite-1 microcrystals (1.3 × 0.5 × 1.7 mm3 ) assembled on Na+ Bu− -coated glass plates by rubbing (a) and the SWS (b) method. (c) Enlarged

structures of mother and parasitic crystals that were drawn based on the inset of (a) and the twinned crystals in (b), and (d) the perspective view of a parasitic crystal [19].

crystals carrying 90◦ -intergrown crystals. Accordingly, the preparation of monolayers of silicalite-1 crystals free from the crystals carrying 90◦ -intergrown parasitic crystals, and of the corresponding continuous silicalite-1 ﬁlms with no a-oriented spots, has been difﬁcult. Now, the dry manual assembly enables the preparation of silicalite-1 crystal monolayers free from parasitic crystals. The dry manual assembly works very well for the monolayer assembly of microcrystals with the sizes between 0.5 and 12 µm. It is expected that this manual method will work equally well for larger microcrystals. The dry manual assembly is also much more effective for the monolayer assembly of microcrystals on large-area substrates at high speed. For instance, a 150 × 150 mm2 glass plate (Figure 15.25) can be covered with a high-quality monolayer of silicalite-1 crystals with an average size around 2 µm within 1 minute. We believe that the rate can be increased further by optimizing the procedure. This feature will enable mass production of microcrystal-coated large substrates, which will provide supported zeolite microcrystal monolayers for various applications.

15.4 Current and Future Applications

(a)

(b)

Figure 15.25 Photographic images of large glass plates (150 × 150 mm2 ) before (a) and after (b) monolayer assembly of silicalite-1 microcrystals (1.3 × 0.5 × 1.7 mm3 ) using PEI as the interfacial hydrogen bonding mediator [19].

15.4 Current and Future Applications

The uniformly aligned zeolite monolayers can be immediately used as excellent precursors for molecular sieve membranes [37]. They can also be used to characterize paramagnetic species in zeolites [16]. The natural and synthetic fabrics and papers coated with Ag+ -exchanged zeolite crystals can be used as novel antibacterial functional fabrics and papers (Figure 15.26) [14]. The optical ﬁbers coated with zeolite microcrystals can be used as high-efﬁciency photocatalysts [69]. The monolayers of vertically oriented ﬂuorophore-incorporating cylindrical zeolite-L crystals gave anisotropic photoluminescence (dichroic ratio = 8.9) [19], which is higher than the previously reported value (4.5) from the ﬂuorescent polymer-incorporating mesoporous silica [70]. Thus, the uniformly aligned mono and multilayers of zeolite crystals can also be used as the media for generation of anisotropic photoluminescence [19], supramolecularly organized light-harvesting systems [42, 71], and nonlinear optical ﬁlms [72]. The micropatterned monolayers can be used as high-throughput combinatorial catalysts and low-dielectric packing materials for integrated circuits [45, 46]. The highly perforated microballs of zeolite nanocrystals can be utilized as effective catalysts and adsorbents. Although the assembly of zeolite microcrystals into 3D supercrystals is still a challenge, we believe that this is the direction to which the zeolite microcrystal organization should move, since the resulting supercrystals could ﬁnd many important optical and other applications. The zeolite microcrystal organization is regarded as one of the future directions of zeolite research [73, 74].

441

442

15 Organization of Zeolite Microcrystals

3 µm (a)

1 µm (b) Figure 15.26 SEM images of a nano-zeolite-Y-coated cotton ﬁber at two different magniﬁcations [14].

15.5 Summary and Outlook

Methods to organize zeolite microcrystals into two- (2D) and three-dimensionally (3D) organized structures [3–23] are described. Since each zeolite microcrystal has millions of regularly spaced nanochannels and billions of nanopores, which can be ﬁlled with a variety of functional molecules or nanoparticles, the organized zeolite microcrystals have great potential to be used as advanced materials. The preformed zeolite microcrystals can be organized into monolayers, micropatterned monolayers, and multilayers on solid substrates, monolayers on water, and surface-aligned microballs. The zeolite micorcrystals can also be organized into arrays with controlled orientations of the crystal axes even during the synthesis of the crystals in the gel by self-assembly. The organization of zeolite microcrystals into monolayers on solid substrates has been most extensively studied. Various types of bonding such as covalent,

15.5 Summary and Outlook

ionic, hydrogen bonding, and the physical adsorption have been employed as for the linkage formation between the functional groups of substrates and zeolite microcrystals. The substrate that has been most widely tested is glass [3, 4, 7–11, 13, 16–21, 23]. Glass ﬁber [8], silica [6], alumina [6], large zeolite [6], vegetable ﬁbers (cotton, linen, and hemp) [14], artiﬁcial ﬁbers (nylon and polyester), conducting substrates (Pt, Au, and ITO glass) [9], plastic (polycarbonate) [35], mesoporous silica supported on porous alumina [36, 37], and Si wafer [38] have also been tested. In fact, any solid materials regardless of the shape and size can be used as substrates. The methods are classiﬁed into two, namely, ‘‘wet self-assembly’’ (organization of the crystals through self-assembly in solution) and ‘‘dry manual assembly’’ (rubbing of the dry crystals on substrates). The upper size limit of the microcrystals by wet self-assembly is ∼3 µm while the lower size limit of the crystals by dry manual assembly seems to be ∼0.5 µm. The dry manual assembly should be employed for the crystals larger than 3 µm. In the overlapping region (∼0.5–3 µm), the dry manual assembly is superior to wet self-assembly with respect to rate, DCP, uniform orientation of the assembled microcrystals, substrate area, and ecological considerations. In the case of monolayer assembly of zeolite microcrystals on substrates by wet self-assembly, the four methods that have been extensively studied are RO, RS, SO, and SWS [3–23, 29, 30]. In this case, the vigorous agitation of the microcrystals is highly necessary to promote the assembly rate, binding strength, and the DCP. The agitation strength varies depending on the type of method. Accordingly, the type of method very sensitively affects the rate, binding strength, and the DCP. The SWS method is the best method to rapidly prepare very closely packed monolayers on glass. The four key processes that occur during monolayer assembly are the dispersion of microcrystals into the solution (dispersion), attachment of microcrystals onto CP-G (surface attachment), migration of the randomly attached crystals leading to lateral close packing between the microcrystals (surface migration), and the replacement of the surface-bound microcrystals with those dispersed in solution (replacement). The surface migration and replacement of the attached crystals with those in solution occur ‘‘linkage-destructively’’ when the type of the linkage is covalent, and ‘‘linkage-nondestructively’’ when the type of the linkage is ionic or hydrogen bonding. Therefore, in the case of monolayer assembly of microcrystals on substrates by covalent linkages, the coverage and %R decrease with time. The characteristic points to monitor the quality of the monolayers were coverage, coverage-t plot, DCP, binding strength, and the %R-t plot. The ﬁnal orientations of the attached crystals are largely determined by the area of the face. The zeolite microcrystals could be organized in the form of monolayers on water by interdigitation of the surface-tethered hydrophobic HC chains. When the water contact angles of HCs were <77◦ , the mode of networking was FTF contact, and when the water contact angles of HCs were >102◦ , it changed to ETE contact mode. This result shows that cubic zeolite microcrystals can also be organized in the form of closely packed monolayers with the threefold (C3 ) axis of each crystal orienting perpendicular to the plane of the support through the ETE contacting mode.

443

444

15 Organization of Zeolite Microcrystals

Using the water droplets dispersed in toluene as templates zeolite nanocrystals (150 nm) can be organized into self-supporting 1–20 µm sized zeolite microballs [22]. In pure water, the microballs of randomly oriented nanocrystals are produced. The addition of sodium dodecylsulfate (SDS) into the system induces the zeolite nanocrystals in the two outermost surface layers of microballs uniformly align. The surface alignment of zeolite nanocrystals occurs due to the self-assembly of dodecylsulfate ions at the water–toluene interface with the negatively charged polar heads pointing to the water droplets to the minimize the electrostatic repulsion between the negative heads and the negatively charged zeolite nanocrystals, since the surface alignment leads to the minimization of the surface area of the zeolite balls. The orientation-controlled synthesis of 2D and 3D arrays of silicalite-1 crystals is possible by using the uniformly aligned polyurethane ﬁlms as the templates [23]. By this method, the closely packed 2D silicalite-1 arrays with a and c orientations, respectively, can be produced. So far, only one axis of the zeolite microcrystals can be uniformly aligned during their mono and multilayer assembly on ﬂat substrates. The obtained monolayers have great potential to be applied in various ﬁelds beyond the examples described in the previous section. However, to be able to use the monolayers for high-precision applications, the methods to precisely control all three axes of the microcrystals should be developed. The methods to prepare very thin self-supporting zeolite ﬁlms should also be developed based on the highly aligned monolayers of zeolite microcrystals in order to develop zeolite membranes with very high ﬂux and selectivity. I believe that the methods to organize zeolite microcrystals into various organized entities described in this chapter will greatly help zeolite science and technology expand into new ﬁelds.

Acknowledgments

I thank the graduate students and the postdoctoral researchers whose name appear in the references for their hard work which made this chapter possible. I also thank the Creative Research Initiatives and Acceleration Research programs of the Ministry of Education, Science and Technology, and the Internal Research Fund of Sogang University for supporting this work. I also thank Yong Su Park and Nak Cheon Jeong for the help to prepare this chapter.

References 1. Bowden, N.B., Weck, M., Choi, I.S., and

4. Choi, S.Y., Lee, Y.-J., Park, Y.S., Ha,

Whitesides, G.M. (2001) Acc. Chem. Res., 34, 231–238. 2. Ozin, G.A. (2000) Chem. Commun., 419–432. 3. Kulak, A., Lee, Y.-J., Park, Y.S., and Yoon, K.B. (2000) Angew. Chem. Int. Ed. Engl., 39, 950–953.

K., and Yoon, K.B. (2000) J. Am. Chem. Soc., 122, 5201–5209. 5. Lee, G.S., Lee, Y.-J., Ha, K., and Yoon, K.B. (2000) Tetrahedron, 56, 6965–6968. 6. Ha, K., Lee, Y.-J., Lee, H.J., and Yoon, K.B. (2000) Adv. Mater., 12, 1114–1117.

References 7. Chun, Y.S., Ha, K., Lee, Y.-J., Lee, J.S.,

8.

9.

10.

11.

12.

13.

14. 15.

16.

17.

18. 19.

20.

21.

22.

23.

24. 25.

Kim, H.S., Park, Y.S., and Yoon, K.B. (2002) Chem. Commun., 1846–1847. Kulak, A., Park, Y.S., Lee, Y.-J., Chun, Y.S., Ha, K., and Yoon, K.B. (2000) J. Am. Chem. Soc., 122, 9308–9309. Ha, K., Park, J.S., Oh, K.S., Zhou, Y.S., Chun, Y.S., Lee, Y.-J., and Yoon, K.B. (2004) Microporous Mesoporous Mater., 72, 91–98. Lee, G.S., Lee, Y.-J., and Yoon, K.B. (2001) J. Am. Chem. Soc., 123, 9769–9779. Park, J.S., Lee, G.S., Lee, Y.-J., Park, Y.S., and Yoon, K.B. (2002) J. Am. Chem. Soc., 124, 13366–13367. Lee, G.S., Lee, Y.-J., Choi, S.Y., Park, Y.S., and Yoon, K.B. (2000) J. Am. Chem. Soc., 122, 12151–12157. Ha, K., Lee, Y.-J., Chun, Y.S., Park, Y.S., Lee, G.S., and Yoon, K.B. (2001) Adv. Mater., 13, 594–596. Lee, G.S., Lee, Y.-J., Ha, K., and Yoon, K.B. (2001) Adv. Mater., 13, 1491–1495. Um, S.H., Lee, G.S., Lee, Y.-J., Koo, K.K., Lee, C., and Yoon, K.B. (2002) Langmuir, 18, 4455–4459. So, H., Ha, K., Lee, Y.-J., Yoon, K.B., and Belford, R.L. (2003) J. Phys. Chem. B, 107, 8281–8284. Park, J.S., Lee, Y.-J., and Yoon, K.B. (2004) J. Am. Chem. Soc., 126, 1934–1935. Lee, J.S., Ha, K., Lee, Y.-J., and Yoon, K.B. (2005) Adv. Mater., 17, 837–841. Lee, J.S., Lim, H., Ha, K., Cheong, H., and Yoon, K.B. (2006) Angew. Chem. Int. Ed. Engl., 45, 5288–5292. Park, J.S., Lee, G.S., and Yoon, K.B. (2006) Microporous Mesoporous Mater., 96, 1–8. Ha, K., Lee, Y.-J., Jung, D.Y., Lee, J.H., and Yoon, K.B. (2000) Adv. Mater., 12, 1614–1617. Kulak, A., Lee, Y.J., Park, Y.S., Kim, H.S., Lee, G.S., and Yoon, K.B. (2002) Adv. Mater., 14, 526–529. Lee, J.S., Lee, Y.-J., Tae, E.L., Park, Y.S., and Yoon, K.B. (2003) Science, 301, 818–821. Li, Z., Lai, C., and Mallouk, T.E. (1989) Inorg. Chem., 28, 178–182. Yan, Y. and Bein, T. (1992) J. Phys. Chem., 96, 9387–9393.

26. Li, J.-W., Pfanner, K., and Calzaferri, G.

(1995) J. Phys. Chem., 99, 2119–2126. 27. Mintova, S., Schoeman, B., Valtchev, V.,

28.

29.

30. 31.

32.

33.

34.

35. 36.

37.

38.

39. 40. 41.

42.

Sterte, J., Mo, S., and Bein, T. (1997) Adv. Mater., 9, 585–589. Boudreau, L.C., Kuck, J.A., and Tsapatsis, M. (1999) J. Membr. Sci., 152, 41–59. Lee, H., Park, J.S., Kim, H., Yoon, K.B., Seeck, O.H., Kim, D.H., Seo, S.H., Kang, H.C., and Noh, D.Y. (2006) Langmuir, 22, 2598–2604. Lee, J.S., Ha, K., Lee, Y.-J., and Yoon, K.B. (2009) Top. Catal., 52, 119–139. Lain´e, P., Seifert, R., Giovanoli, R., and Calzaferri, G. (1997) New J. Chem., 21, 453–460. Ban, T., Ohwaki, T., Ohya, Y., and Takahashi, Y. (1999) Angew. Chem. Int. Ed. Engl., 38, 3324–3326. Lee, J.A., Meng, L., Norris, D.J., Scriven, L.E., and Tsapatsis, M. (2006) Langmuir, 22, 5217–5219. Li, S., Li, Z., Bozhilov, K.N., Chen, Z., and Yan, Y. (2004) J. Am. Chem. Soc., 126, 10732–10737. Zhou, M., Liu, X., Zhang, B., and Zhu, H. (2008) Langmuir, 24, 11942–11946. (a) Wang, X., Zhang, B., Liu, X., and Lin, J.Y.S. (2006) Adv. Mater., 18, 3261–3265; (b) Zhang, B., Zhou, M., and Liu, X. (2008) Adv. Mater., 20, 2183–2189; (c) Lang, L., Liu, X., and Zhang, B. (2009) Appl. Surf. Sci., 255, 4886–4890. Lai, Z.P., Bonilla, G., Diaz, I., Nery, J.G., Sujaoti, K., Amat, M.A., Kokkoli, E., Terasaki, O., Thompson, R.W., Tsapatsis, M., and Vlachos, D.G. (2003) Science, 300, 456–460. ¨ urk, S., and Akata, B. (2009) MicroOzt¨ porous Mesoporous Master., 126, 228–233. Yoon, K.B. (2007) Acc. Chem. Res., 40, 29–40. Lee, J.S. and Yoon, K.B. (2010) J. Nanosci. Nanotechnol., 10, 191–194. Lee, Y.-J., Lee, J.S., and Yoon, K.B. (2005) Microporous Mesoporous Mater., 80, 237–246. Ruiz, A.Z., Li, H., and Calzaferri, G. (2006) Angew. Chem. Int. Ed. Engl., 45, 5282–5287.

445

446

15 Organization of Zeolite Microcrystals 43. Caro, J., Finger, G., Kornatowski, J.,

44.

45.

46.

47.

48.

49. 50. 51. 52. 53.

54.

55.

56.

57.

58. 59. 60. 61.

Richter-Mendau, J., Werner, L., and Zibrowius, B. (1992) Adv. Mater., 4, 273–276. Tsai, T.G., Shih, H.C., Liao, S.J., and Chao, K.J. (1998) Microporous Mesoporous Mater., 22, 333–341. Li, S., Demmelmaier, C., Itkis, M., Liu, Z., Haddon, R.C., and Yan, Y. (2003) Chem. Mater., 15, 2687–2689. Wang, Z., Mirta, A., Wang, H., Huang, L., and Yan, Y. (2001) Adv. Mater., 13, 1463–1466. Yang, P., Deng, T., Zhao, D., Feng, P., Pine, D., Chmelka, B.F., Whitesides, G.M., and Stucky, G.D. (1998) Science, 282, 2244–2246. Huang, L., Wang, Z., Sun, J., Miao, L., Li, Q., Yan, Y., and Zhao, D. (2000) J. Am. Chem. Soc., 122, 3530–3531. Yang, H., Coombs, N., and Ozin, G.A. (1997) Adv. Mater., 9, 811–814. Li, M., Schnablegger, H., and Mann, S. (1999) Nature, 402, 393–395. Pileni, M.P. (2001) J. Phys. Chem. B, 105, 3358–3371. Wang, Z.L. (1998) Adv. Mater., 10, 13–30. Wang, Z.L., Harfenist, S.A., Whetten, R.L., Bentley, J., and Evans, N.D. (1998) J. Phys. Chem. B, 102, 3068–3072. Kiely, C.J., Fink, J., Brust, M., Bethell, D., and Schiffrin, D.J. (1998) Nature, 396, 444–446. Viau, G., Brayner, R., Chakroune, N., Lacaze, E., Fi´evet -Vincent, F., and Fi´evet, F. (2003) Chem. Mater., 15, 486–494. Park, J.S., Jeong, N.C., Lee, Y.-J., Kim, M.J., and Yoon, K.B. (2010) J. Nanosci. Nanotechnol., 10, 370–374. Autumn, K., Liang, Y.A., Hsieh, S.T., Zesch, W., Chan, W.P., Kenny, T.W., Fearing, R., and Full, R.J. (2000) Nature, 405, 681–685. Lee, Y.-J., Lee, J.S., Park, Y.S., and Yoon, K.B. (2001) Adv. Mater., 13, 1259–1263. Lee, Y.-J. and Yoon, K.B. (2005) Microporous Mesoporous Mater., 88, 176–186. Feng, S. and Bein, T. (1994) Nature, 368, 834–836. (a) Petty, M.C. (1996) Langmuir–Blodgett Films: An Introduction, Cambridge University Press; (b) Ulman, A. (1996)

62.

63.

64.

65.

66. 67.

68.

69.

70.

71.

72.

Chem. Rev., 96, 1533–1554; (c) Riklin, A. and Willner, I. (1995) Anal. Chem., 67, 4118–4126. (a) Shen, G., Tercero, N., Gaspar, M.A., Varughese, B., Shepard, K., and Levicky, R. (2006) J. Am. Chem. Soc., 128, 8427–8433; (b) Katz, E., Weizmann, Y., and Willner, I. (2005) J. Am. Chem. Soc., 127, 9191–9200; (c) Liu, D., Gugliotti, L.A., Wu, T., Dolska, M., Tkachenko, A.G., Shipton, M.K., Eaton, B.E., and Feldheim, D.L. (2006) Langmuir, 22, 5862–5866. Keegan, N., Wright, N.G., and Lakey, J.H. (2005) Angew. Chem. Int. Ed. Engl., 117, 4879–4882. Haddour, N., Cosnier, S., and Gondran, C. (2005) J. Am. Chem. Soc., 127, 5752–5753. Bigioni, T.P., Lin, X.-M., Nguyen, T.T., Corwin, E.I., Witten, T.A., and Jaeger, H.M. (2006) Nat. Mater., 5, 265–270. Hay, D.G., Jaeger, H., and Wilshier, K.G. (1990) Zeolites, 10, 571–576. (a) Price, G.D., Pluth, J.J., Smith, J.V., Bennett, J.M., and Patton, R.L. (1982) J. Am. Chem. Soc., 104, 5971–5977; (b) Weidenthaler, C., Fischer, R.X., Shannon, R.D., and Medenbach, O. (1994) J. Phys. Chem., 98, 12687–12694; (c) Geier, O., Vasenkov, S., Lehmann, E., KOrger, J., Schemmert, U., Rakoczy, R.A., and Weitkamp, J. (2001) J. Phys. Chem. B, 105, 10217–10222. Bonilla, G., DPaz, I., Tsapatsis, M., Jeong, H.-K., Lee, Y., and Vlachos, D.G. (2004) Chem. Mater., 16, 5697–5705. Pradhan, A.R., Macnaughtan, M.A., and Raftery, D. (2000) J. Am. Chem. Soc., 122, 404–405. Nguyen, T.-Q., Wu, J., Doan, V., Schwartz, B.J., and Tolbert, S.H. (2000) Science, 288, 652–656. Calzaferri, G., Bossart, O., Bruhwiler, D., Huber, S., Leiggener, C., Van Veen, M.K., and Ruiz, A.Z. (2006) C. R. Chim., 9, 214–225. (a) Kim, H.S., Lee, S.M., Ha, K., Jung, C., Lee, Y.-J., Chun, Y.S., Kim, D., Rhee, B.K., and Yoon, K.B. (2004) J. Am. Chem. Soc., 126, 673–682; (b) Kim, H.S., Lee, M.H., Jeong, N.C., Lee, S.M., Rhee, B.K., and Yoon, K.B. (2006) J. Am. Chem. Soc., 128, 15070–15071; (c)

References Kim, H.S., Sohn, K.W., Jeon, Y., Min, H., Kim, D., and Yoon, K.B. (2007) Adv. Mater., 19, 260–263; (d) Kim, H.S., Pham, T.T., and Yoon, K.B. (2008) J. Am. Chem. Soc., 130, 2134–2135.

73. Bein, T. (2005) MRS Bull., 30, 713–720. 74. Sch¨ uth, F. and Schmidt, W. (2002) Adv.

Mater., 14, 629–638.

447

449

16 Industrial Potential of Zeolites Giuseppe Bellussi, Angela Carati, and Roberto Millini

16.1 Introduction

During the last three decades, the introduction of zeolite catalysts in several industrial processes has brought relevant economical and environmental beneﬁts. By the use of zeolites, it has been possible to phase-out several mineral acids or chloro-containing catalysts, improving the processes yield and selectivity, and products quality, and decreasing the energy consumption. More than 90% of the industrial zeolite catalysts are applied in petrochemistry and reﬁning industries that are technologically mature sectors. In these areas, in which in the past there have been already several radical breakthrough achievements, the innovation is expected to proceed through optimization and incremental improvements. Several good recent reviews have highlighted the trend for industrial innovation [1–5]. In the last few years, some trend-breaking events have modiﬁed the situation. The rapid evolution of the communication media has pushed for the globalization of the markets, at the same time several countries have accelerated their growth and the global energy and goods consumption has increased much more rapidly than in the previous decades. Consequently, we have observed in the last ﬁve years a rapid increase of the price of energy sources and raw materials as well as of the effect of the anthropogenic activities on the environment. These events have generated a strong pressure for reducing the environmental impact and accordingly many countries have adopted more severe regulations for the production of waste and the CO2 emission. For instance, one of the consequences in Europe has been a big change in the demand of passenger cars, promoting the diffusion of diesel engines with respect to the gasoline engines, because of the lower CO2 emission of the former. Moreover, the European legislation will mandate the addition of 10% of biofuel on energy basis to the fossil fuels by year 2020. These changes will have a big impact on European reﬁneries operation in the next years and they will generate opportunities and needs for research of new processes and products. The most important issues appear to be the reduction of energy consumption CO2 and waste emission and the production Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

450

16 Industrial Potential of Zeolites

of cleaner and renewable fuels. In this chapter, we highlight the potential of zeolites in supporting the technology innovation especially in the reﬁning industry.

16.2 Application of Zeolites in Slurry Processes

Many industrial conversion processes occur in the presence of catalysts that can accelerate the speed of conversion and address the selectivity toward the most desired products. The efﬁciency of the process depends on the choice of the reactor, on the process scheme, and on the conditions of reaction, namely on that sum of knowledge generally referred to as ‘‘reaction engineering,’’ as well as on the quality of the catalyst, which always remains the heart of the process. Zeolites are important as solid acid catalysts in several petrochemical and reﬁnery processes where special requirements for shape selectivity exist [5]. Usually, the desired shape selectivity of zeolites is achieved at the expense of slow mass transport to and from the active sites located inside the zeolite micropores. Diffusion can be favored by gas-phase reactions, while in slurry-phase reactions it is much slower. Nevertheless, slurry phase reaction can maximize the shape selectivity effects by several factors: 1)

Increase the pore ﬁlling and the contact time between molecules and active sites inside the pores. 2) Decrease the reaction temperature. 3) Permit an accurate control of the reaction temperature, especially important for exothermic reactions. 4) Modify adsorption properties, which can be tuned by the selection of proper solvent, considering hydrophobic–hydrophilic characters of crystal structure and the solubility of substrates/products. Therefore, slurry phase conditions can permit in many reactions to stress the zeolite properties and to increase the process effectiveness. However, to operate in slurry phase, all the tricks useful to increase the accessibility of the active sites in zeolites must be considered. As a general rule, to decrease the diffusion limitation into the zeolite channels at least one of these approaches should be followed [6]: • • • •

nanocrystalline zeolites materials with hierarchical pore structure ultra large pore zeolites delaminated zeolites.

Since this topic is treated in detail in Section 16.6, attention is paid to the possibility of controlling the porosity of the shaped catalysts for slurry applications. Industrial zeolite-based catalysts are indeed complex materials comprising different building blocks: zeolite, binder and, eventually, metal oxides, and additives. The binder phase, fundamental to shape the catalyst, plays a relevant role in the intercrystalline diffusion limitation. The porosity control of the binder component

16.2 Application of Zeolites in Slurry Processes

is an important approach in the development of catalysts useful for large-molecule transformations. Recent examples are the technologies developed by Grace Davison and Engelhard (now BASF) that permit to speciﬁcally calibrate the binder porosity for the facile transport of heavy feed molecules and the resulting products involved in residue-oil cracking catalysts [7, 8]. Many years before, a mesoporous silica binder was developed for zeolite-based catalyst useful in slurry-phase oxidation processes. This catalyst can be considered an ante litteram hierarchically ordered catalyst, made up of nanosized particles of microporous crystalline titanium-silicalite-1 (TS-1), dispersed in a matrix of mesoporous amorphous silica. It is used since 1985 in industrial slurry-phase oxidation processes. The work on TS-1-based catalyst can be considered as a pivotal development in the use of new or improved catalysts useful for applications in slurry-phase catalytic processes and will be discussed in more detail. 16.2.1 TS-1 Based Catalyst for Liquid-Phase Oxidation Processes

Several liquid phase oxidation processes are performed by using soluble oxometallic compounds as catalysts [9]. These catalysts present three main limitations: (i) the tendency of some oxometallic species to oligomerize, forming µ-oxocomplexes that are catalytically inactive; (ii) the oxidative destruction of the ligands that lead to the destruction of the catalysts; and (iii) the high E-factors, deﬁned as the mass ratio of waste to desired product [9]. An ideal catalyst should have isolated and identical stable sites, in an environment adequate from the point of view of adsorption and geometry. Suitably isomorphous substituted zeolites with their pores and cavities can be good inorganic matrices: they can introduce steric effects, while metal atoms incorporated into the framework may, in some cases, be stable toward leaching. TS-1 [10] is by far the most useful and best-understood zeolite with shape-selective oxidation properties. TS-1 has an MFI structure formed by a three-dimensional system of channels with 0.53 × 0.56 nm and 0.51 × 0.51 nm. Ti(IV) atoms are incorporated into the framework in tetrahedral sites, substituting some nonadjacent Si atoms, as clearly demonstrated by several physicochemical investigations [11]. TS-1 is a hydrophobic material and is suitable for being used as an oxidation catalyst with H2 O2 as oxidant in a large number of reactions. As with several other zeolite catalysts, the performance of TS-1 is strongly related to the optimization of the morphology and, in particular, of the size and shape of the crystallites [12]. In particular, the best performances are obtained with TS-1 samples constituted by aggregates (0.1–0.3 µm) of very small crystallites as those obtainable with the synthesis procedure reported by Clerici et al. [13]. This morphology is favored by a molar ratio Si/Ti = 40 (Type B morphology in Figure 16.1), while sample having a molar ratio Si/Ti > 65 show silicalite-like hexagonal single crystals (Type A morphology in Figure 16.1) [14].

451

16 Industrial Potential of Zeolites Type A 300 KV 50.00 nm

Type B 300 KV 50.00 nm

Figure 16.1 Morphology of TS-1 samples having with molar ratio Si/Ti = 65 (type A) and 40 (type B).

In the industrial catalyst, the sub-micron TS-1 aggregates are kept together by a binder and the shaping procedure was optimized to ensure both the formation of a network of mesopores and sufﬁcient mechanical strength [15–17]. The network of mesopores is due to the use of tetrapropylammonium ion (TPA+ ) during the shaping process. In aqueous solution TPA+ tends to form clusters whose average diameter is deﬁned by the synthesis conditions [18]. The progressive reorganization of silica oligomers around the TPA+ clusters leads, upon calcination, to a solid amorphous ‘‘sponge,’’ in which the average diameter of pores is somewhat deﬁned by the average diameter of organic clusters that have led the porosity. The N2 adsorption–desorption isotherms measured for amorphous silica obtained in presence of TPA+ cluster is reported in Figure 16.2. The irreversible Type IV + (I) isotherms are characteristic of mesoporous MSA-type materials, amorphous (alumina)-silicate prepared via sol–gel from alkaline-free mixture in presence of tetraalkylammonium hydroxide [19]. Adsorbed volume (ml g−1)

452

300 200 100 0 0

0.25 0.50 0.75 Relative pressure (p/p0)

1

Figure 16.2 N2 adsorption/desorption isotherms of amorphous silica obtained in presence of TPA+ cluster.

16.2 Application of Zeolites in Slurry Processes

50 µm Figure 16.3

5 µm

1000 Å

Hierarchical pore structure in TS-1 catalyst.

Therefore, the submicron dimensions of TS-1 crystallites and aggregates favor the intra- and intercrystalline diffusion whereas the mesoporosity of the silica binder facilitates the diffusion through aggregate particles (Figure 16.3). The above mentioned characteristics of the catalyst increases the accessibility of the active slurry phase. In the meantime, the low concentration of binder compared to the active phase ensured a high speciﬁc activity. This is the ﬁrst example of hierarchically ordered micro-mesoporous industrial catalysts, a key property at the base of the development of important industrial processes, including the hydroxylation of phenol [20], the ammoximation of cyclohexanone [21, 22], and the epoxidation of propylene [13, 23, 24]. Since the discovery of TS-1 about three decades ago, quite a number of catalytic applications have been evaluated, including the epoxidation of linear oleﬁns different from propylene, oxidation of linear alkanes to alcohols and ketones, oxidation of alcohols, hydroxylation of aromatics, oxidation of amines, and oxidation of sulfur compounds and ethers [25–33]. Several challenges are still open for a full industrial exploitation of Ti-containing zeolites in both the industrial preparation of the catalyst and the development of process technologies. The cost of hydrogen peroxide is affecting the variable cost of these processes, therefore high process performances are required in order to compensate for the drawback. Large improvements can derive from in situ generation of H2 O2 from molecular oxygen [34], as reported for Lyondell direct-oxidation propylene oxide technology. Other oxidants have also been considered, like organic peroxides and nitrous oxide [35, 36]. 16.2.2 New Advance in Slurry Phase Reaction with Zeolitic Catalysts

A factor limiting the wide applicability of TS-1 in reactions involving bulky substrates is represented by the free dimensions of the MFI pores. For this reason, several efforts have been devoted to incorporate Ti in large pore molecular

453

454

16 Industrial Potential of Zeolites Table 16.1

Selection of last three years works on Ti–zeolites.

Topic

Reference

Propylene epoxidation, in presence of Pt-TS-1 using H2 and O2 as oxidant Reduction of nitrogen oxides using Ti molecular sieves Synthesis of Ti-SSZ-70 SSZ-71 and its application in partial oxidation, reduction of nitrogen oxides, acylation reactions Crystallization of nanosized TS-1 crystals in presence of adding at least one polymer with spheroidal geometry as pore forming agent Propylene epoxidation, in presence of H2 O2 produced in a previous step without washing and puriﬁcation treatments α-Pinene epoxidation with mesoporous titanosilicate synthesized by assembling of colloidal TS-1 precursors with a surfactant Synthesis of Ti-beta/mesoporous silica composite by deposition of Ti-Beta nanoparticles on mesoporous silica Synthesis of Ti-YNU-1, analogous to MWW-type lamellar precursor with expanded interlayer spacing Ammoximation reactions over Ti-MWW Epoxidation reactions over with Ti-MWW Synthesis of Ti-silicate form of BEC polymorph

[46] [47] [48] [47, 49] [51] [52] [53] [54, 55] [56] [57, 58] [59, 60] [61]

sieves, leading to the synthesis of several Ti-containing zeolites including Ti-Beta, Ti-ZSM-12 (MTW), Ti-ITQ-7 (ISV), and Ti-MCM-22 (MWW) [6, 37–43] or to the preparation of micro-mesoporous Ti-zeolites [44] and Ti-zeolite/mesoporous silica composites (see Table 16.1). Also other transition-metal-substituted zeolites have been synthesized and are active and selective for carrying out oxidations in liquid phase using H2 O2 or organic peroxides as oxidants [45]. The scientiﬁc and patent literature in this ﬁeld is very rich, so only a selection of works published in the last three years are reported in Table 16.1. Besides oxidations, the use of zeolitic catalysts in slurry phase condition has been proposed for other reactions: • Parafﬁn alkylation. Shell reported the use of zeolitic catalyst (Beta-zeolite) in a well-stirred slurry reactor for the alkylation of n-butene with iso-butane to produce iso-octanes [62]. The use of slurry phase conditions permits to increase the lifetime of the catalyst at high oleﬁn conversion. Other key elements are the use of a zeolite with high activity/stability and acid site density, the use of a low oleﬁn concentration. • Anisole acylation. Deactivation is the primary roadblock to the application of zeolite in aromatic acylation, due to the slow diffusion of reactants through micropores. Recently papers report the use of nanocrystalline zeolites in the acylation of anisole with acetic anhydride. Nanocrystalline zeolite particles of

16.3 Rebalancing the Reﬁnery Products Slate

Beta [63, 64] and ZSM-5 [65] are characterized by large external surface areas and low diffusion distances in the channels of the particles, thereby a large active surface is exposed for reaction and deactivation is reduced. High-activity catalysts in the acylation of anisole with octanoic acid were also obtained for Beta and Y zeolites containing a large amount of mesoporous area. These zeolites were coated on a ceramic cordierite, resulting in an active and selective integrated catalyst-reactor conﬁguration [66].

16.3 Rebalancing the Reﬁnery Products Slate

The reﬁneries can be classiﬁed according to their complexity. The reﬁnery complexity index (Nelson Complexity Index (NCI)) proposed in the 1960s by W.L. Nelson was aimed to quantify the relative cost of components that make up a reﬁnery, or in other words, to quantify the value of the products generated by a reﬁnery: more complex reﬁneries generate more high-value products. In fact, typical yields from low, medium, and high conversion reﬁneries can be represented in the following way [67]: • Low-conversion (NCI = 2–3): 20% gasoline, 35% middle distillates, 30% fuel oil, 10% other products (including reﬁnery gas, liquiﬁed petroleum gas (LPG), solvents, coke, lubes, wax, and bitumen), and 5% loss; • Medium-conversion (NCI = 5–6): 30% gasoline, 30% middle distillates, 30% fuel oil, 15% other products, and 5% gain; • High-conversion (NCI = 9–10+): 50% gasoline, 30% middle distillates, 15% fuel oil, 15% other products, and 10% gain. The percent of gain and loss reported above are referred to the volume of products in comparison to the volume of total crude oil run. Because of cracking, the volumes of petroleum products, in a complex reﬁnery can exceed total crude oil runs by as much as 5–10%; in contrast, a reﬁnery with a NCI around 3–5 will have a volumetric contraction. The unit operations most important to increase the complexity are the cracking units, and among these, the key technology is the Fluid Catalytic Cracking (FCC). The FCC unit is able to convert vacuum distillates to gasoline and therefore it increases the volume gain and the gasoline production. Over the last 30 years, much attention has been devoted to increase the throughput of gasoline, but recently this trend seems to be less adequate than in the past. Particularly in Europe, from the year 2000, the demand for diesel was steadily increasing, while the demand for gasoline was declining. In Figure 16.4 is presented a projection for the evolution of the demand for EU petroleum products from 2000 until 2020 reported from CONCAWE [68]. Today, there is a net imbalance in Europe between fuel demand and production: gasoline is exported toward the United States and diesel gas-oil is imported from Russia. The need to readjust the imbalance is a strong incentive to modify the reﬁning technologies in order to balance the

455

456

16 Industrial Potential of Zeolites

674

697

726

739

Total demand (Mt yr−1)

739 5

100%

LPG 4

80%

Gasoline Petrochemicals

60%

3

40%

2

20%

1

Middle distillates Residual marine fuel Residual inland fuel Others Middle distillate/ gasoline

0% Year

0 2000

Figure 16.4

2005

2010

2015

2020

Evolution of petroleum products demand in Europe.

product slate by reducing the excess gasoline and increasing the gas-oil production, keeping roughly constant or even increasing the products volume. Since FCC is the technology addressed mainly to the conversion of vacuum gas-oil to gasoline, it appears useful check about the possibility to modify the throughput of FCC. Figure 16.5 reports a synthetic representation of the streams, which are typically generated by an FCC and their ﬁnal destination use. The lighter parafﬁns are used as fuel gas or LPG. Among the oleﬁns, propylene can be addressed to chemical uses, while C4 and eventually C5 oleﬁns can be used for gasoline production by means of alkylation or etheriﬁcation reactions. The cracking naphtha is used for the production of gasoline. Light cycle oil (LCO) is used in part for the production of a low quality diesel-fuel and in part as a ﬂow improver for heavy streams. The heavy cycle oil (HCO) is used to produce heating fuel oil or bunker fuel Products FCC

Final use Fuel gas/LPG

Gas

Olefins

Cat naphtha (LCN, HCN) Light cycle oil (LCO) Heavy cycle oil (HCO) Vacuum gasoil Figure 16.5

FCC products and their use.

Chemistry Gasoline ~80 Diesel gasoil ~20 Bunker, fuel oil, bitumens

16.3 Rebalancing the Reﬁnery Products Slate

oil or bitumen. The demand for fuel oil is in general declining and therefore besides the rebalancing of the gasoline/diesel ratio another important need is the improvement in the conversion of HCO. Reﬁners can use some strategy to move in this direction by means of incremental solutions such as optimize operation conditions, shifts cut point of LCO and gasoline, recycle streams, modify feedstock and pre-treatment operations, and optimize catalysts selection. Other potential more radical ways to improve the LCO/gasoline ratio, without affecting the conversion are improve the cracking of HCO limiting coke and gas formation, improve the LCO upgrading processes for ensuring higher efﬁciency and better product quality, use light oleﬁns when not used for chemistry, for the production of middle distillates for instance through oligomerization. In the following, we will consider in synthesis these three opportunities where zeolite science and technology can make an important contribution. 16.3.1 Bottom Cracking Conversion

The increase in the conversion of bottom FCC is a challenge that involved reﬁners, technology companies, and catalysis researchers for many years. The FCC feed boiling above 430 ◦ C cannot be cracked by the conventional zeolite component of an FCC catalyst since the molecular diameters are too large to pass through the pore opening of the zeolite. By using molecular simulation tools, it has been calculated that parafﬁn boiling in the range 430–700 ◦ C possesses an average carbon number ˚ while for heteroatom of about 25–35 and a dynamic molecular size of 12–20 A, containing poly-aromatics the average carbon atom number is 12–25 and the average dynamic molecular size is about 12–15 A˚ [69]. By assuming the model proposed by Spry and Sawyer [70], for the effective diffusion rate of a molecule in the pore of a solid porous material, Deff = Dbulk (1 − dmolecule /dpore )4 It appears that the pore size needs to be about 10–20 times larger than the size of diffusing molecules. However, we should take into consideration some other effects in order to propose a catalyst suitable to increase the bottom conversion. Some authors have suggested that the cracking of heavy molecules takes place across three stages, as shown in Figure 16.6 [70]. At each stage, the formation of coke is a competing path, but the probability to form coke increases when attempting to convert the PNA. Obviously, this is a simpliﬁed model, which considers only the reactions that are relevant to the cracking of heavy fractions. In order to improve the ability to convert the different molecules that are present in a FCC feedstock, the FCC catalyst have been progressively improved in the last 60 years. The ﬁrst cracking catalyst was AlCl3, introduced by Almer McAfee from Gulf Reﬁning Company around 1920 [71]; few years later Eugene Houdry discovered the catalytic properties of activated clays [72] and then of silica-aluminas [73]. One of the main problems in catalytic cracking was the deposition of coke on the catalyst,

457

458

16 Industrial Potential of Zeolites

Alkylaromatic dealkylation

Coke Precracking/ feed vaporization

Conversion of naphthenoaromatics

Feed

R

R′ Figure 16.6

Cracking of heavy molecules.

which required frequent regenerations. This was the spring for the development of new reactors and processes, such as the semi-continuous cracking operations (Thermofor Catalytic Cracking (TCC)) [74] and the continuous cracking in ﬂuid bed reactors (FCC) [75], which were developed around 1942–1943. Only 20 years later appeared patents claiming the use of zeolites in FCC catalyst [76] almost in the same period in which Donald W. Breck patented the synthesis of a crystalline Y zeolite [77]. Since then, zeolites have become an essential component of catalysts for FCC, either as active phases (coke-selective zeolite Re-HY and high coke-selective US-Y and RE-US-Y) or as pro-oleﬁns and octane-boosting FCC additives (ZSM-5 zeolite). The current FCC catalysts consist of a complicated combination of different materials, which have different functions: acidic functions in amorphous and crystalline matrices, metal impurities traps, combustion promoters, SOx traps, octane boosters, and oleﬁns promoter additives [78]. For the purpose of this discussion, it is worth to consider only the acidic functions. The acidic functions in FCC catalysts are due to the local structure of aluminosilicates either crystalline or amorphous. Acid sites in the crystalline framework of zeolites have a lower tendency to form coke because of the shape selectivity. In addition, the aluminum concentration in the crystalline framework can affect the properties of the acid site: high Si/Al ratios tend to depress the hydrogen transfer attitude and therefore to reduce the oleﬁns saturation, the aromatic formation, and the coke formation [79]. The size of the zeolite pores limits the activity to the molecules that can access the active sites. The porosity of the silica–alumina matrix can be tuned by the preparation procedure and pore size can span from micro-, to meso- to macropores. An interesting schematic representation of the pores architecture of

16.3 Rebalancing the Reﬁnery Products Slate

Macropores

Mesopores

Micropores

Ni

V

Gasoil

Resid

Figure 16.7

Gasoline

LPG

Pore architecture of a FCC catalyst.

a FCC catalyst is reported in Figure 16.7 [79]. In the last 40 years, the zeolite phase has been optimized in order to achieve the maximum conversion and the optimum product quality. From this point of view, the best product is still the modiﬁed Y-type zeolite. As a consequence of the concept reported above and the schemes represented in Figures 16.6 and 16.7, the actual bottom cracking additives are manufactured by optimizing the matrix in order to increase the conversion of heavy molecules in the matrix mesopores. In spite of the improvement, the main problem of this approach is that the selectivity toward the formation of coke is inevitably enhanced because of the lack of shape selectivity. It is clear that it will be impossible to synthesize crystalline alumino-silicate with average pore size diameter 10–20 times larger than reactant molecules, in order to allow diffusion into the pore, but probably this would not be required. In fact, surface hemicages or pore mouth in crystalline zeolites could be sufﬁcient to convert part of the large molecules, for instance, to crack the endocycle C–C bond of a cyclo-parafﬁn ring to an aromatic polynuclear molecule, to allow a selective conversion of a further part of the bottom FCC. From this synthetic picture, it appears that the synthesis of extra large pore zeolites (pore diameter > 12-ring), stable at the FCC condition, is still of great relevance and even more importance now than in the previous years [80]. 16.3.2 LCO Upgrading

The light cycle oils are middle distillates of very poor quality. The typical composition and characteristics of a LCO are reported in Table 16.2. Usually the large part of LCO is fed with other streams to a de-sulfurization unit and it is used for the production of diesel. The presence of LCO depresses the diesel quality and must be limited. The other part of LCO is used to ﬂow the fuel oil and the heavier part of the barrel. In order to increase the possibility to convert LCO, it appears useful to ﬁnd technologies able to hydrogenate this class of compounds until to reach the desired density at about 845 kg/m3 and improve the cetane index above 45 with the minimum hydrogen consumption. The best way to

459

460

16 Industrial Potential of Zeolites Table 16.2

Properties and compositions of a typical LCO.

Density (15 ◦ C)

950−1 050 kg/m3

Distillation <90% Aromatics content Di-aromatics and heavier Sulfur content Nitrogen content Cetane index

350–400 ◦ C 70–80% 40–60% 2 000–7 000 ppm 200–600 ppm 20–25

reach this objective is through a hydrogenation-cracking pathway that maximizes the formation of alkylbenzenes, as represented in Figure 16.8. Further hydrogenation of alkylbenzenes gives rise to cycloparafﬁns and then parafﬁns, which are still good components for diesel but the total hydrogenation requires the highest hydrogen consumption. On the other hand, the alkylbenzenes have a much higher cetane index then the corresponding alkylnaphtalenes and a lower more acceptable density; therefore, it appears convenient to maximize this class of molecules. To move along the pathway indicated by the asterisk in Figure 16.8, the presence of Brønsted acid sites and metallic sites is necessary. In fact, the ring opening of C–C endocyclic bonds can proceed through metal catalysis. Metals such as Pt, Pd, Ir, Ru, and Rh have been found to be selectively active for the ring opening of cyclic parafﬁns, however, these metals lack activity for the ring opening of six-membered rings, which are intrinsically more stable. A Brønsted acid site is required to generate the formation of an intermediate carbenium ion, which isomerizes into ﬁve-membered ring that readily undergoes subsequent ring opening to the metal site [81]. The acid site’s number and strength and the distance between the acid and the metal site are relevant to the catalyst selectivity and stability. The hydrogen spillover supplied by the metal function is locally important in order to prevent the coke formation and growth in correspondence of the acid site [82]. The need for a bifunctional catalyst with a controlled pore size, acid size distribution, and metallic site distribution, constitute a challenge and an opportunity Hydrogenation

∗

Figure 16.8

Selective ring opening

∗

Possible pathways for naphthalene hydrogenation.

16.3 Rebalancing the Reﬁnery Products Slate

for zeolite science and technology. Several reports appeared in the last years on the selective ring opening of aromatic hydrocarbon [83–89], nevertheless the industrial need for an economically viable process is still open, making this subject one of the key area of heterogeneous catalysis. 16.3.3 Oleﬁns Oligomerization

The FCC naphtha (LCN) is a stream, which has a high oleﬁn content. In Table 16.3 a typical composition of a naphtha is reported, from which it is possible to derive the variation of the oleﬁns content with the boiling point [90]. As expected, the lighter fraction is much richer in oleﬁns than the heavier fractions. The oleﬁn fraction is composed of C4 (±10%), C5 (±50%), C6 (±40%), with 1–2% of C7 . This stream could be an interesting feedstock for feeding an oligomerization process aimed at the production of jet or diesel fuels. One of the early catalysts used in this process was phosphoric acid supported on kieselguhr (SPA), developed by Ipatieff in 1935 [91]. Starting from 1950, several solid acid catalysts were demonstrated to be active for this reaction: silica-alumina, clays, sulfonic resins, silica-zirconia and silica-titania, mesoporous silica-alumina, and zeolites. This subject has been extensively discussed and reviewed in a recent publication from the Institute Francais du Petrole [92]. The degree of branching of the oligomers produced by oligomerization of C3 –C4 oleﬁns increases with the size of the pore opening of the zeolites in the following order: Offretite < ZSM-5 < H-Mordenite < H-Y < Mazzite < Mesoporous silicaalumina. Since oligomerization occurs through ionic mechanism, the products are branched, unless effects due to space limitations do not reduce branching. The mesoporous silico-alumina is among those cited above, which promotes the maximum degree of branching and thus are more suitable for the production of jet fuels [93]. For the production of diesel fuel, it is necessary to use catalysts with a high shape selectivity, such as zeolite ZSM-5. The porosity of ZSM-5 is ideal to secure catalyst stability and to promote the synthesis of middle distillates consisting Table 16.3

Typical composition of a FCC naphtha.

Property Fraction, wt% S, ppm N, ppm PONA, wt% BrNo, g/100 ml P+N Oleﬁns Aromatics

IBP-75 ◦ C

75–125 ◦ C

125–150 ◦ C

150 ◦ C-FBP

Full range

22 15 5

30 20 6

16 40 20

32 120 75

(100) 50 35

100 33 65 2

75 38 50 12

45 30 30 40

25 20 20 60

70 30 45 25

461

462

16 Industrial Potential of Zeolites

of oligomers whose degree of branching is low [94, 95]. The addition of longer linear oleﬁns, C4 –C5 , reduces further the branching of the ﬁnal product, increasing the cetane index. Addition of Ni to the ZSM-5 can improve the activity and it offers the possibility to reduce the reaction temperature, thus further reducing the branching degree of the oligomers [96]. Today, oleﬁn oligomerization processes are applied in South Africa. The SASOL Synfuels reﬁneries in Secunda are using a technology based on SPA, while the PetroSA reﬁnery in Mossel-Bay is using a ZSM-5 zeolite-based technology [97]. The oleﬁns oligomerization is not yet a current technology applied in reﬁnery but it can become a useful opportunity in the near future to rebalance product slate and to improve product quality.

16.4 Advanced Separation Technologies

In chemical industry and reﬁnery, separations accounts for more than 40% of total energy consumption [98], thus strongly contribute to processing costs. Improvements in separation technology are key to improve both capital effectiveness and efforts to reconﬁgure process technology to minimize noxious efﬂuents and their environmental impact [99]. Adsorption on solids is applied for the removal of components from either gaseous or liquid mixtures, when there is a little difference in volatility between the components to be separated, or where any of the components are permanent gases or nonvolatile liquids. Adsorption process is based on the preferential partitioning of substances from the gaseous or liquid phase onto the surface of a solid substrate by accumulation or concentration phenomena. Physical adsorption is caused mainly by van der Waals forces and electrostatic forces between adsorbate molecules and the atoms, which compose the adsorbent surface. Thus, adsorbents are characterized ﬁrst by surface properties such as surface area and polarity. The most fundamental consideration regarding the adsorption of chemical species by zeolites is molecular sieving. The micropore size determines the accessibility of adsorbate molecules to the internal adsorption surface: species with a kinetic diameter too large to pass through zeolite pores are effectively ‘‘sieved’’. This ‘‘sieve’’ effect can be utilized to produce sharp separations of molecules by size and shape. These molecular sieving or steric effects can be caused by the shape of the crystal aperture itself, but a relevant role is also played by exchangeable cations, located within the cavities to balance the negative charge introduced onto the framework by the aluminum atoms. Cations can effectively block the pore openings and can also decrease the pore volume [100]. Therefore, zeolite can be speciﬁcally engineered with precise pore size distributions and hence ﬁne-tuned, thanks to a proper combination of structure and extraframework cations. Besides the presence of cations induces high electric ﬁeld gradients within the cavities and the framework itself can possess acidic or basic character. In addition to steric effects, the selectivity of zeolites for a particular adsorbate will also depend on the polarity, magnetic susceptibility, and polarizability of the molecules. Thus,

16.4 Advanced Separation Technologies

excellent separations can be achieved by zeolites even when no steric hindrance occurs [101, 102]. The hydrophilic/hydrophobic behavior of zeolite can be modiﬁed by the silica/alumina ratio. In high aluminum content zeolites (i.e., A type or LTA, X, and Y or Faujasites, Mordenite, other natural zeolites), the strong electrostatic ﬁeld within zeolite cavities results in very strong interactions with polar molecules such as water; while high silica zeolite (i.e., silicalites or ZSMx zeolite) are typical non-polar adsorbents [103]. Adsorption based on molecular sieving, electrostatic ﬁelds, and polarizability is always reversible in theory and usually reversible in practice. Zeolite can be reused many times, cycling between adsorption and desorption. This accounts for the considerable economic value of zeolite in adsorptive applications. The use of zeolite in the adsorption and separation areas has an economic impact similar to their use in reﬁnery and petrochemical catalytic processes [104]. Zeolites are used in many separation and puriﬁcation applications in several industrial ﬁelds: petroleum reﬁning processes, petrochemicals, natural gas treating, industrial gas production and puriﬁcation, specialty and ﬁne chemicals, and pharmaceuticals. Adsorption on zeolites plays also a relevant role in environmental protection applications, ﬁnalized to waste reduction (i.e., recovery of solvents from industrial off-gases, radioactive waste management, builders for phosphate-free

Table 16.4

Representative commercial adsorption separations.

Gas bulk separations (Adsorbate concentrations of about 10 wt.% or higher in the feed) n-Parafﬁns, iso-parafﬁns, aromatics N2 /O2 CO, CH4 , CO2 , N2 , NH3 /H2 H2 O/ethanol Gas puriﬁcations (Adsorbate concentrations generally less than about 3 wt.% in the feed) H2 O/oleﬁn-containing cracked gas, natural gas, air, synthesis gas, and so on. CO2 /C2 H4 , natural gas, and so on. Sulfur compounds/natural gas, hydrogen, liquiﬁed petroleum gas (LPG), and so on. NOx /N2 SO2 /vent streams Hg/chloro-alkali cell gas efﬂuent Liquid bulk separations (adsorbate concentrations of about 10 wt% or higher in the feed) n-Parafﬁns, iso-parafﬁns, aromatics p-Xylene/o-xylene, m-xylene Detergent-range oleﬁns/parafﬁns p-Diethylbenzene/isomer mixture Fructose/glucose Liquid puriﬁcations (adsorbate concentrations generally less than about 3 wt% in the feed) H2 O/organics, oxygenated organics, chlorinated organics, and so on. Sulfur compounds/organics Source: http://ias.vub.ac.be/What%20is%20adsorption.html.

463

464

16 Industrial Potential of Zeolites

laundry detergents) [99, 105]. Representative commercial adsorption separations performed with zeolites are reported in Table 16.4. In reﬁnery, adsorption on zeolites is applied to the puriﬁcation of feeds and petroleum products, thanks to their strong afﬁnity for polar compounds such as water, carbon dioxide, hydrogen sulﬁde, mercaptans, and organic chlorides. The main separation processes are related to removal of CO2 , chlorides, and mercury from a variety of streams; drying and puriﬁcation of liquids and gases; treatment of alkylation unit feed to reduce acid consumption, regenerator use and corrosion, puriﬁcation of reﬁnery H2 to prevent corrosion in downstream equipment, drying and desulfurization of reﬁned products; drying and puriﬁcation of feed and recycle hydrogen in isomerization units [105]. In petrochemical industry, adsorptions on zeolite are used to dry and purify feedstocks (i.e., hydrocarbons, hydrogen, ethane, propane, ethylene, propylene, and butadiene). In natural gas production and puriﬁcation, zeolite are today used to remove H2 O and CO2 from air before liquefaction and separation by cryogenic distillation and to remove sulfur compounds from the natural gas or LPG fractions to prevent corrosion in burners [105]. Adsorption separation processes can be categorized by process equipment features. Two major groups of equipments can be applied: ﬁxed-bed and ﬂuidized-bed systems. One early example of ﬂuidized-bed systems is the MolexTM process using a 5A zeolite adsorbent and light naphtha as desorbent for the separation of linear and branched chain parafﬁns. The ﬁrst Sorbex unit, a Molex unit, came on stream in 1964. The MolexTM process is part of a family of similar processes, which simulate a moving bed of adsorbent with continuous countercurrent ﬂow of a liquid feed over the adsorbent bed, developed by UOP for a variety of difﬁcult industrial separations and referred with the generic name Sorbex processes [106]. They represent a major commercial use of zeolite chemistry to process adsorption applications. SorbexTM separation technologies are most commonly used in conjunction with various catalytic processes. Important separation Sorbex processes applied in reﬁnery and petrochemical industries are as follows [106]: • ParexTM using a cationic form of X and Y zeolites as adsorbent and toluene and p-diethylbenzene as desorbent. The ﬁrst ParexTM unit came on stream in 1971 and revolutionized the method by which p-xylene was produced, rendering the former method of crystallization a secondary route [104]. • MolexTM process using a 5A zeolite adsorbent and light naphtha as desorbent for the linear and branched chain parafﬁns. Separated linear parafﬁns are required to make linear alkylbenzenes (LABs) for detergents. Linear parafﬁns are ﬁrst selectively adsorbed and isolated, either by the Iso-SieveTM process or the MolexTM process, then sent to the catalytic process to be converted in monooleﬁns and after in alkylation unit. • EbexTM process using cationic form of X and Y zeolites as adsorbent and toluene as desorbent for the recovery of ethylbenzene,

16.4 Advanced Separation Technologies

• OlexTM process using CaX zeolite as adsorbent and light naphtha as desorbent for the separation of oleﬁns from saturated hydrocarbon isomers, • CresexTM process for separation of p-cresol or m-cresol from other cresol isomers, • CymexTM process for separation of p-cymene or m-cymene from other cymene isomers. Adsorption separation processes using ﬁxed-bed systems can be divided in two main groups: the ﬁrst requiring the removal of adsorbent for regeneration and the second allowing in situ regeneration, as for pressure swing adsorption (PSA) or thermal swing adsorption (TSA). The PSA process is used to separate some gas species from a mixture of gases under pressure. The higher the pressure, the more gas is adsorbed; when the pressure is reduced, the gas is released or desorbed. Typically, the desired component is not adsorbed and is recovered at high purity. In 1970s, the ﬁrst economical PSA to produce O2 was developed thanks to the use of a zeolite adsorbent (NaX). Further improvements derived from both subsequent generation adsorbents (CaX, LiX, LiCaX, etc.) and improved processes (VSA), permitting a strong reduction of capital, operating costs, and power requirements [99]. One of the primary applications of PSA is in the removal of carbon dioxide as the ﬁnal step in the large-scale commercial synthesis of hydrogen for use in oil reﬁneries and in the production of ammonia. Reﬁneries often use PSA technology in the removal of hydrogen sulﬁde from hydrogen feed and recycle streams of hydrotreating and hydrocracking units. In the natural gas, PSA are mainly applied in puriﬁcation processes. For example, Sorbead Quick-Cycle Process (BASF) permits the simultaneous removal of heavy hydrocarbons, mercaptans, water from natural gas by using a premium silica gel-based adsorbent [107]; Polybed PSA System is applied in puriﬁcation of methane for petrochemicals production. The bulk separation of natural gas from other gaseous components normally found in the dry natural gas, namely, N2 , CO2 , and H2 S, takes usually place by cryogenic treatment (nitrogen) and by extraction with appropriate chemicals (acid gas absorption by amines). For over 60 years, the amine based absorption/stripping process has been used for the puriﬁcation of natural gas and ﬂue streams. Despite its wide commercial use, this technique has several drawbacks including low carbon dioxide loading capacity, high equipment corrosion, amine degradation by SO2 , NO2 , HCl, HF, and O2 in ﬂue gas and high energy consumption during solvent regeneration. Due to all these aspects, amine-based separation treatments are costly and a large improvement could derive by availability of adsorption processes on effective and selective solid adsorbents through changes in pressure or temperature (PSA, TSA). PSA can be an economic choice especially for upgrading small volumes of landﬁll gas or for off-shore applications. In the separation of natural gas from carbon dioxide, the carbon dioxide is preferentially adsorbed with the natural gas passing through the adsorbent bed. By contrast, in the separation of nitrogen from natural gas, adsorbents generally

465

466

16 Industrial Potential of Zeolites

show a preferential afﬁnity for methane over nitrogen: methane is preferentially adsorbed with the nitrogen passing through the adsorbent. Nitrex (UOP) [108] and Nitrotec (Nitrotec Corporation) are the ﬁrst processes commercialized for nitrogen removal from natural gas. They use activated carbons as adsorbents and methane is preferentially adsorbed and recovered depressurizing the adsorbed bed. Natural gas is so obtained at low pressure and the process is not highly efﬁcient in the enrichment of natural gas. A large improvement can be made by the use of adsorbent that is able to adsorb the impurity molecules selectively. In this case methane can be kept at the same pressure, while the impurities can be desorbed at low pressure, avoiding the need for gas recompression and thus allowing energy and cost safe. The Molecular Gate technology from Engelhard Corporation (from 2006 BASF) is based on a peculiar adsorbent that preferentially adsorbs nitrogen over methane. The adsorbent is based on ETS-4, a mixed octahedral/tetrahedral titanium/silicate framework [109]. The effective ETS-4 pore size can be manipulated by cation-exchange and thermal treatment. In its contracted form (CTS material) a pore opening of roughly 3.7 A˚ ˚ can be obtained. This pore opening allows nitrogen (molecular diameter of 3.6 A) ˚ to enter and adsorb while excluding methane (molecular diameter of 3.8 A). For this adsorbent, signiﬁcant co-adsorption of carbon dioxide (molecular diameter of ˚ is also reported. The Molecular Gate technology can be applied for the 3.4 A) removal of N2 [110, 111] or CO2 [112] from CH4 . A selective adsorption of nitrogen from natural gas is also described for clinoptilolite, a zeolite belonging to the heulandite group [113]. A very high selectivity in CO2 /CH4 separation can be obtained using ERS-7 (ESV) zeolite, an original small-pore zeolite. It is characterized by the presence of a one-dimensional system of channels having a section with an average diameter 3.5 × 4.3 A˚ [114, 115]. Along these channels, large cavities are present. While the CO2 molecules can easily pass through the channel’s openings, the larger CH4 molecules have more difﬁculty in diffusing through, so under certain conditions, only CO2 can accumulate in the large cavities. This allows the effective separation of the two different molecular species and to release the CO2 accumulated in the large cavities [116]. Furthermore, the improvement in adsorbed materials, large efforts are going on toward new technological solution. Some examples are here reported: • The QuestAir technology that combines patented QuestAir rotary valve technology and conventional beaded adsorbents with an optimized PSA cycle to deliver higher methane recovery performance than conventional PSA systems [117]. • UOP claims the use of high-pressure rotary adsorbent contractors (monolithic wheels) to purify a high-pressure gas stream. Wheel generally uses thin layers of adsorbent material. The advantage is reduction of adsorbent required, high mass transfer, and low-pressure drop [118].

16.5 Zeolites and Environmental Protection: Groundwater Remediation

• Monoliths can be made from paper like sheets containing polymeric ﬁbers. During the manufacturing process, they can be embedded with active adsorbents such as zeolites and other porous materials [119–123]. Synergistic approach involving several skills is required for improved regenerable solid-bed adsorption processes, permitting reduction in energy requirements and increase in separation selectivity.

16.5 Zeolites and Environmental Protection: Groundwater Remediation

Among the different innovative applications of zeolites, remediation of contaminants in groundwater surely deserves a particular attention. The quality of the groundwater is determined by the nature and concentration of contaminants (inorganic and/or organic), which should be removed for rendering water available for human uses. Traditionally, the remediation is performed using the pump&treat (p&t) technology, which involves the extraction (pump) and the aboveground treatment (treat) of the groundwater before its re-injection. Though widely applied, this technology has some drawbacks due to long operating times and the high costs bounded to the high energy demand of the water pumping. To overpass these limitations, another newer technology is available, based on the use of the so-called permeable reactive barriers (PRBs), in which the remediation is made directly on the groundwater, without the need (and the costs) of pumping it on the ground [124]. Independent of technology, the crucial point is constituted by the material used for treating water. Considering the organic contaminants only, the p&t technology is mainly based on the use of granular activated carbon (GAC), a cheap and efﬁcient material with a wide range of action. GAC, however, presents some relevant drawbacks, the most critical being the adsorption of humic substances and inorganic species often contained in groundwater, which reduce the efﬁciency in removing the contaminants and strongly affect the possibility to regenerate the exhaust material. Furthermore, GAC displays poor effectiveness in removing highly soluble contaminants and/or polar molecules such as alcohols, ethers (the most important being methyl-tert-butyl ether (MTBE)), sugars, starches, and so on. In more innovative systems, GAC is replaced by, for example, ion-exchange resins on polymeric supports or speciﬁc adsorbents [125]. Depending on the type of treatment, PRB can be classiﬁed in different categories: • adsorbent barriers, in which the material used is an adsorbent such as GAC • biological barriers, which take advantage of the presence of autochthonous bacterial colonies able to destroy the organic contaminants without leaving toxic compounds (bioremediation) • chemical barriers, based on the use of materials (e.g., zero-valent metals such as Fe, Sn, Zn or Pd/Fe, Ni/Fe) able to transform the organic contaminants in less

467

468

16 Industrial Potential of Zeolites

harmful compounds and to reduce some heavy metals to non toxic species (e.g., CrVI to CrIII ). With respect to the adsorbent barriers, both the biological and chemical ones are preferable because, in principle, they do not require regeneration treatments. In the case of the chemical barriers, however, the main drawbacks concern the restricted ﬁeld of application (limited to some heavy metal ions and to chlorinated aliphatic compounds), the relatively slow kinetics of the degradation reactions (which implies an increase of the residence time and, consequently, of the thickness of the barrier), the overall life of the material, dependent on the nature and concentration of the chemical species dissolved in the groundwater [126]. In this scenario, zeolites may represent an attractive alternative because of their peculiar ion-exchange and sorption properties. The cheap and abundant natural zeolites (clinoptilolite, heulandite, and chabazite), for instance, are widely used for removing inorganic contaminants from water. They ﬁnd application in several important areas, such as in the treatment of radioactive wastes produced in nuclear plants or contaminated after nuclear accidents, and in perspective, for developing advanced devices (e.g., life-support wastewater system studied by NASA by using clinoptilolite) or for storing highly radioactive wastes in mordenite and clinoptilolite sedimentary deposits [127]. Natural zeolites as such, on the other hand, do not have a great potential for removing organic contaminants because of their high hydrophilicity and the presence of counterions and water molecules in the channels, which ﬁnally reduce the free space available for adsorbing organic molecules. An interesting and promising route for applying natural zeolites in environmental remediation has been opened by the research group directed by R. S. Bowman at the New Mexico Institute of Mining and Technology, who were able to ‘‘activate’’ natural zeolites through their chemical modiﬁcation with surfactants [128]. In particular, they demonstrated the possibility of modifying the surface of sedimentary zeolite (a clinoptilolite-rich zeolitic tuff from the St. Cloud deposit (Winston, New Mexico) containing 74 wt% of clinoptilolite) by using an organic surfactant (e.g., hexadecyltrimethylammonium (HDTMA)), producing the so-called surfactant modiﬁed zeolite (SMZ). Starting from the observation that the surfactant loading capacity of clinoptilolite is roughly double of its external cation-exchange capacity, the bilayer model shown in Figure 16.9 has been developed. In this model, a ﬁrst layer of surfactant cations is produced when they exchange the alkali and earth-alkali metal ions located on the external surface of the crystal. The formation of the second layer is favored by the van der Waals interactions among the long alkyl chains of the surfactant (Figure 16.9). In this way, a deep modiﬁcation of the properties of the external surface of the zeolite crystals occurs, in particular, in the following ways [128]: • The surface is now positively charged since the hydrophilic heads [–N(CH3 )3 ]⊕ of the surfactant ions are exposed to the exterior; in this way, SMZ acquires anion exchange properties. • The modiﬁcation does not alter the cation exchange properties of the zeolite pores.

16.5 Zeolites and Environmental Protection: Groundwater Remediation

CrO4= Cl−

Cl− + +

M+

CrO4= Cl− + + +

CrO4= + + +

Anion exchange

Su

rfa

ce

Organic layer + M+

+

+

+ +

+

+ M+

+

+

M+

Zeolite crystal

Figure 16.9

Schematic representation of the surfactant modiﬁed zeolite (SMZ).

• The long alkyl chains form a hydrophobic region where small, non-polar organic molecules can dissolve. In other words, the functionalization of the surfactant produces materials with a variety of different properties, which can be appropriately used for efﬁciently treating water contaminated both by inorganic and organic substances. For instance, laboratory tests demonstrated that SMZ efﬁciently removes CrO4 2− and Pb2+ ions and to retain benzene and perchloroethylene (PCE) [128], results recently conﬁrmed at a pilot plant level [129]. SMZ proved to be very efﬁcient also for removing organic molecules contaminating oilﬁeld wastewaters and even pathogens (viruses and bacteria) [130]. Interestingly, SMZ can be used in combination with reactive materials. For instance, SMZ and zero-valent iron (ZVI) proved to remove efﬁciently chromate ions and PCE during a pilot test, while a SMZ prepared with a nutrient-amended commercial zeolite was used in combination with appropriate microorganisms for destroying toluene [129]. Overall, the modiﬁcation of natural zeolites by surfactants surely represents an interesting option to improve the capabilities in the remediation of contaminated water. In spite of a large number of materials and technological solutions available, problems still exist when particular contaminants should be removed, particularly from groundwater for which the legislation imposes very low threshold values. It is the case of molecules such as MTBE, which display a certain afﬁnity with water. MTBE is not efﬁciently adsorbed on GAC nor is predicted to have a favorable partition coefﬁcient in the organic bilayer of SMZ. Zeolites may represent valid candidates as demonstrated by the recent experience performed by Eni both in laboratory and in ﬁeld tests.

469

470

16 Industrial Potential of Zeolites

The use of apolar zeolites with different characteristics for the treatment of groundwater contaminated with different hydrocarbons has been claimed in 2005 [131]. The process is based on the use of zeolites with a high SiO2 /Al2 O3 molar ratio (>50) to reduce their hydrophilicity and exchange capacity. The best performances are obtained when two zeolites with different pore openings are used in the sequence. For instance, a possible combination involves, in the ﬁrst stage, the use of a zeolite with pore openings >7 A˚ and with high sorption capacity (e.g., zeolite Y) necessary for reducing the concentration of the contaminants to medium-low level. Successively, the groundwater passes through a second zeolite with pore openings in the range 5–7 A˚ (e.g., ZSM-5, mordenite), allowing the complete removal of the contaminants. Such kind of treatment can be done either in a p&t conﬁguration, with two different columns containing the zeolites formulated in form of extruded cylinders, or as a PRB placed in situ perpendicular to the groundwater ﬂow. In all cases, laboratory tests conﬁrmed the importance of performing the two treatments in succession to remove almost completely different contaminants (Table 16.5). It is worth noting that the efﬁciency of the adsorption in succession is in all cases much higher than those obtained working with a mixture of the two zeolites. This approach has the ﬂexibility as a great advantage. In fact, it is possible to choose the appropriate couple of zeolites, depending on the nature of contaminant(s) to be removed. For instance, when MTBE is present, the use of mordenite and ZSM-5 is preferred [131]. This ﬁnding has been conﬁrmed in two on-site tests performed in a gasoline station and in a reﬁnery [132]. In the ﬁrst case a full-scale p&t plant was operated for two years, using a column of GAC followed by a second column of mordenite in order to remove MTBE, BTEX, and C6 –C28 hydrocarbons. GAC resulted suitable for removing all the hydrocarbons and part of MTBE, while mordenite was employed for removing MTBE below the imposed threshold (10 µg/l), a level maintained for all the operation time. The demonstrative PRB used in the reﬁnery was based on ZSM-5 followed by mordenite. This system proved to be very efﬁcient in maintaining all the Table 16.5 Amount (%) of contaminant adsorbed in the laboratory tests performed with zeolite Y, ZSM-5 individually, in a mixture and in successiona.

Contaminant (ppm)

Y

ZSM-5

Mixture

Succession

Benzene (70) Toluene (60) p-Xylene (40) Chlorobenzene (40) Trichloroethylene (40) Tetrachloroethylene (80)

70 62 86 96 92 94

15 20 72 77 80 90

20 30 75 83 82 92

98 99 99.95 99.98 99.5 99.94

a Tests performed with 20 ml of water containing 5 mg (individual or mixture) or 2.5 mg (succession) of zeolite; contact time = 1 hour.

16.6 New Materials for Emerging Applications

contaminant concentrations below the target limit for the entire duration of the test. ZSM-5 efﬁciently removes BTEX (>99%), GRO’s (Gasoline Range Organics, C6 –C9 , 96%) but only part of DRO’s (Diesel Range Organics, C10 –C28 , 41%) and MTBE. The hydrocarbons and the MTBE escaped from ZSM-5 are ﬁnally adsorbed by mordenite, which maintained the corresponding concentration below the threshold limits [132]. These positive experiences demonstrate also the advantages of using zeolites. They consist of the high adsorption capacity, the inertness toward the high molecular weight substances (e.g., humic acids) and inorganic cations present in the groundwater, the stability of the structure even after a prolonged immersion in water, the possibility of regenerating their initial sorption capacity by simple thermal treatments [133]. These advantages surely compensate the higher costs of the materials with respect to the conventional ones (e.g., GAC), rendering the technologies based on zeolites appealing for future applications.

16.6 New Materials for Emerging Applications

This chapter cannot be concluded without considering the most important ‘‘ingredient’’ in the recipe of innovation in zeolite science and technology: the materials. In the common view, the availability of a large number of materials with different compositional and porous properties is fundamental for expanding the range of practical applications. On this assumption are based the efforts of several research groups (both academic and industrial), which, in the last two decades, have produced impressive accomplishments in the synthesis of porous materials. Focusing on the crystalline porous materials only, in the following, we try to draw a picture of the actual situation, highlighting the most recent achievements and perspectives in the synthesis of new materials. 16.6.1 Zeolites

The synthesis of materials with novel framework topologies is one of the most important research lines in zeolite science. The reasons of the continuous efforts in this ﬁeld are essentially linked to the increasing demand of catalysts with speciﬁc shape selectivity properties, capable of greater productivity in petrochemical and reﬁnery processes, and even materials for novel advanced applications. Today, 191 zeolite and zeotype framework types (as listed in March 2009 in the structural database of IZA), 22 families of disordered structures, and several other zeolites, either ordered or disordered, not included in the atlas, are known. That means that the today’s portfolio consists of literarily hundreds of materials with different properties deriving from the ability of the researchers to vary the Si/Al ratio in the framework, to incorporate heteroatoms (i.e., elements different from

471

472

16 Industrial Potential of Zeolites

Si and/or Al) in the framework, to synthesize the aluminosilicate analogs of AlPO structures, and so on. The progresses in zeolite synthesis essentially derive from the deeper knowledge of the phenomena occurring during the zeolite nucleation and growth, particularly with regard to the role of the organic additives [134]. In fact, the use of increasingly complex (sometime exotic) organic cations is the key factor for the crystallization of several new structures. This topic has recently been reviewed by Burton and Zones [135], who highlighted the main historical trends in the use of the organic structure directing agents (SDAs) in the synthesis of zeolites. What emerges from the analysis of the literature is that an organic additive rarely acts as a real template, since it may favor the crystallization of different zeolite structures, depending on the other synthesis conditions. The general lack of a strict correspondence between the size and shape of the SDA and those of the pores (in other words, the low to moderate selectivity characterizing the organic additives) makes difﬁcult the possibility to synthesize new materials through the rational selection of the most suitable SDA for the target porous structure. In fact, though the most relevant properties that a SDA has to possess are well known, many other parameters may inﬂuence the nature of the crystallized materials. For example, the presence and the nature of the heteroatom, which can stabilize well-deﬁned secondary building units (SBUs) and favor the formation of zeolite structures otherwise not obtainable in the classical aluminosilicate system, is a factor. It is the case of the syntheses performed in the presence of Ge, which stabilizes the double-four ring (D4R) units. This led to the crystallization of several new microporous phases, including three interesting extralarge pore zeolites: IM-12 (IZA code UTL), possessing an intersecting 14-ring and 12-ring channel system [136] and ITQ-33, characterized by intersecting 18-ring and 10-ring channel system [137] and ITQ-37 a zeolite with unprecedented 30-ring windows [138]. The synthesis of zeolites with extra-large pore systems is considered an important task, giving the possibility to perform shape-selective catalytic reactions on large organic substrates. Silica-based materials with 14 rings are known since the end of years 1990s, when the synthesis of CIT-5 (IZA code CFI) [139], and UTD-1 (IZA code for the ordered form DON) [140] have been reported. Later, Strohmaier and Vaughan reported the synthesis of ECR-34 (IZA code ETR), the ﬁrst silica-based material (precisely, an aluminogallosilicate) with 18-ring pore openings [141]. Compared with the other extra-large pore zeolites, both IM-12 and ITQ-33 are less compact, since the 14-ring and 18-ring channels intersect other systems with 12-ring and 10-ring openings, respectively. Preliminary catalytic test in the alkylation of benzene with propylene to cumene and, particularly, in the catalytic cracking of VGO indicated that ITQ-33 is a very promising candidate for acid-catalyzed reactions [138]. However, as for all the Ge-containing zeolites, one of the main factors hampering the industrial use of ITQ-33 is the high cost of Ge, a drawback that can be overcome if the Ge-free analog will be synthesized. There is another drawback limiting the potential use of extra-large pore zeolites: their intrinsic low thermal/hydrothermal stability, which decreases as the framework density (generally indicated as number of tetrahedral atom per 1000 A˚ 3 ) decreases.

16.6 New Materials for Emerging Applications

The stability of the materials can be improved, for instance, by optimizing the Si/Al composition of the framework but their use in processes involving a very high temperature treatment, such as the regeneration of catalysts used in FCC technology, could be problematic. The problem related to the synthesis of zeolite with pore opening large enough to permit the adsorption of bulky molecules and their reaction under steric control has found only a partial solution with the availability of extra-large pore zeolites. In fact, the size of the pore openings (e.g., 10.1 A˚ for ECS-34, 12.2 A˚ for ITQ-33, 4.7 × 19.7 A˚ for ITQ-37) remains in the domain of micropores and diffusion limitations on the reaction rate are still expected. 16.6.2 Hierarchical Zeolites

The discovery of ordered mesoporous materials with tunable pore size in the range 2–10 nm [142] has been considered the answer to this problem, but for several reasons (e.g., acidic strength, thermal–hydrothermal stability, costs [143]) they did not ﬁnd any application yet. Remaining in the ﬁeld of zeolites, an interesting strategy has been proposed for overcoming the limitations imposed by the sole presence micropores: the development of suitable synthesis routes for preparing zeolites with hierarchical pore architecture, in which micro- and mesopores are contemporarily present [144, 145]. The main beneﬁt expected from a hierarchical organization of the pores is the possibility to couple in the same material the advantages given by the regular micropores of the zeolite crystals and the improved mass transport efﬁciency given by the mesopores. To maximize this beneﬁt, it is necessary that the two pore systems are strictly connected in such a way that all the microporous crystals are available for the catalytic reaction. The concept of hierarchically organizing the porosity in zeolite particles is relatively new, even if catalysts with these characteristics are employed at an industrial level since the years 1980s. It is the case of the titanium-silicalite-1 already treated in Section 16.2. Another example of catalyst applied in an industrial process is the dealuminated mordenite, which is employed by DOW Chemicals in its cumene process [146]. In this case, the intracrystalline mesopores are generated by acid leaching and steaming, a process primarily employed for deep dealumination of the mordenite framework in order to increase the strength of the Brønsted acid sites and, more interesting, for connecting the mono-dimensional 12-ring channels to produce a 2D or 3D porous structure. It is worth noting that this approach mainly generates intracrystalline mesoporosity, not necessarily opened to the external surface. A deep impact on mass transport phenomena is therefore not expected, but signiﬁcant improvements of the catalytic performances of the zeolite are produced [146]. Again, the improvement of the mass transport efﬁciency can be achieved by reducing the size of the zeolite crystals. The zeolite beta catalyst employed in the Polimeri Europa’s cumene and ethylbenzene technologies, developed in the early years 1990s, constitutes a clear example [147]. Adopting a proprietary synthesis

473

474

16 Industrial Potential of Zeolites

50 nm

Figure 16.10 TEM micrograph of the zeolite beta catalyst employed in the Polimeri Europa’s cumene and ethylbenzene technologies.

route, developed in Eni’s research laboratories, it has been possible to crystallize zeolite beta in form of blackberry-like aggregates of very small crystals having dimensions in the range 10–20 nm (Figure 16.10). The presence of meso- and macropores assures an efﬁcient transport of reactant and product molecules within the aggregates and this is one of the factors responsible for the high performance of this catalyst. The concepts exempliﬁed here have recently been extended to the preparation of new materials in which the hierarchical pore structure is generated in a more rational way. This topic has been extensively reviewed by Egeblad et al. [148] and by P´erez-Ramirez et al. [149]; therefore we highlight here the most interesting and promising routes, which may lead to materials of industrial interest. According to the scheme reported in Figure 16.11, hierarchical zeolite materials can be classiﬁed into three different categories: hierarchical zeolite crystals, nanosized zeolite crystals, and supported zeolite crystals. Each of these types of materials can be prepared by different synthesis methods, which may or may not involve the use of a templating agent. As far as the nontemplating routes are concerned, the synthesis methods proposed are essentially based on the demetallation (dealumination, desililation, and detitanation) of preformed large zeolite crystals by acid or alkaline leaching and/or steaming to produce materials with an intracrystalline hierarchical pore structure (hierarchical zeolite crystals, Figure 16.11), or on the optimization of the crystallization conditions to produce aggregates of very small crystals, whose packing generates the intercrystalline mesoporosity (nanosized zeolite crystals, Figure 16.11).

16.6 New Materials for Emerging Applications

Hierarchical pore systems

Hierarchical zeolite crystals >1 µm

Intracrystalline mesopores

Nanosized zeolite crystals

Supported zeolite crystals

<50 µm

Intercrystalline mesopores

Intercrystalline mesopores

Templating Solid Supramolecular

Solid Supramolecular

Indirect Supramolecular

Nontemplating Acid leaching Alkaline leaching Steaming

Controlled crystallization

Synthesis methods Figure 16.11 Categorization of hierarchical zeolite materials and corresponding synthesis methods (adapted from [136] and [137]).

The problems inherent the demetallation route are essentially due to the difﬁculty in controlling the characteristics of the mesopores, both in amounts and dimensions. Furthermore, as already underlined for the dealuminated mordenite catalysts of DOW Chemicals, the adverse effect of generating mesopores only in the interior of the crystals, not opened on the surface, is high. The interesting topic related to the preparation of nanosized zeolite crystals has recently been reviewed by Tosheva and Valtchev [150]. There the term ‘‘nanozeolites’’ is adopted, referring to materials composed of crystals of size less than 1000 nm and particular to stable colloidal suspensions of crystal of size less than 200 nm, suitable for a variety of new, emerging applications. In general, nanocrystals are prepared under synthesis conditions that favor the nucleation rather than the growth of the crystals. These conditions vary from a zeolite to another and no general rules have been identiﬁed yet. That means that each zeolite material requires the development of a well-deﬁned synthesis route for being produced in nanocrystalline form. An original strategy for preparing hierarchical nanozeolites involves the use of organosilanes to prevent the growth of the crystals. This approach involves pre-crystallization at low temperature of the clear reactant mixture to produce seeds, whose surface is passivated by silanization prior of the

475

476

16 Industrial Potential of Zeolites

hydrothermal treatment [151]. Phenylaminopropyl-trimethoxysilane (PHAPTMS) proved to be the most efﬁcient sililating agent for the preparation of nanocrystalline ZSM-[152], Beta [153], and mordenite [154]. The average dimensions of the seeds and, consequently, of the nanozeolites crystals as well as the hierarchical pore structures of the materials can be efﬁciently controlled by changing the precrystallization temperature and the concentration of PHAPTMS [155]. In all cases, the hierarchical nanocrystalline zeolite materials showed enhanced catalytic performance in the cracking of polyethylene with respect to conventional zeolites. Nevertheless, the availability of such kind of materials (particularly those with the smaller crystal size) may enable new applications both in traditional and in emerging areas [151–157]. Modiﬁcation of chemical sensors, development of new optical systems, ﬁlms with low dielectric constants (low-k ﬁlms), medical diagnostics are just some of the emerging areas in which nanozeolites are expected to be useful for the development of new devices and technologies. On the contrary, as heterogeneous catalysts, nanozeolites provide the opportunity to catalyze reactions involving large molecules because of the larger external speciﬁc surface where active sites are located. Certainly, this is an advantage, which partially compensates the loss of shape selectivity properties unless the active sites are located at (or close to) the pore mouths where some special shape selective phenomena (e.g., the nest effect, [157]) may take place. On the other hands, even for reactions involving small molecules, nanozeolites present signiﬁcant advantages arising from the shorter residence time of reactant and product molecules in the zeolite channels. In this way, the consecutive reactions are less probable, leading to a reduction of the deactivating effects due to the formation of heavy molecular species in the pores. At the same time, the enhanced mass transport efﬁciency provided by the intercrystalline mesoporosity ensures a more rapid moving away of the products and, consequently, a less pronounced deposition of coke on the external surface of the crystals [147, 158]. Other approaches to the synthesis of hierarchical pore systems are based on the use of ‘‘secondary’’ templates. Interesting is the synthesis route developed by the researchers from Haldor Topsøe for synthesizing hierarchical zeolite crystals [44, 159, 160, 161]. This route involves the impregnation of a carbon source with the reactant mixture, which is successively subjected to a suitable hydrothermal treatment to grow large zeolite crystals incorporating the carbon particles. In the ﬁnal calcination step, both the SDA molecules and the carbon articles are burned off, leaving zeolite crystals with an intra-crystalline micro-meso-porous hierarchical pore structure (Figure 16.12). The ﬁrst example concerns the preparation of ZSM-5 [159] and TS-1 [44] large crystals with intracrystalline mesoporosity produced by using carbon black as a templating agent. These materials are similar to those prepared by acid leaching (i.e., mesopores conﬁned in the interior of the crystals, mainly not exposed to the exterior), unless that the porosity can be better controlled by using the desired amounts of templating agent with well-deﬁned particle size. Connectivity, size, and shape of the mesoporosity can be ﬁne-tuned by changing carbon source and by selecting appropriate synthesis conditions. For instance, when carbon nanotubes or nanoﬁbers are used, the mesopores are mainly opened to the

16.6 New Materials for Emerging Applications

Calc. (air)

Carbon black

Mesopore

Figure 16.12 Schematic representation of the method for generating intracrystalline mesoporosity by using carbon black. The same concept can be applied when other carbon sources (nanotubes, nanoﬁbers) are employed.

exterior, improving the mass transport efﬁciency within the large zeolite crystals [161–163]. Furthermore, it is possible to crystallize hierarchical zeolite crystals or nanosized zeolite crystals, depending on the synthesis conditions favoring the crystal growth or the nucleation are respectively adopted [160]. This route seems to be versatile and of general applicability, considering that it has been successfully applied to prepare ZSM-5 [159, 164, 165], TS-1 [44], Beta [165], ZSM-11 [166], and ZSM-12 [167]. When tested in some relevant reactions, these materials showed also enhanced catalytic performances in some relevant reactions, such as the isomerization of n-hexane and n-heptane [168], the epoxidation of 1-octene and cyclohexene [169], the alkylation of benzene with ethylene [170]. It should be underlined, however, that the comparison was mainly made with zeolite catalysts constituted by large crystals with dimensions close to those of the hierarchical porous samples and not with the state-of-art catalysts. For instance, Christensen et al. compared the catalytic performances of hierarchical zeolite crystals of ZSM-5 prepared by using carbon black with those of ZSM-5 prepared in a conventional way [170]. Both samples were constituted by cofﬁn-shaped particles with very close average dimensions (circa 2 µm) but in one case consisting of single crystals (conventional zeolite), and in the other more similar to blackberry-like agglomerates of small crystals (hierarchical zeolite crystals). Therefore a signiﬁcant difference in the mass transport efﬁciency within the hierarchical zeolite crystals and a more rapid diffusion of the molecules away from the particles with respect to the large crystals are expected. A question arises: What are the real advantages with respect to nanosized ZSM-5? In principle, one should not expect any clear difference in the catalytic performance. On the contrary, advantages can be found in the preparation of the zeolite, in particular in the ﬁltration step, which is expected to be difﬁcult in the case of colloidal dispersions of the zeolite particles in the mother liquor.

477

478

16 Industrial Potential of Zeolites

Other synthesis methodologies, which deserve increasing attention, are those involving mesoporous molecular sieves, such as the M41S family [142], in which mesopores of regular dimensions orderly arranged in the whole particle are present, while the walls around them are constituted by an amorphous silica-based material. Just after the discovery of these materials, their great potentialities have been realized not only in heterogeneous catalysis but also as supports for anchoring active phases or intermediates for the preparation of hierarchical porous materials. For instance, interesting synthesis routes are those categorized as indirect templating methods (Figure 16.11), in which the preformed mesoporous material is transformed (at least partially) in a hierarchical zeolite materials by partial crystallization of the amorphous walls or by controlled deposition of small zeolite crystals (e.g., seeds) on the surface of the walls. In general, the attempts to crystallize the amorphous-ordered mesoporous materials did not lead to encouraging results and the reason is the small thickness of the pore walls. Kloetstra et al. succeeded in transforming the pore walls of MCM-41 in proto-zeolitic ZSM-5 structures by crystallizing the TPA+ -exchanged mesoporous material in glycerol, but their formation was postulated based on the observation of the 550 cm−1 IR band and not by XRD analysis [171]. On the other hand, growing even small zeolite crystals into the amorphous pore walls necessarily implies strong modiﬁcations of the mesoporous framework as in the case of MCM-41/MFI [172] and SBA-15/MFI [173, 174], at worst its complete collapse [175]. Another potential and not yet fully explored synthesis route concerns the deposition of zeolite crystals on preformed mesoporous materials. Only a few examples exist mainly consisting in the impregnation of mesoporous molecular sieves such as SBA-15[176, 177] and MCM-41[178], or meso-structured cellular foams [179] with zeolite seeds followed by hydrothermal treatment. The syntheses seem successful but the effective advantage of such kind of materials is still an open question. Approaches based on supramolecular templating mechanism are attracting increasing attention (Figure 16.11). The idea is to create mesoporosity (either intra- or inter-crystalline) by adding the supramolecular templating agent (i.e., the surfactant) contemporarily with the molecular one in such a way that the crystallization of the zeolite occurs contemporarily with the formation of the mesopores. Alternatively, the supramolecular templating agent is added to a preformed solution of zeolite seeds, which are assembled around the surfactant micelles to produce the hierarchical porous structure (see [145] and [148] and references herein). Certainly, these approaches are much more effective than the indirect ones previously illustrated and that is due to the different role played by the molecular templating agent and by the surfactant: the former drives the crystallization of zeolite in the form of seeds or nanocrystals, and the latter favors their organization around the micelles. Among the different methodologies, it is worth mentioning the methodology proposed by the research groups of Ryoo (KAIST, Danjeon, Korea) because it represents a very nice solution for tuning the dimensions of the mesopores, avoiding, at the same time, the limits imposed by the use of more conventional surfactants,

16.6 New Materials for Emerging Applications

which undergo decomposition at the temperatures required for the crystallization of zeolites. In particular, they use, as supramolecular templating agent, amphiphilic organosilanes of general formula [(CH3 O)3 Si(CH2 )3 N(CH3 )2 Cn H2n+1 ]Cl (n = 12,16,18) [180]. In this way, they are able to crystallize MFI and LTA [180] and even AlPO and CoAPO (AFI, AEL) [181] structures with controlled mesoporosity. When used as catalysts, hierarchical MFI-type zeolites proved to have higher catalytic activity [182, 183] and enhanced stability toward deactivation [184] with respect to analogous structures synthesized in a conventional manner. Moreover, the mesopores walls display a high concentration of silanols, providing anchoring sites for functionalization with various organosilanes [185]. Another supramolecular templating approach to hierarchical porous zeolites has recently been claimed by MIT [186, 187]. It consists in suspending preformed zeolite crystals of Y, ZSM-5 or Mordenite in a basic solution of NH4OH or TMA-OH, adding the required amount of a surfactant (e.g., cetyltrimethylammonium bromide) and hydrothermally treating the resulting mixture for the desired time at 150 ◦ C. The resulting hierarchical porous zeolite crystals show contemporarily the features of the zeolite crystals and of the ordered mesoporous material (e.g., MCM-41). These materials are claimed to possess superior performances as catalysts for the cracking of organic compounds and in the degradation of polymers [186, 187]. On these materials is probably based the announcement recently appeared in MIT Technology Insider of the intention of a small company, Rive Technology, in developing new and improved catalysts for FCC process. In the case of success, it would be the ﬁrst industrial application of a hierarchical porous zeolite prepared by the supramolecular templating approach.

16.6.3 Silica-Based Crystalline Organic–Inorganic Hybrid Materials

What described above summarizes the main trends in the synthesis of zeolites with new and/or improved characteristics with respect to the known materials. There is, however, another interesting topic, which is attracting the interest of the researchers because of the enormous potentialities in both consolidated and emerging areas of applications: the preparation of silica-based crystalline organic-inorganic hybrid materials. The incorporation of organic groups in a microporous framework is seen as a new opportunity for preparing materials with enhanced lipophilic/hydrophobic character of the internal surface of the pores, useful for adsorption and catalytic processes. The lower afﬁnity of these porous materials toward polar molecules, in fact, would provide a more efﬁcient control of the host–guest interactions between the framework and the sorbed organic molecules, facilitating the elution of polar reaction products before their further interaction with the surface. Moreover, through the same approach, it would be possible to incorporate new chemical functions acting as active sites for novel catalytic processes.

479

480

16 Industrial Potential of Zeolites

Three different approaches are followed for synthesizing silica-based crystalline porous materials, all making use of organosilanes: • functionalization with monovalent pendant organic groups • incorporation of organic groups in the framework by using bridged organosilanes as source of Si • the direct synthesis of new, nonzeolitic microporous materials by using complex bridged organosilanes as source of Si. The functionalization of zeolites with monovalent pendant organic groups was ﬁrst approached via postsynthesis treatment of preformed zeolites. This well-known process applied to amorphous silica and periodic mesoporous materials for chromatographic and catalytic purposes proved, however, to be inefﬁcient when performed on zeolites, since the functionalization mainly occurs in the mesoporous regions of the crystals [188–191]. More promising is the direct synthesis route in which the organosilanes is added to the reactant mixture. The group of Davis was the ﬁrst to demonstrate the possibility to synthesize directly what they called silica-based organic functionalized molecular sieves (OFMSs) [192–195]. This approach has, however, a strong limitation, that is, it should be avoided any high-temperature thermal treatment of the as-synthesized zeolites, in order to keep the organic function in the ﬁnal products. Therefore, the candidate zeolite structure should be selected among those which do not require a SDA (e.g., zeolite X, Y, A) or from which the SDA can be removed by extraction. It is the case of pure silica zeolite Beta, whose synthesis procedure was slightly modiﬁed by adding a small fraction of phenylethyltrimethoxysilane (PETMS) to the reaction mixture [192]. The as-synthesized products, containing up to 5 atom% of silane, were ﬁrstly treated with concentrated basic solution to remove the organic groups present on the external surface of the crystals, while successive repeated treatments with acetic acid/water solution at 140 ◦ C were employed for removing the SDA (tetraethylammonium ﬂuoride). Finally the phenyl groups were sulfonated with SO2 vapor at room temperature to produce a catalyst active in the reaction of ethylene glycol with cyclohexanone topentamethylene-1,3-dioxolane, but not with the bulky 1-pyrenecarboxaldehyde [192]. Other OFMSs with polar (e.g., aminopropyltrimethoxysilane, mercaptopropyltrimethoxysilane [193]) or nonpolar (e.g., 3-butenyltriethoxysilane, 2-(3-cyclohexenylethyl)trimethoxysilane [194]) organic groups were successfully synthesized, demonstrating that this approach may effectively lead to the preparation of new shape-selective catalysts. However, no other communications followed the original papers for the simple reason that zeolites with extractable SDAs are very uncommon and therefore there is a lack of candidates for such kind of functionalization. The next logical step is the incorporation of bivalent organic groups in the framework, the simplest being the substitution of the framework oxygen atoms (-O-) by methylene (–CH2 –) groups. Yamamoto et al. demonstrated that the substitution is possible, reporting a new class of organic–inorganic hybrid materials named ZOL (zeolites with organic groups as lattice) comprising some known zeolite structures (MFI, LTA, and Beta) [196, 197]. An immediate effect of the presence of

16.6 New Materials for Emerging Applications

the disilane is the dramatic increase of the crystallization time (e.g., 14 days in the case of ZOL-A instead of one required by LTA in the same conditions) while the C content is signiﬁcantly lower than that theoretically expected (2.2 and 4.3 wt%, respectively). D´ıaz et al. reported the possibility to crystallize hybrid organo-zeolites with the ITQ-21, MFI, and Beta structures, with an unprecedented high organic content [198]. As source of silica, these authors used a mixture of a disilane [bis(triethoxysilyl)methane (BTEM) or 1,2-bis(triethoxysilyl)ethane (BTEE)] and tetraethylortosilicate (TEOS) in different molar ratios and the syntheses were carried out by using the typical SDAs required for the crystallization of the three zeolites and in the presence of HF. Only ZSM-5 and Beta were obtained preferentially using BTEM, while ITQ-21 was never observed, the products being invariably amorphous. The most critical point of the work concerns the effective incorporation of organic groups in the framework. In fact, if the data related to fully crystalline ZSM-5 and Beta (C content of 3.6 and 4.9 wt%, respectively) are reasonable, less convincing is the high C content (8–9 wt%) detected in some Beta samples, because they are referred to materials with a modest degree of crystallinity (35–50%). Doubts about the possibility to incorporate the methylene groups in zeolite framework were recently admitted by other authors [199], who reported the synthesis of ZOL-X, a hybrid zeolite with the FAU structure-type. The C content in this materials was found relatively low (1.3–1.5 wt%) and the presence of Si–C and –CH2 – moieties was conﬁrmed spectroscopically. However, these authors admitted the impossibility to unambiguously ascertain C incorporation into the framework, because they cannot exclude the presence of amorphous impurities in the materials in which the –CH2 – groups may be concentrated. This is probably the main concern, which has not yet found a deﬁnite answer. Even DFT calculations did not contribute to solve this problem. In fact, the substitution of a framework oxygen by a methylene group is possible [200, 201], but the local distortion imposed to the framework (due to the smaller value of the Si–C–Si angle (120–125 ◦ ) with respect to Si–O–Si (140–170 ◦ ), implies a small but signiﬁcant strain contribution to the defect energy of the system, estimated 0.2 eV maximum at signiﬁcant substitution level [200]. In summary, what emerges from the above reported papers is that the incorporation of methylene groups (and, consequently, of larger organic moieties) in the zeolite framework is not easy and not deﬁnitely proved yet. Let us consider some evidences: • When disilane is used as the sole silica source, the crystallization rate becomes very low, compared with what was observed with conventional sources (e.g., TEOS). • Crystallization does not take place in nearly neutral reaction conditions. • All the crystalline phases reported so far display a C content signiﬁcantly lower than the expected theoretical value.

481

482

16 Industrial Potential of Zeolites

These facts indicate that the crystallization occurs only when [SiO4 ] groups are available because added as a second (conventional) silica source or produced via the hydrolysis of the disilane, a very slow reaction that requires strong basic media. The structural and spectroscopic data reported still maintain a certain ambiguity, essentially due to the possible (or certain) presence of even small amounts of amorphous phases in which the organic moieties can be located. Certainly, a detailed structural characterization on well-crystallized materials may provide a deﬁnite answer to this problem. The uncertainties related to the incorporation of organic moieties into the zeolite framework do not hamper the possibility to prepare non-zeolitic crystalline porous materials, as recently demonstrated by the research group operating in the Eni’s laboratories, which was able to synthesize a new class of crystalline porous hybrid materials named Eni Carbon Silicate (ECS) [202]. The idea was developed starting from the fundamental work of Inagaki et al., who reported the preparation of a periodic mesoporous organosilica (PMO), obtained by using 1,4-bis(triethoxysilyl)benzene (BTEB), octadecyltrimethylammonium chloride as a surfactant, NaOH, and water [203]. This material displays a hexagonal array of mesopores and as such can be considered as a hybrid organic–inorganic analog of MCM-41. Differently from this material, characterized by completely amorphous pore walls, PMO exhibits a structural periodicity of 7.6 A˚ along the channel direction, a spacing correspondent to the dimensions of the [O3 Si-C6 H4 -SiO3 ] moiety. Following these results, the role of the surfactant and of the disilane was ﬁrstly examined. In the absence of the surfactant, a mesoporous material named ECS-4 was obtained: its XRD pattern display the 7.6 A˚ periodicity observed in PMO, but no reﬂections in the low-angle region, indicating the lack of ordering of the mesopores. No truly crystalline materials were obtained when a conventional silica source (TEOS) was added to the synthesis. On the contrary, unexpected results were obtained by adding NaAlO2 as aluminum source: three different crystalline phases (ECS-1, ECS-2, and ECS-3) were obtained by varying the Si/Al molar ratio and by replacing NaOH by KOH. Other phases were successively crystallized by using different disilanes: ECS-5 with 4,4 -bis(triethoxysilyl)biphenyl (BETBP), ECS-6 with 1,4-bis(triethoxysilylethyl) benzene (BTEEB), ECS-7 with 1,3-bis(trimethoxysilyl)propane (BTMP). All these materials are constituted by a crystalline phase, only in some cases accompanied by an amorphous component or traces of Sodalite, and all the spectroscopic and analytical evidences indicated the presence of the disilane moieties as well as of the aluminum in tetrahedral coordination. However, the deﬁnite proof of the nature of these materials was achieved when the crystal structure of two of them was solved (ECS-2) or postulated (ECS-7). In the case of ECS-2, the structure consists of aluminosilicate layers held together by the phenylene groups (Figure 16.13). The layers are constituted by [AlO4 ] tetrahedra bonded via O atoms to four different [O3 SiC] tetrahedra; it is worth noting that the structure does not contain any [SiO4 ] units, indicating that the disilane does not undergo signiﬁcant hydrolysis during the hydrothermal treatment. The small fraction of hydrolyzed disilane contributes to the formation of sodalite, contained in trace amounts.

16.6 New Materials for Emerging Applications

(a)

(b)

Figure 16.13 Crystal structure of ECS-2 (a) and the cage deﬁned by six phenylene rings (b) (EtOH molecules, H atoms, and Na+ ions omitted for clarity).

ECS-2 has no open porosity, but only cages surrounded by six phenylene rings, arranged in such a way to hamper the access of any even small molecule. In these cages are located ethanol molecules produced by the hydrolysis of the triethoxy groups; on the contrary, the Na ions are located within the organic layers. For the other ECS materials, the crystal structure has not been clariﬁed so far; it is, however, clear from the HRTEM analysis that all the materials display an alternating stacking of organic and inorganic layers with interlayer distance increasing with the size of the organic moiety. However, the hypothesis that the structure of the other materials is formed by the same inorganic layers composing ECS-2 held together by the different organic moieties has been conﬁrmed for ECS-7 only, where the trimethylene groups replace the phenylene ones. The case of ECS-5 is particularly interesting, because it is composed by much complex and thicker inorganic layers, whose formation requires the presence of [SiO4 ] units deriving from the complete hydrolysis of the disilane precursors. The synthesis of ECS materials is just at the beginning and thanks to the increasing availability of di- and polysilanes precursors, further interesting developments are easily predictable. For the moment, these results conﬁrm the possibility of synthesizing crystalline silica-based organic–inorganic hybrid materials but extensive work is in progress for verifying their properties in view of their application in conventional and emerging ﬁelds.

483

484

16 Industrial Potential of Zeolites

16.7 Summary and Outlook

Today, zeolites constitute one of the most important class of materials used in several relevant industrial applications as heterogeneous catalysts, ion-exchangers, molecular sieves, and adsorbents. Apparently, this research ﬁeld seems to be mature, without any signiﬁcant room for further developments. In reality, it is not true because the worldwide research is continuously providing new materials and technological solutions, which can be suitably exploited for improving existing processes and even for developing new applications. Environmental, economical, and market issues are pushing the innovation in the reﬁnery, petrochemical, and chemical industries and it is undoubted that in the last years signiﬁcant advancements have been achieved in all these ﬁelds. The knowledge and skills in the synthesis of zeolites rapidly increase, leading to the discovery of new, increasingly complex framework structures. For instance, the synthesis of extra-large pore zeolites, until a decade ago considered a very difﬁcult task, today represents one of the most fruitful research lines, which recently led to the crystallization of ITQ-37, the ﬁrst crystalline mesoporous material with pores with 30-ring openings. Considering the deﬁnition of new material as referred to a new crystalline structure is too limiting, because the importance and often the difﬁculties of modifying a basic material in terms of composition and morphology are well known. The case of TS-1 is illustrative of this concept. First, the successful incorporation of Ti in the Silicalite-1 framework is a clear and proved example of how the modiﬁcation of the composition of a zeolite can lead to a material with unprecedented catalytic properties. Secondly, the optimization of the crystallite morphology and the ability to aggregate these crystallites in particles with a hierarchical porous structure opened the possibility to develop catalytic processes in slurry phase. The synthesis of hierarchical porous structures constitutes one of the hot topics in today’s research, because they represent not only a technological solution of the problems often encountered during the industrial preparation of zeolites but even a method for limiting the otherwise severe limitation imposed by the intercrystalline diffusion of reactants and products. The very recent application of this concept to the FCC catalysts, if successfully applied at an industrial level, will represent a signiﬁcant breakthrough. But considering zeolites as heterogeneous catalysts only is somewhat limiting. Other important technological areas, in fact, are taking advantage of the peculiar properties of these materials and of the advancements of zeolite science. It is the case of the gas separation area, where technological solutions are needed for increasing the capabilities and the efﬁciency, for instance, of the natural gas puriﬁcation from N2 , CO2 , and particularly H2 S. The availability of an efﬁcient gas separation technology, based on the use of molecular sieves, would replace the otherwise expensive cryogenic and chemical processes now employed, opening up the possibility to exploit natural gas sources economically unfavorable because of the high content of gases different from CH4 .

References

Again, zeolites can be exploited also for the environmental protection, in particular, in groundwater remediation from organic pollutants. The cost of the zeolite-based systems has limited their practical application in favor of less expensive materials such as GAC or ZVI. Recent on-site tests performed by Eni, in reality, have demonstrated the advantage of using zeolites for the remediation of groundwater polluted by hydrocarbons and, particularly, MTBE, highlighting their ability to keep for a long time the concentration of these pollutants well below the threshold imposed by the legislation. Therefore, also in this ﬁeld, zeolites are demonstrating their effectiveness and interesting developments are expected in a near future. The application of zeolites in conventional and emerging areas requires, however, a deep change of their characteristics and properties. An answer to this need may be given by the modiﬁcation of the inorganic framework by the incorporation of organic groups to produce silica-based crystalline organic–inorganic hybrid materials. In the last decade, several and sometimes successful attempts of grafting or incorporating organic groups onto or in the zeolite framework have been reported but, until now, the advantages of such modiﬁcation have not been demonstrated. The recent report on ECS has unambiguously demonstrated that it is possible to synthesize crystalline silica-based hybrid materials, which, in spite of their non-zeolitic nature, are porous and can be prepared with a variety of different organic groups. Though the research on these materials is just at the beginning and the knowledge on their properties is still insufﬁcient for deﬁning possible applications, the preparation of ECS demonstrates that there are still large unexplored areas in materials science to which the researchers can address their efforts. References 1. Marcilly, C. (2001) Oil Gas Sci. Tech2. 3. 4. 5.

6. 7. 8.

nol.: Rev. IFP, 56 (5), 499–514. Bellussi, G. (2004) Stud. Surf. Sci. Catal., 154, 53–65. Bellussi, G. and Pollesel, P. (2005) Stud. Surf. Sci. Catal., 158, 1201–1212. Degnan, T.F. Jr. (2007) Stud. Surf. Sci. Catal., 170, 54–65. Perego, C. and Carati, A. (2008) Zeolites and zeolite-like materials in industrial catalysis, in Zeolites: from Model Materials to Industrial Catalˇ ysis (eds J. Cejka, J. Perez-Pariente, and W.J. Roth), Transworld Research Network, India, pp. 357–395. Corma, A. (2003) J. Catal., 216, 298–312. Petti, N., Hunt, L., and Yaluris, G. (2006) Petrol. Tech. Q., 11, 43–54. Xu, M. and Madon, R.J. (2005) Petrol. Tech. Q., Q2, 63–78.

9. Sheldon, R.A. (1997) J. Chem. Technol.

Biotechnol., 68, 381. 10. Taramasso, M., Perego, G., and Notari,

11.

12.

13.

14.

B. (1983) US Patent 4,410,501, assigned to Eni S.p.A. Perego, G., Millini, R., and Bellussi, G. (1998) Synthesis and characterization of molecular sieves containing transition metals in the framework, in Molecular Sieves. Science and Technology, Synthesis, Vol. 1 (eds H.G. Karge and J. Weitkamp), Springer-Verlag, Berlin, pp. 188–228. van der Pol, A.J.P.H. and van Hooff, J.H.C. (1992) Appl. Catal. A: Chem., 92, 113. Clerici, M.G., Bellussi, G., and Romano, U. (1991) J. Catal., 129, 159–167. Carati, A., Berti, D., Gagliardi, F., Stocchi, B., Bellussi, G., Mantegazza,

485

486

16 Industrial Potential of Zeolites

15.

16.

17.

18.

19.

20.

21.

22.

23.

24. 25.

26.

27.

28.

29. 30.

M.A., and Rivetti, F. (2005) 3rd FEZA Conference, Book of Abstracts, PO 211. Buonomo, F., Bellussi, G., and Notari, B. (1983) Italian Patent 20115 MI A83, assigned to Eni S.p.A. Bellussi, G., Buonomo, F., Esposito, A., Clerici, M.G., Romano, U., and Notari, B. (1983) Italian Patent 20457 MI A85, assigned to Eni S.p.A. Bellussi, G., Buonomo, F., Esposito, A., Clerici, M.G., and Romano, U. (1986) Italian Patent 22075 MI A86, assigned to Eni S.p.A. Bellussi, G., Perego, C., Carati, A., Peratello, S., Previde Massara, E., and Perego, G. (1994) Stud. Surf. Sci. Catal., 84A, 85–92. Rizzo, C., Carati, A., Barabino, C., Perego, C., and Bellussi, G. (2001) Stud. Surf. Sci. Catal., 140, 401. Esposito, A., Neri, C., and Buonomo, F. (1982) Italian Patent 20262 MI A82, assigned to Eni S.p.A. Rofﬁa, P., Padovan, M., Moretti, E., and De Alberti, G. (1985) Italian Patent 21511 MI A85, assigned to Eni S.p.A. Rofﬁa, P., Leofanti, G., Cesana, A., Mantegazza, M., Padovan, M., Petrini, G., Tonti, S., and Gervasutti, P. (1990) Chim. Ind. (Milan), 72, 598–607. Neri, C., Anfossi, B., Esposito, A., and Buonomo, F. (1982) Italian Patent 22608 MI A82, assigned to Eni S.p.A. Clerici, M.G. and Ingallina, P. (1993) J. Catal., 140, 71–83. Esposito, A., Neri, C., and Buonomo, F. (1984) US Patent 4,480,135, assigned to Eni. S.p.A. Clerici, M.G. and Romano, U. (1987) European Patent 230,949, assigned to Eni S.p.A. Esposito, A., Taramasso, M., and Neri, C. (1983) US Patent 4,396,783, assigned to ANIC S.p.A. Rofﬁa, P., Padovan, M., Moretti, E., and de Alberti, G. (1987) European Patent 208,311, assigned to Montedipe S.p.A. Gontier, S. and Tuel, A. (1994) Appl. Catal. A: Gen., 118, 173–186. Reddy, R.S., Kumar, R., and Ratnasamy, P. (1990) Appl. Catal. A: Gen., 58, L1–L4.

31. Sasidharan, M., Suresh, S., and

32. 33.

34.

35.

36.

37.

38.

39.

40. 41.

42.

43. 44.

45. 46.

47.

48.

Sudalai, A. (1995) Tetrahedron Lett., 36, 9071–9072. Notari, B. (1996) Adv. Catal., 41, 253–334. Bianchi, D., D’Aloisio, R., and Tassinari, R. (2003) WO Patent 3,042,146, assigned to Polimeri Europa. Naqvi, S. PEP Review 2005-05 Lyondell Direct-Oxidation Propylene Oxide Technology, October 2005. Kustov, L.M., Bodgan, V., and Kazansky, V. (2002) US Patent 6,414,197, assigned to General Electric. Kustov, L.M., Tarasov, A.L., Bodgan, V.I., Tyrlov, A.A., and Fulmer, J.W. (2000) Catal. Today, 61, 123–128. Camblor, M.A., Corma, A., and P´erez -Pariente, J. (1992) J. Chem. Soc. Chem. Commun., 589–590. Valencia, S., Camblor, M.A., and Corma, A. (1997) WO Patent 97/33830, assigned to CSIC. Van der Waal, J.C., Lin, P., Rigutto, M.S., and Van Bekkum, H. (1997) Stud. Surf. Sci. Catal., 105, 1093–1100. Tuel, A. (1995) Zeolites, 15, 236–242. D´ıaz-Caba˜ nas, M.J., Villaescusa, L.A., and Camblor, M.A. (2001) Chem. Commun., 761–762. Corma, A., D´ıaz-Caba˜ nas, M.J., Domine, M.E., and Rey, F. (2000) Chem. Commun., 1725–1726. Sasidharan, M., Wu, P., and Tatsumi, T. (2002) J. Catal., 205, 332–338. Schmidt, I., Krogh, A., Wienberg, K., Carlsson, A., Brorson, M., and Jacobsen, C.J.H. (2000) Chem. Commun., 2157–2158. Venuto, P.B. (1997) Stud. Surf. Sci. Catal., 105, 811–852. Whitman, P.J., Miller, J.F., Speidel, J.H. Jr., and Cochran, R.N. (2006) US Patent 7,138,535, assigned to Lyondell Chemical Technology Limited. Chen, C.-Y., Burton, A.W. Jr., and Liang, A.J. (2006) US Patent 7,083,766, assigned to Chevron USA Inc. Zones, S.I. and Burton, A.W. Jr. (2006) US Patent 7,084,305, assigned to Chevron USA Inc.

References 49. Chen, C.-Y., Burton, A.W. Jr., and

50.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

Liang, A.J. (2006) US Patent 7,087,792, assigned to Chevron USA Inc. Chen, C.-Y., Burton, A.W. Jr., and Liang, A.J. (2006) US Patent 7,091,385, assigned to Chevron USA Inc. M¨uller, U., Ma, L., Feng-Shou, X., and Yang, X. (2007) US Patent 7,211,239, assigned to BASF AG. Strebelle, M. and Catinat, J.-P. (2008) US Patent 7,320,779, assigned to Solvay. Eimer, G.A., D´ıaz, I., Sastre, E., Casuscelli, S.G., Crivello, M.E., Herrero, E.R., and Perez-Pariente, J. (2008) Appl. Catal. A: Gen., 343, 77–86. Mazaj, M., Zabukovec Logar, N., Mali, G., Novak Tuˇsar, N., Arˇcon, I., Risti´c, A., Reˇcnik, A., and Kauˇciˇc, V. (2007) Microporous Mesoporous Mater., 99, 3–13. Mazaj, M., Stevens, W.J.J., Zabukovec Logar, N., Risti´c, A., Novak Tuˇsar, N., Arˇcon, I., Daneu, N., Meynen, V., Cool, P., Vansant, E.F., and Kauˇciˇc, V. (2009) Microporous Mesoporous Mater., 117, 458–465. Fan, W., Wu, P., Namba, S., and Tatsumi, T. (2006) J. Catal., 243, 183–191. Song, F., Liu, Y., Wang, L., Zhang, H., He, M., and Wu, P. (2007) Appl. Catal. A: Gen., 327, 22–31. Song, F., Liu, Y., Wu, H., He, M., Wu, P., and Tatsumi, T. (2006) J. Catal., 237, 359–367. Wu, P., Nuntasri, D., Liu, Y., Wu, H., Jiang, Y., Fan, W., He, M., and Tatsumi, T. (2006) Catal. Today, 117, 199–205. Wang, L., Liu, Y., Xie, W., Zhang, H., Wu, H., Jiang, Y., He, M., and Wu, P. (2007) J. Catal., 246, 205–214. Moliner, M., Serna, P., Cant´ın, A., ˇ Sastre, G., D´ıaz-Cabanas, M.J., and Corma, A. (2008) J. Phys. Chem. C, 112, 19547–19554. de Jong, K.P., Mesters, C.M.A.M., Peferoen, D.G.R., van Brugge, P.T.M., and de Groot, C. (1996) Chem. Eng. Sci., 51 (10), 2053–2060.

63. Botella, P., Corma, A., L´ıpez-Nieto,

64.

65. 66.

67. 68.

69.

70.

71.

72. 73. 74. 75.

76.

77.

78.

79.

J.M., Valencia, S., and Jacquot, R. (2000) J. Catal., 195, 161–168. Kantam, M.L., Ranganath, K.V.S., Sateesh, M., Kumar, K.B.S., and Choudary, B.M. (2005) J. Mol. Catal. A: Chem., 225, 15–20. Selvin, R., Hsu, H.-L., and Her, T.-M. (2008) Catal. Commun., 10, 169–172. Beers, A.E.W., Nijhuis, T.A., Kapteijn, F., and Moulijn, J.A. (2001) Microporous Mesoporous Mater., 48, 279–284. Johnston, D. (1996) Oil Gas J., 49. CONCAWE Report No. 8/08, p. 8. http://www.concawe.org/Content/ Default.asp?PageID=31. Zhao, X., Cheng, W.C., and Rudesill, J.A. (2002) NPRA Annual Meeting, NPRA Ed. Paper AM-92-60, March 17–19, San Antonio. Spry, J.C. and Sawyer, W.H. (1975) AIChE 68th Annual Meeting, Paper 16-20, Los Angeles. McAfee, A.Mc.D. (1922) US Patent 1,405,054, assigned to Gulf Reﬁning Co. Houdry, E. (1934) US Patent 1,957,648, assigned to Houdry Process Co. Houdry, E. (1937) US Patent 2,078,945, assigned to Houdry Process Co. Houdry, E. (1945) US Patent 2,387,267, assigned to Houdry Process Co. Campbell, D.L., Martin, H.Z., Murphree, E.V., and Summit, C.W.T. (1948) US Patent 2,451,804, assigned to Standard Oil Dev. Co. Plank, C.J. and Rosinski, E.J. (1964) US Patent 3,140,249, assigned to Socony Mobil Oil Co. Breck, D.W. (1964) US Patent 3,130,007, assigned to Union Carbide Co. Corma, A. (1992) in Zeolite Microporous Solids: Synthesis, Structure and Reactivity, NATO ASI Series, Vol. 352 (eds E.G. Derouane, F. Lemos, C. Naccache, and F. Ramoa Ribeiro), Kluwer Academic Publishers, pp. 373–436. O’Connor, P. and Humphries, A.P. Accessibility of functional sites in FCC (1993) Prepr.– Am. Chem. Soc. Div. Pet. Chem., 38, 598–603.

487

488

16 Industrial Potential of Zeolites 80. Corma, A., Diaz-Cabanas, M.J.,

81.

82.

83.

84.

85.

86.

87.

88.

89. 90.

91. 92.

93.

94.

Martinez-Friguero, J., Rey, F., and Ruiz, J. (2002) Nature, 418, 514–517. McVicker, G.B., Daage, M., Touvelle, M.S., Hudson, W., Klein, D.P., Baird, W.C. Jr., Cook, B.R., Chen, J.C., Hantzer, S., Vaughan, D.E.W., Ellis, E.S., and Feeley, O.C. (2002) J. Catal., 210, 137–148. Du, H., Fairbridge, C., Yang, H., and Ring, Z. (2005) Appl. Catal. A: Gen., 294, 1–21. Santikunaporn, M., Herrera, J.E., Jongpatiwut, S., Resasco, D.E., Alvarez, W.E., and Sughrue, E.L. (2004) J. Catal., 228, 100–113. Arribas, M.A., Corma, A., Diaz-Cabanas, M.J., and Martinez, A. (2004) Appl. Catal. A: Gen., 273, 277–286. Hertl, G., Kn¨ozinger, H., Sch¨uth, F., and Weitkamp, J. (eds) (2008) Handbook of Heterogeneous Catalysis, 2nd edn, vol. 7. Wiley-VCH Verlag GmbH, Weinheim, p. 3133. Calemma, V., Carati, A., Flego, C., Giardino, R., Gagliardi, F., Millini, R., and Bellussi, G. (2008) ChemSusChem, 1, 548–557. Murzin, D.Y., Kubicka, D., Simakova, I.L., Kumar, N., Lazuen, A., M¨aki-Arvela, P., Tiitta, M., and Salmi, T. (2009) Petrol. Chem., 49, 90–93. Kumar, N., Kubicka, D., Garay, A.L., M¨aki-Arvela, P., Heikkil¨a, T., Salmi, T., and Murzin, D.Y. (2009) Top. Catal., 52, 380–386. ˇ KubiSka, D., Salmi, T., Tiitta, M., and Murzin, D.Y. (2009) Fuel, 88, 366–373. Shorey, S.W., Lomas, D.A., and Keesom, W.H. (1999) Hydrocarbon Process., 78, 43–51. Ipatieff, V. (1935) Ind. Eng. Chem., 1067–1071. Marcilly, C. (2006) Acido-basic Catalysis – Application to Reﬁning and Petrochemistry, vol. 2, Technip, Paris, pp. 461–462. Peratello, S., Molinari, M., Bellussi, G., and Perego, C. (1999) Catal. Today, 52, 271–277. Tabak, S. and Krambek, F.J. (1985) Hydrocarbon Process., 64, 72–74.

95. O’Connor, C.T. and Kojima, M. (1990)

Catal. Today, 6, 329–349. 96. Miller, S.J. (1987) in Catalysis (ed.

97. 98.

99.

100.

101.

102.

103.

104.

105. 106.

107.

108.

109.

110.

J.W. Ward), Elsevier, Amsterdam, pp. 187–197. de Klerk, A. (2006) Energy Fuels, 20, 1799–1805. Humphrey, J.L., Seibert, F.A., and Goodpastor, C.V. (1991) U.S. DOE Report DOE/ID/12920-2. Radecki P.P., Crittenden J.C., Shonnard, D.R., and Bulloch, J.L. (eds) (1999) Emerging Separation and Separative Reaction Technologies for Waste Reduction – Adsorption and Membrane Systems, Center for Waste Reduction Technologies, AICHE Ed. Walton, K.S., Abney, M.B., and LeVan, M.D. (2006) Microporous Mesoporous Mater., 91, 78–84. Yang, R.T. (ed.) (2003) Adsorbents: Fundamentals and Applications, John Wiley & Sons, Inc., Hoboken, pp. 157–190. Granato, M.A., Vlugt, T.J.H., and Rodrigues, A.E. (2007) Ind. Eng. Chem. Res., 46, 7239–7245. Cosoli, P., Ferrone, M., Pricl, S., and Fermeglia, M. (2008) Chem. Eng. J., 145, 86–92. Rabo, J.A. and Schoonover, M.W. (2001) Appl. Catal. A: Gen., 222, 261–275. Sherman, J.D. (1999) Proc. Natl. Acad. Sci. U.S.A., 96, 3471–3478. Thomas, W.J. and Crittenden, B.D. (eds) (1998) Adsorption Technology and Design, Butterworth-Heinemann, Oxford, pp. 225–226. Mitariten, M.J. (2008) US Patent 7,442,233, assigned to Basf Catalyst LLC. Burras, R.J., Maritaten, M.J. (1994) University of Oklahoma Laurance Reid Gas Conditioning Conference, pp. 93–101. Kuznicki, S.M., Bell, V.A., Nair, S., Hillhouse, H.W., Jacubinas, R.M., Braunbarth, C.M., Toby, B.H., and Tsapatsis, M. (2001) Nature, 412, 720–724. Butwell, K.F., Dolan, W.B., and Kuznicki, S.M. (2001) US Patent 6,197,092, assigned to Engelhard Corp.

References 111. Dolan, W.B. and Butwell, K.F. (2002)

112.

113.

114.

115.

116.

117.

118.

119.

120.

121.

122.

123.

124.

US Patent 6,444,012, assigned to Engelhard Corp. Dolan, W.B. and Mitariten, M.J. (2003) US Patent 6,610,124, assigned to Engelhard Corp. Jayaraman, A., Yang, R.T., Chinn, D., and Munson, C.L. (2005) Ind. Eng. Chem. Res., 44, 5184. Carluccio, L., Bellussi, G., and Millini, R. (1994) Italian Patent 002037 MI 94, assigned to Eni S.p.A. Campbell, B.J., Bellussi, G., Carluccio, L., Perego, G., Cheetham, A.K., Cox, D.E., and Millini, R. (1998) Chem. Commun., 1725–1726. Carati, A., Rizzo, C., Tagliabue, M., Carluccio, L., Flego, C., and Ciccarelli, L. (2006) Italian Patent 001231 MI A2006, assigned to Eni S.p.A. Connor, D.J., Doman, D.G., Jeziorowski, L., Keefer, B.G., Larisch, B., McLean, C., and Shaw, I. (2002) US Patent 6,406,523, assigned to QuestAir Technologies Inc. Dunne, S.R., Coughlin, P.K., and Sethna, R.H. (2007) US Patent 7,166,149, assigned to UOP. Belding, W.A., Holeman, W.D., Lavan, Z., and Jones, R.L. (1996) US Patent 5,580,369, assigned to LaRoche Ind. Belding, W.A., Lam, C., Holeman, W.D., and Janke, S.L. (1997) US Patent 5,660,048, assigned to LaRoche Ind. Belding, W.A., Delmas, M.P.F., Holeman, W.D., and McDonald, D.A. (1997) US Patent 5,685,897, assigned to LaRoche Ind. Oda, S., Kanayama, R., Mori, K., and Sugimoto, K. (1997) US Patent 5,660,221, assigned to Sintokogio Ltd. Jale, S.R., Fitch, F.R., and Shen, D. (2000) US Patent 6,436,173, assigned to BOC Group. Simon, F.-G., Meggyes, T., and T¨unnermeier, T. (2002) Adsorption of MTBE from contaminated water by carbonaceous resins and mordenite zeolite, in Advanced Groundwater Remediation – Active and Passive Technologies (eds F.-G. Simon, T. Meggyes, and C. McDonald), Thomas Telford, London, pp. 3–34.

125. Shih, T., Wangpaichitr, M., and

126.

127. 128.

129.

130.

131.

132.

133.

134. 135. 136.

137.

138.

139.

Suffet, M. (2005) J. Environ. Eng., 131, 450–460. Sch¨uth, F. (2002) in Handbook of Porous Solids (eds F. Sch¨uth, K.S.W. Sing, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, (BRD), pp. 2719–2745. Colella, C. (2007) Stud. Surf. Sci. Catal., 168, 999–1035. Bowman, R.S., Sullivan, E.J., and Li, Z. (2000) in Natural Zeolites for the Third Millennium (eds C. Colella and F.A. Mumpton), De Frede Editore, Naples, pp. 287–297. Bowman, R.S. (2003) Applications of sufactant – modiﬁed zeolites to environmental remediation Microporous Mesoporous Mater., 61, 43–56. Schulze-Makuch, D., Pillai, S.D., Guan, H., Bowman, R.S., Couroux, E., Hielscher, F., Totten, J., Espinosa, I.Y., and Kretzschmar, T. (2002) EOS, 83, 193–201. Vignola, R., Cova, U., Della Penna G., and Sisto, R. (2005) WO Patent 2005/063631, assigned to Eni S.p.A. Vignola, R., Cova, U., Fabiani, F., Grillo, G., Molinari, M., Sbardellati, R., and Sisto, R. (2008) Stud. Surf. Sci. Catal., 174A, 573–576. Vignola, R., Cova, U., Fabiani, F., Sbardellati, R., and Sisto, R. (2009) WO Patent 2009/000429, assigned to Eni S.p.A. Cundy, C.S. and Cox, P.A. (2003) Chem. Rev., 103, 663–702. Burton, A.W. and Zones, S.I. (2007) Stud. Surf. Sci. Catal., 168, 137–179. Paillaud, J.-L., Harbuzaru, B., Patarin, J., and Bats, N. (2004) Science, 304, 990–992. ˇ Corma, A., Diaz-Cabanas, M.J., Jorda` , J.L., Martinez, C., and Moliner, M. (2006) Nature, 443, 842–845. Sun, J., Bonneau, C., Cantin, A., Corma, A., Diaz-Cabanas, M.J., Moliner, M., Zhang, D., Li, M., and Zou, X. (2009) Nature, 458, 1154–1157. Wagner, P., Yoshikawa, M., Lovallo, M., Tsuji, K., Taspatsis, M., and Davis, M.E. (1997) Chem. Commun., 2179–2180.

489

490

16 Industrial Potential of Zeolites 140. Lobo, R.F., Tsapatsis, M., Freyhardt,

141.

142.

143. 144. 145. 146.

147. 148.

149.

150. 151.

152.

153.

154.

155.

156.

C.C., Khodabandeh, S., Wagner, P., Chen, C.Y., Balkus, K.J., Zones, S.I., and Davis, M.E. (1997) J. Am. Chem. Soc., 119, 8474–8484. Strohmaier, K.G. and Vaughan, D.E.W. (2003) J. Am. Chem. Soc., 125, 16035–16039. Kresge, C.T., Leonowicz, M.E., Roth, W.J., Vartuli, J.C., and Beck, J.S. (1992) Nature, 359, 710–712. Corma, A. (1997) Chem. Rev., 97, 2373–2419. Hartmann, M. (2004) Angew. Chem. Int. Ed., 43, 5880–5882. ˇ Cejka, J. and Mintova, S. (2007) Chem. Rev., 49, 457–509. Meima, G.R., van der Aalst, M.J.M., Samson, M.S.U., Garces, J.M., and Lee, G.J. (1992) in Proceedings of the 9th International Zeolite Conference (eds R. von Ballmoos, J.B., Higgins, and M.M.J. Treacy), Butterworth-Heinemann, Boston, pp. 327–335. Chem. Eng. News, 1995, 12. Egeblad, K., Christensen, C.H., Kustova, M., and Christensen, C.H. (2008) Chem. Mater., 20, 946–960. P´erez-Ramirez, J., Christensen, C.H., Egeblad, K., Christensen, C.H., and Groen, J.C. (2008) Chem. Soc. Rev., 37, 2530–2542. Tosheva, L. and Valtchev, V.P. (2005) Chem. Mater., 17, 2494–2513. Serrano, D.P., Aguado, J., Escola, J.M., Rodriguez, J.M., and Peral, A. (2006) Chem. Mater., 18, 2462–2464. Serrano, D.P., Aguado, J., Escola, J.M., Rodriguez, J.M., and Peral, A. (2008) J. Mater. Chem., 18, 4210–4218. Aguado, J., Serrano, D.P., and Rodriguez, J.M. (2008) Microporous Mesoporous Mater., 115, 504–513. Aguado, J., Serrano, D.P., Escola, J.M., and Peral, A. (2009) J. Anal. Appl. Pyrol., 85, 352–358. Serrano, D.P., Aguado, J., Morales, G., Rodriguez, J.M., Peral, A., Thommes, M., Epping, J.D., and Chmelka, B.F. (2009) Chem. Mater., 21, 641–654. Larsen, S.C. (2007) J. Phys. Chem. C, 111, 18464–18474.

157. Derouane, E.G. (1986) J. Catal., 100,

541–544. 158. Hsu, C.-Y., Chiang, A.S.T., Selvin, R.,

159.

160.

161.

162.

163.

164.

165.

166.

167.

168.

169.

170.

171.

172.

and Thompson, R.W. (2005) J. Phys. Chem. B, 109, 18804–18814. Jaconbsen, C.J.H., Madsen, C., Houzvicka, J., Schmidt, I., and Carlsson, A. (2000) J. Am. Chem. Soc., 122, 7116–7117. Jacobsen, C.J.H., Houˇzvicka, J., Carlsson, A., and Schmidt, I. (2001) Stud. Surf. Sci. Catal., 135, Paper 03-O-02. Boisen, A., Schmidt, I., Carlsson, A., Dahl, S., Brorson, M., and Jacobsen, C.J.H. (2003) Chem. Commun., 958–959. Schmidt, I., Boisen, A., Gustavsson, ˚ E., Stahl, K., Pehrson, S., Dahl, S., Carlsson, A., and Jacobsen, C.J.H. (2001) Chem. Mater., 13, 4416–4418. Janssen, A.H., Schmidt, I., Jacobsen, C.J.H., Koster, A.J., and de Jong, K.P. (2003) Microporous Mesoporous Mater., 65, 59–75. Kustova, M.Y., Kustov, A.L., and Christensen, C.H. (2005) Stud. Surf. Sci. Catal., 158, 255–262. Pavlackova, Z., Kosova, G., Zilkova, N., ˇ Zukal, A., and Cejka, J. (2006) Stud. Surf. Sci. Catal., 162, 905–912. Kustova, M.Y., Hasselriis, P., and Christensen, C.H. (2004) Catal. Lett., 96, 205–211. Wei, X. and Smirniotis, P.G. (2006) Microporous Mesoporous Mater., 89, 170–178. Houˇzvicka, J., Jacobsen, C.J.H., and Schmidt, I. (2001) Stud. Surf. Sci. Catal., 135, Paper 26-O-02. Johannsen, K., Boisen, A., Brorson, M., Schmidt, I., and Jacobsen, C.H. (2002) Stud. Surf. Sci. Catal., 142, 109–117. Christensen, Ch.H., Johannsen, K., Schmidt, I., and Christensen, C.H. (2003) J. Am. Chem. Soc., 125, 13370–13371. Kloetstra, K.R., van Bekkum, H., and Jansen, J.C. (1997) Chem. Commun., 2281–2282. Huang, L., Guo, W., Deng, P., Xue, Z., and Li, Q. (2000) J. Phys. Chem. B, 104, 2817–2823.

References 173. Trong On, D., Lutic, D., and

174. 175.

176. 177.

178.

179. 180.

181. 182.

183.

184. 185.

186. 187.

188.

Kaliaguine, S. (2001) Microporous Mesoporous Mater., 44-45, 435–444. Trong On, D. and Kaliaguine, S. (2001) Angew. Chem. Int. Ed., 40, 3248–3251. Verhoef, M.J., Kooyman, P.J., van der Waal, J.C., Rigutto, M.S., Peters, J.A., and van Bekkum, H. (2001) Chem. Mater., 13, 683–687. Trong On, D. and Kaliaguine, S. (2002) Angew. Chem. Int. Ed., 41, 1036–1040. Trong On, D., Nossov, A., Springuel-Huet, M.-A., Schneider, C., Bretherton, J.L., Fyfe, C.A., and Kaliaguine, S. (2004) J. Am. Chem. Soc., 126, 14324–14325. Mavrodinova, V., Popova, M., Valchev, V., Nickolov, R., and Minchev, C. (2005) J. Colloid Interface Sci., 286, 268–273. Trong On, D. and Kaliaguine, S. (2003) J. Am. Chem. Soc., 125, 618–619. Choi, M., Cho, H.S., Srivastava, R., Venkatesan, C., Choi, D.-H., and Ryoo, R. (2006) Nat. Mater., 5, 718–723. Choi, M., Srivastava, R., and Ryoo, R. (2006) Chem. Commun., 4380–4381. Shetti, V.N., Kim, J., Srivastava, R., Choi, M., and Ryoo, R. (2008) J. Catal., 254, 296–303. Suzuki, K., Aoyagi, Y., Katada, N., Choi, M., Ryoo, R., and Niwa, M. (2008) Catal. Today, 132, 38–45. Choi, M., Srivastava, R., and Ryoo, R. (2006) Chem. Commun., 4489–4490. Lee, D.-H., Choi, M., Yu, B.-W., and Ryoo, R. (2009) Chem. Commun., 74–75. Garcia-Martinez, J. (2006) WO Patent 2006/031259, assigned to MIT. Ying, J.Y. and Garcia-Martinez, J. (2009) US Patent Application 2009/0005236, Applicant: MIT. Corma, A., Iglesias, M., del Pino, C., and Sanchez, F. (1991) J. Chem. Soc. Chem. Commun., 1253–1254.

189. Sanchez, F., Iglesias, M., Corma, A.,

190.

191.

192. 193.

194.

195.

196.

197. 198.

199.

200. 201.

202.

203.

and del Pino, C. (1991) J. Mol. Catal., 70, 369–379. Carmona, A., Corma, A., Iglesias, M., San Jose, A., and Sanchez, F. (1995) J. Organomet. Chem., 492, 11–21. Chauvel, A., Brunel, D., Di Renzo, F., Moreau, P., and Fajula, F. (1994) Stud. Surf. Sci. Catal., 94, 286–294. Jones, C.W., Tsuji, K., and Davis, M.E. (1998) Nature, 393, 52–54. Jones, C.W., Tsuji, K., and Davis, M.E. (1999) Microporous Mesoporous Mater., 29, 339–349. Jones, C.W., Tsuji, K., and Davis, M.E. (1999) Microporous Mesoporous Mater., 33, 223–240. Jones, C.W., Tsuji, K., and Davis, M.E. (2001) Microporous Mesoporous Mater., 42, 21–35. Yamamoto, K., Nohara, Y., Domon, Y., Takahashi, Y., Sakata, Y., Pl´evert, J., and Tatsumi, T. (2005) Chem. Mater., 17, 3913–3920. Yamamoto, K. and Tatsumi, T. (2008) Chem. Mater., 20, 972–980. D´ıaz, U., V´ıdal-Moya, J.A., and Corma, A. (2006) Microporous Mesoporous Mater., 95, 180–186. Su, B.L., Roussel, M., Vause, K., Yang, X.Y., Gilles, F., Shi, L., Leonova, E., Ed´en, M., and Zou, X. (2007) Microporous Mesoporous Mater., 105, 49–57. Astala, R. and Auerbach, S.M. (2004) J. Am. Chem. Soc., 126, 1843–1848. Elanany, M., Su, B.-L., and Vercauteren, D.P. (2007) J. Mol. Catal. A: Chem., 263, 195–199. Bellussi, G., Carati, A., Di Paola, E., Millini, R., Parker, W.O. Jr., Rizzo, C., and Zanardi, S. (2008) Microporous Mesoporous Mater., 113, 252–260. Inagaki, S., Guan, S., Ohsuna, T., and Terasaki, O. (2002) Nature, 416, 304–307.

491

493

17 Catalytically Active Sites: Generation and Characterization Michael Hunger

17.1 Introduction

The different surface sites of zeolites make these materials important catalysts in chemical technology. Besides the ﬂuid catalytic cracking (FCC) [1, 2], acidic zeolite catalysts are applied in various processes in petroleum reﬁning and basic petrochemistry [3], such as isomerization of light gasoline [4], hydrocracking of heavy petroleum distillates [5], catalytic dewaxing [5], alkylation of benzene with ethene or propene [6], disproportionation of toluene [7], isomerization of xylenes [7], and numerous others. In some of these processes, the Brønsted acid sites of zeolites are combined with a hydrogenation/dehydrogenation component, typically a noble metal, to make the catalysts bifunctional [8]. Furthermore, suitable modiﬁcation techniques may create basic properties of zeolites [9]. However, in comparison to their acidic and bifunctional counterparts, basic zeolites have gained very little importance in chemical technology until now. To obtain suitable zeolite catalysts with proper acidic, bifunctional, or basic properties, the parent material has to be modiﬁed in the right manner and tailored for the envisaged reaction. In the case of acidic zeolites, it is important to distinguish between (i) the chemical nature of acid sites in a zeolite, that is Brønsted versus Lewis acid sites, (ii) their respective concentration or density, and (iii) their strength and strength distribution. Furthermore, the (iv) accessibility of acid sites due to various locations, for example, on the external surface or inside the micropores or in large and small cavities of the zeolite framework are relevant to catalysis. Strength and accessibility are also important properties for base sites in zeolites. In the presence of metal clusters, the cluster size depending on the dispersion of the corresponding metal is an important behavior in addition to their location. This chapter introduces the nature of Brønsted and Lewis acid sites, base sites, and metal clusters as well as in their formation and characterization. For deeper insight into these topics, see [10–13].

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

494

17 Catalytically Active Sites: Generation and Characterization

17.2 Acid Sites in Zeolites 17.2.1 Nature of Acid Sites

In aluminosilicate-type zeolites, the 4+ charges on framework silicon atoms at tetrahedral positions (T position) and the 2− charges on the coordinating oxygen atoms lead to neutral SiO4/2 tetrahedra. Upon substitution of silicon atoms in the framework for metal atoms with a 3+ charge, typically by aluminum atoms, the formal charge on the corresponding tetrahedra changes from neutral to 1− (AlO4/2 − ). These negative framework charges are balanced by extra-framework metal cations or hydroxyl protons forming weak Lewis acids site or strong Brønsted acid sites, respectively, responsible for the catalytic activity of the zeolite materials [14–16]. The hydroxyl protons are located on oxygen bridges connecting a tetrahedrally coordinated framework silicon and aluminum atoms (Figure 17.1a). These hydroxyl groups are commonly donated structural or bridging OH groups (SiOHAl) [14, 15]. In addition to aluminosilicate-type zeolites, the Brønsted acidity based on bridging OH groups occurs for a wide variety of crystalline materials that have exchangeable cations, such as silicoaluminophosphates (SAPOs) [17], ferrosilicates [18, 19], and gallosilicates [20, 21]. The Brønsted acid sites in mesoporous materials with amorphous walls, such as MCM-41, MCM-48, SBA-15, and FSM-16, most likely arise from sites with local structures similar to bridging OH groups in zeolites [22–24]. The best description of these Brønsted acid sites are weakly bound protons of SiOH groups at tetrahedrally coordinated silicon atoms interacting with neighboring atoms acting as Lewis acid sites, that means as electron pair acceptors, such as aluminum atoms. In the case of zeolites, the Si–O–Al bridging bonds are closed also in the activated state of the material. In mesoporous materials with amorphous walls, the O–Al distance of the above-mentioned bridging arrangement is much larger than in crystalline zeolite leading to characteristics of SiOH groups with neighboring Lewis sites, especially in the activated state [25]. For zeolites, which are substituted with different metal atoms (SiOHT, T = Al, Ga, Fe, etc.) at framework T positions, the chemical behavior of the substituting atoms inﬂuences the acid strength of bridging hydroxyl protons in a characteristic H O Si (a)

(b)

Al OH (c)

Si OH Al

(d)

Si+

Al

Figure 17.1 Schematic representation of the different types of hydroxyl groups and acid sites in zeolites.

17.2 Acid Sites in Zeolites

manner [26–28]. Furthermore, the Si–O–T bond angle, which depends on the zeolite structure type, affects the partial charge and the acid strength of the hydroxyl protons [29]. In zeolite ZSM-5, the Si–O–T bond angles vary from 137◦ to 177◦ , for mordenite from 143◦ to 180◦ , and for zeolite Y from 138◦ to 147◦ [30, 31]. On the other hand, often, hydroxyl protons are preferably bound to speciﬁc oxygen atoms of the zeolite structure, such as to oxygen atoms at O1 positions in zeolite Y [32], which limits the range of Si–O–Al bond angles in the local structure of bridging OH groups. In addition to the above-mentioned local effects, also the global composition of the zeolite framework inﬂuences the acid strength of bridging OH groups in zeolites, such as the mean electronegativity of the framework, which changes with the aluminum content. The mean Sanderson electronegativity, Sm , of the zeolite framework is deﬁned as the geometric means of the electronegativities Si of the atoms i, for example, in one unit cell. The value of Sm for H-form zeolites can be calculated by [33, 34] 1/(i+j+k+l+m) j Sm = SiH SSi SkAl SlO Sm Si

(17.1)

with SH = 3.55, SSi = 2.84, SAl = 2.22, and SO = 5.21. The value Sm is not related to any long-range order (structure), but can be utilized to correlate the chemical behavior of sites with same local structure or analytical data on these sites with the chemical composition of the materials under study. Generally, the acid strength of zeolitic bridging OH groups increases with increasing mean electronegativity, Sm , of the zeolite framework, that is, with decreasing aluminum content. Therefore, dealumination of the zeolite framework results in zeolites with higher acid strengths [32, 35]. In a similar manner, the concept of next nearest neighbors (NNNs) is utilized [36, 37]. For example, in FAU-type zeolites, each framework aluminum atom is linked via oxygen bridges with four silicon atoms [38]. These four silicon atoms are connected with nine further T atoms in the next coordination sphere, which are the so-called NNNs. According to the NNN concept, the acid strength of SiOHAl groups in aluminosilicate-type zeolites depends on the number of framework aluminum atoms (lower electronegativity in comparison with silicon) on NNN positions. A completely isolated AlO4 tetrahedron (highest acid strength) has a 0-NNN conﬁguration (nSi /nAl 11). An FAU-type zeolite with the maximum number of framework aluminum atoms (nSi /nAl ≈ 1) is characterized by a 9-NNN conﬁguration (lowest acid strength) [36, 37]. The second important type of hydroxyl groups in zeolites are the silanol groups (SiOH), also called terminal OH groups, which are located on the external surface of crystal particles or at framework defects (Figure 17.1b). Dealumination of the zeolite framework, for example, by calcination, hydrothermal treatment, or treatment with strong acids, is the most important reason for the formation of framework defects and silanol groups. Depending on the treatment conditions, healing of framework defects, such as by silicon migration, formation of silanol groups, or formation of hydroxyl groups at extra-framework aluminum species (Figure 17.1c) may occur [16].

495

496

17 Catalytically Active Sites: Generation and Characterization

Often, dealumination of the zeolite framework is accompanied by the formation of Lewis acid sites at extra-framework aluminum species and framework defects (Figure 17.1d). If these Lewis acid sites are located in the vicinity of bridging OH groups, for example, in the framework of weakly steamed zeolites, superacidic Brønsted sites are formed [39, 40]. Lago et al. [41] found a strongly enhanced catalytic activity in the n-hexane cracking on zeolite H-ZSM-5, which was attributed to superacidic Brønsted sites formed upon mild steaming. These sites were explained by a partial hydrolyzation of framework aluminum atoms acting as Lewis acid sites in the vicinity of bridging OH groups (SiOHAl). The framework Lewis sites act as strong electron withdrawing centers for neighboring bridging OH groups and create Brønsted acid sites with very high strength. However, similar effects also occur for bridging OH groups in the vicinity of extra-framework aluminum species acting as Lewis acid sites [42]. 17.2.2 Formation of Brønsted and Lewis Acid Sites

The most important procedures to form bridging OH groups in zeolites are described by Eqs. (17.2) and (17.3). Here, Z− stands for the negatively charged zeolite framework in the local structure of the tetrahedrally coordinated framework aluminum atoms [32, 43] ≈573−673K

+ − + − −−−−−→ H+ Z− NH+ 4 + Na Z −−−→NH4 Z − −Na+

−NH3

(17.2)

An aqueous ion exchange of the alkali metal form of the zeolite with an ammonium salt is followed by thermal decomposition of the ammonium ions inside the zeolite, leading to the desorption of ammonia and the bonding of a hydroxyl proton at the bridging oxygen of an Si–O–Al arrangement. The second procedure starts with an aqueous ion exchange of the alkali metal form of the zeolite with the salt of a multivalent metal cation (often used cations are Mg2+ , Ca2+ , La3+ , and Al3+ or mixed rare-earth cations) followed by thermal dehydration [43–45] ≈573 K [La (H2 O)n ]3+ + 3Na+ Z− − −−−−−→ [La (H2 O)n ]3+ Z− 3 −−−−−−−−→ + −(n−2)H2 O −3Na [(LaOH) (H2 O)]2+ H+ Z− 3 −−−→ [La (OH)2 ]+ H+ 2 Z− 3 (17.3) The reactions in Eq. (17.3) correspond to the Hirschler–Plank mechanism [46, 47]: removing of the hydration shell of multivalent cations, for example, by thermal treatment, leads to the strong electrostatic ﬁelds inside the zeolite pores, because these cations have to neutralize typically two or three negative framework charges in the zeolite with signiﬁcant distances from each other. Residual water molecules dissociate in the strong local electrostatic ﬁelds and form hydroxyl protons bound to a bridging oxygen atoms (SiOHAl) and OH groups bound to the extra-framework cations. The latter type of OH groups is nonacidic. According to Eq. (17.3), maximum two Brønsted acid sites and two metal OH groups can be formed per three-valent cation introduced by cation exchange.

17.2 Acid Sites in Zeolites

Another procedure is the direct ion exchange with mineral acids [43] H+ + Na+ Z− −−−→ H+ Z− −Na+

(17.4)

However, this route is generally less favored, because the exposure of zeolites to strong liquid acids often leads to framework dealumination or, in the case of aluminum-rich zeolites, to a complete framework collapse. However, it is a suitable way for the preparation of the H-form of zeolites with low aluminum contents, such as silicon-rich mordenite, ZSM-5, and MCM-22. Finally, the reduction of noble metal cations in zeolites by molecular hydrogen leads to the formation of bridging OH groups [1, 2] [Pd (NH3 )4 ]2+ + 2Na+ Z− −−−−−−→ [Pd (NH3 )4 ]2+ Z− 2 ≈573 K

−−−→ Pd −4NH3

2+

Z

−2Na+

−

−−−→ Pd0 H+ 2 Z− 2 +H2

2

(17.5)

At ﬁrst, complexes of metals nobler than molecular hydrogen, for example, [Pd(NH3 )4 ]2+ , have to be introduced by ion exchange with the Na-form zeolite. Thermal treatment causes the transformation of the above-mentioned complexes into noble metal cations, which are reduced by molecular hydrogen into neutral noble metals accompanied by the formation of hydroxyl protons at oxygen bridges in the vicinity of the negatively charged framework aluminum atoms. Regardless of the procedure utilized for the formation of Brønsted acidic bridging OH groups (Eqs. 17.2–17.5), their local structure and chemical nature are the same. Strong heat treatment (T ≥ 773 K) or steaming of acidic zeolites causes a dehydroxylation of the Brønsted acid sites and water is split off with the concomitant formation of Lewis acid sites. The chemical nature of Lewis acid sites in zeolites caused by the above-mentioned treatments is a matter of research. In the case of framework dealumination, Lewis acid sites can be attributed to extra-framework aluminum species of octa-, penta-, or tetrahedral coordination [48–50]. Already Scherzer and coworkers [51] suggested AlO+ , Al(OH)2 + , and AlO(OH) as aluminum species on extra-framework positions in dealuminated zeolites. Kuehl [43, 52] concluded from X-ray spectrometry that [AlO]+ units removed from the zeolite framework are transformed into cationic extra-framework species, which act as so-called true Lewis acid sites. The presence of cationic extra-framework aluminum species in dehydrated and dealuminated Y-type zeolites could be recently supported by high-ﬁeld 27 Al solid-state nuclear magnetic resonance (NMR) spectroscopy [50]. The transitional state of aluminum atoms in the process of framework dealumination is distorted tetrahedrally coordinated framework aluminum atoms, like partially dislodged species. For zeolite H-beta, Grobet and coworkers could observe this aluminum transitional state by solid-state NMR spectroscopy [53]. Framework Lewis acid sites are discussed to consist of positively charged silicon ions in the neighborhood of tri-coordinated aluminum atoms. Gonzales et al. [54] studied the formation of framework Lewis sites, for example, via the dehydroxylation route shown in Figure 17.2. The tricoordinated framework silicon species are electron pair acceptors acting as strong Lewis acid sites. Density

497

498

17 Catalytically Active Sites: Generation and Characterization

Si

Si O

O

O

O H

Al O Si

O

O Si

Si O

−H2O

Al OH Si

Si

Si

Si

Si

Si +

O

Al O

O Si

O

O

Si

Al O

O Si

Si

O Si

Figure 17.2 Mechanisms of the dehydroxylation of zeolites and formation of framework Lewis acid sites (adapted from [16]).

functional theory (DFT) calculations indicated that the strength of these framework Lewis acid sites is much higher than the strength of extra-framework AlO+ and Al(OH)2 + species [16].

17.3 Characterization of Acid Sites 17.3.1 Catalytic Test Reactions

Since cracking of heavy petroleum fractions on acidic zeolite catalysts is among the most important commercial processes, also this reaction has been used frequently for characterizing the catalytic activity. For H-forms of zeolites, the cracking rates of hydrocarbons were often found to increase linearly with the framework aluminum content, which corresponds to the density of Brønsted acid sites. For example, for zeolites H-ZSM-5 and H-Y, a linear increase in the rate of n-hexane cracking with the aluminum content has been found over a wide range of framework compositions [55–59]. The frequently utilized alpha test consists of the measurement of the n-hexane cracking rate under speciﬁed conditions [59]. Interestingly, upon mild steaming of zeolite H-ZSM-5, a signiﬁcant increase in the n-hexane cracking rate in comparison with nonsteamed zeolites H-ZSM-5 with equal Brønsted acid site densities was observed [41]. This enhanced activity was attributed to superacidic Brønsted sites formed by a synergism of Lewis acid sites in the vicinity of bridging OH groups. Furthermore, the cracking rates may also depend on the zeolite structure, even for nonsteamed catalysts [60]. These examples demonstrate that a great care must be applied, when drawing conclusions on the density of Brønsted acid sites from results of cracking reactions. Since most of research groups dealing with heterogeneous catalysis are equipped to perform hydrocarbon cracking reactions, however, these test reactions have been quite popular for characterizing the density of acid sites in zeolites. In a similar manner, the disproportionation of ethylbenzene on zeolites is utilized for characterizing the Brønsted acidity [61–66]. Due to the good linear relationship between the density of Brønsted acid sites in acidic zeolites and their disproportionation activity, this reaction was suggested by the Catalysis Commission of the

17.3 Characterization of Acid Sites

International Zeolite Association (IZA) as test reaction [61]. The reaction products are benzene, unconverted ethylbenzene, ortho-, meta-, and para-diethylbenzenes, and at elevated conversions, triethylbenzenes. Interestingly, an induction period occurs for the disproportionation of ethylbenzene on large-pore zeolites, such as acidic Y-type zeolites. In this period, the catalysts gain activity while on stream. After a certain time-on-stream, the catalyst activity is constant and can be used for characterizing the density of Brønsted acid sites. Karge et al. [62–65] investigated the disproportionation of ethylbenzene for a broad variety of acid zeolites. They found that the induction period occurs only for large-pore zeolites, but it is absent for medium-pore zeolites [64, 65]. In addition, higher temperatures are needed for ethylbenzene disproportionation on medium-pore zeolites than on large-pore zeolites. This has been interpreted in terms of different disproportionation mechanisms: (i) a dealkylation/realkylation path with free ethene as an intermediate for medium-pore zeolites, as opposed to (ii) a mechanism via diphenylmethane intermediates for large-pore zeolites, which has been recently supported by in situ 13 C MAS NMR investigations [67]. Several other reactions have been proposed and utilized for characterizing acidic zeolite catalysts, such as the conversion of toluene and xylene [68] and the dehydration of cylohexanol [69]. Also series of test reactions requiring different activation energies or reaction temperatures were suggested in order to measure the different acid strengths of the zeolite catalysts (Table 17.1) [70, 71]. According to Sigl et al. [72], however, also the disproportionation of ethylbenzene is sensitive to the strength of acid sites in zeolites. They employed this reaction to study the acidity of H-[Al]ZSM-5, H-[Ga]ZSM-5, and H-[Fe]ZSM-5 zeolites with almost exactly identical acid site densities under standardized conditions. Under quasistationary conditions, the ethylbenzene conversion reﬂected the acid strength decreasing in the sequence H-[Al]ZSM-5 > H-[Ga]ZSM-5 > H-[Fe]ZSM-5 [72].

Characterization of the strength of Brønsted acid sites in zeolites via catalytic test reactions running at different reaction temperatures [70, 71].

Table 17.1

Reactant

Reaction

Tr (K)

3,3-Dimethyl-1-butene Cyclohexene 2,2,4-Trimethylpentane 2,4-Dimethylpentane o-Xylene 1,2,4-Trimethylbenzene 2-Methylpentane n-Hexane

Skeletal isomerization Skeletal isomerization, hydrogen transfera Cracking Isomerization, cracking Isomerization, disproportionationa Isomerization, disproportionationa Isomerization, cracking Isomerization, cracking

473 473 623 623 623 623 673 673

a Bimolecular

reaction.

499

500

17 Catalytically Active Sites: Generation and Characterization

17.3.2 Titration with Bases

Benesi [73–76], Hirschler and Schneider [77], and Moscou and Mone [78] introduced the titration of acid sites with bases and indicators. This method was very early utilized to collect information on both the density and strength of acid sites in solid catalysts. Generally, the surface sites are titrated with an amine, such as n-butylamine, in a nonaqueous solvent, and a series of Hammett indicators with different pKa values is employed. Upon protonation of the indicators by the Brønsted acid sites of the zeolite catalyst under study, a color change is observed. Typical Hammett indicators are summarized in Table 17.2. In addition, so-called Hirschler indicators are used, which consist of arylcarbinol compounds [77]. However, it must be mentioned that none of the usually applied Hammett and Hirschler indicators are selective for either Brønsted or Lewis acid sites. The application of colored indicators in liquid acid–base titrations is a routine method. Their use for the titration of acid sites is a strong chemistry-related technique for zeolite characterization. Generally, the strength of acids is the capability of the transfer of a proton H+ to a neutral base B leading to the conjugated form BH+ . Quantitatively, the proton transfer can be described by the acidity function H0 in the following manner [71] a + fB H0 = −log H (17.6) fBH+ where aH + is the proton activity and fB and fBH+ are the activity coefﬁcients of B and BH+ , respectively. When dealing with liquid acids, the following assumptions must be fulﬁlled: the chemical equilibrium is achieved at each time and the amount of indicator is much too small to disturb the equilibrium appreciably. In the titration of acid sites on the surface of acidic solids, however, these assumptions have been severely questioned. An additional topic is the effect of different nonaqueous solvents with Table 17.2 Hammett indicators suitable for the visible indication of the end-point in the titration of colorless solid acids, such as acidic zeolites [71, 77].

Indicator

Basic color

Acid color

pKa

Acid strength/wt% H2 SO4

Natural red Phenylazonaphthylamine Butter yellow 4-Benzeneazodiphenylamine Dicinnamalacetone Benzalacetophenone Anthraquinone

Yellow Yellow Yellow Yellow Yellow Colorless Colorless

Red Red Red Purple Red Yellow Yellow

+3.3 +4.0 +3.3 +1.5 –3.0 –5.6 –8.2

8 × 10 –8 5 × 10 –5 3 × 10 –4 2 × 10 –2 48 71 90

17.3 Characterization of Acid Sites

different dielectric constants, εr , since the energy required to separate charged species is inversely proportional to εr [15]. Streitwieser and Kim [79] compared the basicity of a series of amines in tetrahydrofuran (THF, εr = 7.6) with their basicities in dimethyl sulfoxide (DMSO, εr = 46.7), and acetonitrile (εr = 35.9). While the protonation of amines in DMSO and acetonitrile leads to spatially separated ionic species, the corresponding protonation products in THF are similar to ion pairs since the energy required for separating charged species in a medium with low dielectric constant is too high. Finally, the applicability of the indicator method for the characterization of acidic zeolites is limited by the size of the base molecules. In some cases, they cannot enter the pores and cages zeolites and thus react only with sites on the external surface and in the pore mouth region [80]. In spite of the above-mentioned limitations, the standard Hammett indicator method is useful, such as for the characterization of the acidity of mesoporous materials. For MCM-41 materials without aluminum incorporation (MCM-41) and upon aluminum incorporation according to nSi /nAl = 30 ([Al]MCM-41/30) and 10 ([Al]MCM-41/10), the titration with n-butylamine led to densities of acid sites of 0.08, 0.22, and 0.45 mmol g−1 , respectively [81]. Recently, the Hammett indicator method was modiﬁed by Wang and coworkers [82–84] for the characterization of medium-pore zeolites, such as nanoscale H-ZSM-5. The main difference of the procedure is that the method of antititration and the insertion method were adapted. Before the titration, the samples were dehydrated at 613 K for 30–60 minutes. Upon cooling to 473 K, petroleum ether and, subsequently, speciﬁc amounts of n-butylamine were added. An ultrasonic oscillator was used to reduce the equilibration time. Furthermore, the overlap of the indicators were reduced to minimum in comparison to the traditional titration method by applying neutral red (pKa = +6.8), methyl red (pKa = +4.8), p-aminoazobenzene (pKa = +2.27), and dicinnamalacetone (pKa = −3.0) as indicators [82]. For studying the total amount of acid sites and the amount of external acid sites, n-butylamine and cyclohexylamine, respectively, were used for the titration [83]. Comparison of the toluene disproportionation as catalytic test reaction with titration of nanoscale zeolite H-ZSM-5 indicated that only acid sites with H0 ≤ +2.27 are related to the catalytic activity of the corresponding materials [83]. 17.3.3 Temperature-Programmed Desorption of Bases

Temperature-programmed desorption (TPD) of basic molecules is utilized to determine both the density and strength of acid sites in zeolites. The procedure requires at ﬁrst an evacuation of the zeolite under study, often at 773 K. Subsequently, a gaseous base, for example, ammonia or pyridine, is adsorbed via the gas phase, typically at 373 K. The measurement starts by heating the base-loaded zeolite in a stream of inert gas, such as helium, argon, or nitrogen, with a temperature program. The amount of bases desorbed is detected gravimetrically [85], volumetrically [85], by gas chromatography [86], or mass spectrometry [87]. The temperatures at which desorption peaks occur are related to the acid strength of the Brønsted sites, if the

501

502

17 Catalytically Active Sites: Generation and Characterization

a c b

353 403 453 503 553 603 653 703 753 Desorption temperature (K) Figure 17.3 Temperature-programmed desorption of ammonia (TPDA) from the silicoaluminophosphates H-SAPO-5 (a) and H-SAPO-11 (b), and zeolite H-ZSM-5 (c) [88].

presence of Lewis acid sites can be excluded (see below). The area under the desorption peaks is correlated with the number of acid sites. In this case, the total area must be calibrated via a standard sample or by titration of the total amounts of desorbed bases. Typical examples of different TPD curves are shown in Figure 17.3 [88]. These curves were obtained for the temperature-programmed desorption of ammonia (TPDA) from silicoaluminophosphates H-SAPO-5 (nSi /(nAl + nP ) = 0.076) and H-SAPO-11 (nSi /(nAl + nP ) = 0.068) and zeolite H-ZSM-5 (nSi /nAl = 25.3) [88]. The desorption peaks occurring at 438 K (0.35 mmol g−1 ) and 513 K (0.40 mmol g−1 ) for H-SAPO-5 and at 438 K (0.30 mmol g−1 ) and 548 K (0.35 mmol g−1 ) for H-SAPO-11 were assigned to two kinds of Brønsted sites with medium acid strength. For zeolite H-ZSM-5, two desorption peaks were detected at 478 K (0.48 mmol g−1 ) and 673 K (0.50 mmol g−1 ), which were attributed to Brønsted sites of medium and high acid strength, respectively. The values in parenthesis are the concentrations of acid sites calculated via the areas under the corresponding desorption peaks. The above-mentioned examples are quite representative in that the strength of the acid sites from which the probe molecules were desorbed is simply correlated in a qualitative manner with the temperature of the maximum in the desorption peak. Generally, crystalline SAPOs have a signiﬁcantly lower acid strength in comparison with aluminosilicate-type zeolites H-ZSM-5. In spite of the broad application of the TPD method for the characterization of the acid sites of zeolites, some serious limitations have to be considered. Desorption spectra may be affected by a hindered diffusion of the desorbed base molecules and a readsorption on their way out of the pores [89, 90]. This can affect the position of the desorption peaks on the temperature scale. An important difﬁculty is that ammonia interacts with both Brønsted and Lewis acid sites, and it is not possible to distinguish between the nature of the sites from which the base molecules have been desorbed. Since calcined zeolites often contain extra-framework aluminum species and framework defects, the amount of desorbed ammonia is by no means equal to the number of Brønsted acid sites. Woolery et al. [91] suggested a mild steaming

17.3 Characterization of Acid Sites

upon exposing the calcined zeolites to ammonia. The authors demonstrated that the method is suitable to eliminate ammonia desorption from Lewis acid sites for high-silica zeolites yielding a good correlation between the Brønsted acid site densities obtained by TPD of ammonia and other methods. An alternative method for the selective determination of the density of Brønsted acid sites in zeolites is the TPD of reactive amines [92–95]. Alkylammonium ions, which are formed by protonation of amines at Brønsted acid sites, react in a very narrow temperature range via a reaction similar to the Hofmann elimination reaction [15, 95] − R–CH2 –CH2 –NH2 + ZOH −→ R–CH2 –CH2 –NH+... 3 ZO

(17.7)

− R–CH2 –CH2 –NH+... 3 ZO

(17.8)

−→ R–CH=CH2 + NH3 + ZOH

After saturating the activated zeolite with the amine at room temperature and evacuation for 1 hour, all molecules except those that are bound to a framework aluminum atom are desorbed. Using isopropylamine, the probe molecules in excess of one per framework aluminum atom do not react below 500 K and leave the sample until a coverage of 1 : 1 is reached. The remaining probe molecules decompose via Eqs. (17.7) and (17.8) to propene and ammonia between 625 and 700 K, which can be observed, for example, by mass spectrometry [15]. The temperature at which the decomposition occurs depends on the nature of the alkyl group rather than on the type of the solid acid under study. This indicates that the technique is not suitable for determining the strength of Brønsted acid sites [95]. Recently, Simon and coworkers [96] correlated the NH3 -TPD curves of zeolites H-ZSM-5 (nSi /nAl = 60–2000) with the results of impedance measurements. They found up to four characteristic types of proton transport processes depending of the ammonia desorption temperature respectively the number of residual ammonia molecules on the sample materials. In the ﬁrst temperature range (up to about 393 K), a Grotthus-like proton transport along (NH3 )n -chains occurs (range (i)). Near the low-temperature (LT) peak of the TPDA curve at about 453 K, the (NH3 )n -chains are broken and the above-mentioned proton transport mechanism requires more thermal activation (range (ii)). Before reaching the high temperature (HT) peak at about 623 K, a vehicle-like transport process of NH4 + species was observed (range (iii)). At desorption temperatures higher than 623 K, the conductivity decreases signiﬁcantly and is dominated by an intersite proton hopping (range (iv)). An interesting approach is the TPDA in combination with infrared (IR) and mass spectroscopy (Figure 17.4) [97–99]. With this technique, the change of the IR spectra of zeolites beta simultaneously with desorption of ammonia has been studied [97]. The authors distinguished desorption of NH3 weakly bound at Lewis acid sites (LT peak at 1620 cm−1 ), at Brønsted acid sites (HT peak at 1450 cm−1 ), and NH3 strongly adsorbed on dislodged extra-framework aluminum species (HT+ peak at 1320 cm−1 ). The catalytic activity of zeolite H-beta in the octane cracking could be related to the amount of strong Brønsted acid sites responsible for the NH3 adsorptions heats of H = 140 kJ mol−1 [97]. In a similar manner, the reasons of the octane cracking activity of Ba-, Ca-, and La-exchanged zeolites Y was investigated [99]. An increasing Brønsted acid strength (increasing NH3

503

504

17 Catalytically Active Sites: Generation and Characterization

Helium 3

1 2

5

4 Mass spectrometer

6 IR beam Sample wafer

Figure 17.4 Set-up of temperature-programmed desorption of ammonia (TPDA) in combination with infrared and mass spectroscopy (IRMS) consisting of ﬂow meter (1), liquid nitrogen trap with silica gel column (2), vacuum meter (3), sampling loop for calibration (4), vacuum pump (5), and liquid nitrogen trap (6) [97].

adsorption heat H) of zeolites Y was found comparing the Na-form material with the Ba- and Ca-exchanged samples. This enhanced acid strength was explained by the polarizing effect of the multivalent cations in the Ba- and Ca-form zeolites. A suitable technique for systematic studies of the adsorption energies of probe molecules and reactants is the microcalorimetry. 17.3.4 Microcalorimetry

The higher heat of adsorption occurs when acid sites react with probe molecules or reactants with the stronger acid sites. The heat of adsorption of probe molecules and reactants on acidic zeolites, therefore, is utilized to characterize the chemical behavior of Brønsted acid sites. The determination of the heat of adsorption is usually performed by (i) calculating the isosteric heats from adsorption isotherms, measured at different temperatures, or by (ii) measuring the heat of adsorption directly with a calorimeter at a chosen temperature. For acidic zeolites, the adsorption heats obtained by the above-mentioned two procedures differ signiﬁcantly. These discrepancies are due to the fact that the derivative at the experimental curve, which is required for calculating the isosteric adsorption heat via the Clausius–Clapeyron equation, is difﬁcult to obtain. In contrast, direct calorimetric measurements give more reliable results [15, 100–105]. The determination of the characteristics of adsorption on the surface sites of solid catalysts by gas-phase adsorption microcalorimetry requires equipment for the simultaneous determination of the adsorbed amount of gas molecules and of the adsorption heat. Often, the adsorption heat is measured with an isothermal and differential microcalorimeter, the Tian–Calvet type calorimeter [100, 104]. This type of calorimeter has two champers, one containing the adsorbent and the other is an empty reference chamber. Furthermore, there are a pump for evacuating the

17.3 Characterization of Acid Sites

sample and the cells, a system for dosing the adsorbate, and the microcalorimeter allowing the measurement of adsorption heat in the sample chamber in comparison with the empty reference chamber. With this equipment, it is possible to perform adsorption experiments at constant temperature (usually of the order of 423 K). The amount of adsorbed probe molecules is determined by the manometer linked to the calorimeter. With this manometer, the quantity na of the adsorbed gas molecules is measured. The thermal sensors of the calorimeter give the amount of heat Qint developed by the adsorption process. The true differential heat of adsorption is obtained by calculating qdiff = dQint /dna [103]. This differential heat qdiff is plotted as a function of na . Generally, the differential heat curves plotted versus the acid site coverage consist of four ranges: (i) an initial part with high adsorption heats often caused by adsorption at Lewis acid sites, (ii) a plateau of intermediate adsorption heats due to Brønsted acid sites, (iii) a range where the adsorption heats decrease steeply depending on the heterogeneity of the sites, and (iv) the range of low adsorption heats at high coverage characteristic of reversible adsorption or physisorption. As an example, Figure 17.5 shows the differential heats for adsorption of pyridine on four H-ZSM-5 zeolites with aluminum contents of 180, 370, 530, and 600 µmol g−1 [104]. Independent on the aluminum content of the H-ZSM-5 zeolites, the differential heats are constant at about 200 kJ mol−1 (plateau (ii)) until the coverage of one probe molecule per framework aluminum atom is reached. The points of drop down of the differential adsorption heat agree well with the concentration of aluminum atoms in the zeolites under study. The absence of strong adsorption heats in the initial parts (range (i)) of the curves indicates that no Lewis acid sites exist. Generally, for the zeolites with the lower aluminum contents, the higher acid strength of the Brønsted acid sites and the higher differential adsorption enthalpy at low coverage (plateau (ii)) is expected. Owing to the irreversible adsorption of base molecules at Brønsted acid sites for temperatures of T < 600 K, these

q diff (kJ mol−1)

250 200 150 100 50

0

200

400

600

−1

n a (µmol g ) Figure 17.5 Differential heats of adsorption of pyridine on zeolites H-ZSM-5 with aluminum contents of 180 (•), 370 (◦), 530 (), and 600 µmol/g () [104].

505

506

17 Catalytically Active Sites: Generation and Characterization

molecules interact with the Brønsted acid sites, which are ﬁrst available in the pore system of the zeolite. Therefore, the adsorption temperature should be not too low, in order to allow reaching the adsorption equilibrium at all acid sites and the detection of differences among them. For further effects inﬂuencing the results of microcalorimetric measurements, see [106]. The constant values of the plateau (ii) of differential heat curves are characteristic for the chemical behavior of Brønsted acid sites. In zeolites, the concentration and strength of Brønsted acid sites depend on the aluminum content in the framework (see Section 17.2.1). In Figure 17.6, the differential heats of ammonia adsorption at the plateau (ii) are summarized for various zeolites and mesoporous materials with different structures and nSi /nAl ratios [106]. Interestingly, acidic zeolites with kJ mol−1 Plateau (ii)

(n Si /n Al)

200 Mazzite (12) 190 180

USY

170 160 150

Mordenite (16) Erionite (3.5) Offretite (3.9), H-ZSM-5 (14) Ferrierite (15), SAPO-37 (Si/Si+ Al+ P = 0.2)

140 130

H-beta (10) H-Y (2.4), MCM-41 (4.5)

120 110 H-X (1.25) 100 H-A (1) 90 80 70

Li-X, Li-Y Na-X, Na-Y, Ti-silicalite K-X

60 50 40

K-Y, Rb-Y Rb-X, Cs-X Cs-Y Silicalite

Figure 17.6 Differential heats of ammonia adsorption on various zeolites and mesoporous materials with different structures and nSi /nAl ratios (in parenthesis) determined at 423 K [106].

17.3 Characterization of Acid Sites

similar aluminum contents, but different structures, cause adsorption heats, which differ up to about 50 kJ mol−1 (e.g., mazzite with nSi /nAl = 12 and ferrierite with nSi /nAl = 15). This ﬁnding demonstrates the strong effect of the local geometry (bond angles and distances) of bridging OH groups on their acid strength. According to Figure 17.6, the maximum differential heat of ammonia adsorption occurs for mazzite (about 195 kJ mol−1 ). In a more recent study, microcalorimetry of ammonia adsorption on zeolite MCM-22 was combined with Fourier transform infrared (FTIR) spectroscopy of isobutyronitrile and pivalonitrile as probe molecule [107]. Initial ammonia adsorption heats (region (i)) of 364 kJ mol−1 indicated the presence of strong Lewis acid sites. With increasing ammonia uptake (plateau (ii)), differential heats of >150 kJ mol−1 , 120–150 kJ mol−1 , as well as 120–70 kJ mol−1 were determined. By FTIR spectroscopy of isobutyronitrile and pivalonitrile adsorbed on the MCM-22 zeolite, it was evidenced that the strong Lewis acid sites are located at the outer particle surface. Microcalorimetric studies of the gas-phase adsorption of reactants on zeolites were performed, such as for n-hexane, 2-methylpentane, 2,2-dimethylbutane, and 2,3-dimethylbutane on H-ZSM-5 zeolites with nSi /nAl ratios of 50, 100, and [108]. Interestingly, the differential heat of n-hexane adsorption increased slightly with loading for the materials with nSi /nAl ratios of 100 and 1600, while the heat of adsorption of branched hexanes showed a decrease with loading. In contrast, for zeolite H-ZSM-5 with nSi /nAl = 50, the initial heat of n-hexane adsorption was constant and then increased as a pronounced peak in the plateau region (ii). This strong increase in the differential heat of n-hexane adsorption was explained by adsorbate–adsorbate interactions that increase with stronger packing of the molecules in the zeolite pores. Another calorimetric method is the liquid-phase calorimetry (calorimetryadsorption) of zeolites in a solution, for example, of pyridine in cyclohexane [109–112]. In this case, the acidic zeolite is dispersed into a low solvation degree solvent, such as cyclohexane, and the heat evolved upon addition of incremental amounts of a basic molecule is measured. The intervals between the measurements are chosen in such a way that adsorption equilibrium is reached. Often, the obtained calorimetric data are explained by two- or three-site models assuming adsorption of pyridine at Brønsted or Lewis acid sites [111, 112]. In the case of H-USY zeolite, enthalpies of −134.0 and −101.5 kJ mol−1 were obtained for adsorption of pyridine at two different types of Brønsted acid sites [111]. In the case of cerium-impregnated USY zeolite, the introduced Lewis acid sites led to an adsorption enthalpy of −83.6 kJ mol−1 [111]. For zeolite H-beta (nSi /nAl = 50) activated at 723 K, enthalpies of −227.3 and −93.3 kJ mol−1 were assigned as pyridine adsorption at strong and medium Brønsted acid sites, while the enthalpy of −80.4 kJ mol−1 indicated the presence of sites being a combination of Brønsted and Lewis acid sites [112]. The adsorption enthalpies of −83.6 and −62.8 kJ mol−1 obtained upon activation of zeolite H-beta at 823 K were explained by the formation of Lewis acid sites and weak Brønsted acid sites, respectively, due to dealumination of the framework [112].

507

508

17 Catalytically Active Sites: Generation and Characterization

17.3.5 FTIR Spectroscopy

Due to the permanent dipole moment of OH groups, the FTIR spectroscopy allows the direct study of Brønsted acid sites in zeolites, while the investigation of Lewis acid sites requires the use of probe molecules. To characterize Brønsted acid sites in zeolites, often the fundamental stretching vibrations of hydroxyl groups are investigated using the IR transmission technique [113–115]. This technique requires the preparation of zeolite samples as thin wafers with a thickness of about 10 mg cm−2 . These wafers are measured in transmission cells allowing the activation of the zeolite catalysts under vacuum conditions and without or upon adsorption of probe molecules. In some cases, if the material under study cannot be pressed to thin wafers or if the transmission is too weak, the diffuse reﬂection infrared Fourier transform (DRIFT) technique is applied. For principles, advantages, and limitations of this technique, see [114–117]. Table 17.3 gives a survey on the fundamental stretching vibrations, νOH , of hydroxyl groups in zeolites, which cover a range of 3200 – 3800 cm−1 . Further Table 17.3 Wavenumber, ν OH , of the fundamental stretching vibrations and assignments of hydroxyl groups in dehydrated zeolites.

ν OH (cm –1 )

OH group

Assignment

References

3780

MeOH

[113, 114, 120]

3720–3745

SiOH

3665–3690

AlOH

3570–3610

3600–3660

CaOH*, MgOH*, AlOH* SiOHAl

3540–3580

SiOH*Al

3470–3550

SiOH*

3250

SiOH*Al

Terminal metal OH groups in large cages and on the external surface, such as AlOH groups Terminal silanol groups on the external surface or at lattice defects OH groups at extra-framework aluminum species Cation OH groups located in sodalite cages of zeolite Y and in channels of ZSM-5, hydrogen bonded HF band, bridging OH groups in large cages and channels of zeolites LF band, bridging OH groups in small cages of zeolites, hydroxyl protons interacting with framework oxygen Hydrogen-bonded SiOH groups, internal silanols Disturbed bridging OH groups in zeolite H-ZSM-5, H-beta, and H-MCM-22, hydroxyl proton interacting with framework oxygen

‘‘*’’ indicates hydrogen bonding or electrostatic interaction.

[113, 114, 121–123] [113, 114, 124, 125] [113, 126, 127]

[113, 114, 128, 129] [113, 114, 128, 129] [130–132] [133]

17.3 Characterization of Acid Sites

information on the properties and local structure of OH groups can be derived from analysis of the bending vibrations, overtone, and combination bands [105, 113, 118]. A detailed survey on the assignment of the vibrations of hydroxyl groups in H-form and cation-exchanged zeolites of different structures is given in [115]. Generally, for certain zeolites containing large and small cages or pores, the stretching vibrations of bridging OH groups (SiOHAl) are split into two characteristic ranges, the high-frequency (HF) and low-frequency (LF) bands. The HF bands occurring at 3600–3660 cm−1 are due to noninteracting SiOHAl groups in large cages or pores consisting of 10-rings or larger. Examples are bridging OH groups in the supercages of faujasite-type zeolites or 10- or 12-ring pores of other zeolite-types. In contrast, the LF bands occurring at 3540–3580 cm−1 are caused by SiOHAl groups in small structural units, such as sodalite cages of faujasite-type zeolites or eight-ring pores. The latter type of hydroxyl protons interacts with oxygen atoms in their vicinity, for example, via hydrogen bonding or electrostatic interactions. For solid hydrates, a correlation between the H· · ·O distance of the hydroxyl protons to neighboring oxygen atoms and the wavenumber of the OH-stretching vibrations was found [119]. This correlation supports the above-mentioned explanation of the LF wavenumber shift, νOH , of the stretching vibrations of hydroxyl groups located in small structural units of zeolites. An important parameter inﬂuencing the stretching vibrations of noninteracting bridging OH groups in zeolites is the mean Sanderson electronegativity, Sm , of the zeolite framework (see Section 17.2.1, Eq. (17.1)) [33, 34, 134]. In dependence on the nSi /nAl ratio of the zeolite, the mean electronegativity increases with decreasing aluminum content or increasing silicon content in the framework because of the higher electronegativity of silicon in comparison with aluminum atoms. Correspondingly, the stretching vibrations of SiOHAl groups in zeolites characterized by a high nSi /nAl ratio (high mean electronegativity) occur at lower wavenumbers than those of zeolites with low nSi /nAl ratio (low mean electronegativity). Jacobs and Mortier [134] found a linear correlation between the stretching vibrations of most of the noninteracting bridging OH groups in dehydrated zeolites and Sm . In the corresponding plot, the wavenumbers of hydroxyl groups in small structural units of zeolites give parallel shifted curve. In principle, the concentration of different types of hydroxyl groups can be determined by the integral intensities of their IR bands. To calculate the OH concentration, the extinction coefﬁcient has to be determined by an independent measurement. The extinction coefﬁcient is a function of the wavenumber and varies with a wavenumber shift, for example, due to a variation of the aluminum content in the zeolite framework. A detailed survey on the extinction coefﬁcients of the stretching vibration of zeolitic OH groups and of probe molecules adsorbed on surface sites of zeolites, such as ammonia, carbon monoxide, pyridine, benzene, ethylbenzene, and acetonitrile, is given in [115, 135]. Considering these values, the extinction coefﬁcient, ε, of the stretching vibration of bridging OH groups in the supercages of H-Y zeolites (HF band) varies from ε = 3.1−12.2 cm µmol−1 [115]. Similarly, extinction coefﬁcients of ε = 1.3−3.0 and 1.3–3.3 cm µmol−1 were published for evaluating the integral band intensities obtained for pyridine

509

17 Catalytically Active Sites: Generation and Characterization

adsorption on Brønsted (1540 cm−1 ) and Lewis acid sites (1450 cm−1 ), respectively [115, 135]. The accuracy of concentrations of surface sites obtained by quantitative evaluation of IR band intensities, therefore, strongly depends on the accurate determination or choice of the extinction coefﬁcient. As an example, Figure 17.7 shows the stretching vibration spectra of a dealuminated zeolite H-FER (nSi /nAl = 30) recorded before (Figure 17.7a) and after the conversion of n-butene at 623 K (Figure 17.7b) [136]. As evidenced by the computer-simulation of the spectra in Figure 17.7, one band of silanol groups (3740 cm−1 ), up to two different bands of AlOH groups at nonframework aluminum species (3715 and 3673 cm−1 ), and up to four bands of bridging OH groups

Absorbance (arb. units)

0.16

0.12

0.08

0.04

0.00 3800

3700

3600

3500

3400

3300

3400

3300

Wavenumber (cm−1)

(a) 0.12 Absorbance (arb. units)

510

0.09

0.06

0.03

0.00 3800 (b)

3700

3600 Wavenumber

3500 (cm−1)

Figure 17.7 FTIR spectra of fresh (a) and spent (b) zeolites H-FER with nSi /nAl = 30. The spent catalyst was exposed to n-butene at 623 K for 1 hour with pn-butene /ptotal = 0.225.[136].

17.3 Characterization of Acid Sites

Absorbance (arb. units)

100

80

60

40

20

0 0

20 40 60 80 Cation exchange degree (%)

100

Figure 17.8 Integral band intensities obtained by deconvolution of the FTIR spectra of zeolite H-FER with nSi /nAl = 8 and cation exchange degrees of 15–100% (• terminal SiOH, bridging OH groups in 10-membered rings, bridging OH groups in 8-membered rings, bridging OH groups in 6-membered rings) [138].

in 10-, 8, 6-, and 5-membered oxygen rings (3590, 3580, 3528, and 3475 cm−1 , respectively) occur [136, 137]. A quantitative evaluation of the integral intensities of the IR bands of hydroxyl groups of a zeolite H-FER (nSi /nAl = 8) with different cation exchange degrees indicated that at ﬁrst the formation of bridging OH groups in the 10-ring channels occurs (Figure 17.8) [138]. With increasing cation exchange degree, there is an increasing formation of bridging OH groups in the 8- and 6-rings until the number of these hydroxyl groups is comparable with those in the 10-ring channels at the cation exchange degree of 100% [138]. To gain an insight into the different types of acid sites (i.e., Brønsted and Lewis acid sites) and their behavior and distribution in zeolites, probe molecules are applied. Often used molecular probes are pyridine (see above) and acetonitrile [137, 139–145]. In comparison with pyridine, acetonitrile is a weak base interacting with acid sites via the nitrogen lone pair of the C≡N group. The strength of acetonitrile adsorption with acid sites is reﬂected in the shift of the ν(C≡N) stretching vibration to higher frequencies [137]. In liquid, the ν(C≡N) stretching vibration band is at about 2265 cm−1 . Due to Fermi resonance between the ν(C≡N) stretching and δ(CH3 ) + ν(C−C) combination vibrations, this band can split into a doublet at 2294 and 2254 cm−1 [146]. In contrast, the fundamental νs (CH3 ) and νas (CH3 ) modes are not signiﬁcantly changed after acetonitrile adsorption on Lewis and Brønsted acid sites [137]. Generally, adsorption of acetonitrile at bridging hydroxyl groups gives the so-called ABC pattern in the FTIR spectrum. This pattern consists of two intense bands at about 2800 and 2400 cm−1 and one less intense band at about 1600 cm−1 . The ABC pattern has been assigned to very strong quasi-symmetrical hydrogen

511

512

17 Catalytically Active Sites: Generation and Characterization Table 17.4 ν(C≡N) stretching vibrations upon adsorption of acetonitrile on dehydrated zeolites.

ν OH (cm –1 )

Assignment

References

2265 2275–2280 2292–2297

Physisorbed acetonitrile Adsorbed at terminal SiOH groups Adsorbed at bridging OH groups (ε = 2.05 ± 0.1 cm µmol –1 for H-FER [137]) Adsorbed at Fe(II) ions acting as weak Lewis sites Adsorbed at nonframework aluminum species acting as weak Lewis acid sites (ε = 3.6 ± 0.2 cm µmol –1 for FER zeolite [137]) Adsorbed at nonframework aluminum species acting as strong Lewis acid sites (ε = 3.6 ± 0.2 cm µmol –1 for FER zeolites [137])

[137] [137, 144, 147] [137, 144, 147]

2303 2310

2318–2325

[141] [137]

[137, 144, 147]

The extinction coefﬁcients, ε, are valid for FER-type zeolites.

bonding, where the hydroxyl proton is partially but not completely transferred to the base [146]. As an example, Figure 17.9 shows characteristic FTIR spectra of H-FER zeolite calcined at 670 (Figure 17.9a) and 990 K (Figure 17.9b) and loaded with acetonitrile [143]. The bands occurring upon acetonitrile adsorption on the zeolites under study are summarized in Table 17.4. The spectra in Figure 17.9 are dominated by the ν(C ≡ N) stretching bands of acetonitrile adsorbed at Lewis and Brønsted acid sites at 2325 and 2297 cm−1 , respectively. It is obvious that the increase in the calcination temperature from 670 to 990 K leads to a signiﬁcant decrease in the band of acetonitrile adsorbed at Brønsted acid sites and an increase in the band due to probe molecules adsorbed at Lewis acid sites. Using the extinction coefﬁcients, ε, given in literature (Table 17.4, second column), the integral intensities of these bands were utilized for calculating the densities of Brønsted (670 K: 1.50 mmol g−1 ; 990 K: 0.20 mmol g−1 ) and Lewis acid sites (670 K: 0.08 mmol g−1 ; 990 K: 0.73 mmol g−1 ) [143]. In very recent studies, the slightly larger probe molecule pivalonitrile (0.6 nm [148]) was utilized for investigating the nature and accessibility of acid sites in zeolites SSZ-33, SSZ-35, and MCM-22[107, 148]. Bands occurring at 2239 and 2253 cm−1 in the FTIR spectra of pivalonitrile-loaded zeolite SSZ-33 indicate the presence of different types of SiOH groups (3732 and 3744 cm−1 ). A band at 2305 cm−1 was assigned to AlOH groups (3670 cm−1 ), while the main maximum at 2295 cm−1 was due to pivalonitrile interacting with acidic SiOHAl groups (3620 cm−1 ). The studies have shown that virtually all acidic bridging OH groups in the 10- and 12-ring pores of zeolites SSZ-33 are accessible for the rather bulky probe molecule [148]. In the FTIR spectra of zeolite MCM-22 (10-ring pores of 0.41 × 0.51 nm) loaded with pivalonitrile, the interaction of the probe molecules

17.3 Characterization of Acid Sites

Absorbance (arb. units)

1.6

B

1.2

L

0.8

0.4

0.0 2350

2250

2220

L

1.5 Absorbance (arb. units)

2300

Wavenumber (cm−1)

(a)

1.0 B 0.5

0.0 2350 (b)

2300 Wavenumber

2250 (cm−1

2220

)

Figure 17.9 FTIR spectra of zeolites H-FER (1.65 mmol Al per gram of dry zeolite) calcined at 670 K (a) and 990 K (b) (L: acetonitrile adsorbed at Lewis acid sites, B: acetonitrile adsorbed at Brønsted acid sites) [143].

with acidic SiOHAl groups causes bands at 2269 – 2272 cm−1 , while the band at 2295 cm−1 was explained by adsorption on Lewis acid sites [107]. Furthermore, adsorption of probe molecules is also an important approach for the study of the strength of Brønsted acid sites in zeolites. Linear correlations exist between the strength of Brønsted acid sites in solid catalysts and the wavenumber shift, νOH , of the stretching vibrations of OH groups upon adsorption of benzene [149], acetone [150], CO [151, 152], and ethene [153]. For Low Temperature (T ∼ = 100 K) adsorption of CO on different Y-type zeolites, wavenumber shifts of the HF band of bridging OH groups in the following sequence of acid strength were observed: H,Li,Na-Y (νOH = 302 cm−1 ) > H,K,Na-Y (νOH = 220 cm−1 ) > H,Rb,Na-Y (νOH = 68 cm−1 ) > H,Cs,Na-Y (νOH = 160 cm−1 ) [151].

513

514

17 Catalytically Active Sites: Generation and Characterization

Characteristic changes of the stretching vibrations of CO and H2 occur also for their adsorption at Lewis acid sites [154]. Upon adsorption of CO at Lewis acid extra-framework aluminum species in zeolite H-ZSM-5, a doublet at νCO = 2230 and 2220 cm−1 corresponding to HF band shifts in comparison with the vibration mode of the isolated CO of νCO = +87 and +77 cm−1 , respectively, was observed. Using H2 as probe molecule, the spectrum consist of a doublet at νHH = 4027 and 4002 cm−1 corresponding to band shifts of νHH = −133 and −158 cm−1 , respectively. Comparison of these values indicates that H2 is a superior probe molecule for evaluating Lewis acid sites of zeolites since its spectroscopic response is nearly twice that of CO [154]. 17.3.6 NMR Spectroscopy

NMR spectroscopy is a suitable method for characterizing Brønsted acid sites of solids upon adsorption of probe molecules or in a direct manner. For this purpose, zeolite samples dehydrated at elevated temperatures in vacuum are ﬁlled in gas-tight magic-angle spinning (MAS) rotors or sealed in glass inserts [32, 35]. The 1 H MAS NMR signals of hydroxyl groups in dehydrated solid catalysts cover a range of chemical shifts of δ1H = 0−16 ppm (Table 17.5) [32, 155]. The lowest chemical shifts of about 0 ppm have been observed for isolated metal and cation OH groups, such as AlOH, LaOH, and MgOH groups in the large cages of dehydrated zeolites or at the outer surface of solid particles [32, 155]. While the 1 H MAS NMR signals of isolated SiOH groups occur at 1.3–2.2 ppm, strong hydrogen bonding causes signals at chemical shifts of up to 16 ppm [156, 157]. The 1 H MAS NMR spectra of dealuminated zeolites contain signals at δ1H = 2.4–3.6 ppm caused by hydroxyl protons bound to extra-framework aluminum species (AlOH) [45]. This assignment was supported by 1 H/27 Al TRAPDOR NMR experiments (transfer of population in double resonance (TRAPDOR)) [162, 166, 170, 171], which allow the detection of the dipolar coupling between 1 H and 27 Al nuclei. The resonance position of AlOH groups at 2.4–3.6 ppm indicates their location in narrow channels or small cages. Adsorption of probe molecules [32, 35, 155, 172, 173] and application of the 1 H/27 Al TRAPDOR NMR technique [162, 166, 171] evidenced that signals occurring at δ1H = 3.6–4.3 ppm are due to unperturbed bridging OH groups in dehydrated H-form zeolites. In the 1 H MAS NMR spectra of dehydrated zeolite H,Na-Y, the signal of bridging OH groups in the supercages appears at δ1H = 3.6–4.0 ppm, while the signal at δ1H = 4.8–5.2 ppm is due to bridging OH groups in the small sodalite cages [32, 155]. The larger chemical shift of the bridging OH groups in the sodalite cages is explained by an interaction of these hydroxyl protons with neighboring framework oxygen atoms. Similarly to the wavenumbers of the stretching vibrations of bridging OH groups (see Section 17.3.5), also the 1 H NMR shift values of unperturbed SiOHAl groups in large structural units, such as supercages of zeolites X and Y or 10- and 12-ring pores of H-ZSM-5 and H-mordenite, depend on the framework nSi /nAl ratio

17.3 Characterization of Acid Sites Chemical shifts, δ1H , of 1 H MAS NMR signals caused by hydroxyl groups in dehydrated zeolites and their assignments.

Table 17.5

δ1H (ppm)

OH group

Assignment

References

–0.5 to 0.5

Metal OH

[32, 158]

1.3–2.2

SiOH

2.4–3.6

Metal OH* AlOH*

2.8–6.2

Cation OH* CaOH* LaOH*

3.6–4.3

SiOHAl, SiO1HAla

4.6–5.2

SiOHAl SiO3HAla

5.2–8.0

SiOHAl*

10–16

SiOH*

Metal or cation OH groups (e.g., AlOH, MgOH) in the large cages of zeolites or at the outer surface of solid particles Silanol groups at the external surface of solid particles or at framework defects in zeolites Metal OH groups involved in hydrogen bonding, such as extra-framework aluminum species in zeolites Cation OH groups involved in hydrogen bonding, such as OH groups located in the sodalite cages of zeolite Y Bridging OH groups in large structural units, such as in the supercages of zeolite Y Bridging OH groups in small structural units, such as in the sodalite cages of zeolite Y Disturbed bridging OH groups in zeolites H-ZSM-5 and H-beta involved in hydrogen bonding with neighboring framework oxygen atoms Internal SiOH groups involved in strong hydrogen bonding

[44, 45, 159–162]

[44, 45, 159–164]

[32, 44, 45, 165]

[44, 45, 159–165]

[44, 45, 159, 162, 164] [160, 161, 166–169]

[156, 157]

The resonance positions are referenced to tetramethylsilane (δ1H = 0 ppm).‘‘*’’ indicates hydrogen bonding or electrostatic interaction. a Assignment of bridging OH groups in zeolites X and Y.

[32, 35, 155]. Generally, an increasing chemical shift, δ1H , is observed with decreasing framework aluminum content or increasing mean electronegativity, Sm , of the zeolite framework [32, 35, 155]. As an example, Figure 17.10 shows 1 H MAS NMR spectra of dehydrated zeolites [165]. The 1 H MAS NMR spectrum in Figure 17.10a is that of an H-form ferrierite with nSi /nAl = 9.1. It consists of two signals at 1.8 and 4.2 ppm due to unperturbed silanol groups (SiOH) and bridging OH groups (SiOHAl), respectively, the latter acting as Brønsted acid sites. After dealumination of this zeolite by an HCl treatment [165], an additional signal appears at 2.8 ppm (Figure 17.10b), which was

515

516

17 Catalytically Active Sites: Generation and Characterization

SiOHAl

AlOH

SiOHAl

SiOH H-bridged SiOH

SiOH

10 (a)

8

6

4

2

d 1H (ppm)

0

−2

−4

10

8

6

(b)

4

2

0

−2

−4

d 1H (ppm)

Figure 17.10 1 H MAS NMR spectra of dehydrated H-form ferrierites (H-FER) recorded before (a) and after (b) dealumination by treatment with HCl [165].

assigned to hydroxyl groups bound to extra-framework aluminum species (AlOH∗ ). The signal at about 6 ppm indicates the presence of silanol groups interacting with neighboring oxygen atoms (SiOH∗ ). Since 1 H nuclei have a spin I = 12 , the 1 H MAS NMR intensities can be utilized in a direct manner for calculating the concentration of hydroxyl groups in dehydrated zeolites by comparison with the intensity of an external standard, such as a well-characterized H-form zeolite. For quantitative studies, the repetition time of the pulse experiments has to be large (about factor 3 – 5) in comparison with the spin-lattice relaxation times T1 of the different OH species, which have values of 1–10 seconds [174]. The concentration of the SiOHAl groups in the parent H-form ferrierite (Figure 17.10a) was determined to 1.4 mmol H+ g−1 , while the acid leaching of framework aluminum atoms caused a strong decrease in this value to 0.4 mmol H+ g−1 (Figure 17.10b) [165]. The high-ﬁeld shift (larger chemical shift values) of OH groups in small structural units of zeolites is a hint for the presence of hydrogen bonding or electrostatic interactions of the hydroxyl protons under study with neighboring framework oxygen atoms. In Table 17.5, the corresponding hydroxyl groups are marked by an asterisk. Comparison of structural data and 1 H NMR shift values of hydroxyl protons in –O–H· · ·O– arrangements led to the following empirical formula describing the dependence of the low-ﬁeld resonance shift, δ1H , of hydroxyl protons on the oxygen–oxygen distance dOH-O given in ‘‘picometers’’ [175] δ1H /ppm = 79.05 − 0.255 dOH−O /pm

(17.9)

The 1 H MAS NMR spectra of siliceous as-synthesized zeolites ZSM-51, ZSM-58, ZSM-12, SSZ-24, and ZSM-5 prepared with organic quaternary ammonium

17.3 Characterization of Acid Sites

cations consist of a signal at 10.2 ppm [156]. The chemical shift of this signal was explained by ≡SiOH · · · OSi≡ hydrogen bonds between hydroxyl protons of defect silanol groups with neighboring framework oxygen atoms in an O· · ·O distance of 0.27 nm. According to Eq. (17.9), the signal at 15.9 ppm in the 1 H MAS NMR spectra of RUB-18 indicates a strong hydrogen bonding of silanol groups with an O · · · O distance of 0.25 nm [157]. Like for zeolites X and Y, low-ﬁeld shifted signals of bridging OH groups were observed in the 1 H MAS NMR spectra of dehydrated zeolites H-ZSM-5, H-beta, and H-MCM-22 [167, 169, 174, 176]. Also in these zeolites, an interaction of the hydroxyl protons of bridging OH groups with neighboring framework oxygen atoms is assumed (SiOH*Al). A survey on the most important probe molecules applied for studying the properties of Brønsted and Lewis acids sites by solid-state NMR spectroscopy is given in Table 17.6. The ability to protonate strongly basic probe molecules, such as pyridine and trimethylphosphine (TMP), or to form hydrogen bonds to these molecules is utilized to distinguish between acidic and nonacidic surface OH groups. A more quantitative comparison of the acid strength of acid hydroxyl groups is possible using weakly basic probe molecules, such as acetone and acetonitrile, which interact via hydrogen bonding. The adsorbate-induced low-ﬁeld shifts, δ, of the NMR signals caused by the interacting hydroxyl protons (1 H MAS NMR) or due to the functional groups of the probe molecules (13 C, 15 N, or 31 P MAS NMR) is a measure of the strength and accessibility of the corresponding surface OH groups. Some of the probe molecules listed in Table 17.6 are suitable for studying the location of OH groups at the outer surface of the zeolite particles or inside the pores and cages. Perﬂuorotributylamine (diameter ≈ 0.94 nm [160]), which is too large to enter the micropores of zeolites (H-ZSM-5: 0.53 × 0.55 nm; H-Y: 0.74 nm [38]), allows the study of hydroxyl groups at the outer surface of zeolite particles. Upon adsorption of perﬂuorotributylamine at SiOH and AlOH groups, low-ﬁeld resonance shifts of the corresponding 1 H MAS NMR signals by δ1H = 0.25 ppm and δ1H = 0.47 ppm, respectively, indicate a location of these hydroxyl groups at the outer surface or in secondary mesopores [160, 161]. Often, no low-ﬁeld resonance shift was found for the signals of bridging OH groups, which indicates their location inside the pores [160, 161]. Another approach is to utilize triphenylphosphine for investigating the location of bridging OH groups acting as Brønsted acid sites [196]. In this way, it was found that 6% of the Brønsted acid sites of zeolite H-MCM-22 are located at the outer particle surface. In a number of studies, TMP was used as probe molecule for identifying strong Brønsted acid sites [187–192]. The 31 P MAS NMR spectrum of TMP adsorbed on zeolite H,Na-Y is dominated by a signal at −2.5 ppm (referenced to 85% H3 PO4 ). This signal is due to (CH3 )3 PH+ complexes arising from chemisorption of TMP at Brønsted acid sites [188]. Upon strong calcination, additional 31 P MAS NMR signals in the region from −32 to −67 ppm indicate the coordination of TMP at Lewis acid sites [188]. With increasing calcination temperature, a signiﬁcant increase

517

518

17 Catalytically Active Sites: Generation and Characterization Table 17.6

Probe molecules suitable for studying Brønsted and Lewis acid sites in zeolites.

Probe molecule Brønsted sites Pyridine-d5

Acetonitrile-d3

Perﬂuorotributyl amine Deuterated alkanes and aromatics 13 C-2-acetone

Trimethylphosphine (TMP) Trimethylphosphine oxide (TMPO) Triphenylphosphine (PPh3 ) Lewis sites 13 C-2-acetone Carbon monoxidea Trimethylphosphine (TMP) Trimethylphosphine oxide (TMPO)

Resonance/observation

References

1 H: hydrogen-bonded pyridine at about δ1H = 10 ppm (SiOH) and pyridinium ions at δ1H = 12 − 20 ppm (SiOHAl) 1 H: adsorbate-induced low-ﬁeld shift by δ1H = 3.6 (H,Na-X) to 7.9 ppm (H-ZSM-5) 1 H: adsorbate-induced low-ﬁeld shift of accessible OH groups by δ1H = 0.25 (SiOH) to 0.47 ppm (AlOH) 1 H: activation energy of the H/D exchange

[22, 32, 44, 45, 66, 154]

[179–183]

13 C:

[184–186]

hydrogen-bonded acetone at δ13C = 216.8 (H-SAPO-5) to 225.4 ppm (H-ZSM-22) 31 P: δ 31P = −2 to − 3 ppm for TMP protonated by strong acid sites 31 P: hydrogen-bonded TMPO at δ31P = 53 (H-Y) to 63 ppm (USY) 31 P: δ 31P = 11.1 to 14.8 ppm for PPh3 at accessible acid sites

δ13C = 233 ppm for acetone at zeolite USY 13 C: δ 13C ≈ 770 ppm for CO at dealuminated zeolite H-ZSM-5 31 P: δ31P = −32 to −67 ppm for TMP at dealuminated zeolites 31 P: δ 31P = 37 ppm for TMPO at dealuminated zeolites Y and γ -Al2 O3 13 C:

[44, 45, 177–179]

[160, 161]

[187–192] [193–195] [196]

[185, 186] [197, 198] [189, 190, 199] [194, 200]

exchange requires NMR measurements at low temperatures (about T = 120 K). and 13 C NMR shifts are referenced to tetramethylsilane (δ1H = 0 ppm, δ13C = 0 ppm), while the 31 P NMR signals are related to 85% H PO (δ 3 4 13P = 0 ppm). a Rapid 1H

in these signals occurs, which indicates that the preparation of zeolite catalysts with a high concentration of Lewis acid sites requires an optimum calcination temperature, such as 873 K for zeolite H,Na-Y [188]. The local structure of adsorbate complexes consisting of TMP coordinated at Lewis acid sites in dehydroxylated zeolite H,Na-Y was investigated by 31 P/27 Al TRAPDOR NMR experiments [201]. The 31 P MAS NMR signal at −47 ppm could be assigned to TMP directly bound to an

17.3 Characterization of Acid Sites

aluminum Lewis acid site. Furthermore, the 27 Al/31 P insensitive nuclei enhanced by polarization transfer (INEPT) NMR technique was utilized to determine the 27 Al/31 P J-coupling [201, 202]. Upon adsorption of TMP on mixtures of AlCl3 and zeolite H,Na-Y, two different TMP-AlCl3 complexes were found [202]. 31 P/27 Al J-coupling constants and internuclear distances of 299.5 Hz and 0.258 nm and of 260 Hz and 0.296 nm were determined for AlCl3 molecules coordinated to one (fourfold coordinated aluminum) and two (ﬁvefold coordinated aluminum) TMP molecules, respectively [202]. When TMP is used as probe molecule, however, it must be considered that this probe molecule is rather bulky (with a kinetic diameter of 0.55 nm [188]). Therefore, the characterization of surface sites in zeolites by TMP is restricted by the maximum adsorption capacity of the pore system as well as the pore diameter of the zeolite under study [189]. Similar limitations exist for the application of trimethylphosphine oxide (TMPO) as probe molecule [193–195]. Upon adsorption of pyridine at acidic bridging OH groups, the total proton transfer to the probe molecule leads to the formation of pyridinium ions (PyrH+ ). Calculations of the 1 H NMR shifts of PyrH+ protonated by bridging OH groups in 8T zeolite clusters gave a linear correlation between the δ1H values and the proton afﬁnity of the corresponding acid site [203]. In the case of total proton transfer to the pyridine molecule, the smaller δ1H values are obtained for the smaller proton afﬁnities of the SiOHAl groups corresponding to the higher acid strength and vice versa. Experimental studies of the interaction of perdeuterated pyridine with SiOHAl groups yielded signals of PyrH+ at 1 H NMR shifts of 16.5 ppm for zeolite H,Na-Y and at 15.5–19.0 ppm for zeolite H-ZSM-5 [22, 32, 35, 44, 45, 66, 155]. A frequently applied probe molecule for characterizing the strength of Brønsted acid sites in zeolites by solid-state NMR spectroscopy is 13 C-2-acetone [184–186]. Using this molecular probe, Biaglow et al. [184] investigated the acid strength of bridging OH groups in acidic zeolites by observing the 13 C NMR shifts of the carbonyl atom in 13 C-2-acetone and obtained values of δ13C = 216.8 (H-SAPO-5), 219.6 (H-Y), 221.8 (H-MOR), 222.8 (H-[Ga]ZSM-5), 223.4 (H-ZSM-12), 223.6 (H-ZSM-5), and 225.4 ppm (H-ZSM-22). These chemical shift values have to be compared with the shift values of 205 and 245 ppm for carbonyl atoms of 13 C-2-acetone dissolved in CDCl3 and in 100% sulfuric acid solution, respectively [183]. Furthermore, a scale of the Brønsted acid strength was introduced, which is based on the experimentally determined dependence of the resonance positions of carbonyl atoms in 13 C-2-acetone dissolved in aqueous sulfuric acid of varying concentration [204]. According to this scale, the acid strength of bridging OH groups in zeolite H-ZSM-5 (δ13C = 223.6 ppm [184]) corresponds to 80% H2 SO4 in water [204]. A limitation of the application of acetone as probe molecule is its reactivity in the case of an adsorption at strongly acidic sites leading to the formation of mesityl oxide. Recently, the weak base acetonitrile-d3 found a broad application as probe molecule for the characterization of the strength of Brønsted acid sites since it is very sensitive and can be studied by 1 H MAS NMR spectroscopy [44, 45, 177–179]. For an interaction of the probe molecule via hydrogen bonding, the larger adsorbate-induced low-ﬁeld shift, δ 1H , of the 1 H MAS NMR signal of zeolitic hydroxyl groups corresponds to the higher acid strength and vice versa. As an

519

520

17 Catalytically Active Sites: Generation and Characterization

3.6

8.5

∆d1H = 4.9 ppm

2.5

6.0 4.6

1.8

+ CD3CN 14 (a)

12

10

8

6 4 d1H (ppm)

2

0

−2

−4

3.9

9.6

∆d1H = 5.7 ppm 4.8 6.3 1.8

+ CD3CN

14 (b)

12

10

8

6 4 d1H (ppm)

2

0

−2

−4

Figure 17.11 1 H MAS NMR spectra of zeolites La,Na-X/75 (a) and La,Na-Y/74 (b) recorded before (top) and after (bottom) adsorption of deuterated acetonitrile [44].

example, Figure 17.11 shows the 1 H MAS NMR spectra of lanthanum-exchanged zeolites X (nSi /nAl = 1.3) and Y (nSi /nAl = 2.7) recorded before (top) and after (bottom) adsorption of acetonitrile-d3 [44]. The signals at 1.8 and 3.6 – 3.9 ppm are due to SiOH and bridging OH groups (SiOHAl), respectively. The above-mentioned SiOHAl groups are located in the supercages, while bridging OH groups in sodalite cages causes weak shoulders at 4.6 – 4.8 ppm. Hydroxyl groups at lanthanum cations (La(OH)n ) and oxide complexes located in the sodalite cages are responsible for the signals at 5.6 – 6.3 ppm. The signal at 2.5 ppm occurring in the spectrum of zeolite La,Na-X/75 indicates the presence of La(OH)n groups in the supercages. Upon adsorption of deuterated acetonitrile (Figure 17.11, bottom), the 1 H MAS NMR signals of accessible bridging OH groups in the supercages of zeolites La,Na-X/75 and La,Na-Y/74 (La-exchange degrees of 75% and 74%,

17.4 Base Catalysis

respectively) shift by δ1H = 4.9 ppm and δ1H = 5.7 ppm, respectively, to lower magnetic ﬁeld [44]. These adsorbate-induced resonance shifts have to be compared with the low-ﬁeld shifts, δ1H , obtained for acetonitrile-loaded zeolites H,Na-X/62 (3.6 ppm [45]), Al,Na-X/32 (3.8 ppm [45]), La,Na-X/42 (3.8 ppm [44]), Al,Na-X/61 (4.4 ppm [45]), H,Na-Y/91 (5.1 ppm [179]), Al,Na-Y/34 (5.3 ppm [45]), Al,Na-Y/64 (5.3 ppm [45]), H-MOR with nSi /nAl = 10 (6.7 ppm [177]), dealuminated H,Na-Y with nSi /nAl = 18 (7.0 ppm [177]), and H-ZSM-5 with nSi /nAl = 26 (7.9 ppm [179]). The above-mentioned adsorbate-induced resonance shifts, δ1H , indicate that the cation type, exchange degree, and framework nSi /nAl ratio are important parameters for adjusting the acid strength of zeolites.

17.4 Base Catalysis 17.4.1 Nature of Base Sites

In contrast to the broad application of acidic zeolites as solid catalysts in chemical technology [1–7], much less attention has been paid to basic microporous and mesoporous materials, even though these solids do have a considerable potential for a number of industrially important reactions [205]. Analyzing the industrial application of solid acid and base catalysts showed that in only 8% of the processes solid bases are being employed, and none of these reactions is performed with basic zeolite catalysts [206]. It is only with a growing effort in the preparation, characterization, and catalytic investigation of new microporous and mesoporous materials with basic properties that a breakthrough in their application may be achieved. The nature of basic sites in zeolites is less well-deﬁned than that of acid sites. This is related to the fact that most basic zeolites, which contain alkali metal cations, act as weak Lewis acid sites as well as basic framework oxygen atoms. In most reactions catalyzed by basic zeolites, both Lewis acid sites and base sites are involved. The strength of the base sites must be high enough to stabilize anionic or polarized species that take part in the catalytic cycle [9, 14]. According to the Sanderson principle of equalization of electronegativities for framework atoms of zeolites (see Section 17.2.1), the partial charge on framework atoms and cations in zeolites can be calculated [33, 34]. Barthomeuf [207, 208] utilized the Sanderson principle to estimate the charge, –qO , at the framework oxygen atoms in zeolites exchanged with different alkali metal cations. For a given framework aluminum content, Barthomeuf [207, 208] found increasing charges –qO , which was attributed to an increase in the base strength in the sequence Li- < Na- < K- < Rb- < Cs-exchanged zeolites. Generally, the basicity of the framework oxygen atoms increases with decreasing cation electronegativity and, for a given cation-type, with an increasing framework aluminum content. The latter is due to the lower electronegativity of aluminum in comparison with silicon atoms [33].

521

522

17 Catalytically Active Sites: Generation and Characterization

Hence, the zeolite catalyst with the cation of the lowest electronegativity, that is cesium, and the highest framework aluminum content is the catalyst with highest base strength, such as cesium-exchanged zeolite X [208]. 17.4.2 Formation of Base Sites

Alkali-exchange of zeolites in aqueous solution or by solid-state ion exchange leads to materials, which possess basic framework oxygen atoms of relatively low base strength. In some cases, however, this property is required because of their resistance to poisoning by water or carbon dioxide and the rapid desorption of product molecules leading to a suppression of side-reactions [9]. For these applications, not only cesium- and rubidium-exchanged zeolites X and Y [207] but also zeolite Cs-ETS-10 are suitable catalysts [209]. In other cases, the low base strength of alkali-exchanged zeolites limits their applicability in organic syntheses. A suitable way for generating stronger base sites in zeolites is the impregnation with alkali metal salts, for example, cesium hydroxide. Hathaway and Davis [210] prepared intrazeolitic alkali oxide clusters, such as CsOx , by impregnation of cesium-exchanged zeolites X and Y, which are signiﬁcantly stronger than the zeolitic framework. The preparation of intrazeolitic KOx clusters by impregnation of potassium-exchanged zeolite X leads to materials with low basicity [211]. Another method is the impregnation of zeolite Y with NaN3 followed by controlled thermal decomposition yielding tetrahedral Na4 3+ clusters in the sodalite cages, Nay 0 clusters in the supercages, and metallic clusters on the external surface [212]. In oleﬁn isomerization, side chain alkylation of toluene, and aldol condensations, a correlation between the concentration of the Nay 0 clusters in the supercages and the catalytic activity was observed. However, these materials are sensible toward contact with air or water, which limits their application in catalysis [9]. Ono and Baba [213] developed the following procedure for obtaining basic zeolites: alkali-exchanged zeolite Y was immersed in a solution of metallic Na, Yb, or Eu in liquid ammonia, and the solvent was removed by evacuation. Heating in vacuum at about 450 K led to zeolites with a base strength, which strongly depends on the type of additional alkali metal cations and the amount of guest compounds. For zeolite Eu/K-Y, a maximum initial rate in the oleﬁn isomerization occurred at a Eu-loading of 8 wt% [213]. EXAFS investigations of zeolite Yb/K-Y revealed that the local structures of Yb species change drastically upon evacuation at around 500 K, that is, from a highly dispersed state to aggregated particles [214]. Since the micropores of basic zeolites prevent bulky molecules from reaching the active sites, mesoporous MCM-41 materials were used as carriers for basic guest species [215, 216]. By impregnation of MCM-41 with cesium acetate in aqueous or methanolic solutions and subsequent calcination, highly dispersed cesium oxide clusters were obtained in the mesopores of the carrier as long as the cesium content did not exceed 10 wt%. However, the impregnated material did not show a good thermal and chemical stability. After repeated calcination or after the use as catalyst,

17.5 Characterization of Base Sites in Zeolites

aggregation of the cesium oxide particles and a signiﬁcant decrease in the speciﬁc surface area were found. More stable materials were obtained by impregnating MCM-41 simultaneously with cesium acetate and lanthanum nitrate [217, 218]. In this case, a CsLaOx guest compound was formed in the channels of MCM-41, the base strength of which was, however, lower than that of the CsOx guest oxide. Yang et al. [219] investigated the effect of the substitution of framework aluminum by boron and gallium on the basicity of zeolites beta and faujasites, respectively. The observed sequence of basicity was [B]- < [Ga]- < [Al]-zeolites. The same sequence was obtained for quantum-chemical calculations of the binding energy of NO+ ions [220], which are used as FTIR spectroscopic probes for studying the base strength of zeolites. A number of groups utilized the nitridation of the framework of aluminosilicate-type zeolites [221, 222], SAPOs [223], and mesoporous materials [224] for the preparation of basic catalysts. The standard procedure is the treatment of Na- or H-form zeolites or siliceous mesoporous materials in an ammonia ﬂow at temperatures of about 1073 K or higher for 5 – 72 hours. Different surface reactions, such as ≡Si−OH + NH3 −→ ≡Si−NH2 + H2 O

(17.10)

≡Si−(OH)−Al ≡ + NH3 −→ ≡Si−(NH2 )−Al ≡ + H2 O

(17.11)

and the formation of ≡Si-(NH)-Si≡ are discussed [222]. Nitridated zeolite beta was found to have a much higher activity in base-catalyzed reactions in comparison with ammonia-treated aluminophosphates, SAPOs, and mesoporous materials [222].

17.5 Characterization of Base Sites in Zeolites 17.5.1 Test Reactions

The selectivities of several reactions are utilized for studying the presence of base sites and their base strength. Since base-catalyzed reactions have relatively low rates in comparison with acid-catalyzed reactions, minor traces of acidic protons, for example, due to silanol groups, may change the selectivity of the reaction dramatically. In order to overcome this problem, basic zeolite catalysts are often prepared with a slight excess of alkali metal cations. Double bond isomerization of 2,3-dimethylbut-1-ene and 1-butene leading to 2,3-dimethylbut-2-ene and 2-butene, respectively, is utilized for characterizing strong solid bases at low temperatures [211, 225–228]. Handa et al. [226] compared the isomerization of 2,3-dimethylbut-1-ene to 2,3-dimethylbut-2-ene in liquid phase and the decomposition of 2-methyl-3-butyn-2-ol to acetone and acetylene in vapor phase on several basic catalysts, such as on alumina-supported alkali metal compounds, metal oxides, mixed oxides, and zeolite K-Y. It was found that the selectivities in the decomposition of 2-methyl-3-butyn-2-ol do allow the discrimination

523

524

17 Catalytically Active Sites: Generation and Characterization

+H+ −H+

−H+ +H+ −H+

1-Butene +H

cis-2-Butene

+H+

+

−H+

trans-2-Butene Figure 17.12 Scheme of the isomerization of 1-butene to cis- and trans-2-butene on solid catalysts [9].

between acidic and basic catalysts. The isomerization of 1-butene (Figure 17.12) gives a measure of the base strength with high cis/trans ratios of the obtained 2-butenes for strong base sites [211, 225]. While the isomerization is a useful test reaction for determining the relative activities of strong solid bases, such as alkali oxide impregnated zeolites, only very little or no conversion was found for alkali-exchanged zeolites [211, 225]. Two decades ago, Dessau [229] introduced the dehydration of acetonylacetone as catalytic test reaction. Under speciﬁed conditions, the ratio of the selectivities to methylcyclopentenone and dimethylfuran is taken as a measure of the base strength. Alcaraz et al. [230] showed that the dehydration of acetonylacetone allows the catalytic characterization of materials exhibiting acidic as well as basic sites over a broad range of acid and base strengths. The most often utilized test reaction for basic catalysts is the Knoevenagel condensation (Figure 17.13) [221–223, 231–234]. The important advantage of this liquid-phase reaction is that it can be performed with reactants having different acidities. However, problems with diffusional hindrance may limit its application for the characterization of basic zeolites. Corma et al. [231] studied the Knoevenagel condensation of benzaldehyde with cyanoacetate, ethylacetoacetate, and ethylmalonate on alkali-exchanged zeolites X and Y. They found an order of the reactivity, which agrees with the increase in the charges on the framework oxygen atoms of the zeolite catalysts under study as estimated by the mean framework electronegativities: Li- < Na- < K- < Cs- and Y- < X-type zeolites. Corma et al. [231] concluded that most of the base sites in alkali-exchanged zeolites Y and X have pKb ≤ 10.3 and sites with pKb ≤ 13 are present in zeolite Cs-X only. This R1 O R2

+

H2C

CN

R1

CN

Y

R2

Y

+

H2O

R1 = Ph, PhCH=CH, 2-MeOC6H4, 2-furyl R2 = H, Me Y = CN, CO2Et Figure 17.13

Scheme of the Knoevenagel condensation on basic catalysts [9].

17.5 Characterization of Base Sites in Zeolites

catalyst was found to be more active than pyridine (pKb ≤ 8.8) and less active than piperidine (pKb ≤ 11.1). By comparing the Knoevenagel condensation on zeolite Na-X and germanium-substituted faujasite (nAl /nGe = 1.03), it was shown that the latter catalyst is more active [235]. It was concluded that most of the base sites in the germanium-modiﬁed zeolite have pKb ≈ 11.2 and additional sites with pKb ≤ 13.3 exist. By Knoevenagel condensation of benzaldehyde with malononitrile, Ernst and coworkers demonstrated the high basicity of zeolite beta, which was nitridated via an HT treatment with ammonia [222]. Very recently, the basicity of alkylammonium-exchanged zeolites X and Y was studied with the Knoevenagel condensation leading to a higher activity in comparison with the cesium-exchanged form [236, 237]. 17.5.2 Analytical and Spectroscopic Methods

For characterizing base sites in zeolites, the most important method is the use of molecular probes and their study by TPD [211, 238, 239], FTIR spectroscopy [240–243], and NMR spectroscopy [9, 35, 172]. Li and Davis [211, 239] utilized step-wise temperature-programmed desorption (STPD) and adsorption calorimetry of CO2 for determining the number and strength of base sites in zeolites CsOx /Cs-X and CsOx /K-X. The authors found that 80% of the activity of the above-mentioned zeolites in the isomerization of 1-butene is due to base sites responsible for CO2 desorption in the temperature range of 673 – 773 K. However, the content of these base sites was only about 5% of the total amount of base sites in the catalysts. Generally, adsorption of CO2 on alkali-exchanged zeolites is not always straightforward: different adsorbate structures may occur, and on strongly basic guest compounds, surface carbonates are formed [240]. Both these effects complicate the TPD curves and FTIR spectra of CO2 on basic zeolites [244]. Knoezinger and Huber [243] published a survey on the application of carbon monoxide, pyrrole, acetylenes, and deuterated chloroform as FTIR probes for the investigation of basic solids. For carbon monoxide adsorbed on alkali-exchanged zeolites Y, a correlation between the wavenumber shift, ν CO , of the stretching frequency and the cation radius of the exchanged cations was found, which has the inverse sequence of their electronegativity [243]. The main drawback of C–H and N–H acids used as probe molecules is the possible dissociation of the C–H or N–H bonds on strongly basic surface sites limiting the detection of strong base sites. The adsorption of pyrrole at alkali metal ions (Li+ , Na+ , K+ , Rb+ , Cs+ ) in zeolite ZSM-5 has been studied by FTIR spectroscopy in combination with quantum-chemical methods [245]. It was found that this probe molecule interacts via the ring with the metal cations and is additionally stabilized by a hydrogen bond of the NH group with the zeolite framework. Two types of hydrogen bonds were observed: to the oxygen atoms of (i) Si-O-Al and (ii) Si-O-Si arrangements causing band shifts of >150 and 80–150 cm−1 , respectively. The general shape of the FTIR spectra of alkali metal-exchanged and pyrrole-loaded zeolites ZSM-5

525

17 Catalytically Active Sites: Generation and Characterization

1927

1903

depends on the framework nSi /nAl ratio, the size of the alkali metal ion, and the pyrrole coverage [245]. Recently, the stretching vibrations of NO+ ions formed by NO2 disproportionation or adsorption of N2 O4 on basic zeolites are utilized for characterizing the base strength of these materials [220, 246, 247]. As an example, Figure 17.14a shows the FTIR spectra of NO+ ions formed upon adsorption of

Rb61Na39-X

2083

1976

K76Na24-X

Na100-X Li88Na12-X

0.15

2200 2150 2100 2050 2000 1950 1900 1850 Wavenumber n(NO+) (cm−1)

(a) 2120

R2 = 0.9971

2080

Li+ n(NO+) (cm−1)

526

2040 2000 Na+

1960 1920

K+

Rb+

1880 5 (b)

15

25

35

Hardness h (Ve−1)

Figure 17.14 FTIR spectra of the stretching vibrations of NO+ ions formed upon adsorption of NO2 on various alkaline-form zeolites X (a). In (b), the wavenumber ν(NO+ ) is plotted as the function of the chemical hardness η of the alkaline cations [246].

45

17.5 Characterization of Base Sites in Zeolites

NO2 on zeolites Na-X, Li,Na-X, K,Na-X, and Rb,Na-X [246]. Depending on the cation type, the NO+ stretching vibrations cover a range of 2083 (Li,Na-X) to 1903 cm−1 (Rb,Na-X). According to Figure 17.14b, there is a linear correlation between the NO+ stretching vibrations and the hardness, η, of the alkali metal cations. Hard atoms and ions are more difﬁcult to polarize. Harder alkaline metal cations (Li+ , Na+ ) induce a smaller electronic charge on the framework oxygens than softer cations (K+ , Rb+ ) [246]. The latter is the reason for the higher base strength of zeolite Rb,Na-X in comparison with zeolite Li,Na-X. During the past decades, the development of superconducting magnets with magnetic ﬁelds of up to B0 = 22.3 T and new techniques of high-resolution solid-state NMR spectroscopy opened new possibilities of investigating 17 O nuclei in solid materials [248–251]. Using high magnetic ﬁelds, double oriented rotation (DOR) NMR, and two-dimensional multiple-quantum magic-angle spinning (MQMAS) NMR spectroscopy, the strong signal broadening caused by the second-order quadrupolar interaction of 17 O nuclei with a nuclear spin of I = 5/2 can be averaged [248–251]. For low-silica faujasites with nSi /nAl = 1 (LSX) containing Si–O–Al bridges only, the spectra show four lines due to oxygen atoms at the four nonequivalent crystallographic positions. The chemical shift distribution of 17 O nuclei with different numbers of aluminum atoms in the local structure, however, causes a signiﬁcant residual line width of 17 O DOR and MQMAS NMR signals of zeolites with nSi /nAl > 1, which is a serious limitation for obtaining highly resolved spectra [251]. Like for FTIR spectroscopy, there is a number of probe molecules for solid-state NMR spectroscopic studies of base sites in zeolites. A survey on these molecular probes is given in Table 17.7. An important disadvantage of probe molecules may be their reactivity in the presence of strong base sites and the formation of different adsorption structures complicating the assignment of spectra. An interesting method, which overcomes this problem, is the use of methoxy groups as spectroscopic probes, directly formed at the basic framework oxygens by a conversion of methyl iodide. Applying 13 C MAS NMR spectroscopy, Bosacek et al. [256, 258] found a correlation between the isotropic chemical shift of surface methoxy groups bound to zeolite oxygens in bridging positions and the mean Sanderson electronegativity, Sm , of the zeolite framework (Figure 17.15). According to this correlation, a low 13 C NMR shift of methoxy groups corresponds to a high base strength of the framework oxygen atoms. Methoxy groups bound at framework oxygens of alkali-exchanged zeolites Y and X cover a range of 13 C NMR shift of 54.0 – 56.5 ppm. For zeolites Y and X impregnated with alkali metal hydroxides, two high-ﬁeld signals of methoxy groups bound at strongly basic guest compounds were observed at 50.0 and 52.3 ppm [259]. According to Krawietz et al. [261], guest compounds formed by impregnating a zeolitic support with cesium hydroxide or acetate are a mixture of cesium oxide (Cs2 O), peroxide (Cs2 O2 ), and superoxide (CsO2 ). Sanchez-Sanchez et al. [253–255] applied trichloromethane, triﬂuoromethane, chlorodiﬂuoromethane, and pyrrole as NMR probes for basic zeolites. Upon adsorption of pyrrole on zeolites K-X, Na-X, Cs-Y, Li-X, K-Y, Na-Y, and Li-Y, the

527

528

17 Catalytically Active Sites: Generation and Characterization Table 17.7

Probe molecules used for solid-state NMR studies of base sites in zeolites.

Probe molecule

Resonance/observation

References

Trichloromethane

1 H:

[252]

Triﬂuoromethane Chlorodiﬂuoromethane

Pyrrole Methyl iodide

Nitromethane

1H

hydrogen-bonded trichloromethane at δ1H = 7.55 (Li-Y) to 8.23 ppm (Cs,Na-Y) 1 H: hydrogen-bonded trichloromethane at δ1H = 7.45 (H-Y) to 8.70 ppm (Na,Ge-X) 13 C: hydrogen-bonded 13 C-chloroform at δ13C = 77.9 (H-Y) to 81.7 ppm (Na,Ge-Y) 1 H: hydrogen-bonded triﬂuoromethane at δ1H = 6.62 (Li-Y) to 7.6 ppm (Cs,Na-Y) 1 H: hydrogen-bonded chlorodiﬂuoromethane at δ1H = 7.5 (Li-Y) to 8.4 ppm (Cs-X) 19 F: hydrogen-bonded chlorodiﬂuoromethane at δ19F = −77.5 (Li-X) to −72.5 ppm (Cs-X) 1 H: hydrogen-bonded pyrrole at δ1H = 8.4 (Li-Y) to 11.5 ppm (K-X) 13 C: methoxy groups occurring at δ13C = 58.5 (Na-ZSM-5) to 54.0 ppm (Cs,Na-X) 13 C: δ 13C = 102 − 112 ppm for nitromethane at mixed magnesium–aluminum oxides

[253] [253] [252] [254]

[254]

[255] [256–259]

[260]

and 13 C NMR shifts are referenced to tetramethylsilane (δ1H = 0 ppm, δ13C = 0 ppm)

hydrogen atoms at the ring positions are not inﬂuenced by the adsorbents and cause two 1 H MAS NMR signals at 6 – 7 ppm [255]. The 1 H NMR shift of the hydrogen atoms at the nitrogens, on the other hand, covers a range of 8.4 for pyrrole adsorbed on zeolite Li-Y to 11.5 ppm for pyrrole adsorbed on zeolite K-X (most basic zeolite under study). This resonance shift is due to the different base strengths of the framework oxygen atoms in various zeolites contributing to the H-bondings with the pyrrole molecules [255]. An important advantage of pyrrole as probe molecule is its remarkable sensitivity and good resolution of the MAS NMR spectra. Recently, chlorodiﬂuoromethane was demonstrated to be a suitable probe molecule for FTIR as well as solid-state NMR characterization of basic zeolites X and Y [254]. Upon adsorption of this probe molecule, for example, on zeolites Li-Y and Cs-X, the C–H stretching vibrations and the 1 H MAS NMR signal shifted from 3043 to 3020 cm−1 and from 7.5 to 8.4 ppm, respectively. 19 F MAS NMR spectroscopy of chlorodiﬂuoromethane on various basic zeolites led to spectra consisting of doublets due to the spin–spin coupling between 19 F and 1 H nuclei. The 19 F chemical shift values as well as the J(H,F) coupling constants of adsorbed

17.6 Metal Clusters in Zeolites

60

H-ZSM-5 Na-ZSM-5

Na-ZSM-5/Zhd

59 Na-MOR Cu,Na-Y Na-MOR

H,Na-Y/a.L.

d13C (ppm)

58

H-beta Cu,Na-Y

57

Zn-Y H,Na-MOR Mg-Y

Na-Y

H,Mg-Y/50

H,Na-Y/50St

Na-X

H,Na-Y/30

56 K-ZK5 Rb-X

55

Cs-Y

Na-LSX

54

Cs-X

2.2

2.4

2.6

2.8

Mean electronegativity S

3.0

3.2

m

Figure 17.15 Dependence of the 13 C MAS NMR shift, δ 13C , of methoxy groups bound to framework oxygens in bridging positions on the mean Sanderson electronegativity, Sm , of the zeolite framework [258].

chlorodiﬂuoromethane were found to depend on the type of the alkali cations in the zeolites X and Y under study [254]. 17.6 Metal Clusters in Zeolites 17.6.1 Nature of Metal Clusters

Among the most important zeolite catalysts for industrial applications are those with Brønsted acid sites as well as noble metal atoms or clusters. These bifunctional zeolites are utilized, for example, as catalysts for shape-selective hydrogenations [262], hydroalkylation, hydroisomerization, hydrocracking, catalytic reforming [263], and CO hydrogenation [264]. Among the possible metals, platinum, palladium, and rhodium are the most important hydrogenation/dehydrogenations compounds in bifunctional zeolites [8]. The above-mentioned metals form nanoclusters inside the zeolites cavities. DFT theory calculations of Pdn clusters (n = 1–13) in their neutral, negatively, or positively charged state indicated that the stable structures are often independent on the charge of the cluster [265]. For clusters with n ≤ 6, the stable structure

529

530

17 Catalytically Active Sites: Generation and Characterization

of the anionic and neutral clusters follows the basic concept of maximizing the atoms coordination; however, cationic clusters favor a planar conﬁguration. For Pdn clusters with n = 7 – 13, the stable structures converge for charged and neutral clusters. For n = 13, the isomeric cubo-octohedron is favored. In the case of platinum and rhodium, the smallest stable cluster is a 13-atom icosahedron having one atom in the center and 12 atoms in equivalent positions surrounding it or the isomeric cubo-octahedron. For a Pt13 cluster inside a zeolite cage, it is expected that the interactions between the platinum atoms and cage walls are much weaker than the Pt–Pt bonds [8]. Collective electronic properties, such as electric conductivity, thermal conductivity, and ferromagnetic susceptibility, change signiﬁcantly when the size of the metal clusters become smaller than the mean free path of the conducting electrons or the size of the ferromagnetic domain. These behaviors are related to the electronic band structure, which changes with the cluster size [266, 267]. For example, the energy required to remove an electron from a small cluster is larger than the work function of the bulk metal, but smaller than the ionization potential of a metal atom. The chemical characteristics corresponding to the catalytic properties of metal clusters appear less sensitive to the cluster size than electronic effects. However, when isolated metal atoms exist on the support or the metal cluster size reaches a limit, strong changes in the catalytic activity must be expected. In catalysis, metal clusters on a support have been considered as very small pieces of metals, but exposing highly unsaturated atoms. Atoms involved in chemisorption on macroscopic metal pieces may be also coordinatively unsaturated, but the extent of coordinative unsaturation is signiﬁcantly greater for small clusters. For more details concerning the nature and behavior of metal clusters in zeolites, see [8, 267–270]. 17.6.2 Formation of Metal Clusters

The most popular method for the preparation of metal clusters in zeolites is the three-step procedure starting with an ion exchange (see Section 17.2.2, Eq. (17.5)). For metals of high catalytic activity, such as Pt, Pd, and Rh, the ﬁrst step is the exchange of the complexed ions [Pt(NH3 )4 ]2+ , [Pd(NH3 )4 ]2+ , and [Rh(NH3 )5 (H2 O)]3+ in aqueous solution against the alkali-form of the zeolites. Due to the relative small pore diameters in comparison with the rather bulky coordination shell of the above-mentioned complexes, the exchange process often takes several days and temperatures above room temperature are required. If the exchange procedure is stopped before reaching the equilibrium, a gradient of the metal concentration can be obtained with larger amounts of the metal near the particle surface than in the inner part. The second step is the calcination, which is necessary to remove the water and the ligands from the exchanged metal ions. For this purpose, the exchanged zeolite

17.6 Metal Clusters in Zeolites

is heated in a high ﬂow of air or oxygen. Three phenomena may occur during the elimination of water and ammonia [8, 269]: 1) Formation of metal clusters and zeolite protons in the reducing atmosphere of the decomposing amine ligands. The reduction of the metal ions to metal clusters is followed by their growth, and, depending on the metal type, the formation of oxide particles is observed. 2) Ion migration from large cages to smaller cages of the zeolite structure may occur, which can be prevented by limiting the calcination temperature. The driving force is the high negative charge density in the zeolite cages. The migration back to large cages can be initiated by offering attractive ligands, such ammonia. 3) Hydrolysis of the metal ions may lead to a decrease in the positive charge number. Multivalent metal ions, such as Rh3+ , are converted to monovalent complexes, for example, (RhO)+ , and hydroxyl protons. These hydroxyl protons compensate negative framework charges; that means, they form bridging OH groups. Generally, uncontrolled autoreduction can be the reason of agglomeration of clusters leading to large metal particles. This can be minimized by calcination in pure oxygen, at a high ﬂow rate, and using a slow temperature ramp, such as 0.5 K min−1 . The third step is the controlled reduction of the metal ions typically performed in a ﬂow of hydrogen (see Section 17.2.2, Eq. (17.5)) [8, 269]. This process creates hydroxyl protons of bridging OH groups acting as Brønsted acid sites in addition to the metal atoms and clusters. The formation of hydroxyl protons can be minimized by limiting the calcination temperature leading to a sufﬁcient concentration of residual ammonia ligands within the zeolite. In the presence of oxo-ions, such as (RhO)+ , carbon monoxide is used for the reduction in the following manner + (17.12) RhO + CO −→ Rh+ + CO2 This reaction is also used to discriminate between bare ions and oxo-ions in the zeolites [270]. Bimetallic zeolites, such as (Pt+Cu) [271], (Pd+Co) [272], (Pd+Ni) [273], can be prepared utilizing the earlier described procedure. For the combinations (Pt+Re) and (Rh+Fe), the second metal was introduced by chemical vapor deposition of volatile carbonyls [8]. Also stable monometal clusters consisting, for example, of Rh and Ir were prepared by chemical vapor deposition of Rh(CO)2 (acac) and Ir(CO)2 (acac) precursors, respectively, followed by decarbonylation [274, 275]. For deeper insight into preparation of bimetallic clusters in zeolites, see [270]. For zeolites with small pores, especially for those with eight-membered oxygen rings, the solid-state ion exchange is the suitable method for introducing metals [276, 277]. The reason is the geometric constrain of the solvated cations or complexes, such as [Pt(NH3 )4 ]2+ . They are not able to penetrate the narrow pores of zeolites. Studies of introducing Pt, Pd, and Rh in zeolites with eight-membered oxygen rings have been performed for zeolites rho and ZK-5 as well as SAPO-41[278, 279].

531

532

17 Catalytically Active Sites: Generation and Characterization

6

4

3

5

1

2

Figure 17.16 Scheme of size and location of metal clusters in zeolites (see text for explanation) [270].

In these cases, the NH4 -form zeolites were ﬁrst heated at 773 K to obtain the H-form materials. Subsequently, these materials were mixed in a glove box with precalculated amounts of PtCl2 , PdCl2 , or RhCl3 for preparing metal clusters consisting of Pt, Pd, and Rh, respectively. The ion exchange in the mixtures, such as for PtCl2 /2(ZOH), PtCl2 + 2ZOH −→ 2HCl + 2 (ZO)− Pt+

(17.13)

is carried out at 823 – 898 K [278]. Often, the formation of HCl is detected by mass spectrometry. For a survey on the introduction of Pt, Pd, and Rh into 8-, 10-, 12-ring zeolites by solid-state ion exchange, see [277]. In dependence on the preparation procedure and the catalytic application, the size of the metal clusters Mn in and on the zeolite particles may vary across a wide range from n = 1 to 1000. The scheme in Figure 17.16 describes the different step of nucleation [270]: 1) Very small clusters with 1 < n < 4 in small zeolites cages, such as sodalite cages in zeolites X and Y or side pockets of pores. 2) Metal clusters with n < 10 – 40 and diameters up to 1.3 nm, for example, in the supercages of zeolites X and Y or in the crossing intersections of pores. 3) Metal clusters not limited by cages or pores, but ﬁlling the pore volume in a small part of the zeolite crystal. 4) Metal clusters, which grow larger than cages and pores, but still remain encapsulated in the zeolites particles. 5) Metal clusters located in lattice defects or twin boundaries of crystals. 6) Large metal clusters at the external crystal surface formed by migration of metal atoms or small clusters along the pores and their nucleation and growth to large particles. 17.7 Characterization of Metal Clusters in Zeolites 17.7.1 Test Reactions

The mechanism of hydrocarbon conversion on bifunctional zeolite catalysts consists of the following steps: the saturated hydrocarbon is ﬁrst dehydrogenated on the

17.7 Characterization of Metal Clusters in Zeolites

noble metal compound to the corresponding oleﬁns, which are protonated at the bridging OH groups acting as Brønsted acid sites. The resulting carbenium ions undergo skeletal rearrangements and β-scissions. Finally, the obtained carbocations are desorbed from the Brønsted acid sites as alkenes, which are often hydrogenated on the noble metal compounds. There are two test reactions of this type, which also give insight into the porous properties of the bifunctional zeolites under study: (i) isomerization and hydrocracking of n-decane [280, 281] or n-alkanes with shorter or longer chains [282, 283] and (ii) hydrocracking of butylcyclohexane or naphthenes with 10 carbon atoms [284, 285]. If n-alkanes are converted on a bifunctional zeolite in the absence of shape selectivity, that means on large-pore zeolites, the reactants are ﬁrst hydroisomerized to a mixture of all possible alkanes with one branching. For n-decane, for example, the monobranched isomers 2-, 3-, 4-, and 5-methylnonane, 3- and 4-ethyloctane, and 4-propylheptane are formed. Consecutive reactions lead to iso-alkanes with two branchings, such as dimethyloctaens and ethylmethylheptanes. Subsequently, these compounds undergo hydrocracking reactions, either directly or via tribranched intermediates [286]. In bifunctional medium-pore zeolites, pronounced differences in the product distribution in comparison with large-pore zeolites occur. Bulky ethyl- and propylbranched isomers are no longer formed. For example, on zeolite ZSM-5, 2-methylnonane is the preferred isomer. Therefore, the ratio of the yields of 2-methylnonane and 5-methylnonane obtained for the isomerization of n-decane is utilized for calculating the modiﬁed constraint index CI∗ [280]. This index covers a range of about 1 for Y-type zeolites (supercages with 12-ring windows) to about 14 for zeolite ZSM-22 (10-ring pores) [286]. In contrast to the above-mentioned isomerization reactions, the hydrocracking of C10 -naphthenes on large-pore zeolites is very selective. Methylcyclopentane and isobutane are almost exclusively formed [287]. However, the energetically favored reaction route leading to isobutane and methylcyclopentane requires bulky intermediates. On bifunctional zeolite catalysts with medium pores, hydrocracking of C10 -cyclopentanes leads to a much larger number of hydrocracked products [286]. Therefore, the ratio of the yield of isobutane and n-butane is utilized for calculating the spaciousness index SI [284, 285]. This index covers a range of about 1 for zeolites ZSM-5, ZSM-22, and ZSM-23 (10-ring pores) to about 22 for Y-type zeolites (supercages with 12-ring windows) [286]. In recent studies, the hydroisomerization of n-heptane [288], n-octane [289, 290], n-decane [291, 292], and other long-chain n-alkanes [293, 294] was applied for characterizing the catalytic activity of bifunctional zeolite catalysts, in some cases modiﬁed with promoters, such as Ni, Cr, La, Ce, Al, and Zn [288, 289, 291]. Comparing bifunctional zeolites with different structures and particle sizes, nanocrystalline (particle size of 30 nm) zeolite beta with nSi /nAl = 16 was found to be the best catalyst to produce multibranched isomers from n-decane [290]. For reviews on the characterization of bifunctional zeolite catalysts by hydroisomerization of n-alkanes, see [295, 296].

533

534

17 Catalytically Active Sites: Generation and Characterization

17.7.2 Analytical Methods

A number of theoretical studies utilizing the extended Huckel (EH) method or DFT calculations gave insight into the nature and chemical behavior of noble metal clusters [265, 297–300]. DFT calculations of Pt6 clusters in zeolite ZSM-5 indicated that the electronic structure of the metal particles is strongly affected by the interaction with framework oxygen atoms and the Brønsted acid sites of the zeolitic support [300]. Harmsen et al. [301] investigated the interaction of a single Pd atom with the zeolitic Brønsted acid site. While the reduction of Pd2+ to Pd0 was found to be strongly exothermic, the energy for an exchange of a single proton between Pd0 and the zeolite is nearly thermodynamically neutral. Important analytical methods for studying noble metals and their clusters in zeolites are transmission electron microscopy (TEM) [270, 302, 303], extended X-ray adsorption ﬁne structure (EXAFS) [304, 305], X-ray absorption near edge structure (XANES) [304–307], electron paramagnetic resonance (EPR) [308–310], and adsorption/desorption, for example, of H2 and CO [311, 312]. The formation of an irreversibly adsorbed monolayer of gas molecules is the most frequently applied method for characterizing metal atoms and metal clusters in zeolites. The measurement of the quantity of a gas adsorbed at monolayer coverage gives the metal surface area and the metal dispersion D. This metal dispersion D is deﬁned as [311] D = NS /NT

(17.14)

with the number NS of metal atoms at the surface of the clusters and the total number NT of all metal atoms. Determining D by gas chemisorption requires the knowledge of the stoichiometry of this reaction. There are several approaches for measuring the gas uptake, such as static and dynamic methods based on volumetry and gravimetry or on continuous ﬂow and pulsed ﬂow techniques, respectively [311]. In the case of static methods, the isotherms of H2 or CO adsorption are recorded and evaluated. In this case, the sample must be dehydrated and evacuated using a vacuum system. Often, the adsorption is performed at near room temperature. Flow techniques are faster and more convenient than static techniques since they do not require vacuum systems. Upon pretreatment of the catalyst at high temperature in an inert gas ﬂow and cooling down to the adsorption temperature, the ﬂow is switched, for example, to 2 vol% H2 in argon until the detector (often thermal conductivity cell) downstream shows a constant gas-phase composition. The H2 uptake is determined from the difference of the curves obtained without and with metal cluster containing catalyst. In the case of pulsed ﬂow experiments, the adsorbate gas is injected as successive small pulses of known volume into the ﬂow of the inert gas and the number of pulses to reach a complete coverage is determined. Recently, H2 adsorption was utilized for characterizing Pt particles in LTL- and Y-type zeolites [302, 312]. A three-site adsorption model was used to gain insight into the dependence of the H2

References

coverage on temperature, pressure, and support ionicity for zeolites Pt/Na-Y and Pt/H-USY [312]. 17.8 Summary and Outlook

This chapter explains the principles of the formation of the most important active sites of zeolite catalysts, their nature, and chemical behavior in dependence on the structure type and the chemical composition of the framework. By means of examples, it is demonstrated that modern analytical methods are able to provide important and novel information on the properties of surface sites in zeolite catalysts improving our understanding of their role in heterogeneously catalyzed reactions. In the past decade, a number of new techniques were introduced and applied, which allow step by step a deeper insight into zeolites and their surface sites under reaction-near conditions. Considering the function of catalytically active sites, the application of probe molecules is an important approach for investigating the nature, strength, and accessibility of surface centers. Such probe molecules are utilized for adsorption/desorption experiments and several spectroscopic methods. Obtaining relevant information requires a proper choice of the probe molecule applied, which should be near in the chemical behavior and size to the reactants converted in the heterogeneously catalyzed reaction under study. Worldwide research during the past decades and the successful use of acid and bifunctional zeolites in industrial catalysis has led to a general agreement in the understanding of the catalytically active sites in these materials. There is general consensus about the nature, formation, and behavior of bridging hydroxyl groups and their role as catalytically active sites in acid and bifunctional zeolite catalysts. Until now, we have limited knowledge on Lewis sites in working zeolite catalysts, which strongly depend on the type of zeolite and the conditions of formation. In some cases, modern analytical methods were able to give an insight into the nature and behavior of Lewis sites, but often limited to speciﬁc types of zeolites and under speciﬁc conditions, but not in a general manner. Also needed are new preparation routes for base sites in zeolites and the development of novel methods, which lead to a deeper understanding of these sites in zeolites and their role in heterogeneous catalysis. Progress along these lines can be expected in the following years. References 1. von Ballmoos, R., Harris, D.H., and

Magee, J.S. (1997) Catalytic Cracking in Handbook of Heterogeneous Catalysis, vol. 4 (eds G. Ertl, H. Knoezinger, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 1955–1986. 2. Cheng, W.C., Habib, E.T. Jr., Rajagopalan, K., Roberie, T.G., Wormsbecher, R.F., and Ziebarth,

M.S. (2008) Fluid catalytic craching in Handbook of Heterogeneous Catalysis, 2nd edn, vol. 6 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 2741–2777. 3. Blauwhoff, P.M.M., Gosselink, J.W., Kieffer, E.P., Sie, S.T., and Stork,

535

536

17 Catalytically Active Sites: Generation and Characterization

4.

5.

6.

7.

8.

9.

10.

11.

W.H.J. (1999) Zedites as Catalysts in Industrial Processes in Catalysis and Zeolites – Fundamentals and Applications (eds J. Weitkamp and L. Puppe), Springer, Berlin, Heidelberg, New York, pp. 437–538. Sie, S.T. (2008) Isomerization in Handbook of Heterogeneous Catalysis, 2nd edn, vol. 6 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 2809–2829. van Veen, J.A.R., Minderhoud, J.K., Huve, L.G., and Stork, W.H.J. (2008) Hydro Cracking and Catalytic Dewaxing in Handbook of Heterogeneous Catalysis, 2nd edn, vol. 6 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 2778–2808. Clark, M.C., Smith, C.M., Stern, D.L., and Beck, J.S. (2008) Alkylation of Aromatics in Handbook of Heterogeneous Catalysis, 2nd edn, vol. 7 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 3153–3167. Stern, D.L., Brown, S.H., and Beck, J.S. (2008) Isomerization and Transalkylation of Aromatics in Handbook of Heterogeneous Catalysis, 2nd edn, vol. 7 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 3168–3193. Sachtler, W.M.H. and Zhang, Z.C. (2008) Metal Clusters in Zeolites in Handbook of Heterogeneous Catalysis, 2nd edn, vol. 1 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 510–522. Weitkamp, J., Hunger, M., and Rymsa, U. (2001) Microporous Mesoporous Mater., 48, 255–270. Ertl, G., Knoezinger, H., and Weitkamp, J. (eds) (1999) Preparation of Solid Catalysts, Wiley-VCH Verlag GmbH. Weitkamp, J. and Puppe, L. (eds) (1999) Catalysis and Zeolites – Fundamentals and Applications, Springer, Berlin.

12. Weitkamp, J. and Hunger, M. (2007)

13.

14.

15.

16.

17.

18.

19.

20. 21.

22.

23.

Acid and Base Catalysis on Zeolites in Introduction to Zeolite Molecular Sieves, Studies in Surface Science and ˇ Catalysis, Vol. 168 (eds J. Cejka, H. van Bekkum, A. Corma, and F. Schueth), Elsevier, Amsterdam, pp. 787–835. Ertl, G., Knoezinger, H., Schueth F., and Weitkamp, J. (eds) (2008) Handbook of Heterogeneous Catalysis, 2nd edn, Wiley-VCH Verlag GmbH, Weinheim. Lercher, J.A. and Jentys, A. (2002) Application of Microporous Solids as Catalysts in Handbook of Porous Solids, vol. 2 (eds F. Schueth, K.S.W. Sing, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 1097–1145. Gorte, R.J. (2002) Surface Activity in Handbook of Porous Solids, vol. 1 (eds F. Schueth, K.S.W. Sing, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 432–464. Elanany, M., Koyama, M., Kubo, M., Broclawik, E., and Miyamoto, A. (2005) Appl. Surf. Sci., 246, 96–101. Patton, E.M.R.L. and Wilson, S.T. (1988) Structural, Synthetic and Physiochemical Concepts in Aluminophosphate-Based Molecular Sieves in Innovation in Zeolite Materials Science, Studies in Surface Science and Catalysis, Vol. 37 (eds P.J. Grobet, W.J. Mortier, E.F. Vansant, and G. Schulz-Ekloff), Elsevier, Amsterdam, pp. 13–20. Chu, C.T.-W., Kuehl, G.H., Lago, R.M., and Chang, C.D. (1985) J. Catal., 93, 451–458. Szostak, R., Nair, V., and Thomas, T.L. (1987) J. Chem. Soc., Faraday Trans. 1, 83, 487–494. Chu, C.T.-W. and Chang, C.D. (1985) J. Phys. Chem., 89, 1569–1571. Gnep, N.S., Doyemet, J.Y., Seco, A.M., and Ribeiro, F.R. (1988) Appl. Catal., 43, 155–166. Hunger, M., Schenk, U., Breuninger, M., Glaeser, R., and Weitkamp, J. (1999) Microporous Mesoporous Mater., 27, 261–271. Hu, W., Luo, Q., Su, Y., Chen, L., Yue, Y., Ye, C., and Deng, F. (2006)

References

24.

25. 26.

27.

28.

29.

30. 31. 32. 33.

34. 35.

36. 37.

38.

Microporous Mesoporous Mater., 92, 22–30. Xie, X., Satozawa, M., Kunimori, K., and Hayashi, S. (2000) Microporous Mesoporous Mater., 39, 25–35. Corma, A. (2003) J. Catal., 216, 298–312. Kresnawahjuesa, O., Kuehl, G.H., Gorte, R.J., and Quierini, C.A. (2002) J. Catal., 210, 106–115. Wichterlova, B., Vorbeck, G., Fricke, ˇ R., Richtermendau, J., and Cejka, J. (1992) Collect. Czech Chem. Commun., 57, 799–808. Arnold, A., Hunger, M., and Weitkamp, J. (2004) Microporous Mesoporous Mater., 67, 205–213. Martens, J.A., Souverijns, W., Van Rhijn, W., and Jacobs, P.A. (1997) Acidity and Basicity in Zeolites in Handbook of Heterogeneous Catalysis, vol. 1 (eds G. Ertl, H. Knoezinger, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 324–365. Chao, K.-J., Lin, J.-C., Wang, Y., and Lee, G.H. (1986) Zeolites, 6, 35–38. Olson, D.H. and Dempsey, E. (1969) J. Catal., 13, 221–231. Hunger, M. (1997) Catal. Rev.-Sci. Eng., 39, 345–393. Sanderson, R.T. (1976) Chemical Bonds and Bond Energy, 2nd edn, Academic Press, New York. Mortier, W.J. (1978) J. Catal., 55, 138–145. Hunger, M. (2008) NMR Spectroscopy for the Characterization of Surface Acidity and Basicity in Handbook of Heterogeneous Catalysis, 2nd edn, vol. 2 (eds G. Ertl, K. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 1163–1178. Pine, L.A., Maher, P.J., and Wachter, W.A. (1984) J. Catal., 85, 466–476. Corma, A. (1989) Application of Zeolites in Fluid Catalytic Cracking and Related Processes in Zeolites: Facts, Figures, Future, Studies in Surface Science and Catalysis, Vol. 49 (eds P.A. Jacobs and R.A. van Santen), Elsevier, Amsterdam, pp. 49–56. Baerlocher, Ch., Meier, W.M., and Olson, D.H. (2001) Atlas of Zeolites

39.

40.

41.

42.

43.

44.

45.

46. 47.

48. 49.

Framework Types, 5th edn, Elsevier, Amsterdam. Shigeishi, R.A., Chiche, B.H., and Fajula, F. (2001) Microporous Mesoporous Mater., 43, 211–236. Szostak, R. (1991) Modiﬁed Zeolites in Introduction to Zeolite Science and Practice, Studies in Surface Science and Catalysis, Vol. 58 (eds H. van Bekkum, E.M. Flanigen, and J.C. Jansen), Elsevier, Amsterdam, pp. 153–200. Lago, R.M., Haag, W.O., Mikovski, R.J., Olson, D.H., Hellring, S.D., Schmidt, K.D., Kerr, G.T. (1986) The Nature of the Catalytic Sites in HZSM-5 Activity Enhancement in New Developments in Zeolite Science and Technology, Studies in Surface Science and Catalysis, Vol. 28 (eds Y. Murakami, A. Iijima, and J.W. Ward), Elsevier, Amsterdam, pp. 677–684. Brunner, E., Ernst, H., Freude, D., Hunger, M., Krause, C.B., Prager, D., Reschetilowski, W., Schwieger, W., and Bergk, K.H. (1989) Zeolites, 9, 282–286. Kuehl, G.H. (1999) Modiﬁcation of Zeolites in Catalysis and Zeolites (eds Weitkamp J. and Puppe, L.), Springer, Berlin, pp. 81–197. Huang, J., Jiang, Y., Reddy Marthala, V.R., Ooi, Y.S., Weitkamp, J., and Hunger, M. (2007) Microporous Mesoporous Mater., 104, 129–136. Huang, J., Jiang, Y., Reddy Marthala, V.R., Thomas, B., Romanova, E., and Hunger, M. (2008) J. Phys. Chem. C, 112, 3811–3818. Hirschler, A.E. (1963) J. Catal., 2, 428–439. Plank, C.J. (1965) Discussion in Proceedings of the 3rd International Congress on Catalysis (eds W.M. Sachtler, G.C. Schuit, and P. Zwietering), North-Holland, Amsterdam, p. 727. Chen, F.R., Davis, J.G., and Fripiat, J.J. (1992) J. Catal., 133, 263–278. Altwasser, S., Jiao, J., Steuernagel, S., Weitkamp, J., and Hunger, M. (2004) Elucidating the Dealumination Mechanism of Zedite H-Y by solid-state NMR Spectroscopy in Recent Advances in the Science and Technology of Zeolites

537

538

17 Catalytically Active Sites: Generation and Characterization

50.

51.

52. 53.

54.

55. 56. 57.

58.

59. 60.

61.

62.

63.

64.

and Related Materials, Studies in Surface Science and Catalysis, Vol. 154 (eds E. van Steen, L.H. Callanan, and M. Claeys), Elsevier, Amsterdam, pp. 3098–3105. Jiao, J., Kanellopoulos, J., Wang, W., Ray, S.S., Foerster, H., Freude, D., and Hunger, M. (2005) Phys. Chem. Chem. Phys., 7, 3221–3226. Maher, P.K., Hunter, F.D., and Scherzer, J. (1971) Adv. Chem. Ser., 101, 266–274. Kuehl, G.H. (1977) J. Phys. Chem. Solids, 38, 1259–1263. Chen, T.-H., Houthoofd, K., and Grobet, P.J. (2005) Microporous Mesoporous Mater., 86, 31–37. Gonzales, N.O., Bell, A.T., and Chakraborty, A.K. (1997) J. Phys. Chem. B, 101, 10058–10064. Olson, D.H., Haag, W.O., and Lago, R.M. (1980) J. Catal., 61, 390–396. Haag, W.O., Lago, R.M., and Weisz, P.B. (1984) Nature, 309, 589–591. Haag, W.O. (1984) Catasysis by Zeolites-Science and Technology in Zeolites and Related Microporous Materials: State of the Art 1994, Studies in Surface Science and Catalysis, Vol. 84B (eds J. Weitkamp, H.G. Karge, H. Pfeifer, and W. Hoelderich), Elsevier, Amsterdam, pp. 1375–1394. DeCanio, S.J., Sohn, J.R., Fritz, P.O., and Lunsford, J.H. (1986) J. Catal., 101, 132–141. Miale, J.N., Chen, N.Y., and Weisz, P.B. (1966) J. Catal., 6, 278–287. Derouane, E.G. and Chang, C.D. (2000) Microporous Mesoporous Mater., 35-36, 425–433. De Vos, D.E., Ernst, S., Perego, C., O’Connor, C.T., and Stoecker, M. (2002) Microporous Mesoporous Mater., 56, 185–192. Karge, H.G., Ladebeck, J., Sarbak, Z., and Hatada, K. (1982) Zeolites, 2, 94–102. Karge, H.G., Hatada, K., Zhang, Y., and Fiedorow, R. (1983) Zeolites, 3, 13–21. Karge, H.G., Wada, Y., Weitkamp, J., Ernst, S., Girrbach, U., and Beyer, H.K. (1984) A Comparative Study of Pentasil Zeolites and Dealuminated Mordenites

65.

66.

67.

68. 69.

70.

71.

72.

73. 74. 75. 76.

as Catalysts for the Disproportionation of Ethylbenzene in Catalysis on the Energy Scene, Studies in Surface Science and Catalysis, Vol. 19 (eds S. Kaliaguine and A. Mahay), Elsevier, Amsterdam, pp. 101–111. Karge, H.G., Ernst, S., Weihe, M., Weiß, U., and Weitkamp, J. (1994) in Zeolites and Related Microporous Materials: State of the Art 1994, Studies in Surface Science and Catalysis, Vol. 84C (eds J. Weitkamp, H.G. Karge, H. Pfeifer, and W. Hoelderich), Elsevier, Amsterdam, pp. 1805–1812. Weihe, M., Hunger, M., Breuninger, M., Karge, H.G., and Weitkamp, J. (2001) J. Catal., 198, 256–265. Huang, J., Jiang, Y., Reddy Marthala, V.R., Bressel, A., Frey, J., and Hunger, M. (2009) J. Catal. 263, 277–283. Guisnet, M.R. (1990) Acc. Chem. Res., 23, 392–398. Karge, H.G., Koesters, H., and Wada, Y. (1984) Dehydration of cyclohexanol as a Test Reaction for Zeolite Reactivity in Proceedings of the 6th International Zeolite Conference (eds D.H. Olson and A. Bisio), Butterﬁelds, Guilford, pp. 308–316. Bourdillon, G., Gueguen, C., and Guisnet, M. (1990) Appl. Catal., 61, 123–139. Karge, H.G. (2008) Acidity and Basicity in Handbook of Heterogeneous Catalysis, 2nd edn, vol. 2 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 1096–1122. Sigl, M., Ernst, S., Weitkamp, J., and Knoezinger, H. (1997) Catal. Lett., 45, 27–33. Benesi, H.A. (1956) J. Am. Chem. Soc., 78, 5490–5494. Benesi, H.A. (1957) J. Phys. Chem., 61, 970–973. Benesi, H.A. (1973) J. Catal., 28, 176–178. Benesi, H.A. and Winquist, B.H.C. (1979) Surface Acidity of Solid Catalysts in Advances in Catalysis, vol. 27 (eds D.D. Eley, H. Pines, and P.B. Weisz), Academic Press, New York, pp. 97–181.

References 77. Hirschler, A.E. and Schneider, A. 78. 79. 80.

81.

82.

83. 84.

85. 86. 87.

88. 89. 90. 91.

92.

93.

94.

95.

(1961) J. Chem. Eng. Data, 6, 313–318. Moscou, L. and Mone, R. (1973) J. Catal., 30, 417–422. Streitwieser, A. and Kim, Y.-J. (2000) J. Am. Chem. Soc., 122, 11783–11786. Jentys, A. and Lercher, J.A. (2001) Techniques of Zeolite Characterization in Introduction to Zeolite Science and Practice, Studies in Surface Science and Catalysis, Vol. 137 (eds H. van Bekkum, E.M. Flanigen, P.A. Jacobs, and J.C. Jansen), Elsevier, Amsterdam, pp. 345–386. Araujo, R.S., Azevedo, D.C.S., Cavalcante, C.L. Jr., Jimenez-Lopez, A., and Rodriguez-Castellon, E. (2008) Microporous Mesoporous Mater., 108, 213–222. Wang, K., Wang, X., and Li, G. (2006) Microporous Mesoporous Mater., 94, 325–329. Wang, K., Wang, X., and Li, G. (2007) Catal. Commun., 8, 324–328. Ding, C., Wang, X., Guo, X., and Zhang, S. (2007) Catal. Commun., 9, 487–493. Miradatos, C. and Barthomeuf, D. (1979) J. Catal., 57, 136–146. Cvetanovic, R.J. and Amenomiya, Y. (1967) Adv. Catal., 17, 103–118. Parker, M.L., Bibby, D.M., and Meinhold, R.H. (1985) Zeolites, 5, 384–388. Yang, L., Aizhen, Y., and Qinhua, X. (1991) Appl. Catal., 67, 169–177. Gorte, R.J. (1982) J. Catal., 75, 164–174. Demmin, R.A. and Gorte, R.J. (1984) J. Catal., 90, 32–39. Woolery, G.L., Kuehl, G.H., Timken, H.C., Chester, A.W., and Vartuli, J.C. (1997) Zeolites, 19, 288–296. Biaglow, A.I., Gittleman, C., Gorte, R.J., and Madon, R.J. (1991) J. Catal., 129, 88–93. Gricus Kofke, T.J., Gorte, R.J., Kokotailo, G.T., and Farneth, W.E. (1989) J. Catal., 115, 265–272. Biaglow, A.I., Parrillo, D.J., Kokotailo, G.T., and Gorte, R.J. (1994) J. Catal., 148, 213–223. Gorte, R.J. (1999) Catal. Lett., 62, 1–13.

96. Rodriguez-Gonzalez, L.,

97.

98.

99.

100. 101. 102.

103. 104. 105.

106. 107.

108.

109.

110.

111.

Rodriguez-Castellon, E., Jimenez-Lopez, A., and Simon, U. (2008) Solid State Ionics, 179, 1968–1973. Niwa, M., Nishikawa, S., and Katada, N. (2005) Microporous Mesoporous Mater., 82, 105–112. Niwa, M., Suzuki, K., Katada, N., Kanougi, T., and Atoguchi, T. (2005) J. Phys. Chem. B, 109, 18749–18757. Noda, T., Suzuki, K., Katada, N., and Niwa, M. (2008) J. Catal., 259, 203–210. Fubini, B. (1988) Thermochim. Acta, 135, 19–29. Cardona-Martinez, N. and Dumesic, J.A. (1992) Adv. Catal., 38, 149–244. Auroux, A. (1994) Thermal Methods: Calorimetry, Differential Thermal Analysis and Thermogravimetry in Catalyst Characterization: Physical Techniques for Solid Materials (eds B. Imelik and J.C. Vedrine), Plenum Press, New York, p. 611–650. Auroux, A. (1997) Top. Catal., 4, 71–89. Gorte, R.J. and White, D. (1997) Top. Catal., 4, 57–69. Zibrowius, B. and Loefﬂer, E. (2002) Characterization in Handbook of Porous Solids, vol. 2 (eds F. Schueth, K.S.W. Sing, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 935–1015. Auroux, A. (2002) Top. Catal., 19, 205–213. Bevilacqua, M., Meloni, D., Sini, F., Monaci, R., Montanari, T., and Busca, G. (2008) J. Phys. Chem. C, 112, 9023–9033. Ferreira, A.F.P., Mittelmeijer-Hazeleger, M.C., Bliek, A., and Moulijn, J.A. (2008) Microporous Mesoporous Mater., 111, 171–177. Chronister, C.W. and Drago, R.S. (1993) J. Am. Chem. Soc., 115, 4793–4798. Drago, R.S., Dias, S.C., Trombetta, M., and Lima, L. (1997) J. Am. Chem. Soc., 119, 4444–4452. Ghesti, G.F., de Macedo, J.L., Parent, V.C.I., Dias, J.A., and Dias, S.C.L. (2007) Microporous Mesoporous Mater., 100, 27–34.

539

540

17 Catalytically Active Sites: Generation and Characterization 112. de Macedo, J.L., Ghesti, G.F., Dias,

113.

114.

115.

116. 117.

118. 119.

120.

121. 122. 123.

124.

J.A., and Dias, S.C.L. (2008) Phys. Chem. Chem. Phys., 10, 1584–1592. Ward, J.W. (1976) Infrared Spectroscopy of Zeolite Surfaces and Surface Reactions in Zeolite Chemistry and Catalysis, ACS, Monograph, Vol. 171 (ed. J.A. Rabo), American Chemical Society, Washington, DC, pp. 118–306. Karge, H.G., Hunger, M., and Beyer, H.K. (1999) Characterization of Zeolites – Infrared and Nuclear Magnetic Resonance Spectroscopy and X-Ray Diffraction in Catalysis and Zeolites (eds J. Weitkamp and L. Puppe), Springer, Berlin, pp. 198–326. Karge, H.G. and Geidel, E. (2004) Vibrational Spectroscopy in Molecular Sieves: Characterization I (eds H.G. Karge and J. Weitkamp), Springer, Berlin, pp. 1–200. Kustov, L.M. (1997) Top. Catal., 4, 131–144. Saussey, J. and Thibault-Starzyk, F. (2004) Infrared Spectroscopy Classical Methods in In situ Spectroscopy of Catalysts (ed. B.M. Weckhuysen), American Scientiﬁc Publishers, Stevenson Ranch, pp. 15–31. Beck, K., Pfeifer, H., and Staudte, B. (1993) Microporous Mater., 2, 1–6. Berglund, B., Lindgren, J., and Tegenfeld, J. (1978) J. Mol. Struct., 43, 179–191. Loefﬂer, E., Lohse, U., Peuker, Ch., Oehlmann, G., Kustov, L.M., Zholobenko, V.L., and Kazansky, V.B. (1990) Zeolites, 10, 266–271. Jacobs, P.A. and van Ballmoos, R. (1982) J. Phys. Chem., 86, 3050–3052. Datka, J. and Tuznik, T. (1985) Zeolites, 5, 230–232. Jentys, A., Mirth, G., Schwank, J., and Lercher, J.A. (1989) Interaction of Hydrocarbons and Water with ZSM-5 in Zeolites, Facts, Figures, Future, Studies in Surface Science and Catalysis, Vol. 49 (eds P.A. Jacobs and R.A. van Santen), Elsevier, Amsterdam, pp. 847–856. Kazansky, V.B., Minachev, K.M., Nefedov, B.F., Borovkov, V.Y., Kontratev, D.A., Chukin, G.D.,

125.

126. 127.

128.

129. 130.

131.

132.

133.

134. 135.

136.

137.

138.

139.

140.

141.

142.

Kustov, L.M., Bondarenko, T.N., and Konovalchikov, L.D. (1983) Kinet. Katal., 24, 679–682. Stepanova, E.A., Komarov, V.S., Sinilo, M.F., and Shirinskaya, L.P. (1989) Zh. Prikl. Spektrosk., 51, 950–956. Ward, J.W. (1968) J. Phys. Chem., 72, 4211–4223. Uytterhoeven, J.B., Schoonheydt, R., Liengme, B.V., and Hall, W.K. (1969) J. Catal., 13, 425–434. Uytterhoeven, J.B., Christner, L.C., and Hall, W.K. (1965) J. Phys. Chem., 69, 2117–2126. Hughes, T.R. and White, H.M. (1967) J. Phys. Chem., 71, 2112–2114. Zecchina, A., Bordiga, S., Spoto, G., Marchese, L., Petrini, G., Leofanti, G., and Petrini, M. (1992) J. Phys. Chem., 96, 4991–4997. Woolery, G.L., Alemany, L.B., Dessau, R.M., and Chester, A.W. (1986) Zeolites, 6, 14–16. Jentys, A., Rumplmayr, G., and Lercher, J.A. (1989) Appl. Catal., 53, 299–312. Zholobenko, V.L., Kustov, L.M., Borovkov, V.Y., and Kazansky, V.B. (1988) Zeolites, 8, 175–178. Jacobs, P.A. and Mortier, W.J. (1982) Zeolites, 2, 226–230. Makarova, M.A., Ojo, A.F., Karim, K., Hunger, M., and Dwyer, J. (1994) J. Phys. Chem., 98, 3619–3623. van Donk, S., Bus, E., Broersma, A., Bitter, J.H., and de Jong, K.P. (2002) Appl. Catal. A: Gen., 237, 149–159. Wichterlova, B., Tvaruzkova, Z., Sobalik, Z., and Sarv, P. (1998) Microporous Mesoporous Mater., 24, 223–233. Domokos, L., Lefferts, L., Seshan, K., and Lercher, J.A. (2000) J. Mol. Catal. A: Chem., 162, 147–157. Lercher, J.A., Gruendling, C., and Eder-Mirth, G. (1996) Catal. Today, 27, 353–376. Pieterse, J.A.Z., Veefkind-Reyes, S., Seshan, K., Domoskos, L., and Lercher, J.A. (1999) J. Catal., 187, 518–520. Kaucky, D., Sobalik, Z., Schwarze, M., Vondrova, A., and Wichterlova, B. (2006) J. Catal., 238, 293–300. Rakoczy, R.A., Breuninger, M., Hunger, M., Traa, Y., and Weitkamp,

References

143.

144.

145.

146.

147.

148.

149. 150.

151.

152.

153.

154.

155. 156.

J. (2002) Chem. Eng. Technol., 25, 273–275. Wichterlova, B., Zilkova, N., Uvarova, ˇ E., Cejka, J., Sarv, P., Paganini, C., and Lercher, J.A. (1999) Appl. Catal. A: Gen., 182, 297–308. van Donk, S., Bus, E., Broersma, A., Bitter, J.H., and de Jong, K.P. (2002) J. Catal., 212, 86–93. Trombetta, M., Busca, G., Rossini, S., Piccoli, V., Cornaro, U., Guercio, A., Catani, R., and Willey, R.J. (1998) J. Catal., 179, 581–596. Resini, C., Montanari, T., Nappi, L., Bagnasco, G., Turco, M., Busca, G., Bregani, F., Notaro, M., and Rocchini, G. (2003) J. Catal., 214, 179–190. Belhekar, A.A., Ahedi, R.K., Kuriyavar, S., Shevade, S.S., Rao, B.S., Anand, R., and Tvaruzkova, Z. (2003) Catal. Commun., 4, 295–302. Gil, B., Zones, S.I., Hwang, S.-J., ˇ Bejblova, M., and Cejka, J. (2008) J. Phys. Chem. C, 112, 2997–3007. Tanabe, K. (1970) Solid Acids and Bases, Academic Press, New York. Tanabe, K. (1981) Solid Acid and Base Catalysts in Catalysis, Science and Technology, vol. 2 (eds J.R. Anderson and M. Boudard), Springer, Berlin, p. 231–273. Lavalley, J.-C., Anquetil, R., Czyzniewska, J., and Ziolek, M. (1996) J. Chem. Soc., Faraday Trans., 92, 1263–1266. Bevilacqua, M., Montanari, T., Finocchio, E., and Busca, G. (2006) Catal. Today, 116, 132–142. Janin, A., Lavalley, J.C., Benazzi, E., Schott-Darie, C., and H. Kessler (1995) Acidity of Cloverite in Catalysis by Microporous Materials, Studies in Surface Science and Catalysis, Vol. 94 (eds H.K. Beyer, H.G. Karge, I. Kiricsi, and J.B. Nagy), Elsevier, Amsterdam, pp. 124–130. Zecchina, A., Spoto, G., and Bordiga, S. (2005) Phys. Chem. Chem. Phys., 7, 1627–1642. Hunger, M. (1996) Solid State Nucl. Magn. Reson., 6, 1–29. Koller, H., Lobo, R.F., Burkett, S.L., and Davis, M.E. (1995) J. Phys. Chem., 99, 12588–12596.

157. Wolf, I., Gies, H., and Fyfe, C.A. (1999)

J. Phys. Chem. B, 103, 5933–5938. 158. Mastikhin, V.M., Mudrakovsky, I.L.,

159.

160.

161.

162.

163.

164.

165.

166. 167.

168. 169.

170.

171.

172.

173.

and Nosov, A.V. (1991) Prog. NMR Spectrosc., 23, 259–299. Deng, F., Yue, Y., and Ye, C. (1998) Solid State Nucl. Magn. Reson., 10, 151–160. Zhang, W., Ma, D., Liu, X., Liu, X., and Bao, X. (1999) Chem. Commun., 1091–1092. Zhang, W., Bao, X., Guo, X., and Wang, X. (1999) Catal. Lett., 60, 89–94. Simon, A., Gougeon, R.D., Paillaud, J.L., Valtchev, V., and Kessler, H. (2001) Phys. Chem. Chem. Phys., 3, 867–872. Hunger, M., Anderson, M.W., Ojo, A., and Pfeifer, H. (1993) Microporous Mater., 1, 17–32. Kao, H.M., Grey, C.P., Pitchumani, K., Lakshminarasimhan, P.H., and Ramamurthy, V. (1998) J. Phys. Chem. A, 102, 5627–5638. Rachwalik, R., Olejniczak, Z., Jiao, J., Huang, J., Hunger, M., and Sulikowski, B. (2007) J. Catal., 252, 161–170. Beck, L.W. and Haw, J.F. (1995) J. Phys. Chem., 99, 1076–1079. Brunner, E., Beck, K., Koch, M., Heeribout, L., and Karge, H.G. (1995) Microporous Mater., 3, 395–399. Freude, D. (1995) Chem. Phys. Lett., 235, 69–75. Hunger, M., Ernst, S., Steuernagel, S., and Weitkamp, J. (1996) Microporous Mater., 6, 349–353. Ma, D., Deng, F., Fu, R., Han, X., and Bao, X. (2001) J. Phys. Chem. B, 105, 1770–1779. Isobe, T., Watanabe, T., d’Espinose de la Caillerie, J.B., Legrand, A.P., and Massiot, D. (2003) J. Colloid Interface Sci., 261, 320–324. Hunger, M. and Weitkamp, J. (2004) Nuclear Magnetic Resonance in In Situ Spectroscopy of Catalysts (ed. B.M. Weckhuysen), American Scientiﬁc Publishers, Stevenson Ranch, pp. 177–218. Nesterenko, N.S., Thibault-Starzyk, F., Montouillout, V., Yuschenko, V.V.,

541

542

17 Catalytically Active Sites: Generation and Characterization

174.

175.

176.

177.

178.

179.

180.

181.

182.

183. 184.

185. 186.

187.

Fernandez, C., Gilson, J.P., Fajula, F., and Ivanova, I.I. (2004) Microporous Mesoporous Mater., 71, 157–166. Kennedy, G.J., Afeworki, M., Calabro, D.C., Chase, C.E., and Smiley, R.J. Jr. (2004) Appl. Spectrosc., 58, 698–704. Yesinowski, J.P., Eckert, H., and Rossman, G.R. (1988) J. Am. Chem. Soc., 110, 1367–1375. Omegna, A., Vasic, M., van Bokhoven, J.A., Pirngruber, G., and Prins, R. (2004) Phys. Chem. Chem. Phys., 6, 447–452. Jaenchen, J., van Wolput, J.H.C.M., van de Ven, L.J.M., de Haan, J.W., and van Santen, R.A. (1996) Catal. Lett., 39, 147–152. Paze, C., Zecchina, A., Spera, S., Cosma, A., Merlo, E., Spano, G., and Girotti, G. (1999) Phys. Chem. Chem. Phys., 1, 2627–2629. Huang, J., Jiang, Y., Reddy Marthala, V.R., Wang, W., Sulikowski, B., and Hunger, M. (2007) Microporous Mesoporous Mater., 99, 86–90. White, J.L., Beck, L.W., and Haw, J.F. (1992) J. Am. Chem. Soc., 114, 6182–6189. Beck, L.W., Xu, T., Nicholas, J.B., and Haw, J.F. (1995) J. Am. Chem. Soc., 117, 11594–11595. Ernst, H., Freude, D., Mildner, T., and Pfeifer, H. (1999) High Temperature 1 H MAS NMR Studies of the Proton Mobility in Zeolites in Proceedings of the 12th International Zeolite Conference Materials Research Society (eds M.M.J. Treacy, B.K. Marcus, M.E. Bisher, and J.B. Higgins), Warrendale, Pennsylvania, pp. 2955–2962. Haw, J.F. (2002) Phys. Chem. Chem. Phys., 4, 5431–5441. Biaglow, A.I., Gorte, R.J., Kokotailo, G.T., and White, D. (1994) J. Catal., 148, 779–786. Biaglow, A.I., Gorte, R.J., and White, D. (1994) J. Catal., 150, 221–224. Xu, M., Arnold, A., Buchholz, A., Wang, W., and Hunger, M. (2002) J. Phys. Chem. B, 106, 12140–12143. Ma, D., Han, X., Xie, S., Bao, X., Hu, H., and Au-Yeung, S.C.F. (2002) Chem. Eur. J., 8, 162–170.

188. Lunsford, J.H., Rothwell, W.P., and

189.

190.

191.

192.

193. 194.

195.

196.

197.

198.

199. 200.

201. 202.

203.

204.

205.

Shen, W. (1985) J. Am. Chem. Soc., 107, 1540–1547. Lunsford, J.H., Tutunjian, P.N., Chu, P.J., Yeh, E.B., and Zalewski, D.J. (1989) J. Phys. Chem., 93, 2590–2595. Lunsford, J.H., Sang, H., Campbell, S.M., Liang, C.H., and Anthony, R.G. (1994) Catal. Lett., 27, 305–314. Luo, Q., Deng, F., Yuan, Z., Yang, J., Zhang, M., Yue, Y., and Ye, C. (2003) J. Phys. Chem. B, 107, 2435–2442. Zhang, W., Han, X., Liu, X., and Bao, X. (2003) J. Mol. Catal. A, 194, 107–113. Alonso, B., Klur, I., and Massiot, D. (2002) Chem. Commun., 804–805. Sutovich, K.J., Peters, A.W., Rakiewicz, E.F., Wormsbecher, R.F., Mattingly, S.M., and Mueller, K.T. (1999) J. Catal., 183, 155–158. Karra, M.D., Sutovich, K.J., and Mueller, K.T. (2002) J. Am. Chem. Soc., 124, 902–903. Wang, Y., Zhuang, J., Yang, G., Zhou, D., Ma, D., Han, X., and Bao, X. (2004) J. Phys. Chem. B, 108, 1386–1391. Brunner, E., Pfeifer, H., Wutscherk, T., and Zscherpel, D. (1992) Z. Phys. Chem., 178, 173–183. Zscherpel, D., Brunner, E., Koch, M., and Pfeifer, H. (1995) Microporous Mater., 4, 141–147. Zhao, B., Pan, H., and Lunsford, J.H. (1999) Langmuir, 15, 2761–2765. Rakiewicz, E.F., Peters, A.W., Wormsbecher, R.F., Sutovich, K.J., and Mueller, K.T. (1998) J. Phys. Chem. B, 102, 2890–2896. Kao, H.M. and Grey, C.P. (1997) J. Am. Chem. Soc., 119, 627–628. Chu, P.J., de Mallmann, A., and Lunsford, J.H. (1991) J. Phys. Chem., 95, 7362–7368. Zheng, A., Zhang, H., Chen, L., Yue, Y., Ye, C., and Deng, F. (2007) J. Phys. Chem. B, 111, 3085–3089. Haw, J.F., Nicholas, J.B., Xu, T., Beck, L.W., and Ferguson, D.B. (1996) Acc. Chem. Res., 29, 259–267. Feast, S. and Lercher, J.A. (1996) Synthesis of Intermediates and Fire Chemicals using Molecular Sieves in Recent Advances and New Horizons in

References

206. 207.

208.

209. 210. 211. 212.

213.

214.

215.

216.

Zeolite Science and Technology, Studies in Surface Science and Catalysis, Vol. 102 (eds H. Chon, S.I. Woo, and S.E. Park), Elsevier, Amsterdam, pp. 363–412. Tanabe, K. and Hoelderich, W.F. (1999) Appl. Catal. A: Gen., 181, 399–434. Barthomeuf, D. (1994) Basicity in Zeolites in Acidity and Basicity of Solids (eds J. Fraissard and L. Petrakis), Kluwer Academic Publishers, Dordrecht, pp. 181–197. Barthomeuf, D. (1991) Acidity and Basicity in Zeolites in Catalysis and Adsorption by Zeolites, Studies in Surface Science and Catalysis, Vol. 65 (eds G. Oehlmann, H. Pfeifer, and R. Fricke), Elsevier, Amsterdam, p. 157–169. Doskocil, E.J. (2004) Microporous Mesoporous Mater., 76, 177–183. Hathaway, P.E. and Davis, M.E. (1989) J. Catal., 116, 263–278. Li, J. and Davis, R.J. (2003) Appl. Catal. A: Gen., 239, 59–70. Martens, L.R.M., Grobet, P.J., Vermeiren, W.J.M., and Jacobs, P.A. (1986) Sodium Clusters in Zeolites as Active Sites for Carbanion Catalyzed Reactions in New Developments in Zeolite Science and Technology, Studies in Surface Science and Catalysis, Vol. 28 (eds Y. Murakami, A. Iijama, and J.W. Ward), Elsevier, Amsterdam, pp. 935–941. Baba, T., Kim, G.J., and Ono, Y. (1992) J. Chem. Soc., Faraday Trans., 88, 891–897. Yoshida, T., Tanaka, T., Yoshida, S., Hikita, S., Baba, T., and Ono, Y. (2000) Solid State Commun., 114, 255–256. Kloetstra, K.R. and van Bekkum, H. (1995) J. Chem. Soc., Chem. Commun., 1005–1006. Kloetstra, K.R. and van Bekkum, H. (1997) Solid Mesoporous Base Catalysts Comprising of MCM-41 Supported Infraporous Cesium Oxide in Progress in Zeolite and Microporous Materials, Studies in Surface Science and Catalysis, Vol. 105A (eds H. Chon, S.K. Ihm, and K.S. Uh), Elsevier, Amsterdam, pp. 431–438.

217. Kloetstra, K.R., van Laren, M., and

218.

219.

220.

221.

222.

223.

224.

225.

226. 227.

228. 229. 230.

231.

van Bekkum, H. (1997) J. Chem. Soc., Faraday Trans., 93, 1211–1220. Kloetstra, K.R., van den Broek, J., and van Bekkum, H. (1997) Catal. Lett., 47, 235–242. Yang, C., He, N.Y., and Xu, Q.H. (1999) Effect of Trivalent Elements in the Framework on the Basicity of Zeolites in Porous Materials in Environmentally Friendly Processes, Studies in Surface Science and Catalysis, Vol. 125 (eds I. Kiricsi, G. Pal-Borbely, J.B. Nagy, and H.G. Karge), Elsevier, Amsterdam, pp. 457–464. Mignon, P., Geerlings, P., and Schoonheydt, R. (2007) J. Phys. Chem. C, 111, 12376–12382. Ernst, S., Hartmann, M., Sauerbeck, S., and Bongers, T. (2000) Appl. Catal. A: Gen., 200, 117–123. Narasimharao, K., Hartmann, M., Thiel, H.H., and Ernst, S. (2006) Microporous Mesoporous Mater., 90, 377–383. Xiong, J., Ding, Y., Zhu, H., Yan, L., Liu, X., and Lin, L. (2003) J. Phys. Chem. B, 107, 1366–1369. El Haskouri, J., Cabrera, S., Sapina, F., Latorre, J., Guillem, C., Beltran-Porter, A., Betran-Porter, D., Marcos, M.D., and Amoros, P. (2001) Adv. Mater., 13, 192–195. Aramendia, M.A., Borau, V., Garcia, I.M., Jimenez, C., Marinas, A., Marinas, J.M., Porras, A., and Urbano, F.J. (1999) Appl. Catal. A: Gen., 184, 115–125. Handa, H., Fue, Y., Baba, T., and Ono, Y. (1999) Catal. Lett., 59, 195–200. Yamaguchi, T., Zhu, J.-H., Wang, Y., Komatsu, M.-A., and Ookawa, M. (1997) Chem. Lett., 26, 989–990. Stevens, M.G. and Foley, H.C. (1997) Chem. Commun., 519–520. Dessau, R.M. (1990) Zeolites, 10, 205–206. Alcaraz, J.J., Arena, B.J., Gillespie, R.D., and Holmgren, J.S. (1998) Catal. Today, 43, 89–99. Corma, A., Fornes, V., Martin-Aranda, R.M., Garcia, H., and Primo, J. (1990) Appl. Catal., 59, 237–248.

543

544

17 Catalytically Active Sites: Generation and Characterization 232. Ernst, S., Bogers, T., Casel, C., and

233.

234.

235.

236.

237.

238.

239. 240. 241. 242.

243.

244. 245.

Munsch, S. (1999) Cesium-Modiﬁed Mesoporous Molecular Sieves as Basic Catalysts for Knoevenagel Condensations in Porous Materials in Environmentally Friendly Processes, Studies in Surface Science and Catalysis, Vol. 125 (eds I. Kiricsi, G. Pal-Borbely, J.B. Nagy, and H.G. Karge), Elsevier, Amsterdam, pp. 367–374. Kantam, M.L., Choudary, B.M., Reddy, C.V., Rao, K.K., and Figueras, F. (1998) Chem. Commun., 1033–1034. Delsarte, S., Auroux, A., and Grange, P. (2000) Phys. Chem. Chem. Phys., 2, 2821–2827. Corma, A., Martin-Aranda, R.M., and Sanchez, F. (1990) J. Catal., 126, 192–198. Martins, L., Hoelderich, W., and Cardoso, D. (2008) J. Catal., 258, 14–24. Martins, L., Vieira, K.M., Rios, L.M., and Cardoso, D. (2008) Catal.Today, 133, 706–710. Martinez, N.C. and Dumesic, J.A. (2008) Thermochemical Characterization in Handbook of Heterogeneous Catalysis, 2nd edn, Vol. 2 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 1123–1135. Li, J. and Davis, R.J. (2005) J. Phys. Chem. B, 109, 7141–7148. Lavalley, J.C. (1996) Catal. Today, 27, 377–401. Barthomeuf, D. (1996) Catal. Rev., 38, 521–612. Knoezinger, H. (2008) Infrared Spectroscopy for the Characterzation of Surface Acidity and Basicity in Handbook of Heterogeneous Catalysis, 2nd edn, Vol. 2 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 1135–1163. Knoezinger, H. and Huber, S. (1998) J. Chem. Soc., Faraday Trans., 94, 2047–2059. Yagi, F., Tsuji, H., and Hattori, H. (1997) Microporous Mater., 9, 237–245. Kucera, J. and Nachtigall, P. (2004) J. Phys. Chem. B, 108, 16012–16022.

246. Marie, O., Malicki, N., Pommier, C.,

247.

248.

249.

250.

251.

252.

253.

254.

255. 256. 257.

258.

259.

260.

Massiani, P., Vos, A., Schoonheydt, R., Geerlings, P., Henriques, C., and Thibault-Starzyk, F. (2005) Chem. Commun., 1049–1051. Mignon, P., Pidko, E.A., van Santen, R.A., Geerlings, P., and Schoonheydt, R.A. (2008) Chem. Eur. J., 14, 5168–5177. Bull, L.M., Cheetham, A.K., Anupold, T., Reinhold, A., Samoson, A., Sauer, J., Bussemer, B., Lee, Y., Gann, S., Shore, J., Pines, A., and Dupree, R. (1998) J. Am. Chem. Soc., 120, 3510–3511. Pingel, U.-T., Amoureux, J.-P., Anupold, T., Bauer, F., Ernst, H., Fernandez, C., Freude, D., and Samoson, A. (1998) Chem. Phys. Lett., 294, 345–350. Freude, D., Loeser, T., Michel, D., Pingel, U., and Prochnow, D. (2001) Solid State Nucl. Magn. Reson., 20, 46–60. Schneider, D., Toufar, H., Samoson, A., and Freude, D. (2009) Solid State Nucl. Magn. Reson., 35, 87–92. Bosch, E., Huber, S., Weitkamp, J., and Knoezinger, H. (1999) Phys. Chem. Chem. Phys., 1, 579–581. Sanchez-Sanchez, M., Blasco, T., and Rey, F. (1999) Phys. Chem. Chem. Phys., 1, 4529–4535. Sanchez-Sanchez, M., Blasco, T., and Corma, A. (2008) J. Phys. Chem. C, 112, 16961–16967. Sanchez-Sanchez, M. and Blasco, T. (2000) Chem. Commun., 491–492. Bosacek, V. (1995) Z. Phys. Chem., 189, 241–250. Bosacek, V., Ernst, H., Freude, D., and Mildner, T. (1997) Zeolites, 18, 196–199. Bosacek, V., Klik, R., Genoni, F., Spano, G., Rivetti, F., and Figueras, F. (1999) Magn. Reson. Chem., 37, S135–S141. Schenk, U., Hunger, M., and Weitkamp, J. (1999) Magn. Reson. Chem., 37, S75–S78. Lima, E., de Menorval, L.C., Tichit, D., Lasperas, M., Grafﬁn, P., and Fajula, F. (2003) J. Phys. Chem. B, 107, 4070–4073.

References 261. Krawietz, T.R., Murray, D.K., and

262. 263.

264. 265. 266. 267.

268.

269.

270.

271. 272.

273. 274. 275. 276.

277.

Haw, J.F. (1998) J. Phys. Chem. A, 102, 8779–8785. Dessau, R.M. (1982) J. Catal., 77, 304–306. (a) Ribeiro, F., Marcilly, C., and Guisnet, M. (1982) J. Catal., 78, 267–274; (b) Ribeiro, F., Marcilly, C., and Guisnet, M. (1982) J. Catal. 78, 275–280. Trevino, H. and Sachtler, W.M.H. (1994) Catal. Lett., 27, 251–258. Efremenko, I. and Sheintuch, M. (2000) J. Mol. Catal., 160, 445–451. Wertheim, G.K. (1989) Z. Phys. D: At., Mol. Clusters, 12, 319–326. Jena, P., Khanna, S.N., and Rao, B.K. (1992) Int. J. Mod. Phys. B, 6, 3657–3666. Sachtler, W.M.H. and Zhang, Z.C. (1993) Zeolite-Supported Transition Metal Catalysts in Advances in Catalysis, Vol. 39 (eds D.D. Eley, H. Pines, and P.B. Weisz), Academic Press, San Diego, pp. 129–220. Sachtler, W.M.H. (1999) Metal Clusters in Zeolites in Preparation of Solid Catalysts (eds G. Ertl, H. Knoezinger, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 388–405. Gallezot, P. (2002) Preparation of Metal Clusters in Zeolites in Molecular Sieves:Post-synthesis Modiﬁcation I, (eds H.G. Karge and J. Weitkamp), Springer, Berlin, pp. 257–305. Moretti, G. and Sachtler, W.M.H. (1989) J. Catal., 115, 205–216. Zhang, Z.C. and Sachtler, W.H.M. (1990) J. Chem. Soc., Faraday Trans., 86, 2313–2319. Feeley, J.S. and Sachtler, W.M.H. (1991) J. Catal., 131, 573–581. Weber, W.A. and Gates, B.C. (1998) J. Catal., 80, 207–217. Li, F. and Gates, B.C. (2003) J. Phys. Chem. B, 107, 11589–11596. Karge, H.G. and Beyer, H.K. (2002) Solid-state Ion Exchange in Microporous Materials in Molecular Sieves: Post-synthesis Modiﬁcation I (eds H.G. Karge and J. Weitkamp), Springer, Berlin, pp. 43–201. Karge, H.G. (2008) Solid-state Ion Exchange in Zeolites in Handbook

278.

279.

280.

281. 282.

283.

284. 285.

of Heterogeneous Catalysis, 2nd edn, Vol. 1 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 484–510. Weitkamp, J., Ernst, S., Bock, T., Kiss, A., and Kleinschmidt, P. (1995) Introduction of Noble Metals into Small Pore Zeolites via Solid state Ion Exchange in Catalysis by Microporous Materials, Studies in Surface Science and Catalysis, Vol. 94 (eds H.K. Beyer, H.G. Karge, I. Kiricsi, and J.B. Nagy), Elsevier, Amsterdam, pp. 278–285. Weitkamp, J., Ernst, S., Bock, T., Kromminga, T., Kiss, A., and Kleinschmidt, P. (1996) US Patent No. 5,529,964, assigned to Degussa AG. Martens, J.A., Tielen, M., Jacobs, P.A., and Weitkamp, J. (1984) Zeolites, 4, 98–107. Martens, J.A. and Jacobs, P.A. (1986) Zeolites, 6, 334–348. Martens, J.A., Vanbutsele, G., and Jacobs, P.A. (1993) Characterization of Large and Extra-Large Zeolite Pores with the Heptadecane Test in Proceedings of the 9th International Zeolite Conference, Vol. 2 (eds R. von Ballmoos, J.B. Higgins, and M.M.J. Treacy), Butterworth-Heinemann, Stoneham, pp. 355–362. Feijen, E.J.P., Martens, J.A., and Jacobs, P.A. (1996) Isomerization and Hydrocracking of Decane and Hepta-decane on Cubic and Hexagonal Faujasite Zeolites and their Intergrowth Structures in 11th International Congress on Catalysis, Studies in Surface Science and Catalysis Vol. 101 (eds J.W. Hightower, W.N. Delgass, E. Iglesia, and A.T. Bell), Elsevier, Amsterdam, pp. 721–728. Weitkamp, J., Ernst, S., and Kumar, R. (1986) Appl. Catal., 27, 207–210. Weitkamp, J., Ernst, S., and Chen, C.Y. (1989) The Spaciousness Index: A Useful Catalytic Method for Probing the Effective Pore Width of Molecular Sieves in Zeolites, Facts, Figures, Future, Studies in Surface Science and Catalysis, Vol. 49B (eds P.A. Jacobs and R.A. van Santen), Elsevier, Amsterdam, pp. 1115–1122.

545

546

17 Catalytically Active Sites: Generation and Characterization 286. Weitkamp, J., Ernst, S., and Puppe,

287.

288.

289.

290. 291.

292.

293.

294.

295. 296. 297. 298.

299. 300.

301.

302.

L. (1999) Shape-selective Catalysis in Zeolites in Catalysis and Zeolites (eds J. Weitkamp and L. Puppe), Springer, Berlin, pp. 327–376. Weitkamp, J., Ernst, S., and Karge, H.G. (1984) Erd¨ol Kohle Erdgas, Petrochem., 37, 457–462. Liu, P., Wang, J., Wie, R.P., Ren, X.Q., and Zhang, X.G. (2008) Catal. Lett., 126, 346–352. Karthikeyan, D., Lingappan, N., Sivasankar, B., and Jabarathinam, N.J. (2008) Ind. Eng. Chem., 47, 6538–6546. Chica, A. and Corma, A. (2007) Chem. Ing. Tech., 79, 857–870. Karthikeyan, D., Lingappan, N., Sivasankar, B., and Jabarathinam, N.J. (2008) Appl. Catal. A: Gen., 345, 18–27. Soualaha, A., Lemberton, J.L., Chater, M., Magnoux, P., and Moljord, K. (2007) React. Kin. Catal. Lett., 91, 307–313. Soualah, A., Lemberton, J.L., Pinard, L., Chater, M., Magnoux, P., and Mojord, K. (2008) Appl. Catal. A: Gen., 336, 23–28. Wang, G., Liu, Q., Su, W., Li, X., Jiang, Z., Fang, X., Han, C., and Li, C. (2008) Appl. Catal. A: Gen., 335, 20–27. Deldari, H. (2005) Appl. Catal. A: Gen., 293, 1–10. Jun, L. and Wang, E.P. (2008) Progr. Chem., 20, 457–463. Efremenko, I. and Sheintuch, M. (1998) Surf. Sci., 414, 148–158. German, E.D., Efremenko, I., and Sheintuch, M. (2001) J. Phys. Chem. A, 105, 11312–11326. Efremenko, I. and Sheintuch, M. (2005) Chem. Phys. Lett., 401, 232–240. Mikhailov, M.N., Kustov, L.M., and Kazansky, V.B. (2008) Catal. Lett., 120, 8–13. Harmsen, R., Bates, S., and van Santen, R.A. (1997) Faraday Discuss., 106, 443–450. Ji, Y.Y., van der Eeirden, A.M.J., Koot, V., Kooyman, P.J., Meeldijk, J.D.,

303.

304.

305. 306.

307.

308.

309.

310.

311.

312.

Weckhuysen, B.M., and Koningsberger, D.C. (2005) J. Catal., 234, 376–384. Datye, A.K., Hansen, P.L., and Helveg, S. (2008) Election Microscopy Techniques in Handbook of Heterogeneous Catalysis, 2nd edn, Vol. 2 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 803–833. Koningsberger, D.C. and Ramaker, D.E. (2008) Applications of X-Ray Absorption Spectroscopy in Heterogeneous Catalysis: EXAFS Atomic XAFS and Delta XANES in Handbook of Heterogeneous Catalysis, 2nd edn, Vol. 2 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 774–803. Davis, R.J. and Boudart, M. (1994) J. Phys. Chem., 98, 5471–5477. Asakura, K., Kubota, T., Chun, W.J., Iwasawa, Y., Ohtani, K., and Fujikawa, T. (1999) J. Synchrotron Radiat., 6, 439–441. Koningsberger, D.C., Oudenhuijzen, M.K., de Graaf, J., van Bokhoven, J.A., and Ramaker, D.E. (2003) J. Catal., 216, 178–191. Zheng, J., Schmauke, T., Roduner, E., Dong, J.L., and Xu, Q.H. (2001) J. Mol. Catal. A: Chem., 171, 181–190. Liu, X., Dilger, H., Eichel, R.A., Kunstmann, J., and Roduner, E. (2006) J. Phys. Chem. B, 110, 2013–2023. Akdogan, Y., Anantharaman, S., Liu, X., Lahiri, G.K., Bertagnolli, H., and Roduner, E. (2009) J. Phys. Chem. C, 113, 2352–2359. Bergeret, G. and Gallezot, P. (2008) Particle Size Dispersion Measurements in Handbook of Heterogeneous Catalysis, 2nd edn, Vol. 2 (eds G. Ertl, H. Knoezinger, F. Schueth, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, pp. 738–765. Ji, Y.Y., Koot, V., van der Eeirden, A.M.J., Weckhuysen, B.M., Koningsberger, D.C., and Ramaker, D.E. (2007) J. Catal., 245, 415–427.

547

18 Cracking and Hydrocracking Marcello Rigutto

18.1 Introduction

New technologies often develop in and emerge from niches. In contrast, zeolite catalysis conquered catalytic cracking [1] and hydrocracking [2], its largest industrial applications, ﬁrst. The impact turned out to be huge. According to a recent survey [3], worldwide hydrocracking capacity reached just over 5 million barrels of feed intake per day (MMbbl per day, which amounts to roughly 200 million tons per year) in 2008, and catalytic cracking capacity was close to 14.5 MMbbl per day (about 600 million tons per year), which is about 23% of the total reﬁning capacity of 85 MMbbl per day [4] (some 3.5 billion tons per year). Global production of these volumes consumes just over 200 kilotons per year of zeolite [5], in catalysts worth several billion US dollars [6–8]. Synthetic zeolites have ﬁrst emerged from Barrer’s seminal work, and then from the Linde laboratory in search of adsorbents, to make all of this possible [9, 10]; so, as always, the path to invention is tortuous. And, in turn, it is probably not unfair to state that the application of zeolites in reﬁning has not only given rise to, and again fed on, a large body of applied research, but, through this, has also been an important stimulus to fundamental research involving microporous and mesoporous materials in many disciplines, providing a small but ﬁne example of positive feedback between application and science [11].1) 18.1.1 The Oil Reﬁnery – Where to Find Zeolites in It, and Why – and the Place of Hydrocracking and Catalytic Cracking

Every reﬁnery is different – in scale, complexity, crude oil diet, infrastructure, environment, regional demand for its products, and history. For this chapter, it is 1) Mokyr deﬁnes this feedback more precisely

in terms of prescriptive knowledge and propositional knowledge. Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

VDU

DA

FURF

CFHT

Dewax

FCC

Coker

PtR

Unsaturated gas plant

Heavy gasoline desulfurization

HDC

Olefin upgrading

ALKY

ISOM

G

Light Olefins

L P

Figure 18.1 A hypothetical complex reﬁnery, from Ref. [6]. Zeolite-catalyzed processes are shaded. CRU,– crude distillation unit; HDW, hydrodewaxing; CHD, catalytic hydrodesulfurization; PtR, reforming; ISOM, isomerization; CFHT, catalytic feed hydrotreating; FCC , ﬂuid catalytic cracking; HDC, hydrocracking; ALKY, alkylation; VDU, vacuum distillation unit; FURF, furfural extraction; DEWAX, lube hydrodewaxing; and DA, deasphalting.

CRU

HDW

CHD

Saturated gas plant

548

18 Cracking and Hydrocracking

18.1 Introduction

useful to consider a hypothetical world scale, say 400 000 barrels per day (roughly 18 million tons per year) very complex reﬁnery, home to the most important conversion processes. A simple scheme, showing distillation and conversion processes as simple unit operations, is depicted in Figure 18.1 (needless to say, in reality such a complex is tightly integrated) [12]. The main purpose of ﬂuid catalytic cracking (FCC), hydrocracking, and residue hydroprocessing is to convert heavy fractions to liquid fuels – clean and without the need for postprocessing wherever possible. FCC is most suitable for the production of gasoline, whereas hydrocrackers allow for more ﬂexibility, and are mostly geared to produce clean kerosene and diesel fuel in most parts of the world, except for the United States where the primary purpose is production of naphtha as reformer feed for gasoline production. Hydrotreating is applied throughout the boiling range to remove sulfur, nitrogen, aromatics and sometimes alkenes where necessary. Catalytic dewaxing removes n-alkanes from lubricants or diesel by isomerization or, obviously less preferred, selective cracking. Reforming produces gasoline by converting naphtha-range feed, mostly alkanes, to aromatics and hydrogen, by cyclization and dehydrogenation. Isomerization and alkylation convert light fractions (C5−6 and C4 , respectively) to high-octane, branched alkane gasoline components [9, 13]. Zeolites are used in the reﬁnery as strong and thermally stable solid acid catalysts ﬁrst and foremost, drawing on the arsenal of elementary reactions discussed in Chapters 12 and 15, and 17: they crack alkanes, alkenes, and alkylaromatics (in FCC, hydrocracking, residue conversion), they isomerize (in almost all processes), they enable hydrogen transfer between molecules (in FCC), and they oligomerize (intermediately produced) alkenes (in alkylation, FCC). In all uses, shape selectivity (see Chapters 10 and 26) has some role to play: in some processes it is essential, whereas in other processes, it is perhaps a restriction that has to be accepted. Historically, in FCC and to a signiﬁcant extent also in hydrocracking, zeolites have come to replace amorphous silica aluminas primarily if not only because of their much higher activity (interpreted as caused by higher acidity, but as we will see it is only now becoming clear what that means). Better stability and regenerability caused AlCl3 -based catalysts to be gradually replaced by zeolites in alkane isomerization [14]. Environmental reasons – the possibility to avoid use of HF or H2 SO4 – may provide opportunities for zeolite-catalyzed alkylation. Catalytic dewaxing is uniquely enabled by shape-selective conversion. Zeolite catalysts were adopted on different grounds in every case. In and around the reﬁnery, zeolites are also used (and proposed) for selective adsorption. 18.1.2 The Changing Environment for Reﬁning

Long-term trends in reﬁning, some long standing, some more recently established, are • a growing demand for products across the board [15] (see Figure 18.2); • a so-called ‘‘whitening’’ of the hydrocarbon demand – that is, an ever-increasing demand for distillate products (mostly transportation fuels: gasoline, kerosene,

549

18 Cracking and Hydrocracking

18 000

80 000

16 000

70 000

14 000

Kilobarrels per day

90 000

60 000 50 000 40 000 30 000 20 000

12 000 10000 8000 6000 4000 2000

0

0

Others Light distillates Figure 18.2

19

9 19 7 9 19 8 9 20 9 0 20 0 0 20 1 0 20 2 0 20 3 0 20 4 0 20 5 0 20 6 07

10 000 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Kilobarrels per day

550

Middle distillates Fuel oil

Global oil demand and fuels distribution 1997–2007 [4].

diesel fuel) at the cost of heavy products (mostly fuel oil, for marine transport, electricity generation) [16] (see also Figure 18.2); • a gradual shift in demand from gasoline to diesel fuel, especially outside the United States; • an increasing need for the processing of increasingly heavy crudes (although this trend has been interrupted from time to time); • ever tightening legal speciﬁcations on sulfur and other properties of fuels. For a well-known example, in many European countries very low sulfur (10 ppm) diesel fuel is now mandatory (Figure 18.3). In addition, tight speciﬁcations exist on polyaromatics in and density of diesel fuel, sulfur, aromatics, and alkenes in gasoline [17]. The North American situation is not very different; Asian, Middle Eastern, and South American countries do follow, at a distance, but with considerable impact, given the size of some of these markets. In a nutshell, where it concerns zeolite catalysis, this translates into the following: • A growing importance of hydrocracking [18, 19]. In the recent past typical growth rates were 3–5% per year, from a 250 million tonnes per annum base in 2003. These developments will slow down somewhat in the current climate, but the long-term trend will likely remain unchanged. • Generally, a stronger economic incentive for upgrading heavy feeds into transportation fuel-range products. • A need to remedy the limitations of the FCC process, viz. to reduce FCC gasoline sulfur without octane loss, and to cope with the highly aromatic heavier products (cycle oils). • Speciﬁcally, at least for FCC, an increasing heaviness of the feed. • Continuing interest in upgrading of residue, nonconventional crude oils, oil sands, and so on. • A growing importance of gas to liquid (GTL) processes.

18.2 FCC

European Union, Japan, United States – ahead of the pack Fuel type Diesel sulfur limit (ppm)

EU 50 (2005) 10 (<2009) Gasoline sulfur limit (ppm) 30 (2006)

Japan 50 (2004) 10 (2007) 10 (2008)

US 15 (2006) 30 (2007)

Australia, Hong Kong – following first Fuel type Australia Diesel sulfur limit (ppm) 50 (2000) Gasoline sulfur limit (ppm) 150 (2005)

Hong Kong 50 (2000) 150 (2001)

Brazil and India – progress in major cities Fuel type Diesel sulfur limit (ppm)

Brazil–Sao Paulo only 50 (2009)

Gasoline sulfur limit (ppm) 80 (2009)

India – major cities 500 50 (2010) 500 150 (2010)

China, Mexico – following later… Fuel type China Diesel sulfur limit (ppm) 2000 Gasoline sulfur limit (ppm) 800

Mexico 500 300 (Mexico city) 900 (elsewhere)

Figure 18.3 Global development of sulfur speciﬁcations for automotive fuels (International Council on Clean Transportation).

18.2 FCC 18.2.1 The FCC Process

Fluidized catalytic cracking is a core process in many reﬁneries, and produces light alkenes, high-octane gasoline, and aromatic middle distillates from vacuum gasoil and often also from residue. FCC units have a feed intake between 2000 and 10 000 tons per day (about 0.6–3 million tons per annum). The process is the major gasoline-producing process in the reﬁnery: FCC gasoline is a mixture of C5 –C11 alkanes and alkenes, and aromatics, with a relatively high octane number (RON 90–94) [20]. The alkanes and alkenes are mostly branched; the branched alkenes and the aromatics contribute most to the high octane number. Examples of product compositions are given in Figure 18.4 [21, 22]. An FCC unit operator faces a changing market and changing environmental restrictions, some of which concern us here: • Demand for propene is growing faster than demand for ethene, making the FCCU an increasingly attractive source. The share of FCC-produced propene is

551

0

20

40

60

80

1950s

1960s

HCO/SLURRY

LCO

1970s

GAS

1980s

GASOLINE LPG

0.75

C12NH9

C16NH11 0.69 C20NH13 0.65 0.67

0.36

C17H12

C30H20

C33H12

217 267

380

412

0.71

H/C

Possible Formula assignment

0.00

5.00

10.00

15.00

20.00

25.00

216

167

Mass (m/z)

wt %

C5

C6

0.00

5.00

10.00

15.00

20.00

C8

C9

C10

C11 triAr Naph-diAr diAr Naph-moAr moArom diNaph c-Cm i-Cm n-Cm

Cn

C9 C10 C11 C12 C13 C14 C15 C16 C17 C18

C7

Aromatics c-Cm= i-Cm= n-Cm= c-Cm= i-Cm= n-Cm=

Figure 18.4 FCC product slate: historical development through improved catalysts [12] (left), and some typical detailed compositions from PIONA [21] GCxGC and MALDI-TOF [22] (right).

(%)

100

COKE

LPG composition (%) Propane 5.6 30.6 Propane n-butane 4.0 i-butane 16.3 n-butanes 30.8 i-butane 12.7

% w/w

552

18 Cracking and Hydrocracking

18.2 FCC

expected to grow from roughly a quarter to roughly a third of world propene production in 2015 [23]. • FCC gasoline may contain more than 1000 ppm sulfur and with current legislation often requires separate hydrotreatment [24]. • Light cycle oil (LCO, the FCC gasoil-range product) is highly aromatic and has a very low cetane number, and usually a high sulfur content. If not downgraded to a fuel oil component, subsequent hydroprocessing is necessary. LCO is considered a difﬁcult feedstock for hydroprocessing and hydrocracking units [25]. The FCC process is in many ways quite exceptional. Some key aspects are best discussed, in a nutshell, with the help of a schematic representation of the FCC Elevation

Flue gas To expander and waste heat bolier

Compact separator system Shell Swirl tube sepr. Lining

Product vapours to desuperheater/ fractionator

Reactor/ stripper

Regenerator

Steam

Combustion air

Resid. feed Catalyst coolers

BFW 0

Figure 18.5

Shell Resid FCC [26].

HP steam Steam

553

554

18 Cracking and Hydrocracking

unit (Figure 18.5) [26]. For further details about the process the reader is referred elsewhere [12]. The riser is the actual reactor, in which preheated (∼300 ◦ C) feed is contacted with a large amount of hot (>650 ◦ C) ﬂuidized catalyst at a relatively low pressure of 2–3 bar (this happens in a device at the bottom called the liftpot), and the mixture is driven to the top of the riser – as the feed is vaporized and the subsequently converting mixture expands, in the course of several seconds. Much more catalyst than feed has to be fed to the riser to achieve the required conversion: catalyst-to-oil ratios may vary but values are often in the range between 6 and 9 (on a mass basis). Over the length of the riser, endothermic cracking reactions take place, generating a temperature proﬁle (Figure 18.6). As coke is produced alongside the lighter products [27, 28], the catalyst quickly deactivates [29, 30]. The catalyst arrives in the stripper at temperatures between 500 and 540 ◦ C. This device is designed to minimize subsequent contact between catalyst and products, reducing secondary cracking reactions as much as possible. After stripping, the catalyst, which still contains typically 0.8–1.3 wt% coke, is fed to the regenerator where the coke is burnt off at temperatures of about 700 ◦ C. This requires a residence time in the order of minutes. The coke combustion provides the heat needed for cracking in the riser, and this heat is stored in the regenerated catalyst. It also generates steam, because the coke still contains some hydrogen. The coproduction of coke Vapor velocity, molar expansion

+20 °C

High-rare-earth-exchanged zeolite Riser terminal temperature

Conversion of 340 °C feed

Increasing value

+40 °C Mix. temp. Low-rare-earth-exchanged zeolite

Coke, metals on catalyst C5–221 °C gasoline 221–340 °C light cycle oil

Catalyst activity decay

Feed injection

90

60 50 40 30 20

CTO 4

10

CTO 2

Gasoline fraction (wt%)

70

Gasoline (C5–215 °C)

End of the riser disengager

CTO 4

50 CTO 2

40 30 20 10 0

0 0 (a)

Reaction time/riser length

60

HCO (325 °C)

80 HCO fraction (wt%)

Beginning of the riser

1

2

3

4

Residence time (s)

5

0

6 (b)

Figure 18.6 Typical riser reactor temperature proﬁle (top left); yields proﬁles in the riser (top right); and yields versus residence time from microriser studies [12, 29, 30]. (CTO = Cat-to-oil ratio; HCO = Heavy cycle oil.)

1

2

3

4

Residence time (s)

5

6

18.2 FCC

nevertheless allows products to be enriched in hydrogen relative to the feed. This is termed upgrading by carbon rejection. The steam generated by combustion reaches a partial pressure of typically 0.2 bar, which has the essentially undesired effect of dealuminating the framework of the zeolite and degrading zeolite crystallinity. One consequence is that fresh catalyst has to be fed to the unit continuously to replace a bleed stream of spent catalyst. The so-called equilibrium catalyst circulating in the unit at a given time is therefore a mixture with a broad age distribution, and a corresponding distribution of properties. 18.2.2 The FCC Catalyst, and Catalytic Chemistry

A fresh FCC catalyst [20, 31] consists of spray-dried spherical, ﬂuidizable, attrition-resistant particles, typically containing 20–40% ultrastable Y (USY) zeolite, a binder, and optionally a catalytically active, acidic matrix [31, 32], as well as various additives that serve different functions, some of which we will discuss later. The zeolite is either in an NH4 -exchanged form or exchanged with lanthanide ions (a commercial mixture) yielding rare earth-(US)Y or brieﬂy ‘‘RE (US)Y’’ (more on this later), and is usually added as a separate powder early in the preparation process. In situ synthesis of NaY zeolite from a kaolin precursor in a formed particle, with additional beneﬁts of accessibility and strength, is also applied [31, 33, 34]. The matrix contributes to the conversion of the bulkiest molecules in the feed, but it is the zeolite that catalyzes most of the conversion: it cracks alkanes and alkylaromatics in the feed and alkene intermediate products, and one generally assumes this to occur by the same monomolecular and bimolecular acid-catalyzed pathways that have been established for the conversion of model compounds (Figure 18.7, for a detailed discussion of this chemistry see also the next chapter) [35–39]: The bimolecular pathway involves hydrogen-transfer reactions, which progressively lead to the disproportionation of alkenes to alkanes and aromatics. This reaction also contributes to the formation of coke. A key property of the zeolite in the FCC catalyst is its crystallographic unit-cell size, which is a proximate measure of the aluminum content of the zeolite framework, and hence often taken as a proximate measure of its acidity (see Chapters 10–12). As depicted in Figure 18.8, it has a major inﬂuence on yield distributions, coke make, and product quality (i.e., the gasoline Research Octane Number), and, remarkably, only a modest inﬂuence on activity. As the reader may have inferred from the above process description, the zeolite unit-cell size is an average property of the equilibrium catalyst and is determined by operational history as much as by properties of the fresh catalyst. There is, however, a strong relation between the equilibrium unit-cell size – usually between 2.422 and 2.435 nm, versus values closer to 2.455 nm for the fresh catalysts – and the rare-earth (RE) loading of the zeolites. At least a portion of the framework aluminum sites that are charge-compensated by RE3+ ions seems to be removed at a much lower rate than protonated aluminum sites, or aluminum sites that are charge-compensated by cationic extra-framework aluminum species [40].

555

556

18 Cracking and Hydrocracking

HZ

Z

Protolysis

−

HZ

Z

+

H b-scission H-transfer

Z Z

+

Z

Z

−

H-transfer HZ H-transfer

H-transfer

Figure 18.7 Elementary reactions operating in an FCC: (a) monomolecular, protolytic cracking of alkanes; (b) isomerization and β-scission of alkenes; and (c) ‘‘bimolecular’’ cracking involving hydrogen transfer.

Setting rare-earth content of the fresh catalyst, and indirectly thereby steering its equilibrium zeolite unit-cell size, therefore allows for a trade-off between higher activity and higher gasoline selectivity on the one hand, and higher gasoline octane number and lower coke make on the other hand [41].2) The higher activity of the high unit-cell size materials is sufﬁciently explained by a better crystallinity retention of the zeolite [42] but generally assumed to be due to a higher density of acid sites also [43].3) The yield and selectivity effects are more complicated. They are brought about by a more extensive occurrence of hydrogen-transfer reactions with increasing unit-cell size, leading to lower yields of alkenes and higher yields of alkanes and aromatics, and coke [12, 44]. This trend is usually discussed in terms of underlying intrinsic kinetics, hydrogen transfer being a bimolecular reaction. However, there is no convincing mechanistic picture of such a process requiring multiple zeolitic sites, and indeed for model compounds it does not [45]. An alternative explanation in terms of selective adsorption of alkenes at higher unit-cell sizes, promoting their progressive bimolecular reaction relative to competing cracking reactions, may also be formulated [46]. One more comment on the role of zeolite acidity is warranted: since the general view is that cracking reactions in FCC are catalyzed mostly by zeolitic Brønsted-acid sites, one would indeed expect activity of the FCC equilibrium catalyst to correlate with its acidity in some way. As mentioned, and perhaps surprisingly, this is not evident from available experimental data – rather, at least for vacuum gasoil 2) Needless to say this is only a ﬁrst-order ap-

proximation as many catalyst vendors claim various approaches to break this trade-off. 3) Although it has been clearly established that Brønsted acid sites form when RE3+ ions in nondealuminated RE-Y hydrolyze

to RE(OH)2+ , the same phenomenon has not been demonstrated for dealuminated materials that would be more representative of the zeolite present in equilibrium FCC catalysts.

55 50 45 40 24.20 24.25 24.30 24.35 24.40 24.45 24.50

5.5

93 92 91 90 89

5.0 Conversion = 62 v% FFB pilot plant 950 °F, 30 second contact time Feed 02

4.5 4.0 3.5

3.0 24.20 24.24 24.28 24.32 24.36 24.40 24.44 24.48 Unit cell (Å)

KMC BMG Sigma

24.45

MZ-11 MA-7 Super D EKZ

85 80 Mat activity (%)

Coke (%w)

94

88 24.20 24.24 24.28 24.32 24.36 Unit cell size (Å)

6.0

24.30

24.20

Delta 400

0

2

4 6 8 10 12 RE2O3 on 100% Y zeolite

65

7

Sigma Delta plus 455

24.25

70

55 24.20 24.24 24.28 24.32 24.36 Unit cell size (Å)

24.40

24.35

75

60

XL DA

C3 – gas (%w)

Zeolite unit cell size (Å) after 5 h, 815 °C steaming

557

95

60 Research octane no

C5 – gasoline (%v)

18.2 FCC

14

Figure 18.8 Effect of unit-cell size on yields and activity, and effect of rare-earth level on zeolite unit-cell size (bottom left) [12].

16

6 5 4 3 24.20 24.24 24.28 24.32 24.36 Unit cell size (Å)

18 Cracking and Hydrocracking

3 Sample A

2.5

Sample B Sample C

Rate constant (k)

558

2

1.5

1

0.5

0

0

5

10 15 20 Zeolite concentration (wt%)

25

30

Figure 18.9 VGO cracking activity (expressed as pseudo second-order rate constant) depends on equilibrium catalyst (residual) zeolite content with little dependence on unit-cell size [42].

cracking, equilibrium catalyst activity correlates to resulting zeolite content only, and unit-cell size, while determining selectivity, has a much smaller effect on activity (see Figure 18.9, and also Figure 18.8 above) [42]. This is still an unsolved puzzle in the understanding of FCC catalysis. The inﬂuence of nonframework aluminum species on FCC performance remains somewhat elusive [47, 48], perhaps also because such species represent an intermediate stage in (steam) dealumination, which may be too transient to be of importance in real equilibrium catalysts. The importance of acid strength has been addressed, for example, in studies involving SAPO-37 but without a ﬁnal answer in view of the experimental complications [49, 50]. 18.2.3 Residue Cracking and the Effect of Deposited Metals on the Catalyst

Improvements in both catalyst [51] (better yields and lower coke make, active matrices) and hardware (among which better feed vaporization, better stripping, introduction of catalyst coolers) [52, 53] have enabled cracking of feeds containing a signiﬁcant amount of atmospheric distillation residue (also called long residue), allowing the processing of feeds of up to 7% Conradson Carbon (a conventional measure for the coking tendency) or even higher [12, 26, 54]. New short contact time conﬁgurations for residue cracking, such as the MSCC process, have also been introduced [29, 55]. The last decades, and particularly the 1980s and 1990s,

18.2 FCC

have seen an increasing share of residue processing, also in response to a declining demand for fuel oil. Residues also contain signiﬁcant levels of various metals, of which nickel and vanadium (mostly present as porphyrins [56]) are particularly relevant to FCC because they can enhance coke make by acting as dehydrogenation catalysts [57]. Vanadium is moreover harmful because it catalyzes destruction of the zeolite lattice, probably by intermediacy of vanadic acid and possibly VO2 + ions that exchange into the zeolite when vanadium is fully oxidized in the regenerator [58], or by formation of low-melting sodium vanadate when sodium is present [59]. It appears that dealumination and collapse of the siliceous structure are both facilitated [60, 61]. Iron can also damage catalyst performance by reducing accessibility of the catalyst pores (i.e., the larger pores in the binder giving access to the zeolite crystals within the particle) [62, 63]. Nickel may be passivated by compound formation through addition of low levels of antimony [64] or bismuth [65, 66] to the feed, whereas the effect of vanadium is more difﬁcult to remedy, through addition of ‘‘traps’’ to or with the catalyst, compounds that react preferentially with vanadium, to form, for example, stable vanadates or solid solutions of immobilized V(IV) [67]. 18.2.4 Light Alkenes by Addition of ZSM-5

Cofeeding of a second catalyst containing ZSM-5 to an FCC unit signiﬁcantly increases the yield of propene and butenes, mostly at the cost of gasoline yield, but with a gain in gasoline octane [68–70]. The yield shift implies conversion of linear and (mono)branched C6 –C9 alkenes (and, at ﬁrst sight, also alkanes) to mostly propene and butenes, leaving the gasoline fraction enriched in aromatics and isopentenes [21]. Product is also depleted in C6 –C9 alkanes but this is because their formation from primarily formed C6 –C9 alkenes is suppressed [71]. At very high inventories of ZSM-5 additive, yields of propene may in principle be doubled [72, 73]. To put this into perspective, at a high 8% propene yield, a very large FCCU would produce in the order of 200 kilotons per year, which is comparable to the propene output of a small naphtha cracker. The most important outlet for coproduced butenes is alkylation. At ﬁrst, a serious limitation in applying ZSM-5 in the FCCU appeared to be its hydrothermal stability. The structure itself is more stable than that of zeolite Y, but apparently, activity is lost at a higher relative rate by faster dealumination if no special measures are taken. Modiﬁcation of the additive with phosphorus was found to limit the (relative) activity loss and this is now standard practice [68]. The mechanism of phosphorus action is not well understood, but detailed NMR studies indicate the, perhaps reversible, formation of framework-bound, dispersed Al–O–P species, which may retard the formation of condensed alumina phases and a fully dealuminated structure (Figure 18.10) [74–76]. Application of phosphorus treatment to zeolite USY [77, 78] (as well as to zeolite Beta [79] and MCM-80) has been claimed but is to our knowledge not being practiced commercially. There are limits to what can be achieved in an unmodiﬁed FCC unit here. One solution, as in Sinopec and Stone & Webster’s DCC process, is to modify

559

560

18 Cracking and Hydrocracking SiO SiO

(OH2)y O(HPO3)xH

Al

SiO

O

O(HPO3)xH

P

O(HPO3)xH

Where y = 0,1,2; x = 0,1,2 ...

SiO SiO

Al

SiO

(OH2)y O(HPO3)xH2 O(HPO3)xH2

Where y = 0,1; x = 1,2 ...

O SiO SiO SiO

Al

(A) H3PO4

H O Si

SiO SiO SiO

Al

(OH2)y

(B) Steaming 800 °C

HOSi

SiO SiO

Al (OH2)y O

SiOH

O

Where x = 1,2 ... polyphosphates 1–4

O(HPO3)xH

P

Where y = 1,2,3

+ (H2O)(HPO3)x

O(HPO3)xH

P

O

O(HPO3)xH

Where y = 0,1; x = 0,1 ...

SiO SiO

Al

O O(HPO3)xH

P

SiOH O

O(HPO3)xH

Where x = 0,1 ...

SiO SiO

Al

(OH2)y

SiOH O(HPO3)xH Where y = 0,1,2; x = 1,2 ...

Figure 18.10 Proposed structure of framework aluminum-phosphate complexes formed in phosphorus-stabilized ZSM-[74, 75].

process [81, 82] and catalyst [83] to allow for deeper conversion, at somewhat longer (2–8 seconds) residence times and higher reactor (up to 580 ◦ C) temperatures [84–86]. This allows propene yields of some 17% in a typical case, and coproduces some ethene (∼5–6%). Since ideal conditions for heavy feed cracking to distillates and distillates cracking to light alkenes are not identical, separation of these unit operations is another way to stretch yields of the latter. Shell has developed the MILOS process, which uses, among other features, a separate riser/stripper system to allow conversion of recycled gasoline, which can then be made to operate at a higher temperature (but still low enough to minimize thermal reactions). This way, at 17% overall propene yield, signiﬁcantly less coke and dry gas are produced than by deeper cracking of the integral feed at high temperature [87]. Other dedicated catalytic cracking processes for enhanced propene production are UOP’s PetroFCC, Maxoﬁn (KBR), INDMAX (Indian Oil Corp./Lummus), NEXCC (Neste Oy), and HS-FCC (JPEC) [72, 88–90].

18.3 Hydrocracking

18.2.5 Potential Use of Other Zeolites in FCC

Possible application of zeolite Beta as a cracking component has been studied quite extensively, with the aim of increasing isobutane [91] or isobutene [92] yields, or overall butene yields [79]. Hydrothermal stability is poor compared to zeolite USY [91], which is why phosphorus stabilization would be in all likelihood necessary; it is anyway claimed to be feasible [79]. ITQ-7, whose pores are similar in size to those of zeolite Beta but which has differently shaped intersections, yields more gasoline of a higher alkene and lower aromatics content, by virtue of a reduced extent of hydrogen transfer [93]. ZSM-20[94] and ITQ-21[95], both with pore aperture sizes identical to those of USY, do not differ dramatically from USY in yield pattern, except for the higher LPG yield for ITQ-21 (including signiﬁcantly more propene). ITQ-33, with its amazing 12.2 A˚ 18-ring pore aperture, has already been demonstrated to be active in FCC, yielding more diesel from a vacuum gas oil (VGO) than USY at a given conversion, which is likely to translate into better performance for heavier feeds [96]. Medium pore-size zeolites other than ZSM-5 investigated as an FCC additive include MCM-22[97] and ITQ-13[98]. MCM-22 gives a similar increase in propene yield at lower loss in gasoline yield compared to ZSM-5, mainly because it forms less propane at the same time. ITQ-13, with its peculiar 9 × 10-ring pore structure, behaves similarly. The requirement that FCC catalyst components be low cost, in view of the fact that FCC units may be fed up to several tons per day of fresh catalyst, strongly favors application of catalysts based on zeolite Y and ZSM-5. Many of the newer structures require relatively expensive organics and/or long hydrothermal synthesis times, and although it is not inconceivable that low-cost synthesis routes may be developed for alternative materials, cost is likely to remain a signiﬁcant hurdle.

18.3 Hydrocracking 18.3.1 The Hydrocracking Process

Hydrocracking allows the conversion of VGO to middle-distillate products (diesel fuel, jet fuel, and kerosene), naphtha and LPG, with simultaneous hydrogenation [99]. ‘‘Hydrogen addition’’ can be seen as a different upgrading strategy from ‘‘carbon rejection’’ (see Section 16.2) andmany complex reﬁneries either have an FCC unit or a hydrocracker, where the hydrocracker allows for better yields of good-quality diesel fuel and kerosene. In the United States, where hydrocracking was ﬁrst introduced [100],4) hydrocrackers traditionally serve a different primary purpose, viz. to produce naphtha (as a contribution to the gasoline pool) from 4) The origins of hydrocracking may be traced

back to the late 1920s, when IG Farben

developed a process for converting lignite into gasoline.

561

562

18 Cracking and Hydrocracking

Two-stage hydrocracker

Single-stage hydrocracker

Fresh gas

Fresh gas Recycle gas

Recycle gas

Recycle compressor HP separator First stage

Feed

Quench gas

HP separator

Fractionation

370 °C−

Recycle compressor

Fractionation

Second stage

LP separator

(a)

370 °C−

370 °C+ LP separator

∞

Feed

370 °C+

∞

(b) Figure 18.11 (a) Two-stage hydrocracking process scheme hydrocracking process scheme. (b) Single-stage hydrocracking process scheme.

low-value aromatic streams including LCO product from the FCC unit. But also there, hydrocrackers are increasingly shifting to production of middle distillates. Hydrocrackers can be very large units, with feed intakes up to 3 million tons per year. Different hydrocracking processes exist [101–105], with different process conﬁgurations. Most hydrocrackers have ﬁxed-bed reactors in which the liquid oil feedstock and gaseous hydrogen are passed in downﬂow over the catalyst beds (so-called trickle ﬂow), and typically operate at pressures in the range of 80–200 bar and at temperatures in the range of 300–450 ◦ C. Two basic schemes, which are widely applied, can be identiﬁed: two-stage hydrocracking and single-stage hydrocracking (see Figure 18.11). Two-stage hydrocracking uses separate hydrotreating (HDS, HDN, and HDO) and cracking stages with interstage separation of products, H2 S and NH3 . Organic nitrogen and NH3 are poisons to the acidic cracking catalyst and in this lineup, the levels can be kept low, which is a requirement if the cracking catalyst is based on amorphous silica–alumina (the standard before zeolitic hydrocracking catalysts became more common in the 1970s) [106]. Partial pressures of H2 S in the cracking stage can be brought down to levels that allow the use of noble metals as the hydrogenation function, if so desired. Typically, cracking conversions are kept between 30 and 70% per pass to achieve good selectivity to naphtha or middle distillates, and can be brought close to 100% overall by recycling unconverted material. Precracking in the ﬁrst stage can also be achieved by applying stacked beds containing separate pretreating and cracking catalysts [107]. Single-stage (or sometimes series-ﬂow) hydrocracking is a simpler lineup without the interstage separation (Figure 18.11b). Its application has become more widespread after the introduction of zeolitic cracking catalysts, which are not only

18.3 Hydrocracking

more active than the amorphous silica–alumina-based systems but also, remarkably, less sensitive to ammonia. Both once-through, partial conversion units and units applying recycle exist. Hydrowax, the noncracked product from partial conversion units, can be very favorably used as feedstock in ethylene and catalytic crackers, and in the manufacture of lubricating base oils. Various alternative lineups have been applied [101, 108]. One important variety is the so-called mild hydrocracking (MHC) process, which operates at lower hydrogen partial pressures (40–80 bar), and moderate conversion [109–111]. The major reasons for applying the MHC process commercially are its lower cost, and the suitability to use the unconverted oil as a feedstock for the catalytic cracker. Integration of the MHC process with hydrotreating of gas oil fractions from various sources can be a very attractive way of both alleviating the inherent deﬁciencies of MHC, that is, production of relatively low-quality diesel, and upgrading gas oils to diesel meeting ultralow sulfur speciﬁcations [112]. 18.3.2 Feedstocks and Products

The most common feedstock outside North America is straight-run VGO with a typical 350–600 ◦ C boiling-point range. In the United States, lighter feedstocks are generally processed, such as straight-run light and heavy gas oils, coker gas oils (from coker processes) and cycle oils (from catalytic crackers). Also the heavy and VGO fractions of the synthetic crude oil derived from tar sands in Alberta, Canada, are used as hydrocracker feedstock [113]. Other processable feedstocks are deasphalted oils (DAOs) [114] and VGOs from residue conversion units [115]. As a rule, the amount of catalyst poisons (metals, aromatic coke precursors, nitrogen) increases with feedstock heaviness [116]. Polynuclear aromatics inhibit catalyst activity by (initially reversible) adsorption [117, 118] and can moreover serve as coke precursors [119–121]. Metals cause irreversible deactivation of the ﬁrst-stage hydrotreating catalyst, whereas organic nitrogen compounds require deep conversion in the ﬁrst stage [116], because they speciﬁcally reduce the cracking activity of the acidic second-stage catalysts as a result of their strong Effect of nitrogen compounds in feed and of temperature on hydrocracking conversion rate [122].

Table 18.1

N addition to feed

Nonea 1000 wt-ppm N as quinolone 1000 wt-ppm as pyrrole a Feed

contained 2 wt-ppm N.

Relative conversion T = 300 ◦ C

T = 370 ◦ C

1 0.05 0.07

– 1.10 1.15

563

564

18 Cracking and Hydrocracking Table 18.2

Typical nitrogen contents of ﬁrst-stage hydrocracker feedstocks [116].

Feedstock type

Nitrogen content (wt ppm)

Straight-run vacuum gas oil Arabian Light Iranian Heavy Brent Ria Juano Pesado Minas Deasphalted oil from Arabian Light vacuum residues Heavy catalytically cracked cycle oil Vacuum gas oil from thermal cracking of atmospheric residue Vacuum gas oil from hydrotreating of atmospheric residue

1000 2000 900 1700 700 1500 2000 3000 2000

adsorption (see Table 18.1 [122]). Some typical nitrogen contents of hydrocracker feedstocks are given in Table 18.2. Although the effects of NH3 (produced by HDN reactions) on catalyst performance are discussed below in more detail, in general, the inﬂuence on catalyst activity and selectivity is very signiﬁcant. Cofeeding 2000 ppmw NH3 results in an enormous loss of cracking activity of a typical second-stage catalyst, to the extent that a temperature increase in the order of 100 ◦ C is needed to maintain the same level of activity [123]. Concomitantly, the product selectivity to middle distillates is considerably enhanced [123, 124]. Clearly, this large impact of NH3 on catalyst performance provides an opportunity to balance or tune activity and selectivity to speciﬁc process needs. Table 18.3

Product yields in hydrocracking of heavy virgin gas oil [125].

Feedstock properties Density (g.cm –3 ) Sulfur (wt%) Nitrogen (wt ppm) Boiling range (◦ C) Product compound

0.92 2.9 820 316–538 Maximum product objective, yields on feed (vol%) Gasoline Turbine fuel Diesel fuel Heating oil

Butane Light gasoline Heavy gasoline Jet fuel Diesel fuel Heating fuel Chemical H2 consumption (wt%)

14.5 31.7 78.9 – – – 3.4

8.3 17.2 28 64.4 – – 2.8

4.9 11.5 19.6 – 81.1 – 2.5

3.7 11.8 13.8 – – 87.4 2.3

18.3 Hydrocracking Product yields and properties in once-through hydrocracking of a gas oil/deasphalted oil blend [114].

Table 18.4

Feedstock Boiling range (◦ C) 10% 50% Final boiling point

406 556 596

Operating conditions

Moderate conversion

Conversion (vol%) H2 partial pressure (bar) Temperature in cracking catalyst (◦ C) Product yields (vol%) Light naphtha Heavy naphtha, 85–160 ◦ C Diesel fuel, 260–350 ◦ C FCC feed, 350 ◦ C+ H2 consumption (wt%) Product properties Light naphtha Density (g.cm –3 ) RON, clear Heavy naphtha Density (g.cm –3 ) P/N/A (vol/vol/vol) Product properties Turbine fuel Density (g.cm –3 ) Smoke point (mm) Diesel fuel Cetane number FCC feedstock Density (g.cm –3 ) Nitrogen (wt ppm) Aniline pointa (◦ C) a According

Recovery (%) Density (g.cm –3 ) Nitrogen (wt ppm) Sulfur (wt%)

68 900 900 0.3 High conversion

61 148 Base

77 148 7

13.3 21.1 9.2 39.1 1.88

24.1 19.4 6.9 23.2 2.31

0.654 80

0.654 81

0.739 47/49/4

0.735 47/49/4

0.802 28

0.797 30

65

66

0.865 <5 140

0.871 – –

to ASTM D 611-82.

One of the major advantages of hydrocracking compared to, for example, catalytic cracking is the ﬂexibility in the distribution of the various products that can be manufactured (see Table 18.3 [125]) Hydrocracked products are low in sulfur and aromatics, and are regarded as high quality, both from a product performance and an environmental point of view (e.g., compared to FCC products [126]). The tightening of product speciﬁcations in, for example, the United States and the European Union even makes the design, the catalysts, and the operation

565

566

18 Cracking and Hydrocracking

of the hydrocracker more critical. Table 18.4 gives an example of product yields and properties obtained by once-through hydrocracking of a gas oil/deasphalted oil feedstock [114]. Diesel fuel ignition quality (cetane number) is one important product property well served by hydrocracking: it could further beneﬁt from ring-opening reactions [127], and hydrocracking catalysts are also sought to catalyze these, but a recent review [128] details the general ﬁnding that classical hydrocracking catalysts may not be the best choice for this particular reaction. 18.3.3 Hydrocracking Catalyst Systems, and Catalytic Chemistry

Pretreat catalysts (in the ﬁrst stage) usually consist of phosphorus-promoted [129] nickel–molybdenum sulﬁdes on γ -alumina (NiMoP/Al2 O3 ), and for reaction temperatures below some 380 ◦ C, in the presence of sulfur, no systems with higher HDN activity have been found [12]. The proper hydrocracking catalysts (usually in the second stage) are dual functional and therefore choices need to be made as to which (de)hydrogenation and (acidic) cracking function to apply. The common dual-functional alternatives are shown in Table 18.5. In general, Ni/Mo and Ni/W mixed sulﬁdes are applied – the former (Figure 18.12) usually in situations were some extra HDN activity is required, and the latter where hydrogenation performance is of paramount importance. In some cases where the concentrations of sulfur are low (i.e., two-stage conditions), a (noble) metal like Pd may be used, which is a very powerful hydrogenation catalyst. Such Pd-promoted catalysts are generally more active than the mixed-sulﬁde containing ones and the products are also very well hydrogenated. On the debit side are the high sensitivity to sulfur and the cost of the metal (and of keeping the hydrogen sulﬁde pressure reliably low). Alumina, even after ﬂuoridation [130], is obviously the least active in cracking, but has speciﬁc merits as an acidic hydrocracking function in, for example, converting nitrogen-rich feeds, because it is at the same time the best support for the hydrotreating function. Exposed W sites in NiW/alumina are also reported to have cracking activity (presumably not involving Brønsted acidity) [131]. More acidity is provided by amorphous silica–aluminas (ASAs). ASAs can be prepared in a variety of ways, and over a range of alumina contents – they usually have Table 18.5

Bifunctional hydrocracking catalysts. Hydrogenation function

Acidic function (support)

Increasing

  Ni/Mo

Al2 O3

Hydrogenation Power

Ni/W Pt/Pd

Al2 O3 /halogen SiO2 /Al2 O3 Zeolites (low-S conditions)

  Increasing

Acidity

18.3 Hydrocracking

Various nitrogen levels 390

Reactor outlet temperature for 65% conversion (°C)

380 2000 ppm

370 360 500 ppm N

350 340

100 ppm N

330 320 310 300 290

No N

280 270

0

200

400

800

Runtime (h) Figure 18.12

Activity of a NiMo/zeolite catalyst for various ammonia levels [123].

a rather high surface area (say, 400–500 m2 .g−1 ). At ﬁrst, the acidity increases with alumina content, but beyond 25 wt% the cracking activity decreases again, as shown by Ward [132]. Since with varying alumina content also its dispersing power for the hydrogen function (in particular the base-metal ones) changes, it is not easy to deﬁne from ﬁrst principles the optimum ASA. Improved metal-emplacement routes can still lead to improved catalysts [133]. Often ASAs are considered to have a rather small density of acid sites, of high strength (we discuss this later in the text, Figure 18.20) – this makes them less acidic than zeolites, and also more prone to poisoning, by, for example, nitrogen compounds. Zeolites provide in general by far the highest cracking activity. Compared with ASAs they also show a more stable performance. On the other hand, they are in their own right poor supports for NiMo and NiW hydrogenation functions, and may suffer from mass-transport limitations, leading to lower selectivities to desired products and/or increased gas make. These aspects will be further discussed below. A great deal of study has been devoted to unraveling the chemistry of the HC process [99, 134]. A simpliﬁed mechanistic scheme, adequate for basic design purposes, is presented in Figure 18.13. This scheme is primarily derived from model studies related to the hydroconversion of normal alkanes [135–140], mixtures, and cycloalkanes [141, 142], including descriptions of the fundamental kinetics. The ﬁrst step in this mechanism is the dehydrogenation of (normal) alkanes to produce alkene intermediates, at very low near-equilibrium concentrations, that subsequently react on the zeolite Brønsted-acid sites to form zeolate alkoxides, analogous to, for example, the addition of sulfuric acid to an alkene (in which case, the subsequent chemistry is also expressed in the ‘‘shorthand’’ of carbenium ion chemistry). The alkoxides can then undergo the usual acid-catalyzed elementary

567

568

18 Cracking and Hydrocracking

Type F

I

C

F=

I=

F=

I=

Ions involved

A

tert

tert

C=

B1

sec

tert

C=

B2

tert

sec

M

Example + +

+

+

+

+

+

A

+

Coke

C

sec

+

+

sec

+ +

tert = tertiary ; sec = secondary Figure 18.13 (a) Simpliﬁed reaction scheme for hydrocracking. F = feed molecule, I = isomerized molecule, C = cracked molecule, M = metal site, A = acid site. F= , I= , C= = alkenes derived from the respective

molecules. (b) Cracking by b-scission occurs more easily for some isomers than for others, and carbenium ion stability is a good proxy for activation energy [138].

reactions, viz. ﬁrst isomerization and subsequently beta-scission cracking. The isomerized and/or cracked species are then desorbed from the acid site and hydrogenated to the corresponding alkane. In addition, intermediate alkenes can react in bimolecular H-transfer reactions with cycloalkanes in a process that may ultimately lead to coke [143, 144]. The fact that hydrocracking can be carried out in steady state (permitting cycles of several years in the reﬁnery) whereas FCC catalysts deactivate in seconds is largely attributable to the fact that alkene concentrations are kept very low. In practice, coke levels and deposition rates decrease with increasing hydrogen partial pressure, through this mechanism and also by prevention of condensation reactions involving polyaromatic molecules. As a result, with ‘‘real’’ feeds, rates are usually increased with increasing hydrogen partial pressure, in contrast to the negative order expected on the basis of model kinetics for n-alkanes [110], with only very few exceptions. Close proximity of hydrogenation sites and acid sites should also help to reduce reactions contributing to coking [145–148] insofar as they are diffusion limited. From the above scheme, Figure 18.13, an obvious design parameter is the relative importance of the (de)hydrogenation and the cracking functions. When the former dominates over the latter, so-called ‘‘ideal HC’’ conditions exist whereby consecutive reactions (after the isomerization and the primary cracking steps) can be limited to the minimum, since high rates of (de)hydrogenation increase the concentration of feed-derived alkenes relative to primary cracking product alkenes [149]. Hydrogenation and cracking functions should also be in relatively close proximity, so that the equilibrium composition of alkenes is maintained at the acid sites without distortion from diffusion limitations [150, 151]. When the rate of (de)hydrogenation is comparable to that of cracking, multiple cracking events can occur and lighter products form. In this case, a higher iso/normal ratio will be observed in the products than in the ideal case, since secondary isomerization (i.e., of primary cracked products) is now possible. The results of a classical

Moles per 100 moles C16 cracked

18.3 Hydrocracking

140

Catalytic cracking: SiO2– Al2O3– ZrO2 T = 500°C; Conversion = 54%

120 100

Hydrocracking: CoMo – S/ SiO2– Al2O3 T = 400°C; Conversion = 50%

80 60

Hydrocracking: Pt / Ca Y T = 230°C; Conversion = 55%

40 20 0 2

4

6

8

10

12

14

Carbon (number of cracked products) Figure 18.14 Molar carbon number distributions in catalytic cracking and hydrocracking of n-hexadecane at 50% conversion [149].

study of ideal versus nonideal HC are shown in Figure 18.14 (cf. also Sie [152]). Quantitative modeling of this has greatly progressed [153–155]. Qualitatively, the same principles can be applied in commercial practice [106]. For example, a high naphtha selectivity requires a strongly acidic catalyst with a relatively weak (de)hydrogenation function – the resultant relatively high iso/normal ratio being beneﬁcial for the naphtha quality too (higher octane numbers). By contrast, for middle-distillate selective catalysts, the reverse is true, and a relatively high ratio of (de)hydrogenation/acidity function is desired for both yield and product quality requirements. Analogously, the effect of ammonia, which moderates the acidic function, is to lower activity and increase selectivity to primary cracking products as can be seen in Figure 18.15 [123].

C11 – C24 Selectivity (%)

70 60

Zeolite (second stage) Zeolite (first stage) Amorphous (second stage)

50 40 30 20 10 0 320

Figure 18.15

340 360 380 400 Reactor temperature for 60% C24 - conversion (°C)

420

Middle-distillate selectivity versus activity [123].

569

18 Cracking and Hydrocracking

18.3.4 Zeolite Y in Hydrocracking

As indicated in Table 18.5, zeolites generally exhibit relatively high acidity and are therefore a preferred catalyst ingredient, particularly where high activity levels are required. To date, almost all commercial hydrocracking catalysts have been based on zeolite Y (USY or VUSY), simply because it has the most accessible pore system, although many others have been tried, for example, L, mordenite, omega in the late 1960s, and later X, beta, and ZSM-5[156]. Even zeolite Y has problems of access by molecules in the higher boiling range [157, 158], and it is for this reason that sometimes an ASA is added [159]: it is much less acidic but it does not induce any diffusion limitations, thus serving to convert the species that the zeolite cannot cope with [160] (see Figures 18.16 and 18.17). Zeolite Y can be modiﬁed by combinations of steaming and/or calcination, ion-exchange, leaching, and various chemical treatments [161, 162], with the aim of controlling acid-site density, nonframework aluminum species and mesoporosity, while preserving a high crystallinity [99, 163, 164]. The experimental space set up by these techniques (see Figure 18.18) is so enormous that to this day new, more effective modiﬁcations of Y continue to be found.

First-order rate constant for hydrocracking relative values k ′ to boiling point x

570

1.8 Ni / W/ SiO2, Al2O3 (395 °C) 1.4

1.0 Ni / W/ zeolite Y (355 °C) 0.6

0.8

300

400

500

600

Boiling point (°C) Figure 18.16 First-order HC rate constant as a function of boiling-point fraction for NiW/zeolite Y and NiW/ASA catalysts with a Middle East ﬂashed distillate feedstock [160].

18.3 Hydrocracking

3

wt.-%

2

1

0 2

0

−2

(a)

−4 −6 Z - number

−8

75 65 55 r 45 be 35 m 25 nu n 15 bo 5 ar C −10

3

wt.-%

2

1

0 2

0

−2

(b)

−4 −6 Z - number

−8

−10

75 65 55 45 r be 35 m u 25 n 15 on rb 5 a C

3

wt.-%

2

1

0 2 (c)

0

−2

−4

Z - number

−6

−8

75 65 55 45 er 35 mb 25 nu n 15 bo r 5 Ca −10

Figure 18.17 FIMS spectra of the unconverted material after hydrocracking treated VGO over (a) all-zeolite, (b) all-amorphous, and (c) composite catalysts (two-stage operation). Z indicates the hydrogen content in the hydrocarbons Cn H2n + z.

571

18 Cracking and Hydrocracking

Synthesis NaY

SAR

a 0 (Å)

5

24.64

5

24.50

5–7

24.53

IE-1 NH4,NaY Steam calcine H,NaUSY IE-2 USY Steam calcine H-VUSY

24.30 IE-3

VUSY

9.6

24.32

Figure 18.18 Schematic representation of the zeolite Y dealumination process steps. (SAR = SiO2 /Al2 O3 ratio, a = unit-cell dimension, IE = ion exchange, USY = UltraStable zeolite Y, VUSY = Very UltraStable zeolite Y.

Zeolite-Y crystallographic unit cell, as a crude measure of acid-site density [162, 165], has been shown to affect the middle-distillate selectivity (kerosene) of HC catalysts (see Figure 18.19) [126]. Reducing the acid-site density while maintaining a constant hydrogenation activity should reduce secondary cracking, as is observed. Perhaps surprisingly, the overall catalyst activity remains relatively high despite the reduction in acid-site density. This can be understood, at least in part, as being due

Kerosene selectivity

572

Unit cell size Figure 18.19

Inﬂuence of unit-cell size on product selectivity [126].

18.3 Hydrocracking

to a compensating effect of reduced coke formation, as the hydrogenation function now has to keep fewer sites clean. Conversely, if naphtha is the desired product, a high concentration of acid sites is desired, and USY the cracking function of choice. HC catalysts based on zeolite-Y materials with similar unit-cell dimensions can still display signiﬁcantly different performances [165], and obviously for a more detailed picture, NMR, FTIR, back-exchange experiments, and so on can help. The role of extra-framework aluminum (EFAl) in modifying zeolite acidity continues to be discussed [166]. We subscribe to the view that there is an EFAl species that enhances the intrinsic acid strength of zeolite Y Brønsted sites [167]: as indicated in Figure 18.20 EFAl-free zeolite Y contains sites with a lower intrinsic ability to hydroisomerize n-heptane, a Brønsted-acid catalyzed reaction, than does regular (ultrastabilized) Y. This has, of course, consequences for HC catalyst design: when applying materials with enhanced acid strength, one needs to provide a stronger hydrogenation function to balance this acidity. Figure 18.20 also includes some data points pertaining to ASA materials: they tend to indicate that the intrinsic acid strength of ASA sites is as high as that of the ‘‘enhanced’’ Y materials, but their number is just much lower, thus explaining the rather large difference in cracking activity between the two materials. To explore the effect of still lower acid strengths than that available in EFAl-free zeolite Y, but retaining the FAU structure, one could either evaluate a material like SAPO-37 [49, 168] or try and substitute Fe3+ or Ga3+ in USY [169]. The problem with SAPO-37 is (apart from its water sensitivity), however, that it has very large crystallites, leading to mass transfer limitations (see below). And although we were able to synthesize a nice RE–Ga–USY material (RE = rare earth), it turned out to be unstable under HC testing conditions [170]. Indeed, a little hydrogen is already sufﬁcient to dislodge Ga3+ from the faujasite framework.

Faujasites

n (N isomerization)

With enhanced sites

Without enhanced sites

ASAs

n (N OD) Figure 18.20 Log–log plot of n-heptane hydroisomerization activity versus number of strong acid sites (as estimated by the H/D exchange method/mmol.g−1 ) [167].

573

18 Cracking and Hydrocracking

Nonmodified zeolite Y

Pore volume

574

5

10

50

100

500

Pore diameter (Å) Modified zeolite Y

5

10

50

100

500

Pore diameter (Å) Figure 18.21

Creation of a secondary pore structure as a result of dealumination [104].

Mesoporosity brought about by ultrastabilization of zeolite-Y materials [104] (Figure 18.21) has been shown to reduce mass-transport limitations during hydrocracking and thereby suppress secondary cracking. For heavy feeds, the activity is also improved by enhanced accessibility, as has been recently conﬁrmed in a comparison between HC of tetralin and of atmospheric residue [171]. A recent PQ patent describes stabilized Y materials with much enhanced mesoporosity [172], and it turns out that indeed such materials can lead to increased middle-distillate selectivity. It has even been claimed that the selectivity pattern of mesoporous zeolite-Y approaches that of an ASA [104], although this has not been well substantiated [148]. As yet, the inﬂuence of zeolite-Y mesopore structure has not been considered in model studies. The presence of mesoporosity in crystals of several hundreds of nanometers [173] also holds against the approach to apply even smaller crystals of zeolite Y, despite some claims to success [174]. Dispersing noble metals on zeolites is relatively straightforward [175], but for the base-metal catalysts, the marriage between group 6 elements and zeolites is somewhat unhappy: the aqueous chemistry applicable for catalyst preparation causes W and particularly Mo [176] to strongly prefer the alumina or silica–alumina binder, a digression from the idealized homogeneous bifunctional catalyst. Having Mo or W directly on an acidic support, that is, the zeolite, could moreover improve its activity per se, as has been reported for ASAs [177]. Efforts to address this problem using alternative chemistries remain somewhat unpractical on a commercial scale [178]. The effect on catalyst stability may be limited anyway because of the zeolite’s intrinsically better resistance to spatially demanding coke-forming reactions. Indeed, in commercial operation zeolitic catalysts tend to be signiﬁcantly more stable than ASAs (see Figure 18.22). Recently, ZSM-12 has been singled out as being especially advantageous in this respect [179].

Temperature required for 50% conversion (°C)

18.3 Hydrocracking

400 Amorphous

380 360 340

Zeolite 320 300

0

100

200

300 400 Time (h)

500

600

Figure 18.22 Deactivation rates of amorphous and zeolite catalysts in the hydrocracking of a hydrotreated vacuum gas oil [126].

18.3.5 New Catalyst Developments

Recent developments have been mainly incremental improvements in the ASA and Y base materials and the control of catalyst architecture, sometimes augmented by improved metal-emplacement routes. And better catalysts can still be found in this way. Improvement of the traditional mixed-sulﬁde hydrogenation functions has also been sought through promotion (by, e.g., Nb) or even through replacement (by, e.g., RuSx -based systems), without apparent breakthroughs so far. Beneﬁts have been claimed for mixtures of zeolites (e.g., USY/VUSY) [125] and zeolite beta is sometimes canvassed as well [180]. Another development involves shifting of cracking duty to the ﬁrst stage, enabled by use of zeolitic catalysts in the bottom bed there. It is sometimes advantageous to have NiMo rather than NiW as the metal function on such bottom bed catalysts to improve the denitrogenation activity. However, this is normally achieved at the cost of some loss in aromatics hydrogenation performance [116]. To improve our understanding of where gains in activity selectivity are still possible, we would need to be able to determine the molecular composition of the feed, and to model the HC process in sufﬁcient detail. In both these areas, good progress is being made [181–184], but we are not there yet. Wider-pore zeolites have perennial promise but the hypothetical very open pore architectures needed to improve on the performance of zeolite Y have so far remained beyond reach. This is despite quite amazing progress in the ﬁeld, having brought a number of 1D 14-ring pore structures, and most recently a novel 18 × 10-ring pore material, ITQ-33 [96]. In all likelihood, materials with effective 1D pore systems likely to display single ﬁle diffusion are of limited potential in hydrocracking. Mesoporous aluminosilicate materials with pore diameters in the range of 2–10 nm like MCM-41/-48 and SBA-n [185], which do not possess short-range order, only appear to be very expensive ASAs as far as HC is concerned. A recent development, however, is to have the walls

575

576

18 Cracking and Hydrocracking

of the regular pores consist of small zeolitic nuclei (beta, Y) [186, 187] or indeed larger, continuous crystals [188], and these materials certainly merit close scrutiny. However, the combination of 12-ring zeolites, however small their crystallites, with larger mesopores, does not bode too well for their stability, even if they were to achieve higher selectivities. Going to smaller mesopores (say, 1–1.5 nm) might also be interesting [189]. Pillared clays have also come in for their share of attention [190], but did not emerge as commercial catalysts, one problem being that they coke up rather easily. 18.3.6 Residue Conversion – Some Notes

There is a major economic incentive to extend the current HC processes to enable heavier feedstocks to be converted to lighter, higher-value transportation fuels, but it is not obvious that zeolitic catalysts offer the best prospects here. Earlier studies by Idemitsu indicate that iron-modiﬁed zeolite catalysts signiﬁcantly enhance conversion when heavy oils such as atmospheric residue are hydroprocessed [9]. Nevertheless, high conversions and product selectivities are difﬁcult to achieve with truly heavy feeds (end boiling points beyond 620 ◦ C), because such feeds contain relatively high concentrations of polyaromatics and heavy nitrogen compounds, leading to severe catalyst deactivation, exacerbated by the fact that the higher operating temperatures required with heavier feeds tend to shift the hydrogenation equilibria toward the aromatic side (this is obviously not good for the product properties either). Higher hydrogen pressures could be applied as a remedy, but this generally imposes some severe economic penalties (e.g., beyond 170 bar). Furthermore, although zeolite-based catalysts offer the best prospects for heavy feeds conversion owing to their relatively low coking propensity, their activity is suppressed owing to the (very) restricted access of the heavier molecules to the zeolite pores – which may also entail a loss of product selectivity, as the zeolites now tend to be more active for cracking the initial products than the feedstock. One answer has been to switch to a rather different process design, viz. performing the HC in an ebullated bed, in which continuous withdrawal of deactivated catalyst from the system is practiced (LC-ﬁning, H-Oil, T-Star etc.). Because of the shorter average catalyst age it is easier to cope with catalyst deactivation. But as these processes generally proﬁt also from a relatively high thermal contribution to the cracking, further discussion of such designs is outside the scope of this chapter [191].

18.4 Summary and Outlook

Technological progress in reﬁning has fed on zeolite science and is in turn feeding zeolite science. A quote from Haw et al. [192] appropriately illustrates this relationship from the scientiﬁc viewpoint, even though it relates to the methanol-to-oleﬁns chemistry which is outside our scope:

18.4 Summary and Outlook

‘‘MTO (is) a unifying problem in catalysis. The active site for MTO catalysis is not a proton on a rareﬁed piece of clay. It is an organic–inorganic hybrid that will come to be understood through a fusion of structure activity and topological concepts. The detailed reactions of the hydrocarbon pool encompass alkylation and isomerization of methylbenzenes, homologation, oligomerization, cracking, and isomerization of oleﬁns, and cyclization and hydride transfer reactions inherent in disproportionation to aromatics and alkanes. Thus, a fuller understanding of MTO catalysis will necessarily beneﬁt other important, fundamental problems in hydrocarbon catalysis on solid acids. MTO hydrocarbon pool mechanisms involve distinct supramolecular species with clearly deﬁnable chemical properties. Thus, MTO catalysis will continue to be a very attractive test case for the application and reﬁnement of theoretical modeling and in situ spectroscopy to catalysis. The molecular vocabulary of MTO catalysis invites synergy between homogeneous and heterogeneous catalysis.’’ Study of industrially relevant problems has thus revealed unimagined complexity in zeolite-catalyzed systems, and yet in some cases allowed their understanding to a level of detail – though often still less than fully satisfactory – that is rare in the ﬁeld of heterogeneous catalysis. So, what challenges remain here? On a scientiﬁc level, one would ideally like to have all pieces of the puzzle linking theory, spectroscopy, catalysis in model reactions, and catalysis in the industrial systems. As there is still a struggle to understand what deﬁnes rates of very simple hydrocarbon conversion reactions in zeolites [193], we are still some way away from that. In FCC catalysis, a better understanding of the nature of catalyst activity, as discussed in Section 16.2.2, would be helpful. Hydrocracking is more thoroughly understood and the link to the more fundamental level is less problematic there. As mentioned in the opening of this chapter, the FCC process faces several challenges, and, in concert with modiﬁcations of the process, catalyst improvements have helped turn it from a gasoline factory into a more ﬂexible unit operation, enabling the global gradual shift from gasoline to distillate fuels and steadily growing demand for petrochemical feedstocks. As outlined above, the requirement that FCC catalyst components be low cost remains a difﬁcult hurdle for application of zeolite structures other than Y and ZSM-5. Nevertheless, the steady increase in the number of structures with very large pores will undoubtedly lead to new ﬁndings on how such structures can best be exploited to convert heavy feeds, which can then provide incentive for further research to take that hurdle, for example, through lower-cost synthesis or improved stability. The same research will undoubtedly beneﬁt hydrocracking, where process conditions are more forgiving, and chances of implementing new materials concomitantly higher. Doubtless, in the challenge that we face to make the best possible use of our conventional oil, nonconventional oil, and natural gas supplies, society will need this increased understanding and mastery of zeolites and zeolite catalysis.

577

578

18 Cracking and Hydrocracking

References 1. Rosinski, E.J. (1992) The Origin and

2. 3. 4.

5.

6. 7. 8.

9. 10. 11. 12.

13. 14.

Development of the First Zeolite Catalyst for Petroleum Cracking in Inventive Minds: Creativity in Technology (eds R.J. Weber and D.N. Perkins), Oxford University Press, pp. 166–177. Rabo, J.A. and Schoonover, M.W. (2001) Appl. Catal. A, 222, 261. Nakamura, D. (2008) Oil Gas J., 106 (48), 46. BP Statistical Review of World Energy (2008), at www.bp.com/statisticalreview (accessed in 2009). (a) Maesen, Th. (2007) The Zeolite Scene–An Overview in Stud. Surf. Sci. Catal., vol. 168 (eds J. ˇ Cejka, H. van Bekkum, A. Corma, F. Sch¨uth), Elsevier, Amsterdam, pp. 1–12.; (b) Lauriente, D.H. and Inoguchi, Y. in The Chemical Economics Handbook, SRI international, pp. 599.1000A– 599.1002.K; see also online summaries of The Economics of Zeolites Roskill, (2003) on http://www.roskill.com/reports/zeolites and Zeolites – Industry Trends and Worldwide Markets in 2010, Hewin International Report (2001). Degnan, T.F. (2001) Top. Catal., 13, 349. Stell, J. (2005) Oil Gas J., 103 (47), 50. BCC Research (2007) Report Summary, Catalysts for Environmental and Energy Applications. http://www.bccresearch.com/report/ CHM020C.html (accessed in 2008). Maxwell, I.E. and Stork, W.H.J. (1991) Stud. Surf. Sci Catal., 58, 571. Sherman, J.D. (1999) Proc. Natl. Acad. Sci. U.S.A., 96, 3471. Mokyr, J. (2002) Gifts of Athena, Princeton University Press, Princeton . Blauwhoff, P.M.M., Gosselink, J.W., Kieffer, E.P., Sie, S.T., and Stork, W.H.J. (1999) in Catalysis and Zeolites – Fundamentals and Applications (eds J.Weitkamp and L. Puppe), Springer, Berlin, p. 437. Maxwell, I.E. and Stork, W.H.J. (2001) Stud. Surf. Sci Catal., 137, 747. Weyda, H. and K¨ohler, E. (2003) Catal. Today, 81, 51.

15. (a) International Energy Agency, Oil

16. 17.

18. 19. 20. 21. 22.

23.

24.

25.

26.

27.

28.

29.

Market Report, Paris, France, May 2006, p. 14, http://omrpublic.iea.org/ (accessed in 2008). (b) Energy Information Agency, U.S. Department of Energy (2006) International Energy Outlook 2006 . Report #:DOE/EIA-0484, http://www.eia.doe.gov/oiaf/ieo/ index.html (accessed in 2008). Ross, J., Lepage, J.P., and Confuorto, N. (2003) Hydrocarbon Process., 82, 47. Directive 2003/17/EC of the European Parliament and of the Council, English Version, online on http://eur-lex.europa.eu/LexUriServ/site/ en/oj/2003/l 076/l 07620030322en 00100019.pdf. Silvy, R.P. (2004) Oil Gas J., 102 (16), 58. Stell, J. (2004) Oil Gas J., 102 (16), 26. Scherzer, J. (1993) Stud. Surf. Sci. Catal., 76, 145. Adewuyi, Y.G. (1997) Appl. Catal. A, 163, 15. Cerqueira, H.S., Sievers, C., Joly, G., Magnoux, P., and Lercher, J.A. (2005) Ind. Eng. Chem. Res., 44, 2069. (2005) Europe/Middle East Report Oleﬁns and Derivatives, vol. 224, CMAI, London, p. 14. Brunet, S., Mey, D., P´erot, G., Bouchy, C., and Diehl, F. (2005) Appl. Catal. A, 278, 143. Thakkar, V.P., Abdo, S.F., Gembicki, V.A., and McGehee, J.F. (2005) Petr. Quarterly, Q3, 41. Naber, J.E., Barnes, P.H., and Akbar, M. (1988) Japan Petroleum Institute, Petroleum Reﬁning Conference, October 19-21, 1988, Tokyo. Cerqueira, H.S., Caeiro, G., Costa, L., and Ramˆoa Ribeiro, F. (2008) J. Mol. Catal. A, 292, 1. Nishihara, H., Yang, Q.H., Hou, P.X., Unno, M., Yamauchi, S., Saito, R., Paredes, J.I., Mart´ınez -Alonso, A., Tasc´on, J.M.D., Sato, Y., Terauchi, M., and Kyotani, T. (2009) Carbon, 47, 1220. Harding, R.H., Peters, A.W., and Nee, J.R.D. (2001) Appl. Catal. A, 221, 389.

References 30. Dupain, X. (2006) Fluid Cat-

31. 32. 33.

34. 35.

36.

37. 38.

39.

40.

41.

42.

43. 44.

45. 46.

alytic Cracking – Feed Stocks and Reaction Mechanism, Ph.D. Thesis, Delft University of Technology, ISBN 90-902-0417-2, http://repository.tudelft.nl/ﬁle/96711/ 078220 (accessed in 2008). Scherzer, J. (1989) Catal. Rev. – Sci. Eng., 31, 215. Davison Catalagram, No 76 (1987). Stockwell, D.M., Liu, X., Nagel, P., Nelson, P.J., and Gegan, T.A. (2004) Stud. Surf. Sci. Catal., 149, 257. Liu, H., Ma, J., and Gao, X. (2006) Catal. Lett., 110, 229. Haag, W.O. and Dessau, R.M. (1984) Proceedings of the 8th International Congress on Catalysis, July 2–6, 1984, Berlin, Germany, vol. 2, Verlag Chemie, Weinheim, p. 305. Wojciechowski, B.W. and Corma, A. (1986) Catalytic Cracking, Marcel Dekker, New York. Abbott, J. and Wojciechowski, B.W. (1988) Can. J. Chem. Eng., 66, 817. Williams, B.A., Ji, W., Miller, J.T., Snurr, R.Q., and Kung, H.H. (2000) Appl. Catal. A, 203, 179. Kotrel, S., Kn¨ozinger, H., and Gates, B.C. (2000) Microporous Mesoporous Mater., 35-36, 11. Iyer, P.S., Scherzer, J., and Mester, Z.C. (1988) Perspectives in Molecular Sieve Science, ACS Symposium Series, Vol. 368, American Chemical Society, Washington, DC, pp. 48–65. van de Gender, P. (2005) Albemarle Catalyst Courier, 61, p. 8, http://www.albemarle.com/Products and services/Catalysts/Courier/ Catalysts Courier 61.pdf (accessed in 2009). Chen, N.Y., Mitchell, T.O., Olson, D.H., and Pelrine, B.P. (1977) Ind. Eng. Chem., Prod. Res. Dev., 16, 247. Bolton, A.P. (1971) J. Catal., 22, 9. There are many model studies, see e.g. Corma, A., Mocholi, F., Orchilles, V., Koermer, G.S., and Madon, R.J. (1991) Appl. Catal., 67, 307. Rigby, A.M., Kramer, G.J., and van Santen, R.A. (1997) J. Catal., 170, 1. (a) Effects of selective adsorption phenomena that are not based on shape

47.

48.

49.

50. 51.

52.

53.

54. 55.

56.

57.

selectivity on catalysis are surprisingly poorly documented. See for relevant adsorption studies and references: Papaioannou, C., Petroutsos, G., and Gunßer, W. (1997) Sol. St. Ionics, 101-103, 799; (b) Daems, I., M´ethivier, A., Leﬂaive, P., Fuchs, A.H., Baron, G.V., and Denayer, J.F.M. (2005) J. Am. Chem. Soc., 117, 11600; (c) see for a discussion of selective adsorption in zeolite catalysis: Espeel, P.H.J., Parton, R., Toufar, H., Martens, J.A., H¨olderich, W., and Jacobs, P.A. (1999) Zeolite Effects in Organic Catalysis in Catalysis and Zeolites – Fundamentals and Applications, Chapter 6 (eds J.Weitkamp and L. Puppe), Springer, Berlin, p. 377. Beyerlein, R.A., McVicker, G.B., Yacullo, L.N., and Ziemiak, J.J. (1988) J. Phys. Chem., 92, 1967. Corma, A., Forn´es, V., Mochol´ı, F.A., Mont´on, J.B., and Rey, F. (1991) ACS Symp. Ser., 452, 12. Corma, A., Forn´es, V., Franco, M.J., Mochol´ı, F.A., and P´erez-Pariente, J. (1991) ACS Symp. Ser., 452, 79. Corma, A. and Martinez- Triguero, J. (1994) Appl. Catal. A, 118, 153. Mitchell, M.M. Jr., Hoffman, J.F., and Moore, H.F. (1993) Stud. Surf. Sci. Catal., 76, 293. Dries, H., Muller, F., Willbourne, P., Fum, M., and Williams, C.P. (2005) Hydrocarbon Process., 82 (2), 69. (a) Tan, S. and Satbhai, P. (1998) Design and operation of residue catalytic crackers; paper presented at the Asia Paciﬁc Reﬁning Technology Conference (Singapore 4/20-20/98), Pet. Technol. Q., 3 (4), 47. Philips, G. and Liu, F. (2003) Hydrocarbon Eng., 8 (9), 69. UOP MSCC Process, Brochure, Available on http://www.uop.com UOP 4223–18 (2008). (Accessed in 2008). (a) Czernuszewicz, R.S. (2000) J. Porphyrins Phtalocyanines, 4, 426; (b) Mitchell, P.C.H., Scott, C.E., Bonnelle, J.P., and Grimblot, J.G. (1985) J. Chem. Soc., Faraday Trans., 81, 1047. Habib, E.T. Jr., Owen, H., Snyder, P.W., Streed, C.W., and Venuto, P.B.

579

580

18 Cracking and Hydrocracking

58.

59. 60. 61. 62.

63.

64. 65. 66. 67. 68.

69.

70.

71. 72. 73. 74.

75.

(1977) Ind. Eng. Chem., Prod. Res. Dev., 16, 291. Truhillo, C.A., Uribe, U.N., Knops-Gerrits, P.P., Oviedo, L.A., and Jacobs, P.A. (1997) J. Catal., 168, 1. Biswas, J. and Maxwell, I.E. (1990) Appl. Catal., 58, 1. Wormsbecher, R.F., Peters, A.W., and Maselli, J.M. (1986) J. Catal., 100, 130. Pine, L.A. (1990) J. Catal., 125, 514. Rainer, D.R., Vadovic, C., Rautiainen, E., Nelissen, B., and Imhof, P. (2004) Stud. Surf. Sci. Catal., 149, 165. Yaluris, G., Cheng, W.C., Peters, M., Hunt, M.L.J., and Boock, L.T. (2001)NPRA Annual Meeting Papers 2001, Annual Meeting – National Petrochemical and Reﬁners Association, New Orleans, p. 26. McKay, D.L. and Bertus, B.J. (1979) ACS Div. Pet. Chem. Prepr., 24, 645. Ramamoorthy, P. (1988)NPRA Annual Meeting, March 20-22, San Antonio. Heite, R.S., English, A.R., and Smith, G.A. (1990) Oil Gas J., 88 (23), 81. Nielsen, R.H. and Doolin, P.K. (1993) Stud. Surf. Sci. Catal., 76, 339. Degnan, T.F., Chitnis, G.K., and Schipper, P.H. (2000) Microporous Mesoporous Mater, 35-36, 245. Yanik, S.J., Demmel, E.J., Humphries, A.P., and Campagna, R.J. (1985) Oil Gas J., 83 (19), 108. Chen, N.Y., Degnan, T.F. Jr., and Smith, C.M. (1994) Molecular Transport and Reaction in Zeolites, VCH Publishers, New York. Buchanan, J.S. (2000) Catal. Today, 55, 207. O’Connor, P., Kuehler, C., and Imhof, P. (1983) Hydrocarbon Eng., 8 (1), 35. Zhao, X. and Roberie, T.G. (1999) Ind. Eng. Chem. Res., 38, 3847. Cabral de Menezes, S.M., Lam, Y.L., Damodaran, K., and Pruski, M. (2006) Microporous Mesoporous Mater., 95, 286. Damodaran, K., Wiench, J.W., Cabral de Menezes, S.M., Lam, Y.L., Trebosc, J., Amoureux, J.-P., and Pruski, M. (2006) Microporous Mesoporous Mater., 95, 296.

76. The subject matter is further reviewed

77.

78. 79.

80. 81.

82.

83.

84.

85.

86.

87. 88.

89.

90.

91.

92.

93.

thoroughly in Blasco, T., Corma, A., and Mart´ınez-Triguero, J. (2006) J. Catal., 237, 267. Corma, A., Fornes, V., Kolodziejski, W., and Martinez-Triguero, L.J. (1994) J. Catal., 145, 27. Pine, L.A. (1985)US Patent 4,504,382. Xie, C.X., Zhao, J., Pan, H.F., and Ning, S.-K. (2002) Petrochem. Technol., 31 (9), 691. Huang, L. and Li, Q. (1999) Chem. Lett. (8), 829. Li, Z., Liu, L.S., and Ge, X. (1988)European Patent Application 305,720(A2), to Research Institute of Petroleum, Processing SINOPEC. Letsch, W.S. and Earl, G. (1994)US Patent 5,662,868, to Stone & Webster Engineering Corporation. Shi, Z., Zhang, F., and Liu, S. (1998)European Patent Application 909,582(A1), to Research Institute of Petroleum, Processing SINOPEC. Hulet, C., Briens, C., Berruti, F., and Chan, E.W. (2005) Int. J. Chem. Reactor Eng., 3, 19. Meng, X., Gao, J., Li, L., and Xu, C. (2004) Petr. Sci. Technol., 22 (9-10), 1327. Bowen, C., Santner, C., and Dahria, D. (2008)AIChE Spring Meeting Proceedings, pp. 597–608. Nieskens, M. (2008)NPRA Annual Meeting, March 2008, Paper AM-08-54. Lauhresen, C.J., Clark, P.D., and Du Plessis, M.P. (2006) J. Can. Petr. Technol., 45 (8), 1. Tallman, M.J. and Eng, C.N. (2008)AIChE Spring Meeting Proceedings, pp. 584–596. Soni, D., Angevine, P.J., Rao, R., Saidulu, G., Battacharyya, D. and Krishnan, V. (2008)AIChE Spring Meeting Proceedings, pp. 575–583. Bonetto, L., Camblor, M.A., Corma, A., and P´erez-Pariente, J. (1992) Appl. Catal., 82, 37. Mavrovouniotis, G.M., Cheng, W.C., and Peters, A.W. (1994) ACS Symp. Ser., 571, 16. Corma, A., Mart´ınez-Triguero, J., and Mart´ınez, C. (2001) J. Catal., 197 (1), 151.

References 94. Haas, A., Harding, D.A., and Nee,

95.

96.

97. 98.

99.

100.

101.

102. 103.

104.

105. 106. 107.

J.R.D. (1999) Microporous Mesoporous Mater., 28 (2), 325. (a) Corma, A., Diaz-Caba˜ nas, M.J., Marti’nez-Triguero, J., and Rey, F. (2002) J. Rius Nat., 418 (6897) 514; (b) Corma, A., Diaz-Caba˜ nas, M.J., Marti’nez-Triguero, J., and Rey, F. (2002)European Patent Application 1,445,297 A1. Corma, A., D´ıaz-Caba˜ nas, M.J., Jord´a, J.L., Mart´ınez, C., and Moliner, M. (2006) Nature, 443, 842. Corma, A. and Martinez-Triguero, J. (1997) Appl. Catal. A, 165, 102. Casta˜ neda, R., Corma, A., Fornes, V., Martinez-Triguero, J., and Valencia, S. (2006) J. Catal., 238, 79. Scherzer, J. and Gruia, A.J. (1996) Hydrocracking Science and Technology, Marcel Dekker, New York, 305 pp. Gary, J.H. and Handwerk, G.E. (2001) Petroleum Reﬁning: Technology and Economics, 4th edn, Marcel Dekker, New York, 441 pp. Kalnes, T., Fleming, P., Wang, L., Thakkar, V., and Antos, G. (2001)NPRA Paper AM-01-30. Law, D.V. (2000/2001) Pet. Technol. Q, Winter, 55. Hagan, A.P., Ioannou, M.G., Cackett, S.J., and Shivaram, A. (1999)PetroTech-99, Paper presented at the 3rd International Petroleum Conference and Exhibition, January 9–12, 1999, New Delhi. Hennico, A., Billon, A., and Bigeard, P.-H. (1992) Hydrocarbon Technol. Int., 19. Chongren, H. and Xiangchen, F. (2002) Pet. Technol. Q., Autumn, 93. Sullivan, R.F. and Scott, J.W. (1983) Am. Chem. Soc. Symp. Ser., 222, 293. Esener, A.A. and Maxwell, I.E. (1989) Improved Hydrocracking Performance by Combining Conventional Hydrothreating and Zeolitic Catalysts in Stacked Bed Reactors in Advances in Hydrotreating Catalysts, Annual AIChE meeting, November 27 – December 2, 1988, Washington DC(eds M.L.Occelli and R.G. Anthony), Elsevier, Amsterdam, p. 263.

108. Mukherjee, U., Mayer, J., and

109.

110. 111. 112.

113.

114. 115. 116. 117.

118. 119. 120.

121. 122. 123. 124.

125. 126.

127.

Srinivasan, B. (2005) Pet. Technol. Q. Catal., 10 (2), 49. Gosselink, J.W., Van De Paverd, A., and Stork, W.H.J. (1989) Stud. Surf. Sci. Catal., 53, 385. Dufresne, P., Bigeard, P.-H., and Billon, A. (1987) Catal. Today, 1, 367. Hunter, M.G., Pappal, D.A., and Pesek, C.L. (1994)NPRA Paper AM-94-21. Sarrazin, P., Bonnardot, J., Wambergue, S., Morel, F., and Gueret, C. (2005) Hydrocarbon Process., 85 (2), 57. Brunn, L.W., Karapakos, N.E., Plesko, R.W., and Schoenroth, B.E. (1984) Oil Gas J., 82 (13), 90. Frazer, S. and Shirley, W. (1999) Pet. Technol. Q., Autumn, 25. Ladeur, P. and Bijwaard, H. (1993) Oil Gas J., 91 (17), 64. Minderhoud, J.K. and van Veen, J.A.R. (1993) Fuels Process. Technol., 35, 87. Law, D.V. (1993)Paper presented at the 7th Reﬁnery Technology Meeting, Bombay, December 6-8, 1993. Mignard, S. and B´eroudiaux, O. (1998) React. Kinet. Catal. Lett., 65, 185. Boduszynski, M.M. (1988) Energy Fuels, 2 (6), 597. (a) Gruia, A.J. (1987)US Patent 4 698 146, assigned to UOP Inc.; (b) Gruia, A.J. (1990) US Patent 4 921 595, assigned to UOP Inc; (c) Gruia, A.J. (1990)US Patent 4 954 242, assigned to UOP Inc. Abdul Latif, N. (1989) Stud. Surf. Sci. Catal., 53, 349. Sie, S.T. (1994) Stud. Surf. Sci. Catal., 85, 587. Nat, P.J. (1989) Erd¨ol Kohle Erdgas Petrochem., 42, 447. Dufresne, P., Quesada, A., and Mignard, S. (1989) Stud. Surf. Sci. Catal., 53, 301. Ward, J.W. (1993) Fuels Process. Technol., 35, 55. Hoek, A., Huizinga, T., Esener, A.A., Maxwell, I.E., Stork, W.H.J., van de Meerakker, F.J., and Sy, O. (1991) Oil Gas J., 89 (16), 77. Santana, R.C., Do, P.T., Santikunaporn, M., Alvarez, W.E., Taylor, J.D.,

581

582

18 Cracking and Hydrocracking

128. 129. 130. 131.

132. 133.

134.

135. 136. 137. 138.

139. 140.

141.

142.

143.

Sughrue, E.L., and Resasco, D.E. (2006) Fuel, 85, 643. Du, H., Fairbridge, C., Yang, H., and Ring, Z. (2005) Appl. Catal. A, 294, 1. Prins, R. (2001) Adv. Catal., 46, 299. Douwes, C.T. and T’Hart, M. (1968) Erd¨ol Kohle Erdgas Petrochem., 21, 202. (a) Roussel, M., Lemberton, J.L., Guisnet, M., Cseri, T., and Benazzi, E. (2003) J. Catal., 218, 427; (b) Roussel, M., Norsic, S., Lemberton, J.L., Guisnet, M., Cseri, T., and Benazzi, E. (2005) Appl. Catal. A, 279, 53. Ward, J.W. (1983) Stud. Surf. Sci. Catal., 16, 587. Minderhoud, J.K., van Veen, J.A.R., and Hagan, A.P. (1999) Stud. Surf. Sci. Catal., 127, 3. Gates, B.C., Katzer, J.R., and Schuit, G.C.A. (1979) Chemistry of Catalytic Processes, McGraw-Hill, New York, 464. Schulz, H. and Weitkamp, J. (1972) Ind. Eng. Prod. Res. Dev., 11, 46. Weitkamp, J. (1978) Erd¨ol Kohle Erdgas Petrochem., 31, 13. Weitkamp, J. and Dauns, H. (1987) Erd¨ol Kohle Erdgas Petrochem., 40, 111. (a) Martens, J.A., Jacobs, P.A., and Weitkamp, J. (1986) Appl. Catal., 20 239; (b) Martens, J.A., Jacobs, P.A., and Weitkamp, J. (1986) Appl. Catal., 20, 283; (c) Jacobs, P.A. and Martens, J.A. (1991) Stud. Surf. Sci. Catal. 58, 445. Jacobs, P.A. and Martens, J.A. (2001) Stud. Surf. Sci. Catal., 137, 633. Guisnet, M., Alvarez, F., Giannetto, G., and Perot, G. (1987) Catal. Today, 1, 415. (a) Denayer, J.F., Baron, G.V., Souverijns, W., Martens, J.A., and Jacobs, P.A. (1997) Ind. Eng. Chem. Res., 36, 3242; (b) Martens, G.G., Marin, G.B., Martens, J.A., Jacobs, P.A., and Baron, G.V. (2000) J. Catal., 195, 253. (a) Martens, G.G., Thybaut, J.W., and Marin, G.B. (2001) Ind. Eng. Chem. Res., 40, 1832; (b) Berger, R.J. (2001) CatTech, 5, 30, esp. Intermezzo 6. Sullivan, R.F., Boduszinski, M.M., and Fetzer, J.C. (1981) Energy Fuels, 3, 603.

144. Guisnet, M. and Magnoux, P. (2001)

Appl. Catal. A, 212, 83. 145. Welters, W.J.J. (1994) Ze-

146.

147.

148.

149.

150. 151. 152. 153. 154.

155.

156.

157.

olite Supported Metal Sulﬁdes as Catalysts for Hydrocracking, Ph.D. Thesis, Eindhoven. (a) Lemberton, J.L., Touzeyido, M., and Guisnet, M. (1989) Appl. Catal., 54, 101; (b) Lemberton, J.L.J.L., Touzeyido, M.M., and Guisnet, M.M. (1991) Appl. Catal. A, 79, 115. Sato, K., Iwata, Y., Miki, Y., and Shimada, H. (1999) J. Catal., 186, 45. Habib, M.M., Bezman, R.D., Dahlberg, A.J., and Krishna, A.S. (2000)Paper presented at The XIth Reﬁnery Technology Meeting, February 9-11, 2000, Hyderabad. Weitkamp, J. and Ernst, S. (1990) Factors Inﬂuencing the Selectivity of Hydrocracking in Zeolites in Guidelines for Mastering the Properties of Molecular Sieves (eds E.G. Derouane, D. Barthomenf, and W., H¨olderich), Plenum Press, New York, p. 343. Weisz, P.B. (1962) Adv. Catal., 13, 137. Cornet, D., El Qotbi, M., and Leglise, J. (1997) Stud. Surf. Sci. Catal., 106, 147. Sie, S.T. (1993) Ind. Eng. Chem. Res., 32, 403. Girgis, M.J. and Tsao, Y.P. (1996) Ind. Eng. Chem. Res., 35, 386. Thybaut, J.W., Laxmi Narasimhan, C.S., Denayer, J.F., Baron, G.V., Jacobs, P.A., Martens, J.A., and Marin, G.B. (2005) Ind. Eng. Chem. Res., 44, 5159. (a) Benazzi, E., Leite, L., Marchal-George, N., Toulhoat, H., Raybaud, P. (2003) J. Catal., 217, 376; (b) Toulhoat, H., Raybaud, P., and Benazzi, E. (2004) J. Catal., 221, 500. Ward, J.W. (1984) Molecular Sieve Catalysts in Applied Industrial Catalysis, vol. 3 (ed. B.E. Leach), Academic Press, Orlando, p. 272. (a) Reyes, S.C., Sinfelt, J.H., and DeMartin, G.J. (2000) J. Phys. Chem. B, 104, 5750; (b) Isoda, T., Maemoto, S., Kusakabe, K., and Morooka, S. (1999) Energy Fuels, 13, 617 (c) Kuehne, M.A., Babitz, S.M., Kung, H.H., and Miller, J.T. (1998) Appl. Catal. A, 166, 293.

References 158. Yan, T. (1983) Ind. Eng. Chem. Proc. 159.

160. 161. 162. 163.

164.

165. 166.

167.

168.

169.

170.

171.

172.

Res. Dev., 22, 154. Huizinga, T., Theunissen, J.M.H., Minderhoud, J.K., and van Veen, J.A.R. (1995) Oil Gas J., 93 (26), 40. Maxwell, I.E. (1987) Catal. Today, 1, 385. Scherzer, J. (1984) ACS Symp. Ser., 284, 157. Szostak, R. (2001) Stud. Surf. Sci. Catal., 137, 261. (a) Beaumont, R. and Barthomeuf, D. (1972) J. Catal., 26, 218 (b) Beaumont, R. and Barthomeuf, D. (1972) J. Catal., 27, 45; (c) Beaumont, R. and Barthomeuf, D. (1973) J. Catal., 30, 288. Wachter, W.A. (1990) The Role of Next Nearest Neighbours in Zeolite Acidity and Activity in Theoretical Aspects of Heterogeneous Catalysis (ed. J.B. Moffat), Van Nostrand Reinhold, New York, p. 10. Bezman, R. (1992) Catal. Today, 13, 143. van Bokhoven, J.A., Williams, B.A., Ji, W., Koningsberger, D.C., Kung, H.H., and Miller, J.T. (2004) J. Catal., 224, 50, references cited therein. Hensen, E.J.M., Poduval, D.G., Ligthart, D.A.J.M., van Veen, J.A.R., and Rigutto, M.S. J. Phys. Chem. C, submitted. Mostad, H.B., St¨ocker, M., Karlsson, A., and Rørvik, T. (1996) Appl. Catal. A, 144, 305. (a) Chu, C.T.W. and Chang, C.D. (1985) J. Phys. Chem., 89, 1569; (b) Post, M.F.M., Huizinga, T., Emeis, C.A., Nanne, J.M., and Stork, W.H.J. (1989) Stud. Surf. Sci. Catal., 46, 365. Maesen, Th.L.M., van Veen, J.A.R., Cooper, D.A., van Vegchel, I.M., and Gosselink, J.W. (2000) J. Phys. Chem. B, 104, 716. (a) Sato, K., Nishimura, Y., and Shimada, H. (1999) Catal. Lett., 60, 83; (b) Sato, K., Nishimura, Y., Honna, K., Matsubayashi, N., and Shimada, H. (2001) J. Catal., 200, 288. Cooper, D.A., Hastings, T.W., and Hertzenberg, E.P. (1997)US Patent 5 601 798, assigned to ZI.

173. Janssen, A.H., Koster, A.J., and de

174. 175. 176.

177.

178. 179. 180. 181.

182. 183. 184.

185. 186.

187.

188. 189.

Jong, K.P. (2001) Angew. Chem. Int. Ed., 40, 1102. Krishna, R. (1989) Erd¨ol Kohle Erdgas Petrochem., 42, 194. Geus, J.W. and van Veen, J.A.R. (1999) Stud. Surf. Sci. Catal., 123, 459. Vissenberg, M.J., Joosten, L.J.M., Heffels, M.M.E.H., van Welsenes, A.J., de Beer, V.H.J., van Santen, R.A., and van Veen, J.A.R. (2000) J. Phys. Chem. B, 104, 8456. (a) Robinson, W.R.A.M., van Veen, J.A.R., de Beer, V.H.J., and van Santen, R.A. (1999) Fuel Proc. Technol., 61, 103; (b) Hensen, E.J.M., Kooyman, P.J., van der Meer, Y., van der Kraan, A.M., de Beer, V.H.J., van Veen, J.A.R., and van Santen, R.A. (2001) J. Catal., 199, 224. Hensen, E.J.M. and van Veen, J.A.R. (2003) Catal. Today, 86, 87. Zhang, W. and Smirniotis, P.G. (1999) J. Catal., 182, 400. Ward, J.W. (1993) Fuel Proc. Technol., 35, 55. Barman, B.N., Cebolla, V.L., Mehrotra, A.K., and Mansﬁeld, C.T. (2001) Anal. Chem., 73, 2791. Martens, G.G. and Marin, G.B. (2001) AIChE J., 47, 1607. Laxmi Narasimhan, C.S. (2004)Ph.D. Thesis, Universiteit Gent, Chapter 8. Fu, J., Kim, S., Rodgers, R.P., Hendrickson, C.L., and Marshall, A.G. (2006) Energy Fuels, 20, 661. Sch¨uth, F. (2001) Stud. Surf. Sci. Catal., 135, 1. (a) Lin, Y., Zhang, W., and Pinnavaia, T.J. (2000) J. Am. Chem. Soc., 122, 8791; (b) Lin, Y., Zhang, W., and Pinnavaia, T.J. (2001) Angew. Chem., 113, 1295. (a) Zhang, Z., Han, Y., Xiao, F.-S., Qiu, S., Zhu, L., Wang, R., Yu, Y., Zou, B., Wang, Y., Sun, H., Zhao, D., and Wei, Y. (2001) J. Am. Chem. Soc., 123, 5014; (b) Zhang, Z., Han, Y., Zhu, L., Wang, R., Yu, Y., Qiu, S., Zhao, D., and Xiao, F.-S. (2001) Angew. Chem., 113, 1298. Fang, Y. and Hu, H. (2006) J. Am. Chem. Soc., 128, 10636. Bagshaw, S.A. and Hayman, A.R. (2001) Adv. Mater., 13, 1011.

583

584

18 Cracking and Hydrocracking 190. Martens, J.A., Benazzi, E., Brendl´e, J.,

Lacombe, S., and LeDred, R. (2000) Stud. Surf. Sci. Catal., 130, 293. 191. (a) Chrones, J. and Germain, R.R. (1989) Fuel Sci. Technol. Int., 7, 783; (b) Eccles, R.M. (1993) Fuel Proc. Technol., 35, 21. 192. (a) Marcus, D.M., McLachlan, K.A., Wildman, M.A., Ehresmann, J.O., Kletnieks, P.W., and Haw, J.F. (2006) Angew. Chem. Int. Ed., 45 (19), 3133; (b) Haw, J.F., Song, W., Marcus, D.M.,

and Nicholas, J.B. (2003) Acc. Chem. Res., 36, 317. 193. Rigutto, M.S. (2008) Hydrocarbon Conversion with Zeolites: a Clair-obscur in Zeolites and Related Materials: Trends, Targets and Challenges, Studies in Surface Science and Catalysis, Elsevier, Amsterdam, Vol. 174(B), Proceeding of the 4th International FEZA Conference, 2–6 September 2008, Paris (eds A. G´ed´eon, F. Massiani, and F. Babonneau), p. 43 .

585

19 Naphtha Reforming and Upgrading of Diesel Fractions Carlo Perego, Vincenzo Calemma, and Paolo Pollesel

19.1 Introduction

Gasoline and diesel are by far the most important and valuable fractions produced by petroleum reﬁneries. In the last few years, the demand growth rates have changed for gasoline and diesel, with a global increase for diesel two times higher than for gasoline distillate fractions (Figure 19.1) [1]. Nevertheless, global petroleum product demand has grown since 1995 and it is expected to continue to grow till 2015 (Figure 19.2). Diesel and gasoline, despite their different growth rates, will remain the most important petroleum products [2]. World demand for gasoline is still forecast to grow 1.8% per year through 2010 and about 1.3% per year to 2015 [3]. Catalytic reforming, therefore, although one of the oldest of reﬁnery processes, continues to be a major process for producing high-octane gasoline, hydrogen, and mixture of benzene, toluene, and xylene (BTXs). Worldwide reﬁning capacity corresponds to 13% of crude oil distillation capacity. Reformers are located in almost 70% of the 662 reﬁneries in the World [4]. Ninety-six of the reﬁneries with reformers are located in North America, 82 are in Western Europe, and 102 are in the Asia-Paciﬁc region. The geographical distribution of reforming capacity is summarized in Table 19.1. Figure 19.3 represents a simpliﬁed scheme of a modern reﬁnery. The upper part of the drawing contains the upgrading processes for the naphtha fractions, which produces the streams composing the gasoline pool. Catalytic reforming section is one of the most important sections. The gasoline pool is obtained by blending gasoline from various origins: straight run, reforming, ﬂuid catalytic cracking (FCC), coking, hydrocracking, isomerization, alkylation, polymerization, and additives (butane, ethers), and the reformate gasoline usually accounts for about 34% in the United States and 40% in Western Europe of the whole gasoline pool [5]. Global demand growth rate for BTXs will be higher than for gasoline, ranging from 2.4% per year for benzene to 5.6% per year for xylenes [3]. Reported market trends together with regulatory issues (e.g., limits to benzene content in gasoline) require technological innovations both for the process and for the reforming catalyst. Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

586

19 Naphtha Reforming and Upgrading of Diesel Fractions

Thousands of barrels per day

4000 3500 3000 2500 2000 1500 1000 500 0 Gasoline

Figure 19.1

Diesel

Kero/Jet fuel

Incremental transportation fuel demand 2008–2015.

Diesel fuel is the reﬁnery product with the highest market demand growth rate. Most of the recent investments in reﬁneries have covered processes for production or upgrading of fractions to make them suitable for diesel blendstock. In Figure 19.3, it can be seen that diesel pool is mainly fed by straight-run diesel and by diesel fractions coming from FCC, hydrocracking, and coking units. The maximization and upgrading of these diesel fractions are strongly driving process 90 Residual fuel oil

Diesel

Jet/kerosene

Gasoline

Naphtha

80

Million of barrels per day

70 60 50 40 30 20 10 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Figure 19.2

Global demand for reﬁned products.

19.2 Catalytic Reforming Table 19.1

Reforming capacity by geographic region. Capacity (thousands bbl/day)

North America Western Europe Central and Eastern Europe Asia-Paciﬁc South and Central America Former Soviet Union Middle East Africa Total

Capacity (% of world capacity)

4 174 2 058 359

36.9 18.2 3.2

1 974 427

17.4 3.8

1 135 718 476 11 319

10.0 6.3 4.2 100

innovation and catalyst development. Main processes involving diesel upgrading are hydrotreating processes aimed at desulfurization and dearomatization of gasoil fractions, which are treated in speciﬁc chapters of this book. In the present chapter, we discuss mainly about catalytic dewaxing of middle distillates, a process that involves more zeolite-based catalysis than other upgrading reactions. Among the speciﬁcations of diesel fuel, low-temperature properties refer to the ability to operate under cold weather conditions. On cooling, high-melting-point compounds tend to precipitate, thereby affecting the ﬂow characteristics of diesel fuel, and eventually this may lead to its solidiﬁcation. Catalytic dewaxing is a process to remove the waxy component (mainly n-parafﬁns and slightly branched parafﬁns) to improve the cold-ﬂow properties of gasoil.

19.2 Catalytic Reforming 19.2.1 Process

Catalytic reforming is one of the oldest of reﬁnery processes, commercialized in 1940. In a modern reﬁnery, catalytic reforming has two main roles: produce high-octane gasoline blendstock and produce hydrogen. As shown in Figure 19.3, catalytic reforming converts low-octane straight-run, cracked and hydrocracked naphthas boiling in the range 65–205 ◦ C into high-octane aromatic reformate. Naphthenes and, to a lesser extent parafﬁns, are converted largely to aromatics. Reforming is also a major source of highly aromatic fractions for the extraction of BTXs for petrochemical feedstocks. The aromatics yield

587

588

19 Naphtha Reforming and Upgrading of Diesel Fractions Gases H2

Isomerization unit

Isomerate

Light straight-run naphtha H2 Naphtha splitter

Reformate offgas

Naphtha hydrotreating

Catalytic reforming unit

Heavy straight-run naphtha

i -Butane

Kerosene/ jet fuel

FCC C1/C4 FCC offgas

Light gas oil FCC gasoline

Vacuum residue

Diesel pool Hydrocracking unit

Light cycle oil Slurry oil

Heavy coker gas oil Light coker gas oil Coking unit

Vacuum distilation unit

Vacuum gas oil

Fluid catalytic cracking unit

Atmospheric crude distilation unit

Diesel hydrotreating

Alkylate

Hydrocracked naphtha

Kerosene Kerosene hydrotreating

H2

Gasoline pool

Butanes

Alkylation unit

H2

Crude oil

H2

H2 unit

Hydrocracked distilate

Gas oil hydrotreating Cutter stock

H2 unit Residual fuel oil

H2 unit

Residue hydrotreating

Figure 19.3

Coke

Schematic of a modern reﬁnery.

depends upon the feedstock composition, catalyst, and reactor severity. Catalytic reforming also accounts for about 60% of BTX production worldwide. In the US, catalytic reforming sources represent about 69% of BTX and 70% of benzene, including associated toluene conversion processes [6]. About 70% of world’s toluene supply and about 88% of US production and about 88% of US production originate from catalytic reforming. Catalytic reforming accounts for 80% of the mixed xylenes produced both worldwide and in the United States. The most common method of

19.2 Catalytic Reforming

589

benzene recovery and puriﬁcation is aromatic extraction. Reformate from catalytic reforming of naphtha, typically composed of 65–82 wt% aromatics and 18–35 wt% nonaromatics, consists essentially of C5–C7 parafﬁns, mostly i-parafﬁns. The ﬁrst reforming processes were nonregenerative reforming. These units use a molybdenum oxide catalyst which is unloaded and replaced by fresh catalyst when deactivated. These processes later evolved to semiregenerative reforming (SR), characterized by the periodic shutdown of the unit for in situ catalyst regeneration. SR processes are usually based on a ﬁxed-bed reactor technology. Cyclic regeneration reformers also have ﬁxed catalyst beds, but since a swing reactor is provided, the reactors are individually regenerated at short periods compared to SR units. Cyclic regeneration reforming is characterized by continual in situ regeneration of one of the several reactors by isolation of that reactor for regeneration while the other reactors are in service. The ﬁnal evolution is represented by continuous catalyst regeneration (CCR) processes where a platinum-containing catalyst is continually withdrawn from the last reactor bed, regenerated, and returned to the ﬁrst reactor bed, without plant shutoff. Most new plants are CCR units, whereas the most common type of reformer is still SR. Catalytic reforming currently accounts for about 35–40% of gasoline production and, most signiﬁcantly, for a substantial amount of reﬁnery hydrogen production required for hydrotreating and hydrocracking units. In regions dominated by less complex reﬁneries, reformate comprises an even larger percentage of gasoline production. Figure 19.4 reports a UOP PlatformingTM CCR process scheme [7]. Typical operating conditions in SR, cyclic, and CCR reforming processes are summarized in Table 19.2. Historically, naphtha reforming performances increased over time with both catalysts and process innovations, increasing the yields of reformate and hydrogen while raising the octane rating (see, for example, Figure 19.5 [3]). Leading gasoline speciﬁcations have put constraints on several octane primary sources, including reformate, due to benzene and total aromatics requirement [8]. CCR Stacked reactor Naphtha feed from treating

CCR regenerator

Net gas compressor

Combined feed exchanger

Net H2 rich gas Recovery section

Separator Light ends Stabilizer Regenerated catalyst

Fired heaters Spent catalyst Reformate

Figure 19.4

Schematic of UOP CCR of PlatformingTM process.

19 Naphtha Reforming and Upgrading of Diesel Fractions Table 19.2

Operating conditions in catalytic reforming.

H2 /HC (mol/mol) Pressure (bar) Temperature (◦ C) Catalyst life

SR

Cyclic

CCR

10 15–35 470–510 0.5–1.0 years

4–8 7–15 470–510 Days to weeks

4–8 3–4 495–535 Days to weeks

European gasoline speciﬁcations limit the benzene content in gasoline to maximum 1 vol% and total aromatics to 35%. In the United States, the limits are 1 vol% for benzene (0.7% in California) and 20% for aromatics. Reforming processes tailored for gasoline production can use speciﬁc arrangements in order to comply with these limits. Dehexanising the feed reduces the reformate benzene content to below the 1% gasoline speciﬁcation. Reformate benzene can also be reduced by hydrogenation to saturate the benzene downstream of the reformate splitter. For example, the Axens Benfree process integrates hydrogenation with the reformate splitting in a reactive distillation column. Unlike conventional reactive distillation that holds the catalyst in structured packing on distillation trays, the catalyst is held in a side ﬁxed-bed reactor (Figure 19.6 [9]). Full-range reformate from either a semiregenerative or CCR reformer is split with C6 and lower boiling components, including benzene (but not toluene) taken overhead. A benzene-rich light fraction is withdrawn just above the feed tray and pumped to the top of the hydrogenation reactor, which operates at a higher pressure than the column for increased hydrogen solubility. The low hydrogen ﬂow minimizes gasoline loss to the off gas. The essentially benzene-free reactor efﬂuent returns to the splitter column. The catalyst is a standard non-noble-metal hydrogenation catalyst. 94 C5 + yield - vol% on feed

590

Theoretical

90 1960s (Semiregeneration) 300 psig

86 82 78

1960s technology 500 psig

1970s (CCRSM) 125 psig

1990s (CCRSM) 50 psig

74 86

Figure 19.5

90

94 98 Research octane number

Evolution of PlatformingTM units.

102

106

19.2 Catalytic Reforming Light gas

Splitter C5–C6 C5–C6 reformate H2

Light reformate

Heavy reformate

Figure 19.6

Axens BenfreeTM process.

The boiling range of the feedstock is a signiﬁcant factor to drive the process toward preferred products. The feed may be straight-run, cracked, or hydrocracked naphtha having a boiling range as wide as 65–205 ◦ C. End points higher than about 200 ◦ C contain desirable long-chain parafﬁns (C10 + ) that are hydrocracked into the gasoline range, which in turn can both isomerize to branched parafﬁns and dehydrocyclize to aromatics. However, above 205 ◦ C polycyclic aromatics formation occurs, which can lead to catalyst coking under reforming conditions [10]. When gasoline blendstock is the reforming objective, the trend is to further cut the light end to avoid benzene precursors and charge the C7–C9 fraction (90–160 ◦ C). When BTX for petrochemical feedstock is the main reforming objective, a narrow naphtha cut boiling in the range of 71–104 ◦ C, which is rich in benzene precursors (i.e., 6-carbon atom molecules), is the feed. For hydrogen production, the optimal cut is a C6–C9 fraction (60–160 ◦ C) that contains the highest naphthenes concentration. 19.2.2 Reforming Chemistry

In the presence of hydrogen, catalytic reforming converts low-octane naphtha feedstock into aromatic, high-octane gasoline blending stock or into BTXs for petrochemical feedstock. Hydrogen is an important by-product. Ideally, no other gas or coke is produced. Reforming involves several main reactions that convert naphthenes and parafﬁns into aromatics and several side reactions. Coking is the main cause of catalyst deactivation. The catalyst is regenerated by controlled burning of the coke followed by rejuvenation. The catalytic reformer feedstocks are saturated hydrocarbons, normally straight-run naphtha from the atmospheric distillation unit. Naphtha from an FCC

591

19 Naphtha Reforming and Upgrading of Diesel Fractions

or from a hydrocracker is also reformed. Some naphtha from delayed coking or other units may sometimes be reformed. Regardless of source, naphtha ﬁrst must be hydrotreated to saturate oleﬁns that are coke precursors and to remove sulfur and nitrogen compounds along with metals that poison the reforming catalyst. Reformer conversion and selectivity are functions of the feedstock composition, catalyst, and operating conditions that determine the thermodynamics and kinetics of the reaction network. Reforming catalysts are solids containing both metal and acid active sites. Figure 19.7 illustrates a simpliﬁed, generalized reforming reaction network that also identiﬁes the catalytic active sites involved [11]. Speciﬁcally, the reaction network for C6 hydrocarbons is shown in Figure 19.8 [12]. The main desirable reactions include the following: 1) Dehydrocyclization of parafﬁns to aromatics and hydrogen (endothermic, slow); 2) Isomerization of alkylcyclopentanes to cyclohexanes (quite endothermic, rapid); C1–C3 fragments

Isohexanes

Metalic function

592

Isohexanes

Coke precursors A

( )

Coke

Coke Acidic function

Figure 19.7

Coke precursors B

Reforming reaction network.

( )

19.2 Catalytic Reforming

593

n-Paraffins M/A

M or A

A Cracked products

A

Cyclopentanes

Naphthene isomerization

Isoparaffins II

Dehydrogenation

A = acid M = Metal I = Hydrocracking and demethylation (M) II = Paraffin isomaerization III = Dehydrocyclization

Reaction network for C6 hydrocarbons.

3) Dehydrogenation of cyclohexanes to aromatics and hydrogen (quite endothermic, very fast); 4) Isomerization of linear parafﬁns into isoparafﬁns (mildly exothermic, rapid). The most prominent side reactions include the following: 1) 2) 3) 4)

Lighter aromatics

Dealkylation and demethylation

III Predominant active sites

Figure 19.8

Aromatics

M/A

M or A

I

M/A

M Cyclohexanes

Hydrocracking of naphthenes and parafﬁns (quite exothermic, slowest); Hydrodealkylation of aromatics; Alkylation and transalkylation of aromatics; Coke formation.

All desirable reactions lead to an increase in the octane number of the obtained products (Figure 19.9). Rates of individual reactions depend upon the catalyst but they are generally fast, virtually reaching equilibrium. According to the classical bifunctional mechanism, the metal function catalyzes hydrogenation and dehydrogenation reactions, while the acid function in combination with the metal sites catalyzes skeletal isomerization, dehydrocyclization, and cyclization reactions. The reactions of skeletal isomerization and cyclization of n-alkanes start on a metal site and are completed on an acid site. The mechanisms of reforming over platinum on alumina catalysts have been studied since the 1950s. Mechanisms evolved from the classical bifunctional acid–metal sites mechanism that attributed all dehydrogenation reactions to the metal sites and all skeletal rearrangements to carbonium ion rearrangements on the acid sites. The metal site reactions were considered to be the fast reactions; the acid site reactions the slow, rate-determining steps. This mechanism led ﬁrst to better understanding and process optimization and later on to several catalyst and process innovations [13]. So far, the chemistry of reforming has not resolved into a

594

19 Naphtha Reforming and Upgrading of Diesel Fractions

∆H°298 (kJ mol−1) Isomerization

C C C C C C C

C C C C C C 73 C

−4

26 Cyclization

C + H2

C C C C C C C

+33

50

Aromatization

40 Rearrangement/ combination

+205

C + 3 H2

C

100

C C

C + 3 H2

C

+177

Octane number Figure 19.9

Octane increase with reforming reactions.

set of widely recognized reaction mechanisms that can explain experimental results as a function of several parameters and lacks well-established interpretations. Adding to the complexity of catalytic reforming reactions, different active sites may be required for the various skeletal reactions. When more than one chemical bond forms between an intermediate and the surface, active centers consisting of more than one atom are required. Atomization requires active centers with up to three noble-metal atoms. Active sites for C5 cyclic isomerization consist of more surface metal atoms than those for bond-shift isomerization. Furthermore, not only the atoms of the active centers but also the adjacent atoms may be important for reforming reactions. The environment of the active metal, especially of single-atom sites, can modify its catalytic properties [11]. Reforming reaction is performed in the presence of hydrogen. Hydrogen may affect the fresh catalyst (reduction) by inducing surface reconstruction, including sintering, particle migration, growth or coalescence, and particle shape changes. The presence of substantial hydrogen prevents the rapid coke formation and deactivation of the catalyst. The optimal hydrogen partial pressure is generally different for the different reactions. Maximum rates appear at higher hydrogen pressures with increasing temperature. Higher hydrogen favors skeletal isomerization as opposed to C5 cycles from their common intermediate. Lower hydrogen pressure favors hydrogenolysis. Two types of surface hydrogen species may be assumed in explaining the different hydrogen responses: one is a deeply dissociated precursor for benzene and coke; the other one is a less dissociated intermediate for saturated products produced via C5 cyclic intermediates.

19.2 Catalytic Reforming

19.2.3 Catalyst

The reforming process, developed during World War II, has been improved immensely through the development of new catalysts. The catalysts typically are spheres or extrudates of 1 mm or more in size. Reforming catalysts progressed from monometallic to bimetallic to trimetallic and even multimetallic catalysts. Reforming catalyst is one of the most complicated catalysts. New reforming catalysts are required to produce reformate not only with a more appropriate composition for gasoline speciﬁcations requiring less benzene and total aromatics, but also to reduce energy consumption through reforming at lower temperatures and hydrogen pressure. Reforming catalysts contain both catalytic metal and acid components. Pt is the best and most common metal component, often promoted by Re, Ir, or Sn. Chlorinated alumina is the most common acid component. Zeolites are also used as the acidic support [14]. The activity of Pt for desirable isomerization and cyclization is greater than obtained with other transition metals. Pt also is more active for the undesirable hydrogenolysis (C–C bond cleavage) reaction. The catalytic properties of Pt can be improved by the addition of one or more metals. Rhenium and tin are the most common promoter metals. Re and Cu can catalyze aromatization to some extent. C5 cyclization and isomerization of 3-methylpentane is catalyzed by only Pt, Pd, Ir, and Rh. Reforming catalysts are commonly classiﬁed as monometallic, bimetallic, trimetallic, or multimetallic on the basis of the number of active metals they contain. Presulﬁding is required to reduce highly exothermic hydrogenolysis and the associated temperature spike may damage the reactor and catalyst. Presulﬁding is conducted with at about 370 ◦ C to about 0.025 wt% sulfur on the catalyst [15]. A balance between the metal function and the acid function of the catalyst must be maintained during the life of the catalyst. Too strong a metal function promotes hydrogenolysis to C1–C4 gases and dehydrogenation to polyoleﬁnic coke precursors. Too weak a metal function causes the catalyst to deactivate rapidly by coking. Excessive hydrocracking results from too strong an acid function. With too weak an acid function, the rate-determining dehydroisomerization and dehydrocyclization reactions are too slow, causing increased light gases and decreased liquid reformate yield. Under actual reforming conditions, moisture in the feedstock and chlorine in platinum-alumina catalysts are the main controls of the acidity. Chlorine strengthens the acidity of the alumina. Water enhances the Broensted acidity of alumina but also strips off chlorine, particularly during regeneration. Commercially chlorine is restored by the addition of organic chloride, at the start of cycle or continuously [13]. Monometallic catalysts historically were the ﬁrst commercial catalysts. Molybdenum oxide on alumina was the catalyst used in the ﬁrst catalytic reforming process. The process was jointly developed in 1939 by Standard Oil of New Jersey (now ExxonMobil), Standard Oil of Indiana (now BP-Amoco), and M.W. Kellogg Company. In 1949, UOP introduced the Platforming process, which initially used

595

19 Naphtha Reforming and Upgrading of Diesel Fractions

C5 + yield (LV%)

6 5 4 Pt–Re 3 2 Pt

1 0 0

20

40

60

80

100

40 Reaction temperature

596

30 Pt

Pt–Re

20 10 0 0

Figure 19.10

20

40 60 80 Catalyst age (days on stream)

100

Pt–Re catalyst performance compared to monometallic Pt catalyst.

platinum on alumina catalyst. Pt (0.3–0.35 wt%) on alumina is more sulfur tolerant (<5 ppm allowable in feed) than common bimetallic catalysts but requires higher pressure and a higher hydrogen/feed ratio. Monometallic catalysts were replaced in the 1970s by bimetallic catalysts of rhenium, iridium, or tin, which had greatly improved stability (from 2–4 times for Pt–Re) and selectivity [3]. Bimetallic catalysts produce higher yields for longer catalyst life than monometallic Pt catalysts. Figure 19.10 shows the reformate yield and catalyst life improvements of a Pt–Re bimetallic catalyst [16]. Bimetallic reforming catalysts contain a promoter to modify the activity of Pt or to supplement Pt. The main advantages of bimetallic catalysts over monometallic catalysts are improved selectivity and stability. A wide range of element combinations for the metal function make active reforming catalysts. Pt promoters include Re, Sn, Ge, and Ir. The accessibility to the reactants and the reactivity of the Pt are modiﬁed by the additives. The effects of these promoters on the metal function and on the acid function are given below: • The deep dehydrogenation activity of Pt is decreased, which decreases the formation of unsaturated coke precursors. • The hydrogenolysis activity is decreased, which decreases the production of light gases. • The concentration of surface hydrogen is modiﬁed, which affects the relative rates of different reaction intermediates and the ﬁnal selectivity.

19.2 Catalytic Reforming

• The amount and strength of the acid sites of the support are modiﬁed by the higher oxidation state metal additives. The mechanism by which promoter metals modify the activity and selectivity of the Pt is attributed to the dilution of Pt atoms by promoter atoms (geometrical effect) and/or to a change in electronic conﬁguration of Pt due to the relative electron acceptor or donor properties of the promoter metal (electronic effect). Trimetallic catalysts contain three metal elements and are commercially quite common. The main advantages of trimetallic catalysts are increased selectivity and stability compared to bimetallic or monometallic catalysts. Pt is the main component of trimetallic catalysts, supplying the metallic function. Alumina is a common acidic support. Examples are Pt–Re–Sn; Pt–Ir–Sn; Pt–Sn–M where M is Bi, Te, Au, Ir, or Pd; Pt–Ir–Ge; and Pt–Re–Ge. The third metal is added to Pt–Re to reduce the high hydrogenolysis activity in order to control the production of light products. Reduction of cracking by the alumina is secondary. The oxides of Ge, Sn, and Pb are quite refractory and they are only partially reduced to the metal during reforming. Only a fraction of these metals interacts with the Pt. The higher oxidized species remain on the support and modify the acidity of the support [17]. Trimetallic catalysts of Pt, Re, and Sn on chlorinated alumina generally have lower catalytic activity and lower sulfur tolerance than Pt monometallic and the bimetallic catalyst of these metals. The order of the Pt activity generally decreases in the order monometallic > bimetallic > trimetallic catalysts. This suggests an additive effect of Re and Sn on Pt. The rate of deactivation during sulfur addition with the trimetallic catalyst is intermediate between that of the bimetallic Pt–Re and Pt–Sn. This can be explained by simultaneous poisoning of Pt–Re and Pt–Sn sites [18, 19]. A 0.3% Pt–0.3% Re–0.6% Sn on chlorinated γ -alumina catalyst, when reforming n-heptane, was similar in activity and selectivity to the best bimetallic catalyst (0.3% Pt–0.3% Re with 0.06% sulfur) on the same support, but did not need presulﬁding. The trimetallic catalyst is more stable and shows less hydrogenolysis activity as measured with cyclopentane and toluene. Tin simultaneously affected the metal and acid functions but inhibited them to different degrees, resulting in improved activity, selectivity, and stability of the catalyst. These characteristics are commercially very relevant: most of the industrial reforming catalysts are trimetallic. Reforming catalysts are reversibly deactivated by coke. They are permanently poisoned by various elements (arsenic, copper, lead, cobalt, molybdenum, sodium, phosphorus, and silicon) that may be contained in the feedstock either naturally or through upstream additives. The catalysts are especially sensitive to sulfur, which can be either irreversibly or reversibly adsorbed on the metal sites. Organic nitrogen compounds neutralize acid sites. To protect the catalyst, the feedstock is hydrodesulfurized to less than 0.5 ppmw or less depending upon the catalyst. Catalytic reforming is not very sensitive to coke. Different catalysts differ in coke forming behavior. Two types of coke forms on metallic sites: reversible coke (H/C atomic ratio of 1.5–2.0) and a more graphitic, irreversible coke (H/C atomic ratio about 0.2) that is much harder to remove with hydrogen. The very reactive, reversible

597

19 Naphtha Reforming and Upgrading of Diesel Fractions

Pt

Pt–Ir Carbon (%)

598

Pt–Ge

0.5

Pt–Re

Pt–Sn

0

0.5

1.0

1.5

2.0

Time on stream (h) Figure 19.11

Effect of promoter metals on coke deposition.

coke associated with the metal sites is rapidly burned off. A less reactive coke forms on the acid sites and burns at higher temperatures during temperature programmed combustion. Comparing coke formation on commercial monometallic Pt and bimetallic Pt–Re catalysts under industrial conditions simulated in a ﬁxed-bed laboratory reactor, the main difference is the nature and distribution of the coke and not its quantity. The distribution between metal and acid sites also depends upon the type of catalyst. Bimetallic catalysts form small grain, ﬁnely dispersed ‘‘soft’’ coke located mainly on the support. This coke does not disturb the reforming reactions as evidenced by lack of a sharp activity decline in the presence of relatively great quantities of coke. Bimetallic alumina catalysts resist coking compared to monometallic Pt catalysts. In a laboratory experiment, coke deposition increased during the ﬁrst 2 hours of heptane reforming in the order Pt–Sn < Pt–Re < Pt–Ge < Pt–Ir < Pt. Figure 19.11 shows that bimetallic catalysts present increased ability to resist coking [15]. The trends in semiregenerative commercial catalysts are trimetallic catalysts to reduce hydrogenolysis activity of Pt–Re, modiﬁcation of the support acidity to decrease cracking, staged loading of multiple catalysts, and the decreased catalyst density. Trends in continuous catalysts include increased activity and selectivity to aromatics (allows lower reactor pressure), decreased deactivation rate, lower Pt loading, and higher resistance to attrition. New manufacturing techniques produce higher and more controlled Pt dispersion. 19.2.3.1 Zeolite Catalysts Various molecular sieves, in combination with a hydrogenation–dehydrogenation metal component, can catalyze naphtha reforming. Hydrogenation–dehydro genation metals are generally the noble metals Pt, Pd, or Ir.

19.2 Catalytic Reforming

Quite a number of patents claim the use of zeolites and/or zeotypes as support for reforming reaction. Selected zeolites considered for this application include erionite, Beta, MCM-22, SSZ family (i.e., SSZ-47, SSZ-53, SSZ-55, SSZ-57, SSZ-58, SSZ-59, SSZ-60, SSZ-63, SSZ-64), SAPO-37, mordenite, ZSM-48, ZSM-5, ZSM-12, USY, L (see [3] for an extensive list of patents). The ﬁrst zeolite introduced in reforming process was erionite (i.e., small pore zeolite with eight rings), which was placed in the last bed of the reactor. This post-reforming process was called Selectoforming (Mobil). Only normal parafﬁns can react inside the small pore (reactant shape selectivity (RSS)) and the octane number increase is due to their selective cracking producing LPG and to the increase in the concentration of i-parafﬁns and aromatics [14]. However, the most relevant zeolite for catalytic reforming is probably zeolite L (i.e., large pore zeolite with 12 rings) [3]. Platinum on potassium-exchanged L zeolite is a commercial catalyst that shows increased activity for aromatics in parafﬁnic feedstocks [20]. A process for reforming naphtha containing at least 25 wt% C6 –C9 aliphatic and cycloaliphatic hydrocarbons using an L zeolite, platinum, halogenated catalyst where the zeolite crystals are cylindrical with an average length of 0.6 µm or less has been patented. The reformate has a high C6 –C8 light aromatics content and a reduced content of heavier C9 and C10 aromatics [21]. Zeolite L provides the highest activity during reforming of a parafﬁnic commercial feedstock, compared to large pore zeolites like Beta and USY [15]. Pt on potassium-exchanged L zeolite (Pt/K-L zeolite) catalyst produces very high selectivity for benzene when reforming n-hexane. This contrasts with platinum-containing alumina catalysts, which are effective for producing xylenes and other aromatics with more than eight carbon atoms from C8 + parafﬁnic hydrocarbons. The Pt/K-L zeolite catalyst family is modiﬁed by alkali or alkali earth to modulate acidity [22]. Hybrid catalysts composed typically of HZSM-5 zeolite modiﬁed with a metal component such as Ga, Zn, Y, or Ag are potentially good catalysts for aromatizing parafﬁns. These hybrid catalysts are very selective for conversion of light parafﬁns and oleﬁns to aromatics. Competition occurs between the cracking processes that consume hydrogen and the dehydrogenation reactions that release hydrogen. The enhanced production of oleﬁns and aromatics is theorized to be due to a long-distance hydrogen transfer from the zeolite’s acid sites to the metal sites, long-distance hydrogen back-spillover. ZSM-5, a medium pore zeolite with 10 rings, is used in a reforming catalyst in the Mobil M-Forming process [23]. The channel size of ZSM-5 catalyst can admit singly branched parafﬁns as well as simple aromatics such as benzene and toluene. The second lowest octane components, the single branched parafﬁns, are removed by hydrocracking. Furthermore, the aromatic species are alkylated by the oleﬁnic component of the cracked product. The alkylaromatics contribute to octane number and reduce the loss of cracked products to gas, thus increasing the liquid yield [20]. Zeolite catalysts can also be used to treat reformate or stabilized liquid products from the reformer to change their composition. BTX content can be increased through cracking of ethylbenzene and heavier aromatics. An acidic zeolite is used

599

600

19 Naphtha Reforming and Upgrading of Diesel Fractions

to crack the aromatics to the BTX range when it is economically desirable, taking into account that this is accompanied by a reduction in gasoline yield and hydrogen [16]. A process was initially presented by Mobil, named Mobil Reformate Upgrading (MRU). The zeolite is placed at the bottom of the last reactor in a ﬁxed-bed reformer, where it reacts with reformate prior to separation of the light gases. A combination of M-Forming and reforming has been commercialized by ExxonMobil (BTXtra): the zeolite catalyst is placed in the last bed of a semiregenerative reformer to improve product quality, increasing the concentration primarily of toluene and xylenes by transalkylation of the methyl groups from C9 aromatics with benzene. Some additional BTXs are produced as a result of dealkylation. This process was ﬁrst applied at ExxonMobil’s Chalmette (Louisiana, USA) reﬁnery in 1996 [24]. 19.2.3.2 Commercial Catalysts Naphtha reforming catalysts represent about 7% of the total value of all petroleum catalysts, the same percentage as hydrocracking catalyst [10]. Naphtha reforming catalysts contain about 0.2% to over 0.6% precious metals, platinum, and sometimes rhenium, on alumina or zeolite matrix. Platinum content is typically about 0.25 or 0.375%. Some formulations use platinum along with palladium, ruthenium, or iridium. Commercial reforming catalysts surface area is large, approximately 200 m2 g – 1 . Only 2–4 m2 g – 1 is the Pt or Pt–Re metal surface, the remainder is alumina. In commercial manufacture, the catalyst support for ﬁxed-bed processes may be extrudate cylinders. Alumina supports for moving-bed processes are spherical to reduce attrition and are usually produced by the ‘‘oil dropping’’ method [13, 25]. Catalyst life generally is long, over four years with seven to nine years obtained in some cases. The attrition or make-up rate in continuous catalytic reformer units is less than 2% per year [3]. When worn out, catalyst in either ﬁxed-bed or moving-bed units is replaced and the platinum can be recovered through state-of-the-art metal reclaiming technologies. The reformer catalyst market is competitive. More than 60 formulations are offered to meet the customers’ varying needs. Companies offering naphtha reforming catalysts in 2005 included the following [26]:

• • • • • • •

Axens BASF Criterion Catalyst Co. LP ExxonMobil Indian Petrochemicals Corp. Ltd. RIPP, Sinopec UOP.

A total of 22 naphtha reforming catalysts offered by companies listed on the 1997 list also appear on the 2005 list. Catalyst manufactures responded with 29 new catalyst formulations to increasingly stringent gasoline speciﬁcations that impacted reforming units and to the need to maximize proﬁtability of the existing reformers. These catalysts improved the octane, reformate, and hydrogen yields and selectivity. Attrition losses of CCR catalysts were reduced.

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing

The term cold-ﬂow properties of gasoils, often associated with the words good or poor, refers to the ﬂow characteristics of a diesel fuel after lowering its temperature; in other words, it refers to the ability to operate under cold weather conditions. From a chemical stand point, petroleum derived products are a mixture of innumerable compounds that fall into different classes such as alkane, isoalkane, naphthenes, and aromatics. Among these, normal and slightly branched parafﬁns present relatively high melting points. Considering a typical 250–360 ◦ C cut, the melting point of normal parafﬁns ranges between 10 and 40 ◦ C, while the melting point of the heaviest methyl parafﬁn present (C21 ) is, depending on the position of the branching, between –4 and 13 ◦ C. Figure 19.12 shows the relationship between molecular weight and melting points for normal and methyl branched parafﬁns [27]. The extent of decrease in the melting point signiﬁcantly depends on the degree, the position, and the length of the branching. Branching positions in the middle of the chain have a higher effect than those near the end of the chain, while longer side chains lead to a higher decrease of melting point [28]. Engines running on diesel should be able to operate even in cold weather conditions where temperature can reach values well below 0 ◦ C. On the basis of the data presented above it is evident that such temperatures may cause the crystallization of those compounds presenting the highest melting points, thereby affecting the ﬂow characteristics, and ultimately may lead to solidiﬁcation of the product. Standard test methods for determining these properties are cloud point (CP; ASTM D2500), pour point (PP; ASTM D97), and cold ﬁlter plugging point (CFPP); IP 309). The CP is the temperature at which a haze of wax crystals is formed; PP is the highest temperature at which the product still ﬂows, while CFPP is the highest temperature at which the wax crystals severely reduce the ﬂow through a ﬁlter. 100 80

Melting point (°C)

60 40 20 0 −20 −40 −60

n -Paraffin

2 Methyl

3 Methyl

5 Methyl

−80 −100 10

Figure 19.12

15

20

25 30 Carbon number

35

Melting points of normal and branched parafﬁns.

40

45

601

602

19 Naphtha Reforming and Upgrading of Diesel Fractions

Cold-ﬂow speciﬁcations change according to the country and period of the year. As example, maximum CFPP values allowed for winter diesel fuel range from –5 ◦ C in Greece to –12 ◦ C in Italy up to –32 ◦ C in Sweden.1) Different options are available to improve the cold-ﬂow properties of gas oil: • The simplest is to reduce the backend cutpoint and add kerosene blendstock. This action removes the heavier parafﬁnic molecules that tend to precipitate at higher temperatures. However, reduction of the end point results in signiﬁcant reduction of diesel yield from the crude. • Removal of wax-forming components by crystallization after dilution of the feedstock with a suitable solvent and chilling (i.e., solvent dewaxing). This option is generally economical only for feedstocks with a high content of parafﬁns, and it is mainly used for the production of lubricating oil base stocks. • Use of cold-weather additives to meet seasonal low-temperature pumpability requirements for diesel fuel and furnace oil. • Catalytic dewaxing process wherein the waxy parafﬁn are selectively cracked or isomerized. The catalysts, the chemistry, and the process for the last approach to dewaxing are the subject of this paragraph. Catalytic dewaxing is a process where the products responsible for the deterioration of the cold-ﬂow properties are removed by selective cracking to lighter compounds, by boiling in the gasoline and LPG range (i.e., cracking dewaxing), or by isomerization into more branched lower melting point structures minimizing as much as possible the cracking (i.e., isomerizing dewaxing). It should be noted that as consequence of the operating mechanism with the former route, part of the feedstock will be cracked to naphtha and LPG, thereby leading to a decrease of yields; differently by isomerising dewaxing where the improvement of cold-ﬂow properties is achieved by converting the waxy compounds to more branched compounds of the same carbon number, the process is characterized by higher yield retention in the diesel cut. 19.3.1 Shape Selectivity

Whether it occurs by selective cracking or isomerization, the possibility of improving the cold-ﬂow properties of a given feedstock is based on the so-called shape selectivity behavior of suitable zeolitic materials where the reactivity of molecules is strongly affected by their size and conformation. The concept of shape selectivity was proposed for the ﬁrst time by Weiz and Frilette [27, 29, 30] at the beginning of 1960s to explain the unexpected results 1) Directive 2003/17/EC of the European Par-

liament and of the Council of 3 March 2003, amending Directive 98/70/EC relating to

the quality of petrol and diesel fuels, Ofﬁcial Journal of the European Communities, L76/10, 22.3.2003.

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing

obtained with calcium-ion-exchanged zeolite. Since then, the numerous studies carried out in this ﬁeld have led to the discovery of new forms of shape selectivity and allowed gaining a deeper understanding of the complex interactions taking place between the porous structure of zeolites and the reacting system. Current theories proposed to explain different ﬁndings are clearly described and discussed by Weitkamp et al. [31], Marcilly [32], and more recently in a review by Degnan [33]. The most widely accepted theories for explaining shape selectivity are reported below. • Reactant Shape Selectivity (RSS): This type of selectivity occurs when the difference in reactivity between two competing molecules is caused by their different accessibility of the internal active sites of the zeolite. When the pore opening of the zeolite is similar to the size of the reactant molecule, a small variation in the ‘‘diameter’’ of the latter results in a change of its diffusion by orders of magnitude and in the extreme case hinders the access to the internal surface [27, 34]. In this case, as depicted in Figure 19.13, only those molecules having a diameter small enough to enter the porous structure will react, while the bulkier ones bypass the internal pore structure of zeolite leaving the system unchanged. RSS is the operating mechanism during catalytic dewaxing with ZSM-5-based catalysts. • Product shape selectivity (PSS): In this case, the reactants are small enough to diffuse easily through the channels of the zeolite, but among all the possible products that can be formed, only those with a size small enough in relation to the pore size can diffuse more or less rapidly and leave the zeolite structure. An example of product shape selectivity is reported in Figure 19.14. • Transition state shape selectivity (TSS): It refers to the situation where the lack of space around the active sites hampers the formation of transition states spatially constrained either by their size or their orientation. Besides the example reported in Figure 19.15, other examples are the inhibition of coke formation within

+

Figure 19.13 Example of reactant shape selectivity. Selective cracking of n-hexane in presence of i-hexanes.

+

Figure 19.14 Example of product shape selectivity. Formation of para-ethyl toluene with ethylene. Adapted from [31].

603

+

19 Naphtha Reforming and Upgrading of Diesel Fractions

+

No toluene and trimethyl-benzene

+

604

Figure 19.15 Example transition shape selectivity. Suppression of trimethylbenzene during meta-xylene isomerization.

ZSM-5 [35, 36], long-chain parafﬁn hydroisomerization [36], and most likely dewaxing by isomerization [37]. Besides the classical and well-accepted concepts reported above, new forms of shape selectivity have been advanced to explain different experimental results. These concepts are molecular trafﬁc control, pore-mouth, and key–lock selectivity; inverse shape selectivity; window or cage effect; and nest effect. A clear description and discussion of the more recent theories, which still continue to be a matter of debate, have been presented by Degnan [33] and Weitkamp et al. [31]. Over the last few years, a signiﬁcant help to gain a deeper understanding of the shape selectivity phenomenon has come from application of computational simulations to analyze the behavior of molecules within the porous structure of zeolites and to estimate the probability of formation of a given compound. A particularly active ﬁeld concerns the simulation of normal and branched parafﬁns, aimed at understanding how topology of the internal pore structure of a zeloite can affect the selectivity with which particular isomers are formed during the hydroisomerization process. A recent review by Smit and Maesen describes the most recent advances with the so-called free-energy landscape approach [38].2) A central premise of this approach is that by ignoring the detailed chemical characteristics of a zeolite and simply quantifying how its topology affects the free energies of formation of the various reactants, intermediates, and products involved (that is the free-energy landscape of the reacting system), it is possible to identify the fundamental interactions and processes that control the shape selectivity of a particular transformation. Using the above-mentioned method, the authors [38, 39] were able to offer an explanation of the different products distributions during hydroisomerization of n-decane [40] over zeolites Y, ZSM-5, ZSM-11, and ZSM-22 and the different i-C4/n-C4 ratio in the cracking products exhibited by ZSM-11and ZSM-5 zeolites. Different cases of shape selective transformation have been reanalyzed (window effect, pore-mouth, and key–lock) by using this approach, leading to different 2) Free-energy landscape approach consists of

calculating the effect of zeolite topology on

the free energies of formation of the various molecules involved in catalytic reactions.

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing

explanations of the mechanism originally proposed [41, 42]. A case particularly interesting in this contest is the behavior of SAPO-11 during hydroisomerization of n-parafﬁns whose isomer products distribution were explained by assuming that the branching reaction occurred at the pore mouth [43]. However, recent works [37, 44] have challenged this hypothesis: they explain isomers distribution by geometrical restriction within the channels to form the transition state of dibranched isomers. 19.3.1.1 Catalytic Dewaxing via Shape Selective Cracking The preferential cracking to lighter compounds of normal and slightly branched parafﬁns in comparison with the more branched ones occurs when the zeolite has a pore size that selectively allows the normal and near-normal parafﬁns to enter the pore structure at the expense of highly branched parafﬁns [31, 32, 45, 46]. When the zeolite presents channel dimensions approaching those of the reactant molecules, the diffusivities of the latter change rapidly as a function of their size. In a regime controlled by conﬁgurational diffusion, relatively small increases of molecular size may results in a decrease of diffusivity by orders of magnitude as shown in Figure 19.16. As a result, the rate of cracking of alkanes strictly depends on their steric hindrance as shown in Figure 19.17. As observed by Marcilly [32], the trends of parafﬁn cracking rate and diffusivity in ZSM-5 zeolite as a function of branching degree are similar but the relative differences of cracking rate are lower than the relative differences of diffusivity. The explanation came from the higher reactivity of branched parafﬁns, which in part compensate the lower diffusivity [27, 47]. A clear example of catalytic dewaxing of gasoil via selective cracking of n-parafﬁns is reported in Figure 19.18 [48], showing the GC trace of a gasoil before and after dewaxing step. Since selective cracking of n-parafﬁns occurs in a regime controlled by conﬁgurational diffusion, the observed reaction rates and selectivity should be affected 1.0E-01 Alkanes, 315 °C

Diffusivity (cm2 s−1)

1.0E-03

Aromatics, 500 °C

1.0E-05 1.0E-07 1.0E-09 1.0E-11 1.0E-13 0.3

Figure 19.16

0.4

0.5

0.6 0.7 0.8 Molecular size (nm)

0.9

1

Diffusivity of n-hexane, i-hexanes, and alkyl benzenes in ZSM-5.

605

19 Naphtha Reforming and Upgrading of Diesel Fractions

0.8 0.7 Relative cracking rate

606

0.6 0.5 0.4 0.3 0.2 0.1 0 0

Figure 19.17

1

1 Degree of branching

2

2

Relative cracking rates on ZSM-5 of n-hexane and i-hexanes.

by crystallite size while energy of activation should be relatively low. The results obtained by Sivasanker et al. and Kennedy et al. [49–51] conﬁrm this picture. A decrease in crystallite size increases the activity as well as the selectivity owing to the reduction of secondary cracking. The observed activity is adequately described by using effectiveness factors, which, according to the average crystallite diameters of the sample used, reach values as low as 0.1. The poisoning by a relatively bulky nitrogen-containing molecule (e.g., 5,6 benzoquinoline), which diffuses more slowly than the normal parafﬁn used, was found to be well described by a shrinking core model. The latter is typical of diffusion-limited situations. Furthermore, it was found that de-alumination and passivation of the outer surface of crystallite led to lower activity and higher selectivity. Changing the C20

C21

C22

C23 C24 C25

Heavy gas oil Pour point = +6 ° C

Dewaxed gas oil Pour point = −29 ° C Figure 19.18

Selective removal of n-parafﬁn in gas oil dewaxing. Adapted from Ref. [48].

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing

silica–alumina ratio between 20 and 80 did not affect the performance in dewaxing except for the lowest aluminum content [49, 52]. Isomorphous substitution of Al in ZSM-5 with boron, iron, and gallium was found to increase signiﬁcantly the selectivity. This result was mainly attributed to the lower strength of acid sites in the ‘‘substituted’’ material resulting in a decreased secondary cracking activity [46]. In this connection, another point that should be considered is the effect of isomorphous substitution of Al on the unit cell volume and hence on the pore size [31]. Tielen et al. [53] investigated the effect of isomorphous substitution on the shape selective behavior of MFI-like materials in hydrocracking of n-decane. They found that beryllosilicate MFI gives considerably lower yields of i-pentane formed by hydrocracking of n-decane. Tielen et al. took this result as indicative of a reduced intercrystalline space available for catalytic reaction and hence of the pore width. However, the results mentioned above should be examined carefully because isomorphous substitution changes also drastically the nature of active sites and for this reason it is difﬁcult to separate the inﬂuence of both factors. Some interest has been shown by Mobil [54], Amoco [55], and Chevron [56] for boron-substituted zeolites [45, 46]. The performances of ZSM-5-like materials strictly depend on the presence and type of the metal function [32]. With a catalyst without a metal phase, the naphtha formed as by-product is characterized by a high content of oleﬁns, the presence of aromatics, and consequently good octane numbers. However, under such operating conditions, a more or less fast decrease of activity, caused by coke buildup, is observed. The presence of a more or less strong hydrogenating phase makes the catalyst more stable, increases the selectivity, but leads to lower octane number of the naphtha produced. From a general stand point, the improvement of cold-ﬂow properties by cracking of waxy compounds leads to a reduction of yields and a decrease of cetane numbers. 19.3.1.2 Dewaxing via Isomerization The operating mechanism in the case of ZSM-5 and similar molecular sieves inevitably leads, depending on feed composition, to more or less high cracking yields of lighter products [32, 45, 46]. An alternative and potentially more efﬁcient approach to improve the cold-ﬂow properties of petroleum cuts results from converting the high-melting-point compounds to valuable products by isomerization and simultaneously minimizing the yield loss associated with the cracking reactions of both normal and i-parafﬁns. Extensive work has been done worldwide to overcome the intrinsic limitations associated with the use of ZSM-5-like materials. A possible approach [32, 45] is to use materials characterized by a wider porous structure that should allow a faster diffusion of branched parafﬁns out of the material pore structure thereby preventing the subsequent cracking. Indeed, large pore zeolites present higher isomer selectivity in comparison with medium pore zeolites [57]; however, it should be noted that materials with wider pore openings exhibit lower reactant selectivity and do not hamper the formation of bulkier multibranched parafﬁns

607

608

19 Naphtha Reforming and Upgrading of Diesel Fractions

that are particularly prone to cracking reaction via the β-scission mechanism [31, 32]. As a consequence, it is not be possible to avoid signiﬁcant cracking. In these circumstances, high isomerizing selectivity is obtained by an optimal choice of density and strength of acid sites with the presence of a metal phase with a strong hydrogenating/dehydrogenating activity, such as Pt or Pd. Along this line, different materials (i.e., zeolite beta (BEA), MCM-22 (MWW) with high silica–alumina ratio, mesoporous silica alumina (i.e., MCM-41, HMS, MSA), amorphous silica–alumina, halogenated alumina, borosilicate, boro-aluminosilicate isostructural with zeolite beta) have been investigated [55, 58–66]. An alternative solution for isomerizing normal parafﬁns minimizing the cracking reaction is to exploit the transition shape selectivity, or, in other words, to carry out the isomerization reaction, over a suitable zeolite having a topology that allows the isomerization of normal parafﬁns but which hampers the formation of those multibranched parafﬁns that have a structure (i.e., α-α-γ ) particularly prone to cracking [67]. Work carried out at Chevron Research and Technology led to the development of an isomerization dewaxing process, most likely based on SAPO-11, commercialized in 1994 [68]. After the ﬁrst patents issued at the end of 1980s [69–71], a number of patents were successively issued where criteria considered important for the high performance of the catalyst were claimed [46, 72–74]. Hydroconversion studies carried out with model compounds (n-octane and n-hexadecane) on bifunctional catalysts made up of noble metal (Pt, Pd) loaded on different acidic supports (SAPO-11, SAPO-5, ZSM-5, HY, and ASA) [75–77] showed that the best materials in terms of selectivity to isomers were ASA and SAPO-11. However, for conversion higher than 87%, SAPO-11 maintained a high selectivity whereas ASA showed a consistent decrease. The experimental evidences presented by Miller [75, 77] and results obtained by Maesen with computer modeling [37] strongly suggest that the high selectivity for isomerization presented by the SAPO-11 results from transition state selectivity phenomena occurring inside the pore structure. Alternatively the high isomer selectivity and their distribution can be interpreted in terms of pore-mouth and key–lock catalysis [43, 78]. A remarkable feature of isomerization products is the scarce presence or absence of bulky gem-dimethyl branching with preference for methyl branching separated by more than one carbon. If the formation of gem-dimethyl ramiﬁcations is blocked, the rate of cracking of isomers is considerably lowered owing to the less energetically favorable β-scission route available as reported below. As shown in Figure 19.19, the well-known mechanism of n-parafﬁn conversion on bifunctional catalyst proceeds through the formation of oleﬁn at the metal site and the subsequent formation of carbenium ion at the Brønsted acid sites to form a secondary carbenium ion, the latter rearranging to a tertiary carbenium ion which can either desorb to give the corresponding iso-oleﬁn or undergo β-scission with the production of a smaller carbenium ion and oleﬁn. The oleﬁn and/or isooleﬁn are subsequently hydrogenated at the metal site with formation of the corresponding alkane or isoalkane. As shown in Figure 19.20, from a macroscopic point of view the process can be viewed as a series of consecutive reaction where the

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing

Hydrogenolysis

n -Alkane

Cracking products

n -Alkenes

n-Alkyl secondary C cations

Cracking products b –scission

Rearragement

i -Alkenes

609

Cracking products

i-Alkyl tertiary C cations Hydrogenolysis

i -Alkanes

Cracking products

Figure 19.19 Reaction scheme for hydroisomerization and hydrocracking of n-alkanes on bifunctional catalyst.

n-Paraffin

Monobranched paraffin

r1

Dibranched paraffin

r2

r3

Multibranched paraffins

r4

Cracking products

r1<
Reaction scheme for the formation of isomers and cracking products.

n-parafﬁn is converted into monobranched parafﬁns, which in turn are transformed into dibranched, and so on. Along with the isomerization reaction, the reactant molecules are cracked by a β-scission mechanism as well with increasing relative rate passing from n-parafﬁns to multibranched isomers. As shown in Table 19.3, β-scission of type A, which is by far the fastest, requires a carbenium structure where the branching groups are in ααγ position on the parafﬁnic chain. Remarkably slower is the cracking via the B1 mode, which implies the presence of structures of gem-type and occurs through a less energetically favorable route in comparison with the type A. Always in decreasing order, other modes of cracking are B2 concerning the dibranched parafﬁn with isolated branching groups, monobranched parafﬁns, and, ﬁnally, normal parafﬁns. In these circumstances, the low selectivity of SAPO-11 for the gem-dimethyl species, with preference for methyl branching separated by more than one carbon, which can crack only via the B2 or C mode, can explain the high selectivity for isomerization of this catalyst. 19.3.2 Commercial Applications

The ﬁrst example of exploitation of the shape selectivity at the industrial level is due to Mobil (now ExxonMobil) with the development of the selectoforming process for removing normal and slightly branched parafﬁns from gasoline in order to increase the octane number (see also paragraph 2.3.1.) [79]. The catalyst used was a shape

610

19 Naphtha Reforming and Upgrading of Diesel Fractions Table 19.3

Type

Different β-scission modes of carbenium and relative scission rates [67]. Carbenium structure

Products of β-scission

Carbenium ions involved

Relative reactivity

A

+

+

Ter→Ter

1 050(170a)

B1

+

+

Sec→Ter

2.8

B2

+

+

Ter→Sec

1

C

+

+

Sec→Sec

0.4

B2

+

+

Sec→Prim

≈0

a

Martens, J.A., Tielen, M., and Jacobs, P.A. (1987) Cat. Today 1, 435.

selective zeolite, erionite, with a small amount of Ni [27]. Later on, the performance of the process named M-forming was improved by using the zeolite ZSM-5 without a hydrogenating function but in the presence of hydrogen. These two processes were developed to improve the octane of gasoline but are hardly used nowadays. In 1970, British Petroleum (BP) commercialized the ﬁrst process for heavier feedstock based on mordenite to selectively crack normal and slightly branched parafﬁns. Subsequently in 1977, Mobil announced the development of processes for the catalytic dewaxing of gas oil (MDDW, Mobil Distillate DeWaxing) and Heahier feedstock (Mobil Lube DeWaxing, MLDW) using a ZSM-5-based catalyst. More recently, in 1996, Mobil began licensing the Mobil Isomerization DeWaxing (MIDW), which uses a dual-function, noble-metal, proprietary catalyst. In this case, improvement of cold-ﬂow properties is achieved by isomerizing high-melting-point compounds rather than cracking them into lighter products. Chevron was the ﬁrst to introduce an isomerization dewaxing process in 1993 using a catalyst presumably based on SAPO-11 [80, 81]. Other dewaxing technologies commercially available have been developed by Akzo/Fina [79, 82], UOP [83], S¨ud Chemie [84], and Shell Global Solution/Criterion [85]. 19.3.2.1 Commercial Processes From a general point of view, processes for catalytic dewaxing of gasoil can be divided in two categories depending on the type of metal present in the catalytic system. With sulfur-tolerant catalysts, whose metal phase is generally made up of Ni or bimetallic systems like Co/Mo or Ni/Mo, the process conﬁguration is single stage with stacked beds of hydrotreating, catalyst to remove sulfur and nitrogen compounds, and dewaxing catalysts, as depicted in Figure 19.21.

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing

Feed with decreasing levels of sulfur and nitrogen

HDT

HDT

DW

HDT

DW

HDT

DW

HDT

HDT

For a given feed catalyst with increasing tolerance to sulfur and nitrogen

Figure 19.21

Single-stage conﬁguration for dewaxing and deep HDS [45, 48].

With noble-metal-based catalysts requiring a feedstock with low level of contaminants (i.e., sulfur and nitrogen), a double-stage conﬁguration is used. In this case, the feedstock is ﬁrst subjected to severe hydrotreating, the H2 S and NH3 so formed are removed, and the clean feedstock is sent to catalytic dewaxing unit. A simpliﬁed scheme of the double-stage conﬁguration is shown in Figure 19.22.

HDT DW HDT

DW

H2S, NH3 Figure 19.22

Double-stage conﬁguration for dewaxing and deep HDS [45, 48].

611

19 Naphtha Reforming and Upgrading of Diesel Fractions

100 MDIW Operating range 90 Diesel yield % (150+ °C)

612

80 70 60 MDDW (ZSM-5) Operating "line" 50 40 0

10

20

30

40

50

60

Cloud point reduction (°C) Figure 19.23

Diesel cold-ﬂow improvement [86].

A comparison has been made between the cracking and isomerization routes [79] in catalytic dewaxing. The former can be carried out in a single stage and hence is less expensive, but some portion of diesel-range material will be cracked to naphtha and LPG. On the other hand, the hydrodewaxing process needs to strip H2 S and NH3 between the two stages to provide a clean environment for the noble-metal isomerization catalyst in the second d stage. The n-parafﬁns are isomerized, resulting in a lower CFPP without converting the components in the diesel range. Furthermore, the produced gasoil has a lower density, higher cetane number, and lower polyaromatic content than the gasoil deriving from cracking dewaxing. Figure 19.23 compares the diesel yield at increasing CFPP reduction by these two processes [86]. ABB LumusGlobal/Criterion Catalyst/Shell Global Solution SynFlow Syn Technologies is a family of hydrotreating processes based on a combination of different catalysts to address speciﬁc problems connected with middle distillate upgrading [48, 79, 85, 87]. Among these, SynFlow is designed to produce ultralow-sulfur diesel HDS and to meet cold-ﬂow speciﬁcations. The SynCat catalyst for SynFlow can be activated or deactivated depending on reﬁnery’s desired product characteristics (see Table 19.4). Three catalysts are available for improving cold-ﬂow properties.

• SDD-800 is a sulfur- and nitrogen-tolerant base-metal dewaxing used in combination with HDS catalyst in a single-stage process where the reactor is loaded with stacked beds of HDS and dewaxing catalyst. • SDD-801 is a noble-metal dewaxing combining hydrodewaxing and hydrogenation of aromatic structures. It requires feedstock with low levels of sulfur and

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing Product characteristics of summer and winter operating mode with single-stage conﬁguration.

Table 19.4

Operating modes

Summer

Catalyst loading Liquid yield (C5 + ) (vol%) Diesel yield (150 ◦ C + ) (vol%) Product properties Sulfur (ppm) Nitrogen (ppm) Cloud point (◦ C) Pour point (◦ C)

DN-190 (HDT) SDD-800 (DW) DN-190 (HDT) SDD-800 (DW) >100 >100 >100 >90 <50 <1 –6 –9

Winter

<50 <1 –35 –45

Feed properties – density, 0.852 g cm –3 at 15 ◦ C; cetane index, 52; sulfur, 7500 ppm; nitrogen, 121 ppm; cloud point, –6 ◦ C; ASTM D-86 (◦ C) 10/50/90, 233/288/353. Adapted from [48].

consequently it is used in combination with HDS/HDN catalyst to remove sulfur and nitrogen as a ﬁrst stage. The H2 S and NH3 so formed are removed and the clean feedstock is sent to the second stage for dewaxing and partial saturation of aromatics. • SDD-821 is a noble-metal dewaxing catalyst that presents signiﬁcantly higher yield retention than SDD-801 [85].

Akzo-Fina – CFI Dewaxing (Cold-Flow Improvement) Albemarle together with Total Research offers technology for the improvement of cold-ﬂow properties (CFI) by selective cracking of long-chain parafﬁns [79, 82]. CFI technology combines a sulfur-tolerant, shape selective dewaxing catalyst to improve cold-ﬂow properties (pour point, cloud point, CFPP, freezing point) with HDS for production of ultralow-sulfur diesel. The technology usually operates under moderate hydrogen pressure but can be tailored to meet the speciﬁc needs of the reﬁner by proper selection of catalyst systems and operating conditions. Improvement of 18 ◦ C pour point and diesel yield higher than 90% have been reported in ultralow-sulfur diesel production. ExxonMobil MDDW, MIDW The MDDW process was commercialized in 1978 and since then it has been widely applied worldwide [27, 79, 88]. More than 30 MDDW units are in operation [32, 45, 79]. The process uses a single-stage reactor loaded with a ZSM-5-based catalyst to improve the cold-ﬂow properties of MD. The catalyst can be tailored according to the application in terms of silica/alumina ratio, extrusion procedure, or the presence or not of a metal phase. Typical operating conditions are 260–455 ◦ C; inlet pressure 20–50 kg m – 2 ; and LHSV (Liquid Hourly Space Velocity) 1.0–2.5 per hour. Overall yields in middle distillate and ﬁnal cold-ﬂow properties depend on parafﬁn content of the feed as shown in Table 19.5.

613

614

19 Naphtha Reforming and Upgrading of Diesel Fractions Table 19.5

Commercial MDDW product yields and properties. Feed parafﬁn content Low Medium

Yields (wt% feed) C6 + gasoline Distillate Distillate properties (◦ C) Pour point Cloud point T95 (◦ C)

High

7.2 87.4

9.2 82.9

7.8 71.4

–31 –28 342

–12 –2 416

–12 –6 412

Adapted from [79].

Mobil Isomerization DeWaxing (MIDW) The ﬁrst MIDW unit was started up in 1990 at Jurong reﬁnery in Singapore, while the technology was made available for licensing in 1996 [79, 89, 90]. The process consists of a ﬁrst-stage conventional hydrocracking unit where conversion is limited to 60% to avoid overcracking of light fractions. The bottom fraction from the ﬁrst stage is sent to a second stage hydroisomerization hydrocracking unit to increase the overall conversion level up to 75–80%. The second-stage catalyst is a dual-function noble-metal-on-zeolitic support to remove the waxy component via isomerization and with a high selectivity for the middle distillate. The properties of atmospheric distillates derived from the Jurong unit are illustrated in Table 19.6. Depending on feed characteristics and contaminant level (i.e., S and N), different conﬁgurations can be used: (i) Direct conversion in an MIDW unit without any prior pretreatment; (ii) Hydrotreating followed by MIDW in a single-stage unit with stacked beds; (iii) Hydrotreating with interstage removal of H2 S and NH3 and conversion with an MIDW unit [79]. Table 19.6

Atmospheric distillate properties from two-stage.

Boiling range (◦ C) Yield (%) Gravity (◦ API) Sulfur (ppmw) Smoke point (mm) Freeze point (◦ C) Pour point (◦ C) Ceatane index P/N/A (wt%) Adapted from [91].

Feedstock

Naphtha

Kerosene

Diesel

Fuel oil

350–510 100 32.0 260 – – >38 – 44/39/17

C5–150 22.9 73.0 <1 – – – – –

150–255 21.5 49.5 2 32 <–54 – 52 –/–/10

255–388 19.2 34.7 20 – – –43 56 45/31/24

388 + 36.8 29.5 40 – – <–7 – 36/42/22

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing Table 19.7

Comparison between MDDW and MIDW performance.

Catalyst activity Distillate yield (%wt) Product properties Gravity (◦ API) Viscosity at 40 ◦ C (cSt) Cetane index T95 (◦ C) Sulfur (ppm)

MDDW

MIDW

Base 55.5

Base –63 ◦ C 95.0

Base 7.6 52 423 1200

Base + 8 5.7 63 408 200

Feed: 36◦ API, 8.7 cSt at 40 ◦ C, 32 ◦ C pour point, 800 ppm sulfur, 140 ppm nitrogen, 63% parafﬁns (total), 35% n-parafﬁns. Adapted from [79].

The process is characterized by high ﬂexibility in terms of the feed that can be processed with a wide range of feeds from kerosene to highly parafﬁnic hydrocracker bottoms. Typical operating conditions are 260–440 ◦ C and inlet hydrogen partial pressure 25–50 kg cm – 2 . Compared with MDDW, MIDW exhibits higher retention of parafﬁns in the distillate range, which translates into higher yields especially for highly parafﬁnic feedstock [45]. A comparison between MDDW and MIDW process for a LVGO to produce dewaxed product with a pour point of –7 ◦ C is given in Table 19.7 [79]. Unicracking/DW The UOP Catalytic dewaxing process is a ﬁxed-bed process for improving the cold-ﬂow properties of various hydrocarbon feedstocks [79, 92–94]. The process uses two kinds of catalysts. The ﬁrst is a high-activity HDS and HDN catalyst, while the second is a proprietary dewaxing catalyst. The latter is a dual-function, non-noble-metal zeolite catalyst to selectively hydrocrack the long-chain, high-melting-point parafﬁnic components in the feedstock. The zeolite support used for the dewaxing catalyst is most likely a ZSM-5-like material. Typical yields and properties for vacuum gasoil and diesel applications are reported in Table 19.8. UOP–MQD Unionﬁning The MQD Uninonﬁning precess has been developed to address different aspects related with the upgrading of middle distillates such as hydrodesulfurization and reduction of aromatics [79, 83, 94]. As in the case for improving of cold-ﬂow properties, two catalysts are available. HC-80 is a base-metal zeolitic catalyst that achieves dewaxing by selective cracking of the waxy normal parafﬁns to naphtha. It is used mainly in the single-stage mode (see Figure 19.24) in combination with an HDS catalyst, where the reactor is set up with stacked loadings of two catalysts. DW-10 is a catalyst made up of a noble metal on a proprietary zeolitic support, which improves cold-ﬂow properties mainly by isomerization of waxy parafﬁns. The presence of a noble metal requires the use of clean feedstock. This catalyst

615

616

19 Naphtha Reforming and Upgrading of Diesel Fractions Performance of UOP’s catalytic dewaxing process.

Table 19.8

VGO feed

Diesel feed

27.7 9500 690 4.25 30

35.1 1.7 1.0 – 21

0.5 24.5 75.0

2.5 24.5 73.0

27.4 20 20 3.63 –20.5

37.5 1.0 1.0 – –12.0

Feed properties Gravity (◦ API) Sulfur (ppm) Nitrogen (ppm) Viscosity at 100 ◦ C Pour point (◦ C) Product yield (wt%) C1–C3 C4–260 ◦ C naphtha Dewaxed product Dewaxed product properties Gravity (◦ API) Sulfur (ppm) Nitrogen (ppm) Viscosity at 100 ◦ C, cSt Pour point (◦ C) Adapted from: [79, 92].

is generally placed in the second stage of two-stage conﬁguration, as shown in Figure 19.25. In this case, the retention is higher than with the single-stage arrangement. Chevron-ICR 410 Chevron does not licence a speciﬁc process for dewaxing diesel streams [95]. The Chevron isodewaxing process is aimed at low-pour-point base Fresh feed

Recycle gas compressor

Reactor Quench

Heater

Wash water Separation Light ends

Hydrogen makeup

Stripper Makeup compressor

Sour water

Steam Diesel product

Figure 19.24

Single-stage MQD Unionﬁning process for base-metal catalyst.

19.3 Upgrading Diesel Fractions: Catalytic Dewaxing

Reactor charge furnace Recycle gas compressor

Amine scrubber

Lean amine Rich amine

Makeup gas First stage reactor

Quench gas

Hugh pressure cold separator

Hot stripper

Oil to low pressure cold separator

Wash water

Figure stage reactor

Water

Stripping gas Stripping oil Quench gas

Feed Figure 19.25

Two-stage MQD Unionﬁning process for the noble-metal catalyst system.

oil production. High-quality middle distillates are by-products of the isodewaxing process. Sud Chemie-Hydex-G Hydex-G is a sulfur-tolerant zeolite catalyst for dewaxing gasoil and middle distillates [79]. It can be used as a stand-alone catalyst or together with an HDS catalyst in single-stage conﬁguration, with the reactor loaded with stacked beds of DW and HDS catalysts. The range of cloud point improvement that can be expected with Hydex-G in different conﬁgurations is given in Table 19.9. Table 19.9

Cloud point improvement for different hydroprocessing schemes.

Application Stand-alone dewaxing Hydrotreater with dewaxing catalyst Mild hydrocracker with dewaxing catalyst Conventional hydrocracker with dewaxing catalyst Adapted from [79].

617

Cloud point improvement (◦ C) 20–50 5–20 30–50 >5

618

19 Naphtha Reforming and Upgrading of Diesel Fractions

On the basis of data available on the brochures and the published compositional data of naphtha and retention yields, it seems that the catalyst is a ZSM-5-like material without the presence of a metal phase [84].

19.4 Summary and Outlook

Catalytic reforming of naphtha fractions is a mature technology that is still intensively used in modern reﬁneries. Even though there will be only a limited increase in reforming capacity in the coming years (about 1% of current capacity), the process will remain a crucial technology in the oil reﬁning industry. Improvements in the catalytic system and process optimization to maximize overall efﬁciency will be main topics. Current trends in the development of catalytic naphtha reforming are increased hydrogen production, longer cycle length, and increased reformate octane. Recent research and development emphasis is on catalyst and process improvements to maximize catalyst life along with selectivity to hydrogen and BTX aromatics. Other areas of R&D include the following: • reforming of ultralow sulfur, low-water naphtha, which requires special process and metallurgy considerations; • improved radial ﬂow reactor internals allowing higher catalyst loading; • improved regeneration procedures; • revamping of existing units to improve efﬁciency. For example, semiregenerative units, which account for 55% of global reformer capacity, can be improved by adding a new reactor with continuous regeneration. With regard to ‘‘pro-diesel’’ technology, as mentioned above diesel fuel market demand is increasing faster than for any other fuel and in Europe diesel market share has exceeded that for gasoline. Large investments in reﬁneries involve technologies to maximize the production of diesel and processes to improve the quality of diesel fraction or to upgrade low-quality fraction to make them suitable for the diesel pool. New high-pressure hydrocracking units have been recently installed in several reﬁneries, or old hydrocracking plants have been revamped, in order to increase high-quality diesel production. Hydrotreating technologies also play a major role to increase the quality of gasoil fractions in order to satisfy the constant need to produce cleaner transportation fuels. In this context, catalytic dewaxing in conjunction with deep desulfurization and aromatic saturation will play an important role in providing a cost-effective way of producing increasing amounts of high-quality and environmentally friendly transportation fuels. Studies aimed at better understanding shape selectivity in the ﬁeld of catalytic dewaxing are going on, and those concerning dewaxing by isomerization will certainly lead to improvements of performance in terms of selectivity and characteristics of the products. As previously shown, melting point of isomers strictly depends on length and position of the branching. In this regard [45], a possible development

References

in the isomerization dewaxing reaction could be based on the concept of selective isomerization, that is, to guide the reaction toward those speciﬁc compounds that can give the desired characteristics to the ﬁnal products.

References 1. (2006) Global Petroleum Market Outlook, 2.

3.

4. 5. 6.

7. 8. 9. 10. 11.

12.

13. 14.

15.

16.

Purvin & Gertz Inc. Hedrick, B.W., Seibert, K.D., and Crewe, C. (2006) New Approach to Heavy Oil and Bitumen Upgrading, Technology & More, www.UOP.com, Summer 2006. Nielsen, R. (2006) Advances in Catalytic Reforming. Report No. 129B, SRI Consulting, October. Stell, J. (2005) Oil Gas J. 103, 47–54. Marcilly, C. (2001) Oil Gas Sci. Technol.–Rev. IFP, 56 (5), 499–514. Netzer, D. and Ghalayini, O.J. (2003) NPRA Annual Meeting, March, San Antonio. CCR PlatformingTM www.uop.com Swaty, T.E. (2005) Hydrocarbon Process., 84 (9), 35. Benfree, T.M. (2004) Axens Brochure, August 2004. Silvy, R.P. (2004) Oil Gas J., 102 (16), 58. Paal, Z. (2004) Catalytic Naphtha Reforming, 2nd edn, Marcel Dekker Inc., New York, pp. 35–74. Lapinski, M.P., Baird, L., and James, R. (2004) in Handbook of Petroleum Reﬁning Processes, 3rd edn, McGraw-Hill, New York, pp. 4.3–4.31. Menon, P.G. and Paal, Z. (1997) Ind. Eng. Chem. Res., 36, 3282. Perego, C. and Carati, A. (2008) in Zeolites: From Model Materials to Industrial Catalysts, Transworld Research Network ˇ (India) (eds J. Cejka, J. Perez-Pariente, and W.J. Roth), pp. 357–389. Novaro, O., Li, C.-L., and Wang, J.-A. (2004) Catalytic Naphtha Reforming, 2nd edn, Marcel Dekker Inc., New York, pp. 391–431. Antos, G.J., Moser, M.D., and Lapinski, M.P. (2004) Catalytic Naphtha Reforming, 2nd edn, Marcel Dekker Inc., New York, pp. 335–351.

17. Mazzieri, V.A., Grau, J.M., Vera, C.R.,

18.

19.

20. 21.

22.

23. 24. 25. 26. 27.

28. 29. 30.

31.

32.

Yori, J.C., Parera, J.M., and Pieck, C.L. (2005) Appl. Catal., 296, 216. Carvalho, L.S., Pieck, C.L., Rangel, M.C., F`ıgoli, N.S., Grau, J.M., Reyes, P., and Parera, J.M. (2004) Appl. Catal. A: Gen., 269, 91. Carvalho, L.S., Pieck, C.L., Rangel, M.C., F`ıgoli, N.S., Grau, J.M., Vera, C.R., and Parera, J.M. (2004) Appl. Catal. A: Gen., 269, 105. Corma, A. (1993) Catal. Lett., 22, 33. Kao, J.-L. and Ramsey, S.A. (1997) US 5980731, assigned to Exxon Chemical Patents Inc. Sugimoto, M., Murakawa, T., Hirano, T., and Ohashi, H. (1993) Appl. Catal. A:Gen., 95, 257. Bonacci, J.C. and Patterson, J.R. (1981) US 4,292,167, assigned to Mobil Co. Degnan, T.F. (2000) Top. Catal., 13, 349. Llorens, D. (2004) Hydrocarbon Eng., 9 (11), 75. OGJ–International Reﬁning Catalyst Compilation (2005) Oil Gas J., 17. Chen, N.Y., Garwood, W.E., and Dwyer, F.G. (1989) Shape Selective Catalysis in Industrial Applications, Chapter 5, Marcel Dekker, p. 157. NIST http://webbook.nist.gov/chemistry/. Weiz, P.B. and Frilette, V.J. (1960) J. Phys. Chem., 64, 382. Weiz, P.B., Frilette, V.J., Maatman, R.W., and Mower, E.B. (1962) J. Catal., 1, 307. Weitkamp, J., Ernst, S., and Puppe, L. (2001) in Catalysis and Zeolites Fundamentals and Application (eds J. Weitkamp and L. Puppe), Springer, Berlin, pp. 327–376. Marcilly, C. (2003) Catalyse Acido-basique Application au Rafﬁnage et a` la P´etrochimie, Vol. 2, Chapter 12, 13, Editions Technip.

619

620

19 Naphtha Reforming and Upgrading of Diesel Fractions 33. Degnan, T.F. Jr. (2003) J. Catal., 216, 34.

35. 36. 37.

38. 39.

40.

41.

42. 43.

44. 45.

46.

47.

48.

49.

32. Olson, D.H., Kokotailo, G.T., Lawton, S.L., and Meier, W.M. (1981) J. Phys. Chem., 85, 2238. Buchanan, J.S. (2000) Catal. Today, 55, 207. Marcilly, C.R. (2000) Top. Catal., 13, 357. Maesen, T.L.M., Schenk, M., Vlugt, T.J.H., de Jonge, J.P., and Smit, B. (1999) J. Catal., 188, 403. Smit, B. and Maesen, T.L.M. (2008) Nature, 451, 671. Schenk, M., Smit, B., Vlugt, T.J.H., and Maesen, T.L.M. (2001) Angew. Chem. Int. Ed., 40, 736. Jacobs, P.A., Martens, J.A., Weitkamp, J., and Beyer, H.K. (1981) Faraday Discuss. Chem. Soc., 72, 353. Dubbelddam, D., Calero, S., Maesen, T.L.M., and Smit, B. (2003) Angew. Chem. Int. Ed., 42, 3624. Schenk, M., Calero, S., and Maesen, T.L.M. (2003) J. Catal., 214, 88. Souverijns, W., Martens, J.A., Uytterhoeven, L., Froment, G.F., and Jacobs, P.A. (1997) Studies in Surface Science and Catalysis, Vol. 105, Elsevier Science, p. 1285. Sastre, G., Chica, A., and Corma, A. (2000) J. Catal., 195, 227. van Veen, J.A.R., Minderhoud, J.K., Huve, L.G., and Stork, W.H.J. (2008) in Handbook of Heterogeneous Catalysis, Vol. 6 (eds G. Ertl, H. N¨ozinger, F. Sch¨uth, J. Weitkamp), 2nd edn, Wiley-VCH Verlag GmbH & Co., p. 2794. Rigutto, M.S., van Veen, R., and Huve, L. (2007) in Introduction to Zeolite Science and Practice (eds J. Eejka, H. van Bekkum, A. Corma, and F. Sch¨uth), 3rd revised edn, Elsevier, p. 855. Smith, K.W., Starr, W.C., and Chen, N.Y. (2005) Oil Gas J., May 26 (2005), 75–88. Huve, L.G. (2001) 2nd European Catalyst Technology Conference ECTC Antwerp, Belgium. Sivasanker, S., Waghmare, K.J., Reddy, K.M., Kothasthane, A.N., and Ratnasamy, P. (1990) J. Chem. Tech. Biotechnol., 48, 261.

50. Sivasanker, S. and Reddy, K.M. (1989)

Catal. Lett., 3, 49. 51. Kennedy, C.R., LaPierre, R.B., Pereira,

52.

53. 54.

55.

56.

57.

58.

59. 60.

61.

62.

63.

64. 65.

66.

67.

C.J., and Mikovsky, R.J. (1991) Ind. Eng. Chem. Res., 30, 12. Beyer, H.K., Feher, P., and Jakob, K. (1986) Hungarian J. Ind. Chem., 14, 345. Tielen, M., Geelen, M., and Jacobs, P.A. (1985) Acta Phys. Chem., 31, 1. Degnan, T.F., Hanlon, R.T., Mills, G., Karsner, G.G., and Mazzone, D.N. (2001) US 6,231,749, assigned to Exxon-Mobil. Unmuth, E.E., Bertolacini, R.J., and Mahoney, J.A. (1988) US 4,728,415, assigned to Amoco. Zones, S.I., Holtermann, D.L., Jossens, L.W., Santilli, D.S., Rainis, A., and Ziemer, J.N. (1997) US 5,693,215, assigned to Chevron. Soualah, A., Lemberton, J.L., Pinard, L., Charter, M., Magnoux, P., and Molijord, K. (2008) Appl. Catal. A: Gen., 336, 23. La Pierre, R.B., Partrige, R.D., Chen, N.Y., and Wong, S.S. (1983) US 4,419,220, assigned to Mobil. Degnan, T.F. and Le, Q.N. (1994) US 5,302,279, assigned to Mobil. Carati, A., Flego, C., and Calemma, V. (1999) EP 0635556 B1, assigned to Eni S.p.A. Degan, T.F. Jr., Hanlon, R.T., Karsner, G.G., and Mazzone, D.N. (1993) EP 0575077, assigned to Mobil. Perego, C., Bellussi, G., and Calemma, V. (1997) EP EP 0590714, assigned to Eni S.p.A. Perego, C., Flego, C., Delbianco, A., and Bellussi, G. (1996) EP 0582347, assigned to Eni S.p.A. Calemma, V., Peratello, S., and Perego, C. (2000) Appl. Catal. A: Gen., 190, 207. Corma, A., Martinez, A., Pergher, S., Peratello, S., Perego, C., and Bellussi, G. (1997) Appl. Catal. A: Gen., 152, 107. Calemma, V., Peratello, S., Perego, C., Moggi, A., and Giardino, R. (1999) Polym. Prepr. Am. Chem. Soc. Div. Polym. Chem., 44 (3), 241. Ref. 7, Vol. 1, Chapter 4, p. 211.

References 68. Miller, S.J., Shippeyand, M.A., and

69. 70. 71. 72. 73. 74.

75.

76. 77. 78.

79.

80. 81.

82.

83.

Masada, G.M. (1992) NPRA National Fuels and Lubricants Meeting, November 5–6, Houston. Miller, S.J. (1987) US 4,689,138, assigned to Chevron. Miller, S.J. (1989) US 4,859, 311, assigned to Chevron. Miller, S.J. (1990) US 4,921,594, assigned to Chevron. Miller, S.J. (1992) US 5,135,638, assigned to Chevron. Miller, S.J. (1992) US 5,149,421, assigned to Chevron. Santilli, D., Habib, M.M., Harris, T.V., and Zones, S.I. (1994) US 5,282,958, assigned to Chevron. Miller, S.J. (1994) in Studies in Surface Science and Catalysis, Vol. 84 (eds J. Weitkamp, H.G. Karge, H. Pfeifer, and W. Holderich), p. 2319. Miller, S.J. (1993) Am. Chem. Soc. Div. Petr. Chem., 38 (4), 788. Miller, S.J. (1994) Microporous Mater., 2, 439. Martens, J.A., Souverijns, W., Verrelst, W., Parton, R., Froment, G.F., and Jacobs, P.A. (1995) Angew. Chem. Int. Ed. Engl., 34 (22), 2528. (2004) Advanced Hydrotreating and Hydrocracking Technologies to Produce Ultra-Clean Diesel Fuel, Chapter VII-233, Hydrocarbon Publishing Company. Zakarian, J.A., Robson, R.J., and Farrel, T.R. (1987) Energy Process, 1, 59. Wilson, M.W., Eiden, K.L., Mueller, T.A., Case, S.D., and Kraft, G.W. (1994) NPRA National Fuel and Lubricants Meeting, November 1994. Homan Free, H.W., Schockaert, T., and Sonnemans, J.W.M. (1993) Fuel Proc. Technol., 35, 111. Heckel, T., Thakkar, V., Behraz, E., Brierley, G., and Simpson, S. (1998) NPRA, Annual Meeting Paper AM-98-24.

84. Weyda, H. and Koler, E. (2002) 12th

85.

86.

87.

88. 89.

90.

91.

92.

93.

94. 95.

Saudi Arabia–Japan Symposium on Catalyst in petroleum Reﬁning and Petrochemicals, KFUPM Institute, Dharan, Saudi Arabia. Huve, L.G., Robertson, M., van der Linde, B., Pankratov, L., Kalospiros, N., and Gitau, M. (2006) RRTC, 27-28 September, 2006, Moscow. Kamienski, P.W., Hilbert, T.H., Novak, W.J., and Lewis, W.E. (2006) World Reﬁning Association CE/EE 9nd Annual Round Table, Budapest, Hungary. van der Linde, B., Woolley, H., Swain, J., and High Allen, L.A. (2001) SynFlow/Syn Shift–Commercial Experience on European Diesel, European Reﬁning Technology Conference, Madrid. Smith, F.A. and Bortz, R.W. (1990) Oil Gas J., Aug 13 (1990), 51. Tracy, W.J., Chitnis, G., Novak, W., Helton, T., Macris, A., Papal, D., and Nagel U. (2002) European Catalysis Technology Conference, Amsterdam. Mc Gihon, R.D., Hilbert, T.L., Patel, V., and Subramaniam, A. (2008) 9th Annual Middle East Reﬁning Conference, Abu Dhabi, February 2008 (ExxonMobil website). Hilbert, T.L., Chitnis, G.K., Umansky, B.S., Kamienski, P.W., Patel, V., and Subramanian, A. (2008) Hydrocarbon Process., 47–56. Gala, H. (2004) in Handbook of Petroleum Reﬁning Processes 3rd edn, Chapter 8.5 (ed. R.A. Meyers), Mc Graw-Hill, p. 53. Bertram, R. et al. (1994) Catalytic Dewaxing of VGO an Diesel at OMV’s Schwechat Reﬁnery, San Antonio, NPRA Papers AM-94-50. http://www.uop.com/reﬁning/1060.html. http//www.chevron.com/products/sitelets/ reﬁningtechnology.

621

623

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites ˇ Sulaiman Al-Khattaf, Mohammad Ashraf Ali, and Jiˇrı´ Cejka

20.1 Introduction

Aromatic hydrocarbons are extremely important starting materials for the production of numerous important products in chemical, polymer, agriculture, perfume, civil engineering, and other industries [1]. Their chemistry over zeolites is principally based on three main reactions, namely, alkylation, isomerization, and disproportionation (transalkylation) [2–5]. To carry them out effectively, zeolites as catalysts are the ﬁrst choice because of their high activity, extreme selectivity to the desired products, restricted deactivation, regenerability, and environmental tolerance [6–10]. Among the reactions dealing with transformations of aromatic hydrocarbons over zeolites, most of the recent papers have focused on toluene alkylation with methanol to xylenes including different types of zeolites [11], modiﬁcation of zeolite parameters [12–15], or process variables aimed at obtaining of the maximum para-selectivity [16]. In principle, the objectives of most of these studies can be divided as follows: 1) optimization of the catalyst formulation including different approaches of zeolite modiﬁcation; 2) improvement of process parameters, type of reactor, and modeling of kinetic parameters; 3) investigation of novel zeolites with respect to their structural and chemical parameters.

20.2 Zeolites under Study

The main advantage of zeolites related to transformations of aromatic hydrocarbons stems from the fast that they are microporous crystalline aluminosilicates with pore sizes from 0.4 to about 1.0 nm and controlled acidity [17, 18]. Zeolite structure is formed from the so-called primary building units, tetrahedra of SiO4 or AlO4 , Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

624

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites

which are connected with one oxygen bridge [17]. While up to four SiO4 can be connected to the central SiO4 tetrahedron, direct connection of AlO4 tetrahedra is not allowed as a result of the Loewenstein rule. Further organization of SiO4 and AlO4 2 tetrahedra leads to the formation of secondary building blocks, and through their connections pore systems of different sizes and shapes are formed. These pores are most frequently built from 8, 10, 12, and 14 rings, while even some odd rings or larger rings than 14 have been described [19]. Zeolite channels can be one-dimensional (1D), two-dimensional (2D), or three-dimensional (3D) depending on their connectivity [18]. In some zeolites, pores of different channel sizes or even cavities at channel intersections or in 1D channels are present [19]. The size of the zeolite channels plays a decisive role in their application in adsorption and catalysis. The size of zeolite channels is the same as the kinetic diameters of many organic reactants, products, or transition states. Thus, when those are larger than available space in the zeolite channels, reactants cannot penetrate into the zeolite system, transition states cannot be formed inside, and products (if already formed) cannot leave the channels. This phenomenon is called shape selectivity and there are numerous papers discussing it particularly in relation to transformation of aromatic hydrocarbons [20–22]. The individual types of shape selectivity can signiﬁcantly control the activity and also selectivity of zeolites in different reactions [23, 24]. Although around 200 zeolites and zeotypes have been recognized by IZA, only a few of them have been investigated in the transformation of aromatic hydrocarbons. Among them, ZSM-5, mordenite, Beta, zeolite Y, and MCM-22 represent the most typical zeolite catalysts, while novel zeolites like MCM-58, MCM-68 [25, 26], TNU-9 [27], and IM-5 [28] were also recently synthesized and are tested. Figures 20.1 and 20.2

(a)

(b)

(c)

(d)

Figure 20.1 Structural motifs of zeolites mostly investigated in aromatic transformations: ZSM-5 (a), Beta (b), Y (c), and mordenite (d).

20.3 Toluene Disproportionation Structural characteristics of zeolites studied in transformation of aromatic hydrocarbons.

Table 20.1

Catalyst

Code

Channel structure

Channel entrances

Channel dimension (nm)

Beta

BEA

3D

12-12-12

IM-5

IMF

3D

10-10-10

Mordenite

MOR

2D

12-8

MCM-22

MWW

3D

10-10-10

MCM-58 MCM-68

IFR MSE

1D 3D

12 12-10-10

SSZ-33

CON

3D

12-12-10

SSZ-35 TNU-9

STF TUN

1D 3D

10 10-10-10

ZSM-5

MFI

3D

10-10-10

0.64 × 0.76 0.56 × 0.56 0.55 × 0.56 0.53 × 0.54 0.53 × 0.59 0.65 × 0.70 0.26 × 0.57 0.40 × 0.55 0.41 × 0.51 0.62 × 0.72 0.64 × 0.68 0.52 × 0.58 0.52 × 0.52 0.64 × 0.70 0.59 × 0.70 0.45 × 0.51 0.54 × 0.57 0.52 × 0.60 0.51 × 0.55 0.53 × 0.56 0.51 × 0.55

provide schematic pictures of some of these zeolite structures, and their structural properties are collected in Table 20.1. From the catalytic point of view, in addition to structural features, the chemical composition is of crucial importance. The concentration of AlO4 tetrahedra determines the overall negative charge of the zeolite framework, which in the case of compensation by protons provides a solid acid. The activity of zeolites in acid-catalyzed reactions depends both on the concentration and the type of acid sites (Br¨onsted and Lewis types) and the acid strength [29].

20.3 Toluene Disproportionation

Toluene disproportionation is an industrial process for upgrading less valuable toluene to a mixture of xylenes and benzene, in which the highest selectivity to p-xylene is desired. Xylenes are one of the important petrochemical feedstocks for producing a number of valuable commodities (o-xylene for phthalic anhydride and phthalate plasticizers, m-xylene for isophthalic acid and unsaturated polyester

625

626

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites

resins, and the most desired p-xylene for terephthalic acid and polyethylene terephthalate or polybutylene terephthalate). In general, when toluene disproportionation is carried out in absence of shape-selective catalysts, the following selectivities to individual xylene isomers are obtained: p-xylene ∼25%, o-xylene ∼25%, and m-xylene ∼50%. This clearly demonstrates why technological processes for increasing the selectivity to p-xylene at the expense of m-xylene are so important. Toluene disproportionation was mostly performed over the acid zeolites ZSM-5, MCM-22, mordenite, and USY, while the selective formation of p-xylene is usually connected with ZSM-5 zeolite, particularly after its surface modiﬁcation. Selective formation of p-xylene is attributed to decreasing pore size of medium pore zeolites together with a selective blocking of active sites on the external surface of zeolite crystals [8]. Mostly, both these factors are applied simultaneously, as it is not possible to annihilate surface sites with at least a partial blocking of pores and vice versa. One of the methods to control the diffusivity of xylenes inside the zeolite structure is the use of the ZSM-5 zeolite with large crystal sizes which produces a higher amount of p-xylene than those with small crystal sizes [30, 31]. Based on the mechanism of toluene disproportionation, it is strongly believed that p-xylene formation takes place inside the zeolite pores. Then p-xylene diffuses from the pores of the zeolite, resulting in initial high p-xylene selectivity [32]. However, on the external surface of the zeolites, p-xylene undergoes isomerization to form meta and ortho isomers. To limit this secondary isomerization, a number of methods, including silicon deposition, precoking of the catalyst, selective dealumination of the external active sites, and loading of metal oxide on the zeolite, have been proposed [8, 12, 13, 16]. 20.3.1 Zeolite Modiﬁcation by Silicon Deposition

The external surface of zeolites can be modiﬁed by a variety of silicon-containing compounds by chemical vapor deposition or chemical liquid deposition. Silicon species (larger than size of the channels) react with bridging and probably also with terminal hydroxyl groups on the external surface of the zeolite. During calcination, silicon oxide particles attached to the zeolite surface are formed resulting in an annihilation of acid sites on the external surface and a partial or complete blocking of the pore entrances. In contrast, the zeolite void volume is preserved [33]. Modiﬁcation of external surfaces and pore mouth regions of HZSM-5 samples with different crystal sizes has been achieved by chemical liquid deposition, with Si(OEt)4 passivating the unselective acid sites [34]. The modiﬁcation was found to be more effective for zeolite samples with larger crystal sizes. Previous dealumination of the external surface of ZSM-5 crystals enhanced the silylation effects by primarily removing acid sites associated with the extra-framework alumina species. Organosilicon compounds such as MePhSiOH, EtPhSiNH2 , MeNH2 SiOH, and MePhSiNH2 have been used for surface coating to improve p-xylene selectivity [35, 36]. Chemical vapor deposition provided by Niwa’s group [13] showed the importance of in situ and ex situ preparation conditions

20.3 Toluene Disproportionation

using tetramethyl orthosilicate. It was shown that when tetramethyl orthosilicate was added in situ, much better coverage of the external surface was achieved, as evidenced by the 1,3,5-triisopropyl benzene. 20.3.2 Zeolite Modiﬁcation by Precoking

Precoking of the zeolite-based catalyst is frequently discussed and investigated with the purpose of increasing the para selectivity. Despite a number of very careful studies, the understanding of this effect is still far from deﬁnite and many results are rather controversial. For toluene disproportionation, some enhancement in ˇ p-xylene selectivity was achieved by Haag and coworkers [37]. In contrast, Cejka et al. reported that coking did not increase the selectivity to p-xylene in toluene alkylation with methanol [38]. On the other hand, precoking has been successfully carried out using 1,3,5-triisopropylbenzene on ZSM-5 catalyst prior to the reaction. Deactivation of the external surface resulted not only in a signiﬁcant increase in the p-xylene selectivity but simultaneously also in a decrease in the formation of higher aromatics such as trimethybenzenes [39]. During toluene disproportionation, the rate of coke formation on ZSM-5 decreases with reaction time, while the actual amount of coke formed continuously increases. Higher temperature and lower Si/Al ratio enhanced the rate of coke formation. In the early stages of the reaction, coke formation was found advantageous to toluene disproportionation. When coke was built up, toluene conversion and the yield of xylene fell remarkably [40]. It was proposed that the number of strong Br¨onsted acid sites decreased rapidly during the initial coking period, reaching a plateau of about 7 wt% of coke with a corresponding 80% decrease in total acidity [40]. 20.3.3 Zeolite Modiﬁcation by Dealumination

Two types of aluminum species can substantially deteriorate the p-xylene selectivity in toluene disproportionation over ZSM-5 zeolites. The presence of extra-framework alumina can enhance p-xylene selectivity because of blockage of zeolite pores; however, this is usually connected with coke formation. On the other hand, aluminum on the external surface of zeolite crystals forming both Br¨onsted and Lewis acid sites enhances the rate of consecutive (nonselective) reactions. Dealumination of MCM-22 selectively removes the acid sites from the external surface and suppresses the secondary isomerization of p-xylene, enhancing the para selectivity. This shows that dealuminated MCM-22 can be a promising catalyst for the selective formation of p-xylene from toluene disproportionation [41]. The MCM-22 zeolite having small crystallites has shown higher toluene disproportionation activity, whereas zeolites with large crystallites were more selective for p-xylene formation with a beneﬁcial effect of boron [42, 43]. Using Y-zeolite, high reaction temperatures and high acidity of catalysts favored dealkylation, while low temperature and medium acidity favored toluene disproportionation [44]. The TNU-9 zeolite, with the recently announced

627

628

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites

3D medium pore channel structure, exhibited a unique shape selectivity in toluene disproportionation reaction as compared with ZSM-5, MCM-22, mordenite, and Beta zeolites [45]. 20.3.4 Zeolite Modiﬁcation by Metal Deposition

Toluene conversion by disproportionation can be improved using zeolites loaded with metal in a deﬁned reduced state to provide Lewis acidity [46]. A novel catalyst comprised of H-ZSM 5 zeolite treated with CeNO3 and calcined at 500 ◦ C possessed an enhanced p-xylene selectivity of 42.7% and conversion of 12.4%. [47]. Toluene disproportionation over Ni-, Cr-, Mg-, Bi-, and Zn-ion-exchanged ZSM-5 zeolite catalysts exhibited an increase in toluene conversion as well as higher p-xylene selectivity especially for Ni-exchanged catalyst [48]. At 550 ◦ C, Ni-exchanged catalyst gave toluene conversion of 28.5% and the p-xylene selectivity of 35%. Coke deposits formed during the reaction are responsible for pore-mouth reduction allowing the passage of the small p-xylene molecules mainly. Multiple zeolite-based catalysts have been used for disproportionation of toluene to improve the activity and selectivity to p-xylene [49]. Nickel-loaded, dealuminated, mordenite-based catalyst with Si/Al ratio between 10 and 50 exhibited toluene conversion 46% at LHSV 2 hour−1 at 400 ◦ C [50]. It is seen that added metals or metal oxides, when located on the external surface, enhance the diffusional constraints of HZSM-5 for increasing p-xylene selectivity. In a similar way, antimony oxide was deposited on the external surface of the HZSM-5 crystals by solid-state reaction at 500 ◦ C. This modiﬁcation resulted in a strong interaction with silanol and bridging hydroxyl groups of the zeolite. The dispersion of antimony oxide on the surface of HZSM-5 completely removed unselective Br¨onsted acid sites, inducing enhanced para selectivity for toluene disproportionation [51]. Molybdenum oxide impregnated on H-ZSM-5 zeolite was thoroughly investigated in toluene disproportionation to provide a marked difference in toluene conversion and xylenes selectivity using nitrogen and hydrogen carrier gases and at variable reaction pressures [52]. Some of results obtained are given in Tables 20.2 and 20.3 Initial toluene conversion under nitrogen pressure was found to be higher than under hydrogen pressure. Toluene conversion also increased with increase in the pressure and reaction temperature for both carrier gases. Toluene conversion under hydrogen carrier gas at 375 ◦ C was found to be lower as compared with nitrogen. With increasing reaction temperature, the differences in toluene conversions were reduced and reached similar values at 425 ◦ C for both nitrogen and hydrogen carrier gases. Toluene conversion under nitrogen carrier gas was higher than under hydrogen. Toluene conversion increased with increasing pressure from 1 to 30 bars under both nitrogen and hydrogen carrier gases. There was very little change in toluene conversion while increasing the reaction pressure from 30 to 50 bars. As expected, ZSM-5 zeolite was deactivated much faster under nitrogen compared to hydrogen carrier gas as indicated from the decrease in toluene conversion as well as by the higher amount of coke formed.

20.3 Toluene Disproportionation Results (mol%) for 3 wt% molybdenum loaded ZSM-5(13)a catalyst in the temperature range 375–425 ◦ C and under hydrogen and nitrogen carrier gases at 30 bar pressure.

Table 20.2

Temperature (◦ C)

Hydrogen carrier gas 375 400 425

Nitrogen carrier gas 375 400 425

Benzene Toluene Ethylbenzene m-Xylene p-Xylene o-Xylene C9 + aromatics Total xylenes Xylene selectivity m-Xylen selectivity p-Xylen selectivity o-Xylen selectivity Benzene selectivity Toluene conversion

20.2 61.8 0.0 8.8 4.4 4.2 0.6 17.4 45.5 23.0 11.5 11.0 52.9 38.2

21.4 56.7 0.0 11.2 5.4 5.1 0.2 21.7 50.1 25.9 12.5 11.8 49.4 43.3

a ZSM-5(13)

27.7 47.0 0.3 11.9 6 5.8 1.3 23.7 44.7 22.5 11.3 10.9 52.3 53.0

28.3 44.0 0.4 12.5 6.3 6.2 2.3 25 44.6 22.3 11.3 11.1 50.5 56.0

29.6 48.9 0.1 11.2 4.7 4.8 0.7 20.7 40.5 21.9 9.2 9.4 57.9 51.1

31.1 47.6 0.2 10.3 4.3 4.5 2.0 19.1 36.5 19.7 8.2 8.6 59.4 52.4

has Si/Al molar ratio of 13.5 and surface area of 600 m2 /g.

20.3.5 Factors Affecting Toluene Disproportionation

On the basis of the numerous studies on toluene disproportionation, the major factors inﬂuencing the activity and selectivity of the zeolite catalysts in toluene disproportionation can be summarized as follows: • Toluene conversion increases with increasing concentration of acid sites (lowering Si/Al ratio) and increasing reaction temperature. The increase in the reaction temperature results also in a higher rate of toluene dealkylation and coke formation. • Presence of extra-framework alumina and loading of metal oxides of Si, P, Mg, Ce, Sb, and B decrease the channel size of ZSM-5; as a result selectivity to p-xylene improves. • Application of ZSM-5 zeolites with large crystal sizes resulted in a higher p-xylene selectivity owing to the elongation of diffusion pathways, whereas small crystals showed a higher activity as a result of much higher concentration of acid sites on the external surface. • p-Xylene is formed inside the zeolite channels and undergoes isomerization to form other isomers on external surface of zeolites. Therefore, deactivation of the external active sites improves p-xylene selectivity.

629

630

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites Table 20.3 Catalytic reaction results for 3 wt% molybdenum loaded ZSM-5(15)a for 40 hours on stream at 400 ◦ C, 1.5 LHSV and 30 bar pressure (mol%).

Hydrogen carrier gas Hours on stream Light hydrocarbons Benzene Toluene Ethylbenzene m-Xylene p-Xylene o-Xylene C9 + aromatics Total xylenes Benzene selectivity Xylene selectivity m-Xylene selectivity p-Xylene selectivity Toluene conversion a ZSM-5(15)

6 1.7 22.4 46.6 1.1 13.5 4.7 5.4 4.5 23.6 42.0 44.3 25.3 8.7 53.4

16 2.6 25.1 47.3 1.0 11.6 4.8 4.8 2.9 21.2 47.6 40.1 22.0 9.0 52.7

22 2.4 29.4 46.9 0.9 9.8 4.2 4.0 2.4 18.0 55.4 33.9 18.5 7.9 53.1

Nitrogen carrier gas 30 0.0 31.6 48.6 0.8 9.3 4.0 3.8 1.9 17.2 61.4 33.4 18.1 7.8 51.4

40 1.7 26.6 47.7 0.9 11.0 4.5 4.5 3.1 19.9 51.0 38.2 21.0 8.6 52.3

6 0.3 26.3 51.3 1.6 9.6 4.2 4 2.8 17.8 54.0 36.6 19.7 8.6 48.7

16 2.0 18.1 60.3 1.5 7.8 3.3 3.2 3.8 14.3 45.6 36.0 19.6 8.3 39.7

22 2.6 15.1 64.9 1.3 6.6 2.8 2.7 4.0 12.1 43.0 34.5 18.8 8.0 35.1

30 0.9 11.9 72.8 1 4.9 2.3 2.1 4.1 9.3 43.8 34.2 18.0 8.5 27.2

40 4.1 8.7 77.4 0.7 3.1 1.4 1.3 3.4 5.8 38.5 25.7 13.7 6.2 22.6

has Si/Al molar ratio of 15 and surface area of 400 m2 /g.

• Selective dealumination of MCM-22 and other zeolites removing external acid sites resulted in enhancement of para selectivity. It is a similar effect to the annihilation of these sites by various metal oxides. • Early stage of coke formation is advantageous for toluene disproportionation; however, when coke builds up, toluene conversion and xylene yield drop. • A higher p-xylene selectivity but lower toluene conversion was achieved over ZSM-5 modiﬁed with polyalkylsiloxanes due to the deactivation of the active sites external to the zeolite pores. • Higher initial toluene conversions were reached over ZSM-5 when nitrogen was used as carrier gas as compared to hydrogen. In contrast, catalyst deactivation was more rapid in nitrogen.

20.4 Ethylbenzene Disproportionation

Disproportionation of ethylbenzene reaction is similar to toluene disproportionation but ethylbenzene is more reactive, and by dealkylation of ethylbenzene ethylene can be easily formed. The products of ethylbenzene disproportionation proceeding via a transfer of ethyl group from one ethylbenzene molecule to another are a mixture of diethylbenzene (DEB) isomers and benzene. Depending on the reaction conditions, some undesired reactions such as dealkylation producing gases or

20.4 Ethylbenzene Disproportionation

further transalkylation to higher aromatics can proceed. p-DEB is used as a striping agent in p-xylene production units. Disproportionation of ethylbenzene can be successfully achieved over the acidic forms of different structural types of zeolites including ZSM-5, mordenite, beta, Y, MCM-22, MCM-40, and their modiﬁed and impregnated forms [53–56]. 20.4.1 Effect of Crystal Size and Surface Modiﬁcation

Like other disproportionation reactions, ethylbenzene conversion and DEB product distribution during disproportionation of ethylbenzene are strongly dependent on the size of the crystals related to external acidity [57]. In their study, Melson and Sch¨uth attempted to correlate the selectivity to p-DEB to the concentration of acid sites on the external surface of ZSM-5 crystals possessing crystal sizes from 0.1 up to 80 µm. In general, larger crystals exhibited much lower external acidity and only p-DEB was formed. When the size of the crystals was decreased with simultaneous increase in the external surface acidity, then m-DEB and o-DEB were also formed. These data are in a good agreement with results on toluene disproportionation showing that the para isomers are almost exclusively formed inside the zeolite channels, whereas the undesired isomerization proceeds on the unrestricted acid sites located on the external surface. H-ZSM-5 zeolite used for EB disproportionation reaction showed that an increase in the reaction temperature resulted in the formation of coke residues consisting of condensed polycyclic aromatic compounds. On the other hand, a change in space velocity and H2 /EB molar ratio did not lead to any notable change in the composition of the coke. However, increases in the reaction temperature and space velocity or decrease in the H2 /EB ratio tended to enhance para selectivity using the H-ZSM-5 zeolite. The induced shape selectivity is linked to the preferential deposition of the coke on the external surfaces of the H-ZSM-5 crystallites. Thus, the use of coke as a modifying agent for selectivation of ZSM-5 zeolite in an EB disproportionation reaction was clearly demonstrated [58]. ZSM-5 zeolites were modiﬁed to increase the shape-selective disproportionation of ethylbenzene to p-DEB by chemical liquid deposition and chemical vapor deposition of silica and MgO loading [59]. The results showed that ZSM-5 modiﬁed via chemical liquid deposition performed much better than ZSM-5 modiﬁed with chemical vapor deposition. MgO modiﬁcation improves para-DEB selectivity up to 98.6% at ethylbenzene conversion of 28.1%. The modiﬁcation with MgO reduced the rate of dealkylation reaction signiﬁcantly [59]. 20.4.2 Kinetic Investigations of Ethylbenzene Disproportionation

Experimental and computational approaches were used to understand the shape-selective behavior of HZSM-5 zeolites in ethylbenzene disproportionation. It was observed that geometric constraints imposed on the product molecules

631

632

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites

inside the pores of HZSM-5 zeolite favored the formation of p-DEB. Klemm and coworkers [60, 61] concluded that under the chosen experimental conditions the para isomer is exclusively formed inside the pore structure and subsequent isomerization occurs on the external surface only. These results are in a good agreement with conclusions of Melson and Sch¨uth [57]. Sharanappa and coworkers [62] reported disproportionation of EB in the presence of m- and p-xylene over pore-size-regulated HZSM-5 catalyst. It was observed that irrespective of the different feed compositions the concentration of the xylene isomers was intact in the product. This indicates that there is no interchange of alkyl groups between xylenes and DEBs. Kinetic investigation of ethylbenzene disproportionation over large pore zeolites at low reaction temperatures evidenced that the reaction rate was strongly but reversibly retarded by the formation of the product DEB. In contrast, no product inhibition occurred over medium pore ZSM-5 [63]. In agreement with results discussed above, para selectivity in the DEB increases substantially with increasing crystal size and also with increasing reaction temperature, yielding 100% of p-DEB. These results strongly suggest diffusion control of the product selectivity [63]. The differences in product inhibition between large pore and medium pore zeolites were explained in terms of differences between the adsorption constant of the inhibiting product and that of the feed. They are expected to be much more pronounced in large than in medium pore zeolites [64]. A stronger accumulation of the DEB isomers is proposed in faujasite than in H-ZSM-5. Sorption measurements, carried out by means of in situ infrared spectroscopy, with parent ethylbenzene and DEB isomers in H-ZSM-5 and Y-zeolites conﬁrmed the assumption that with decreasing pore size the product is not sorbed so strongly in the zeolite channels as the feed. In contrast, the results of Klemm et al. [60, 61] on the FTIR investigations provided evidence that DEB isomerization occurs in the interior of H-ZSM-5. Thus, the para-selective features of H-ZSM-5 reﬂect the interplay of catalytic reaction and mass transfer phenomena. Additionally, it was shown that p-DEB and m-DEB are reversibly sorbed in H-ZSM-5 in contrast to the ortho isomer, which cannot enter the zeolite pore system in the temperature range of 150–250 ◦ C examined [65]. One of the factors inﬂuencing the length of the prestationary period during disproportionation of ethylbenzene is the establishment of the sorption equilibrium of the reactants. This is strongly inﬂuenced by the pore structure of the zeolite and its catalytic activity and by the reaction conditions, that is, the reaction temperature, the feed rate, and the concentration [63]. Based on the detailed investigations of ethylbenzene disproportionation, the major factors inﬂuencing zeolite activity catalysts and selectivity to the reaction products are as follows: • Ethylbenzene conversion strongly depends on the size of zeolite crystals. The smaller the crystal size, the higher ethylbenzene conversion. • The para isomers are exclusively formed inside the pore structure of medium pore zeolites, and the subsequent isomerization to meta and ortho isomers occurs on the external surface of the zeolite.

20.5 Disproportionation and Transalkylation of Trimethylbenzene

• Ethylbenzene conversion and product distribution in ethylbenzene disproportionation are strongly dependent on the external acidity which in turn correlates well with the particle size. • Dealumination of mordenite increases the rate of the reaction and improves the life time of the catalysts for EB disproportionation. • Modiﬁcation of ZSM-5 by silica deposition or MgO on the zeolite external surface increased the p-DEB selectivity. A higher selectivity was achieved by chemical liquid deposition of silica compared with chemical vapor deposition. • The use of coke as a modifying agent for selectivation of H-ZSM-5 zeolite in ethylbenzene disproportionation reaction increased the para selectivity as a result of the preferential deposition of the coke on the external surfaces of the H-ZSM-5 crystallites.

20.5 Disproportionation and Transalkylation of Trimethylbenzene

Transformation of trimethylbenzenes via transalkylation with toluene or disproportionation reactions represents the most important ways to upgrade the low-cost C9 fraction consisting mainly of all trimethylbenzenes and ethyltoluenes. In both reactions, the desired products are xylenes; para and ortho isomers being the preferred ones. Major commercial transalkylation processes are TransPlus developed by Mobil-CPC, Xylene-PlusSM process developed by ARCO-IFP, and TatoraySM process developed by UOP. Transalkylation processes are very valuable for reﬁneries having naphtha-cracking units where C9+ aromatics are produced. As a feed, mainly trimethylbenzene isomers are used, being in thermodynamic equilibrium, and the fraction of 1,2,4-trimethylbenzene in total trimethylbenzenes is around 60% [66]. In contrast to toluene and ethylbenzene disproportionation employing ZSM-5 zeolites, trimethylbenzene transformations proceed over large-pore zeolites as the catalysts of choice. The main reason is that the pore size of medium-pore zeolites is too small for penetration of trimethylbenzene isomers and the reactions can proceed only on their external surface. This has been demonstrated for NU-87, which can accommodate only 1,2,4-trimethylbenzene having smaller kinetic diameter as compared to 1,3,5- and 1,2,3-trimethylbenzenes [67]. One of the most important issues to improve the transalkylation process is the long-term stability of the zeolite catalyst used. USY-zeolite has been reported to have a reasonable initial conversion, but deactivation due to coke formation proceeded rather fast. The lifetime of the catalysts has been increased by metal incorporation. It was observed that by loading 1.4% of nickel to the USY catalyst, catalyst stability was enhanced signiﬁcantly [66, 68]. Modiﬁcation of zeolite Y with Pt/La led to the yield of xylenes in trimethylbenzene disproportionation in the range of 17–35% at the reaction temperature of 400 ◦ C [69]. The authors showed that the conversion decreases in the order 1,2,4-trimethylbenzene > 1,3,5-trimethylbenzene ˇ >1,2,3-trimethylbenzene. This is in contrast to the observation of Cejka et al. [70] showing practically the same conversions of 1,2,4- and 1,3,5-trimethylbenzenes

633

634

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites

under the same reaction conditions. In this report, a fast isomerization of starting trimethylbenzenes into a mixture that is close in composition to the thermodynamic one is expected. Thus, all isomers are the real feed. Mordenite, Y, and beta zeolites possess pore sizes large enough for transformation of trimethylbenzenes. In the case of transalkylation of trimethylbenzenes with toluene, zeolite beta exhibited excellent stability and transalkylation selectivity. As compared with medium pore zeolites, a higher yield of xylenes can be obtained using zeolite beta at 400 ◦ C using 1 : 1 toluene to trimethylbenzene molar ratio in the feed [71, 72]. In trimethylbenzene disproportionation, blending toluene into feed stocks can shift the product selectivity of tetramethylbenzene (TeMB) to xylene, whereas in toluene disproportionation, mixing trimethylbenzene into the feedstock not only reduced the reaction temperatures but also raised the xylene yield. Adsorption studies of 1,2,4- and 1,3,5-trimethylbenzenes and 1,2,3,5-tetramethylbenzene over zeolites beta, Y, and mordenite showed substantial differences in the accessibility of acid sites of individual zeolites. The accessibility of Br¨onsted acid sites decreases in the sequence Y > beta > mordenite. The diffusion coefﬁcient of 1,2,4-trimethylbenzene was found to be two times higher than those of 1,3,5-trimethylbenzene and 1,23,5-tetramethylbenzene despite the structural type of the zeolite [70]. It was proposed that transalkylation and disproportionation reactions occur on stronger acid sites, while the isomerization of xylenes and trimethylbenzenes predominates on weaker acid sites [73]. The concentration of active sites and the zeolite structure have most signiﬁcant impact on the activity and selectivity in trimethylbenzene transalkylation with toluene [74]. Addition of toluene to trimethylbenzene signiﬁcantly reduces the rate of trimethylbenzene dealkylation. Furthermore, a toluene/TMB molar ratio in the feed equal to 1 seems to be the optimum for xylene production [71]. Zeolites H-beta, mordenite, and H-ZSM-5 loaded with up to 6 wt% of Mo, Pd, or Ni demonstrated a high transalkylation activity for C9+ aromatics blended with toluene to produce benzene and a mixture of xylenes [74–76]. Addition of a small amount of Pd or Ge (0.2 wt%) in the H-mordenite zeolite resulted in a reduction in aromatic ring saturation and formation of light products [77, 78]. Introduction of tin and germanium along with rhenium also suppressed undesirable reactions such as aromatic ring saturation and formation of light products [79, 80]. Recent investigations using the new zeolites SSZ-74 and SSZ-75 (STI topology) synthesized using tetramethylene-1,4-bis-(N-methylpyrrolidinium) di-quarternary cations as structure directing agents have demonstrated a high activity in transalkylation of aromatic hydrocarbons [81, 83]. Major factors inﬂuencing the toluene disproportionation reaction of the catalysts are given below. • Large pore zeolites (Mordenite, USY, Beta) are more suitable for transalkylation than medium pore zeolites. • Generally, the highest activity for transalkylation of trimethylbenzenes with toluene is at 400 ◦ C.

20.6 Alkylation of Aromatics

• Transalkylation reaction is thermodynamically controlled, and therefore increase in trimethylbenzene concentration in the feed enhances xylene yield. • Transalkylation reaction preferentially occurs on stronger acid sites, whereas isomerization predominates on weaker acid sites. • Transalkylation reaction under H2 atmosphere minimizes coke formation; also, addition of Pt, Rh, or Sn decreases the rate of deactivation. 20.6 Alkylation of Aromatics

Alkylations of benzene, toluene, and ethylbenzene are large industrial processes leading to alkyl and di-alkyl aromatic hydrocarbons which are important intermediates for the preparation of phenol and various intermediates for polymer production. While alkylations of benzene with ethylene or propylene are reactions providing mono-alkylbenzenes, alkylation of toluene and ethylbenzenes leads to dialkylbenzene products, in which particularly para isomers are the desired products [3, 8, 9]. In this connection, large pore zeolites are preferred for benzene alkylation as the size of their channels does not limit substantially the diffusion of reactants and products. In contrast, medium pore zeolites are chosen when para isomers (p-xylene, p-ethyltoluene, p-DEB) are to be obtained with high selectivities. There is no doubt that zeolites are the most important industrial catalysts for alkylation of aromatic hydrocarbons; however, a number of industrial processes still employ Lewis acids such as AlCl3 [9]. The main disadvantage of Lewis acid catalysts in alkylation reactions stems from the contamination of the alkyl aromatic products, corrosion problems in the operation units, and disposal of the spent catalyst [3]. From the beginning of the investigation of zeolites in aromatic alkylations, ZSM-5 zeolite gained considerable attention [84]. Shape selectivity enhancing the formation of desired para isomers, long-term stability of conversion, and low coke formation are the main advantages of this zeolite [8]. Haag and coworkers [85] suggested that the activity of ZSM-5 zeolites is proportional to the concentration of their Br¨onsted acid sites. Although ZSM-5 zeolites show high selectivity for the para isomers of di-alkylbenzenes, a substantial amount of this isomer which is selectively formed in the pores of the zeolites is isomerized on the external surface of the zeolite crystals to the less favorable meta and ortho isomers. As a result, numerous ways to modify ZSM-5 zeolite have been studied. For example, enlargement of the crystal size [86], impregnation with phosphorus and boron [87], and a thermal treatment at high temperatures have been proposed to enhance the selectivity of ZSM-5-based catalyst used in these processes. 20.6.1 Ethylation of Benzene

Ethylbenzene is an important raw material in the petrochemical industry for the manufacture of styrene with a worldwide capacity of production of about 23

635

636

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites

million metric tons per year [88]. Benzene alkylation with ethylene is an electrophilic substitution reaction proceeding on the aromatic ring. Most of the current processes of ethylation of benzene employ ethylene as the alkylating agent. Direct use of ethanol, instead of ethylene, as the alkylating agent for this reaction can be rather advantageous in countries producing cheap bioethanol. The reaction mechanism of benzene alkylation with ethylene to produce ethanol is commonly considered via carbenium-ion-type mechanisms including Br¨onsted acid sites protonating ethylene to give the active species [9]. Further behavior of this active species (most probably ethoxy group attached to the surface) is critical for the selectivity of the reaction. It can directly alkylate benzene to ethylbenzene but also ethylbenzene to DEB or dimerize (oligomerize) to higher oleﬁns. Higher oleﬁns (the smallest being butenes) can isomerize, oligomerize, alkylate, or crack, but in all cases undesired products are formed. Although most of the studies employed ZSM-5 zeolite, the use of other zeolites types like Y, Beta, mordenite, ZSM-12, and MCM-22 has also been reported [8, 9]. It has been shown that long and stable catalyst life is observed when alcohol, rather than oleﬁn, is used as the alkylating agent [89] because of the formation of water which enhances the proton-transfer reactions. At present, one of the most popular zeolite-based processes for the production of ethylbenzene is the Mobil–Badger process which involves the ethylation of benzene over ZSM-5 zeolite. Excellent selectivity of 98% at benzene conversion of up to 20% can be attained [90]. Among the modiﬁed zeolite catalysts for benzene alkylation, Ca- and Mg-modiﬁed ZSM-5 zeolite-based catalysts (Si/Al ratio = 75–100) exhibit more than 90% para selectivity at 370 ◦ C and 1.0 hour−1 WHSV [91]. Ethylation of benzene with ethanol over ZSM-5 (Si/Al ratio = 100) modiﬁed with boric acid and Mg(NO3 )2 resulted in a high benzene conversion and selectivity to p-ethylbenzene at the optimum temperature range of 400–450 ◦ C. Impurities in ethanol could decrease the activity of the same zeolite catalysts when compared to pure ethanol or a 95% aqueous solution of ethanol [92, 93]. The following major factors affect the activity of zeolites in the ethylation of benzene and the product selectivity. • High para selectivity can be achieved over medium pore zeolites, mainly ZSM-5. • Zeolite catalyst stability is higher when using ethanol rather than ethylene as alkylating agent. • B-, Ca-, and Mg-modiﬁed ZSM-5 zeolite exhibited >90% para selectivity at high benzene conversion at the reaction temperatures 400–450 ◦ C. 20.6.2 Methylation of Toluene

Methylation of toluene providing a mixture of xylenes is a reaction being frequently used as a model reaction to characterize the activity and selectivity of zeolite catalysts though their industrial potential is still rather limited because of the

20.6 Alkylation of Aromatics

price of methanol. Toluene disproportionation is economically favored at present. However, methanol is expected as a strategic substance for the future and it can play a role in methylation reactions [94]. ZSM-5 zeolite was mostly investigated for toluene methylation, although some modiﬁcation is needed to improve the para selectivity because p-xylene is the most important of the three xylene isomers. The increase in the selectivity to p-xylene is directly connected with shape selectivity of ZSM-5 zeolite; in this case, the product shape selectivity is under operation [4, 8]. The model of Haag and coworkers, thoroughly discussed in [4, 95], assumes the formation of all three xylene isomers inside of the channel system of ZSM-5 zeolite. The composition of xylenes is assumed to be close to the thermodynamic one (25% para, 50% meta, 25% ortho). p-Xylene is the isomer with the smallest kinetic diameter and its diffusion coefﬁcient is 2–3 orders of magnitude larger than those of m- and o-xylenes [32]. To increase the selectivity to p-xylene, two approaches (or their combination) are needed. First, active sites on the external surface should be passivated to prevent secondary isomerization of p-xylene to m-xylene and further to o-xylene via 1,2-methyl shift, and second, pore entrances should be slightly decreased to further increase the differences among the kinetic diameters of individual xylene isomers. Experimentally, it is not easy to separate these two approaches, because during the surface annihilation of active sites some of channel entrances are partially or completed plugged. Generally, the activity and selectivity of zeolites in the alkylation of toluene with methanol strongly depend on the nature, number, and strength of the acid sites. However, the exact manner in which zeolite acidity affects toluene alkylation with methanol is still not completely understood. It is reported in the literature that catalysts with strong acid centers generally exhibit more activity but a lower selectivity and a shorter operation time due to easy deactivation by coke formation. The acidity of zeolites stems from the presence of aluminum or other trivalent cations in the zeolite framework which provide a good way to tune the acidic properties such as the concentration of acid sites (by changing Al content) or their strength (via replacing silicon with other trivalent cations). The following order was found for the strength of acidity of ZSM-5-based catalysts on the basis of catalytic and physicochemical assessment: B ≤ Fe < Ga < Al [63, 96, 97]. Thus the weak Br¨onsted acid boron, though not very active, is likely to have higher selectivity. It has been shown that p-xylene selectivity is enhanced during the gas-phase alkylation of toluene with methanol when aluminum in the zeolite is replaced by gallium [98]. The side reaction of toluene disproportionation usually requires stronger acid sites than for toluene methylation (in general, aromatic alkylations) and can be limited by modifying the acid strength of MCM -22 using lanthanum [99]. Methylation of toluene using methanol over USY zeolite-based catalyst in the temperature range 375–450 ◦ C provided toluene conversion of 12% at 450 ◦ C, but the para selectivity decreased with increasing reaction temperature [100]. Methylation reaction conducted over a mordenite-based catalyst resulted in toluene conversion of 30% and xylene selectivity of 81%. Some amount of

637

638

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites Table 20.4 Toluene methylation results (wt%) for mordenite/aluminaa catalyst at 400 ◦ C temperature, LHSV 1.5 hour−1 and 30 bar hydrogen pressure using a feed containing 80 wt% toluene and 20 wt% methanol.

Nonaromatics (parafﬁns and cycloparafﬁns) Benzene Toluene Ethylbenzene m-Xylene p-Xylene o-Xylene C9+ aromatic hydrocarbons Total xylenes Xylene selectivity Toluene conversion

6.9 1.7 50.0 0.6 13.6 5.1 5.7 16.3 24.4 81.5 30.0

a Contains 75 wt% mordenite (Si/Al ratio of 9.2 obtained from Tosoh Chemicals, Japan) and 25 wt% alumina binder.

water was also produced. The composition of the organic phase of the reaction products is provided in Table 20.4, which shows the formation of signiﬁcant amounts of C9 + aromatic hydrocarbons and nonaromatics. The formation of C9 + aromatic hydrocarbons was attributed to the presence of large channels of mordenite. 20.6.2.1 Modiﬁcation of the External Surface of Zeolites Methylation of toluene with methanol using up to 10 wt% of B- and Mg-modiﬁed ZSM-5 (Si/Al = 40) was carried out under atmospheric pressure at 440 ◦ C temperature with H2 or N2 as carrier, water vapor : MeOH ratio of 9 : 1, toluene:MeOH ratio of 8, and WHSV of 11.1 hour – 1 . It was observed that high selectivity to p-xylene can be achieved using an ultralow contact time of less than 0.3 seconds. Higher toluene : methanol feed ratio was found to be beneﬁcial to minimize the dehydration reaction of methanol. It was found that B/ZSM-5 can be produced in situ, which resulted in >99.9% selectivity to p-xylene. The effect of the contact time and space velocity on p-xylene selectivity has been clearly demonstrated [16, 101, 102]. Niwa and coworkers investigated in situ and ex situ modiﬁcation of ZSM-5 with chemical vapor deposition aiming to increase the para selectivity [13]. The authors showed that in situ modiﬁcation provided a ﬁne structure of the external surface, enabling ﬁne control of channel entrances. Thus, in situ chemical vapor deposition was preferred to obtain a high selectivity to p-xylene. It should be noted that the group of Niwa proposed chemical vapor deposition in 1982 and still this area provides new, interesting results [103, 104]. The other method was investigated by Ghiaci et al. [105], who modiﬁed the ZSM-5 zeolite with phosphoric acid to adjust the strength of the acid sites. This

20.6 Alkylation of Aromatics

resulted in a maximum 46% conversion of methanol and 100% para selectivity in the temperature range 350–500 ◦ C. Another modiﬁcation of ZSM-5 was carried out by extruding with Ludox AS-40 silica, and silylating with 10 wt% poly(methylphenyl) siloxane and tetraethyl orthosilicate. This modiﬁcation resulted in 100% selectivity to p-xylene at 100% conversion of methanol. The methylation reaction was carried out at 70–80 psig pressure and 600 ◦ C using a toluene : methanol molar ratio 20 : 80 and H2 as carrier [106]. Phosphorus-treated ZSM-5 zeolite catalyst was patented by Saudi Basic Industries Corporation (SABIC) for toluene alkylation with methanol in the presence of hydrogen and water [107]. Another possibility of ZSM-5 modiﬁcation for toluene methylation is precoking using 1,3,5-tri-isopropylbenzene. This method was investigated in a riser simulator at 375–450 ◦ C for 3–15 seconds of contact time with a toluene : methanol molar ratio of 1 : 1. The resulting of p-xylene/o-xylene ratio obtained was rather high (3.5–6.0) at toluene conversion of 16.8% and showed high para selectivity under the studied reaction conditions [100]. In addition to ZSM-5 zeolite, a variety of other zeolites including mordenite and MCM-22 or zeotypes such as SAPO-34, SAPO-11, and SAPO-5 have been studied for the methylation of toluene [108, 109]. The methylation reaction was performed under atmospheric pressure and 440 ◦ C temperature at a space velocity WHSV 2.0 hour−1 and hydrogen as a carrier gas with an H2 : reactant ratio of 10 : 1. The catalyst activity was found to be proportional to the acid strength of the catalysts used [110]. The authors studied MCM-22 and delaminated ITQ-2 zeolites at 250 ◦ C under atmospheric pressure and a toluene : methanol ratio of 1 : 1 and obtained a high p-xylene selectivity of 74%. The surfaces of both zeolites were modiﬁed using collidine. On the basis of a comparison of toluene conversions and selectivities to p-xylene, it was concluded that in MCM-22 the reaction proceeds in a 10-ring system and this was the reason for the para selectivity enhancement [110]. The major factors inﬂuencing the activity and selectivity in the methylation of toluene can be summarized as follows: • Increase in the reaction temperature favors the conversion of toluene with a simultaneous decrease in para selectivity. • Large pore zeolites (e.g., mordenite) produce more C9 + hydrocarbons compared to medium pore zeolites. • Catalysts with strong acid sites are generally more active for methylation of toluene but this is connected with a low selectivity and rapid deactivation. • Boron-containing ZSM-5 resulted in 99% para selectivity at ultrashort contact times of <0.3 seconds. • p-Xylene selectivity is enhanced during toluene alkylation with methanol when aluminum in the zeolite is replaced by gallium. • N-Cetylpyridinium surfactant modiﬁed the H3 PO4 /ZSM-5 catalyst resulting in 46% conversion of methanol and 100% para selectivity. • Surface modiﬁcations of the ZSM-5 zeolite (e.g. with TEOS, coke deposition, bulky organic compounds) usually increase the para selectivity, but at the expense of toluene conversion, which decreases.

639

640

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites

20.6.3 Ethylation of Toluene and Ethylbenzene

Alkylation of toluene and ethylbenzene with ethylene or ethanol are aromatic reactions of immense industrial importance. Toluene ethylation yields isomers of ethyltoluene, which are very important starting materials for the production of a wide range of useful products. The most important of the three isomers, p-ethyltoluene, is mainly dehydrogenated to p-methyl styrene, which is the monomer for poly-p-methyl styrene production. Similarly, direct ethylation of ethylbenzene produces a mixture of DEB isomers. Diethylbenzenes are used as solvent and precursors for cross-linking agents in producing resins. p-DEB is an important desorbent for the Parex process of UOP used in separating p-xylene from a mixture of xylenes. Similar to other aromatic alkylation reactions, particular attention is now centered on the use of zeolite-based catalyst for these reactions. The main challenge is developing a catalyst system that can enhance the yield of the p-ethyltoluene or p-DEB while maintaining reasonable levels of aromatic conversion. The use of large pore zeolites like X, Y, and mordenite in these alkylation reactions has been reported as well, but catalyst aging due to coking is severe. In contrast, medium pore zeolites such as ZSM-5 show high stability, low coke formation, and longer life [111]. When comparing the para selective properties of 10-ring zeolites, mainly ZSM-5, the comparison of diffusion coefﬁcients and rates of isomerization plays very important role. Particularly, when comparing three basic homologs – xylenes, ethyltoluenes, and DEBs [8, 9] – one can expect that the kinetic diameters of individual isomers increase in the following sequence: xylenes < ethyltoluenes < di-ethylbenzenes. The rate of isomerization of these di-alkylbenzenes will also increase in the same order, but the diffusion coefﬁcients should decrease in this series. On the basis of this, no clear conclusions can be drawn without a detailed experimental study on the para selectivity of alkylation reactions over medium pore zeolites. In ethylbenzene alkylation with ethylene, Arsenova and coworkers observed the almost complete absence of the ortho isomer compared to the other xylene isomers [63]. In adsorption studies [65], o-DEB was shown to have restricted access to ZSM-5 channels. In addition, the differences between the meta and para isomers are larger compared to those of the respective xylene isomers. This clearly indicates the importance of restricted transition-state selectivity during the alkylation step followed by the differences in the diffusion rates of individual isomers from the zeolite channels. Finally, the active sites on the external surface of zeolite crystals with an unrestricted reaction environment have to be considered for isomerization of the para isomers leaving the zeolite channels and for providing unrestricted ethylation and disproportionation reactions. The ﬁrst step considered in the ethylation of toluene is the formation of p-ethyltoluene inside the channels which isomerizes to a certain extent to the meta and ortho isomers on the external acid sites [112]. The relative contributions of the differences in the transport rates of the individual dialkylbenzene isomers and the surface active sites have been analyzed by a number of investigators [113, 114]. It has

20.6 Alkylation of Aromatics

been suggested that the rate of diffusion-controlled, shape-selective transformation inside the zeolite crystals is lower than that on the unrestricted surface sites. Accordingly, with small crystals containing relatively high concentrations of surface acid sites, low para selectivity is achieved, while large crystals yield high para selectivity but at low conversion values. Many attempts have been made at Mobil Oil to increase the zeolite selectivity for p-ethyltoluene by modifying ZSM-5 with phosphorus and by impregnation with Mn, Mg, and B salts [115, 116]. A dramatic increase in the para selectivity of upto 98% is ascribed mainly to spatial restrictions in the narrowed zeolite channels rather than differences in the zeolite acidity. In addition, selective silylation of the surface acid sites of H-ZSM-5 by TEOS has shown to increase the zeolite para selectivity (from 32% to 64%) in the ethylation of toluene [117]. At 30% annihilation of the surface acid sites compared to the parent ZSM-5 zeolite, the para selectivity dramatically increased. The effect of zeolite coking on the para selectivity in the ethylation reactions of toluene and ethyl benzene was analogous to that observed with xylenes. Coke was formed preferentially on the surface sites; moreover, its deposits most probably resulted in a narrowing of the channels. Accordingly, coke plays a positive role in the production of p-ethyl toluene and p-DEB up to a certain level which dramatically decreases the conversion. It can be inferred that at this stage most of the surface is covered by coke and there is very limited access to the zeolite channels. Further attempts aiming at the enhancement of para selectivity in toluene and benzene alkylation were reported recently. Ethylation of ethylbenzene with ethanol over nitrided ZSM-5 exhibited much higher selectivity to p-DEB and signiﬁcantly enhanced the lifetime than the parent ZSM-5. The nitridation was carried out with a stream of ammonia at high temperature, and the para selectivity improvement was attributed to the decrease in concentration of strong acid sites causing isomerization of p-DEB [118]. Vapor-phase alkylation of benzene, toluene, and ethylbenzene with ethanol, 2-propanol, methanol, and t-butanol over Mn-substituted aluminophosphate molecular sieves shows a high initial activity at reaction temperatures ranging from 350 to 400 ◦ C. The rate of deactivation accelerated with increasing length of the carbon chain and bulkiness of the alkylating agents [119]. Major factors affecting ethylation of toluene and ethylbenzene are as follows: • Activity of zeolites in toluene and ethylbenzene ethylations increases with the size of channels and their dimensionality and with concentration of the acid sites. • Phosphorous-modiﬁed ZSM-5 impregnated with Mn, Mg, or B resulted in para selectivity of upto 98% during ethylation of toluene. • Silylation of ZSM-5 with TEOS increased para selectivity from 32% to 64% during ethylation of toluene. • Nitrided ZSM-5 zeolite exhibited much higher selectivity to p-DEB and high stability during ethylation of ethylbenzene.

641

642

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites

20.7 Miscellaneous

Recently, the SSZ-33 (structure of this zeolite is schematically depicted in Figure 20.2) zeolite-based catalyst modiﬁed with Mo and alumina binder was evaluated in long-run tests in toluene conversion and compared with ZSM-5 zeolite catalyst prepared under the same conditions [120]. Much higher toluene conversion was achieved with the SSZ-33 catalyst comprising the 12-12-10-ring channel system as a result of high acidity of this zeolite and increased mass transport in the large channels. In addition to the increasing number of industrial applications of zeolites in aromatic chemistry and the increasing amount of units under operation, these reactions can help us to understand the structural features of new zeolites whose structures have not been determined yet. Early ideas came from Csicsery using reactions of 1-methyl-2-ethylbenzene to discriminate between large pore and medium pore zeolites and different channel architectures [121]. Not only the transformations of aromatic hydrocarbons but also their formation from, for example, methanol could be very helpful in the assessment of the structural properties of unknown zeolites [122]. Distribution of C7 –C10 aromatics over zeolites IM-5 and SSZ-57 matched nicely with those over ZSM-5 zeolite, with adsorption studies carried out with 2,2-dimethylbutane [123] evidencing the 10-ring channel systems of both zeolites. In the case of IM-5, it was already conﬁrmed by structure determination [45]. p-Xylene alkylation with isopropyl alcohol is not a frequently studied reaction, although it can be applied as the ﬁrst reaction step for the synthesis of xylenols and used as model reaction to characterize the inner channel volume of zeolites [124]. In particular, zeolite architecture and channel dimensionality of the novel zeolites SSZ-33 and SSZ-35 and the traditional ZSM-5, Beta, and mordenite were studied using this reaction. The structural motif of SSZ-35 is provided in Figure 20.2. We observed that p-xylene conversion increases with increasing pore size and connectivity of the channel system of individual zeolites with the exception of SSZ-35. Zeolite SSZ-35 possesses one-dimensional 10-ring channels that periodically open into wide, shallow cavities circumscribed by 18 rings. In p-xylene alkylation, SSZ-35 exhibited the highest conversion among all zeolites at the reaction temperature of

(a)

Figure 20.2

(b)

(c)

Structural motifs of novel zeolites SSZ-33 (a), SSZ-35 (b), and MCM-22 (c).

20.8 Summary and Outlook

150 ◦ C and also the highest selectivity to 1-isopropyl-2,5-dimethyl-benzene. Molecular modeling conﬁrmed that the dimensions of the 18-ring cages represent an optimum reaction space for the formation of 1-isopropyl-2,5-dimethyl-benzene [125]. 20.8 Summary and Outlook

The conversion of low-value aromatics into value-added aromatics such as xylenes, DEB, and ethyltoluene is very efﬁcient when carried out over a wide variety of zeolite-based catalysts. Both medium pore and large pores zeolites including ZSM-5, mordenite, beta, Y, MCM-22, SSZ-33, and SSZ-35 can be successfully employed depending on the type of reaction. The number of tested zeolites with a combination of 10- and 12-ring channel systems has increased substantially [126–128]. The choice of the zeolite depends upon the type of aromatics to be produced. Generally, p-dialkylbenzenes are preferentially formed over medium pore zeolites, mainly ZSM-5 because of the para shape selectivity. It means that for alkylation of toluene or ethylbenzene or their disproportionations, zeolite ZSM-5 is the ﬁrst choice. The selectivity to the para isomers is enhanced by deactivation of the active sites located on the external surface of zeolite crystals and by narrowing the zeolite channels. Deactivation and pore narrowing of the active acid sites are usually performed using following methods: • impregnation and/or ion-exchanging of the metal species, • loading of high molecular weight silicon-containing compounds followed by calcinations, • selective precoking of the catalysts or steaming of the zeolites for dealumination. Other factors inﬂuencing the para selectivity are as follows: • crystal size of zeolites, • presence of extra-framework alumina in the zeolites, • optimization of the reaction conditions for temperature, pressure, feedstock composition, and space velocity. In contrast, large pore zeolites are clearly favored when benzene is alkylated with ethylene or propylene or for trimethylbenzene disproportionation/transalkylation. For benzene alkylation, a fast diffusion is preferred and even some undesired products can be almost eliminated (like n-propyltoluene) as they are formed in three-dimensional medium pore zeolites. Kinetic diameters of trimethylbenzenes do not allow their penetration to medium pore zeolites; therefore zeolites Beta or Y are the highly active catalysts. Recent advances in the synthesis of new structural types of zeolites provide huge opportunities to adjust the reaction conditions in transformations of aromatics to these new structural types and to optimize the current processes. The standard ZSM-5 now competes with other 10-ring zeolites like TNU-9, IM-5, SSZ-74 with already solved structures, while another (most probably also 10-ring) zeolite SSZ-57 still lacks detailed structural information.

643

644

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites

With increasing number of new zeolite structures, more and more examples are provided showing the special catalytic behaviors of zeolites which are not expected on the basis of their structures. Typical example is SSZ-35 (10-ring zeolite with shallow 18-ring cages). Its catalytic behavior is much closer to that of 12-ring zeolites and even the long-term stability of its activity is excellent. A particular example of catalytic behavior of SSZ-35 is p-xylene alkylation with isopropyl alcohol providing a selectivity to 2,5-dimethylcumene over 90% at a high conversion level. This was ascribed to easy formation and desorption of the product in 18-ring cages although located in 10-ring channels. To understand in more detail the behavior of zeolites in transformations of aromatic hydrocarbons, it is necessary to address several key issues concerning the properties of zeolites, but at least the following three are of a particular importance. 1) The evaluation of the accessibility of the channel system of zeolites for structures of different dimensionality of the channels; when 1D and 3D zeolites are of the same crystals size, the channel accessibility is different as only some planes for 1D zeolite have channel entrances. 2) Acid sites are located in different positions, particularly in zeolites with a large amount of different crystallographic sites, and this is necessary to properly understand, 3) Different synthesis conditions providing the same structural type of zeolite with different distribution of aluminum in the framework [129]. Thus, we are still far from a complete understanding of all of the phenomena that inﬂuence and control the activity of zeolites and their selectivity in transformations of aroamtic hydrocarbons. On the other hand, zeolites are indispensable in chemical industry, which clearly evidences the maturity of these catalysts at present.

Acknowledgments

The Support provided by the Center of Research Excellence in Petroleum Reﬁning & Petrochemicals at King Fahd University of Petroleum & Minerals (KFUPM) is highly appreciated. J.C. thanks the Academy of Sciences of the Czech Republic for the ﬁnancial support to this study (1QS400400560). References 1. Franck, H.G. and Stadelhofer, J.W.

(1988) Industrial Aromatic Chemistry, Springer, Berlin. 2. Kaeding, W.W., Barile, G.C., and Wu,

M.M. (1984) Catal. Rev. – Sci. Eng., 26, 597–612. 3. Meima, G.R. (1998) CATTECH, 3,

5–12.

4. Hansen, N., Br¨ uggemann, T., Alexis,

T.B., and Frerich, J.K. (2008) J. Phys. Chem. C, 112, 15402. 5. Tanabe, K. and H¨ olderich, W.F. (1999) Appl. Catal. A, 181, 399–434. 6. Tsai, T., Liu, S., and Wang, I. (1999) Appl. Catal. A, 181, 355–398. 7. Degnan, T.F. Jr. (2000) Top. Catal., 13, 274–279.

References ˇ J. and Wichterlov´a, B. (2002) 8. Cejka, 9. 10.

11.

12.

13.

14.

15.

16.

17. 18.

19.

20. 21. 22.

Catal. Rev. Sci. Eng., 44, 375. Perego, C. and Ingallina, G. (2004) Green Chem., 6, 274. (a) Bellussi, G. (2004) Proceedings of 14th International Conference on Zeolites, April, 2004, Cape Town, South Africa; (b) Bellusi, G. (2004) Zeolite catalysts for the production of chemical cammodities: BTX derivatives in Studies in Surface Science and Catalysis, 154 (1), 53–65. Inagaki, S., Kamino, K., Kikuchi, E., and Matsukata, M. (2007) Appl. Catal. A, 318, 22–27. Vu, D.V., Miyamoto, M., Nishiyama, N., Egashira, Y., and Ueyama, K. (2009) Catal. Lett., 127, 233–238. Vu, D.V., Miyamoto, M., Nishiyama, N., Ichikawa, S., Egashira, Y., and Ueyama, K. (2007) Microporous Mesoporous Mater., 115, 106–112. Tominaga, K., Maruoka, S., Gotoh, M., Katada, N., and Niwa, M. (2009) Microporous Mesoporous Mater., 117, 523–529. Joshi, P.N., Niphadkar, P.S., Desai, P.A., Patil, R., and Bokade, W. (2007) J. Nat. Gas Chem., 16, 37–41. Breen, J.P., Burch, R., Kulkarni, M., Mc Laughlin, D., Collier, P.J., and Golunski, S.E. (2007) Appl. Catal. A, 316, 53–60. Corma, A. (1995) Chem. Rev., 95, 559–614. ˇ Cejka, J., van Bekkum, H., Corma, A., and Schth, F. (eds) (2007) Introduction to Zeolite Science and Practice, Studies in Surface Science and Catalysis, Vol. 168, 3rd edn, Elsevier, Amsterdam. McCusker, L.B. and Baerlocher, C. (2007) Stud. Surf. Sci. Catal., 168, 13–38. Weisz, P.B. (1980) Pure Appl. Chem., 52, 2091–2103. Csicsery, S.M. (1984) Zeolites, 4, 202–213. ˇ ˇ Bejblov´a, M., Zilkov´ a, N., and Cejka,

J. (2008) Res. Chem. Intermed., 34, 439–454. 23. Venuto, P.B. (1994) Microporous Mater., 2, 297–411.

24. De Vos, D.E. and Jacobs, P.A. (2005)

25.

26. 27. 28.

29.

30. 31. 32. 33.

34.

35.

36.

37.

38.

39. 40.

41. 42.

Microporous Mesoporous Mater., 82, 293–304. ˇ ˇ Cejka, J., Krejˇc´ı, A., Zilkov´ a, N., Kotrla, J., Ernst, S., and Weber, A. (2002) Microporous Mesoporous Mater., 53, 121–130. Koˇsov´a, G. (2005) Stud. Surf. Sci. Catal., 158, 59–66. Hong, S.B. (2008) Catal. Surv. Asia, 13, 131–144. Lee, S.H., Lee, D.K., Shin, C.H., Park, Y.K., Wright, P.A., Lee, W.M., and Hong, S.B. (2003) J. Catal., 215, 151–170. Gil, B., Zones, S.I., Hwang, S.J., ˇ Bejblov´a, M., and Cejka, J. (2008) J. Phys. Chem. C, 112, 2997–3007. Al-Khataff, S. and de Lasa, H. (1999) Ind. Eng. Chem. Res., 38, 1350–1356. Al-Khataff, S. and de Lasa, H. (2001) Can. Chem. Eng., 79, 341–348. ˇ Mirth, G., Cejka, J., and Lercher, J.A. (1993) J. Catal., 139, 24. Beck, J.S., Cheng, J.C., McCullen, S.B., Olson, D.H., and Stern, D.L. (2003) U.S. Patent 6,576,582. Heng, S., Heydenrych, H.R., Roeger, H.P., Jentys, A., and Lercher, J.A. (2003) Top. Catal, 22, 101–106. Xie, Z., Zhu, Z., Kong, D., Li, W., Yang, W., and Chen, Q. (2008) Chin. Patent CN101121139 A. Zhu, Z., Li, W., Kong, D., Chen, Q., Hou, M., and Zhang, H. (2008) Chin. Patent CN101121140 A. Haag, W.O., Olson, D.H., and Rodewald, P.G. (1985) US Patent 4,508,836. ˇ ˇ Cejka, J., Zilkov´ a, N., Wichterlov´a, B., Eder-Mirth, G., and Lercher, J.A. (1996) Zeolites, 17, 266. Rabiu, S. and Al-Khattaf, S. (2008) Ind. Eng. Chem. Res., 46, 39–47. Chen, W.H., Huang, S.J., Lai, C.S., Tsai, T.C., Lee, H.K., and Liu, S.B. (2003) Res. Chem. Intermed., 29, 761–772. Park, S.H. and Rhee, H.K. (2003) React. Kinet. Catal. Lett., 78, 81–89. Mihalyi, M.R., Kolev, I., Mavrodinova, V., Minchev, C., Kollar, M., and Valyon, J. (2007) React. Kinet. Catal. Lett., 92, 345–354.

645

646

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites 43. Kolev, I., Mavrodinova, V., Mih´alyi,

44. 45.

46.

47.

48.

49.

50. 51. 52. 53.

54.

55.

56.

57. 58.

59.

60. 61.

M.R., and Koll´ar, M. (2009) Microporous Mesoporous Mater., 118, 258–266. Al-Khattaf, S. (2007) Energy Fuels, 21, 646–652. Hong, S.B., Min, H.K., Shin, C.H., Cox, P.A., Warrender, S.J., and Wright, P.A. (2007) J. Am. Chem. Soc., 129, 10870–10885. Holmgren, J.S., Galloway, D.B., Galperin, L.B., and Willis, R.R. (1999) US Patent 6,008,423. Sugi, Y., Kubota, Y., Hayashi, S., Sugiyama, N., and Tawada, S. (2003) Jap. Patent JP 2003292462. Kareem, M.A.A., Chand, S., and Mishra, I.M. (2002) J. Inst. Eng., Chem. Eng. Div., 83, 6–8. Verduijn, J.P., Mohr, G.D., Clem, K.R., Mortier, W.J., Mertens, M.M., and Bradford, M.C. (2003) US Patent 504,074. Kelly K.P. and Butler J.R. (2002) US Patent 6,462,247. Zheng, S., Jentys, A., and Lercher, J.A. (2003) J. Catal., 219, 310–319. Al-Khattaf, S., Ali, M.A., and Al-Amer, A. (2008) Energy Fuels, 22, 243–249. Pai, S.M., Sharanappa, N., Anilkumar, M., Kadam, S.T., and Bokade, V.V. (2004) Chem. Eng. Res. Des., 82, 1391–1396. Raj, K.J.A., Padma Mala, E.J., and Vijayaraghavan, V.R. (2006) J. Mol. Catal. A, 243, 99–105. Raj, K.J.A., Meenakshi, M.S., and Vijayaraghavan, V.R. (2007) J. Mol. Catal. A, 270, 195–200. Sharanappa, N., Pai, S., and Bokade, V.V. (2007) J. Mol. Catal. A, 217, 185–191. Melson, S. and Sch¨uth, F. (1997) J. Catal., 170, 46–53. Pradhan, A., Liu, T.S., Chen, W.H., Jong, S.J., Wu, J.F., Chao, K.J., and Liu, S.B. (1999) J. Catal., 29, 184. Zhu, Z.R., Chen, Q.L., Xie, Z.K., Yang, W.M., Kong, D.J., and Li, C. (2006) J. Mol. Catal. A, 248, 152–158. Klemm, E., Scheidat, H., and Emig, G. (1997) Chem. Eng. Sci., 52, 2757–2768. Klemm, E., Wang, J.G., and Emig, G. (1997) Chem. Eng. Sci., 52, 3173–3182.

62. Sharanappa, N., Pai, S., and Bokade,

63.

64.

65.

66.

67. 68. 69. 70. 71.

72. 73.

74.

75.

76. 77. 78. 79.

80.

81.

V.V. (2004) J. Mol. Catal. A, 217, 185–191. Arsenova, N., Haag, W.O., and Karge, H.G. (1997) Stud. Surf. Sci. Catal., 105, 1293–1300. Arsenova-Hartel, N., Bludau, H., Haag, W.O., and Karge, H.G. (2000) Microporous Mesoporous Mater., 35- 36, 113–119. Arsenova-Hartel, N., Bludau, H., Schumacher, R., Haag, W.O., Karge, H.G., Brunner, E., and Wild, U. (2000) J. Catal., 191, 326–331. Wang, I., Tsai, T.C., and Huang, S.T. (1990) Ind. Eng. Chem. Res., 29, 2005–2012. Park, S.H. and Rhee, K.H. (2000) Catal. Today, 63, 267–273. Shamshoum, E.S., Ghosh, A.K., and Butler, J.R. (1995) US Patent 5475180. Mikhail, S., Ayoub, S.M., and Barakat, Y. (1987) Zeolites, 7, 231–234. ˇ Cejka, J., Kotrla, J., and Krejˇc´ı, A. (2004) Appl. Catal. A, 277, 191–199. Das, J., Bhat, Y.S., and Halgeri, A. (1994) Ind. Eng. Chem. Res., 33, 246–250. Al-Khattaf, S. (2007) Energy Fuels, 21, 646–652. Dumitriu, E., Hulea, V., Kaliaguine, S., and Huang, M.M. (2002) Appl. Catal. A, 237, 211–221. Zhu, Z., Xie, Z., Qi, X., Kong, D., Yang, W., and Chen, Q. (2008 Chin. Patent CN 101121144. Zhu, Z., Xie, Z., Chen, Q., Kong, D., Li, W., Yang, W., and Li, C. (2007) Microporous Mesoporous Mater., 101, 169–175. Qi X., Xie Z., Zhu Z., Kong D., and Zuo Y. (2008) Chin. Patent, CN 101121132. Bricker M.L. and Modica F.S. (2008) US Patent 20080026930 A1. Bricker, M.L. and Modica, F.S. (2008) US Patent 20080027257. Boldingh, E.P., Negiz, A., Gregory, F., Bogdan, P.L., and Rende, D.E. (2007) US Patent 20070185356. Boldingh, E.P., Bricker, M.L., Larson, R.B., Modica, F.S., and Rekoske, J.E. (2008) Eur. Patent, EP 1882728. Zones, S.I., Burton, A.W., and Ong, K. (2007) World Patent WO2007079038.

References 82. Zones, S.I., Burton, A.W., Maesen,

83.

84. 85. 86.

87. 88.

89.

90.

91. 92. 93. 94.

95. 96. 97.

98.

99.

100.

T.L.M., Smit, B., and Beerdsen, E. (2007) US Patent 20070284284. Zones, S.I., Burton, A.W., Ong, K., Maesen, T.L.M., Smit, B., and Beerdsen, E. (2007) World Patent, WO 2007146622. Lee, C.S. and Park, T.J. (1993) Appl. Catal., 96, 151–161. Haag, W.O., Lago, R.M., and Weisz, P.B. (1984) Nature, 309, 589–591. Chen, N.Y., Keading, W.W., and Dwyer, F.G. (1979) J. Am. Chem. Soc., 101, 6783–6784. Young, L.B., Butter, S.A., and Keading, W.W. (1982) J. Catal., 76, 418–432. Degnan, T.F., Smith, C.M., and Venkat, C.R. Jr. (2001) Appl. Catal. A, 221, 283–294. Sridevi, U., Bhaskar Rao, B.K., and Pradhan, N.C. (2001) Chem. Eng. J., 83, 185–189. Beck, J.S. and Haag, W.O. (1997) Alkylation of Aromatics in Handbook of Heterogeneous Catalysis (eds G. Ertl, H. Kn¨ozinger, and J. Weitkamp), Wiley-VCH Verlag GmbH, New York, p. 2123. Pop, G.R., Ganea, R., Ivanescu, D., and Tamas, I. (1998) Progr. Catal., 7, 11–14. Jun-Jun, Y. and Boerje, G. (2006) Ind. J. Chem. Technol., 13, 334–340. Jun-Jun, Y. and Boerje, G. (2004) Ind. J. Chem. Technol., 11, 337–345. Olah, G.A., Goeppert, A., and Surya Prakash, G.K. (2006) Beyond Oil and Gas: The Methanol Economy, Wiley-VCH Verlag GmbH & Co KGaA, Weinheim. Olson, D.H. and Haag, W.O. (1984) ACS Symp. Ser., 248, 275–280. Chu, C.T.-W. and Chang, C.D. (1985) J. Phys. Chem., 89, 1569–1571. ˇ Cejka, J., Vondrov´a, A., Wichterlov´a, B., Vorbeck, G., and Fricke, R. (1994) Zeolites, 14, 147–153. Fechete, I., Caullet, P., Dumitriu, E., Hulea, V., and Kessler, H. (2005) Appl. Catal. A, 280, 245–254. Zhu, Z.R., Chen, Q.L., Xie, Z.K., Yang, W.M., Kong, D.J., and Li, C. (2006) J. Mol. Catal. A, 248, 152–158. Rabiu, S. and Al-Khattaf, S. (2008) Ind. Eng. Chem. Res., 46, 39–47.

101. Breen, J.P., Burch, R., Kulkarni, M.,

102.

103.

104.

105.

106. 107. 108.

109. 110.

111. 112.

113.

114. 115. 116.

117.

118.

119.

120.

McLaughlin, D., Collier, P.J., and Golunski, S.E. (2005) J. Am. Chem. Soc., 127, 5020–5021. Breen, J.P., Burch, R., Collier, P.J., and Golunski, S.E. (2008) US Patent 7,321,072 B2. Niwa, M., Itoh, H., Hattori, T., and Murakami, Y. (1982) J. Chem. Soc. Chem. Commun., 819–820. Niwa, M., Kato, S., Hattori, T., and Murakami, Y. (1984) J. Chem. Soc., Faraday Trans., 80, 3135–3145. Ghiaci, M., Abbaspur, A., Arshadi, M., and Aghabarari, B. (2007) Appl. Catal. A, 316, 32–46. Wu, H. and Drake, C.A. (2008) US Patent 7,323,430 B2. SABIC (2006) US Patent 7,060,864. Arandes, J.M., Abajo, I., Fernandez, I., Azkoiti, M.J., and Bilbao, J. (2000) Ind. Eng. Chem. Res., 39, 1917–1924. Zhu, Z. (2004) Catal. Today, 93, 321–325. Inagaki, S., Kamino, K., Kikuchi, E., and Matsukata, M. (2007) Appl. Catal. A, 318, 22–27. Bhandarkar, V. and Bhatia, S. (1994) Zeolites, 14, 439–449. Paparatto, G., Moretti, E., Leofanti, G., and Gatti, F. (1987) J. Catal., 105, 227–232. Shiralkar, V.P., Joshi, P.N., Eapen, M.J., and Rao, B.S. (1991) Zeolites, 11, 511–516. Beschman, K., Riekert, L., and Miller, V. (1994) J. Catal., 145, 243–253. Kaeding, W.W. (1985) J. Catal., 95, 512–519. Kaeding, W.W., Barile, G.C., and Wu, M.M. (1984) Catal. Rev. – Sci. Eng., 26, 597–603. Negiz, A., Boldingh, E.P., Gajda, G.J., and Gurevich, S.V. (2004) US Patent 6,815,570. Guan, X.X., Li, N., Wu, G.J., Chen, J.X., Zhang, F.X., and Guan, N.J. (2006) J. Mol. Catal. A, 248, 220–225. Raj, K.J.A., Meenakshi, M.S., and Vijayaraghavan, V.R. (2007) J. Mol. Catal. A, 270, 195–200. Al-Khattaf, S., Musilov´a-Pavlaˇckov´a, ˇ Z., Ali, M.A., and Cejka, J. (2009) Top. Catal., 52, 140–147.

647

648

20 Recent Development in Transformations of Aromatic Hydrocarbons over Zeolites 121. Csicsery, S.M. (1987) J. Catal., 108,

433–443. 122. Zones, S.I., Chen, C.Y., Corma, A., Cheng, M.T., Kibby, C.L., Chen, I.Y., and Burton, A.W. (2007) J. Catal., 250, 41–54. 123. Chen, C.Y. and Zones, S.I. (2007) Microporous Mesoporous Mater., 104, 39–45. 124. Musilov´a-Pavlaˇckov´a, Z., Kubu, M., Burton, A.W., Zones, S.I., Bejblov´a, M., ˇ and Cejka, J. (2009) Catal. Lett., 131, 393–400. ◦

125. Patra, C.R. and Kumar, R. (2002) J.

Catal., 212, 216–223. 126. Serra, J.M., Guillon, E., and Corma, A.

(2004) J. Catal., 227, 459–469. 127. Llopis, F.J., Sastre, G., and Corma, A.

(2004) J. Catal., 227, 227–241.

ˇ 128. Zilkov´ a, N., Bejblov´a, M., Gil, B., Zones, S.I., Burton, A.W., Chen, C.-Y., Musilov´a-Pavlaˇckov´a, Z., Koˇsov´a, G., and Eejka, J. (2009) J. Catal., 266, 79–91. ˇ 129. G´abov´a, V., Dˇedeˇcek, J., and Cejka, J. (2003) Chem. Commun., 1196–1197.

649

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves for the Conversion of Natural Gas to Fuels and Chemicals Agustı´n Martı´nez, Gonzalo Prieto, Andr´es Garcı´a-Trenco, and Ernest Peris

21.1 Introduction

The known worldwide natural reservoirs of natural gas would provide a potential supply for more than a century. Foreseeing a postpetroleum era, methane (from natural gas) stands as a very promising source of fuels and value-added platform (petro)chemicals. Its high H/C ratio and low heteroatom (O, N, S) contents are the main reasons for its attractiveness as feedstock for producing clean fuels, while functionalization is required to produce raw chemicals. As most of the abundant deposits of natural gas are situated in remote zones, far away from the populated regions where its use is required, an onsite conversion into transportable forms is the most suitable way to monetize these natural gaseous sources. Activation of methane is a very demanding task, due to the stability and symmetry of the molecule, having a C–H bond energy of 425 kJ mol−1 ; and catalysis is commonly required to work under technologically less expensive conditions. As schematically depicted in Figure 21.1, there are two main options to catalytically convert methane into fuels and chemicals. Direct routes involve a single-step conversion of methane, while in the indirect routes the carbon in methane is previously ‘‘stored’’ in a more reactive molecule such as CO, in synthesis gas or syngas (CO + H2 ), which then opens a much wider spectrum of conversion pathways. Among the direct conversion routes, nonoxidative paths directly convert methane into higher molecular weight products, having a lower H/C ratio, with the concomitant release of H2 , which is a valuable subproduct. Nevertheless, unfavorable thermodynamics hampers high per-pass yields in nonoxidative routes, and continuous selective removal of at least one reaction product (commonly H2 ) is required in order to increase the efﬁciency. Methane dehydroaromatization (MDA) and nonoxidative coupling of methane (NOCM) are the principal nonoxidative direct conversion routes. Alternatively, direct oxidative routes employ an external oxidant (commonly oxygen), which continuously removes hydrogen as water, providing a driving force to avoid thermodynamic limitations. In these cases, however, kinetic considerations also limit the maximum yields, as methane is commonly converted Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

650

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves MTG

Gasoline DTO

Diesel, lubricants

Olefins

DTG Dehydration

Hydrocracking Hydroisomerization

Waxes

DME Modified Methanol (bifunctional) STD synthesis FTS

Fischer–Tropsch Synthesis (FTS)

MTO Selective synthesis of olefins

Methanol

Syngas – Homologation (EtOH) – Carbonylation (CH3COOH) –Selective hydrogenation –Oxo-synthesis

Indirect routes Direct routes

Alcohols C2+ Carbox. acids C2+

Oxydative alkylation with CH4

Aromatics

MDA

CH4 -OCM -NOCM

Homologationof CH4 with olefins

Hydrocarbons C2+

POM

Methanol, Formaldehyde

Olefin MTG

oligomerization

Gasoline Olefin

Hydrocarbons C3+

oligomerization

Figure 21.1 Main direct and indirect (through syngas) catalytic routes from methane to fuels and raw chemicals.

into less refractory products, displaying weaker C–H bonds, and are thus more prone to undergo complete oxidation into COx when the reaction is run at high conversion rates. Avoiding the kinetic limitations in this case also requires the ‘‘chemical protection’’ or selective removal of the formed products from the reaction medium. Oxidative coupling of methane (OCM), oxidative methane ﬁxation by alkylation, and methane partial oxidation (MPO) are among these direct oxidative conversion routes. Generally, the direct routes are energy demanding and afford yields that are not currently economic, while the feasible modiﬁcations to be implemented in order to increase the per-pass yields are commonly technologically complex, keeping these processes, at the present level of development, out of proﬁt for a large-scale operation [1]. On the other hand, indirect catalytic routes employ the reactive CO in syngas as a building block in the synthesis of more complex products. Methane can be converted into syngas by steam reforming, dry reforming (using CO2 as coreactant), and partial oxidation with air. The steam-reforming process is the most widely employed process, and has been in operation since the 1930s. The steam-reforming

21.2 Direct Conversion of Methane

process is highly endothermic and constitutes a main energy sink in the indirect methane conversion pathways. However, its combination with the exothermic partial oxidation route in a hybrid process, which is commonly referred to as autothermal reforming, is possible and it reduces the energy duties of the syngas production. Indirect (syngas-based) conversion routes are currently signiﬁcantly more developed and lead to a much wider spectrum of products than the direct counterparts. At the moment, two processes such as the selective synthesis of methanol and the Fischer–Tropsch synthesis (FTS) of long-chain hydrocarbons are being proﬁtably operated at an industrial scale [2, 3]. Employing methanol and FTS-waxes as feedstocks, secondary catalytic processes create and cleave C–C bonds in order to upgrade them into fuels or open the spectrum of syngas-derived platform chemicals. Among all the catalytic routes schematized in Figure 21.1, acid, base, and metal catalyzed steps take place. During the last two decades, micro- and mesoporous molecular sieves have been endowed with a prominent role in some of these catalytic processes. The present chapter summarizes the role of zeolites (and zeotypes) as well as ordered silica mesostructures in providing well-deﬁned, active, and selective sites in the catalytic routes to convert methane (direct routes) or synthesis gas (indirect routes) into clean fuels and platform chemicals of industrial interest. How the chances of tuning the active sites’ nature and their spatial location in the coherent inorganic structure of the micro- and mesoporous molecular sieves allow increasing the selectivity and/or the yields in the methane conversion processes are discussed. Besides, though the catalytic role of zeotypes, in particular, is crucial in ‘‘secondary’’ processes such as the conversion of methanol to hydrocarbons or the upgrading of FTS primary products into fuels, these processes are out of the scope of the present chapter, as extensive compilations and reviews have been already reported. Thus, exclusively one-step synthetic processes are dealt with in detail.

21.2 Direct Conversion of Methane 21.2.1 Oxidative Conversion: OCM and Methylation Processes

The feasible combination of two molecules of methane under an oxidative atmosphere to yield C2 products became apparent in the early 1980s owing to the studies of Keller and Bashin [4], Baerns and coworkers [5], and Lunsford and coworkers [6]. The latter contributed through a series of works to demonstrate that ethane is formed through a mechanism involving an oxygen-mediated generation of surface CH3 radicals and a subsequent gas-phase combination, while ethylene is produced from ethane through a secondary oxygen-mediated radical route [7]. Best catalysts for the OCM are strongly basic, generally unsupported oxides comprising alkalines, alkaline earths, and lanthanides [7], while only a very marginal role is played by zeolites and related molecular sieves in this ﬁeld. Very few studies have applied

651

652

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

basic oxidic clusters supported on zeolites in the OCM [8]. On the other hand, the OCM reaction can be coupled with additional acid-catalyzed steps where zeolites have been proposed to catalyze oligomerization, isomerization, and/or dehydrocyclization of the C2 –C3 oleﬁns obtained by coupling of methane to produce either gasoline [9] or aromatics [10]. Besides oxidative self-coupling, methane can be ﬁxated in oleﬁns and aromatics through an oxidative cross-coupling route to yield a number of chemicals of great industrial signiﬁcance. In this sense, of special interest are the synthesis of propylene through methylation of ethylene and the production of ethylbenzene and styrene through methylation of toluene. Owing to the great industrial relevance of the products ethylbenzene and styrene, much research effort has been devoted to the oxidative methylation of toluene with methane after the initial experiments were ﬁrst reported by Khcheyan et al. [11] around 1980. Unlike in the OCM, zeolite-based catalysts have been explored in the oxidative alkylation of toluene with methane. The highly basic character required for the active sites led Kovacheva and coworkers to explore a number of alkali-loaded zeolites and microporous aluminophosphates in the oxidative alkylation of toluene with methane [12–14]. The reaction is commonly performed in a coﬂow mode, employing air as an oxygen carrier, at atmospheric pressure and temperatures in the range 600–800 ◦ C. Interestingly, industrially signiﬁcant side-chain alkylation products such as ethylbenzene and styrene are the major products. It has been concluded that extraframework alkaline (Li, Cs) species are the optimal active species, while Cs/AlPO-5 is the most suitable catalyst among those studied, yielding ethylbenzene and styrene with a selectivity of 75%, though at relatively low toluene conversions (<20%) [14]. An alternative high-pressure catalytic route for the alkylation of aromatics with methane was proposed by a number of researchers [15, 16], employing molecular sieves as catalysts, a high excess of CH4 , and high pressure (typically >4 MPa), at temperatures of 400–500 ◦ C, in a batch reactor conﬁguration. Though initially this high-pressure reaction was considered nonoxidative, later on, systematic works by Adebajo et al. [17–19] highlighted the crucial role of oxygen, unintentionally introduced in the autoclaves during the reactant feeding, in the reaction mechanism. Their experimental data ﬁrmly support a reaction mechanism, which indeed involves the partial oxidation of CH4 to methanol with oxygen as a stoichiometric reactant ﬁrst, and, subsequently, the alkylation of aromatics with the in situ generated methanol on the acidic catalysts (Figure 21.2). Several protic and metal exchanged zeolites as well as ordered mesoporous silicates have been tested in the high-pressure oxidative alkylation of benzene and toluene with methane [18–20]. Regarding benzene, it has been shown that molecular sieves bearing acidic sites of moderate strength are responsible for the alkylation step, employing the endo-generated methanol as alkylating agent. By contrast, when strong acidity is brought into play, as in the case of H-beta zeolite, an alternative reaction pathway is able to operate by using fragments derived from the cracking of benzene as alkylating agents, and, in this case, the reaction can proceed in the absence of oxygen [17].

21.2 Direct Conversion of Methane

CH4

+

1/2

+

CH3OH

O2

CH3OH

Figure 21.2 Proposed two-step reaction for the high-pressure oxidative alkylation of benzene with toluene [17].

The need for a moderate acidity is appraised from the results from Adebajo et al. [19, 20] in the alkylation of benzene with CH4 at a pressure of 6.9 MPa and a temperature of 400 ◦ C in a batch reactor. Thus, as shown in Table 21.1, MCM-41 mesoporous silica in its pure silica form or containing Cu do not hold acid sites with the required strength as to catalyze the reaction, while this mesoporous molecular sieve becomes active when it is synthesized as an aluminosilicate, owing to the presence of Brønsted acid sites. In any case, the catalytic activity displayed by the amorphous MCM-41 catalysts is substantially lower than that shown by the more acidic zeolitic catalysts such as ZSM-5, either in its protic form or exchanged with Co or Cu. According to the results from these authors, signiﬁcant production of ethylbenzene takes place. As side-chain alkylation of the primary product toluene is unlikely to occur through acid catalysis, it has been shown that this product actually forms as a result of benzene alkylation with contaminant ethylene existing in CP-grade ( > 99%) methane. Indeed, as also shown in Table 21.1, formation of ethylbenzene is largely suppressed when ultra-high-purity methane is employed as a reactant. Finally, using a more acidic catalyst like H-beta zeolite, both a higher benzene conversion and selectivity to ethylbenzene are obtained as compared to H-ZSM-5 at the same reaction conditions, which might be explained by the activation of the alternative reaction path involving alkylation of benzene with fragments derived from its acid-catalyzed cracking. The high-pressure acid-catalyzed oxidative alkylation of toluene with methane has also been reported, though, in this case, considerable yields of benzene and xylenes are obtained through the undesired toluene disproportionation [21]. The alternative processes based on alkylation of aromatics with methane have been claimed to compare favorably with the conventional alkylation with methanol in terms of process ‘‘greenness’’ [22]. Nevertheless, at the actual level of development, the extremely refractory character of methane leads to yields that are several times lower than those obtained with more reactive methanol as alkylating agent. The combination of methane-derived aromatics (such as benzene or naphthalene by MDA) with oxidative alkylation with methane might constitute a global catalytic route to important (petro)chemicals such as toluene or ethylbenzene. Current yields are, however, far from being competitive. Consideration of renewable methane produced by animal and vegetable metabolism would lower the cost of the oxidative methylation process and might lead to proﬁtable local solutions [22].

653

654

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves Table 21.1 Oxidative alkylation of benzene with methane over several molecular sieve-based catalysts.

Catalyst

H-ZSM-5a Co/ZSM-5 Cu/ZSM-5 MCM-41 AlMCM-41b CuMCM-41c H-ZSM-5d H-betad,e

Benzene conversion (%)

9.3 8.1 8.0 0.1 2.5 0.1 5.5 4.1

Product selectivity (%) Toluene

Ethylbenzene

Xylenes

65.7 48.1 41.4 100 56.0 75.9 93.0 89.8

17.4 31.5 37.7 – 28.2 24.1 4.0 7.4

15.5 18.5 18.5 – 15.8 – 2.9 2.8

= 70. = 30. c Si/Cu = 30. d Ultra-high-purity CH used as reactant. 4 e Si/Al = 56. Reaction conditions: T = 400 ◦ C, P = 6.9 MPa, CP-grade methane, reaction time = 4 hours, reactants loaded with no removal of residual oxygen [19, 20]. a Si/Al b Si/Al

21.2.2 Nonoxidative Methane Homologation and Alkylation Processes

Activating CH4 always requires relatively demanding conditions (i.e., high temperature). Thus, working under nonoxidative conditions is desirable in order to avoid side reactive paths leading to overoxidation products such as CO or CO2 . NOCM is a two-step process in which methane is initially dissociated on a metal catalyst at temperatures in the range 300–600 ◦ C, and the resulting metal-adsorbed carbonaceous species are later hydrogenated at lower temperatures (typically below 200 ◦ C) to yield C2+ hydrocarbons [23]. This cycling process ensures that the CHx (x = 1–3) species derived from CH4 activation are not aged at high temperatures, which promotes their progressive evolution toward poorly reactive (and deactivating) forms. Zeolite-hosted metallic clusters are employed in the two-step NOCM reaction as they provide very high metallic surface areas as well as coordinatively unsaturated surface sites required for CH4 activation at relatively soft conditions [24–26]. The speciﬁc connections between the properties of the metal clusters (size, composition, electronic density) and the catalytic behavior (C2+ selectivity) have resulted in the use of the NOCM reaction as an additional characterization technique for zeolite-hosted metal clusters. Using Co/NaY catalysts, Shen and Ma showed that the cluster size (i e., two vs three Co atoms) plays a crucial role in determining the catalytic selectivity [27]. Thus, the methane-derived CHx adspecies are more mobile and reactive on a Co2 cluster

21.2 Direct Conversion of Methane

than on the Co3 counterpart, showing enhanced conversion and selectivity to C2+ products through chain-growth events. The chemical composition of the metal clusters can also deeply modify their interaction with the intermediate CHx adspecies. In this respect, Co clusters have been found to be more efﬁcient than their equivalent Ru counterparts [28] due to the increased metal–C interaction in the latter, which hampers CHx recombination into C2+ products in the second step. In a similar way, a synergistic effect of Co and Pt was reported for bimetallic PtCo/NaY, favoring the formation of weakly bonded (more reactive) CHx intermediate species [29]. Finally, additional ligand-like effects due to the zeolitic host can also modify the catalytic behavior of the metal clusters in the homologation of methane [28]. Generally, though interesting at an academic level, this two-step nonoxidative process for CH4 conversion is not suitable for a large-scale operation. Besides CH4 homologation (self-coupling), nonoxidative cross-coupling of methane with oleﬁns and aromatics is catalyzed by metal-zeolite catalysts. In this respect, it has been demonstrated that the conversion of CH4 in the presence of ethylene or benzene in a continuous ﬂow system at temperatures in the range 300–500 ◦ C can be accomplished by using Ag and In-exchanged NH4 -ZSM-5 catalysts [30–32]. Continuous nonoxidative alkylation of benzene with methane has also been recently outlined, employing a bifunctional Pt/H-ZSM-5 [33], and it has been claimed the synergistic effect of both metallic and acidic sites in the formation of toluene. 21.2.3 Nonoxidative Methane Dehydroaromatization (MDA)

In 1993, the group of Wang at the Dalian Institute in China reported for the ﬁrst time the nonoxidative conversion of methane to benzene using transition metals, particularly Mo, supported on an acidic HZSM-5 zeolite [34]. Since then, several reviews on this subject have been published [35–38]. An additional advantage of the MDA process is the coproduction, besides aromatics, of H2 which might be used, for instance, in fuel cells. The nonoxidative conversion of methane to aromatics, typically abbreviated as MDA, is limited by thermodynamics with equilibrium conversions to a nearly equimolar mixture of benzene and naphthalene of about 12% at 700 ◦ C and atmospheric pressure, which are typical conditions for MDA. In the initial papers on MDA, methane conversions close to equilibrium (6–10% at 700 ◦ C) and almost 100% selectivity to benzene, the most desired aromatic MDA product, were reported over Mo/HZSM-5 catalysts. However, by improving the analytics of the reaction products and by introducing an inert gas as an internal standard (typically N2 ) together with methane in the gas feed, it was realized that benzene selectivities were lower than those initially reported (actually, about 60–70% on a carbon basis) and that a signiﬁcant part of the carbon atoms (about 20–30%) introduced as methane ended into carbonaceous deposits (coke) located on the catalyst surface [39, 40]. In fact, a fast catalyst decay, particularly during the initial reaction stages, due to the accumulation of carbon deposits on the Mo/HZSM-5 catalyst, is one

655

656

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

major drawbacks of the MDA process hindering its commercialization. On the other hand, naphthalene, the other major aromatic product of the MDA reaction besides benzene, was produced in amounts much lower than those predicted by the thermodynamics owing to the steric constraints imposed by the 10-ring channels of the ZSM-5 zeolite. Since these pioneering works, much effort has been devoted in the past few years toward improving the catalytic performance of zeolite-based MDA catalysts. The efforts have been mainly focused on understanding the role of acid sites, the nature and location of Mo species, and their speciﬁc interaction with the zeolitic acid sites, and the nature, location, and deactivating effect of different types of carbon deposits as an attempt to enhance the catalyst lifetime. Besides HZSM-5, other zeolite structures have also been applied as acidic support in the preparation of MDA catalysts. Generally, better catalytic performance has been observed when zeolites possessing a 2D pore system and pore sizes close to the dynamic diameter of benzene molecules are used, in particular, MCM-22 (MWW topology). Thus, the Mo/MCM-22 catalyst has been shown to be more selective to benzene than Mo/ZSM-5 owing to the peculiar topology of MCM-22 having pore entrances (0.40 × 0.59 nm in the sinusoidal interlayer 10-ring system, and 0.40 × 0.54 nm in the channel containing the 12-ring supercages) smaller than those in ZSM-5. Moreover, Mo/MCM-22 was also more coke tolerant (leading to enhanced lifetime) than Mo/HZSM-5 under similar reaction conditions, which has been related to the presence of the large 12-ring cavities in MCM-22 acting as coke reservoirs [41, 42]. Other zeolites belonging to the MWW family, such as MCM-49 [43], MCM-56 [44], and the delayered ITQ-2 material, particularly when the acid sites on the external surface are selectively removed by dealumination with oxalic acid [45], have also been explored for the MDA reaction with interesting results. Typically, Mo-based MDA catalysts are prepared by impregnation of the acid zeolite with a solution of the (NH4 )6 Mo7 O24 precursor, followed by calcination at temperatures of 500–700 ◦ C. Similar catalytic behavior was also observed by treating in air physical mixtures of MoO3 and zeolite. During the calcination step, the bulky molybdate species initially located on the external surface decompose into mobile MoOx species and at least a part of them migrates and disperses inside the zeolite channels via surface and gas-phase transport, exchanges at acid sites, and reacts to form H2 O [46]. By carefully measuring the amount of H2 O evolved and the amount of residual OH groups detected by isotopic equilibration with D2 , besides the requirement for charge compensation, Iglesia and coworkers proposed that the exchanged species consist of (Mo2 O5 )2+ ditetrahedral structures interacting with two cation exchange sites, as illustrated in Figure 21.3 [46]: When a Mo/HZSM-5 catalyst is contacted with the methane feed at the MDA conditions, the exchanged (Mo2 O5 )2+ species reduce and form MoCx /MoCx Ox clusters, which are believed to be the active sites for the activation of methane and formation of C2 intermediates (i.e., ethylene). In the initial reaction stages, an induction period is observed (corresponding to the reduction/carbidization of Mo6+ ions), during which almost no aromatics are produced. This period coincides with fast catalyst decay, as seen in Figure 21.4, adapted from [39]. After the induction

21.2 Direct Conversion of Methane

O

O Mo6+

657

O Mo6+

O

O O

O

Al3+

Al3+

Si4+

Figure 21.3 Structure of Mo species at exchange zeolitic sites, as proposed by Iglesia and coworkers [46]. 20 70 18 60

16

Product selectivity (%C)

CH4 conversion (%C)

14 12 10 8 6

50

COx C2+C3 Benzene Toluene Naphthalene

40

30

20 4 10 2 0 (a)

0

2

4

6 TOS (h)

8

10

0

12 (b)

0

2

4

6

8

TOS (h)

Figure 21.4 Methane conversion (a) and product selectivity (b) as a function of TOS for MDA reaction over 2 wt% Mo/ZSM-5 catalyst at 700 ◦ C, 1 atm, and GHSV = 800 hour−1 (adapted from [39]).

period, the C2 H4 intermediate is further converted into benzene and naphthalene through oligomerization, cyclization, and dehydrogenation reactions on the zeolite Brønsted acid sites. Thus, Mo/HZSM-5 catalysts are bifunctional in nature as both Mo carbide/oxycarbide and H+ are required to produce the desired aromatics. Direct evidence of the participation of the Brønsted acid sites in the formation of aromatics on Mo/HZSM-5 has been obtained by in situ 1 H MAS NMR spectroscopy under reaction conditions (700 ◦ C) [47].

10

12

658

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

In general, the best catalyst performance is attained for HZSM-5 samples with a Si/Al ratio of about 20–30 and intermediate Mo loadings (2–6 wt%). In this respect, Su et al. [48] found an optimum Mo/[H+ ] ratio per unit cell of about 1, giving maximum methane conversion and aromatics yield. Iglesia and coworkers, on the other hand, found an optimum CH4 aromatization rate for Mo/HZSM-5 samples with Mo/Al ratios of about 0.4 [46]. At Mo loadings exceeding those required to form a monolayer of MoOx on the external zeolite surface, part of the Mo sublimes and/or extracts Al from the zeolite framework, leading to the formation of extraframework Al2 (MoO4 )3 species that is easily detected by 27 Al MAS NMR [49]. This process is also favored at high calcination temperatures. Besides the Si/Al ratio and Mo loading, the size [50] and morphology [51] of the HZSM-5 crystallites also play a role in catalyst performance. As mentioned earlier, accumulation of carbonaceous deposits on the catalyst surface, leading to fast catalyst decay, is one of the major drawbacks of the MDA reaction. Different approaches have been applied in order to enhance the lifetime of the Mo/HZSM-5 catalyst. Some of them are based on the inhibition of coke formation through the addition of CO/CO2 [52, 53], O2 [54], and water [55] to the methane feed. It is worth mentioning here that addition of water may cause structural damage to the zeolite at the high temperatures applied in MDA. Other strategies that are habitually employed are the addition of a metal promoter, such as Pt [56], Fe, and Co [57], and, more typically, submitting the parent zeolite or Mo-containing catalyst to controlled postsynthesis treatments aimed at attaining a proper balance between the Mo and H+ sites in the bifunctional Mo/zeolite catalyst. In fact, it has been proposed that an ‘‘excess’’ of strong Brønsted acid sites promotes the formation of coke on the zeolite surface, poisoning the acid sites and blocking the pore entrances impeding the access to the active sites inside the pores. Effective postsynthesis treatments reported in the literature leading to a lower coking tendency and thus to enhanced catalyst lifetime are, for instance, dealumination by steaming [58], treatment with Al(NO3 )3 aqueous solution [59], treatment in basic solutions [60], partial exchange of H+ in HZSM-5 with alkali metal cations (i.e., Na+ , Cs+ ) [61], and surface silanation [62]. Recently, an improvement in catalyst durability has been reported for Mo/HZSM-5 catalyst in which the zeolite was synthesized following a carbon-template route using size-uniform Black Pearls (BP2000) as carbon primary particles leading to the formation of intracrystalline mesovoids (see representative TEM image in Figure 21.5a) [63]. The improvement in catalyst stability during MDA as compared to an equivalent Mo/ZSM-5 sample prepared from a commercial zeolite (Figure 21.5b) may be related to the presence of intracrystalline mesovoids that act as coke reservoirs (similar to the supercages in MCM-22), leaving a higher fraction of sites in the independent sinusoidal 10-ring system of channels active for the aromatization reaction. Despite the advances achieved in the past few years from the viewpoint of both catalyst performance and level of understanding of the MDA reaction on Mo/zeolite catalysts, much more intensive work would be required in order to improve single-pass benzene yields and catalyst durability for the process to have a real chance of industrial implementation. In order to accomplish this, improvements

21.3 Syngas Conversion Processes

50 nm

Total aromatics yield (%C)

6

659

Mo/ZSM-5-BP Mo/ZSM-5-ret

5 4 3 2 1

0

1

2

3

4

5

6

7

TOS (h)

Figure 21.5 Representative TEM image of ZSM-5 synthesized with BP2000 carbon particles (a), and evolution of the aromatics yield with TOS for the corresponding Mo/ZSM-5-BP catalyst and the reference one prepared from commercial zeolite with similar Si/Al ratio (Si/Al ∼ 25) and Mo loading (3 wt%) (b) [63].

in catalyst development should be accompanied by the design of appropriate integrated reactor/regeneration systems and cost-effective separations [38]. In this line, promising preliminary results have been reported when a H2 -selective transport membrane is used together with the Mo/ZSM-5 catalyst, allowing the removal of H2 product (i.e., by oxidation to water), thus overcoming the thermodynamic limitations for the formation of aromatics [64], though the practical application of this type of membrane technology is still far from being commercially viable.

21.3 Syngas Conversion Processes 21.3.1 Selective Synthesis of Short-Chain (C2 –C4 ) Oleﬁns

Light oleﬁns (C2 –C4 ) are very versatile platform chemicals, which are employed as building blocks in a large number of high-volume petrochemical processes toward derivatives such as fuels, specialties, and polymers. World production of ethylene has increased from 92.7 to nearly 110 million tons per year from 2000 to 2006 and is forecasted to increase at a rate of about 4–5% in the near future [65, 66]. Meanwhile, the demand for propylene derivatives (mainly polypropylene) has recently boomed and, consequently, the industry must face this increasingly prominent role of propylene within the light oleﬁn market [65]. On the other hand, butadiene production is also expected to increase over the next years, though at a lower rate than ethylene and propylene production.

8

9

660

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

Light oleﬁns are currently produced by pyrolysis of liqueﬁed petroleum gas (LPG) and ﬂuid catalytic cracking (FCC) of vacuum distillates. In addition, the increasing demand for propylene is prompting the development of technologies such as oleﬁn metathesis to boost propylene production from ethylene and butene [67]. As in the case of other base chemicals, the declining reserves of crude oil stimulate the search for alternative material sources such as natural gas, whose reserves are forecasted to remain unexhausted for more than a century. Nevertheless, most of the potential processes for the conversion of natural gas to light oleﬁns remain inferior to the crude-oil-related conventional technologies and, thus, they have scarcely been operated on a large scale. The exceptions are the widely established indirect routes through methanol (methanol-to-oleﬁns (MTO) [68]) or dimethyl ether (DME)-to-oleﬁns (DMO), whose costs and yields are competitive with conventional processes. Though zeolites and related ‘‘zeotypes’’ (such as silicoaluminophosphates (SAPOs)) play a key role in the MTO and DMO processes, these subjects have been already extensively reviewed [69–71] and are out of the scope of the present chapter, which is focused on one-step processes from methane (direct routes) or syngas (indirect routes) catalyzed by zeolites and other molecular sieves. In this sense, the selective synthesis of light oleﬁns can be accomplished directly from syngas [72]. The FTS, which is known from the early 1920s, has more recently assumed great industrial signiﬁcance as the core of the gas-to-liquids (GTLs) processes, which convert syngas into transportable fuels [73]. Though the GTL process is commonly run at conditions that favor liquid products, low molecular weight oleﬁns are also produced through the FTS. Tailoring the Fischer–Tropsch chemistry to selectively produce light oleﬁns represents a feasible natural gas–based catalytic route, which deserves academic and industrial attention. As a matter of fact, the direct conversion of syngas to light oleﬁns is currently a subject of interest for some leading chemical companies and, in this respect, Dow and S¨ud-Chemie have announced the constitution of a joint venture to develop and manufacture catalysts for the direct conversion of syngas to oleﬁns in January 2009 [74]. One of the greatest limitations of the Fischer–Tropsch chemistry is the operating polymerization mechanism, which leads to an unavoidable wide spectrum of products, following the so-called Anderson–Schulz–Flory (ASF) distribution [75], which comprises both oleﬁns and parafﬁns (O/P) with molecular weights from methane to waxes. In addition, though α-oleﬁns are primary products, readsorption, chain insertion, and hydrogenation secondary events tend to lower the overall selectivity to oleﬁns in favor of longer parafﬁnic products [76]. Thus, modifying the chemistry of CO hydrogenation to selectively produce light oleﬁns requires bypassing, or at least limiting, the chain-growth mechanism as well as avoiding readsorption and moderating the hydrogenation functionalities. Alkali and/or Mn-promoted Fe-based FTS catalysts, either bulk or supported on suitable carriers such as carbon, MnO or SiO2 , are known for their high selectivities toward oleﬁns [77]. The promotion by alkalines, essentially potassium, has been shown not only to increase the O/P ratio but also to shift the product pattern toward higher molecular weights owing to a decreased hydrogenating function [78, 79]. Reducing the size of the metal catalyst to the subnanometer range, conﬁning

21.3 Syngas Conversion Processes

the reaction inside constrained spaces [80], and intercepting reaction intermediates before they can undergo readsorption, chain insertion, or hydrogenation [81] have been suggested as feasible strategies to overcome the ASF product distribution and to selectively synthesize light oleﬁns. Owing to their hydrothermal stability, their crystalline microporous structure, displaying pores and cavities of molecular dimensions, and the chance for tuning their acid–base properties, zeolites have been employed as supports for heterogeneous catalysts applied in the selective conversion of syngas to light oleﬁns. As a result of its higher intrinsic selectivity to oleﬁns, iron has been the most widely chosen active metal for the preparation of zeolite-based catalysts to be applied in the selective production of light oleﬁns (mainly C2 –C5 ). Impregnation has been one of the most commonly employed preparative routes [82, 83], while other more sophisticated methods such as gas-phase metal-carbonyl adsorption have also been explored to produce well-deﬁned zeolite-hosted metal clusters [84, 85]. Most of these zeolite-based metal catalysts lack acidic sites in order to avoid secondary acid-catalyzed reactions of oleﬁns. Thus, zeolites, either in their pure silica forms, such as silicalite-1 (MFI), or as alkali-exchanged silicoaluminates, are preferred as supports. In this respect, Das et al. [86] applied (Mn)Fe and (Mn)Co catalysts supported on either HZSM-5 or silicalite-1 and found that, in spite of a higher reaction rate, the catalysts supported on the acidic form of the MFI zeolite display O/P ratios up to 2 orders of magnitude lower than the corresponding counterparts supported on silicalite-1, at the same reaction conditions. In agreement, Calleja et al. [87] found that progressively increasing the Si/Al ratio in MFI zeolites from 29 to ∞ (silicalite-1) increases the selectivity to C2 –C4 oleﬁns obtained on Fe/MFI catalysts. It has been suggested that the spatial location of iron species in the zeolitic support determines both their chemical nature and catalytic properties. It is well known that Fe carbides develop upon exposure of Fe-based catalysts to the syngas reductive atmosphere and are indeed the active species in the iron-catalyzed FTS [77]. Separating from the classical FTS behavior, which leads to long-chain products and low o/p, to attain high selectivity to short-chain oleﬁns, should, consequently, imply different active sites. Marchetti and coworkers [88, 89] employed Fe/K-LTL catalysts and found that larger iron species located on the external surface of the zeolitic support undergo carburization and display a classical Fischer–Tropsch behavior following the ASF product statistical distribution, while, by contrast, K-promoted Fe species hosted inside the zeolite microporous structure remain as very small metallic Fe0 clusters and show higher selectivity to light oleﬁns. These authors found that Fe/K-LTL catalysts show enhanced activity and selectivity to short alkenes as compared to equivalent catalysts supported on SiO2 , Al2 O3 , or C, remarking that higher conversion rates and selectivities to short-chain products are not incompatible, provided the active sites are conveniently tailored. The inconvenience of Fe carbides for a high selectivity toward light oleﬁns has also been proposed by other authors [86], though, in this case, small Mn-promoted iron oxide (FeOx ) clusters were proposed as the selective sites for producing oleﬁns.

661

662

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

Not only the porous structure of the zeolitic support but also the presence of certain promoters modiﬁes the nature and location of the Fe active sites, and serves to enhance their catalytic selectivity. Gallegos et al. [83] found that Cs+ ions occupying exchange positions at the hexagonal prisms of the LTL structure strongly interact with the Fe species, hampering their reduction, and leading to an eightfold increase in the O/P ratio as compared to the pristine potassic support. This increase in selectivity is, however, accompanied by a fourfold decrease in the reaction rate, pointing to the lower activity of the oleﬁn-selective FeOx sites. Ravichandran et al. [82] also remarked on the key role of Mn as a promoter in Fe/silicalite-1 catalysts. The addition of Mn enhances the dispersion of the Fe species, hampering carburization. According to their results of M¨ossbauer spectroscopy, the proportion of H¨agg-type Fe carbide (χ − Fe5 C2 ) is lower, in favor of Fe(III)Ox species, in those catalysts which show enhanced selectivity toward short oleﬁns. As shown in Table 21.2, promotion of Fe/silicalite-1 catalysts by Mn leads to an increase in the selectivity to C2 –C4 oleﬁns (20–46%) and the overall O/P ratio (up to a 50%), interestingly, with a simultaneous increase in catalytic activity (46%) at high Mn loading (10 wt%). The detrimental effect of the acidity in the zeolitic MFI support is also appraised from the results collected in Table 21.2, which show that the oleﬁn content in the product drops dramatically when the acidic ZSM-5 zeolite is used as support for the Mn–Fe bimetallic catalyst. Besides Fe, other metals that show activity in the hydrogenation of CO have been employed to prepare zeolite-supported clusters. Thus, Co [84, 90], Ru [91], Ir [92], and Os [85] clusters have been prepared through adsorption of metal carbonyls or ion exchange with ammonia-based complexes on NaY zeolite and have been tested in the conversion of syngas. In all the cases, a non-ASF product distribution is reported, with enhanced selectivities toward short-chain (C2 –C4 ) oleﬁns. In Table 21.2

Catalytic results in the conversion of syngas with (Mn)Fe/MFI catalysts.

Catalyst

CO conversion (%)

10%Fe/Sil-1a

5%Mn10%Mn10%Mn10%Fe/Sil-1a 10%Fe/Sil-1a 10%Fe/ZSM-5b

7.1

5.2

10.4

13.2

11.2 44.6 5.3 39.0 8.4

9.3 65.3 6.1 19.4 10.7

9.3 53.6 4.3 32.7 12.5

27.4 4.3 44.4 23.9 0.1

Hydrocarbon distribution (wt%) CH4 Oleﬁns C2 –C4 Parafﬁns C2 –C4 Hydrocarbons C5+ O/P ratio a Supported

on silicalite-1 zeolite. on ZSM-5 zeolite. Reaction conditions: T = 275 ◦ C, P = 2.1 MPa, GHSV = 1200 hour−1 , H2 /CO = 1, [86]. b Supported

21.3 Syngas Conversion Processes

particular, Nazar et al. [90] reported a high selectivity to butenes by employing Co/NaY catalysts. Finally, other works employ acidic zeolites as supports for Fe-based catalysts to selectively produce short oleﬁns from syngas. Conceptually, these catalytic systems should be seen as a modiﬁed (bifunctional) FTS, which combines FTS catalyst and an acidic function responsible for secondary reactions such as (hydro)cracking, (hydro)isomerization, cyclization, and so on. The details are not given here, as an additional section of the present chapter is devoted to this concept. It is remarkable, however, that in certain cases the acidic catalyst can be chosen to maximize the fraction of C2 –C4 alkenes in the products [93]. In general, the direct conversion of synthesis gas to light oleﬁns should be considered as a suitable alternative to the two-step route, which comprises the synthesis of methanol and the subsequent MTO process. In the latter case, around 30–35% of the total costs arise from the methanol synthesis step [72]; these costs would be avoided by employing a direct route. Nevertheless, the competitiveness of the direct synthesis of light oleﬁns from syngas seems still far, mainly due to selectivity issues. High selectivity to the desired oleﬁns has been attained by exclusively employing catalysts that show only modest activity (low conversion levels), and this would impose a high dilution of oleﬁns in the product stream and undesirably large recycles. In addition, CO hydrogenation catalysts should display an active lifetime of at least half a year, and there is still no proof that the catalytically selective zeolite-hosted subnanometric metal clusters would display enough hydrothermal stability as to retain their unique location and structure for so long under demanding reaction conditions. Thus, the development of newer catalysts showing high selectivity at higher conversion rates is required for the one-step route to become economically competitive. The breakthrough of this process requires attaining a higher catalytic activity while retaining the desired selectivity to light oleﬁns. In turn, this goal calls for catalysts bearing a high homogeneity in their active sites, and molecular sieves will probably play a main role in the search for this site-coherency. 21.3.2 Fischer–Tropsch Synthesis (FTS) 21.3.2.1 Conventional FTS The FTS is a well-known process catalyzed by different transition metals (Co, Fe, and Ru being the most active ones) leading to a mixture of liquid hydrocarbons that can be upgraded to produce high-quality synthetic fuels. In particular, Fischer–Tropsch (FT)-derived diesel has excellent blending properties as it contains almost no sulfur and (poly)aromatics and has an extremely high cetane number (CN> 70) owing to the parafﬁnic and linear nature of the FT products. FT processes aimed at producing diesel are typically operated at low temperatures (200–240 ◦ C) in the presence of a cobalt-based catalyst in order to maximize the formation of high molecular weight

663

664

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

n-parafﬁns (waxes) that are subsequently upgraded into high-quality diesel through selective hydrocracking. Conventional Co-based FT catalysts are typically prepared by impregnation of a porous inorganic oxide, among which SiO2 , Al2 O3 , and, to a lesser extent, TiO2 are the most frequently employed, with a cobalt precursor salt such as cobalt nitrate. After impregnation, the solid is dried and submitted to thermal activation (calcination and reduction or direct reduction) to obtain metallic cobalt (the active phase for FTS) dispersed on the support surface. An extensive review on the different methodologies applied in the preparation and characterization of cobalt FT catalysts has been recently published [94]. Despite previous works from Iglesia and coworkers, which showed that the intrinsic activity (typically referred to as turnover frequency or TOF) of Co0 sites remains invariant for particles above about 9 nm in size [95], recent results point toward a dramatic decrease in TOF for Co0 nanoparticles with sizes that are less than about 8 nm [96, 97]. Such a nonclassical particle size effect has been related to the formation of less active (or inactive) Coδ+ sites at the Co-support interface as a consequence of a Co0 nanoparticle ﬂattening taking place in the presence of syngas at the typical FT reaction temperatures [97]. In a very recent paper from the group of de Jong, by using steady-state isotopic transient kinetic analysis (SSITKA) and assuming a cuboctahedral geometrical model for the Co0 nanoparticles, the lower TOF of Co particles < 6 nm in size is ascribed to both blocking of edge/corner sites (whose proportion increases with decreasing particle size) and to a lower intrinsic activity of Co0 atoms at the small terraces [98]. Despite a different explanation offered by both groups to the same effect, it has to be taken into account that the model Co catalysts employed in their studies and the thermal activation treatments to which the catalysts are submitted are signiﬁcantly different, that is, Co/ITQ-2 previously calcined at 500 ◦ C before reduction in [97] and Co/CNF (CNF stands for carbon nanoﬁber) is obtained by direct reduction in [98]. Anyway, it is clear that from the point of view of maximizing catalyst activity it would be desirable to produce catalysts with Co0 dispersion as close as possible to the optimum and homogeneous particle size distribution at high metal loadings. This goal can hardly be attained by using conventional nonstructured supports having relatively low surface areas and wide pore size distributions. Therefore, ordered mesoporous silicas (OMSs) have attracted great interest in the last decade as supports for cobalt with the hope of being able to produce highly dispersed Co0 nanoparticles with controlled particle size and size distribution owing to their extremely high surface area (> 700 m2 g−1 ), their narrow pore size distribution, and the possibility to tune the size of the mesopores a priori by careful selection of the synthesis conditions [99, 100]. However, it was soon realized that the picture was not as simple as it initially seemed and that a careful control of the preparation and activation conditions is required in order to take full advantage of the particular properties offered by OMS when used as support for cobalt. More speciﬁcally, the most relevant parameters affecting cobalt dispersion are the method of cobalt incorporation, the nature of the cobalt precursor, the total metal loading, and the activation conditions applied to decompose the metal precursor [94, 101].

21.3 Syngas Conversion Processes

Different sophisticated methodologies have been applied to produce highly dispersed Co0 nanoparticles in OMS, such as direct incorporation into the synthesis gel [102], ion exchange of the cationic surfactant [103], and atomic layer chemical vapor deposition (CVD) from carbonyl precursors [104], though in general they do not allow to obtain relevant (≥ 10 wt%) Co loadings. Moreover, these methodologies typically lead to very small cobalt nanoparticles displaying a strong interaction with the silica walls which tend to form hardly reducible phases (e.g., different types of cobalt silicates). Consequently, impregnation (generally to incipient wetness) of the OMS support with a solution containing the cobalt precursor is the most simple and widely applied procedure. Nevertheless, even when prepared by impregnation, the nature of the solvent and cobalt precursor salt are factors that have to be taken into account as they may have a strong impact on the ﬁnal cobalt dispersion and, indirectly, on the reducibility of the supported cobalt oxide phases. For instance, the use of aqueous impregnation may produce a partial collapse of the mesostructure, particularly in the case of the less hydrothermally stable MCM-41 material [105], where the use of less polar solvents, such as ethanol, is preferred. By contrast, impregnation with aqueous solutions results in no signiﬁcant loss of pore ordering for the more hydrothermally stable SBA-15 mesoporous silica having thicker pore walls [105]. On the other hand, the nature of the cobalt precursor in the impregnating solution has an affect on the cobalt-support interaction strength and, thus, on the cobalt dispersion and reducibility, which, at the end, determines the number of active Co0 sites in the ﬁnal catalyst. Thus, impregnation of SBA-15 silica with a solution containing organic cobalt precursor salts, such as cobalt acetate and acetylacetonate, leads to very small but hardly reducible Co nanoparticles displaying very low FT activity and C5+ selectivity as compared to that prepared from cobalt nitrate at an equivalent Co loading (20 wt%) [106]. After impregnation and drying, the thermal activation conditions applied to decompose the metal precursor have been recently shown to be crucial for attaining a high dispersion of the formed oxidic phases inside the mesochannels. Thus, in Co/SBA-15 (18 wt% Co) prepared from the nitrate precursor, thermal activation in the presence of a ﬂow of inert gas (e.g., He) containing a small amount of NO (1 vol%) leads to smaller nanoparticles conﬁned within the mesopores and to catalysts displaying a signiﬁcantly higher speciﬁc FT reaction rate than the equivalent sample calcined in stagnant air and containing large metal particles on the outer SBA-15 surface [107]. Besides the above preparation parameters, the pore size of the OMS has been shown to largely determine the average size (thus the dispersion) of the conﬁned cobalt particles [108]. As a general trend, large-pore OMS silicas, such as SBA-15, lead to larger and more reducible Co particles and to catalysts displaying higher FT reaction rates and C5+ selectivity than small-pore mesostructures of the MCM-41 type [109]. Moreover, Co supported on the monodirectional MCM-41 and tridirectional MCM-48 structures with similar pore size leads to alike Co dispersions, thus emphasizing the key role of pore size rather than pore connectivity on the properties of the supported Co phases (A. Mart´ınez, C. L´opez, unpublished results). The effect of size and connectivity of the mesopores in OMS on the ﬁnal metal

665

666

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves Table 21.3

Physicochemical and catalytic properties of Co/OMS samples (20 wt%Co).

Catalyst

Co/MCM-41a Co/MCM-41b Co/MCM-48 Co/SBA-15

SBET

PD (BJH)

d(Coo)a

ER400 b

CO conversion

C5+ selectivity

(m2 /g)

(nm)

(nm)

(%)

(%)

(%C)

711 563 975 508

3.2 3.6 2.9 8.2

4.1 6.3 3.8 8.5

38 54 42 62

25.6 17.5 15.7 27.7

51.3 46.7 50.2 65.5

a Obtained

from d(Co3 O4 ) values derived from XRD. of reduction estimated by H2 -TPR in samples pre-reduced in pure H2 at 400 ◦ C for 10 hours. FT reaction conditions: T = 220 ◦ C, P = 20 bars, H2 /CO = 2, GHSV = 13.5 lsyngas /(gcat hours). b Extent

dispersion and reducibility, as well as on the catalytic properties of Co/OMS (about 20 wt% Co loading), is exempliﬁed in Table 21.3. As seen in the table, Co/MCM-41 and Co/MCM-48 samples having smaller pores (3–4 nm) are less active and C5+ selective than Co/SBA-15 (PD ∼ 8 nm). This is due to both, a lower reducibility of the supported Co3 O4 phase (lower ER values) and to a signiﬁcantly higher contribution of very small nanoparticles (below 6–8 nm in size) displaying a lower TOF, in the former samples. The reducibility of Co/OMS materials can be increased by applying different strategies, such as incorporation of Zr into the silica matrix [110], surface silylation [111], and by incorporating small amounts of easy reducible noble or seminoble metals, such as Pt, Pd, Ru, or Re. For instance, addition of 1 wt% Re to a 20% Co/SBA-15 sample increases the reducibility (at 400 ◦ C) from 62 to 96% with the corresponding beneﬁt in conversion (from 23 to 43%) and C5+ selectivity (from 65 to 74%C) [106]. Besides the Co0 particle size (and reducibility), diffusional issues are very important in FTS, as reported by Iglesia and coworkers in the [76]. In fact, a working catalyst has its pores completely ﬁlled with liquid hydrocarbons (waxes) and this imposes a barrier to the diffusion of reactants (especially for CO) and products (α-oleﬁns) from the gas phase to the active Co0 sites inside the pores and vice versa, respectively. For catalyst pellet sizes typically applied in ﬁxed-bed reactors, the CO diffusion rate through the liquid phase may control the selectivity to C5+ hydrocarbons by changing the local H2 /CO ratio around the Co0 particles. Moreover, a decrease in catalyst activity may be experienced under severe CO diffusion-controlled conditions. Therefore, it is likely that an enhanced CO diffusional barrier in small-pore MCM-type silicas as compared to SBA-15 might contribute to the lower activity and C5+ selectivity in the former catalysts. Even in the case of SBA-15, for which the mesopores are commonly highly curved at a mesoscopic scale, making the pore length almost double the size of the primary particles, the relatively large CO diffusion paths may negatively affect the catalyst activity. Therefore, in order to take full advantage of the properties offered by this

21.3 Syngas Conversion Processes

mesostructured silica as support for Co, it would be desirable to synthesize it with the appropriate pore size (leading to Co0 particles near the optimum size of 8–10 nm from the point of view of TOF) and short pore length. Some additional advantages with regard to Co dispersion and reducibility owing to a faster evacuation of deleterious NOx during decomposition of the metal nitrate precursor and of water generated in the reduction step, respectively, may also be envisaged when having short pores as compared to large and highly curved pores as in conventional SBA-15 silicas. Indeed, very recent results from our group show a clear beneﬁt in dispersion and FT catalytic activity when the length of the SBA-15 pores is decreased at a constant pore diameter of 11 nm [112]. Figure 21.6 shows representative TEM images for the long-curved and short-pore SBA-15 silicas and the mean Co0 particle size and catalytic properties for FTS of the corresponding CoRu/SBA-15 catalysts. It is worth mentioning that the Co time yield obtained for the catalyst based on the short-pore SBA-15 with PD = 11 nm is among the highest, to the best of our knowledge, reported up to now in the literature under similar FTS conditions. Contrary to the mesostructured silicas, zeolites are less suitable as supports for preparing Co-based catalysts for diesel-oriented FT processes as very small nanoparticles conﬁned within the micropores would be hard to reduce because of

500 nm

200 nm SBA-15 support textural properties Pore diameter = 11.1 nm Average pore length = 5.70 µm

Pore diameter = 11.2 nm Average pore length = 0.33 µm

RuCo/SBA-15 catalysts metal dispersion d(Co0) = 17.8 nm d(Co0) = 10.2 nm Catalytic properties (T = 220 ºC, P = 2.0 MPa,H2/CO = 2, Xco = 55%) Co-time-yield =154 × 10−3mol CO/gCo h Selectivity C5+ = 76.0 %C

Co-time-yield = 269 × 10−3mol CO/gCo h Selectivity C5+ = 81.9 %C

Figure 21.6 Representative TEM images and textural characterization of SBA-15 silica with long and curved pores (a) and short pores (b) and Co0 particle size and catalytic properties of the corresponding CoRu/SBA-15 catalysts for FTS.

667

668

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

a very strong metal-support interaction and, even if reduced, they would display very poor FT activity because of the particle size–TOF dependence discussed earlier. Moreover, the presence of micropores and Brønsted acid sites, associated with framework AlIV species and promoting cracking, would limit chain-growth processes, decreasing the selectivity to long-chain hydrocarbons that are the precursors for high-quality diesel. There is, however, a remarkable exception: the family of zeolites known as delaminated or delayered zeolites developed by Corma and coworkers at ITQ about 10 years ago, among which ITQ-2 and ITQ-6 are the most representative [113, 114]. These materials can be synthesized in their pure silica form (and thus are free from acid sites) and are characterized by a very high (> 600 m2 g−1 ) and accessible surface area and an open interparticle porosity, mostly in the mesopore range, associated with the random arrangement of the individual nanolayers. These properties make these peculiar class of zeolites very interesting as catalytic supports, in particular, for preparing Co-based FT catalysts. Thus, Co/ITQ-2 and Co/ITQ-6 catalysts (20 wt% Co loading) combine a good metal dispersion, with average metal particle size close to the optimum, especially in Co/ITQ-6, with short CO diffusion paths owing to their open porosity, resulting in catalysts that display a higher CO conversion rate and C5+ selectivity as compared to Co supported on a low surface area nonordered commercial SiO2 [115]. 21.3.2.2 Modiﬁed (Bifunctional) FTS An interesting potential application of zeolites in the production of high-quality fuels directly from syngas is the so-called modiﬁed FTS using hybrid or composite catalysts. These hybrid catalysts typically comprise a mechanical mixture of a FTS catalyst (either iron or cobalt based) and a solid acid (i.e., a zeolite) in order to promote the in situ conversion of primary FTS products (mostly α-oleﬁns and n-parafﬁns) into high-octane gasoline components through cracking, isomerization, oligomerization, and cyclization reactions occurring on the zeolitic Brønsted acid sites. The concept of the hybrid catalyst for modiﬁed FTS is illustrated in Figure 21.7. Although this principle has been shown to work, the main drawback of this approach is the relatively short lifetime of the acid component due to deactivation Fischer–Tropsch synthesis

CO + H2

In situ upgrading of primary FT products

α-Olefins Co or Fe based

Aromatics Zeolite

n-Paraffins

Branched olefins and paraffins

Figure 21.7 Schematic representation for the in situ upgrading of primary FTS products into high-octane gasoline components (the so-called modiﬁed FTS).

21.3 Syngas Conversion Processes

by formation of soft coke molecules (mostly alkylated mono- and diaromatics) inside the micropores [116]. Even if it has been shown that the stability of the zeolite can be signiﬁcantly improved by decreasing its crystallite size [117] or by the addition of small amounts of noble metals, such as Pt or Pd, that provide a fast hydrogenation of coke precursors [117, 118], the lifetimes are still far from those typical of FTS catalysts, and thus an efﬁcient regeneration of the spent zeolite in the presence of the FTS catalyst is not straightforward. Moreover, it has to be remarked that, in order to be effective, the acid cocatalyst must be able to produce the desired hydrocarbon rearrangements at the typical reaction conditions applied in conventional FTS. In this respect, the use of iron-based FT catalysts working at higher temperatures (250–350 ◦ C) than those based on cobalt (200–240 ◦ C) might be advantageous from the point of view of zeolite activity. Besides the nature of the FTS catalyst used (Fe vs Co), the ﬁnal structure and carbon-number distribution of the hydrocarbons produced are strongly affected by the topology, acidity, and crystallite size of the zeolite cocatalyst. Thus, for HZSM-5, it was shown that the extent of the acid-restructuring reactions was enhanced by increasing the zeolite acidity, that is, by increasing the framework Al content [117]. The zeolite topology has a great effect on the hybrid catalyst performance. In the case of KFeCo-zeolite hybrids, a higher gasoline selectivity and a lower deactivation rate was observed for the medium-pore HZSM-5 as compared to zeolites containing 12 rings or large supercages such as MCM-22, its delaminated counterpart ITQ-2, and ITQ-22 (comprising interconnected 8-, 10-, and 12-ring channels) [119]. Moreover, aromatization of the short-chain oleﬁns formed on the Fe-based FT catalyst was favored on HZSM-5, while gasoline-range isoparafﬁns formed by cracking of the heavier (C13+ ) n-parafﬁns were the predominant hydrocarbons initially formed on MCM-22 and ITQ-2. In the case of physical mixtures of a Co/SiO2 FT catalyst and an acid zeolite, it was found that the medium-pore HZSM-5 produced higher gasoline selectivities and iso-/n-parafﬁn ratios than large-pore USY, beta, and mordenite [116] as well as the complex IM-5 zeolite [120] because of a higher stability of the former against deactivation by coking. The extent of the acid-catalyzed reactions taking place on the zeolite and its stability with TOS also depends on the relative spatial location of the FTS and zeolite components in the hybrid systems. Thus, a more intimate contact between the zeolite and the FT catalyst, as in physical mixtures of powdered solids prior to shaping, may also alter the product distribution coming from the own FT component by intercepting the primary α-oleﬁns formed on the FT catalyst by the nearby acid sites, where they are isomerized and cracked, preventing their participation in chain-growing processes through readsorption at the FT metal sites [121]. Recently, the preparation of hybrid catalysts having a sophisticated core-shell structure has been reported [122]. In this approach, the FT catalyst (core) is homogeneously surrounded by a zeolitic shell, which is commonly developed through a hydrothermal synthesis on the preshaped FT catalyst pellet. The FT products formed in the core catalyst are forced to diffuse out through the zeolitic shell and thus have a higher probability for undergoing secondary acid-catalyzed reactions, as schematically shown in Figure 21.8. Core-shell hybrids comprising

669

670

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

Core (Fischer –Tropsch catalyst) CO + H2

Upgraded products Shell (Acidic zeolite)

Figure 21.8

Schematic representation of a core-shell hybrid FT-acid catalyst conﬁguration.

Co/Al2 O3 at the core and a beta zeolitic shell displayed an increased branching in C4+ parafﬁns and a lighter product distribution as compared to a physical mixture of the individual catalysts [122]. The relative size of the core and the shell in this hybrid conﬁguration has also been shown to inﬂuence the product selectivity by changing the residence time of the products through the FT core (α-oleﬁn readsorption and chain-growth) and through the zeolitic shell (chance for suffering acid catalysis) [123]. 21.3.3 Synthesis of Oxygenates 21.3.3.1 One-Step Synthesis of Dimethyl Ether (DME) from Syngas DME is a potential intermediate in the production of important chemicals, such as dimethyl sulfate, methyl acetate (a precursor to acetic acid), and lower oleﬁns. Moreover, it has also been increasingly used as an aerosol propellant to replace harmful chloroﬂuorocarbons (CFCs), which are known to have an ozone-depleting effect in the atmosphere, contributing to global warming. DME can also be used as a fuel additive and as a substitute for LPG for domestic heating/cooking purposes, and as a hydrogen carrier for fuel cells. Nonetheless, DME has recently attracted special interest because of its potential as an alternative fuel for compression-ignition engines, particularly for diesel-fueled engines, owing to its high CN, low auto-ignition temperature, and reduced emission of contaminants [124, 125]. Traditionally, DME is produced through a two-step process in which syngas is converted to methanol on a Cu-based catalyst in the ﬁrst step, and methanol is subsequently dehydrated to DME in the second step using an acid catalyst. In the ﬁrst step, the synthesis of methanol is limited by thermodynamics and becomes favored at high pressures and low reaction temperatures. Recently, a new route involving the direct (one-step) production of DME from syngas, generally known as syngas-to-dimethyl ether (STD), has attracted interest from both academia and industry as an alternative route to DME with higher economic value than the

21.3 Syngas Conversion Processes

traditional two-step process [126]. The main reactions involved in STD are the following: Methanol synthesis: CO + 2H2 ↔ CH3 OH

(21.1)

CO2 + 3H2 ↔ CH3 OH + H2 O (relevant for biomass-derived syngas)

(21.2)

Methanol dehydration reaction: 2CH3 OH ↔ CH3 OCH3 + H2 O

(21.3)

Water gas shift reaction (WGSR): CO + H2 O ↔ CO2 + H2

(21.4)

Therefore, the overall reaction can be written as follows: 3CO + 3H2 ↔ CH3 OCH3 + CO2

(21.5)

In the STD process, the methanol formed in reactions (Eqs. (21.1) and (21.2)) is consumed in the production of DME and water through reaction (Eq. (21.3)), shifting the equilibrium toward the right-hand side and increasing CO per-pass conversion; moreover, the water formed in reaction (Eq. (21.3)), which can slow down the methanol dehydration reaction, is consumed in reaction (Eq. (21.4)) to give CO2 and H2 (WGSR), which are reactants for the methanol synthesis reactor. The overall STD process is highly exothermic (H = −244.45 kJ mol−1 ), and this has to be considered for a proper reactor design in order to rapidly dissipate the heat of reaction, particularly when using ﬁxed-bed reactors in which the formation of hot spots, which may accelerate catalyst deactivation, should be avoided or minimized. From this standpoint, slurry bed reactors appear to be more suitable for STD, though they also exhibit some drawbacks mostly related to diffusional issues due to their complex hydrodynamics and higher costs. According to the set of reactions (Eqs. (21.1)–(21.4)) involved in STD, it becomes clear that STD catalysts must comprise a dual functionality, that is, a CO/CO2 hydrogenation function for the methanol synthesis step, and a methanol dehydration function. The hydrogenation function is generally given by a Cu-based methanol synthesis catalyst (MSC), typically CuO-ZnO-Al2 O3 , while acidic γ -Al2 O3 and zeolites are the most widely applied dehydration functions. As in any bifunctional catalysis, the two catalytic functions must work in complete synergy (i.e., to be balanced) in order to maximize DME productivity during the STD process. Typically, high catalytic activity can be obtained when the STD catalyst is prepared as a simple physical mixture of the hydrogenation and dehydration components, while lower activities have been reported for catalysts in which the methanol synthesis function is directly impregnated onto the acidic dehydration component or when both components are coprecipitated. However, good catalytic results have been reported for catalysts prepared by the so-called coprecipitating sedimentation method [127, 128]. In short, this method basically consists in adding

671

672

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

a suspension containing the MSC precursor, obtained previously by simultaneous coprecipitation of the individual components, into another suspension containing the solid acid dehydration component. Then, the mixture is stirred, ﬁltered, dried, and ﬁnally calcined [127]. This preparative route warrants a close contact between the two catalytic functions, enhancing the ‘‘synergetic effect’’ and leading to higher CO hydrogenation rates. Because the Cu-based MSC is industrially applied and is, therefore, highly optimized, most of the research efforts that are being devoted to the development of active, selective, and stable STD catalysts are focused on the optimization of the methanol dehydration component. In this respect, zeolites and related microporous molecular sieves have deserved special attention as the dehydration function in STD catalysts as they generally display higher catalytic activity than γ -Al2 O3 (bearing weaker acid sites), allowing to operate at lower reaction temperatures where the methanol synthesis step (Reaction (21.1)) becomes thermodynamically more favored. Moreover, as the temperature required in STD on alumina dehydration catalysts is typically above the optimum temperature required in the methanol synthesis step, the Cu-based component is more prone to deactivate, typically by sintering, and the overall activity of the bifunctional catalyst becomes lowered [129]. Among different zeolites, HZSM-5 has been, so far, the most widely investigated for the STD process [127, 128, 130], though good catalytic results have also been reported for other zeolites such as HY [127], H-mordenite (HMOR) [131], HMCM-22 [132], and H-ferrierite (HFER) [133], as well as for microporous SAPOs [134]. While there is a general consensus in the literature that the acidity of the zeolite is the most critical parameter determining its catalytic performance for the STD reaction, discrepancies appear regarding the required amount and strength of the acid sites in the zeolite dehydration function. In the case of alumina, it has been known since a long time that alcohol dehydration can take place on Lewis acid–base pairs generated on the alumina surface upon dehydration [135]. Regarding zeolites, different mechanisms involving a Brønsted acid site [136] or both a Brønsted acid site (H+ ) and its adjacent Lewis base site (O2− ) [137] have been proposed for DME formation from methanol. Part of the uncertainty concerning the inﬂuence of the zeolite acidity on the STD activity arises from the fact that in some of the studies the overall rate of syngas-to-DME reaction is controlled by the syngas-to-MeOH step, while in others the rate-determining step is that of the MeOH dehydration reaction. In the ﬁrst case, changes in the acidity of the dehydration catalyst may not result in appreciable changes in CO conversion but it mainly affects product selectivity, while in the latter case the level of CO conversion is strongly affected by the acidity, which at the end determines the MeOH-to-DME conversion rate by shifting the equilibrium of the syngas-to-MeOH step to the right side [138]. Nevertheless, it is possible to extract some conclusions from the results published in the literature regarding the zeolite acidity requirements for STD. For instance, Kim et al. [130] found that NaZSM-5 (being in principle free of H+ ) synthesized with low Si/Al ratio (30) was active for methanol dehydration, though its activity

21.3 Syngas Conversion Processes

was much lower than HZSM-5 with an equivalent Si/Al ratio, suggesting that both Brønsted (associated with framework Al species) and Lewis (related to Na+ cations) acid sites may be active for methanol dehydration, though the former would display a much higher intrinsic activity. On the other hand, these authors observed that the activity of HZSM-5 increased with decreasing Si/Al ratio (in the range of 30–100), which was related to a concomitant weakening in the acid strength, according to NH3 -temperature-programmed desorption (TPD) data. Nevertheless, the Brønsted acid sites in HZSM-5 samples are already relatively strong and one may not expect drastic changes in their acid strength when varying the Si/Al ratio in the 30–100 range, as all tetrahedral Al species in this range of composition would be completely ‘‘isolated’’ in the zeolitic framework. Thus, at variance with the results of Kim et al., it has been shown that a very high amount of strong Brønsted acid sites in HZSM-5 can be detrimental for the syngas-to-DME reaction because these sites are active for catalyzing the conversion of DME into hydrocarbons, as it occurs during methanol-to-gasoline (MTG) reactions, decreasing DME selectivity in the overall STD process [139–141]. Therefore, it is not surprising that different postsynthesis modiﬁcations have been applied to weaken the acid strength of the Brønsted sites in H-zeolite as an attempt to improve DME selectivity without paying a very high penalty on catalyst activity. In this sense, modiﬁcation of HZSM-5 (Si/Al = 38) with the appropriate amount of MgO (< 5 wt%) introduced by ‘‘dry’’ impregnation decreases the extent of formation of side products (CO2 and hydrocarbons) with the parallel increase in DME selectivity, as observed in Table 21.4 adapted from [142]. The improvement in DME selectivity was related to the decrease in the amount of strong acid sites catalyzing secondary DME reactions upon modiﬁcation of the zeolite with the optimum amount of MgO. Very interestingly, on the basis of their results, these authors proposed a mechanism for the dehydration of methanol on the MgO-modiﬁed HZSM-5 zeolite that involves, as the initial steps, a strong Inﬂuence of the modiﬁcation of HZSM-5 with MgO on the catalytic properties of Cu-Zn-Al/HZSM-5 (2/1 weight ratio) for the direct syngas-to-DME reaction.

Table 21.4

MgO content (wt%)

– 0.5 1.25 2.5 5.0 10.0

XCO (%)

95.8 96.3 96.0 95.6 67.6 64.7

Selectivity (C mol%) DME

MeOH

CO2

Hydrocarbons

49.1 64.5 64.4 64.1 21.4 15.5

4.5 4.6 4.8 4.8 48.9 53.3

37.1 30.5 30.7 30.9 29.4 30.9

9.30 0.37 0.08 0.12 0.32 0.19

Conditions: T = 260 ◦ C, P = 4 MPa, H2 /CO/CO2 = 66/30/4 (vol%), GHSV = 1500 ml/(g h).

673

674

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

adsorption of MeOH on an acidic site and a weak adsorption of alcohol on a neighboring basic site [142]. Modiﬁcation of the zeolite with other basic oxides, such as CaO [143], and by impregnation with solutions containing certain transition metals (e.g., Zr, Ni, Al) [131, 144, 145] leads to redistribution of the zeolite acid strength toward weaker sites with the consequent improvement in DME selectivity and on-stream stability. Other strategies that have been applied to reduce the strong acidity in the dehydration zeolite component and thus to improve DME selectivity (by suppressing secondary reactions) involve dealumination of the parent zeolite by steaming and acid treatments [132]. To conclude this part, it can be said that the one-step synthesis of DME from syngas (STD) is a promising alternative to the actual two-step processes and that zeolites and zeotypes are good candidates as the dehydration component of bifunctional STD catalysts. From the point of view of zeolite requirements, it appears that acidity is a more determinant parameter than topology, and that a very high amount of strong Brønsted acid sites (associated with framework Al) is detrimental from the standpoint of DME selectivity because they are also quite effective for catalyzing undesired DME secondary reactions. However, the role of extraframework Al species (i.e., Lewis acid sites), which can be generated during removal of the organic material by calcination of the as-synthesized zeolite or during steam dealumination treatments, on the catalytic behavior of the zeolite during STD is still not clearly understood, and a more systematic work in this direction would help in attaining an optimum zeolite dehydration catalyst. In fact, the conversion of methanol over HZSM-5 was shown to be sensitive to the presence of extralattice Al species, probably by forming Aln OCH3 species (detected by IR spectroscopy), which may enhance hydrocarbon formation reactions [136]. 21.3.3.2 Syngas to Higher (C2+ ) Oxygenates The selective synthesis of higher (C2+ ) oxygenates (i.e., alcohols, aldehydes, carboxylic acids, etc.) by catalytic conversion of syngas requires rather complex active sites (or multisite arrangements) affording the simultaneous dissociative activation of CO and hydrogen to give adsorbed hydrogenated alkyl species, Cx Hy , as well as the chain insertion of at least one nondissociatively activated CO molecule. It is known that certain organometallic complexes based on Co, Ru, or Rh are able to effectively catalyze the selective synthesis of oxygenates such as methanol, ethanol, and methylacetate from synthesis gas [146, 147]. In these homogeneous catalytic systems, the cationic metal sites are electronically tuned by the organic ligands, providing the required active sites for the insertion of oxygenated functions in the products, while the extreme homogeneity of the active sites allows for high selectivities. The preparation of active heterogeneous catalysts, which preserves the high efﬁciency toward oxygenates provided by the homogeneous organometallic catalysts remains a scientiﬁc challenge. In an attempt to mimic the highly selective metal sites present in the organometallic systems, the support and/or catalytic promoters can be employed to tailor the electronic properties of well-deﬁned metal sites, and therefore the way in which they interact with CO. In this sense, the concomitant occurrence of both dissociative and nondissociative CO activation is

21.3 Syngas Conversion Processes

inﬂuenced by the adsorption mode of the CO reactant on the metal sites. While linear and bridge adsorption modes have been reported on the surface of cobalt and iron-based Fischer–Tropsch catalysts, and thus these adsorption modes appear to be related to the dissociation of CO [148], the nondissociative adsorption of CO is related to a decreased electron back-donation from the metal sites to the adsorbate molecule. This can be achieved by the formation of bimetallic alloys causing rehybridization in d orbitals of some metals or due to the nearby presence of electron-deﬁcient promoters operating via charge polarization [149]. These electron-deﬁcient neighboring species might also act as Lewis acid sites and might promote the tilting of the metal-adsorbed CO molecule through stabilization of its electron-donating oxygen end. For promoted Rh catalysts [150, 151], tilted-adsorbed CO, as ascertained by in situ FTIR spectroscopy, has been directly related to the selective formation of higher oxygenates. The engineering of the active sites required for the selective synthesis of oxygenates from syngas can be approached by using zeolites as catalytic supports. Thus, metal clusters having a tailored size, electronic properties, and spatial location can be supported in the conﬁned microporous environment of the zeolite crystalline structure. Shen and Ichikawa showed that the composition of the metal clusters signiﬁcantly determines the catalytic selectivity in the hydrogenation of CO [152]. They found a synergetic behavior, which boosts the selectivity to higher oxygenates, principally ethanol, for bimetallic Rux Coy /NaY catalysts as compared to the monometallic counterparts, which produce mainly hydrocarbons. Besides the chemical composition of the supported clusters, zeolite-related Brønsted acid sites have a detrimental effect on the selectivity to oxygenates as they catalyze secondary undesired dehydration reactions [150, 153]. Owing to this, zeolites are commonly employed in an alkali-exchanged form when employed to prepare catalysts for the selective synthesis of oxygenates. Nevertheless, in certain cases, unintentional Brønsted acid sites might develop in situ by reaction of Lewis acid sites (exchange metal cations) with the water formed during the catalyst activation (namely reduction) or under reaction conditions. A careful choice of metal precursors and thermal activation protocols [153] or a neutralization step after the catalyst reduction [150] are required to afford high selectivity toward oxygenates in these cases. The methodological route employed during the synthesis of the catalyst also plays a determinant role in the spatial location, chemical properties, and, ultimately, the catalytic behavior of the zeolite-hosted metallic clusters. In this sense, the exposure or not to moisture during the thermal activation of PdCo/NaY bimetallic catalysts [154] and the pH at which the ion exchange process is performed to synthesize zeolite-entrapped Fe clusters [155] have been shown to strongly determine whether the catalyst displays a Fischer–Tropsch-like behavior (high selectivity to hydrocarbons) or selectively produces oxygenates in the catalytic conversion of synthesis gas. Rhodium is the most widely used active metal to produce higher oxygenates. The presence of metal oxide promoters (Fe, Mn, Ce, V, Nb) in close contact

675

676

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

with Rh is indispensable to ensure a high selectivity to the higher oxygenates, as selective sites have been claimed to be formed at the contact boundary between rhodium and the electron-withdrawing promoter [156]. This promotion effect has been localized in bimetallic clusters occluded in the nanosized zeolitic cavities as in the case of RhFe/NaY catalysts, which display moderately high selectivity (up to 15 mol%) to ethanol, though at very low CO conversions of about 1% [157]. Additionally, unpromoted Rh/NaY catalysts can form Rh carbonyl clusters under reaction conditions (250 ◦ C, 100 bar, H2 /CO = 1), showing high selectivity to acetic acid (up to 40%), again at modest CO conversions below 2% [158]. Thus, despite the academic interest in zeolite-entrapped (promoted) Rh clusters, ideally displaying deﬁned and very homogeneous active sites, allowing to gain fundamental knowledge on the site requirements, their modest catalytic activities compare disfavorably with other Rh-based catalytic systems comprising promoted Rh nanoparticles supported on amorphous SiO2 [159] or CNTs [160], which show higher time yields to C2+ oxygenates. To date, no heterogeneous-catalyzed commercial processes for the direct selective synthesis of C2+ oxygenates from syngas are being operated, and zeolite-supported metal clusters are not considered among the promising catalyst alternatives. 21.3.3.3 Carbonylation of MeOH and DME About 60% of the worldwide acetic acid production is commercially accomplished through methanol carbonylation using homogeneous Rh or Ir complexes in the presence of an iodide cocatalyst at low temperatures and atmospheric pressure, as it is performed in the Monsanto and BP CativaTM processes [161]. Acid zeolites are also known to catalyze the carbonylation of alcohols and ethers at mild conditions [162–164], prompting the development of a greener halide-free heterogeneous catalytic route to produce important chemicals, such as acetic acid. Carbonylation of DME to produce methyl acetate, a precursor to acetic acid, presents some advantages over the direct carbonylation of methanol to acetic acid, in part, because the water generated from the alcohol through parallel dehydration reactions is known to strongly inhibit carbonylation processes and may also adsorb competitively on CO binding sites. The kinetically relevant steps in the DME carbonylation on acid zeolites involve reactions of gas phase and/or adsorbed CO with methyl-saturated surface species (initially formed by the reaction of DME with a zeolite Brønsted acid site) to form adsorbed acetyl intermediates. Then, in the ﬁnal reaction steps, the acetyl moieties react with DME to produce methyl acetate and restores the surface methyl species [165, 166]. Interestingly, H-MOR and, though less active, H-FER, both possessing eight-ring zeolite channels, displayed outstanding selectivity (> 99%) and stability for DME carbonylation at low temperatures (150–190 ◦ C) [165]. Following this observation, Iglesia and coworkers presented unambiguous evidence for the much higher reactivity of zeolite protons located within the eight-ring channels toward CH∗3 –CO reactions (the kinetically relevant step in DME carbonylation) as compared to H+ present in the 12- and 10-ring channels of H-MOR and H-FER, respectively [167]. In fact, these authors found a good linear correlation between the methyl acetate synthesis rate and the number of H+

21.3 Syngas Conversion Processes 2.0 12-ring channels (MOR)

Methyl acetate synthesis rate (10−3 mol (g h−1))

8-ring channels (MOR) 8-ring channels (FER)

1.5

1.0 y = 1.52× R2 = 0.99

0.5

Co FER

0.0 0.0

BER, MFI, USY SiO2-Al2O3

0.1 0.2 0.3 0.4 0.5 0.6 0.7 Number of sites (10−3 mol H+ g−1)

0.8

Figure 21.9 DME carbonylation rates per unit mass against the number of H+ sites per unit mass in 8-ring channels of MOR (Si/Al = 10, Zeolyst) and FER (Si/Al = 33.5, Zeolyst) and 12-ring channels of MOR. Reaction conditions: 438 K, 3.34 cm3 s−1 g−1 , 0.93 MPa CO, 20 kPa DME, 50 kPa Ar (adapted from [167]).

sites per unit mass in the eight-ring channels of MOR and FER, while no evident correlation was seen when the total number of H+ in the respective zeolites were considered (Figure 21.9, adapted from [167]). Such unique speciﬁcity of zeolite protons in eight-ring channels for CH∗3 –CO reactions was suggested to arise from their ability to stabilize the acetyl-like ionic transition states when conﬁned within the restricted space of the narrow eight-ring pores [167]. However, in a recent theoretical work, the group of Corma proposed that the unique selectivity of H-MOR for MeOH/DME carbonylation is not only due to the size of the eight-ring channels but also due to the speciﬁcity of a particular site within the channel [168]. In particular, these authors proposed that, among the two possible positions (T3-O31 and T3-O33), only the T3-O33 is selective in the carbonylation reaction because of the unusual orientation of the methoxy group lying parallel to the cylinder axis of the eight-ring channel, as schematically shown in Figure 21.10, adapted from [168]. Only in this situation, the transition state formed by the attack of CO to the adsorbed methoxy group ﬁts perfectly within the eight-ring channels of MOR, thus constituting one of the most striking examples of enzyme-like speciﬁcity in zeolites. From the viewpoint of a commercial process, it would be more advantageous to perform the direct carbonylation of methanol to acetic acid instead of starting from DME. However, owing to its much lower reactivity as compared to DME and the mentioned rate-inhibition effect of water, methanol carbonylation would require the use of higher reaction temperatures that are more favorable for the

677

678

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

O C O

C

C

O C

O (a)

(b)

Figure 21.10 Schematic representation of the relative orientation of the Oframework − CH3 bond and the channel axis at (a) the T3-O33 position of MOR and (b) any other position in an eight-ring channel (adapted from [168]).

formation of hydrocarbons with the consequent negative impact on acetic acid selectivity [169]. Very interestingly, Corma and coworkers found that the methanol carbonylation rate could be enhanced by designing a metal-acid zeolite bifunctional catalyst (Cu-HMOR) comprising an active site composed of two adjacent sites, a Brønsted acid site (bridged OH group), and a Cu+ site located in a next nearest neighbor aluminum framework [170]. By using IR operando and in situ MAS NMR spectroscopy, these authors provided experimental evidence for a bifunctional carbonylation mechanism occurring on Cu-HMOR in which methanol is activated on the zeolite Brønsted acid site to form an adsorbed methoxy species, while CO is activated on an adjacent Cu+ site. Besides activation of CO, Cu+ sites were also shown to adsorb DME (formed from MeOH upon reaction with the zeolite H+ ) preferentially to MeOH and water. Then, the Cu+ –CO complex and the adjacent methoxy group react to form an adsorbed acylium cation intermediate (CH3 CO+ ), which subsequently reacts with the adsorbed DME to produce methyl acetate as the primary carbonylation product, in contrast to the preferential formation of acetic acid when the reaction takes place on a purely acidic H-MOR [170]. Despite the fact that the above works from the groups of Iglesia and Corma constitute outstanding contributions to the fundamental knowledge of the site requirements and reaction mechanisms of zeolite-catalyzed MeOH/DME carbonylation reactions, thus providing valuable information for the design of improved catalysts, the reported activities are still far from those that would be required for hypothetical process commercialization.

21.4 Summary and Outlook

Direct catalytic conversion of methane into fuels and chemicals requires a breakthrough to become industrially attractive. The few possible products and the limited yields attained at the current development level are not competitive against the indirect syngas-based catalytic routes. Combining reaction and selective membrane-based separation technologies is a promising aspect, which

21.4 Summary and Outlook

might relaunch catalytic processes such as OCM or nonoxidative MDA, bringing them closer to industrial competitiveness. Zeolites have played a marginal role in most of the oxidative pathways for the direct conversion of methane, though they play a key role in nonoxidative MDA. In this respect, though the last decade has witnessed a marked increase in the knowledge of the strict site requirements to attain high selectivity and improved catalyst lifetime in MDA, a breakthrough in the process is not expected to be related to further improvements in the catalyst but to a novel design of the process, overcoming the limits that the thermodynamics imposes to the per-pass yields. Some other direct methane conversion pathways, where molecular sieves might play an important role, are to be further developed, though in a longer term. In nature, the paradigm of selective methane conversion into liquid products at mild conditions is exempliﬁed by enzymes such as methane monooxygenase (MMO), existing at the membrane of methanotrophic bacteria such as Methylococcus capsulatus (Bath, England). This enzyme displays di-iron active sites capable of selectively oxidizing methane to methanol with molecular oxygen at mild conditions. Scientists have developed organometallic iron complexes, as mimics of MMO, which show activity in the selective oxidation of methane [171, 172], especially when conﬁned in inorganic mesoporous solids [172]. In the future, molecular sieves might help scientists in bridging the gap between homogenous and heterogeneous catalysis and thus prepare novel materials displaying a high density of well-deﬁned, biomimetic, organometallic active sites belonging to a crystalline open porous architecture, which might show activity in the selective oxidation of methane under mild conditions. Indeed, the preparation of ideally ‘‘single-site’’ heterogenous catalysts by the successful incorporation of organometallic active centers in metal-organic frameworks (MOFs) porous structures has been recently reported [173–176]. The intensive work that is in progress worldwide in developing novel metal-organic or hybrid organic–inorganic molecular sieves, in some cases with enhanced thermal stabilities, such as the recently discovered zeolite imidazolate frameworks (ZIFs) [177], might be envisaged to open a ﬁeld plenty of possibilities in this sense. Additionally, photocatalytic routes can also be applied to convert methane into useful chemicals, but, at the moment, these processes are at an early stage of development [178], and the singular features of molecular sieves might be useful in preparing novel photocatalysts having well-deﬁned active sites. On the other hand, micro- and mesoporous molecular sieves have been protagonists in the indirect catalytic routes, which convert syngas into fuels and raw chemicals. In this respect, acidic zeolites and other crystalline microporous zeotypes have been explored in one-step hybrid processes such as the modiﬁed FTS, which produces branched hydrocarbons in the gasoline range or the syngas-to-DME process (STD). Future work is expected on these subjects in order to improve the selectivity and the catalyst durability, which, to date, are the main drawbacks in this intensiﬁed one-step process as compared to the two-stage conventional ﬂow-schemes. Besides the well-established Fischer–Tropsch and methanol synthesis, other interesting routes from syngas such as the selective synthesis of short

679

680

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves

oleﬁns, the synthesis of higher oxygenates, or the heterogeneous selective carbonylation of methanol or DME with CO might become feasible as large-scale processes in the future. Their short- and medium-term development lies, in most of the cases, on the potential of inorganic molecular sieves to allow for nearly single-site selective catalysts, owing to their unique structural coherence. In particular, a good example of site-selectivity has been recently found for the carbonylation of methanol or DME on mordenite zeolites. These ﬁndings point to a very speciﬁc site, located in the eight-ring zeolitic channels, as the only one capable of selectively carrying out the bimolecular reaction. Beyond this knowledge, on-going and future developments in the synthesis of novel zeolitic structures or in the control on the site location in known zeolitic materials might be envisaged as future drivers for the commercial competitiveness of the process. In line with this concept, the rising number of zeolite and related zeotype structures will surely constitute short- to medium-term drivers for the development of novel, more selective, ‘‘secondary’’ processes for the upgrading of primary products obtained from syngas such as methanol (or DME) and Fischer–Tropsch products. A very recent example, which might serve to justify this prediction, is the discovery of a remarkable speciﬁcity displayed by some large-pore zeolites, especially H-BEA, in the solid acid-catalyzed homologation of DME to the high-octane triptane (2,2,3-trimethylbutane) [179]. Unraveling the active site requirements for selective conversions such as this one, assisted by the increasing capacity to create novel zeotype structures, will probably lead to new or improved processes in the syngas-based indirect routes to convert natural gas into liquid fuels or platform chemicals.

Acknowledgments

Financial support by the Comisi´on Interministerial de Ciencia y Tecnolog´ıa (CICYT) of Spain through the Project CTQ2007-66614/PPQ is gratefully acknowledged. G.P and A. G-T. thank the Ministerio de Ciencia e Innovaci´on of Spain for their Ph.D. scholarships.

References 1. (a) Lunsford, J.H. (2000) Catal. Today,

2. 3. 4. 5.

63, 165; (b) Holmen, A. (2009) Catal. Today, 142, 2. Tijm, P.J.A., Waller, F.J., and Brown, D.M. (2001) Appl. Catal. A, 221, 275. Dry, M.E. (2001) J. Chem. Technol. Biotechnol., 77, 43. Keller, G.E. and Bhasin, M.M. (1982) J. Catal., 73, 9. Hinsen, W., Bytyn, W., and Baerns, M. (1984) Proceedings 8th International

6.

7. 8. 9.

Congress on Catalysis, Berlin, 3, p. 581. Ito, T., Wang, J.-X., Lin, C.-H., and Lunsford, J.H. (1985) J. Am. Chem. Soc., 109, 5062. Lunsford, J.H. (1995) Angew. Chem. Int. Ed., 34, 970. Kovacheva, P. and Davidova, N. (1994) React. Kinet. Catal. Lett., 53, 277. Choudhary, V.R., Sansare, S., and Chaudhari, S.T. (1994) US Patent 5336825.

References 10. Qiu, P., Lunsford, J.H., and Rosynek, 11.

12.

13.

14.

15.

16.

17.

18. 19. 20. 21. 22. 23.

24. 25. 26. 27. 28.

29. 30. 31.

M.P. (1997) Catal. Lett., 48, 11. Khcheyan, K.E., Shatalova, A.N., Borisoglebskaya, A.V., and Fishman, L.I. (1976) Zh. Org. Khim., 12, 467. Arishtirova, K., Kovacheva, P., and Davidova, N. (1998) Appl. Catal. A, 167, 271. Kovacheva, P., Arishtirova, K., and Davidova, N. (1999) React. Kinet. Catal. Lett., 67, 261. Kovacheva, P., Arishtirova, K., and Davidova, N. (1999) Appl. Catal. A, 178, 111. He, S.J.X., Long, M.A., Attalla, M.I., and Wilson, M.A. (1992) Energy Fuels, 6, 498. He, S.J.X., Long, M.A., Wilson, M.A., Gorbaty, M.L., and Maa, P.S. (1995) Energy Fuels, 9, 616. Adebajo, M., Long, M.A., and Howe, R.F. (2000) Res. Chem. Intermed., 26, 185. Adebajo, M.O., Howe, R.F., and Long, M.A. (2001) Catal. Lett., 72, 221. Adebajo, M., Long, M.A., and Frost, R.L. (2004) Catal. Commun., 5, 125. Adebajo, M.O. and Frost, R.L. (2005) Energy Fuels, 19, 783. Adebajo, M.O., Howe, R.F., and Long, M.A. (2001) Energy Fuels, 15, 671. Adebajo, M.O. (2007) Green Chem., 9, 526. Koerts, T., Leclercq, P.A., and van Santen, R.A. (1992) J. Am. Chem. Soc., 114, 7272. Wei, J. and Iglesia, E. (2004) J. Catal., 225, 116. Wei, J. and Iglesia, E. (2004) Angew. Chem. Int. Ed., 43, 3685. Liu, Z. and Hu, P. (2003) J. Am. Chem. Soc., 125, 1958. Shen, J.G.C. and Ma, Q. (2004) Catal. Lett., 93, 19. Shen, J.G.C., Liu, A.M., Tanaka, T., and Ichikawa, M. (1998) J. Phys. Chem. B, 102, 7782. Guczi, L., Sarma, K.V., and Bork´o, L. (1996) Catal. Lett., 39, 43. Baba, T. and Sawada, H. (2002) Phys. Chem. Chem. Phys., 4, 3919. Baba, T. and Abe, Y. (2003) Appl. Catal. A, 250, 265.

32. Baba, T., Abe, Y., Nomoto, K., Inazu,

33. 34.

35. 36. 37. 38. 39. 40. 41. 42. 43.

44.

45. 46.

47.

48.

49.

50.

51.

K., Echizen, T., Ishikawa, A., and Murai, K. (2005) J. Phys. Chem. B, 109, 4263. Lukyanov, D.B. and Vazhnova, T. (2009) J. Mol. Catal. A, 305, 95. Wang, L., Tao, L., Xie, M., Xu, G., Huang, J., and Xu, Y. (1993) Catal. Lett., 21, 35. Xu, Y. and Lin, L. (1999) Appl. Catal. A, 188, 53. Xu, Y., Bao, X., and Lin, L. (2003) J. Catal., 216, 386. Shu, Y. and Ichikawa, M. (2001) Catal. Today, 71, 55. Skutil, K. and Taniewski, M. (2006) Fuel Process. Technol., 87, 511. Wang, D., Lunsford, J.H., and Rosynek, M.P. (1997) J. Catal., 169, 347. Wang, D., Lunsford, J.H., and Rosynek, M.P. (1996) Top. Catal., 3, 289. Ma, D., Shu, Y., Cheng, M., Xu, Y., and Bao, X. (2000) J. Catal., 194, 105. Shu, Y., Ma, D., Xu, L., Xu, Y., and Bao, X. (2000) Catal. Lett., 70, 67. Wang, D.Y., Kan, Q.B., Xu, N., Wu, P., and Wu, T.H. (2004) Catal. Today, 93-95, 75. Xing, H., Zhang, Y., Jia, M., Wu, S., Wang, H., Guan, J., Xu, L., Wu, T., and Kan, Q. (2008) Catal. Commun., 9, 234. Mart´ınez, A., Peris, E., and Sastre, G. (2005) Catal. Today, 107-108, 676. Borry, R.W.III, Kim, Y.H., Huffsmith, A., Reimer, J.A., and Iglesia, E. (1999) J. Phys. Chem. B, 103, 5787. Ma, D., Shu, Y., Zhang, W., Han, X., Xu, Y., and Bao, X. (2000) Angew. Chem. Int. Ed., 39, 2928. Su, L., Yan, Z., Liu, X., Xu, Y., and Bao, X. (2002) J. Nat. Gas Chem., 11, 180. Liu, W., Xu, Y., Wong, S., Qiu, J., and Yang, N. (1997) J. Mol. Catal. A, 120, 257. Zhang, W., Ma, D., Han, X., Liu, X., Bao, X., Guo, X., and Wang, X. (1999) J. Catal., 188, 393. ¨ Sarioglan, A., Savasci, O.T., Erdem-Senatalar, A., Tuel, A., Sapaly, G., and Ben Taˆarit, Y. (2007) J. Catal., 246, 35.

681

682

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves 52. Liu, S., Wang, L., Dong, Q., Ohnishi,

53. 54.

55. 56. 57.

58.

59. 60. 61.

62. 63.

64.

65. 66.

67. 68.

69. 70. 71. 72. 73.

R., and Ichikawa, M. (1998) Chem. Commun., 1217. Lacheen, H.S. and Iglesia, E. (2005) J. Catal., 230, 173. Yuan, S., Li, J., Hao, Z., Feng, Z., Xin, Q., Ying, P., and Li, C. (1999) Catal. Lett., 63, 73. Liu, S., Ohnishi, R., and Ichikawa, M. (2003) J. Catal., 220, 57. Chen, L., Lin, L., and Xu, Z. (1996) Catal. Lett., 39, 169. Liu, S., Dong, Q., Ohnishi, R., and Ichikawa, M. (1997) Chem. Commun., 1445. Lu, Y., Ma, D., Xu, Z., Tian, Z., Bao, X., and Lin, L. (2001) Chem. Commun., 2048. Song, Y., Sun, C., Shen, W., and Lin, L. (2007) Appl. Catal. A, 317, 266. Song, Y., Sun, C., Shen, W., and Lin, L. (2006) Catal. Lett., 109, 21. Mart´ınez, A., Peris, E., and Vidal-Moya, A. (2008) Stud. Surf. Sci. Catal., 174B, 1075. Ding, W., Meitzner, G.D., and Iglesia, E. (2002) J. Catal., 206, 14. Derewinski, M., Burkat-Dulak, A., Mart´ınez, A., and Peris, E. (2009) communication presented at the 6th World Congress on Catalysis by Acids and Bases (ABC-6), May 10–14, 2009, Genova . Kinage, A.K., Ohnishi, R., and Ichikawa, M. (2003) Catal. Lett., 88, 199. Plotkin, J.S. (2005) Catal. Today, 106, 10. Khadzhiev, S.N., Kolesnichenko, N.V., and Ezhova, N.N. (2008) Petrol. Chem., 58, 325. Mol, J.C. (2004) J. Mol. Catal. A: Chem., 213, 39. Kaeding, W.W. and Butter, S.A. (1975) US Patent 3911041, assigned to Mobil Oil. Chang, C.D. (1983) Catal. Rev., 25, 1. Hutchings, G.J. and Hunter, R. (1990) Catal. Today, 6, 279. St¨ocker, M. (1999) Microporous Mesoporous Mater., 29, 3. Janardanarao, M. (1990) Ind. Eng. Chem. Res., 29, 1735. Dry, M.E. (2002) Catal. Today, 71, 227.

74. http://www.sud-chemie.com (accessed

January 2010). 75. Friedel, R.A. and Anderson, R.B. (1950)

J. Am. Chem. Soc., 72, 1212. 76. Iglesia, E., Reyes, S.C., and Madon, R.J.

(1991) J. Catal., 129, 238. 77. de Smit, E. and Weckhuysen, B.M.

(2008) Chem. Soc. Rev., 37, 2758. 78. Dictor, R.A. and Bell, A.T. (1986) J.

Catal., 97, 121. 79. Li, S., Li, A., Krishnamoorthy, S., and

Iglesia, E. (2001) Catal. Lett., 77, 197. 80. Fraenkel, D. and Gates, B.C. (1979) J.

Am. Chem. Soc., 102, 2478. 81. Caesar, P.D., Brennan, J.A., Garwood,

82.

83.

84.

85. 86.

87.

88.

89.

90.

91.

92. 93.

W.E., and Ciric, J. (1979) J. Catal., 56, 274. Ravichandran, G., Das, D., and Chakrabarty, D.K. (1994) J. Chem. Soc. Faraday Trans., 90, 1993. ´ Gallegos, N.G., Alvarez, A.M., Cagnoli, M.V., Bengoa, J.F., Cano, L.A., Mercader, R.C., and Marchetti, S.G. (2005) Catal. Today, 107-108, 355. McMahon, K.C., Suib, S.L., Johnson, B.G., and Bartholomew, C.H. (1987) J. Catal., 106, 47. Zhou, P.L., Maloney, S.D., and Gates, B.C. (1991) J. Catal., 129, 315. Das, D., Ravichandran, G., and Chakrabarty, D.K. (1997) Catal. Today, 36, 285. Calleja, G., de Lucas, A., van Grieken, R., Pe˜ na, J.L., Guerrero-Ruiz, A., and Fierro, J.L.G. (1993) Catal. Lett., 18, 65. ´ Marchetti, S.G., Alvarez, A.M., Bengoa, J.F., Cagnoli, M.V., Gallegos, N.G., Yerami´an, A.A., and Mercader, R.C. (1999) Hyperﬁne Interact. C, 4, 61. ´ Cagnoli, M.V., Gallegos, N.G., Alvarez, A.M., Bengoa, J.F., Yerami´an, A.A., Schmal, M., and Marchetti, S.G. (2002) Appl. Catal. A, 230, 169. Nazar, L.F., Ozin, G., Hughes, F., Godber, J., and Rancourt, D. (1983) Angew. Chem. Int. Ed., 22, 624. Nijs, H.H., Jacobs, P.A., and Uytterhoeven, J.B. (1979) J. Chem. Soc. Chem. Commun., 180. Kawi, S., Chang, J.R., and Gates, B.C. (1993) J. Am. Chem. Soc., 115, 4830. Lee, Y.-J., Park, J.-Y., Jun, K.-W., Bae, J.W., and Viswanadham, N. (2008) Catal. Lett., 126, 149.

References 94. Khodakov, A.Y., Chu, W., and

95. 96.

97.

98.

99.

100. 101. 102.

103.

104.

105.

106. 107.

108.

109.

Fongarland, P. (2007) Chem. Rev., 107, 1692. Iglesia, E., Soled, S.L., and Fiato, R.A. (1992) J. Catal., 137, 212. Bezemer, G.L., Bitter, J.H., Kuipers, H.P.C.E., Oosterbeek, H., Holewijn, J.E., Xu, X., Kapteijn, F., van Dillen, A.J. and de Jong, K.P. (2006) J. Am. Chem. Soc., 128, 3956. Prieto, G., Mart´ınez, A., Concepci´on, P., and Moreno-Tost, R. (2009) J. Catal. 266, 129. den Breejen, J.P., Radstake, P.B., Bezemer, G.L., Bitter, J.H., Frøseth, V., Holmen, A., and de Jong, K.P. (2009) J. Am. Chem. Soc., 131, 7197. de Soler-Illia, G.J.A.A., Sanchez, C., Lebeau, B., and Patarin, J. (2002) Chem. Rev., 102, 4093. Taguchi, A. and Schueth, F. (2004) Microporous Mesoporous Mater., 77, 1. Mart´ınez, A. and Prieto, G. (2009) Top. Catal., 52, 75. ˚ Vralstad, T., Øye, G., Rønning, M., Glomm, W.R., St¨ocker, M., and Sj¨oblom, J. (2005) Microporous Mesoporous Mater., 80, 291. Ohtsuka, Y., Takahashi, Y., Noguchi, M., Arai, T., Takasaki, S., Tsubouchi, N., and Wang, Y. (2004) Catal. Today, 89, 419. Suvanto, S., Hukkam¨aki, J., Pakkanen, T.T., and Pakkanen, T.A. (2000) Langmuir, 16, 4109. Khodakov, A.Y., Zholobenko, V.L., Bechara, R., and Durand, D. (2005) Microporous Mesoporous Mater., 79, 29. Mart´ınez, A., L´opez, C., M´arquez, F., and D´ıaz, I. (2003) J. Catal., 220, 486. Sietsma, J.R.A., Meeldijk, J.D., den Breejen, J.P., Versluijs-Helder, M., Jos van Dillen, A., de Jongh, P.E. and de Jong, K.P. (2007) Angew. Chem., 119, 4631. Khodakov, A.Y., Griboval-Constant, A., Bechara, R., and Zholobenko, V.L. (2002) J. Catal., 206, 230. Khodakov, A.Y., Girardon, J.-S., Griboval-Constant, A., Lermontov, A.S., and Chernavskii, P.A. (2004) Stud. Surf. Sci. Catal., 147, 295.

110. Wei, M., Okabe, K., Arakawa, H., and

111.

112.

113.

114.

115.

116.

117. 118.

119.

120.

121.

122.

123.

124.

125.

126.

Teraoka, Y. (2004) Catal. Commun., 5, 597. Kim, D.J., Dunn, B.C., Cole, P., Turpin, G., Ernst, R.D., Pugmire, R.J., Kang, M., Kim, J.M., and Eyring, E.M. (2005) Chem. Commun., 1462. Prieto, G., Mart´ınez, A., Murciano, R., and Arribas, M.A. (2009) Appl. Catal. A, 367, 146. Corma, A., Forn´es, V., Mart´ınez-Triguero, J., and Pergher, S.B. (1999) J. Catal., 186, 57. Corma, A., Diaz, U., Domine, M.E., and Forn´es, V. (2000) Angew. Chem. Int. Ed., 39 (8), 1499. Concepci´on, P., L´opez, C., Mart´ınez, A., and Puntes, V. (2004) J. Catal., 228, 321. Mart´ınez, A., Roll´an, J., Arribas, M.A., Cerqueira, H.S., Costa, A.F., and Sousa-Aguiar, E.F. (2007) J. Catal., 249, 162. Mart´ınez, A. and L´opez, C. (2005) Appl. Catal. A, 294, 251. Tsubaki, N., Yoneyama, Y., Michiki, K., and Fujimoto, K. (2003) Catal. Commun., 4, 108. Mart´ınez, A., L´opez, C., Peris, E., and Corma, A. (2005) Stud. Surf. Sci. Catal., 158, 1327. Mart´ınez, A., Valencia, S., Murciano, R., Cerqueira, H.S., Costa, A.F., and Sousa-Aguiar, E.F. (2008) Appl. Catal. A, 346, 117. Brennan, J.A., Caesar, P.D., Ciric, J., and Garwood, W.E. (1981) US Patent No. 4304871. Bao, J., He, J., Zhang, Y., Yoneyama, Y., and Tsubaki, N. (2007) Angew. Chem. Int. Ed., 47, 353. Yang, G., He, J., Yoneyama, Y., Tan, Y., Han, Y., and Tsubaki, N. (2007) Appl. Catal. A, 329, 99. Semelsberger, T.A., Borup, R.L., and Greene, H.L. (2006) J. Power Sources, 156, 497. Arcoumanis, C., Bae, C., Crookes, R., and Kinoshita, E. (2008) Fuel, 87, 1014. Shikada, T., Ohno, Y., Ogawa, T., Ono, M., Mizuguchi, M., Tomura, K., and Ujimoto, K. (1998) Stud. Surf. Sci. Catal., 119, 515.

683

684

21 Advanced Catalysts Based on Micro- and Mesoporous Molecular Sieves 127. Ge, Q., Huang, Y., Qiu, F., and Li, S. 128. 129. 130.

131.

132.

133.

134.

135. 136.

137. 138.

139.

140.

141.

142.

143. 144.

145.

146.

(1998) Appl. Catal. A, 167, 23. Sun, K., Lu, W., Qiu, F., Liu, S., and Xu, X. (2003) Appl. Catal. A, 252, 243. Chen, J.G. and Niu, Y.Q. (1997) Nat. Gas Chem. Ind., 22, 6. Kim, J.-H., Park, M.J., Kim, S.J., Joo, O.-S., and Jung, K.-D. (2004) Appl. Catal. A, 264, 37. Khandan, N., Kazemeini, M., and Aghaziarati, M. (2009) Catal. Lett., 129, 111. Xia, J., Mao, D., Tao, W., Chen, Q., Zhang, Y., and Tang, Y. (2006) Microporous Mesoporous Mater., 91, 33. Sai Prasad, P.S., Bae, J.W., Kang, S.-H., Lee, Y.-J., and Jun, K.-W. (2008) Fuel Process Technol., 89, 1281. Yoo, K.S., Kim, J.-H., Park, M.-J., Kim, S.-J., Joo, O.-S., and Jung, K.-D. (2007) Appl. Catal. A, 330, 57. Jain, J.R. and Pillai, C.N. (1967) J. Catal., 9, 322. Kubelkov´a, L., Nov´akov´a, J., and Nedomov´a, J. (1990) J. Catal., 124, 441. Bandiera, J. and Naccache, C. (1991) Appl. Catal., 69, 139. Ramos, F.S., Duarte de Farias, A.M., Borges, L.E.P., Monteiro, J.L., Fraga, M.A., Sousa-Aguiar, E.F., and Appel, L.G. (2005) Catal. Today, 101, 39. Wang, L., Qi, Y., Wei, Y., Fang, D., Meng, S., and Liu, Z. (2006) Catal. Lett., 106, 61. Takeguchi, T., Yanagisawa, K., Inui, T., and Inoue, M. (2000) Appl. Catal. A, 192, 201. Xu, M., Lunsford, J.H., Goodman, D.W., and Bhattacharyya, A. (1997) Appl. Catal. A, 149, 289. Mao, D., Yang, W., Xia, J., Zhang, B., and Song, Q. (2005) Q. Chen. J. Catal., 230, 140. Xu, Q.-L., Li, T.-C., and Yan, Y.-J. (2008) J. Fuel Chem. Technol., 36, 176. Fei, J., Hou, Z., Zhu, B., Lou, H., and Zheng, X. (2006) Appl. Catal. A, 304, 49. Kang, S.-H., Bae, J.W., Jun, K.-W., and Potdar, H.S. (2008) Catal. Commun., 9, 2035. Maitlis, T.M. (2003) J. Mol. Catal. A, 204-205, 55.

147. Mahajan, D. (2005) Top. Catal., 32

(3-4), 209. 148. Song, D., Li, J., and Cai, Q. (2007) J.

Phys. Chem. C, 111, 18970. 149. Guczi, L. (1990) Catal. Lett., 7, 205. 150. Trevi˜ no, H., Lei, G.-D., and Sachtler,

W.M.H. (1995) J. Catal., 154, 245. 151. Sachtler, W.M.H. and Ichikawa, M.

(1986) J. Phys. Chem., 90, 4752. 152. Shen, J.G.C. and Ichikawa, M. (1998) J.

Phys. Chem. B, 102, 5602. 153. Cavalcanti, F.A.P., Stakheev, A.Y., and

154. 155.

156.

157.

158. 159.

160.

161. 162.

163.

164.

165.

166.

Sachtler, W.M.H. (1992) J. Catal., 134, 226. Yin, Y.-G., Zhang, Z., and Sachtler, W.M.H. (1993) J. Catal., 139, 444. L´az´ar, K., Manninger, I., and Choudary, B.M. (1991) Hyperﬁne Interact., 69, 747. Boffa, A., Lin, C., Bell, A.T., and Somorjai, G.A. (1994) J. Catal., 149, 149. Sch¨unemann, V., Trevi˜ no, H., Lei, G.D., Tomczak, D.C., Sachtler, W.M.H., Fogash, K., and Dumesic, J.A. (1995) J. Catal., 153, 144. Xu, B.-Q. and Sachtler, W.M.H. (1998) J. Catal., 180, 194. Yin, H., Ding, Y., Luo, H., Zhu, H., He, D., Xiong, J., and Lin, L. (2003) Appl. Catal. A, 243, 155. Pan, X., Fan, Z., Chen, W., Ding, Y., Luo, H., and Bao, X. (2007) Nature Mater., 6, 507. Sunley, G.J. and Watson, D.J. (2000) Catal. Today, 58, 293. Xu, Q., Inoue, S., Tsumori, N., Mori, H., Kameda, M., Tanaka, M., Fujiwara, M., and Souma, Y. (2001) J. Mol. Catal. A, 170, 147. Stepanov, A.G., Luzgin, M.V., Romannikov, V.N., and Zamaraev, K.I. (1995) J. Am. Chem. Soc., 117, 3615. Smith, W.J. (1995) US Patent No. 5420345, assigned to BP Chemicals Limited. Cheung, P., Bhan, A., Sunley, G.J., and Iglesia, E. (2006) Angew. Chem. Int. Ed., 45, 1617. Cheung, P., Bhan, A., Sunley, G.J., Law, D.J., and Iglesia, E. (2007) J. Catal., 245, 110.

References 167. Bhan, A., Allian, A.D., Sunley, G.J.,

168.

169.

170.

171. 172.

173.

Law, D.J., and Iglesia, E. (2007) J. Am. Chem. Soc., 129, 4919. Boronat, M., Mart´ınez-Sanchez, C., Law, D., and Corma, A. (2008) J. Am. Chem. Soc., 130, 16316. Ellis, B., Joward, M.J., Joyner, R.W., Reddy, K.N., Padley, N.B., and Smith, W.J. (1996) Stud. Surf. Sci. Catal., 101, 771. Blasco, T., Boronat, M., Concepci´on, P., Corma, A., Law, D., and Vidal-Moya, J.A. (2007) Angew. Chem. Int. Ed., 46, 3938. Labinger, J.A. (2004) J. Mol. Catal. A, 220, 27. Knops-Gerrits, P.-P.H.J.M. and Goddard, W.A.III (2003) Catal. Today, 81, 263. Chui, S.S.Y., Lo, S.M.F., Charmant, J.P.H., Orpen, A.G., and Williams, I.D. (1999) Science, 283, 1148.

174. Noro, S., Kitagawa, S., Yamashita, M.,

175.

176. 177.

178. 179.

and Wada, T. (2002) Chem. Commun., 222, 41. Hwang, Y.K., Hong, D.Y., Chang, J.S., Jhung, S.H., Seo, Y.K., Kim, J., Vimont, A., Daturi, M., Serre, C., and Ferey, G. (2008) Angew. Chem. Int. Ed., 46, 4144. Zhang, X., Llabr´es-i-Xamena, M., and Corma, A. (2009) J. Catal., 265, 155. Yaghi, O.M., O’Keeffe, M., Ockwig, N.W., Chae, H.K., Eddaoudi, M., and Kim, J. (2003) Nature, 423, 705. Yuliati, L. and Yoshida, H. (2008) Chem. Soc. Rev., 37, 1592. Ahn, J.H., Temel, B., and Iglesia, E. (2009) Angew. Chem. Int. Ed., 48, 1.

685

687

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG) Michael St¨ocker

22.1 Introduction

Large amounts of natural gas (about 60%) are recovered at remote gas ﬁeld sites. Taking into account the total known worldwide natural gas reserves (about 141 trillion cubic meters), which is quite impressive compared to the crude oil resources, one might understand the strong focus on the conversion of natural gas into liquids (GTLs) and higher added value products during the past few year. This can be performed either via the direct route based on the Fischer–Tropsch technology (GTL) or via the indirect route, which consists of the production of methanol from synthesis gas (made by steam reforming of natural gas, gasiﬁcation of coal, or lignocellulosic biomass), and the consecutive formation of oleﬁns and/or gasoline. At the end of the last century, the proven crude oil reserves worldwide were in the range of 145 GT, and with the current consumption of about 3.4 GT per year these reserves will be utilized in roughly 30 years from now, taking into account an increase in the consumption of about 2% per year worldwide. Including the gas-associated liquids, a reserve of hydrocarbons will cover a period of roughly 100–150 years, depending on which scenario the estimations are based on (International Energy Agency, IEA, or US Department of Energy, DoE) [1]. The above-mentioned routes are alternatives to the chemical conversion of methane, either via oxidative coupling, a route not successful so far from an industrial point of view, or via direct coupling, which, however, is thermodynamically not favorable. The interest in manufacturing oleﬁns and gasoline from natural gas is driven by the more or less required direct application of this technology, for example, at remote natural gas ﬁeld sites, ﬁrst of all with respect to minimizing gas burning at the recovery sites and the transportation costs. In the early stages, the methanol-to-gasoline (MTG) process was mainly considered as a powerful technology to convert coal into high-octane gasoline. However, this concept has been expanded since, both with respect to the formation of other fuels and to other chemicals as well. This development has been governed by the importance of light oleﬁns as valuable components for the petrochemical industry Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

688

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

and by the increasing demand for high-quality gasoline. In a way, with this new technology, one can make almost anything out of natural gas, coal, or lignocellulosic biomass that can be made out of crude oil. Methanol is manufactured from synthesis gas, which is made by steam reforming from either natural gas or coal or even gasiﬁcation of lignocellulosic biomass. The methanol is then processed to an equilibrium mixture of methanol, dimethyl ether, and water, which can be converted catalytically to either oleﬁns (MTO) or gasoline (MTG), depending on the process conditions applied and/or the catalyst used. Although methanol itself is a potential motor fuel or can be blended with gasoline, it would require large investments to overcome the technical problems connected with the direct use of methanol as fuel. In addition, the increasing attention with respect to the production of biofuels (bioethanol, biodiesel) implies a reconsideration of the future transportation fuel market as well [1–4]. The industrial MTG process applies a ZSM-5 catalyst and operates at temperatures of about 400 ◦ C at a methanol pressure of several atmospheres. The addressed conditions represent an optimum for converting oleﬁns, which are formed in the pores of the microporous zeolite catalyst into aromatics and parafﬁns. However, at one point during the MTG process, the product mixture consists of about 40% of light oleﬁns. It was recognized early that the formation of light oleﬁns as intermediates is important with respect to the conversion of MTG. Therefore, several attempts were made to form light oleﬁns selectively from methanol, using both medium-pore and small-pore zeolites as well as SAPO-type molecular sieves. Interrupting the reaction at the point of about 40% light oleﬁn formation, C2 –C4 oleﬁns can be saved. By adjusting the reaction conditions (like, for example, raising the temperature to 500 ◦ C) as well as the catalyst applied, one can signiﬁcantly increase the oleﬁn yield. This discovery led to the development of the MTO process, which mostly generates propene and butene, with high-octane gasoline as a by-product. However, the catalyst can be modiﬁed to selectively produce more ethene [5]. A strong interest in synfuels and other chemicals could be observed in 1973 in connection with the worldwide energy crisis. This focus favored, among others, the continuation of the research devoted to the methanol to oleﬁns and gasoline technology [6]. Both the MTG and the MTO processes represent a type of petrochemical plants, to be brought onstream as the technological and/or economic demands arise. As a continuation, one can convert the oleﬁns to an entire spectrum of products, applying another ZSM-5-based process: Mobil’s oleﬁn-to-gasoline and distillate process (MOGD), originally developed as a reﬁnery process, operates well connected to the MTO process. In the MOGD process, ZSM-5 oligomerizes light oleﬁns, from either reﬁnery streams or MTO, into higher molecular weight oleﬁns that fall into the gasoline, distillate, and lubricant range (Figure 22.1) [7]. The ﬁrst commercial MTG process was established in New Zealand in 1979, in competition with the SASOL process (based on the Fischer–Tropsch technology), for the conversion of natural gas from the Maui ﬁeld to gasoline. At that time, Mobil’s ﬁxed-bed MTG process was unproven commercially, whereas the SASOL technology was already commercialized [7]. In the beginning, the plant in New Zealand produced about 14 500 barrels of gasoline per day from April 1986,

22.1 Introduction

Natural gas

Natural gas liquids

Gasification

Steam reform

Syn gas CO + H2

Methanol synthesis

Methanol

Coal

689

MTG Moblie methanol to gasoline MTO Moblie methanol to olefins

Gasoline

MOGD Moblie olefins to gasoline and distillate

Dehydrogenation

Figure 22.1 Gasoline and distillate production via methanol and Mobil’s ZSM-5 technology [5]. Reproduced by permission of Elsevier, Amsterdam.

supplying one-third of the nation’s gasoline demand [8]. Currently, the gasoline production part of the factory is closed, due to the price available for gasoline versus the price of methanol, and only the methanol production part is in operation. The MTO process is ready for commercial use. UOP has announced their SAPO-34-based MTO process to be realized for construction of a 250 000tons-per-year plant using a natural gas feedstock for the production of ethene in cooperation with Norsk Hydro. The ﬁrst commercial factory is under construction in Nigeria. A 0.5-tons-per-year demonstration unit operated by Norsk Hydro has veriﬁed the oleﬁn yields and catalyst performance. SAPO-34 is extremely selective toward ethene and propene formation with the ﬂexibility of altering the ratio between the two oleﬁns by varying the reactor conditions [8]. As ZSM-5 and SAPO-34 are the main representatives for the methanol conversion to, respectively, gasoline or oleﬁns, these two structures are brieﬂy presented here. ZSM-5 is a medium-pore-sized zeolite (three-dimensional 10-ring system with pores in the range of 5.3–5.6 A˚ in diameter, space group Pnma), whereas SAPO-34 is a small-pore-sized zeotype material (three-dimensional eight-ring system with pores in the range of 3.8 A˚ in diameter, space group R-3m). Their schematic structures are shown in Figure 22.2. The discovery of the MTG reaction was an accident. One group at Mobil was trying to convert methanol to other oxygen-containing compounds over a ZSM-5 catalyst. Instead, they received unwanted hydrocarbons. Somewhat later, another Mobil group, working independently, was trying to alkylate isobutane with methanol over ZSM-5 and identiﬁed a mixture of parafﬁns and aromatics boiling in the gasoline range – all coming from methanol [5]. Although the discovery of MTG was accidental, it occurred due to a balanced effort in catalysis over many years. The MTO reaction seems to beneﬁt from this development, although independent research has been performed since. The evolution of the methanol-to-hydrocarbons (MTHs) technology, from their discovery to the demonstrative and/or commercial

Gasoline Distillate

690

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

Y Z

N Y

X

(a)

X

(b)

Figure 22.2

Schematic structures of ZSM-5 (a) and SAPO-34 (b).

realization, has been accompanied by extensive research related to the basic question of the mechanism of formation of the initial C–C bond [5].

22.2 Mechanism and Kinetics of the MTO and MTG Reactions

The main reaction steps of the methanol conversion to oleﬁns, gasoline, and other hydrocarbons can be summarized as follows: −H2 O −H2 O (22.1) 2CH3 OH ←→ CH3 OCH3 −→ +H2 O lights oleﬁns −→ n/iso − parafﬁns, higher oleﬁns, aromatics, naphthenes

Methanol is ﬁrst dehydrated to dimethylether (DME). The formed equilibrium mixture consisting of methanol, DME, and water is then converted to light oleﬁns. In the last step of this scheme, the light oleﬁns react to form parafﬁns, aromatics, naphthenes, and higher oleﬁns by alkylation and polycondensation reactions (Figure 22.3). There is a consensus that the intermediate in the dehydration of methanol to DME (step 1 in Equation (22.1)) over solid acid catalysts is a protonated surface methoxyl, which is subject to a nucleophilic attack by methanol [5]. The subsequent conversion of light oleﬁns to parafﬁns, aromatics, naphthenes, and higher oleﬁns (step 3 in Equation (22.1)), which proceeds via classical carbenium ion mechanisms with concurrent hydrogen transfer, is well known from hydrocarbon chemistry in acid media [9]. However, the second step in Equation (22.1), which represents

22.2 Mechanism and Kinetics of the MTO and MTG Reactions 70 60

Methanol

Product yield (wt%)

Water Dimethyl ether

50 40

Paraffins (and C6 + Olefins)

30 20

Aromatics 10 0 10−4

C2−C5 Olefins 10−3

10−2

10−1

1

10

1/LHSV (hours)

Figure 22.3 Methanol-to-hydrocarbons reaction path [5]. Reproduced by permission of Elsevier, Amsterdam.

the initial C–C bond formation from the C1 reactants, has been the topic of an extensive discussion throughout the years [5]. A large number of papers are available presenting more than 20 possible mechanistic proposals for the formation of the ﬁrst C–C bond. The most discussed and/or relevant of these mechanisms can be broadly classiﬁed and brieﬂy summarized according to the following [9–11]: Strong attention has been devoted to the oxonium ylide mechanism [12], postulating that DME interacts with a Brønsted acid site of the solid catalyst to form a dimethyloxonium ion, which reacts further with another DME to form a trimethyloxonium ion. This trimethyloxonium ion is subsequently deprotonated by a basic site (oxygen atom) to form a surface-associated dimethyloxonium methylide species [9, 10]. The next step is either an intramolecular Stevens rearrangement [12], leading to the formation of methyl–ethyl ether, or an intermolecular methylation [5], leading to the formation of ethyl-dimethyloxoniumion. In both the cases, ethene is formed via β-elimination [9]. Extensive research has been concentrated on the questions about the existence of the oxonium ylides, their ability to rearrange according to Stevens, and the zeolite’s ability to abstract a proton from the oxonium ions to form the desired ylides [5]. Suitable information was obtained by running different experiments; however, they do not give a deﬁnitive answer with respect to the ylide question due to their shortcomings. Nevertheless, if those oxonium ylides do exist and are involved in the C–C bond formation, there is still a question about the zeolite’s ability to abstract a proton from the oxonium ions to form the desired ylides. Chao and Huarng [13] studied the conversion of MTHs over ZSM-5 with mixtures of labeled and unlabeled methanol and DME at elevated temperatures. Their results indicate that ethene was the primary hydrocarbon product through bimolecular reactions of DME with or without methanol on the catalyst’s surface. The C–C bond was formed following

691

692

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

an acidic mechanism via oxonium ions with intermolecular transfer of H and D. An isotope distribution analysis of the products showed that propene were most likely produced by the reaction of ethene and C1 species from methanol adsorbed on the ZSM-5 surface [13]. In summary, the oxonium ylide mechanism involves the formation of a surface-bound intermediate as the initial reaction step. The zeolitic surface OH-group is methylated to form the methyloxonium intermediate, which gives rise to a surface-bound methylene oxoniumylide due to deprotonation. The surface-bound methylene oxoniumylide is isoelectronic with a surface-associated carbene [14, 15]. Further methylation results in the initial C–C bond formation [9]. The carbine–carbenoid mechanism involves the α-elimination of water from methanol, followed by either polymerization of the resultant carbene to oleﬁns or by concurrent sp3 insertion of the carbene into methanol or DME [16–18]. The formation of the carbene by the cooperative action of acid and basic sites in mordenite was suggested by Swabb and Gates [18] and can be summarized as follows [5]: Zeo − O− ← H − CH2 − OH → H−O−Zeo → H2 O+ : CH2 (22.2) According to Salvador and Kladnig [19], carbenes are generated through decomposition of surface methoxyls formed originally upon chemisorption of methanol on the zeolite (zeolite Y) [5, 10]. So far, experimental evidence with respect to the intermediate carbene formation has been indirect, for example, by trapping the reactive C1 intermediate from 13 C-labeled methanol decomposition over ZSM-5 using unlabeled propane. The analysis of the product isomer and isotope distributions gave indications that the reactive C1 intermediate was carbenic and the mode of attack on propane was sp3 insertion into H–C bonds. Evidence in favor of carbene intermediacy in the conversion of MTG over H-ZSM-5 was reported by Dass et al. [20]. Hutchings et al. [5, 21] contributed with a number of papers dealing with the mechanistic aspects of the MTH reaction; among them was the reaction of methanol with added hydrogen using WO3 /alumina and H-ZSM-5 catalysts, which did not signiﬁcantly alter the product distribution from that of normal methanol conversion. The authors considered their results as clear evidence against the involvement of a gas-phase methylene intermediate. In addition, the behavior of the methoxylmethyl radical in the gas phase has been studied and the results demonstrate that this species is also not involved as an intermediate in initial C–C bond formation in the MTH reaction [5, 21]. Although the oxonium ylide mechanism essentially involves the formation of a surface-bound intermediate as the initial reaction step, the carbene mechanism involves only surface-associated intermediates [5, 9], as outlined in the following: Z − OH + CH3 OH → Z − O − CH3 + H2 O Surface − incorporated ylide :Z − O − CH3 → H+ + Z − O − CH− 2 Surface − associated carbene :

Z − O :: CH2 (22.3)

22.2 Mechanism and Kinetics of the MTO and MTG Reactions

The carbocationic mechanism has been favored by other groups [21–26]. According to Ono and Mori [22], surface methoxyls may function as free methyl cations, which adds to the C–H bond of DME to form a pentavalent carbonium transition state. Abstraction of a proton completes the reaction. However, one may argue whether the C–H bond of methanol or DME is sufﬁciently nucleophilic to undergo substitution as proposed by Ono and Mori [22]. As an alternative for the formation of methyloxonium intermediates, the free radical mechanism has been introduced [27]. This mechanism can be summarized as follows: Z − OH + CH3 OH → Z − O − CH3 + H2 O Z − O − CH + R• → • CH O − Z + RH 3

2

• CH O − Z + ZO → Z − O :: CH 2 2

+ZO] • Z − O − CH− 2 HR + ZO• → R• + Z − OH

(22.4)

The participation of free radicals in the conversion of MTH over natural mordenite has been suggested by Zatorski and Krzyzanowski [28]. However, the free radical pathway has been discussed by several other authors concluding with experimental evidence against a radical mechanism. Most of the mechanistic studies related to the MTH reaction were carried out using ZSM-5 as catalyst. The mechanisms proposed so far may broadly be classiﬁed into two groups: The consecutive-type mechanism, which means that one carbon from methanol is added during each step. Addition and cracking reactions of the alkene molecules may take place as illustrated as follows [29]: 2C1 → C2 H4 + H2 O C2 H4 + C1 → C3 H6 C3 H6 + C1 → C4 H8 . . .

(22.5)

A parallel-type mechanism, known as hydrocarbon-pool mechanism was suggested by Dahl and Kolboe [29, 30], who studied the MTH conversion by applying SAPO-34 as a catalyst and 13 C-labeled methanol as feed (and (12 C) ethene (made in situ from ethanol)) (Scheme 22.1). The ‘‘hydrocarbon-pool’’ = (CH2 )n represents an adsorbate, which may have many characteristics in common with ordinary coke, and which might easily contain less hydrogen than indicated. It would perhaps be better represented by (CHx )n with 0 < x < 2 [30]. Once the C–C bond formation has started, an induction period for oleﬁns formation is observed, because oligocyclization has to produce a signiﬁcant amount of aromatics (‘‘hydrocarbon pool’’), a prerequisite for a closed loop, in which they behave as catalysts [31]. When using SAPO-34 in the title

693

694

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

C2H4

CH3OH

(CH2)n

C3H6 Saturated hydrocarbons

C4H8

Coke

Scheme 22.1 Hydrocarbon-pool mechanism [5]. Reproduced by permission of Elsevier, Amsterdam.

reaction, the product pattern is thus simpler than, for example, ZSM-5, where a much wider range of products was found. Therefore, it might be easier to obtain a picture of the reaction pathway using SAPO-34, and Dahl and Kolboe showed that the consecutive mechanism, insofar as propene formation was concerned, did not turn out to be valid. They concluded that only a minor part of the propene molecules may have been formed by addition of methanol to ethene as this would imply a 12 C/13 C ratio larger than one. Most of the propene molecules were formed directly from methanol [29, 30]. During the last few years, it has become clear that methylbenzenes play central roles in the ‘‘hydrocarbon-pool’’ mechanism and they are regarded as important species taking part during the course of the reaction. Marcus et al. demonstrated that the only pathway for the MTO conversion is based on the hydrocarbon-pool mechanism, which describes the primary synthesis of oleﬁns on organic reaction centers, such as methylbenzenes, coupled with well-known secondary reactions of oleﬁns [32]. Arstad and coworkers used a zeolite cluster model to describe the relative energies of the MTO reaction accurately, reproducing experimental structure-reactivity and structure-selectivity data for the methylbenzene hydrocarbon-pool mechanism [33]. Experimental and theoretical work has ﬁrmly established that methanol and dimethyl ether react on cyclic organic species contained in the cages or channels of the inorganic hosts. These organic reaction centers act as scaffolds for the assembly of light oleﬁns so as to avoid the very high-energy intermediates required by all direct mechanisms [34]. A number of recent investigations have been concentrated on the role of methylbenzenes in connection with the hydrocarbon-pool mechanism. For example, isotopic scrambling in all polymethylbenzenes has been found after switching from a 12 C-methanol feed to a 13 C-methanol feed in the organic material trapped inside a SAPO-34 catalyst [35–38]. Arstad and Kolboe observed the oleﬁns distribution after the isotope switch (12 C-methanol/13 C-methanol), and they concluded that the polymethylbenzenes are the active species in the hydrocarbon pool of SAPO-34, due to the fact that the isotopic distribution of the oleﬁnic products changed only gradually [37, 38]. Further, the role of polymethylbenzenes as the major ‘‘hydrocarbon-pool’’ species appears to be independent of the microporous zeotype catalyst used. Detailed studies on the reactivity of polymethylbenzenes

22.2 Mechanism and Kinetics of the MTO and MTG Reactions

have been carried out over H-BEA zeolite, as this 12-MR zeolite allows direct feeding of polymethylbenzenes. The reaction of 13 C-methanol and 12 C-benzene over zeolite H-BEA resulted in the formation of methylbenzenes consisting of a 12 C-benzene ring and 13 C-methyl groups, indicating a ‘‘paring’’-type mechanism (alkyl side-chain growth by ring contraction/expansion) responsible for the oleﬁn formation over this zeolite. Studies performed by Svelle et al., dealing with the methylation of 12 C-ethene, -propene, and -butene with 13 C-methanol over a H-ZSM-5 catalyst, conﬁrmed that methylation, oligomerization, polymethylbenzene formation (‘‘hydrocarbon-pool’’) and cracking reactions took place in parallel, again underlining the complex reaction pathways of the MTH conversion [36]. Isotopic scrambling in the oleﬁnic products was also observed by Mikkelsen et al. when cofeeding 12 C-toluene and 13 C-methanol over H-ZSM-5 [39]. Similar conclusions have been drawn for H-Beta zeolite as well, showing that polymethylbenzenes are the major hydrocarbon-pool species, independent of the zeotype system chosen [40]. The complexity of the MTH reaction due to rapid secondary reactions as well as limitations of suitable techniques to follow this reaction has always been a problem while ﬁnding conclusive evidence for the chemistry of the methanol conversion. Besides the topic of formation of the ﬁrst C–C bond, the number and nature of intermediates have been a matter of great interest. 13 C MAS NMR spectroscopy turned out to be a powerful tool with respect to the identiﬁcation of the formed intermediates, and a number of investigations have been conducted mainly by applying ZSM-5 and SAPO-34 catalysts [5]. Besides solid-state NMR spectroscopy, a variety of other techniques like in situ FTIR spectroscopy, ﬂow reactor/GC-MS spectrometry, temperature-programmed desorption (TPD), as well as differential scanning calorimetry (DSC) have been used to follow the complexity of the MTH reaction [5]. A review exists covering the use of spectroscopic techniques to investigate the MTG process. Examples are quoted of catalyst characterization, investigation of the ﬁrst carbon–carbon bond formation, alkene oligomerization, and catalyst deactivation through coke formation [41–43]. Currently, the importance of the initial C–C bond formation is considered to be minor and the ‘‘hydrocarbon-pool’’ mechanism is gaining general acceptance. More details about the product formation can be found in the recent review by Kvisle et al. [44]. Obviously, a large number of experiments have been carried out to resolve the question of C–C bond formation from methanol. Although the answer still remains elusive, these experiments tell us at least what is probably not involved in the bond formation, particularly in the presence of zeolite catalysts [10]. Recently, McCann et al. reported a working catalytic cycle for the conversion of methanol into oleﬁns, in consistency with both experimental and theoretical investigations [45]. For each cycle step, rate constants are presented, which were obtained by quantum chemical simulations on a supramolecular model of both the H-ZSM-5 zeolite and the cocatalytic hydrocarbon-pool species. This work not only represents the most robust computational analysis of a successful MTO route to date but it also succeeds in tying together the many experimental clues. The ﬁrst clue toward identiﬁcation

695

696

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

of a successful hydrocarbon-pool route is the cocatalytic effect of toluene and higher methylbenzenes. Scrambling of labeled carbon atoms from the methanol feed into these methylbenzenes as well as into the oleﬁn products further illustrates the catalytic activity of the methylbenzene pool. Additional clues are provided by in situ NMR spectroscopy, with which crucial cationic intermediates can be identiﬁed. An additional important observation is the large variation in product distribution depending on the zeolite catalyst applied. All experimental clues are combined in a theoretically proposed catalytic cycle, which greatly enhanced previous research and developed a fundamental understanding of the actual steps involved in producing oleﬁns from aromatic reaction centers [45]. Kinetic investigations related to the MTH conversion normally regard the methanol/DME mixture as a single species. This is conﬁrmed by the observation that the DME formation is much faster than the subsequent reactions, so that the oxygenates are at equilibrium. An overall ﬁrst-order rate constant of roughly 250 m3 gas m−3 catalyst s−1 has been reported for SAPO-34 applied in the MTO reaction at 450 ◦ C [9, 46]. On the basis of the autocatalytic nature of the methanol conversion over ZSM-5, Chen and coworkers applied a kinetic model, which takes into account that the rate of disappearance of oxygenates is accelerated by their reaction with oleﬁns. Fitting experimental data obtained on H-ZSM-5 with varying concentrations of acid sites showed a linear correlation between the rate constant of the reaction of oxygenates with oleﬁns and the intrinsic acid activity of the catalyst [9, 47]. Ono and Mori [22] proposed the ﬁrst step in the model of Chen and coworkers [47] to be bimolecular. Assuming this reaction to be of second order with respect to the concentration of the oxygenates and the autocatalytic reaction to be of ﬁrst order with respect to both the oxygenates and the oleﬁns, reasonable ﬁts of the kinetic model were obtained at 219 and 239 ◦ C, indicating that the autocatalytic effect decreases with increasing temperature [5, 9, 22]. The model of Chen et al. [47] was modiﬁed by Chang by adding a bimolecular step accounting for the carbene insertion in the primary oleﬁns [48]. Chang’s description of the carbene insertion in oleﬁns predicts a redistribution of the oleﬁnic species, but not a net increase in the production of oleﬁns [48]. A revision of the modiﬁed model of Chang [48] was suggested by Sedran et al., who were able to predict the product distribution at various conversion levels and temperatures [49]. On the basis of the carbene mechanism, Mihail and coworkers developed a kinetic model for the MTO reaction up to C5 H10 containing 33 reactions [50]. The formation of carbene from DME turned out to be the rate-determining step among the kinetic parameters evaluated. This is in line with the ﬁndings of Chen et al. [47] and Ono and Mori [22]. Later on, the model was extended to comprise 53 reactions, including the formation of aromatics and C5 + aliphatic compounds [5]. A kinetic model for the methanol conversion to oleﬁns with respect to methane formation at low conversion numbers has been discussed in the literature [51]. The study of Hutchings et al. revealed that at low conversion hydrogen formation from methanol decomposition and the water gas shift reaction was signiﬁcant over the zeolite ZSM-5, and, therefore, these reactions should not be included in the derivation of kinetic models for

22.3 Methanol to Oleﬁns (MTO)

this reaction. In view of this, the authors considered it unlikely that methane formation predominantly proceeds via reaction of carbene intermediate with molecular hydrogen [51]. Declined reaction rates have been reported as coke deposits are accumulated in the structure of porous catalysts [46]. A kinetic model for simulation of the MTO reaction over SAPO-18 has been suggested by Gayubo et al., taking into account the initiation reaction period and the maximum in oleﬁn production [52]. Bos and Tromp developed different deactivation models and observed that an exponential model gave the best ﬁt to experimental data obtained using SAPO-34 [46]. However, Chen et al. reported that a linear dependence gave the best representation of the experimental data [44, 53]. There is a discussion on whether the product selectivity in SAPO-34 is controlled by steric, diffusion, or kinetic effects. Evidence of intracrystalline diffusion limitations has been reported at crystal sizes of >2.5 µm and for intraparticle diffusion limitations for particles >1 mm [44]. Dahl and coworkers investigated the effect of crystal size on the conversion of ethanol and 2-propanol over SAPO-34, receiving ethene and propene, respectively [54]. They suggested that the product selectivity in SAPO-34 is controlled by diffusion limitations [44]. Song and Haw [55] modiﬁed the cages of SAPO-34 and observed an improved ethene selectivity, which was explained by the transition state selectivity suggested by Chen et al. [56] and by stronger diffusion limitations as outlined by Barger [57] and Dahl et al. [44, 54].

22.3 Methanol to Oleﬁns (MTO) 22.3.1 Catalysts and Reaction Conditions

Catalysts mainly applied for the MTO reaction are ZSM-5 and SAPO-34. Although strong acid sites are those that are mainly responsible for aromatization reactions, moderate acidity is required for the MTO process. ZSM-5 and SAPO-34 have different properties and behavior due to their different compositions and topologies. ZSM-5 is an aluminosilicate possessing a three-dimensional pore structure consisting of 10-ring pore openings (medium-pore zeolite), whereas SAPO-34 is a small-pore silicoaluminophosphate consisting of an eight-ring system with a cage structure. The SAPO-34 catalyzed MTO reaction yields both ethene and propene in variable amounts with very low formation of heavier by-products (Figure 22.4). Further, the MTO process can be designed for an ethene-to-propene ratio between 0.75 and 1.5 – at nearly complete methanol conversion and with ethene formation favored at higher severity. A high selectivity to ethene (48%) gives SAPO-34 a signiﬁcant advantage over other types of catalyst systems, like ZSM-5 or SSZ-13 (synthetic aluminosilicate with chabasite (CHA) structure).

697

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

698

22.3.2 Deactivation

The mechanism of deactivation and coke formation has not yet been completely accepted. The formation of naphthalene derivatives during the coreaction of methanol and different methylbenzenes over BEA zeolite was observed by Sassi and coworkers [40]. The rate of coke formation varies among the different zeotype catalysts used, and the inﬂuence of catalyst topology, acid site density, and acid strength has been investigated [44]. Comparative tests of catalysts with CHA topology with varying acid strength (SAPO-34 and SSZ-13, respectively) and acid site density have been performed by Yuen et al. [59]. They recognized an intermediate acid site density in SSZ-13, advantageous for the stability of the catalyst. In comparison, a SAPO-34 catalyst with the same structure but lower acid strength was far more stable than the corresponding CHA sample. Yuen et al. suggested a relationship between the hydride transfer ability and the acid strength [44, 59]. SAPO-34 has a signiﬁcantly better stability due to a lower rate of coke formation than the other catalytic systems with comparable acid site densities. Further, the catalyst operates best at a certain equilibrium level of coke, which again reﬂects the improved stability of SAPO-34. Besides coke, heavier oleﬁns are formed via cationic polymerization. Coking, which is responsible for nonirreversible catalyst deactivation, is many times faster during the induction period, then it decreases to a constant rate, roughly with a linear dependence to space velocity [44, 59]. Owing to this fact, the catalyst life time is inversely related to the total methanol 50

C2 = / SAPO-34 C2 = / ZSM-5 C3 = / SAPO-34 C3 = / ZSM-5 C4 = / SAPO-34 C4 = / ZSM-5 C5+ / SAPO-34 C5+ / ZSM-5 C1−C5 paraffins / SAPO-34 C1−C5 paraffins / ZSM-5 Coke and COx / SAPO-34 Coke and COx / ZSM-5

45 40 35

Yield (%)

30 25 20 15 10 5 0

Figure 22.4 MTO product yields for SAPO-34 and ZSM-5 [58]. Reproduced by permission of DGMK, Hamburg.

22.3 Methanol to Oleﬁns (MTO)

fed. Nonrecoverable deactivation, mainly due to water formation (dealumination), has also been reported[44, 59]. The deactivated catalyst can be regenerated by combustion with air. The vapor phase reaction is carried out at temperatures between 425 and 500 ◦ C within a pressure range of 1–3 bar. Catalyst deactivation is quite fast for SAPO-34 producing mainly aromatic coke. In addition, SAPO-34 is more stable than the corresponding small-pore zeolite structures toward coking, whereas ZSM-5 reveals slower deactivation by aromatic coke. ZSM-5 produces propene as the major oleﬁn; however, signiﬁcantly higher amounts of C5 + and aromatic by-products are formed, compared to SAPO-34 (see Figure 22.4) [31]. 22.3.3 Process Technology and Design

The choice of reactor technology and process design depends strongly on the behavior of the catalysts used. As the MTO conversion is an exothermic reaction, the need to remove the high exothermic heat as well as the need for frequent regeneration led to a ﬂuidized-bed reactor and regenerator design – last but not least to cope with the fast catalyst decay. The heat of the reaction is controlled by steam generation. The catalyst is sent continuously to the regenerator, where the coke is burned off and steam is generated to remove the heat resulting from burning. After heat removal, the reactor efﬂuent is cooled, and some of the water is condensed. After compression, the efﬂuent passes through a caustic scrubber to remove carbon dioxide and to a dryer to remove water [60]. The selectivity to ethene (48%) and propene (33%) is about 81% (90% if butenes are included as well). SAPO-34 is extremely selective toward ethene and propene formation with the ﬂexibility of altering the ratio between the two oleﬁns by varying the reactor

Conversion Conversion or selectivity (%)

100 80 60 Selectivity to C2 = 40 Selectivity to C3 =

20 0

0

10

20

30

40

50

60

Days

Figure 22.5 Stability of the MTO process using methanol from a demo plant [58]. Reproduced by permission of DGMK, Hamburg.

70

80

90

699

700

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

Quench Reactor Regenerator tower

Caustic wash

De-C2

De-C1

C3 C2 splitter De-C3 splitter De-C4 Tail gas Ethylene

Regen gas

Propylene Dryer

Mixed C4

DME removal C2H2 reactor

Air

Water

C5 + Propane Ethane

MeOH

Figure 22.6 Process ﬂow scheme of the UOP/Norsk Hydro MTO process [58]. Reproduced by permission of DGMK, Hamburg.

conditions [5]. A 0.5-tons-per-day demo unit operated by Norsk Hydro in Porsgrunn (Norway) has veriﬁed the oleﬁn yields and catalyst performance (Figure 22.5). The ﬂuidized-bed technology provides all advantages in terms of increased product yield, better quality, and a very efﬁcient heat recovery. UOP and Norsk Hydro have commercially manufactured the MTO catalyst (MTO-100), based on SAPO-34, which has shown the type of attrition resistance and stability suitable to handle multiple regeneration steps and ﬂuidized-bed conditions. The MTO technology is now ready for commercial use. In 1996, UOP announced their SAPO-34-based MTO process in cooperation with Norsk Hydro to be realized for construction of a 250 000 tons-per-year plant using a natural gas feedstock for production of ethene [5]. The ﬁrst commercial plant is scheduled to start in Nigeria [31, 58]. The process ﬂow scheme of the UOP/Hydro MTO process is shown in Figure 22.6. The UOP/Norsk Hydro MTO process design includes recovery of oxygenates, which are produced in minor quantities [44]. Finally, Mobil Corp., Union Rheinische Braunkohlen Kraftstoff AG and Uhde GmbH jointly designed, engineered, and operated a 4000-tons-per-year demo plant at Wesseling (Germany) in order to prove the ﬂuidized-bed MTG process. However, this concept was also demonstrated for the MTO process by applying a ZSM-5 zeolite as catalyst [60]. 22.3.4 Commercial Aspects/Economic Impact

There is a strong industrial interest in the MTO technology: Besides the MTO process owned by UOP/Norsk Hydro, ExxonMobil has a strong patent position;

22.3 Methanol to Oleﬁns (MTO)

however, limited information is available on the ExxonMobil MTO technology. Further, the Casale Group has rights to a Romanian MTO catalyst and Atoﬁna/UOP have certain perspectives concerning the gas-to-oleﬁns in the Middle East. In addition, Lurgi has developed a methanol-to-propene (MTP) process (see Section 22.4) demonstrated successfully in a pilot unit. An economical evaluation between an MTO unit and a naphtha cracker – based on a 500 kilo metric tons per annum (kMTA) production of ethene – has been performed: On the basis of a methanol price of US$90 per ton, the breakeven point between a naphtha cracker and an MTO unit is around a naphtha price of US$150 per ton, corresponding to a crude oil price of about US$15–16 per barrel. This can be taken as an indication of the competitiveness of an MTO unit with a naphtha cracker – with a delivered methanol price in the order of US$90 per ton [61]. 22.3.5 Future Perspectives

The MTO process provides an economically attractive alternative to naphtha cracking for the production of ethene and propene, and the corresponding polyoleﬁns (gas-to-polyoleﬁns – GTP) from methanol that can be produced at low cost and in large quantities from natural gas or other hydrocarbon sources. The UOP/Norsk Hydro MTO process and catalyst have been successfully demonstrated and are available for license. MTO/gas-to-oleﬁns – GTO/GTP is the most economical gas monetization alternative and the plant capacity is ﬂexible with respect to the gas ﬁeld size. MTO is very ﬂexible with respect to the ethene/propene ratio and the increased future propene demand can be met without an excessive coproduction of ethene. The oleﬁn market is large, has a healthy growth, and can consume large amounts of gas without being disturbed [58]. The UOP/Norsk Hydro MTO process can be economically viable in different scenarios: 1) Production of methanol at a remote gas ﬁeld site (less sensitive to the oil price) and transportation of the methanol to an MTO plant or to other downstream facilities located at the oleﬁns user’s site. 2) An integrated GTO complex at the gas ﬁeld site and transportation of oleﬁns or polyoleﬁn products to customers. 3) Increased oleﬁns production and feedstock ﬂexibility at an existing naphtha or ethane–propane cracker facility by installing an MTO reactor section and feeding into a revamped cracker fractionation section. 4) A smaller MTO unit using methanol produced in a single-train methanol plant to meet the local demand for oleﬁns or polyoleﬁns or both [58, 60]. Furthermore, MTO is also considered as a valuable option for the valorization of stranded gas reserves [31]. Finally, the market for ethene and propene is large compared to that of methanol and, therefore, the impact of MTO/GTO on methanol could be strong. Currently, several MTO alternatives are being studied. The abundance of natural gas and the

701

702

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

economy of large methanol plants lead to the conclusion that the MTO technology is an attractive alternative to naphtha cracking for the production of ethene and propene.

22.4 Methanol to Gasoline (MTG) 22.4.1 Catalysts and Reaction Conditions

As already mentioned, the MTG process was selected by the New Zealand government over the Fischer–Tropsch (SASOL process) in 1979 for converting natural gas from their extensive Maui and Kapuni ﬁelds to gasoline. At that time, Mobil’s ﬁxed-bed MTG process was unproven commercially, whereas the SASOL technology was already commercialized. The New Zealand plant started to produce about 14 500 barrels of gasoline per day from April 1986, supplying one-third of the nation’s gasoline demand. The gasoline production part of the factory was later closed down due to the price available for gasoline versus the price of methanol; however, the methanol production part is still in operation [5]. The commercial MTG reaction runs at temperatures of around 400 ◦ C at a methanol pressure of several atmospheres and uses a ZSM-5 catalyst. The gasoline produced is fully compatible with conventional gasoline. The conversion of MTHs and water is virtually complete and essentially stoichiometric. The reaction is exothermic. A simpliﬁed block diagram of the MTG process is shown in Figure 22.7 [60]. 22.4.2 Deactivation

During the MTG process, the ZSM-5 catalyst becomes less effective due to three forms of catalyst deactivation: 1) the deposition of a carbonaceous residue on and in the ZSM-5 catalyst; 2) an irreversible loss of activity due to the effect of steam on the zeolite structure; and 3) during regeneration, when coke is removed by oxidation, the high temperatures involved may effect the zeolite structure. These processes can be controlled by minimizing the regeneration temperature, the reaction temperature, and partial pressure of steam and adjusting operating procedures to minimize coke formation. The total number of cycles before a new ZSM-5 catalyst charge is required for the MTG process has not yet been reported, but catalyst life appears to be longer than two years, the period once anticipated [5].

22.5 Methanol to Propene (MTP) C2−

Superheat vaporize preheat

DME reactor

ZSM-5 reactors

HP sep.

Regeneration

Water to treating

LPG

Distillation

703

Light gasoline

HGT

Heavy gasoline Crude methanol

Figure 22.7 Block diagram of the ﬁxed-bed MTG process [60]. Reproduced by permission of Elsevier, Amsterdam.

22.4.3 Process Technology

Bench-scale studies of the MTG process have been performed both with ﬁxed-bed and ﬂuidized-bed reactors. Consequently, the major objective of the demo plant was to verify the bench-scale results. The only different variable between a bench-scale unit and a commercial-size reactor is the linear velocity of the reactants. Further, the ﬂuidized-bed MTG process has been proved in the 4000-tons-per-year demo plant at Wesseling (Germany) [60]. Finally, Mobil’s ﬁxed-bed MTG process was in operation from April 1986 at New Zealand, as mentioned earlier – with a production of about 14 500 barrels of gasoline per day to begin with.

22.5 Methanol to Propene (MTP)

The future demand of propene is much larger than that for ethene, and this gap can probably only be bridged by the application of the MTP technology. Currently, worldwide about 70% of the propene is produced via steam cracking, followed by ﬂuidized catalytic cracking (FCC) with about 28% (Figure 22.8). However, a ZSM-5-based catalyst developed by S¨ud-Chemie AG is used in Lurgi’s MTP process. A slower deactivation rate of this catalyst allows the application of a ﬁxed-bed reactor technology, which is adiabatically operated. After a DME production intermediate step, the DME/methanol/water mixture enters the MTP section, which includes three reactors in series with intermediate cooling (Figure 22.9). Almost thermodynamic equilibrium is obtained. The DME/methanol/water mixture is directed to the MTP reactor at the same time with steam and recycled oleﬁns. The process operates at about 425 ◦ C and a low pressure of about 1.5 bar, with propene as the main product (about 70% yield and 97% selectivity) and gasoline-range compounds as by-products – which are recycled. Two reactors are operating in parallel, whereas the third reactor is in standby mode or is used for

Finished gasoline

704

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG) 70 70 60 50

(%)

40 28 30 20 10

2

0 Propene production Dehydrogenation/MTO/MTP FCC Steam cracking

Figure 22.8 Current propene production ﬁgures. Reproduced by permission of Statoil and Lurgi.

Methanol Fuel gas internal use Propylene DME pre reactor

Product conditioning

MTP reactors (2 operating + 1 regenerating) Olefin recycle

LPG

Gasoline Product fractionation

Water recycle Process water for internal use

Figure 22.9 MTP process: simpliﬁed process ﬂow diagram. Reproduced by permission of Statoil and Lurgi.

22.6 TIGAS Process

705

regeneration, which is necessary after about 500–600 hours of operation. The low coke formation allows these long cycle times and regeneration in situ using diluted air at almost the reaction temperature, thus avoiding thermal stress on the catalyst. The process has been optimized with more than 8000 hours of pilot plant operation and has been demonstrated at Tjeldbergodden, Norway, in cooperation with Statoil [31].

22.6 TIGAS Process

The MTG plant in New Zealand has been combined with a methane steamreforming unit for the production of synthesis gas and a methanol plant to produce gasoline from natural gas. In this so-called Topsøe integrated gasoline synthesis (TIGAS) process (developed by Haldor Topsøe AS), the two process steps – the methanol synthesis and the MTG process – are integrated into one single synthesis loop without isolation of methanol as an intermediate (Figure 22.10). The process economics have been improved by a clever combination and close integration of the different steps. The purpose of the process development work on the integrated gasoline synthesis was to modify the three process steps: synthesis gas production, oxygenate synthesis, and the MTG conversion – in order to be able to operate all steps at the same procedure and the last two steps in one single synthesis loop [60, 62–64]. By selecting combined steam reforming and autothermal reforming for the synthesis gas production, and by using a multifunctional catalyst system (not based on zeolites or related microporous materials), producing a mixture of oxygenates instead of only methanol, the processes can operate at the same pressure of about 20 bar. The TIGAS process was developed due to the fact that future synthetic fuel plants will be built in remote areas where the price of natural gas is very low and not related to gasoline [60, 62–64].

Recycle gas

Natural gas Steam Oxygen

Purge Synthesis gas production

Oxygenate synthesis

Gasoline synthesis

Separation unit

LPG Water Gasoline

Figure 22.10 Haldor Topsøe’s TIGAS process [62]. Reproduced by permission of Elsevier, Amsterdam.

706

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

22.7 Mobil’s Oleﬁn-to-Gasoline and Distillate Process (MOGD)

1973 marked the beginning of the energy crisis, and new interest in synfuels and other chemicals favored the continuation of the MTH research. Already the MTG and MTO processes represent a sort of chemical factory, to be brought onstream as the technological and/or economic demands arise. One can go a step further and convert the oleﬁns to an entire spectrum of products, through yet another ZSM-5-based process: Mobil’s oleﬁn-to-gasoline and distillate process (MOGD), originally developed as a reﬁnery process, which works well coupled with the MTO process. In the MOGD reaction, ZSM-5 oligomerizes light oleﬁns, from either reﬁnery streams or the MTO process, into higher molecular weight oleﬁns that fall into the gasoline, distillate/diesel, and lubricant range (Figure 22.1) [5]. 22.7.1 Catalyst and Process Operation

Both processes use ZSM-5 as catalyst and the combined MOGD process offers a high valuable gasoline (C5 –C10 ) and distillate/diesel (C10 –C20 ) yield in various proportions – starting from light oleﬁns (C3 = to C4 =). This process is conducted by a product pattern constrained by both the shape selectivity of the zeolite catalyst as well as the thermodynamics governing the oligomerization reaction. In the MOGD process, the gasoline and distillate/diesel selectivity is larger than 95% of the oleﬁns in the feed and the gasoline/distillate product ratios range from 0.2 to >100. The distillate product is mostly isoparafﬁnic and is an exceptionally good blending stock due to its high cetane index, low pour point, and negligible sulfur content. The physical properties, such as ﬂash point, boiling range, and viscosity, are comparable with conventional distillate/diesel fuels. While the ZSM-5 pore architecture determines the product pattern (branched hydrocarbons), the conditions of the reaction (temperature, pressure, space velocity) have an inﬂuence on the product’s molecular weight. In general, the reaction window for oleﬁns over ZSM-5 is large, with reactions having been demonstrated for temperatures as low as 40 ◦ C and pressures as low as 7 kPa and as high as 14 000 kPa. There is no upper limit concerning the reaction temperature; however, equilibrium constrains become signiﬁcant at about 330 ◦ C [60, 65, 66]. Commercial interest with respect to this reaction is mainly under conditions where there is essentially a complete conversion, and the primary effect of the process variables is to alter the average product molecular weight [66]. 22.7.2 Thermodynamic Considerations

The conditions that can be used for condensation of light oleﬁns overlap those for cracking of large parafﬁns or oleﬁns. Thus, the consideration of chemical equilibrium can clarify the effects of process conditions on yields of heavy hydrocarbons

22.8 Summary and Outlook

obtainable. The problem in calculating equilibrium yields is that an extremely large number of compounds is involved and very few of the required free energy data are available. The computations can be greatly simpliﬁed by the fact that when a group of isomers are in equilibrium with each other they can be treated as a single compound while calculating their equilibrium with other compounds [66]. 22.7.3 Technical Process

The transformation of these oleﬁn reactions into a large-scale process requires that a number of technical aspects be taken into account. The two main concerns for the MOGD process are the heat control of the reaction and maximization of the yield of either the gasoline- or the distillate/diesel-range products. The solution chosen consists of four ﬁxed-bed reactors – three on-line and one in the regeneration mode – during the course of technical operation. The three on-line reactors are operated in series with interstage cooling and condensed liquid recycle to control the heat of reaction. The oleﬁns feed is mixed with a gasoline recycle stream and passed, after heating, through the three reactors. The conceptual design allows that both the maximum gasoline as well as the maximum diesel modes be envisaged by shifting of the reactor temperature and recycle composition. In order to generate a gasoline-rich stream for recycle to the reactors, a fractionation is applied. The recycle also improves the distillate/diesel selectivity [60, 66].

22.8 Summary and Outlook

Light oleﬁns like ethene and propene are the key building blocks in the petrochemistry, with annual productions (2006 ﬁgures) of about 110 × 106 and 65 × 106 tons per annum, respectively. Almost all ethene production comes from steam cracking of naphtha and natural gas liquids, whereas about two-third of the propene production is obtained as a coproduct of steam cracking. The second largest source of propene, again as a coproduct, is from the FCC units of the crude oil reﬁneries (Figure 22.8). The future need of ethene and propene is expected to increase, especially with respect to propene. A signiﬁcant challenge is the availability of feedstocks and whether these feedstocks can meet the future demand, again especially regarding propene. Traditional naphtha steam crackers and reﬁnery FCC units cannot meet the future propene demand. This gap has to be bridged via propane dehydrogenation and the MTO technology. In addition, the costs of the conventional oleﬁn feedstocks are strongly connected to the crude oil price [44]. Finally, both lignocellulosic biomass and coal emerge as potential feedstocks for light oleﬁns through MTO because both can be used as raw materials in the production of methanol [44]. Currently, several commercial MTP and MTO projects are at different levels of development. These technologies are usually connected to the utilization of remote

707

708

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

natural gas often located far from the market. Therefore, the gas must be converted into an easily transportable product in order to be cost effective. Consequently, the logistic aspects are important with respect to the operation of MTO and MTP units, in particular, for the UOP/Norsk Hydro MTO process and the Lurgi MTP technology. The overall driving forces for the realization of these new technologies are the search for alternative, lower-cost feedstocks for the oleﬁn production, monetization of remote natural gas, and developments in methanol technology. Both technologies are based on innovative catalyst systems [44]. In conclusion, a broad variety of well-proven technologies for the production of hydrocarbons from methanol is established; however, their future perspectives depend strongly on the gasoline, methanol, and natural gas prices besides the logistic aspects [60]. Regarding the catalysts investigated, there is no doubt that the ZSM-5 and SAPO-34 systems have shown to be the best catalysts so far for the MTG and MTO processes, respectively. Although SAPO-34 deactivates faster than ZSM-5, it is, however, regenerable and more resistant to deactivation than the isomorphous zeolites due to the lower acid strength and acid site density [44]. Furthermore, a number of other catalysts have been tested for the MTO reaction, including Beta-zeolite, MeAPSOs, MeAPOs, and others; however, none of them have resulted in any commercial importance [44]. Anyway, the monetization of natural gas, crude oil, tar sand, shale bitumen, coal, and renewable energy sources with respect to the methanol conversion will be driven by the technology available, the fuel and light oleﬁns demand worldwide, and secure energy supply (among others), as outlined by G.A. Olah et al. in the recently published monograph entitled ‘‘Beyond Oil and Gas: The Methanol Economy’’ [67].

22.9 Outlook

Even though the described catalytic systems for the MTG and MTO processes are already well developed, there is still room for improvement, like the enhanced hydrothermal stability of SAPO-34 by ammonia treatment [68]. Completely new microporous materials are under development, like the recently introduced inorganic–organic hybrid materials (MOFs, COFs, ZIFs, etc.), and tailoring these new type of porous materials toward application in catalysis (by introducing proper acidity, thermal stability, etc.) might be one of the future challenges, not necessarily with respect to the MTG and MTO reactions. References 1. Marcilly, C. (2001) Evolution of Re-

ﬁning and Petrochemicals. What is the place of zeolites in Studies in Surface Science and Catalysis, Vol. 135, Zeolites and Mesoporous Materials at the Dawn of the 21st Century (eds

A. Galarneau, F. Di Renzo, F. Fajula, and J. Vedrine), Elsevier, Amsterdam, p. 37. 2. Chheda, J.N., Huber, G.W., and Dumesic, J.A. (2007) Angew. Chem. Int. Ed. Engl., 46, 7164.

References 3. Huber, G.W. and Corma, A. (2007) 4. 5. 6. 7. 8.

9.

10.

11. 12.

13.

14.

15.

16. 17. 18.

Angew. Chem. Int. Ed. Engl., 46, 7184. St¨ocker, M. (2008) Angew. Chem. Int. Ed. Engl., 47, 9200. St¨ocker, M. (1999) Micropor. Mesopor. Mater, 29, 3. Chang, C.D. and Silvestri, A.J. (1987) Chemtech, 10, 624. Meisel, S.L. (1988) Chemtech, 1, 32. Vora, B.V., Marker, T.L., Barger, P.T., Nilsen, H.R., Kvisle, S., and Fuglerud, T. (1997) Natural Gas Conversion in Studies in Surface Science and Catalysis, Vol. 107, Economic Route for Natural Gas Conversion to Ethylene and Propylene (eds M. de Pontes, R.L. Espinoza, C.P. Nicolaides, J.H. Scholz, and M.S. Scurrell), Elsevier, Amsterdam, p. 87. Froment, G.F., Dehertog, W.J.H., and Marchi, A.J. (1992) Catalysis. Rev. Lit., London, 9, 1. Chang, C.D. (1988) Mechanism of Hydrocarbon Formation from Methanol in Methane Conversion, Studies in Surface Science and Catalysis, Vol. 36 (eds D.M. Bibby, C.D. Chang, R.F. Howe, and S. Yurchak), Elsevier, Amsterdam, p. 127. Hutchings, G.H. and Hunter, R. (1990) Catal. Today, 6, 279. van den Berg, J.P., Wolthuizen, J.P., and van Hooff, J.H.C. (1980) The Conversion of Dimethylether to Hydrocarbons on Zeolite H-ZSM-5. The Reaction Mechanism for Formation of Primary Oleﬁns in Proceedings 5th International Zeolite Conference (Naples) (ed. L.V. Rees), Heyden, London, p. 649. Chao, K.-J. and Huarng, L.-J. (1984) Proceedings 8th International Congress on Catalysis (Berlin), Verlag Chemie, Weinheim, p. V–667. Hutchings, G.J., Jansen van Rensburg, L., Pickl, W., and Hunter, R. (1988) J. Chem. Soc. Faraday Trans. I, 84, 1311. Hutchings, G.J., Gottschalk, F., Hall, M.V.M., and Hunter, R. (1987) J. Chem. Soc. Faraday Trans. I, 83, 571. Chang, C.D. and Silvestri, A.J. (1977) J. Catal., 47, 249. Venuto, P.B. and Landis, P.S. (1968) Adv. Catal., 18, 259. Swabb, F.A. and Gates, B.C. (1972) Ind. Eng. Chem. Fundam., 11, 540.

19. Salvador, P. and Kladnig, W. (1977) J.

Chem. Soc. Faraday Trans. I, 73, 1153. 20. Dass, D.V., Martin, R.W., Odell,

21.

22. 23. 24. 25.

26.

27. 28. 29. 30. 31.

A.L., and Quinn, G.W. (1988) A re-examination of evidence for carbene (CH2 :) as an intermediate in the conversion of methanol to gasoline. The effect of added propane in Methane Conversion, Studies in Surface Science and Catalysis, Vol. 36 (eds D.M. Bibby, C.D. Chang, R.F. Howe, and S. Yurchak), Elsevier, Amsterdam, p. 177. Hutchings, G.J., Hunter, R., Pickl, W., and Jansen van Rensburg, L. (1988) Hydrocarbon formation from Methanol Using WO3 /AL2 O3 and Zeolite ZSM-5 Catalyst: a Mechanistic study in Methane Conversion, Studies in Surface Science and Catalysis, Vol. 36 (eds D.M. Bibby, C.D. Chang, R.F. Howe., and S. Yurchak), Elsevier, Amsterdam, p. 183. Ono, Y. and Mori, T. (1981) J. Chem. Soc. Faraday Trans. I, 77, 2209. Nagy, J.B., Gilson, J.P., and Derouane, E.G. (1979) J. Mol. Catal., 5, 393. Kagi, D. (1981) J. Catal., 69, 242. Kolboe, S. (1988) On the mechanism of hydrocarbon formation from methanol over protonated zeolites in Methane Conversion, Studies in Surface Science and Catalysis, Vol. 36 (eds D.M. Bibby, C.D. Chang, R.F. Howe, and S. Yurchak), Elsevier, Amsterdam, p. 189. Mole, T. (1988) Isotopic and Mechanistic Studies of Methanol Conversion in Methane Conversion, Studies in Surface Science and Catalysis, Vol. 36 (eds D.M. Bibby, C.D. Chang, R.F. Howe, and S. Yurchak), Elsevier, Amsterdam, p. 145. Chang, C.D., Hellring, S.D., and Pearson, J.A. (1989) J. Catal., 115, 282. Zatorski, W. and Krzyzanowski, S. (1978) Acta Phys. Chem., 29, 347. Dahl, I.M. and Kolboe, S. (1994) J. Catal., 149, 458. Dahl, I.M. and Kolboe, S. (1993) Cat. Lett., 20, 329. Bellussi, G. and Pollesel, P. (2005) Industrial applications of zeolite catalysts: production and uses of light oleﬁns in Molecular Sieves: From Basic Research to Industrial Applications, Studies in Surface Science and Catalysis, Vol. 158 (eds

709

710

22 Methanol to Oleﬁns (MTO) and Methanol to Gasoline (MTG)

32.

33. 34.

35.

36. 37. 38. 39.

40.

41.

42. 43. 44.

45.

46.

ˇ J. Cejka, N. Zilkov´a, and P. Nachtigall), Elsevier, Amsterdam, p. 1201. Marcus, D.M., McLachlan, K.A., Wildman, M.A., Ehresmann, J.O., Kletnieks, P.W., and Haw, J.F. (2006) Angew. Chem. Int. Ed. Engl., 45, 3133. Arstad, B., Nicholas, J.B., and Haw, J.F. (2004) J. Am. Chem. Soc., 126, 2991. Haw, J.F., Song, W., Marcus, D.M., and Nicholas, J.B. (2003) Acc. Chem. Res., 36, 317. Olsbye, U., Bjørgen, M., Svelle, S., Lillerud, K.-P., and Kolboe, S. (2005) Catal. Today, 106, 108, references cited therein. Svelle, S., Rønning, P.O., and Kolboe, S. (2004) J. Catal., 224, 115. Arstad, B. and Kolboe, S. (2001) Catal. Lett., 71, 209. Arstad, B. and Kolboe, S. (2001) J. Am. Chem. Soc., 123, 8137. Mikkelsen, Ø., Rønning, P.O., and Kolboe, S. (2000) Micropor. Mesopor. Mater., 40, 95. Sassi, A., Wildman, M.A., Ahn, H.J., Prasad, P., Nicholas, J.B., and Haw, J.F. (2002) J. Phys. Chem. B, 106, 2294. Howe, R.F. (1988) Methanol to gasoline: spectroscopic studies of chemistry and catalyst in Methane Conversion, Studies in Surface Science and Catalysis, Vol. 36 (eds D.M. Bibby, C.D. Chang, R.F. Howe, and S. Yurchak), Elsevier, Amsterdam, p. 157. Hunger, M., Seiler, M., and Horvath, T. (1999) Catal. Lett., 57, 199. Seiler, M., Schenk, U., and Hunger, M. (1999) Catal. Lett., 62, 139. Kvisle, S., Fuglerud, T., Kolboe, S., Olsbye, U., Lillerud, K.P., and Vora, B.V. (2008) Methanol-to-Hydrocarbons in Handbook of Heterogeneous Catalysis, 2nd edn (eds E. Ertl., H. Kn¨ozinger., F. Sch¨uth., and J., Weitkamp), Wiley-VCH Verlag GmbH & Co KGaA, Weinheim, p. 2950. McCann, D.M., Lesthaeghe, D., Kletnieks, P.W., Guenther, D.R., Hayman, M.J., Van Speybroeck, V., Waroquier, M., and Haw, J.F. (2008) Angew. Chem. Inter. Ed. Engl., 47, 5179. Bos, A.N.R. and Tromp, P.J.J. (1995) Ind. Eng. Chem. Res., 34, 3808.

47. Chen, N.Y. and Reagan, W.J. (1979) J.

Catal., 59, 123. 48. Chang, C.D. (1980) Chem. Eng. Sci.,

35, 619. 49. Sedran, U., Mahay, A., and De Lasa,

H.I. (1990) Chem. Eng. Sci., 45, 1161. 50. Mihail, R., Straja, S., Maria, G., Musca,

51.

52.

53.

54.

55. 56.

57.

58.

59.

60. 61.

62.

63.

G., and Pop, G. (1983) Ind. Eng. Chem. Process Res. Div., 22, 532. Hutchings, G.J., Gottschalk, F., and Hunter, R. (1987) Ind. Eng. Chem. Res., 26, 637. Gayubo, A.G., Aguayo, A.T., Alonso, A., Atutxa, A., and Bilbao, J. (2005) Catal. Today, 106, 112. Chen, D., Rebo, H.P., Grønvold, A., Moljord, K., and Holmen, A. (2001) 6th World Congress of Chemical Engineering, September 23–27, Melbourne. Dahl, I.M., Wendelbo, R., Andersen, A., Akporiaye, D., and Mostad, H. (1999) Microporous Mesoporous Mater., 29, 159. Song, W. and Haw, J.F. (2003) Angew. Chem. Int. Ed. Engl., 42, 892. Chen, D., Moljord, K., Fuglerud, T., and Holmen, A. (1999) Microporous Mesoporous Mater., 29, 191. Barger, P. (2002) Methanol to oleﬁns (MTO) and beyond in Zeolites for Cleaner Technologies, Catalysis Science Series, Vol. 3 (eds M. Guisnet and J.-P. Gilson), Imperial College Press, Danvers, p. 239. Kvisle, S., Nilsen, H.R., Fuglerud, T., Grønvold, A., Vora, B.V., Pujado, P.R., Barger, P.T., and Andersen, J.M. (2002) Erd¨ol Erdgas Kohle, 118, 361. Yuen, L.-T., Zones, S.I., Harris, T.V., Gallegos, E.J., and Auroux, A. (1994) Microporous Mater., 2, 105. Keil, F.J. (1999) Microporous Mesoporous Mater., 29, 49. Heber, J. (2000) 3rd Asia Oleﬁns and Polyoleﬁns Markets Conference 2000, January 20–21, Bangkok. Topp-Jørgensen, J. (1988) Topsøe Integrated gasoline synthesis – The TIGAS process in Methane Conversion, Studies in Surface Science and Catalysis, Vol. 36 (eds D.M. Bibby, C.D. Chang, R.F. Howe, and S. Yurchak), Elsevier, Amsterdam, p. 293. Maxwell, I.E. and Stork, W.H.J. (2001) Hydrocarbon processing with zeolites in Introduction to Zeolite Science and

References Practice, Studies in Surface Science and Catalysis, Vol. 137, 2nd edn (eds H. van Bekkum, E.M. Flanigen, P.A. Jacobs, and J.C. Jansen), Elsevier, Amsterdam, p. 747. 64. St¨ ocker, M. (2005) Microporous Mesoporous Mater., 82, 257. 65. Avidan, A.A. (1988) Gasoline and Distillate Fuels from methanol in Methane Conversion, Studies in Surface Science and Catalysis, Vol. 36 (eds D.M. Bibby, C.D. Chang, R.F. Howe, and S. Yurchak), Elsevier, Amsterdam, p. 307.

66. Tabak, S.A., Krambeck, F.J., and

Garwood, W.E. (1986) AIChE J., 32, 1526. 67. Olah, G.A., Goeppert, A., and Surya Prakash, G.K. (2006) Beyond Oil and Gas: The Methanol Economy, Wiley-VCH Verlag GmbH & Co KGaA, Weinheim. 68. Mees, F.D.P., Martens, L.R.M., Janssen, M.J.G., Verberckmoes, A.A., and Vansant, E.F. (2003) Chem. Commun., 44.

711

713

23 Metals in Zeolites for Oxidation Catalysis Takashi Tatsumi

23.1 Introduction

Zeolites have long been used as solid acid catalysts. The ﬂuid catalytic cracking of heavy fractions of petroleum by the use of Y-type zeolites included in catalyst matrices is the world’s largest catalytic process. Acidic zeolites have also widely replaced mineral and Lewis acids in large-scale chemical manufacturing such as alkylation of aromatics and the Beckmann rearrangement. Zeolites are also endowed with the redox property by the incorporation of a variety of metals, and this chapter deals with incorporation of metals and the resultant oxidation catalysis. The ways to introduce heteroatoms into zeolites are classiﬁed into two categories: heteroatoms can be introduced into the framework as well as into voids as extraframework species. Most zeolites have an intrinsic ability to exchange cations [1]. This ability to exchange is a result of isomorphous substitution of a cation of trivalent (mostly Al) or lower charges for Si as a tetravalent framework cation. As a consequence of this substitution, a net negative charge develops on the framework of the zeolite, which has to be neutralized by cations present within the channels or cages that constitute the microporous part of the crystalline zeolite. These cations may be any of the metals, metal complexes, or alkylammonium cations. If these cations are transition metals with redox properties, they can act as active sites for oxidation reactions. As a pioneering work, Wacker-type reactions were catalyzed by Y zeolite into which Pd2+ and Cu2+ were incorporated by ion exchange [2]. Research on coordination chemistry in zeolites started in 1970s and early work was summarized by Lunsford [3]. A metal complex of the appropriate dimensions can be encapsulated in a zeolite, being viewed as a bridge between homogeneous and heterogeneous systems. Complexes that are smaller than the free diameters of the channels and windows have access to the cavities. On the other hand, complexes that are larger than the diameters of the windows must be synthesized in situ, namely, by the adsorption of the ligands into the zeolites containing transition metal ions or by the synthesis of the ligands in those zeolites [4–6]. Herron et al. ﬁrst referred to such zeolite guest molecules as ship-in-a-bottle complexes [7]. Cationic complexes can be tethered to zeolites through the electrostatic interaction. However, Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

714

23 Metals in Zeolites for Oxidation Catalysis

it is noteworthy that ship-in-a-bottle complexes, even if they are neutral, can be stabilized in zeolite pores [8]. This type of complex encapsulation does not require interaction between the host zeolite and the guest complex, allowing the catalytic properties of homogeneous complexes to be mimicked closely. Since the ﬁrst report on the synthesis of a metal phthalocyanine inside zeolite Na-Y in 1977, numerous examples of encapsulation of metal phthalocyanine complexes have been provided. Related heme-, polyaza-type, and N,N -bis(salicylidene)ethylenediimine (SALEN) complexes have also been trapped in a zeolite cavity that has restricted apertures. These are typical examples of ship-in-a-bottle complexes, being biomimetic models for dioxygen binding, oxygenase, and photosystems [9]. The other way to introduce heterometals is their isomorphous substitution for Si in the framework, in a similar manner to the isomorphous substitution of Al. The heteroatoms should be tetrahedral (T) atoms. In hydrothermal synthesis, the type and amount of T atom, other than Si, that may be incorporated into the zeolite framework are restricted due to solubility and speciﬁc chemical behavior of the T-atom precursors in the synthesis mixture. Breck has reviewed the early literature where Ga, P, and Ge ions were potentially incorporated into a few zeolite structures via a primary synthesis route [10]. However, until the late 1970s, exchangeable cations and other extraframework species had been the primary focus of the researchers. The isomorphous substitution of Ti for Si was claimed by Taramasso et al. in 1983 [11]. The resulting material has the structure of silicalite-1 (pure silica MFI) with Ti in the framework positions and named titanium silicalite-1. or (TS-1) The new ﬁndings including the claim that other metals can be inserted into the zeolite lattice met with skepticism. Ione et al. predicted the probability of isomorphous substitution of metal ion (Mn+ ) and the stability of the Mn+ position in the tetrahedrally surrounding oxygen atoms by using the Pauling criterion [12]. On the basis of the ratio of ionic radii ρ of the cation and anion, the value for Ti and O (ρ = 0.515) falls out of the range (ρ = 0.225–0.414) for which tetrahedral coordination is expected [13]. The allowed cation would include only Al3 + , Mn4 + , Ge4 + , V5 + , Cr6 + , Si4 + , P5 + , Se6 + , and Be2 + . Presumably this type of estimate is surely effective, which can explain the preference of B3 + for trigonal coordination and the resultant instability of B3+ in the zeolite matrix. However, it is a very rough approximation since the completely ionic character of T–O bond is not the case and the model assumes that the atoms have a perfect round shape. In this chapter, the oxidation catalysis exhibited by isolated metals incorporated into the framework of zeolites will mainly be dealt with. TS-1 proved to be a very good catalyst for liquid-phase oxidation of various organic compounds using H2 O2 as oxidant, and several industrial processes utilizing TS-1 are being operated [14]. Most of this chapter is devoted to titanium-containing zeolites, although zeolites incorporating vanadium, chromium, cobalt, and tin are brieﬂy mentioned. The success of TS-1 has encouraged the researchers to synthesize other titanosilicates with different zeolite structures, especially those with larger pores, since TS-1 encountered with a limitation of inapplicability to bulky molecules owing to the medium pores of 10-ring. Table 23.1 lists the representative titanosilicates prepared by various techniques.

23.2 Titanium-Containing Zeolites Table 23.1

Representative titanosilicates.

Material

Structure code

Channel system

Preparation methodsa

References

TS-1 TS-2 Ti-ZSM-48 Ti-beta TAPSO-5 Ti-ZSM-12 Ti-MOR Ti-ITQ-7 Ti-MWW T-YNU-2

MFI MEL N.A.b *BEA AFI MTW MOR ISV MWW MSE

10-10 10-10 10 12-12-12 12 12 12-8-8 12-12-12 10, 10 12-10-10

HTS HTS HTS HTS, F – , DGC HTS HTS PS HTS HTS, PS PS

[10] [13] [14] [15–17] [18] [19] [20] [21, 22] [23, 24] [25]

a HTS, hydrothermal synthesis in alkali media; dry gel conversion (DGC); PS, postsynthesis; F− , ﬂuoride media method. b Not assigned.

23.2 Titanium-Containing Zeolites 23.2.1 TS-1

A very comprehensive review was made on TS-1 and other titanium-containing molecular sieves [26]. TS-1 was synthesized by the hydrothermal crystallization of a gel obtained from Si(OC2 H5 )4 and Ti(OC2 H5 )4 [11, 27] (Enichem method, hereafter named method A). The incorporation of Ti into the framework of MFI structure was demonstrated by the increase in unit-cell size in XRD pattern, as shown in Figure 23.1 [28], and the appearance of tetrahedral Ti species in the UV–vis spectra. The maximum amount of Ti that can be accommodated in the framework positions is claimed to be limited to x = Ti/(Ti+Si) of 0.025. TS-1 is capable of serving as a highly efﬁcient catalyst for the oxidation of various organic substrates, for example, alkanes, alkenes, alcohols, and aromatics, with H2 O2 as an oxidant under mild conditions [15, 29–32]. The epoxidation reaction catalyzed by TS-1 may be performed under mild conditions in dilute aqueous or methanolic solution. The active oxygen content of H2 O2 , 47 wt% (16/34), is much higher than that of organic peracids and hydroperoxides; water is the only co-product. Besides epoxidation, TS-1 catalyzes a broad range of oxidation reactions with hydrogen peroxide as the oxidant, as shown in Table 23.2. It is to be noted that, in the presence of alkalis, extraframework Ti species are formed, giving rise to inferior catalytic properties [15]. It has been widely believed that the presence of an alkali metal, even in very small amounts, cramps the activity of TS-1 for the oxidations with H2 O2 by preventing the insertion of titanium into the silicalite framework. However, Khouw and Davis reported that the presence

715

20.16

20.00

20.14

19.95

Axis b (Å)

Axis a (Å)

23 Metals in Zeolites for Oxidation Catalysis

20.12 20.10 20.08 0.000

0.010 0.005

0.015

Unit-cell volume V (Å3)

13.40 13.38 13.36

(c)

0.010 0.005

0.020

0.005 0.015 0.025 Atomic ratio Ti/(Ti + Si) (d)

0.020 0.015

0.025

Atomic ratio Ti/(Ti + Si)

(b)

13.42

0.010

19.95

0.025

13.44

0.000

19.90

19.80 0.000

0.020

Atomic ratio Ti/(Ti + Si)

(a)

Axis c (Å)

716

5400 5380 5360 5340 5320 0.000

0.010 0.020 0.005 0.015 0.025 Atomic ratio Ti/(Ti +Si)

Figure 23.1 The increase in unit-cell size against the incorporation of Ti into the framework of MFI structure [26]. Table 23.2

Catalytic chemistry with TS-1.

Substrate

Product

Oleﬁns Oleﬁns and methanol Dioleﬁns Phenol Benzene Parafﬁns Primary alcohols Secondary alcohols Ammoximation of cyclohexanone N,N-Dialkylamines Thioethers

Epoxides Glycol monomethyl ethers Monoepoxides Hydroquinone and catechol Phenol Alcohols and ketones Aldehydes Ketones Oximes N,N-Dialkylhydroxylamines Sulfoxides

of alkali-metal ion in the preformed TS-1 does not have any signiﬁcant effect on the activity [33]; although neither sodium-exchanged TS-1 nor TS-1 synthesized in the presence of high alkali-metal concentrations (Si/Na < 20) is active for alkane oxidation, the catalytic activity can be restored by washing the solid with acid solution. The restoration of the activity may be ascribed to the conversion of the Na-exchanged TS-1 into its original form as shown in Scheme 23.1 No satisfactory explanation has yet been offered for the lack of activity of sodium-exchanged Ti species. If this acid treatment is generally applicable, it will be useful for

23.2 Titanium-Containing Zeolites Si

Si O

O

NaOH

O− O Na+

H2SO4

O

Ti O

O

Si

Scheme 23.1

Si

H

Si

Si

Si O Ti O Si

Possible interconversion between TS-1 and Na-exchanged TS-1.

synthesizing a variety of titanium silicate structures that require the presence of alkali-metal ions for their crystallization. The catalytic properties of TS-1 depend on the lattice Ti content, which is usually less than 2 wt% [34, 35]. The effective way to increase the Ti content in the framework of TS-1 is still a big challenge. Thangaraj and Sivasanker reported that eight Ti ions could be incorporated in the lattice sites per unit cell (Si/Ti = about 10) by an improved method (method B) in which titanium tetra-n-butoxide was ﬁrst dissolved in isopropyl alcohol before addition to the aqueous solution of hydrolyzed tetraethyl orthosilicate for the purpose of avoiding the formation of TiO2 precipitate by reducing the hydrolysis rate of the alkoxide [36], but Schuchardt and his coworkers could not reproduce it, and found that there was no difference in the framework Ti content between the samples synthesized by the methods A and B [37]. To synthesize Ti-rich TS-1, it is necessary and helpful to make its crystallization mechanism clear. However, very few reports have devoted to the study of this subject [38]. The crystallization process of titanosilicates is much more complex than that of aluminosilicates, because Ti4+ has a weak structure-directing role and is much more difﬁcult to be incorporated into the framework than Al3+ . Isomorphous substitution of metal atoms for Si in zeolites is not only related to zeolite structures/framework composition ﬂexibility and the chemical nature of metals but also strongly inﬂuenced by the crystallization mechanism. The framework composition ﬂexibility of zeolites is chemically important. Ti K-edge EXAFS studies have shown that the Ti–O bond length of tetrahedral Ti(OSi)4 species is about 1.80 A˚ in contrast to 1.61 A˚ for Si–O [39, 40]. The Ti–O bond is much longer than the Si–O bond, probably making the local structure around Ti seriously distorted. This results in the slow inclusion of Ti into the framework, compared to Si ions. If crystallization proceeded too fast, Ti ions would not have enough time to be incorporated into the lattice. However, crystallization that is too slow would possibly lead to the formation of transition metal oxides, preventing metal cations from being incorporated into the framework. In addition, the difﬁcult crystallization may also result from the strong competition between the interaction of soluble silicate ions and mother liquor and the condensation of silicate ions. Thirdly, a mismatch among hydrolysis of Ti and Si alkoxides, polymerization of Ti4+ and/or Si4+ ions, nucleation, and crystal growth would lead to much difﬁculty in the inclusion of Ti in the framework. Since the chemical nature of Ti and the rigidity of the framework of TS-1 cannot be altered, ﬁnding an effective crystallization-mediating agent would be the sole way to increase the lattice Ti content in TS-1 by harmonizing the hydrolysis rate of Ti alkoxide with that of silicate species as well as the nucleation and crystal growth rates.

717

718

23 Metals in Zeolites for Oxidation Catalysis

In this respect, a new route to the synthesis of TS-1 has been developed by using (NH4 )2 CO3 as a crystallization-mediating agent (the Yokohama National University (YNU) method) [41]. By this method, the framework Ti content can be signiﬁcantly increased without forming extraframework Ti species. The prepared catalyst has a Si/Ti ratio as low as 34, although under the same synthesis conditions the ratio of 58 was only achieved by the methods A and B established by the Enichem group [27] and Thangaraj and Sivasanker [36], respectively. The YNU samples are well crystalline, containing less defect sites than the samples synthesized by the other two methods. The 29 Si NMR spectra indicates that the YNU-50 sample has a higher Q4 /Q3 ratio of 23.1 than the Enichem method A-50 sample (4.9), which indicates that more defective sites are present in the A samples. This is consistent with the ﬁnding that the silanol groups or the defect sites decreased with increasing Ti content in the framework [34, 39], since Ti has a mineralizing effect and more Ti cations are incorporated into the YNU samples than in the A samples. Compared to the sample crystallized in the method-A system, TS-1 synthesized in the YNU system was more stable against the hydrolysis under the crystallization conditions after the crystallization was complete. This may be due to the incorporation of more Ti cations into the framework, associated with a reduction in defect sites, resulting in the higher hydrophobicity. This is also evidenced by the 1 H and 29 Si MAS NMR spectroscopic ﬁndings that silanol protons of H-bonded siloxy oxygens at defect sites proportionally decreased with the Ti atoms per unit cell [42]. Therefore, compared with the YNU samples, the A samples are not only relatively hydrophilic but might contain more ((SiO)3 TiOH) species. The higher hydrophilicity of the A samples than that of the YNU samples is further corroborated by the fact that the weight loss below 150 ◦ C due to the desorption of water was about 1.0 wt% for as-synthesized A-50 in contrast to 0.3 wt% for as-synthesized YNU-50. Although the catalytic properties of titanosilicates are related to their crystalline structures and/or characters, the hydrophilicity caused by the presence of more defect sites in titanosilcates would be unfavorable for the oxidation of hydrocarbon reactants, as suggested by the following facts: (i) compared with TS-1, Ti-MWW and Ti-beta with more defect sites showed very low activity for the oxidation of hexane, styrene, and benzene; (ii) Ti-MCM-41, with a low Q4 /Q3 ratio, showed very low activity in the oxidation of hexane and 1-hexene; however, silylation resulted in a remarkable improvement in the oxidation activity [43]. Furthermore, it has been proven that ((SiO)3 TiOH) species are much less active than ((SiO)4 Ti) species for the liquid-phase oxidation of organics over TS-1 [44, 45]. As a result, the YNU sample showed a much higher activity for the oxidation of various organic substrates, such as linear alkanes/alkenes and alcohols, styrene, and benzene, than the sample A, because of their higher hydrophobicity, as shown in Figure 23.2. However, it is to be noted that a contrary view on the activity of the ((SiO)3 TiOH) species was presented by Thomson et al. based on a DFT study [46]. Figure 23.3 shows the Si/Ti ratio of the solid samples collected in the whole crystallization process. The Si/Ti ratio in the solid samples synthesized in the YNU system was kept almost constant, being in the range of 48–50.5. This constant

23.2 Titanium-Containing Zeolites

80

A-50 YNU-20

60

40

20

0

n -Hexene

1-Hexene

Benzene

Styrene

Substrates Figure 23.2 Catalytic results of the oxidation of various organic substrates over the YNU-20 and A-50 catalysts (H2 O2 -based conversion (%): percent is used for the oxidation of substrates. Reaction conditions: for 1-hexene, 0.05 g catalyst, 10 ml methanol, 10 mmol substrate, 10 mmol H2 O2 , 60 ◦ C,

2 hours; for n-hexane, 0.1 g catalyst, 10 ml methanol, 10 mmol substrate, 20 mmol H2 O2 , 60 ◦ C, 4 hours; for benzene, 0.1 g catalyst, 10 ml sulfolane, 10 mmol substrate, 1 mmol H2 O2 , 100 ◦ C, 2 hours; for styrene, 0.1 g catalyst, 10 ml acetone, 5 mmol substrate, 2.5 mmol H2 O2 , 60 ◦ C, 4 hours).

150 Si/Ti molar ratio in the solid

YNU system Method-A system

120

90

60

30

0

A

B

C

D

E

F

G

Crystallization conditions Figure 23.3 Dependence of the Si/Ti molar ratio of the solid fraction on the crystallization conditions in the YNU and method-A systems. (A) 30 ◦ C, one day; (B) 60 ◦ C, one day; (C) 80◦ C, one day; (D) 100 ◦ C, one day; (E) 140 ◦ C, one day; (F) 170 ◦ C, one day; (G) 170 ◦ C, two days; and (H) 170 ◦ C, three days.

H

719

720

23 Metals in Zeolites for Oxidation Catalysis

Si/Ti ratio is similar to that found in the synthesis of TS-1 from amorphous wetness-impregnated SiO2 –TiO2 xerogels and the nonaqueous synthesis of zeolites with solid reaction mixtures where solid-phase transformation mechanism predominated [38, 47]. In addition, the solid yields recovered after calcination of all the YNU samples were always higher than 95% based on the added SiO2 and TiO2 . This is also consistent with the occurrence of a solid-phase transformation mechanism, which would be favorable for the incorporation of Ti cations into the framework, since dissociation, coalescence, and reorganization are the main processes during nucleation and crystal growth. On the other hand, as shown in Figure 23.3, the Si/Ti ratio of the samples synthesized by method A drastically increased during the period of crystal growth; the solid samples in the induction period were directly obtained by drying the liquid at 100 ◦ C since no solid was formed during this period. This suggests that during the period of rapid crystal growth in the method-A system, silicic and/or silicate species (≡Si–OH, ≡Si–O− ) were polymerized with each other at a much higher rate than the condensation of silicic or silicate species and titanic or titanate (≡Ti–OH, ≡Ti–O− ) species to form TS-1 crystals. The Ti content in the solid remarkably increased after the sample reached the highest crystallinity. The crystallization in the method-A system occurred via a homogeneous nucleation mechanism although precursor aggregates might be formed before nucleation. It can be also seen from Figure 23.3 that at this stage the Ti content in the as-synthesized YNU products was higher than in the A products, suggesting the incorporation of more Ti cations into the framework. In contrast, the evolution of the Ti/Si ratio in the mother liquid during the whole crystallization process is in agreement with that in the solid samples. In the method-A system, the Ti/Si ratio of the liquid phase sharply increased from the point when the nuclei formed, and reached the highest value when crystallization was almost completed. This also shows that in the method-A system Ti cations began to be inserted into the framework after the zeolitic architecture was almost built. In the YNU system, however, the Ti/Si ratio in the liquid was almost constant (about 0.02) during the induction period, being the same as that in the solid fraction. In the synthesis of TS-1 in the method-A system, the hydroxide ion of tetrapropylammonium hydroxide (TPAOH) plays a major role in the acceleration of the hydrolysis of silica source and Ti alkoxide, and the oligomerization of Si–OH and Ti–OH species (Eq. (23.1)). These processes would lower down the pH value. Therefore, a high pH value could promote these processes, and thus decrease the time for reaching the critical concentration of soluble silica and titanium species from which zeolites are ultimately precipitated. However, a very high alkaline condition would make the synthesis gel unreactive, but dissolve the silicate/titanosilicate species (Eq. (23.2), from right to left), resulting in a lower crystallinity of the products crystallized in the method-A system and the presence of more defect sites. In contrast, for the YNU system, the synthesis gel was quickly solidiﬁed after the addition of (NH4 )2 CO3 . Thus, most of the TPA+ species were embedded in the solid with the amount of free OH− species drastically decreasing and with the pH

23.2 Titanium-Containing Zeolites

value of the liquid lowering down. The lower the pH value, the higher the yield of crystalline material, or the better the crystalline product because the condensation reaction (Eq. (23.2), from left to right) proceeds to a larger extent. This accounts for the higher crystallinity of the TS-1 samples achieved by the YNU method than by the method A. + − − x(TPA+ OH− ) + T(OR)4 + (4 − x)H2 O − −− − − (TPA )x ( O)x − T)(OR)4−x + − − +xROH + (4 − x)H2 O − −− − − (TPA )x ( O)x − T)(OH)4−x )

+(4 − x)ROH(T = Si, Ti)

(23.1)

− − ≡ Si − O− + ≡ TOH − −− − − ≡ Si − O − T ≡ +OH (T = Si, Ti)

(23.2)

As expected, an increase in the pH value during the crystal growth period was observed for both synthesis systems, but the degree of increase was much more appreciable for the method-A system than for the YNU system. The pH value increased from 11.9 to 12.8 in the method-A system, while it increased only slightly from 10.4 to 10.7 in the YNU system. The increase in the pH value is attributed to the incorporation of silicate species into the framework of TS-1, which releases free OH− (Eq. (23.2)). This increase continued till the completion of crystal growth. The higher OH− concentration in the method-A system also accelerates the crystal growth. Such a high crystallization rate is not beneﬁcial for the incorporation of Ti into the framework since this process would make the local structure around Ti distorted. In addition, the high alkalinity of the liquid is unfavorable to the condensation of Ti–OH and silicate species. In contrast, the presence of (NH4 )2 CO3 appropriately slows down the crystallization rate by signiﬁcantly decreasing the pH value, buffering the synthesis gel and introducing NH4 + , which is a structure-breaking cation in water [48], and consequently reducing the polymerization rate of silicic/silicate/titanic/titanate species. This would provide enough time for Ti species to be inserted into the lattice during the crystallization process, as indicated by the much lower Si/Ti ratio in the solid samples obtained during the crystal growth period in the YNU system than in the method-A system. Thus, it is concluded that the presence of (NH4 )2 CO3 not only drastically lowered pH, slowing down the crystallization process and making the incorporation of Ti into the framework agree well with nucleation and crystal growth, but also modiﬁed the crystallization mechanism. It seems that the solid-phase transformation mechanism predominated in the crystallization process initiated by dissociation, reorganization, and recoalescence of the solidiﬁed gel although a small amount of nongelatinated Ti shifted to the solid during the crystal growth period. In contrast, a typical homogeneous nucleation mechanism occurred in the method-A system. The most widely accepted mechanism for TS-1-catalyzed epoxidation is the peracid-like mechanism, which involves a hydroperoxo rather than a peroxo species, and coordination of an alcohol or water molecule to the site (Scheme 23.2) [26, 31]. Ti isomorphously substituted for Si in the zeolite framework in tetrahedral coordination is much more resistant against hydrolysis compared to titanium species on amorphous silica. Thus, Ti species in TS-1 are negligibly leached out from

721

722

23 Metals in Zeolites for Oxidation Catalysis R

OSi SiO

H2O2

Ti OSi

Ti OOH HOSi

ROH (R = H, Me, Et)

Ti

OSi −H2O +H2O2

OH

−H2O

C

−ROH

C

C

R

HOSi O

O

H O H HOSi

O Ti

Scheme 23.2

O

Ti

O

C

C

H O H HOSi C

Mechanisms of epoxidation of titanosilicates.

the framework during the oxidation reactions in the liquid phase, in contrast to those on amorphous silica. A key characteristic of TS-1 zeolites is their relatively high hydrophobicity, resulting in the favorable adsorption of alkanes and other hydrocarbons, as described above. Thus, the low concentration of hydrogen peroxide present at all times in the catalyst results in its efﬁcient use. The strong hydrophobicity also enables fast desorption of the oxygenated products. Therefore, oxidations can occur up to high conversions with high H2 O2 selectivities and high efﬁciencies. TS-1 zeolite has the MFI structure and medium-pore 10-ring channels. The structure of TS-1 prevents all compounds having cross-sections larger than about 0.55 nm from diffusing inside TS-1 channels and, therefore, from reacting or interfering with reactions occurring at Ti sites. Thus, TS-1 shows remarkable shape selectivity. The rate of oxidation of the branched or cyclic hydrocarbons is much lower than that of the straight one [29, 49]. Hayashi et al. have reported that oxidation of 2-propanol over TS-1 is strongly retarded in the presence of 1-propanol [50]; 1-propanol coordinates strongly on the active site followed by slow oxidation, while 2-propanol coordinates weakly but is oxidized quickly. The importance of coordination of the group to be oxidized to the active site present in the sterically constrained environment in the zeolite pores has been revealed in the intra- and intermolecular competitive oxidation [51]. Since oxidations are irreversible in principle, and usually not accompanied by reactions that can carry out the interconversion of products, only reactant selectivity or restricted transition state selectivity can be practically important. 23.2.2 Ti-Beta

Since TS-1 is a medium-pore zeolite, the pore size of the zeolite-type framework restricts its use, even for small molecules such as simple cyclic alkenes. Thus, after the success of TS-1 as a liquid-phase oxidation catalyst, great efforts have been devoted to the synthesis of zeolites containing Ti in the framework for different zeolitic structures. One of the interesting titanium-containing zeolites is Ti-beta

23.2 Titanium-Containing Zeolites

(BEA*), a large-pore zeolite, which has been hydrothermally synthesized from gels containing tetraethylammonium hydroxide (TEAOH) as a structure-directing agent (SDA) [52]. Due to its large pore size, Ti-beta was shown to be more active than TS-1 for the oxidation of bulky substrates such as cyclic and branched molecules [53]. Ti-beta was usually obtained in very low yield [53], in contrast to TS-1. Moreover, an additional factor, viz., the presence of aluminum in the framework of Ti-beta, can contribute to the different catalytic behavior observed between Ti-beta and TS-1. In contrast to the hydrophobic characteristics of TS-1, the presence of Al and large concentration of internal and external silanol groups confers a rather hydrophilic character on Ti-beta. There is therefore a strong incentive for the preparation of Ti-beta with low Al content in a better zeolite yield by using new methods. Improved methods for the synthesis of Ti-beta, for example, use of special SDA [54], the ﬂuoride method [55], and the dry gel conversion method [56], have been developed to obtain the Ti-beta zeolites active for H2 O2 oxidation in high yields. The improvement of epoxide selectivity by selectively poisoning the acid function without spoiling the oxidation activity is attained by modiﬁcation by ion exchange with quaternary ammonium ions [57]. It is noteworthy that cleavage of the oxirane ring is promoted by the presence of H2 O2 , indicating that the acidity of titanosilicates is generated by the contact of Ti sites with H2 O2 [57]. The ﬂuoride method gives rise to a very hydrophobic Ti-beta free of framework Al and silanol defects. Although this material has been expected to have high catalytic activity, in the epoxidation of alkenes, the turnover number of Ti-beta synthesized by the ﬂuoride method was only comparable with that of hydrothermally synthesized Ti-beta. One reason might be the larger crystal size obtained by the ﬂuoride method decreasing the effectiveness of the solid catalyst. Koller et al. have reported the existence of ﬁve-coordinated Si species, (SiO)4 SiF− , in high-silica zeolites (MFI, BEA, MTW, etc.) synthesized in the ﬂuoride media, as conﬁrmed by the 19 F MAS NMR and 29 Si{19 F} CP MAS NMR techniques [58]. It has been proved that the presence of the ﬂuorine attached to the zeolite framework would be harmful to the catalytic activity of Ti-beta [59]. Treatment of Ti-beta synthesized by the ﬂuoride method with basic quaternary ammonium solution followed by calcination enhanced the epoxidation activity with hydrogen peroxide, which seems to be due to the reduction of the amount of ﬂuorine contained in the catalyst. FT-IR and UV-visible spectra revealed the increase in the Si–OH and Si–O–Ti groups and the decrease in the Ti species with high coordination number after the treatment followed by calcination. There is a different approach to substitution of metal atoms into the framework, namely, the secondary synthesis or postsynthesis method. This is particularly effective in synthesizing metallosilicates that are difﬁcult to be crystallized from the gels containing other metal atoms or hardly incorporate metal atoms by the direct synthesis method. Actually, substitution of Ti for Al goes back to the 1980s; the reaction of zeolites with aqueous solutions of ammonium ﬂuoride salts of Ti or Fe under relatively mild conditions results in the formation of materials that are dealuminated and contain substantial amounts of either iron or titanium

723

724

23 Metals in Zeolites for Oxidation Catalysis

and are essentially free of defects [60]. However, no sufﬁcient evidence for the Ti incorporation has been provided. The postsynthetic incorporation of titanium into the framework of zeolite beta has been achieved by liquid-phase treatment employing ammonium titanyl oxalate [61]. Ti-beta has also been prepared by treating Al-containing beta with a concentrated solution of perchloric or nitric acid in the presence of dissolved titanium [62]. Although Ti(OBu)4 and TiF4 were efﬁcient sources for the incorporation of tetrahedral Ti, use of TiO2 as the Ti source gives rise to octahedral Ti as well as tetrahedral Ti. In this case, the extraction of Al from the framework simultaneously occurred, giving almost Al-free Ti-beta. Postsynthesis gas–solid isomorphous substitution methods have also been known [63]. Ti-beta essentially free of trivalent metals can be prepared from boron-beta. However, the gas-phase method is not efﬁcient for Ti incorporation and could have some disadvantages such as the deposition of TiO2 [64]. 23.2.3 Ti-MWW

MWW aluminosilicate (generally known as MCM-22) is hydrothermally synthesized without difﬁculty; however, the synthesis of MWW titanosilicate (Ti-MWW) had been a challenge until it was shown for the ﬁrst time that Ti is effectively incorporated into the MWW framework when boric acid coexists in the synthesis media [22a]. Since titanosilicates generally require speciﬁc synthesis conditions in comparison to silicalites and aluminosilicates, much effort made to synthesize numerous zeolite structures has led to a very limited success. This has been also the case with the MWW zeolite. Although it is possible to hydrothermally synthesize MWW silicalite in alkali-free media using a speciﬁc organic SDA of trimethyladamantylammonium hydroxide, the addition of other metal cations such as Al and Ti into the synthesis gel results in failure [65]. When boric acid, with its amount even more than that of silicon, and a Ti source were coexistent in the synthesis gel composed of fumed silica and cyclic amine SDA such as hexamethyleneimine (HM) or piperidine (PI), Ti-MWW was crystallized readily by autoclaving the gel at 403–443 K [22a, b]. These are designated as Ti-MWW-HM or Ti-MWW-PI. All the as-synthesized samples of Ti-MWW–PI and Ti-MWW–HM showed the XRD patterns totally consistent with those of the lamellar precursor of MWW topology, generally designated as MCM-22(P) [66, 67]. Upon calcination at 803 K, all the samples were converted to the porous three-dimensional (3D) MWW structure with a good quality. The amount of B incorporated into the products was in the Si/B range of 11–13, far lower than that in the gel with the Si/B ratio of 0.75. In contrast, there was little difference in the Si/Ti ratios between the gel and the solid product except for the gel of Si/Ti = 100, indicating that the synthesis system is very effective for Ti incorporation. The UV-visible spectra of Ti-MWW–PI without calcination are shown in Figure 23.4a. No obvious band around 330 nm was observed even at a very

D

B

C B A

E

F

220

C D

330

260

A Absorbance

Absorbance

220

260

23.2 Titanium-Containing Zeolites

200 (a)

300

400

Wavelength (nm)

500

200 (b)

300

400

500

Wavelength (nm)

Figure 23.4 UV-visible spectra of Ti-MWW-type materials. (a) As-synthesized Ti-MWW-PI with the Si/Ti ratio of (A) 100, (B) 50, (C) 30, and (D) 10. (b) Acid-treated and further calcined Ti-MWW-PI with Si/Ti ratio of (A) 170, (B) 116, (C) 72, (D) 59, (E) 38, and (F) 17.

high Ti level, indicating that the anatase-like Ti phase was hardly formed during crystallization. However, the spectra are quite different from those reported for TS-1 and Ti-beta that generally show only a narrow band around 210 nm. Irrespective of the Si/Ti ratio and the SDA used, all as-synthesized Ti-MWW samples exhibited a main band at 260 nm together with a weak shoulder around 220 nm. The 220-nm band, resulting from the charge transfer from O2− to Ti4+ , has been widely observed for Ti-substituted zeolites and is characteristic of tetrahedrally coordinated Ti highly dispersed in the framework [30]. The 260-nm band has been attributed to octahedral Ti species, related to a kind of extraframework Ti species probably with Ti–O–Ti bonds in the case of Ti-beta [17]. Upon calcination the dehydroxylation between the lamellar sheets occurred to form the MWW structure. This recrystallization process also led to a change in the nature of the Ti species; a new band around 330 nm (Figure 23.4b) is ascribed to the anatase. The anatase-type Ti species are not active and may cause unproductive decomposition of the oxidant H2 O2 when employed as an oxidation catalyst. Once octahedral Ti species are converted to anatase, it could not be removed by washing with a HNO3 or H2 SO4 solution under reﬂuxing conditions. However, when as-synthesized Ti-MWW was ﬁrst reﬂuxed with an acid solution and then calcined, the octahedral Ti species were eliminated selectively (Figure 23.4b); only the narrow band at 220 nm due to tetrahedral Ti species was observed for the samples prepared by acid-treating the precursors with the Si/Ti ratio of 100−30. Extraframework Ti species, both octahedral and anatase-like, still remained to a certain level for the samples obtained from the precursors with Si/Ti of 20 and 10 because they contained very high concentrations of octahedral Ti. Thus, it should be emphasized that the pretreatment sequences are essential for obtaining Ti-MWW with tetrahedrally substituted Ti species. Together with extraframework Ti, a part of framework boron was also extracted to a level corresponding to the Si/B ratio of about 30. Large-pore titanosilicates developed after TS-1, for example, Ti-beta and Ti-ITQ-7 [17–19, 68], and mesoporous titanosilicates Ti-MCM-41 and Ti-MCM-48 [69] have

725

726

23 Metals in Zeolites for Oxidation Catalysis

been claimed to have advantages for the oxidation of bulky alkenes because of their pore size. However, none of them is intrinsically more active than TS-1 in the reactions of small substrates that have no obvious diffusion problem for the medium pores. Therefore, in parallel with developing large-pore titanosilicates, the search for more intrinsically active ones than TS-1 is also an important research subject. The catalytic performance of Ti-MWW is compared in the oxidation of 1-hexene with H2 O2 with that of TS-1 and Ti-beta in Table 23.2. Consistent with the results reported elsewhere [17–19], TS-1 showed higher conversion than Ti-beta with a similar Ti content. However, Ti-MWW exhibited activity about three times as high as TS-1 based on the speciﬁc conversion per Ti site. A more unique property exhibited only by Ti-MWW is that it shows shape selectivity in the epoxidation of geometric alkene isomers. Table 23.3 shows the results of various titanosilicates in the epoxidation of 2-hexenes with a cis/trans ratio of 41 : 59. The products were 2,3-epoxyhexanes with both cis and trans conﬁgurations, and diols formed by a successive hydrolysis of epoxides over acid sites. Although the Ti content varied greatly with the titanosilicates, obviously Ti-MWW exhibited the highest speciﬁc activity in the conversion of 2-hexenes with high H2 O2 efﬁciency. Interestingly, Ti-MWW showed totally different behavior in the epoxide distribution. That is, only Ti-MWW exhibited a peculiar selectivity as high as 81% for the trans-epoxide. In contrast, the other titanosilicates selectively promoted the epoxidation of the cis-isomer to give selectivity for the corresponding epoxide higher than the percentage in the starting substrates. TS-1 is generally a cis-selective catalyst for alkene stereoisomers, selectively producing the cis-epoxide in the epoxidation of cis/trans 2-butenes or cis/trans 2-hexenes with H2 O2 [70]. Thus, there was an essential difference in the geometric selectivity between Ti-MWW and conventional titanosilicates in the epoxidation of cis/trans alkenes. Table 23.3 Epoxidation of hex-2-ene isomers with hydrogen peroxide over various titanosilicatesa .

Catalyst

Ti-MWW TS-1 TS-2 Ti-beta Ti-MOR Ti-Y Ti-MCM-41 SiO2 –TiO2 a Catalyst,

Si/Ti

46 42 95 40 79 43 50 85

Conversion (mol%)

50.8 29.1 13.6 15.9 2.6 3.8 3.1 0.8

Product selectivity (mol%)

Epoxide selectivity (mol%)

H2 O2 (mol%)

Epoxides

Diols

cis

trans

Conv.

Eff.

99 96 96 91 99 40 36 37

1 4 4 9 1 60 64 63

19 66 67 73 52 55 62 61

81 34 33 27 48 45 38 39

55.1 32.5 18.0 35.8 3.9 8.4 21.0 7.6

92 89 77 45 66 46 15 10

0.05 g; hex-2-enes (cis/trans = 41 : 59), 10 mmol; H2 O2 , 10 mmol; MeCN, 10 ml; temperature, 333 K; time, 2 hours.

23.2 Titanium-Containing Zeolites

As described above, despite a relatively high content of boron (generally corresponding to a Si/B ratio of 30) contained in the framework, hydrothermally synthesized Ti-MWW proved to be an extremely active catalyst for alkene epoxidation. Then, B-free Ti-MWW is expected to exhibit excellent activity, much higher than that of B-containing Ti-MWW. Since the direct synthesis of Ti-MWW without using the crystallization-supporting agent of boric acid is still a matter of challenge, the postsynthesis is an alternative choice for the preparation of B-free catalysts. The treatment with TiCl4 vapor at elevated temperatures is a usual method well employed for modifying MOR and BEA zeolites of 12-MR channels [20, 59]. Actually, the preparation of Ti-MCM-22 by the reaction of dealuminated MCM-22 with TiCl4 vapor has been patented [71]. However, it is suspected that the TiCl4 treatment method is actually ineffective for the MWW zeolite, because TiCl4 , 6.7 × 6.7 A˚ in molecular size, is expected to suffer serious steric restriction when penetrating ˚ 4.0 × 5.4 A) ˚ of MWW and thus might give rise to the 10-ring pores (4.0 × 5.9 A, uneven Ti distribution. As a totally different postsynthesis method that is ﬁrmly based on the structural characteristics of MWW, reversible structural conversion between 3D MWW silicate and its corresponding 2D lamellar precursor MWW (P) has been developed to construct more active Ti species within the framework [72]. Figure 23.5 illustrates the strategy of this postsynthesis method, ‘‘reversible structural conversion.’’ First, highly siliceous MWW is prepared from hydrothermally synthesized MWW borosilicate by the combination of calcination and acid treatment. Second, the MWW silicalite is treated with an aqueous solution of HM or PI and a Ti source. A reversible structure conversion from MWW to the corresponding lamellar precursor occurred as a result of Si–O–Si bond hydrolysis catalyzed by OH− , which is supplied by basic amine molecules. This is accompanied by the intercalation of the amine molecules. Calcination and acid treatment of the as-synthesized borosilicate MWW not only removed the framework boron but also converted the lamellar precursor into a MWW silicate (Si/B > 500). When this deboronated MWW was treated Defect site Ti Ti

and

Ti

Ti

amine Ti

solution

Ti

Amine Ti

MWWsilicate

Ti

Ti

Ti Ti Ti Ti-containing MWW lamellar precursor

Figure 23.5 Reversible structural conversion from MWW to MWW(P) lamellar precursor as the method of postsynthesizing Ti-MWW.

727

23 Metals in Zeolites for Oxidation Catalysis

with Ti(OBu)4 (tetrabutyl orthotitanate, TBOT) in the presence of HM or PI, the incorporation of Ti was achieved and more interestingly, the lamellar structure was simultaneously restored, namely, Ti species entered the interlayer space freely through the pore entrance of expanded layers to ﬁll up the defect sites such as hydroxyl nests. Extraction of the extraframework Ti species by acid treatment (see the next section) followed by calcination caused the layers to dehydroxylate, resulting in B-free Ti-MWW. For the Si/Ti ratio range of 20–100 in the starting gel, nearly all the Ti was incorporated into the solid product, indicating that this postsynthesis method was highly effective in introducing Ti. It should be noted that this structural conversion occurred only in the presence of HM or PI, the two typical SDAs for the crystallization of MWW zeolites, but was never caused by pyridine or piperazine although these cyclic amines have similar molecular shapes. This means that there is a ‘‘molecular recognition’’ of amine molecules given by the layered MWW sheets and that HM or PI molecules stabilize the lamellar MWW (P) structure. The catalytic properties of postsynthesized PS-Ti-MWW were compared with directly hydrothermally synthesized HTS-Ti-MWW and TS-1 in the epoxidation of 2-hexenes with H2 O2 (Figure 23.6). For reasonable comparison, the reactions were carried out in the most suitable solvents for the two titanosilicates, in acetonitrile for Ti-MWW and in methanol for TS-1, respectively. HTS-Ti-MWW showed much higher intrinsic activity than TS-1 for 2-hexenes. PS-Ti-MWW further proved to be 90 80 70 1-Hexene conversion (mol%)

728

PS-Ti-M WW

60 50 40

HTS-TiMWW

30 20 TS-1

10 0

0

0.1

0.2

0.3

0.4

Ti content (mmol g −1) Figure 23.6 The catalytic properties of postsynthesized PS-Ti-MWW, directly hydrothermally synthesized HTS-Ti-MWW, and TS-1 in the epoxidation of 1-hexene with H2 O2 .

0.5

23.2 Titanium-Containing Zeolites

about two times as active as HTS-Ti-MWW. The efﬁciency of H2 O2 utilization was also very high on PS-Ti-MWW. Thus, in terms of the activity, epoxide selectivity and H2 O2 efﬁciency, PS-Ti-MWW has so far been the most efﬁcient heterogeneous catalyst for liquid-phase epoxidation of linear alkenes. At ﬁrst, the main difference between PS-Ti-MWW and HTS-Ti-MWW seemed to be the boron content. A further treatment with reﬂuxing 2 M HNO3 was able to deboronate HTS-Ti-MWW to produce a sample nearly free of boron (Si/B > 500), while removing the Ti species only slightly. However, this did not result in any substantial increase in turnover number (TON). Therefore, the low boron content cannot account for the high activity of PS-Ti-MWW. Since HTS-Ti-MWW is prepared by using boric acid as a structure-supporting agent, the coexisting boron would preferentially occupy the speciﬁc framework position, which would hinder the uniform incorporation of titanium. On the other hand, PS-Ti-MWW was prepared by inserting the Ti species mainly into defect sites formed by elimination of boron atoms. Thus, the Ti species occupying crystallographically different tetrahedral sites may account for the different catalytic behavior between PS-Ti-MWW and HTS-Ti-MWW. Ti-MWW catalysts have also been found to exhibit catalytic performance superior to conventional titanosilicates such as TS-1 and Ti-beta in other oxidation reactions, for example, epoxidation of allyl alcohol to glycidol [73], epoxidation of diallyl ether to allyl glycidyl ether [74], epoxidation of allyl chloride to epichlorohydrin [75], epoxidation of 2,5-dihydrofuran to 3,4-epoxytetrahydrofuran [76], hydroxylation of 1,4-dioxane to 1,4-dioxane-2-ol [77], and ammoximation of cyclohexanone to cyclohexanone oxime [78]. The catalytic performance of Ti-MWW in the liquid-phase ammoximation of cyclohexanone depends greatly on the operating conditions of the reaction, especially the method of adding H2 O2 . Only when H2 O2 is added slowly into the reaction system, high yield of cyclohexanone oxime is attained. The reaction consists of reaction of NH3 with H2 O2 to produce hydroxylamine and subsequent noncatalytic reaction of hydroxylamine with cyclohexanone to afford cyclohexanone oxime. In the presence of excess H2 O2 , the extensive oxidation of hydroxylamine occurs on the Ti species of Ti-MWW with extremely high oxidation ability. Thus, under optimized conditions Ti-MWW is capable of catalyzing the ammoximation of cyclohexanone to cyclohexanone oxime at a conversion and selectivity >99%, proving to be a catalyst superior to TS-1, the currently used industrial catalyst. Reversible structural conversion did not occur when as-synthesized Ti-MWW (P) with the Si/Ti ratio of >100 was calcined subsequent to washing with 2 M HNO3 . Thus, the obtained novel titanosilicate with the structure analogous to the MWW precursor, designated as Ti-YNU-1, shows much higher oxidation ability, epoxide selectivity, and stability than Ti-beta in the oxidation of bulky cycloalkenes [79, 80]. Ti-YNU-1 has proved to have a large interlayer pore space corresponding to 12-ring zeolites [81]. Whereas direct condensation of the layers results in the formation of the MWW structure having a 10-ring interlayer pore, apparently monomeric Si species have been inserted into the interlayer spaces followed by condensation to provide a 12-ring pore. Since no Si source has been added, it is assumed that silica ‘‘debris’’ formed by the decomposition of a part of the MWW layer acted as the Si source.

729

730

23 Metals in Zeolites for Oxidation Catalysis

This assumption has been tested by deliberately adding a Si source to the as-synthesized MWW(P); by using silylating agents such as SiMe2 (OR)2 and SiMe2 Cl2 , silylene units can be inserted between the layers, which was followed by removal of the organic moieties to give the material that is exactly similar to Ti-YNU-1. Furthermore, it has been revealed that this method of inserting monomeric Si sources into the interlayer spaces can be widely applied to the conversion of a variety of 2D lamellar precursors into novel 3D crystalline metallosilicates with expanded pore apertures between the layers [82, 83]. Figure 23.7 depicts the scheme of expansion of interlayer pores following this general procedure. On the basis of the knowledge well established about ITQ-2 [65], the delaminated titanosilicate, Del-Ti-MWW, which essentially consists of thin sheets, has been synthesized [84]. Subsequent treatment of the acid-treated Ti-MWW precursor in a basic solution of TPAOH and cetyltrimethylammonium bromide (CTMABr) cleaved the interlayer linkages and made the surfactant molecules enter easily, to be intercalated between the layers. This resulted in the formation of a swollen material with expanded interlayer space exhibiting a diffraction peak at lower 2◦ with a corresponding d spacing of 3.9 nm. After the swollen material was treated in an ultrasound bath and further calcined at 823 K, the expanded layered structure substantially collapsed, leading to a sample with extensively weakened diffractions due to the MWW structure. Nevertheless, this sample exhibited an enlarged surface area, over 1000 m2 g−1 , compared to Ti-MWW, and contained a large amount of silanol groups as revealed by a strong band at 3742 cm−1 in the IR spectra. 29 Si MAS NMR spectra also veriﬁed the increase in the concentration of silanol groups. The catalytic properties of Del-Ti-MWW have been compared with those of other titanosilicates in the epoxidation of cyclic alkenes (Table 23.4). The TON decreased Me2SiX2 (X = OR or halogen

SDA

Interlayer

Lamellar precursor

Silylation Calcination Calcination

Large or medium pore

Interlayer-expanded zeolites IEZ-XXX

Medium Figure 23.7 The strategy for converting zeolite lamellar precursors to new zeolite structures with the interlayer pore space expanded.

23.2 Titanium-Containing Zeolites Table 23.4

731

Epoxidation of cycloalkenes with H2 O2 over various titanosilicatesa .

Catalyst

Si/Ti

Surface area (m2 g–1 )

Alkene epoxidationb

Cyclopentene

Del-Ti-MWW 3D PS-Ti-MWW TS-1 Ti-beta Ti-MCM-41

42 46 34 35 46

1075 520 525 621 1144

Cyclooctene

Cyclododecenes

Conversion

TON

Conversion

TON

Conversion

TON

58.9 15.7 16.3 9.9 3.5

306 89 69 43 20

28.2 4.3 1.6 4.6 5.1

147 24 7 20 29

20.7 3.3 1.2 1.9 4.1

57 9 3 4 12

a Reaction

conditions: catalyst, 10–25 mg; alkene, 2.5–10 mmol; H2 O2 , equal to the alkene amount; CH3 CN, 5–10 ml; temperature, 313 K for cyclopentene and 333 K for other substrates; time, 2 hours. b Conversion in mol%; TON in mol (mol Ti)−1 .

sharply for TS-1, Ti-beta, and 3D Ti-MWW with increasing molecular size of cyclic alkenes. Ti-MCM-41 with mesopores, on the other hand, showed higher TON for cyclooctene and cyclododecene than these titanosilicates. This implies that the reaction space is extremely important for the reactions of bulky molecules. The delamination of Ti-MWW increased the TON greatly for not only cyclopentene but also bulkier cycloalkenes. Especially, the catalytic activity of Del-Ti-MWW was about six times as high as that of Ti-MWW for cyclooctene and cyclododecene. Del-Ti-MWW turned out to be even superior to Ti-MCM-41 in the epoxidation of bulky substrates. This should be the beneﬁt of high accessibility of Ti active sites in Del-Ti-MWW. Thus, the delamination was able to change Ti-MWW into an effective catalyst applicable to reactions of bulky substrates. 23.2.4 Other Titanium-Containing Zeolites

It is possible to directly incorporate Ti into the framework of other zeolites. The structure of UTD-1, which can be synthesized using bis(pentamethylcyclopentaienyl)cobalt (III) hydroxide, Cp*2CoOH, as a structure-directing ˚ pores agent, has one-dimensional extra-large elliptically shaped (7.5 × 10 A) circumscribed by 14-ring [85]. Despite the existence of the extra-large pores, the thermal stability of the framework is remarkably high. The addition of titanium to the UTD-1 synthesis gel results in the isostructural titanosilicate Ti-UTD-1 [86]. The Ti-UTD-1 molecular sieves are active in the oxidation of cyclohexane at room temperature with tert-butyl hydroperoxide (TBHP) as oxidation agent to give cyclohexanone as the major product with lesser amounts of cyclohexanol and adipic acid. Cyclohexene oxidation with H2 O2 results in allylic oxidation and epoxidation followed by hydrolysis [87]. Tuel has synthesized Ti-ZSM-12

732

23 Metals in Zeolites for Oxidation Catalysis

using a diquaternary ammonium structure-directing agent (Et2 MeN+ C3 H6 )2 [21]. Ti-ZSM-48 [16], TAPSO- 5[20], Ti-ITQ-7 [23, 24] were also synthesized and used as oxidation catalysts. Postsynthesis modiﬁcations have been successful in preparing titaniumcontaining molecular sieves active in oxidation. The method for the postsynthesis incorporation of Ti into the zeolite beta was also applied to the incorporation of Ti into MOR and FAU [62]. Yashima et al. proposed the ‘‘atom planting’’ method as the development of the alumination through the reaction of internal silanol groups with aluminum halides. By using halides of other metals, metallosilicates with the MFI structure (Ga, In, Sb, As, and Ti) and with the MOR structure (Ga, Sb, and Ti) have been prepared [88]. The incorporation of Ti into the MOR structure was conﬁrmed by the appearance of the speciﬁc absorption band in the IR spectra [20]. Very recently Kubota et al. synthesized Ti-YNU-2 (MSE) [89] by the postsynthesis modiﬁcation of YNU-2 (P) that has a large number of defect sites [90]. Ti-YNU-2 has proven to be a very active catalyst in the liquid-phase oxidation using H2 O2 as oxidant [89]. 23.2.5 Solvent Effects and Reaction Intermediate

The role of the solvent is complex: polarity, solubility of reactants and products, diffusion and counterdiffusion effects, and also interaction with the active centers [91]. Using a triphase system (solid–liquid–liquid) in the absence of any cosolvent, a considerable increase in the conversion of various water-immiscible organic compounds (toluene, anisole, benzyl alcohol, etc.) can be achieved [92]. Similarly, increased conversion in the absence of cosolvent has already been reported for the benzene oxidation [93, 94]. The solvent may be competing with reactant for diffusion in the channels and adsorption at the active sites of TS-1 catalyst. It has been shown that the activity of Ti-beta for 1-hexene and cyclohexanol oxidations is the highest in acetonitrile, a polar and nonprotic solvent [95]. This forms a contrast with the observed enhancement of the activity of TS-1 by methanol and protic solvents [70]. These differences have been ascribed to the hydrophilic character of Ti-beta compared to TS-1. The presence of Al and a large concentration of internal and external silanol groups (beta is composed of intergrowth of polymorphs A and B) give Ti-beta its hydrophilicity. Corma et al. pointed out that Ti-beta is more active than TS-1 for 1-hexene oxidation when the reaction is carried out in acetonitrile as solvent, suggesting that the active species in Ti-beta is a cyclic species, in which a water molecule coordinates the Ti atoms instead of an alcohol molecule [95]. Species I with a stable ﬁve-membered ring structure, formed by the coordination of ROH to Ti centers and hydrogen bonding to Ti–peroxo complex, was believed to be the active intermediate in protic alcohol solvents (Scheme 23.3), while species II was assumed to contribute to the oxidation of substrates in aprotic solvents. Recently, Lamberti et al. found that an end-on η2 (dihapto-coordinated) Ti–hydroperoxo complex was generated probably by the reversible rupture of Ti–O–Si bridge under anhydrous H2 O2 conditions, and this complex was reversibly

23.2 Titanium-Containing Zeolites R

H

O

O

H

Ti

H

Ti O

O

O

H

Species I

Scheme 23.3

O

H

Species II

Structural scheme for the proposed intermediate Ti species.

transformed to a side-on η2 Ti–peroxo complex after the addition of water, indicating that water molecules play an active role in determining the relative concentration of Ti–peroxo to -hydroperoxo species present on the working catalyst [39, 96, 97]. In spite of such ﬁndings and interpretations, aprotic acetone was reported to be the solvent of choice in terms of both activity and selectivity for the epoxidation of styrene and allyl alcohol on TS-1 [98, 99]. Comprehensive investigation of the solvent effect on the catalytic performance of three types of representative titanosilicates, viz. TS-1, Ti-MWW, and Ti-beta, revealed that solvent effect was highly dependent on substrates [100]. Figure 23.8 shows the catalytic results for the oxidation of 1-hexene over TS-1. In agreement with the results reported in literature [25c, 27, 95], methanol as solvent resulted in a substantial enhancement in catalytic activity, compared with acetonitrile. As expected, the difference in activity became small with decreasing Ti content since the number of active sites reduced. However, with respect to the oxidation of cyclohexene, a converse solvent effect was observed (Figure 23.9). When acetonitrile 30 25 Conversion (%)

MeOH 20 15 10

MeCN

5 0

0

0.004

0.008

0.012

0.016

0.02

Ti/(Si +Ti) molar ratio Figure 23.8 Dependence of 1-hexene conversion on the Ti content of TS-1 in the oxidation of 1-hexene in methanol and acetonitrile solvents. (Reaction conditions: 60 ◦ C, 2 hours, 0.05 g catalyst, 10 ml solvent, 10 mmol substrate, and 10 mmol H2 O2 (31% aqueous solution)).

733

23 Metals in Zeolites for Oxidation Catalysis

6 MeCN 5 Conversion (%)

734

4 3 2

MeOH

1 0

0

0.004

0.008

0.012

0.016

0.02

Ti /(Si+Ti) molar ratio Figure 23.9 Dependence of conversion on the Ti content of TS-1 in the oxidation of cyclohexene in methanol and acetonitrile solvents. (Reaction conditions: 60 ◦ C, 4 hours, 0.05 g catalyst, 10 ml solvent, 10 mmol substrate, and 10 mmol H2 O2 (31% aqueous solution)).

was used as a solvent, the conversion was nearly four times as high as that obtained with methanol as the solvent. Almost no dependence of the conversion on the Ti content was observed when the Ti/(Ti+Si) molar ratio was larger than 0.0086. This is probably due to the occurrence of the reaction mainly on the external surface and near the pore mouth; the increase in the Ti content incorporated into the internal surface does not signiﬁcantly contribute to the enhancement in the activity in the oxidation of cyclohexene. Compared with species II (Scheme 23.2), species I is likely to form on TS-1. This is believed to be due to its hydrophobic character [101], which would make methanol approach Ti sites more easily than water, leading to the formation of a large amount of active sites of species I [101]. Therefore, when 1-hexene is oxidized on TS-1, methanol should be the solvent of choice. However, because cyclohexene has a large molecule size, its oxidation would occur mainly on the exterior surface and/or near the pore mouth, where much more Si–OH and Ti–OH groups are present than inside the channels making these areas relatively hydrophilic. In order to conﬁrm this hypothesis, a TS-1 sample was poisoned by adding 2,4-dimethylquinoline to the reaction mixture. The conversion of cyclohexene drastically declined in acetonitrile solvent, while the conversion of 1-hexene was not signiﬁcant in methanol solvent. This shows that the oxidation of cyclohexene was indeed primarily catalyzed by Ti sites on the exterior surface and/or near the pore mouth. Species II should be formed more easily on the external surface and act as an active intermediate when acetonitrile is used as the solvent [95]. Additionally, species II is more intrinsically active than species I due to its higher electrophilic character [95]. This makes acetonitrile a better solvent in the oxidation of cyclohexene over TS-1, which mainly occurs on the external surface of TS-1, than in

23.2 Titanium-Containing Zeolites

the oxidation of 1-hexene. This is also supported by the fact that the silylation slightly increased the conversion of cyclohexene over TS-1 when methanol was used as the solvent. The selective silylation of the external surface of TS-1 by bulky 1,1,1,3,3,3-hexamethyldisilazane could increase the hydrophobicity, favoring the formation of species I. It was shown that the silylation of Ti-MCM-41 led to a remarkable increase in the Q4 /Q3 ratio, signiﬁcantly increasing its hydrophobicity and enhancing its activity for the oxidation of 1-hexene and hexane [43]. Nevertheless, after silylation, considerable amounts of OH groups were still present on the exterior surface of TS-1 as a result of the steric limitation originating from the large molecular size of 1,1,1,3,3,3-hexamethyldisilazane. This makes the external surface of TS-1 still relatively hydrophilic even after being silylated. Methanol as solvent would strongly compete with water to adsorb on the Ti sites, because it is preponderant in amount, resulting in a lower activity than acetonitrile as solvent. In contrast to TS-1, Ti-MWW exhibited much higher activity in acetonitrile solvent than in methanol for the oxidation of 1-hexene, whereas for the oxidation of cyclohexene, acetonitrile as solvent just gave slightly higher conversion than methanol. Ti-MWW was much more active in acetonitrile than in methanol in the epoxidation of 1-hexene (Figure 23.10), because of its hydrophilicity [80, 102]. The silylation had no marked effect on the catalytic results due to the reaction mainly occurring inside the channels with distorted 10-ring pore openings; the silylation of internal Ti species by bulky 1,1,1,3,3,3-hexamethyldisilazane would be hard to occur. It is believable that the epoxidation of cyclohexene over Ti-MWW, in a similar manner to its epoxidation over TS-1, mainly took place on its external surface since the pore openings of MWW-type materials are slightly smaller than those of the MFI-type [103], and the poisoning by 2,4-dimethylquinoline resulted in a drastic 80 MeCN

Conversion (%)

60

MeOH

40

20

0

0

0.005

0.01 0.015 0.02 Ti /(Si+Ti) molar ratio

0.025

Figure 23.10 Dependence of conversion on the Ti content of Ti-MWW in the oxidation of 1-hexene in methanol and acetonitrile solvents. (Reaction conditions: 60 ◦ C, 2 hours, 0.05 g catalyst, 10 ml solvent, 10 mmol substrate and 10 mmol H2 O2 (31% aqueous solution)).

735

736

23 Metals in Zeolites for Oxidation Catalysis

reduction in the activity for oxidation of cyclohexene regardless of the employed solvent but did not signiﬁcantly affect the activity in 1-hexene epoxidation. The presence of a lot of 12-ring side-pockets on the external surface of Ti-MWW led to a moderate cyclohexene conversion between those obtained with TS-1 and Ti-beta. It should be noted that the Ti species inside the side-pockets are located on the intracrystalline surface, and consequently have an environmental state similar to that of the Ti species in TS-1; these Ti species are relatively hydrophobic. As a result, methanol as solvent gave almost the same activity as acetonitrile probably because both the hydrophilic Ti species on the external surface and the hydrophobic ones in the side-pockets could serve as active sites for the epoxidation of cyclohexene. Slylation of Ti-MWW with 1,1,1,3,3,3-hexamethyldisilazane, which is too large to enter the 10-ring, results in the silylation of the external surface of silanol only. Therefore, the sites active for the cyclohexene epoxidation become totally hydrophobic, resulting in methanol being the solvent of choice. However, the silylation close to the entrance of side-pockets with such a large-molecule agent would result in a considerable steric constraint on the diffusion of substrate and product molecules into and out of these side-pockets, giving rise to a slight decrease in activity. When 1-hexene was oxidized over Ti-beta, the conversion in acetonitrile solvent was about two times as high as that attained in methanol. However, for cyclohexene oxidation, methanol as solvent was highly superior to acetonitrile, which conﬂicts the results reported by Corma et al. [95], but is consistent with our previous ﬁndings [57]. As has long been known, acetonitrile should be the solvent of choice for the oxidation of 1-hexene over Ti-beta and Ti-MWW while methanol is preferred by TS-1. In contrast, for cyclohexene oxidation, methanol is favorable for Ti-beta, whereas acetonitrile is the best for TS-1 and Ti-MWW. Thus, the effect of hydrophilicity /hydrophobicity is clearly demonstrated by a series of catalytic results, but this cannot interpret the solvent effect on the oxidation of cyclohexene over Ti-beta.

23.3 Other-Metal-Containing Zeolites

Vanadium is claimed to be incorporated into the zeolite MEL framework to give VS-2 [104]. A conspicuous characteristic of VS-2 is that it can catalyze the oxidation of terminal methyl groups of linear alkanes in the presence of H2 O2 as oxidant to produce 1-alkanol and aldehydes, whereas TS-1 and TS-2 catalyze only the oxidation of internal carbons (methylene groups) [29]. Comparative study on alkane oxidation on VS-2 and TS-2 has been conducted [105]. Spin-trapping experiments have revealed that the VS-2–H2 O2 –hexane system generates primary hexyl radical species, which could not be observed in the TS-2 system. It is proposed that the oxidation of internal carbons and that of terminal carbons proceed by different mechanisms. Gallot et al. have provided another explanation on the complete absence of oxidation of terminal carbon based on the agostic interaction of Ti4+ and C–H bond of a terminal methyl [106].

23.3 Other-Metal-Containing Zeolites

Detailed study on the form of the vanadium species in the V-MEL samples has been carried out [107]. As-synthesized V-MEL samples contain two different V species, a framework V5+ in a distorted tetrahedral environment and an extraframework V4+ in an octahedral environment. Upon calcination the V4+ species transforms into two types of V5+ species. However, these V5+ species are easily extracted by NH4 OAc, and only the nonextractable vanadium species are active in the oxidation of toluene and phenol. V-beta can be prepared in a similar two-step manner. The method consists of ﬁrst creating a vacant site by dealumination of the beta zeolite with nitric acid and then contacting it with an NH4 VO3 solution [108]. Reddy et al. have prepared V-NCL-1 using a diquaternary ammonium ion (Et3 N+ C3 H6 )2 [109]. It is to be noted, however, that in terms of H2 O2 utilization efﬁciency and resistance against leaching, vanadosilicates synthesized so far are inferior to titanosilicates. Tin-containing silicalite-2, Sn-Sil-2, has been synthesized [110]. From 119 Sn NMR study, however, it has been suggested that Sn4+ ions are mostly in octahedral coordination. It has been discovered that Sn-beta catalyzes the Baeyer–Villiger oxidation of cyclic ketones to lactones without using peracids but using H2 O2 [111]. Notably, this Sn-beta catalyst selectively promotes the Baeyer–Villiger oxidation when the substrate contains a carbon–carbon double bond besides the carbonyl bond. Two possible mechanisms have been considered, namely, (i) one pathway involving the activation of the carbonyl group by Sn and (ii) the other one in which H2 O2 is activated to form a tin–hydroperoxo intermediate. From the computational and kinetic studies, it has been concluded that pathway 1 is followed and that the catalytically active site consists of two centers: the Lewis acid Sn atom to which ketone coordinates and the oxygen atom of the Sn–OH group that interacts with H2 O2 through a hydrogen bond [112]. The Sn-beta catalyst can also be employed for the Baeyer–Villiger oxidation of aromatic aldehydes [113]; aldehydes containing 4- or 2-alkoxy substituents are oxidized to the corresponding formate esters, which can be hydrolyzed to the corresponding phenols. Chromium-containing silicalite-2, CrS-2, has been synthesized and shown to catalyze similar reactions using tert-butyl hydroperoxide (TBHP) as an oxidant [114]. It is to be noted that TBHP has been reported to be an ineffective oxidant in TS-1-catalyzed oxidations [26]. However, it has been claimed that TS-1–TBHP combination exhibits activity in the oxidative cleavage of the C=C double bond of silyl enol ethers to produce dicarboxylic acids [115]. Lempers and Sheldon have reported that small amounts of chromium that are leached from CrAPO-5, CrAPO-11, and CrS-1 catalyze the liquid phase oxidation of bulky alkenes with TBHP [116]. The leaching seems to be caused by TBHP that extracts chromium from the micropores. They emphasize that experiments demonstrating that heterogeneous catalysts can be recovered and recycled without apparent loss of activity are not a deﬁnite proof of heterogeneity. Fe ion is easily incorporated into zeolites. Ferrosilicates and ferrisilcates are often used as acid catalysts. Direct oxidation of benzene to phenol over Fe-MFI zeolites has been shown to be practical when N2 O is used as an oxidant (AlphOx

737

738

23 Metals in Zeolites for Oxidation Catalysis

process) [117]. This process is suitable for the abatement of highly concentrated N2 O (20–40%) gas discharged from adipic acid plants, which will in turn help in the tighter regulation of emission levels of N2 O. However, the structure of the active Fe site is still in dispute, and it is unclear whether the Fe species in the zeolite framework is actively involved in this reaction.

23.4 Conclusion

The isomorphous substitution of metal heteroatoms with redox properties for the framework silicon atoms leads to the catalytic activity in the oxidation. Signiﬁcant developments have been made in the ﬁeld of new metal-substituted zeolites, which are applicable to industrially important catalytic processes. They include the synthesis and characterization of new metal-substituted zeolites and new methods of synthesizing metal-substituted zeolites useful for oxidation reactions. In particular, a great development has been made in the ﬁeld of titanosilicate zeolite catalysts for liquid-phase oxidation of various organic compounds using H2 O2 as the oxidant. So far, titanosilicate zeolites have been superior in catalytic performance in oxidation reactions to any other metallosilicate. In addition to high activity and selectivity in oxidation, H2 O2 utilization efﬁciency and resistance against leaching would be critically important in putting the metallosilicate zeolites to practical use. Ammoximation of cyclohexanone to produce cyclohexanone oxime by using TS-1 is already in use in industries, and propylene oxide has been industrially manufactured from propylene and H2 O2 in the presence of TS-1 since 2008. Ti-beta and Ti-MWW (MCM-22) as well as TS-1 proved to be promising catalysts for a variety of oxidation reactions. Here, future prospects in the research of metallosilicate zeolites have been pointed out. Although H2 O2 is an excellent oxidant that has high active oxygen content and produces water as the by-product, molecular oxygen is ideal for oxidation reactions. Therefore, strenuous efforts to utilize molecular O2 as oxidant are needed. One approach would be in situ preparation of H2 O2 from O2 and H2 , whereas the catalytic synthesis of H2 O2 from O2 and H2 itself is extremely difﬁcult. Direct synthesis of metal-substituted zeolites is an ideal method. However, since the postsynthesis modiﬁcations can be made under the wide-ranging conditions (temperature, solvent, atmosphere, pH, etc.) far from those for the zeolite synthesis, the modiﬁcations of zeolites present us powerful indirect methods for manipulating the properties of zeolites. Therefore, it is believed that the ﬁne-tuning of the properties of zeolites will continue to be achieved by developing a variety of postsynthesis modiﬁcation procedures as well as direct synthetic techniques. Solvents are often critically important in liquid-phase oxidation reactions from the practical viewpoint, because separation of the products from the solvent could be energy consuming if there is only a small difference in the boiling point or solubility between the products and the solvent. Solvent might affect the life of the catalyst; both the leaching of the active component and the deactivation due to

References

carbonaceous deposits could be avoided by choosing suitable solvents. Therefore, more attention must be paid to the choice of the solvent and modiﬁcation of metallosilicate zeolites to match the solvent with the substrate and zeolite. The feasibility of synthesis of metallosilicate zeolites with pores comprising 12and larger ring will continue to be pursued to overcome the problems of diffusion limitation/inaccessibility to the intracrystalline active sites. In this context, the research on ‘‘mesoporous zeolites’’ containing active metal elements is clearly expanding. However, there could be some differences in the microenvironmental structure of incorporated metal and its catalytic property among the well-crystallized zeolites and ‘‘mesoporous zeolites.’’ Another approach is to attempt the synthesis of metallosilicate zeolites of a nanometer order size to address the diffusion limitation problem. ‘‘All surface’’ zeolite sheets of an ITQ-2 type [65] are promising materials, although the synthesis on a massive scale might be difﬁcult. Activation of alkane, CH4 in particular, is an important subject. Application of shape selectivity to industrially important redox catalysis has not yet been developed. It should be re-emphasized that the catalytic performance is greatly inﬂuenced by hydrophilicity/hydrophobicity and crystal size as well as the number of active species particularly in the liquid-phase oxidation. The stability problems due to structure collapse, leaching, and blocking by products of high molecular weight/carbonaceous deposits must also be tackled. The development of computer modeling and quantum chemical calculations is rapidly growing, which will increase their contribution to design synthesis, structural analysis, and design of catalysis of metallosilicate zeolites and provide insight into reaction mechanism over metallozeolites. References 1. K¨ uhl, G.H. (1999) in Catalysis and Zeo-

2. 3. 4.

5.

6.

lites, chapter 3 (eds J.Weitkamp and L. Puppe), Springer, Berlin, p. 81. Kubo, T., Arai, H., and Tominaga, H. (1971) Bull. Chem. Soc. Jpn., 44, 2297. Lunsford, J. (1977) ACS Symp. Ser., 40, 473. Weitkamp, J. (1993) in Proceedings from the Ninth International Zeolite Conference, vol. 1 (eds R. von Ballmoos, J.B. Higgins, and M.M.J. Treacy), Buttherworth-Heinemann, Boston, p. 13. Balkus, K.J. Jr. and Gabrielov, A.G. (1995) J. Inclusion Phenom. Mol. Recognit. Chem., 21, 159. Schulz-Ekloff, G. and Ermst, S. (1999) in Preparation of Solid Catalysts (eds G.Ertl, K. Kn¨ozinger, and J. Weitkamp), Wiley-VCH Verlag GmbH, Weinheim, p. 405.

7. Herron, N., Stucky, G.D., and Tolman,

8. 9.

10.

11. 12.

13. 14.

C.A. (1985) Inorg. Chim. Acta, 100, 135. McMorn, P. and Hutchings, G.J. (2004) Chem. Soc. Rev., 33, 108. Dioos, B.M.L., Sels, B.F., and Jacobs, P.A. (2007) Stud. Surf. Sci. Catal., 168, 915. Breck, D.W. (1974) in Zeolite Molecular Sieves, chapter 4 John Wiley & Sons, Inc., New York, p. 320. Taramasso, M., Perego, G., and Notari, B. (1983) US Patent 4,401,051. Ione, K.G., Vostrikova, L.A., and Mastikhin, V.M. (1985) J. Mol. Catal., 31, 355. Tielen, M., Geelen, M., and Jacobs, P.A. (1985) Acta Phys. Chem., 31, 1. Ylimaz, B. and Mueller, U. (2009) Top. Catal. 52, 888, doi: 10.1007/s11244-009-9226-0.

739

740

23 Metals in Zeolites for Oxidation Catalysis 15. Reddy, J.S., Kumar, R., and Ratnasamy, 16.

17.

18.

19. 20. 21. 22.

23.

24.

25.

26.

P. (1990) Appl. Catal., 58, L1. Serrano, D.P., Li, H.-X., and Davis, M.E. (1992) J. Chem. Soc., Chem. Commun., 745. (a) Camblor, M.A., Corma, A., Mart´ınez, A., and P´erez-Pariente, J. (1992) Chem. Commun., 589; (b) Camblor, M.A., Constantini, M., Corma, A., Gilbert, L., Esteve, P., Mart´ınez, A., and Valencia, S. (1996) Chem. Commun., 1339. (a) Blasco, T., Camblor, M.A., Corma, A., Esteve, P., Mart´ınez, A., Prieto, C., and Valencia, S. (1996) Chem. Commun., 2367; (b) Blasco, T., Camblor, M.A., Corma, A., Esteve, P., Guil, J.M., Mart´ınez, A., Perdigon-Melon, J.A., and Valencia, S. (1998) J. Phys. Chem. B, 102, 75. Jappar, N., Xia, Q., and Tatsumi, T. (1998) J. Catal., 180, 132–141. Tuel, A. (1995) Zeolites, 15, 228. Tuel, A. (1995) Zeolites, 15, 236. (a) Wu, P., Komatsu, T., and Yashima, T. (1996) J. Phys. Chem., 100, 10316; (b) Wu, P., Komatsu, T., and Yashima, T. (1997) J. Catal., 168, 400; (c) Wu, P., Komatsu, T., and Yashima, T. (1997) Stud. Surf. Sci. Catal., 105, 663; (d) Wu, P., Komatsu, T., and Yashima, T. (1998) J. Phys. Chem. B, 102, 9297. D´ıaz-Caba˜ nas, M.J., Villaescusa, L.A., and Camblor, M.A. (2000) Chem. Commun., 761. ˜ az-Caba˜ Corma, A., D´ın nas, M.J., Domine, M.E., and Rey, F.Z. (2000) Chem. Commun., 1725. (a) Wu, P., Tatsumi, T., Komatsu, T., and Yashima, T. (2000) Chem. Lett., 774; (b) Wu, P., Tatsumi, T., Komatsu, T., and Yashima, T. (2001) J. Phys. Chem. B, 105, 289; (c) Wu, P., Tatsumi, T., Komatsu, T., and Yashima, T. (2001) J. Catal., 202, 245; (d) Wu, P. and Tatsumi, T. (2001) Chem. Commun., 897; (e) Wu, P. and Tatsumi, T. (2002) J. Phys. Chem. B, 106, 748. Notari, B. (1996) in Advances in Catalysis, vol. 41 (eds D.D.Eley, W.O. Haag, and B.C. Gates), Academic Press, San Diego, p. 253.

27. Clerici, M.G., Bellussi, G., and

28.

29.

30. 31.

32.

33. 34.

35.

36. 37.

38.

39.

40. 41.

42. 43.

44.

Romano, U. (1991) J. Catal., 129, 159. Millini, R., Previde-Massara, E., Perego, G., and Bellussi, G. (1992) J. Catal., 137, 497. Tatsumi, T., Nakamura, M., Negishi, S., and Tominaga, H. (1990) J. Chem. Soc., Chem. Commun., 476. Bellussi, G. and Rigutto, M.S. (1994) Stud. Surf. Sci. Catal., 105, 177. Ratnasamy, P. (2004) in Advances in Catalysis, vol. 48 (eds B.C.Gates and H. Knoezinger), Elsevier, Amsterdam, pp. 1–179. Perego, G., Bellussi, G., Corno, C., Taramasso, M., Buomono, F., and Esposito, A. (1986) Stud. Surf. Sci. Catal., 28, 129. Khouw, C.B. and Davis, M.E. (1995) J. Catal., 151, 77. Lamberti, C., Bordiga, S., Zecchina, A., Artioli, G., Marra, G., and Spano, G. (2001) J. Am. Chem. Soc., 123, 2204. Bordiga, S., Damin, A., Bonino, F., Ricchiardi, G., Zecchina, A., Tagliapietra, R., and Lamberti, C. (2003) Phys. Chem. Chem. Phys., 5, 4390. Thangaraj, A. and Sivasanker, S. (1992) J. Chem. Soc., Chem. Commun., 123. Schuchardt, U., Pastore, H.O., and Spinace, E.V. (1994) Stud. Surf. Sci. Catal., 84, 1877. Serrano, D.P., Uguina, M.A., Ovejero, G., van Grieken, R., and Camacho, M. (1996) Microporous Mater., 7, 309. Bordiga, S., Bonino, F., Damin, A., and Lamberti, C. (2007) Phys. Chem. Chem. Phys., 9, 4854. Thomas, J.M. and Sankar, G. (2001) Acc. Chem. Res., 34, 571. Fan, W., Duan, R.G., Yokoi, T., Wu, P., Kubota, Y., and Tatsumi, T. (2008) J. Am. Chem. Soc., 130, 10150. Parker, W.O. and Millini, R. (2006) J. Am. Chem. Soc., 128, 1450. Tatsumi, T., Koyano, K.A., and Igarashi, N. (1998) Chem. Commun., 325. Srinivas, D., Manikandan, P., Laha, S.C., Kumar, R., and Ratnasamy, P. (2003) J. Catal., 217, 160.

References 45. Zhuang, J., Ma, D., Yan, Z., Deng, F.,

46.

47.

48. 49.

50.

51.

52.

53.

54.

55.

56. 57.

58.

59. 60. 61. 62.

63.

Liu, X., Han, X., Bao, X., Guo, X., and Wang, X. (2004) J. Catal., 221, 670. Wells, D.H. Jr., Delgass, W.N., and Thomson, K.T. (2004) J. Am. Chem. Soc., 126, 2956. Fan, W., Li, R., Ma, J., Fan, B., Dou, T., and Cao, J. (1997) Microporous Mater., 8, 131. Gabelica, Z., Blom, N., and Derouane, E.G. (1983) Appl. Catal., 5, 227. Tatsumi, T., Nakamura, M., Yuasa, K., and Tominaga, H. (1990) Chem. Lett., 297. Hayashi, H., Kizawa, K., Murei, U., Shigemoto, N., and Sugiyama, S. (1996) Catal. Lett., 36, 99. Tatsumi, T., Yako, M., Nakamura, M., Yuhara, H., and Tominaga, H. (1993) J. Mol. Catal., 78, L41. Camblor, M.A., Corma, A., and Perez-Pariente, J. (1993) Zeolites, 13, 82. Corma, A., Esteve, P., Martinez, A., and Valencia, S. (1995) J. Catal., 152, 18. van del Waal, J.C., Lin, P., Rigutto, M.S., and van Bekkum, H. (1997) Stud. Surf. Sci. Catal., 105, 1093. Camblor, M.A., Corma, A., and Valencia, S. (1998) J. Phys. Chem. B, 102, 75. Tatsumi, T. and Jappar, N. (1998) J. Phys. Chem. B, 102, 7126. (a) Goa, Y., Wu, P., and Tatsumi, T. (2001) Chem. Commun., 1714; (b) Goa, Y., Wu, P., and Tatsumi, T. (2004) J. Phys. Chem. B, 108, 8401. Koller, H., Woelker, A., Villaescusa, L.A., Diaz-Cabanas, M.J., Valencia, S., and Camblor, M.A. (1999) J. Am. Chem. Soc., 121, 3368. Goa, Y., Wu, P., and Tatsumi, T. (2004) J. Phys. Chem. B, 108, 4242. Skeels, G.W. and Flanigen, E.M. (1989) ACS Symp. Ser., 398, 420. Reddy, J.S. and Sayari, A. (1995) Stud. Surf. Sci. Catal., 94, 309. Di Renzo, F., Gomez, S., Teissier, R., and Fajula, F. (2000) Stud. Surf. Sci. Catal., 130, 1631. Rigutto, M.S., de Ruiter, R., Niederer, J.P.M., and van Bekkum, H. (1994) Stud. Surf. Sci. Catal., 84, 2245.

64. Tatsumi, T., Nakamura, M., Yuasa, K.,

65.

66.

67.

68.

69. 70. 71.

72. 73. 74. 75.

76.

77. 78.

79.

80. 81.

82.

and Tominaga, H. (1991) Catal. Lett., 10, 259. Camblor, M.A., Corma, A., D´ıaz-Cabanas, M.J., and Baerlocher, C. (1998) J. Phys. Chem. B, 102, 44. Roth, W.J., Kresge, C.T., Vartuli, J.C., Leonowicz, M.E., Fung, A.S., and McCullen, S.B. (1995) Stud. Surf. Sci. Catal., 94, 301. Corma, A., Forn´es, V., Pergher, S.B., Maesen, Th.L.M., and Buglass, G. (1998) Nature, 396, 353. Blasco, T., Navarro, M.T., Corma, A., and P´erez-Pariente, J. (1995) J. Catal., 156, 65. Koyano, A.K. and Tatsumi, T. (1997) Chem. Commun., 145. Clerici, M.G. and Ingallina, P. (1993) J. Catal., 140, 71. Levin, D., Chang, A.D., Luo, S., Santiestebana, G., and Vartuli, J.C. (2000) US Patent 6,114,551. Wu, P. and Tatsumi, T. (2004) Catal. Surv. Asia, 8, 137. Wu, P. and Tatsumi, T. (2003) J. Catal., 214, 317. Wu, P., Liu, Y., He, M., and Tatsumi, T. (2004) J. Catal., 228, 183. Wang, L., Liu, Y., Xie, W., Zhang, H., Wu, H., Jiang, Y., He, M., and Wu, P. (2007) J. Catal., 246, 205. Wu, H., Wang, L., Zhang, H., Liu, Y., Wu, P., and He, M. (2006) Green Chem., 8, 78. Fan, W., Kubota, Y., and Tatsumi, T. (2008) ChemSusChem, 1, 175. (a) Song, F., Liu, Y., Wu, H., He, M., Wu, P., and Tatsumi, T. (2005) Chem. Lett., 34, 1436; (b) Song, F., Liu, Y., Wu, H., He, M., Wu, P., and Tatsumi, T. (2006) J. Catal., 237, 359. Fan, W., Wu, P., Namba, S., and Tatsumi, T. (2003) Angew. Chem. Int. Ed., 43, 236. Fan, W., Wu, P., Namba, S., and Tatsumi, T. (2006) J. Catal., 243, 183. Ruan, J., Wu, P., Slater, B., and Terasaki, O. (2005) Angew. Chem. Int. Ed., 44, 6719. Wu, P., Ruan, J., Wang, L., Wu, L., Wang, Y., Liu, Y., Fan, W., He, M., Terasaki, O., and Tatsumi, T. (2008) J. Am. Chem. Soc., 130, 8178.

741

742

23 Metals in Zeolites for Oxidation Catalysis 83. Inagaki, S., Yokoi, T., Kubota, Y., and

84.

85.

86.

87.

88.

89.

90.

91. 92. 93.

94.

95. 96.

97.

Tatsumi, T. (2007) Chem. Commun., 5188. Wu, P., Nuntasri, D., Ruan, J., Liu, Y., He, M., Fan, W., Terasaki, O., and Tatsumi, T. (2004) J. Phys. Chem. B, 108, 19126. Freyhardt, C.C., Tsapatsis, M., Lobo, R.F., Balkus, K.J., and Davis, M.E. (1996) Nature, 381, 295. Balkus, K.J., Gavrielov, A.G., and Zones, S.I. (1995) in Zeolites: A Reﬁned Toolfor Designing Catalytic Sites (eds L.Bonneviot and S. Kalliaguine), Elsevier, Amsterdam, pp. 519–525. Balkus, K.J., Khanmamedova, A., Gavrielov, A.G., and Zones, S.I. (1996) in 11th International Congress on Catalysis (eds J.W.Hightower, W.N. Delgass, and E. Iglesia), Elsevier, Amsterdam, pp. 1341–1348. Yashima, T., Yamagishi, K., and Namba, S. (1991) Stud. Surf. Sci. Catal., 60, 171. Kubota, Y., Koyama, Y., Yamada, T., Inaki, S., and Tatsumi, T. (2008) Chem. Commun., 6224. Koyama, Y., Ikeda, T., Tatsumi, T., and Kubota, Y. (2008) Angew. Chem. Int. Ed., 47, 1042. Clerici, M.G. (2001) Top. Catal., 15, 257. Bhaumik, A. and Kumar, R. (1995) J. Chem. Soc., Chem. Commun., 349. Tatsumi, T., Asano, K., and Yanagisawa, K. (1994) Stud. Surf. Sci. Catal., 84, 1861. Tatsumi, T., Yanagisawa, K., Asano, K., Nakamura, M., and Tominaga, H. (1994) Stud. Surf. Sci. Catal., 83, 417. Corma, A., Esteve, P., and Martinez, A. (1996) J. Catal., 161, 11. Bonino, F., Damin, A., Ricchiardi, G., ` G., D’Aloisio, R., Ricci, M., Spano, Zecchina, A., Lamberti, C., Prestipino, C., and Bordiga, S. (2004) J. Phys. Chem. B, 108, 3575. Prestipino, C., Bonino, F., Usseglio, S., Damin, A., Tasso, A., Clerici, M.G., Bordiga, S., D’Acapito, F., Zecchina, A., and Lamberti, C. (2004) ChemPhysChem, 5, 1799.

98. Bhaumik, A., Kumar, R., and

99.

100. 101.

102. 103.

104.

105.

106.

107.

108.

109.

110.

111.

112.

113.

114.

Ratnasamy, P. (1994) Stud. Surf. Sci. Catal., 84, 1883. Kumar, S.B., Mirajkar, S.P., Pais, G.C.G., Kumar, P., and Kumar, R. (1995) J. Catal., 156, 163–166. Fan, W., Peng, W., and Tatsumi, T. (2008) J. Catal., 256, 62. Bellussi, G., Carati, A., Clerici, M.G., Maddinelli, G., and Millini, R. (1992) J. Catal., 133, 220. Wu, P. and Tatsumi, T. (2002) Chem. Commun., 1026. Baerlocher, Ch., McCusker, L.B., and Olson, D.H. (2007) Atlas of Zeolite Framework Types, 6th Revised edn, Elsevier, p. 213–235. Hari Prasad Rao, P.R., Kumar, R., Ramaswany, A.V., and Ratnasamy, P. (1992) J. Catal., 137, 225. Tatsumi, T., Hirasawa, Y., and Tsuchiya, J. (1996) ACS Symp. Ser., 638, 374–383. Gallot, J.E., Fu, H., Kapoor, M.P., and Kaliaguine, S. (1996) J. Catal., 161, 798–809. Sen, T., Ramaswamy, V., Ganapathy, S., Rajamohanan, P.R., and Sivasanker, S. (1996) J. Phys. Chem., 100, 3809–3817. Dzwigaj, S., Peltre, M.J., Massiani, P., Davidson, A., Che, M., Sen, T., and Sivasanker, S. (1998) Chem. Commun., 87. Reddy, K.R., Ramaswany, A.V., and Ratnasamy, P. (1992) J. Chem. Soc., Chem. Commun., 1613. Mal, N.K., Ramaswany, V., Ganapathy, S., and Ramaswamy, A.V. (1995) Appl. Catal., A, 125, 233. Corma, A., Nemeth, L.T., Renz, M., and Valencia, S. (2001) Nature, 412, 423–425. Boronat, M., Corma, A., Rez, M., Gastre, S., and Valencia, S. (2005) Chem. Eur. J., 11, 6905. Bare, S.R., Kelly, S.D., Sinker, W., Low, J.J., Modica, F.S., Valencia, S., Corma, A., and Nemeth, L.T. (2005) J. Am. Chem. Soc., 127, 12924. Jayachandran, B., Sasidhran, M., Sudalai, A., and Ravindranathan, T. (1995) J. Chem. Soc., Chem. Commun., 1341.

References 115. Raju, S.V.N., Upadhya, T.T.,

Ponrathanam, S., Daniel, T., and Sudalei, A. (1996) J. Chem. Soc., Chem. Commun., 1969. 116. Lempers, H.E.B. and Sheldon, R.A. (1997) Stud. Surf. Sci. Catal., 105, 1061.

117. Kharitonov, A.S., Sheveleva, G.A.,

Panov, G.I., Sobolev, V.I., Paukshtis, Y.A., and Romannikov, V.N. (1993) Appl. Catal. A, 98, 33.

743

745

24 Environmental Catalysis over Zeolites Gabriele Centi and Siglinda Perathoner

24.1 Introduction

The ﬁrst widespread use of the term environmental catalysis to indicate catalysts and catalytic technologies for environment protection began around the beginning of 1990s, when the ﬁrst large meetings on this speciﬁc topic were organized [1–6]. The original concept of environmental catalysis referred only to technologies for reducing polluting emissions. However, the applications addressed by environmental catalysis include a broad range of questions currently: • catalytic clean-up technologies for gas (elimination of nitrogen oxides, conversion of volatile organic compounds – VOCs, etc.) and liquid phase (elimination of nitrate, toxic, and biorecalcitrant chemicals, etc.) emissions from stationary sources; • catalysis for sustainable mobility (cleaner fuels, reduction of NOx , CO, HC, and particulate in vehicle emissions, etc.); • catalytic approaches for converting or recycling solid waste (conversion of polymer waste, recycling industrial solid waste, etc.); • catalytic technologies for greenhouse gas reduction (reduction of N2 O, CH4 , CO2 , and ﬂuorocarbon emissions); • catalysis for in-house applications (improvement of air and water quality, self-cleaning surfaces and devices, etc.); • catalytic approaches to clean energy production (H2 production, use of renewables and biomass, fuel cells, energy storage, etc.); and • new catalytic processes for sustainable production and eco-compatible technologies. We restrict the discussion here, however, only to the original concept, because most of the other aspects are analyzed in other sections of this book. From the beginning, a signiﬁcant part of the communications presented at the cited conferences on environmental catalysis regarded the use of zeolite catalysts or

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

746

24 Environmental Catalysis over Zeolites

related materials. The largest part of these communications concerned the use of zeolites for NOx and VOC removal.

24.2 A Glimpse into Opportunities and Issues

One of the ﬁrst discoveries that has largely stimulated research in this ﬁeld was the ﬁnding that Cu ions exchanged into the faujasite (FAU) and MFI microporous matrix exhibit higher activity among metal ions exchanged into zeolites in the decomposition of NO [7–9], in particular, the ‘‘overexchanged’’ Cu/MFI (Cu2+ /Al > 0.5). In the presence of O2 and water vapor, the activity was signiﬁcantly depressed, and thus these materials never reached practical applications. However, there is still active research in this ﬁeld. Kustova et al. [10, 11] reported recently that mesoporous Cu/ZSM-11 (MEL structure characterized by a two-dimensional 10-ring pore) and Cu/ZSM-12 (MTW structure, one-dimensional 12-ring pore), in addition to the original and most studied Cu/ZSM-5 (MFI structure; similar to MEL, but one set of pores is zig-zag, or sinusoidally shaped), are active catalysts for the direct decomposition of NO. They also reported that Cu/ZSM-5 has been recognized as a unique catalyst for direct NO decomposition for many years, but their ‘‘discover that both Cu–ZSM-11 and Cu–ZSM-12 are about twice as active as Cu–ZSM-5 indicate that the special pore structure of Cu–ZSM-5 is not a decisive factor for catalytic activity in NO decomposition.’’ A primary issue in Cu/zeolites, but of general relevance for all the ﬁeld of transition-metal-ion-containing zeolites for environmental applications, concerns therefore the role of zeolite as a host material. The question is whether the zeolite provides only a good dispersion of the metal ions, the formation of speciﬁc species not being present on other oxides, or there is a speciﬁc effect of the zeolite pore structure (‘‘shape-selectivity’’ or analogous aspects). To discuss this issue in relation to the recent results of Christensen et al. [10, 11], it is useful to start with the observation that patents issued already 10 years earlier claimed the use of these zeolites for preparing catalysts active in the decomposition of NO. For example, Price and Kanazirev [12] claimed that ‘‘the zeolite should be preferably a ZSM-5, ZSM-11, or ZSM-12 zeolite.’’ Kagawa and Teraoka [13] also claimed ZSM-11 and ZSM-12 among the active zeolites for the decomposition of NO. Many other patents indicated these and other zeolitic structures among those relevant for preparing active catalysts. However, these materials were tested but showed lower activity with respect to ZSM-5, which is in contrast with the recent ﬁnding of Christensen et al. [10, 11]. To clarify this point, it is necessary to recall that the basic idea of using transition-metal-containing zeolites was the possibility to realize extremely well-dispersed single active species within the zeolite cavities [14]. However, it was soon realized that multiple species usually form and that the nature of the species present depends greatly on the preparation. The distribution of these

24.2 A Glimpse into Opportunities and Issues

species depends on many parameters beyond the zeolite structure itself, such as the modality of preparation (including precursor compounds), zeolite Si/Al ratio, thermal treatment, and so on. Different reactivity orders in a zeolite structure series have been often observed when the preparation method was changed. Even though a unique behavior of Cu/MFI (ZSM-5) catalyst in this reaction has been indicated [15], it was known that this peculiar characteristics referred to a nature of the active sites (transition-metal ions) different to that present in oxide-supported catalysts, and not to the special properties associated with the zeolite pore structure (monoor tridimensional, linear or sinusoidal channels, size of the channels, etc.) [16]. In fact, nitrogen oxide (NO) is a small molecule and the dimensions of possible reaction intermediates (dimeric species, for example) are small compared to the channel and cages sizes. Therefore, shape-selectivity effects cannot be present. The possible inﬂuence of the electrostatic ﬁeld within the zeolite (in a zeolite structural series such as ZSM-5, ZSM-11, and ZSM-12) on the stabilization of the reaction intermediates, or the inﬂuence of zeolite pore structure on the diffusivity of molecules (NO, N2 ), is also minimal. Therefore, the characteristics of zeolite pore structure (for equivalent nature of the active sites) are not expected to play a major role in the decomposition of NO, although they could be relevant in other reactions of nitrogen oxide conversion, such as for the selective reduction of NO with hydrocarbons. On the other hand, it is known for a decade now that different sites for isolated species of transition-metal ions are possible in the zeolite structure [17–19]. The charge balance at the cationic sites in metal-ion-exchanged zeolites can occur (Figure 24.1a) as given below: 1) Bare cations coordinated exclusively to the framework oxygen atoms and thus exhibiting open coordination sphere. 2) Metal-oxo species coordinated to the framework, but simultaneously bearing extra-framework oxygen atom(s). 3) Metal-oxide-like species supported in the zeolite inner volume or mostly at the outer surface of the crystals. The structure of the metallo-center depends on the type of the metal ion, the procedure of cation introduction into the zeolite, and the host zeolite matrix (Si/Al ratio and location of Al ions). Transition-metal ions such as Cu2+ or Co2+ are present as divalent cations predominantly coordinated only to framework oxygens. For higher charged ions such as Fe species, the lack of sufﬁcient local negative framework charge, particularly in high-silica zeolites, to balance trivalent cations may instead induce the formation of Fe ions bearing an extra-framework oxygen ligand, that is, FeO+ or dinuclear Fe–O–Fe type complexes. Different metal ion sites in zeolites have been identiﬁed for the bare cations (Figure 24.1b) [17–19]. They were indicated as α, β, and γ positions. In the ﬁrst two sites, the bivalent metal ions are coordinated to the framework by six-membered rings (α- and β-types), while in γ -type cations are in a boat-shaped site. The population of these different sites depends on Si/Al ratio, type of zeolite, method of preparation (both of zeolite itself and of loading of the metal ions), and so on.

747

748

24 Environmental Catalysis over Zeolites

MxOy

Metal-oxide nanoparticle (inside pore structure)

Mn+ bare cation

O

Mn+ oxo- cation (a)

g b g b

b b a

g b g a

a

a FER

MFI (ZSM-5)

g Al

Si

Si

b

Cu

Si Al

a

(b) Figure 24.1 (a) Schematic model of cation location in zeolites. (b) Simpliﬁed sites and local framework structures of the α-, β-, and γ -type Me(II) ions in MFI and FER zeolites. Adapted from [19]. The model of localization of copper ions in ferrierite (FER) (bottom part of the ﬁgure) has been adapted from [33].

24.2 A Glimpse into Opportunities and Issues

It is also known that, during ionic exchange, the electrostatic ﬁeld within the zeolite causes precipitation of metal-hydroxo species in the larger zeolite cages. Depending on the thermal treatment, these precursor species may form isolated cationic copper species, oxocations, or metal-oxide nanoparticles [20]. On the other hand, by conventional impregnation, the low rate of diffusion within the zeolites causes a preferential precipitation of these metal-hydroxo species on the external surface of the zeolite crystals. It is, thus, known from the recent ﬁndings that the preparation of Cu/MFI by impregnation leads mainly to CuO particles on the external zeolite surface [21]. Therefore, the preparation method is a key parameter to maximize the concentration of the active copper species and their performance in the decomposition of NO [22–25]. However, the great sensitivity of the nature and distribution of the active species to the zeolite preparation and their characteristics, coupled with their difﬁcult characterization associated with the low amount of transition-metal ions present in the zeolite (typically, few percentage by weight or less), has resulted in the often contradictory ﬁndings in the literature, which still are not completely resolved. For this reason, the identiﬁcation of the nature of active sites was for a long time, and still is, under discussion. Miyamoto et al. [26] using molecular dynamics (MDs) simulations and molecular orbital (MO) calculations suggested the presence of Cu(II)–O–Cu(II) species, when two Al atoms occupy T8 sites in the six-membered ring, and indicated this species as the active one. However, the Takaishi rule [27] indicates that two Al atoms cannot be placed in one pentasil ring. Catlow et al. [28, 29] also proposed by computational methods that the active site in Cu/MFI catalysts is associated with two copper ions bridged with extra-framework OH species. Figure 24.2 reports the model of active sites proposed by Catlow et al. [29]. Wichterlov´a et al. [30] using Cu+ photoluminescence and infrared (IR) spectroscopy identiﬁed two main Cu sites (denoted as Cuα and Cuβ ). The Cuβ site is preferentially occupied at low Cu loadings and exhibits a more packed environment and higher positive charge. The Cuα site is occupied in the whole concentration

OH CuI

CuII Al

Figure 24.2 Model of active sites in Cu/MFI for NO decomposition proposed by Catlow et al. [29]. Adapted from [29].

749

750

24 Environmental Catalysis over Zeolites

range, but predominantly at high loadings, approaching and exceeding 100% degree of ion exchange. It is in a more open coordination, possesses lower positive charge, and was suggested to be balanced by a single framework Al atom, in contrast to two framework Al atoms for Cuβ . The reducibility of these Cu sites differs substantially and depends on the local negative framework charge. The latter depends on both the local Si–Al ordering adjacent to the Cu ion and the total framework charge given by the Si/Al ratio. The analysis of the catalytic activity and reducibility versus the population of the Cuα and Cuβ sites allowed the identiﬁcation of the Cuα site as the active center for NO decomposition. The results of Wichterlov´a et al. [30] thus indicate a single, bare cation as the active site in the decomposition of NO. Various other authors have discussed the nature of active sites and the reaction mechanism in the decomposition of NO over copper-containing zeolites (see the review of Centi and Perathoner on the conversion of nitrogen oxides over copper-based catalysts [31]) using transient reactivity studies, spectroscopic investigations, and theoretical modeling. Speciﬁc overviews of the theoretical studies of the reaction of NO on metal-exchanged zeolites have been reported recently by Pietrzyk and Sojka [32], McMillan et al. [33], and Schneider [34]. General aspects of the reaction mechanism of conversion of nitrogen oxides have been discussed also by Busca et al. [35] and Centi and Perathoner [36]. The recent theoretical study of Pulido and Nachtigall [37] indicated a cyclic dinitrosyl complex (hyponitrite-like) as the key intermediate. The model is reported in Figure 24.3. Three types of dinitrosyl complexes with different coordination on the Cu+ cation were identiﬁed by Pulido and Nachtigall [37]: (i) fourfold tetrahedral, (ii) fourfold square planar, and (iii) threefold trigonal planar. The most stable dinitrosyl complex, formed when the two NO molecules interact with Cu+ via the N atom, has a tetrahedral coordination on Cu+ . The cyclic adsorption complex, having a square-planar arrangement of ligands on Cu+ and interaction via O atoms, is only about few kilocalories per mole less stable than the N-down

O1 1.925 Cu 1.926

1.300 N1 1.317 N2 1.299

O2

Figure 24.3 Active key intermediate (hyponitrite-like) in DeNOx reaction over Cu/FER proposed by Pulido and Nachtigall [37].

24.2 A Glimpse into Opportunities and Issues

dinitrosyl complex. This cyclic dinitrosyl complex is suggested to be the key intermediate in the DeNOx process taking place in Cu/zeolites. This model thus proposes that an isolated Cu+ ion is the active site. The formation of gem dinitrosyl species on isolated copper ions has been identiﬁed from the earlier studies by IR spectroscopy and considered a key characteristic aspect. However, different ideas have been proposed about their role on the reactivity. Bell et al. [38] suggested the mechanism reported in Figure 24.4a, based on the evidences by IR spectroscopy of the gradual substitution of Cu+ (NO)2 dinitrosyl species by Cu2+ (NO) and Cu2+ (O− )(NO) species with formation also of Cu2+ (NO2 ) and Cu2+ (NO3 − ) species. The key intermediate according to Bell et al. [38] is thus a Cu2+ (NO2 − )(NO) or Cu+ (N2 O3 ) species, supporting the original indications of Li and Hall [39] while studying the kinetics of decomposition of NO that the rate-limiting step is the adsorption of NO by an extra-lattice oxygen (ELO) containing site (e.g., Cu2+ O− ). Zecchina et al. [40] proposed a slightly different reaction mechanism based on the role of single copper species and the key formation of a Cu2+ (NO2 − )(NO) intermediate (Figure 24.4b). They explicitly exclude the role of copper dimeric species. Schay et al. [41] also indicate Cu2+ (O)(NO)(NO2 ) as the key intermediate in NO decomposition. Ramprasad et al. [42] reported, in agreement with this hypothesis, density functional theory (DFT) results showing that the single-step, symmetric, concerted decomposition reaction of NO in the vicinity of Cu ion sites in zeolites is forbidden by orbital symmetry. On the contrary, metastable hyponitrite complexes display N–N coupling and may be precursors for multistep decomposition of NO. Schneider [34] observed that in the ‘‘hyponitrite’’ structure (Figure 24.5, left) the bonding resulting from charge transfer from Cu to the adsorbate drives the formation of a short N–N bond resulting in the formation of N2 O and a ZCuO copper-oxo species, where Z indicates the zeolite. NO

Cu+

Cu+(NO)2

(O)

–N2O NO

+ – O−

Cu2

Cu+

N2

Cu2+

N

Cu2+

O−

NO

O Cu+

N N O

NO

N2

Cu2+

NO2−

NO2−

NO

O O

(a)

NO

N2

O2

O (b)

Figure 24.4 Reaction mechanisms in the decomposition of NO on Cu/zeolites proposed by Bell et al. [38] (a) and Zecchina et al. [40] (b).

Cu2+

N2 O−

O2

751

752

24 Environmental Catalysis over Zeolites

O N

O N

N N N

N

O O

O

O

Cu

Cu

Cu

Hyponitrite intermediate

Transition state E ‡ = 15 kcal mol−1

ZCuO + N2O E ‡ = −2 kcal mol−1

Figure 24.5 Conversion of metastable ZCu-bound hyponitrite intermediate to a copper-oxo (ZCuO) species and N2 O as proposed by Schneider [34]. Adapted from [34].

The copper-oxo species may further react with N2 O according to the following mechanism: ZCuO + N2 O → [ZCuOO-NN]‡ → ZCuO2 + N2

(24.1)

Subsequent desorption of O2 from ZCu would complete the cycle. The second O-atom transfer reaction has an energy barrier of about 36 kcal mol−1 and is thus the more difﬁcult step. Pietrzyk and Sojka [32], however, observed that the conversion of {CuI N2 O2 }Z transient species is kinetically constrained by the intersystem crossing, because the spin singlet {1 CuI N2 O2 }Z intermediate is converted to {3 CuO}Z center which has a spin triplet ground state. The alternative mechanism to this outer-sphere coupling leading to dimeric {CuI N2 O2 }Z intermediate is an inner-sphere nitrosonium route [32]. The oxidative adsorption, giving rise to the bound NOδ – species, deﬁnes a nitroside pathway of activation, while the reductive adsorption leading to NOδ+ species deﬁnes a nitrosonium pathway of activation. In the latter case, the M–N–O moiety is highly bent (130–140◦ ), the N–O bond length shortened, and its polarization increased by about three times in comparison to the free NO molecule. The second NO molecule can coordinate to the metal center and, when the spin density is largely on the metal, the dinitrosyl complex may thermally decompose giving rise to the N2 O formation via an inner-sphere route. These two pathways thus largely depend on the characteristics of dinitrosyl conformation, which in turn depends on the characteristics of the metal and its charge as well. As discussed in detail by Pietrzyk and Sojka [32] in analyzing the characteristics of mononitrosyl complexes in various transition-metal-ion-exchanged zeolites, the coordination of NO leads to a pronounced redistribution of the electron and

24.2 A Glimpse into Opportunities and Issues

spin densities, accompanied by modiﬁcation of the N–O bond order and its polarization. The zeolite framework participates in the electron density and spin-density repartitions. Therefore, not only the nature and valence state of the transition-metal ion determine the characteristics of the NO coordination complexes (and possible pathways of transformation) but also the site of the transition metal in the zeolite (Figure 24.1) and the zeolite framework, which determines the effectiveness of the zeolite in charge and spin-density redistribution. The situation is therefore more complex than that usually considered from theoretical modeling. In addition, the question is whether the model of a single (isolated) metal complex is valid for this type of catalysts. The presence of multiple copper sites, as well as their possible electronic interaction that modiﬁes the charge and spin-density redistribution, is a clearly issue. In addition, an unsolved question is also the coupling effect, that is, the inﬂuence of chemisorption coverage. As pointed out earlier, there is a relevant role of the zeolite framework itself in determining the conﬁguration of the mononitrosyl complexes. This is also demonstrated by IR studies, which have provided evidence that NO chemisorption on Cu/MFI modiﬁes the skeletal vibrations of the zeolite. There is thus a relaxation effect due to chemisorption, that is, the conﬁguration of the sites at low chemisorption coverage is not exactly the same of that at high chemisorption coverage. It is still an unsolved question whether this effect could be also relevant in terms of reaction mechanism. In addition to single copper sites, multinuclear copper sites could be present and be the active sites. In particular, much attention has been focused on the possibility of having two copper ions in a close enough proximity to act in concert for catalysis [43]. A number of possible single-O-bridged and di-O-bridged Cu pairs can be identiﬁed by theoretical modeling [34] (Figure 24.6) and are quite strongly bound: ZCuO + ZCu → ZCu-O-CuZ,

E = −60 kcal mol−1

ZCuO2 + ZCu → ZCu-O2 -CuZ, E ∼ −40 kcal mol−1

(24.2)

(24.3)

Oxygen desorption being the rate-limiting step in the decomposition of NO, a copper pair would facilitate this reaction. Kuroda et al. [44, 45], while studying the X-ray absorption spectroscopy of the oxidation–reduction processes during NO adsorption on Cu/MFI, concluded that zeolite having an appropriate Si/Al ratio, in which it is possible for the copper ion to exist as dimer species, may provide the key to the redox cycle of copper ion as well as catalysis in NO decomposition. The role of the multi-ionic structure of the active copper centers in Cu/MFI has also been indicated recently [46]. It is out of the scope of this review to go into the details of the different observations and conclusions on the reaction mechanism of the decomposition of NO. The comments above evidence the complexity of the problem and the still on-going discussion after many years even for such a simple reaction as the decomposition of NO on apparently well-deﬁned copper ions in an ordered environment (the zeolite host).

753

754

24 Environmental Catalysis over Zeolites

O Cu

Cu

O

Cu

Cu O

0.345 nm

0.289 nm

Single O-bridged Cu pair

µ di-oxo bridged Cu pair

µ-η2–η2 di-oxo-bridged Cu pair

Peroxo bridge Cu pair

O

O Cu

Cu O

0.342 nm

Cu

Cu O

0.418 nm

Figure 24.6 Spin-polarized local density approximation structures of O-bridged Cu pairs, calculated for a single T-site (AlO4 tetrahedra). Adapted from [34].

The idea behind most of the hypotheses of the reaction mechanism, and the stimulus of several theoretical investigations, is that in Cu/zeolite a single well-deﬁned site is present. As commented above, different copper sites are present, and they can change reversibly during the catalytic reaction or during the various procedures necessary for their characterization. The relocation of copper ions depending on temperature and gas composition, as well as the easy and spontaneous reduction of copper ions, is known. Furthermore, the site of the copper ions depends on the coordination of chemisorbed molecule. For example, IR results show that the coordination of NO to Cu+ ions moves them to more open positions in the cage. Broclawik et al. [47] reported recently that Cu+ in site II binds NO molecule more weakly than it does in site III. It is due to the very stable planar threefold coordination to the framework O atoms that is achieved in site II. However, the heats of adsorption of the second NO molecule are only slightly higher at site III than at site II, because upon the ﬁrst NO molecule adsorption on site II, Cu(I) ions are displaced to the position more resembling site III. Similarly, CO coordination on Cu+ ions in ZSM-5 induces a displacement of copper ions, and the coordination of NO oxidizes Cu+ ions as was observed [48]. As mentioned, by IR it is observed that the overtone bands of the skeletal vibration of the zeolite structure depend on the chemisorption coverage, indicating that the adsorption of a molecule on the copper site does not have only a local (short-range) inﬂuence. The local coordination of copper ions would depend on the chemisorption coverage. All these indications point to the highly mobile situation of the copper ion species in the zeolite and thus limit considering the presence of a well-deﬁned type of copper species, as made in theoretical approaches. We should also consider

24.2 A Glimpse into Opportunities and Issues

that a synergetic effect between the various copper species could be also present. It is known, for example, that extra-framework Al ions inﬂuence the acidity of the nearby Br¨onsted sites. Extra-framework copper ions may thus inﬂuence the behavior of isolated copper species either directly (e.g., participating in the reaction mechanism) or indirectly (e.g., acting as a sink for charge and spin density, thus inﬂuencing the stability of the different nitrogen oxide complexes or, instead, mediating the oxygen desorption and thus increasing the rate-limiting step). All these aspects should be considered in discussing the reaction mechanism. As mentioned, the starting observation from reactivity tests is that the speciﬁc activity is higher in overexchanged Cu/MFI, for example, when the amount of copper ions is higher than the amount to balance all Al ions (one Cu2+ each for two Al ions). The typical maximum speciﬁc activity was observed for the level of exchange in the 100–140% range. This observation was the starting point to postulate that dimeric copper ions (with one or two bridging oxygen atoms) are the active sites. In overexchanged zeolites, the deposition of [CuOH]+ species occurs during ion exchange and these species could form by dehydration of oxygen-bridged copper species or chains. The question, however, is whether they are the active sites. In fact, other factors could explain the need of overexchange: (i) multinuclear copper species (such as the Cu pairs shown in Figure 24.6) are the active sites and they form predominantly in overexchanged Cu/zeolites; (ii) isolated copper ions in more open positions (Cuα sites), which show higher activity than the other isolated copper species, form only when the other positions are occupied and at high exchange level; and (iii) there is a synergism between isolated and multinuclear copper sites, but only when they are in close proximity (as occurs in overexchanged samples). More interpretations are also possible, but already these three show the difﬁculty in reaching unique conclusions. We may conclude, as is often remarked, that these zeolites containing metal ions in extra-framework positions can be considered as enzyme-like materials because of the presence of an isolated metal ion surrounded by a 3D environment that orients the adsorption and reactivity of the incoming molecules. The brief discussion reported above shows that the situation is far more complex, with multiple sites that could be in a dynamic equilibrium. Also from the characterization perspective, the system is more complex and far more difﬁcult that initially supposed. This is one of the reasons for the still on-going debate on the mechanism of one of the (apparently) simplest reactions that can be considered, namely the decomposition of NO. From the application point of view, the main characteristic is the possibility to have isolated and well-accessible metal ions, but typically for very low loadings, which can determine a low productivity per total catalyst weight. The redox and reactivity characteristics of these metal ions, however, are different from those of the same ions supported over conventional oxides (silica, alumina, etc.). The relatively high mobility, however, determines a general low stability especially under hydrothermal conditions. In the case of decomposition of NO on Cu/MFI, the presence of gaseous O2 decreases the reaction rate because it competes with NO for the chemisorptions on reduced copper ions. The self-reduction of copper ions is thus the rate-limiting step.

755

756

24 Environmental Catalysis over Zeolites

Recently, however, alternative zeolitic materials have been proposed that apparently overcome this problem. Shi et al. [49] investigated the direct decomposition of NO in the presence of oxygen over a series of Fe–Mn/H-beta catalysts with Mn/Fe = 1. When Mn is incorporated to Fe/H-beta, NOx conversion is improved, and the active temperature window is lowered. The Fe(5%)–Mn(5%)/H-beta exhibits the highest activity. Neither O2 nor CO2 inhibits the reaction. These results thus prove that there is still ample research space to develop improved Me-zeolite catalysts for the decomposition of NO.

24.3 Fields of Applications

Transition-metal-containing zeolites have been very successfully applied in reﬁnery and petrochemistry, but a signiﬁcant interest has also raised their application in the ﬁeld of environmental protection although their practical use is still much more limited as shown by the number of studies on these catalysts. There are two main areas of applications of zeolites in the ﬁeld of environmental catalysis, within the limits discussed in the introduction: • conversion of nitrogen oxides (NOx and N2 O) and • conversion of VOCs. Other areas of applications include wastewater treatment, photocatalytic conversion of pollutants, air puriﬁcation, and soil remediation, although often it is not well demonstrated that zeolite-based catalysts offer clear advantages over alternative catalysts such as oxide-supported materials. An overview of the applications of zeolites in environmental catalysis has been reported by Larsen [50] and earlier by Delahay and Coq [51]. More speciﬁc reviews to be cited are the followings: • the use of zeolite catalysts for dehalogenation processes [52]; • the use of metallo-zeolites (particularly, Co/MFI, Fe/MFI, ferrierite (FER), and mordenite (MOR)) for NOx selective catalytic reduction (SCR) with hydrocarbons [19]; • the use of metal-ion-exchanged zeolites as DeNOx catalysts for lean-burning engines [53]; • the reaction mechanisms of lean-burn hydrocarbon SCR over zeolite catalysts [54], and the use of zeolites in the ﬁeld of nitrogen monoxide removal [55, 56]; • the quantum mechanical modeling of the properties of transition-metal ions in zeolites, with particular reference to the SCR of nitrogen oxides in the presence of NH3 [57]; • the use of zeolites in pollution control, in particular the abatement of NOx and N2 O emissions from stationary sources [58]; • the characterization, modeling, and performances (particularly for NOx removal from the tail gas of nitric acid plants) of CuI /CuII -Y zeolites [59]; • the SCR of NOx by ammonia (NH3 -SCR) over metal-exchanged zeolites (especially Fe/MFI) for diesel engine exhaust applications [60];

24.3 Fields of Applications

• the use of transition-metal oxides (Ti, V, Mo, Cr) incorporated within the framework of zeolites as well as transition-metal ions (Cu+ , Ag+ ) exchanged within the zeolite cavities for the photocatalytic conversion of NOx (NO, N2 O) or the reduction of CO2 with H2 O [61], or as photoanode [62]; • the use of microporous and mesoporous materials to prepare single-site photocatalysts [63, 64] or to enhance the photocatalytic activity of TiO2 through spatial structuring and particle size control [65]; • the use of zeolite and other novel materials for wastewater treatment [66]; • the use of natural zeolites for environmental applications in water puriﬁcation [67] and greywater treatment [68]; • the use of zeolites in the wet hydrogen peroxide catalytic oxidation of organic waste in agro-food and industrial streams [69]; and • the use of zeolite ﬁlms for trace pollutant removal from air, and other applications [70]. Selected examples of the uses of zeolites as environmental catalysts are reported in Table 24.1 to show the very broad range of materials and applications that have been investigated. However, it should be remarked that their commercial application is still dominated by their adsorption properties rather than by their use as catalysts [71], at least for for niche applications, although we are probably close to seeing their use in large-scale applications. As an example, Zeolyst International (one of the world’s leading producers of zeolite catalysts) reports in its web site (www.zeolyst.com) the use of zeolite for environmental waste reduction and secondary treatment of efﬂuents (reduction of NOx and VOCs, including automotive cold-start emissions), as well as the removal of automotive combustion products (particularly for lean NOx ), even though only their ‘‘potential’’ use is indicated. BASF Catalysts (www.catalysts.basf.com, the world leader in catalyst manufacture) has recently announced (November 2008) at the SAE (Society of Automotive Engineers) International Commercial Vehicles Congress in Chicago a new copper-zeolite SCR catalyst that provides more than 90% NOx (nitrogen oxides) removal for on-road and off-road heavy-duty diesel vehicles. The catalytically active component is coated on a ceramic honeycomb and will be used with urea as a reductant. On January 2009, the opening of a new R&D facility focused on commercializing the new zeolite catalysts that would boost diesel and gasoline yields compared to conventional zeolites was also announced. A number of papers at the SAE (the most important society in this ﬁeld) meetings from leading car manufacturers indicate that Cu/MFI or Fe/MFI zeolites are close to being introduced commercially for the control of diesel emission. For example, researchers of Ford Motor Co. (USA) [101] reported that the SCR is a viable option for the control of oxides of nitrogen (NOx ) from diesel engines using urea as a reductant, in particular to meet certiﬁcation under Tier 2 Bin 5 (USA) emission requirements (passenger and light-duty diesel vehicles will require up to 90% NOx conversion over the Federal Test Procedure – FTP). Currently, copper-zeolite (Cu/zeolite) SCR catalysts are favored for conﬁgurations where the exhaust gas temperature is below 450 ◦ C for the majority of operating

757

758

24 Environmental Catalysis over Zeolites Table 24.1 Selected examples of the use of zeolites as environmental catalysts (Y, FAU structure; ZSM-5, MFI structure; Beta, BEA structure; ferrierite, FER structure).

Catalyst

Application

References

Cu/FAU Fe/MFI Cu/MFI Cu/BEA Cu/MFI Fe/MFI, Cu/MFI Ag/MFI, Ag/BEA Ag/FAU Ir/MFI Zeolites (as additives)

NOx removal from tail gas of nitric acid plants N2 O abatment in nitric acid plants NO-assisted N2 O decomposition NOx conversion with propane NOx conversion with decane Urea-SCR (diesel vehicles) NO reduction with propane in the presence of H2 NOx reduction with oxygenates Simultaneous removal of soot and NOx NH3 transient storage to enhance NOx conversion in storage-reduction catalysts NOx SCR with methane Low-temperature catalytic combustion (CO, HC) VOC removal

[51] [72] [73] [74, 75] [76] [77] [78] [79] [80] [81]

Deep oxidation of 1,2-dichloroethane Catalytic combustion of trichloroethylene Catalytic ozonation of toluene

[85] [86] [87]

VOC catalytic ozonation Wet oxidation of pollutants with H2 O2 Hydrodehalogenation of chlorocarbons in water

[88] [69, 89–92] [52]

Soil amendment, wastewater puriﬁcation (Cr/VI) photoreduction for water and soil remediation Photocatalytic elimination of pollutants (air)

[93] [94]

Co, Mn/FER Pd/MCM-41 Cu/ and Pt/MFI on cordierite CeO2 /FAU Pt/P-MCM-41 Zeolites and MCM-41 Pd/BEA, Pd/FAU Fe/MFI, Cu/MFI Pd/hydrophobic FAU Natural zeolites Fe/MFI Ti- and TiO2 -micro and mesoporous materials FAU, MFI Fe/MFI Ag/clinoptilolite

Elimination of nitrosamines in cigarette smoke Oxidation of lignin with H2 O2 (pulp and paper mill wastewater) Water disinfection

[82] [83] [84]

[62–64, 95, 96]

[97, 98] [99] [100]

SCR, selective catalytic reduction.

conditions, while iron zeolite (Fe/zeolite) SCR catalysts are preferred where NOx conversion is needed at temperatures above 450 ◦ C. The key issue in the control of NOx emissions of light-duty diesel engines is to have activity at very low temperatures (the catalytic converter has a temperature below 350 ◦ C for most of the emission cycle), while higher temperature performances and stability is more critical in heavy-duty diesel engines.

24.3 Fields of Applications

The selection of Cu/zeolite or Fe/zeolite SCR catalysts is based on the different performance characteristics of these two catalyst types. Cu/zeolite catalysts are generally known for having efﬁcient NOx reduction at low temperatures with little or no NO2 , and they tend to selectively oxidize ammonia (NH3 ) to N2 at temperatures above 400 ◦ C, leading to poor NOx conversion at elevated temperatures. Fe/zeolite catalysts are very efﬁcient at NOx conversion at temperatures as high as 600 ◦ C or even higher, but they are not as efﬁcient as Cu/zeolite catalysts at lower temperatures in the absence of NO2 . A combined SCR system consisting of an Fe/zeolite catalyst in front of a Cu/zeolite catalyst could thus widen the operating temperature range of the SCR catalyst. At low temperatures, the Cu/zeolite improves NOx conversion efﬁciency versus an Fe-only system. At elevated temperatures, the Fe/zeolite is more active. In addition, one can overdose NH3 at elevated temperatures with the combined Fe–Cu system without NH3 slip, while the Fe-only system leads to substantial NH3 slip when overdosing. One of the key aspects of the performance is the ability of the zeolitic materials to store both hydrocarbons and ammonia [102]. Ammonia storage on zeolite has a beneﬁcial effect on NOx conversion; hydrocarbons, however, compete with ammonia for storage sites and may also block access to the interior of the zeolites, where the bulk of the catalytic processes take place. Another issue is the hydrothermal stability. The high-temperature SCR deactivation is unavoidable due to the requirements necessary to actively regenerate diesel particulate ﬁlters and purge SCRs from sulfur and hydrocarbon contamination. Careful temperature control of these events is necessary to prevent unintentional thermal damage, which is not always possible. As a result, there is a need to develop thermally robust SCR catalysts. Fe/zeolite formulations are known to exhibit superior hydrothermal stability over Cu/zeolite formulations [103]. However, current Fe/zeolite formulations are not very active for NOx conversion in the desired 200–350 ◦ C temperature regime under conditions of low NO2 /NOx ratios. Cu/zeolite, however, may be stabilized, and the latest state-of-the-art Cu/zeolite formulations show remarkable high-temperature hydrothermal stability up to 950 ◦ C while maintaining stable low-temperature NOx activity [103]. The deactivation by sulfur is another issue [104]. Even with the use of ultralow-sulfur fuel, sulfur poisoning is still a durability issue for base metal/zeolite SCR catalysts. The impact of sulfur is more severe on Cu/zeolite than on Fe/zeolite SCR catalysts for the NOx activity, but the sensitivity of thermal aging status to the sulfur poisoning impact is different [104]. The most thermally durable SCR catalysts are not necessarily the most resistant to sulfur poisoning. Control of sulfur poisoning and strategy for DeSOx might be different depending on the formulation and thermal aging status. The alternative catalytic system for urea-SCR is based on vanadium oxide supported on titania. These SCR catalysts are now available in mass production for heavy-duty vehicles in Europe. The SCR-active material can either be applied as a coating on an inert carrier honeycomb or be worked up to a homogeneous honeycomb by extrusion. The homogeneously extruded catalysts have the advantage that they contain 100% of the active material. Especially in the lower temperature

759

760

24 Environmental Catalysis over Zeolites

range, higher NOx conversions can be achieved compared to coated systems. In addition, they feature superior resistance against poisoning (e.g., by sulfur). However, to establish this technology for the new US 2010 and EURO 6 systems containing a particulate ﬁlter and to meet the demands of the US market for a catalyst that does not contain vanadium, there is the need to use alternative zeolite-based catalysts. The reason for not using vanadium in these future systems is that ﬁlter regeneration demands high-temperature stability of the SCR catalyst and that there are concerns about vanadium loss during operation. Researchers of SINOx Emission Control (former Argillon GmbH, and now Johnson Matthey Catalysts Germany) [105] have developed a homogeneous catalyst honeycomb catalyst consisting of 100% active material based on zeolite and containing no vanadium. The catalyst features superior low-temperature activity and provides high-temperature stability to withstand particulate ﬁlter regenerations upstream in the exhaust. Additionally, the low bulk density helps to reduce efforts in canning and system design. Another relevant example of zeolites as environmental catalysts, which will probably become a major application in the near future, concerns the abatement of N2 O emissions in the production of nitric acid. The industrial production of nitric acid (HNO3 ) involves oxidizing ammonia (NH3 ) with air over a platinum/rhodium gauze catalyst to produce nitrogen oxides. This process yields nitrogen monoxide (NO), which then reacts with oxygen and water to form nitric acid. However, it also produces nitrous oxide (N2 O) – a powerful greenhouse gas and ozone killer (N2 O has a greenhouse effect about 300 times higher than that of CO2 [106]) – as an undesired by-product. Unlike NO, the nitrous oxide is not involved in the HNO3 production process and is emitted into the atmosphere with the tail gas. The N2 O emissions in nitric acid plants vary from about 3 to 4 kg of N2 O per metric ton of HNO3 to as much as 20 kg of N2 O per metric ton of HNO3 depending on the type of nitric acid plant. An estimated 400 000 metric tons of nitrous oxide is emitted each year by nitric acid plants worldwide. Nitric acid plants are now the largest single source of greenhouse gas emissions among industrial manufacturing facilities. Although still very few countries have introduced limits on emissions of N2 O from nitric acid plants, the recent agreement on greenhouse gas emissions will soon extend these limits to more countries or there will be the need to include N2 O emissions in the implementation of mechanisms of Kyoto Protocol carbon trading protocols (Emissions Trading, Joint Implementation, and Clean Development Mechanism). Uhde (a world leader in nitric acid technology) in collaboration with S¨ud-Chemie AG (a world leader in special zeolite manufacture) has developed a technology for removing N2 O from the nitric acid production process (in particular from the tail gas), which is based on the use of Fe/MFI catalysts [107, 108] (see also Uhde publications no. 5000008.00 – EnviNOx Setting Emission Standard for Nitric Acid Plants – which can be downloaded from the site www.uhde.eu). The Fe/MFI-based catalyst is active either in decomposing N2 O into N2 and O2 – an effect increased signiﬁcantly by the presence of NOx in the tail gas (cocatalytic NOx effect; the NOx acts as scavenger for the oxygen left on the iron site as a result

24.3 Fields of Applications

of the N2 O decomposition and thus accelerates the reaction rate, because oxygen desorption is the rate-limiting step in the reaction), or by reducing N2 O using various reducing agents, such as hydrocarbons [109, 110]. In addition, the iron zeolites have also proved to be excellent DeNOx catalysts in a wide temperature window. The special iron-exchanged zeolite catalysts for use in its EnviNOx process are indicated EnviCat-N2 O and EnviCat-NOx, and are commercialized in the pellet form. The advantages of these zeolite catalysts compared to conventional DeNOx catalysts are their resistance to typical catalyst poisons, such as sulfur or chlorine, and their operating range over a wide temperature window of approximately 200–600 ◦ C. There are different possible EnviNOx process variants: (i) catalytic decomposition of N2 O and the catalytic reduction of NOx , (ii) catalytic reduction of N2 O and NOx , and (iii) catalytic decomposition of N2 O. Figure 24.7 shows the simpliﬁed ﬂow sheet of the ﬁrst two process variants [111]. In the ﬁrst option, the reactor is usually located between the ﬁnal tail gas heater and the tail gas turbine and contains two catalyst beds ﬁlled with iron zeolite catalysts operating at the same pressure and temperature, and a device for the addition of NH3 between the beds. In the ﬁrst O stage, the N2 O abatement is effected simply by the catalytic decomposition of N2 O into N2 and O2 . Owing to NOx promotion of the decomposition of N2 O, the DeNOx stage is downstream. This process variant is especially applicable for tail gas temperatures between about 425 and 520 ◦ C. The ﬁrst commercial-scale plant using this option was realized in 2003 by Uhde in Linz (Austria) for Agrolinz Melamine International (AMI), which produces about 1000 t per day of HNO3 and in which the tailgas has a ﬂow rate of 120 000 Nm3 h−1 at a temperature of 430 ◦ C. A conversion rate of 98–99% is achieved. In the second variant, N2 O is removed by catalytic reduction with a hydrocarbon such as natural gas or propane. Unlike with N2 O decomposition, the NOx content of the tail gas inhibits the N2 O reduction reaction. It is, therefore, necessary to completely eliminate the NOx in the tail gas. Depending on the tail gas composition and the particular operating conditions, this can be accomplished in a DeNOx unit located upstream of the DeN2 O stage or, preferably, simultaneously with the N2 O reduction in a single common stage. Either the common stage process or the two-stage process with its hydrocarbon reducing agent feed mixer can be accommodated in a single reactor vessel. This option is suitable for temperatures between about 300 and 520 ◦ C depending on the speciﬁc conditions in the nitric acid plant. The ﬁrst commercial plant using this option was realized in 2006 in Egypt for the company Abu Qir Fertilizers. It uses ammonia for NOx SCR and methane for N2 O SCR. The plant has a nitric acid production of 1870 t per day and the tail gas ﬂow rate is 225 000 Nm3 h−1 (temperature = 410 ◦ C). The N2 O conversion is 99%, while NOx outlet concentration is below 1 ppm. Several other plants were realized by Uhde/SCAG for a total removal of about 8 million tons per year of equivalent CO2 . Other companies have also developed similar processes using Fe/zeolite catalysts.

761

24 Environmental Catalysis over Zeolites

2N2O → 2N2 + Ο2

To atmosphere

H2O

Process gas

Tailgas turbine

DeN2O®

EnviNOx® reactor

Absorption tower

cw

catalyst Ammonia DeNOx catalyst

cw

(N2O, NOx)

HNO3 aqueous

(a)

To atmosphere Ammonia

Process gas (N2O, NOx)

(b)

cw

EnviNOx® reactor

H2O

Absorption tower

762

Tailgas turbine

DeNOx catalyst Methene DeN2O® catalyst

cw

CH4 + 3 − 4N2O → COx + 2H2O + 3 − 4N2

HNO3 aqueous

Figure 24.7 EnviNOx (Udhe) process of catalytic decomposition of N2 O and catalytic reduction of NOx (a), and catalytic reduction of N2 O and NOx (b). DeN2 O catalyst = Fe/MFI, DeNOx catalyst = Fe/MFI or V2 O5 /TiO2 /WO3 . Adapted from [111].

The choice of the hydrocarbons for nitrous oxide abatement depends on various criteria [112]: (i) DeN2 O activity and operation temperature, (ii) hydrocarbon utilization, that is, the selectivity to react with N2 O in O2 excess, (iii) emission of CO and CO2 , (iv) sensitivity to NO and NH3 , and (v) cost. Alkanes are generally more effective reducing agents than unsaturated hydrocarbons. In particular, alkynes require a higher temperature to activate N2 O and are unselective because

24.3 Fields of Applications

of their proneness to react with O2 . Ethane is an optimal reductant, featuring a high DeN2 O activity, high selectivity, and compared to methane, a lower degree of inhibition by NO. Usually, steam activation is necessary to promote the performances of Fe/MFI catalysts, but the preparation method has also a relevant effect. Other zeolite structures have also been shown to be active in the reaction. Recently, Sobalik et al. [113] showed the superiority of Fe/FER over Fe/BEA and Fe/MFI with Fe/Al ratios below 0.15 in the decomposition of nitrous oxide in the absence of NO. The higher performances were tentatively attributed to the presence of a unique FER structure containing two close collaborating Fe(II) cations accommodated in the cationic sites of FER. This local structure with unique spatial properties could, due to their distance and orientation, provide for N2 O splitting by mutual action of two adjacent iron cations in Fe/FER. These isolated cationic sites are located in two adjacent β sites facing each other across the channel (Figure 24.8). The distance ˚ A strong attraction between the oxygen between two iron cations is about 7.5 A. ˚ atom of Fe . . . NNO complex and the adjacent iron cation (distance O . . . Fe <3 A) could be expected. Regarding the probability of the existence of the active sites with such an arrangement of close Fe(II) cations, the occupation of various cationic positions by divalent cations indicates that the two close Al atoms forming mainly β sites prevail in the FER framework [114]. If the iron exchange leads to a random occupation of the cationic sites, then at least 5% contains two Al atoms in the β site balancing Fe(II) in the arrangement that a neighboring unit cell also accommodates Fe(II) in the β site, thus forming the active site containing two closely collaborating Fe(II) cations accommodated in the cationic sites. Such an arrangement of two Fe(II) cations located in the β sites of two neighboring unit cells allowing the formation of collaborating Fe(II) cations in Fe . . . Fe FER

Al

Al

Fe N N – O

Al

Figure 24.8 Periodic DFT-optimized structure of Fe/FER including N2 O interacting with two Fe cations occupying adjacent beta cationic positions. Adapted from [113].

Fe

Al

763

764

24 Environmental Catalysis over Zeolites

pairs is quite unique and exists only for the FER sample used because of two reasons: the distribution of Al atoms in this FER framework as well as the optimal structural arrangement of the β sites in the FER framework. Sobalik et al. [113] noted that for highly loaded Fe/zeolites and for the NOx -assisted N2 O decomposition, the catalytic activity could be connected to other types of iron sites. Comparing Fe/MFI and Fe/BEA, P´erez-Ram´ırez et al. [115] claimed that the microporous matrix does not play a decisive role in the decomposition of N2 O, provided similar forms of iron are present in the ﬁnal catalyst. However, the results appear less convincing, but it is not the aim here to discuss in detail the nature of the active sites. It is to be mentioned, however, that many groups have investigated iron in MFI and other zeolitic structures [116–121]. A review on the properties of iron in various zeolite structures has been recently published by Nov´akov´a and Sobal´ık [122]. A recent review of Pirngruber [123] has compared in detail the properties of ironand copper-containing zeolites in the decomposition of N2 O and NO, and other reactions. In fact, a common aspect of the various mechanisms also discussed here is the formation of a Me–O active species, which can be usefully applied in selective oxidation reactions, such as the benzene hydroxylation to phenol and methane to methanol reactions. Pirngruber [123] observed that iron zeolites work best as selective oxidation catalysts at low iron concentrations. Moreover, the active sites for selective oxidation react speciﬁcally with N2 O, but not with O2 . This indicates that site isolation is an important premise for obtaining selective iron zeolite catalysts, although this aspect is in contrast with the conclusions reported above by Sobalik et al. [113]. The activity of iron zeolites increases when they are pretreated under steaming conditions, which causes the partial reduction of Fe3+ to Fe2+ and probably creates lattice defects. The Fe2+ sites in the vicinity of lattice defects are most probably the sites where the surface oxygen atoms, which were generated by dissociation of N2 O, are stabilized in a highly reactive form. In agreement with this conclusion, we earlier reported [124] the model of active sites in benzene hydroxylation reported in Figure 24.9 based on (−Si–O)2 Fe2+ species located at zeolite defect sites which react with N2 O forming a (−Si–O)2 Fe4+ =O in equilibrium with the (−Si–O)2 Fe3+ –O− species which was proposed to be the effective hydroxylating species. Pirngruber [123] also observed that copper zeolites catalyze similar reactions as their iron analogs, but there are some distinct differences. Cu zeolites work best at high copper loadings and their catalytic activity is strongly related to the formation of Cu dimers. He concluded that the active site is a bis(µ-oxo)dicopper species that is generated by the reaction of a Cu+ dimer with N2 O, NO, and O2 . However, as commented before, there is less than conclusive evidence on this statement. It is true, however, that the ability to react with O2 strongly distinguishes the active sites in copper zeolites from the selective oxidation sites in iron zeolites. There are thus not only analogies but also relevant differences between the two systems that have to be taken into account in studying their use as catalysts. As shown in Table 24.1, there are many types of zeolites investigated other than Cu/ and Fe/zeolites for the conversion of nitrogen oxides. To look at the general

24.3 Fields of Applications

O O O

Fe O

Al

Figure 24.9 Model of the iron active sites ((−Si–O)2 Fe2+ species located at zeolite defect sites) for benzene hydroxylation in Fe/MFI. Adapted from [124].

interest on these systems, Table 24.2 reports a short survey made using SCI-Finder on the number of publications (limiting to journals/review and English as the language) in the last 10 years using keywords as ‘‘Zeolites and . . . ’’. In general, it may be observed that there is still a large interest on the use of zeolites as environmental catalysts and they represent a signiﬁcant fraction of the publications on the use of zeolites as catalysts (see last items in the Table 24.2). It may be also observed that the new and more correct name of ‘‘microporous materials’’ is much less used. In terms of topics, the area of nitrogen oxide conversion is the dominant, even though already a large part of the activity was made before the year 2000, as shown also in this contribution. VOC and catalytic combustion is a second large area of interest, but the potential for practical applications is much more limited, as discussed later. The general area of water treatment (which includes wet oxidation, peroxidation, etc.) is also quite signiﬁcant in terms of the number of publications. It should be noted, however, that a large part of the publications deals with the use of zeolites (synthetic and natural) as adsorbent rather than as catalysts. Two major ﬁelds of application in this area are given below: 1) Use of Cu/ or Fe/zeolites as heterogeneous Fenton catalysts to activate H2 O2 and convert organic pollutants (dyes, recalcitrant, or toxic organics, etc.) to CO2 . The technique can be used, in particular, as pretreatment to remove toxic chemicals and facilitate consecutive biological treatment (aerobic, anaerobic)

765

766

24 Environmental Catalysis over Zeolites Table 24.2 Number of publications (SCI-Finder) in the period 1999–2009 (limited to journals/review and English) ﬁnd using as keywords ‘‘Zeolites and . . . ’’ or ‘‘Microporous materials and . . . ’’, where the second

keyword(s) are those reported in the following list. The results refer to the number of publications where the keywords are present as entered, or when both are found as concepts. Analysis updated on Jan. 15th , 2010.

Topic

Both of the Concepts

As entered

Zeolite and . . . . NOx N2 O Nitrogen oxides DeNOx SCR VOC Catalytic combustion Dehalogenation OR hydrodehalogenation Wastewater Water treatment Wet oxidation Wet peroxidation Photocatalytic OR photocatalyst Photodegradation Photo Ozone Nitrate Membrane Organic pollutants Non thermal plasma Environmental catalysis OR environmental catalyst Catalysis Catalytic Catalyst

867 633 1270 112 552 272 565 62 1124 1424 84 47 989 88 1848 108 1367 1662 95 41 688 14944 14640 14640

6 5 1 9 48 1 1 5 69 1 – – 112 1 2 9 11 747 – – 3 123 211 3856

Microporous materials and . . . . Catalysis Catalytic Catalyst Environmental catalysis OR environmental catalyst

813 756 751 38

11 6 1 –

or as post-treatment to the biological step to decolorize the water [69, 91, 125]. In terms of application, however, the major limitation is the leaching of the transition metal. Zeolites offer the advantage of a better protection of the active center from the fouling by humic and other macromolecular substances, but in general the advantages over oxide-based catalysts are limited. 2) Use of Pd (or other noble metal) zeolite for hydrodehalogenation of organohalogen compounds of environmental concern [52]. Pd/zeolite, in comparison with other Pd-based catalysts, may have the advantage of better hydrophobicity (in high Si/Al materials), and this could be an advantage for a higher stability

24.3 Fields of Applications

against ionic poisons. An example is the use of Pd-ZSM-23 (1 wt% Pd) for catalytic hydrodehalogenation of bromobenzene [126]. The activity is good as well as the recyclability, but there is no cost effectiveness in using these catalysts. The presence of alkali ions is important for the activity. Kanyi et al. [127] have discussed the dehalogenation versus dehydrohalogenation reactions of alkyl halides in zeolite NaX (Faujasite). Balance between substitution and elimination can be understood in terms of various factors: (i) Relative rates of C–X (X = Cl, Br, I) and C–H cleavage, which, upon substitution, is predominant when the C–X cleavage rate is greater than the C–H cleavage rate. C–X cleavage rates decrease in the order I > Br > Cl. (ii) Relative stability of the framework alkoxy species as a result of steric hindrance in the zeolite. This stability decreases in the order primary > secondary > tertiary. The area of use of micro- and mesoporous materials (containing isolated ions such as Ti or nanosized oxide particles such as TiO2 ) is a very fast growing area of interest. Many reviews have appeared on this topic [61–65, 95, 96 128–130] and the number of publications is exponentially rising, as also seen in Table 24.2. However, from the application point of view, this large interest does not correspond to signiﬁcant advantages that may be provided by these systems. It is true that the compartmentalized intracrystalline void space of zeolites allows incorporating and organizing photoactive guests that can be used as photocatalysts. The rigid micropores allow assembly of multicomponent systems comprising antennas and relays reminiscent of natural photosynthetic centers. Besides inorganic metal-oxide clusters, zeolites as host are particularly attractive to construct organic photocatalysts since the guest becomes signiﬁcantly stabilized by incorporation [96]. However, this potential has still to be demonstrated to lead to practical applications as catalysts, besides in perhaps niche cases. Large relevance has also been given to photocatalytic solids, in which the absorption occurs at isolated, spatially well-separated centers, such as isolated Ti ions in micro- or mesoporous materials (single-site photocatalysts) [63]. Their use for the decomposition of NO to N2 and O2 and for the selective oxidation of CO in the presence of H2 has been widely discussed [95]. However, the reaction rates are too low and they do not offer advantages over alternative photocatalytic materials. It should be noted that isolated Ti ions absorb in the UV region, and thus these systems go in the opposite direction of the recent large effort in developing photocatalysts active in the visible region. The same is true when the dimension of the TiO2 particles is reduced by incorporation in micro- or mesoporous materials [65]. There is an advantage of an increase in the surface area, but a blue shift of the absorption band overshadows often the ﬁrst advantage. However, the inclusion of photoactive centers in an ordered porous matrix may have the potential to allow shape-selective catalysis. The classic example is the use of the microporous titanosilicate ETS-10 [131, 132]. The photoactivated ETS-10 shows catalytic activity driven by the size and polarity of the substrates. ETS-10 efﬁciently catalyzes the conversion of substrates of a size larger than the pore diameter of ETS-10. In contrast, the reactivity of small substrates depends strongly

767

768

24 Environmental Catalysis over Zeolites

on the substrate polarity; less polar substrates show higher reactivity on ETS-10. Large substrates or less polar substrates hardly diffuse inside the highly polarized micropores of ETS-10 and, hence, react efﬁciently with the hydroxyl radicals formed on titanol (Ti-OH) groups exposed on the external surface of ETS-10. In contrast, small polar substrates diffuse easily inside the micropores of ETS-10 and hardly react with hydroxyl radicals, resulting in low reactivity. The photocatalytic activity of ETS-10 may be applicable to selective transformations of large reactants or less polar reactants to small polar products, enabling highly selective dehalogenation and hydroxylation of aromatics. Zecchina et al. [133] have investigated in detail how to tailor the selectivity of Ti-based photocatalysts (TiO2 and microporous ETS-10 and ETS-4) by modifying the surface morphology and electronic structure. While microporous ETS-4 and ETS-10 exhibit signiﬁcant selectivity in the photodegradation of various molecules using both UV and visible lights, TiO2 (P25) selectivity is observed with visible light only. This means that besides the inverse shape-selectivity effect already observed for the microporous materials [132], selectivity may be achieved also by selecting the excitation light in accordance with the electronic transition of the adsorbed molecule. In such a case, the photodegradation may occur if the conduction band of the Ti-based material is opportunely matched with the lowest unoccupied molecular orbital (LUMO) level of the adsorbed molecule so that it can receive the electron of the excited adsorbate (concept of band alignment). These materials, thus, offer interesting opportunities for selective degradation of some speciﬁc organic compounds in solution, but the practical cases where these properties could be used in a cost-effective way with respect to alternative solutions (for example, removal using selective adsorption, membranes, etc.) are quite limited. Therefore, notwithstanding the large research interest on the use of micro- and mesoporous-material-based photocatalysts for environmental protection, there is no apparent match, at least at the current state of the art, between research effort and potential of application. The elimination of VOCs from gas emissions is another area in which there is a mismatch between research effort and potential of application. However, in this case, there are some examples of commercial application. We limit the discussion here to the use of zeolite (or related materials) as catalysts, not as adsorbents. Zeolites are extremely good adsorbents for many applications involving the adsorption of VOCs from polluted water or air. The main characteristic, besides the high surface area and controlled porosity, is the possibility to tailor the hydrophobic/hydrophilic properties for applications in particular chemical environments. For example, MFI could be functionalized with octamethylsilane to drastically increase the hydrophobicity and greatly enhance the adsorption of aromatics in aqueous solution. Zeolite rotor concentrators are commercially available (for example, by Munters – www.munters.us) for the removal of diluted VOCs from the air stream by adsorption onto the hydrophobic zeolite. After passing through the rotor, the cleaned air is discharged into the atmosphere. The rotor continuously

24.4 Summary and Outlook

rotates at several revolutions per hour, transporting VOC-laden zeolite into the regeneration zone and the regenerated zeolite back into the process zone. In the isolated regeneration section, a small, heated stream of air is drawn through the rotor to desorb the VOCs from the zeolite, forming a highly concentrated VOC-laden air stream. This concentrated stream is normally only 5–10% of the process volume and is typically sent to a small oxidizer. The zeolite operates only as an adsorbent. Although, in principle, it is possible to introduce a catalytic active component (i.e., to use Pd-MFI, for example), the rate of desorption is typically faster than the rate of heating and, thus, it is not possible to avoid the downstream oxidizer (usually a catalytic converter). In terms of commercial use of zeolite or related materials as catalysts in VOC abatement, the cost effectiveness is typically not in favor of these materials. However, their hydrophobic properties (in high-silica materials) could be useful in some cases. Zeochem (www.zeochem.ch), for example, commercializes various strongly hydrophobic high-silica zeolites for the removal of VOCs from air, although their main use is in regenerative adsorber systems. The most common metals introduced in zeolites (typically by ion exchange) to increase VOC oxidation activity are noble metals (Pt and Pd) and some transition metals (Sr, Co, Cu, Fe, Mn, and V), but the activity varies greatly with the type of VOC used. Zeolites offer a better dispersion with respect to metal oxides, and thus, for example, the noble metal loading in zeolites is typically lower than in equivalent metal-oxide-supported materials. This could appear an advantage, but the higher substrate cost and greater difﬁculty in preparing in a technically suitable form (for example, to deposit on a ceramic monolith) make zeolitic catalyst less cost effective in the end. However, in some cases, they show a higher resistance to deactivation than oxide catalysts [134]. An interesting area is chlorinated VOC abatement. The problem is related to the deactivation of protonic zeolites in the catalytic oxidation of these compounds [135]. In general terms, coke formation is the main reason for zeolite catalyst deactivation, but the coke derives from the reaction intermediates (for example, vinyl chloride, which results from a ﬁrst dehydrochlorination step of 1,2-dichloroethane [135]). Strong acid Br¨onsted sites are necessary for the reaction. In Y zeolites, the development of strong acidity due to dealumination leads to a large increase in the activity [136]. Likewise, 50% dealuminated sample showed an improved catalytic behavior for the destruction of other typical chlorinated pollutants, namely, dichloromethane (DCM) and trichloroethylene (TCE). The ease of destruction was found to follow the trend 1,2-dichloroethane > dichloromethane > trichloroethylene [136].

24.4 Summary and Outlook

A large variety of zeolite materials have been studied for use in environmental applications from the conversion of nitrogen oxides over transition-metal-exchange

769

770

24 Environmental Catalysis over Zeolites

zeolites to the conversion of VOCs on noble metal or non-noble-metal containing micro- or mesoporous materials (typically hydrophobic). The presence of a well-deﬁned pore structure and topology provides a unique environment for catalysis, although most of the expectations to have well-deﬁned (enzyme-like) materials have to be reconsidered. Some of the problems and issues in terms of the nature of the active sites and reaction mechanism have been discussed, not with the aim to provide a complete review or to arrive to deﬁnitive conclusions but to present the complexity of the problem and the limitations of some conclusions. In terms of applications, Cu/ and Fe/zeolites (MFI, FER) are the most interesting and relevant, and very soon their large use in the control of emissions of NOx from light-duty diesel engines will appear. A second relevant area concerns their use in the removal of N2 O or combined removal of N2 O and NOx from the tail gas of nitric acid plants. Although already in use, a marked expansion of this application is expected as a consequence of the increasing concerns and regulations on greenhouse gas emissions. More limited is the use of these materials in VOC combustion, although some opportunities in particular related to high-silica zeolite materials with strong hydrophobic properties exist. Lower perspectives are instead suggested for their use as photocatalysts, notwithstanding the large research effort. A general observation is the mismatch between areas of investigation and the more promising areas of applications. Also in the area of nitrogen oxides removal, the largest effort has been done on the identiﬁcation of the active sites and the reaction mechanism, but it should be remarked that there appears to be limited transferability of these fundamental studies to the development of improved catalysts.

References 1. Armor, J.N. (ed.) (1994) Environmental

2.

3.

4.

5.

Catalysis, ACS Symposium Series, Vol. 552, ACS Publications, Washington, DC. Centi, G., Cristiani, P., Forzatti,P., and Perathoner, S. (eds) (1995) Environmental Catalysis, Societa` Chimica Italiana (SCI) Publications, Rome. Centi, G. and Forzatti, P. (eds) (1996) Catal. Today (special issue on Environmental Catalysis), 27 (1–319). Cheng, S. and Wan, B.-Z. (eds) (1995) Catal. Today (special issue on Catalysis in Environmental Applications), 26 (1–96). Misono, M. and Kikuchi, E. (eds) (1996) Pure Appl. Chem. (special issue on Second International Forum on Environmental Catalysis), 68 (357–385).

6. Ruiz, P., Thyrion, F., and Delmon, B.

7. 8.

9.

10.

11.

(eds) (1993) Catal. Today (special issue on Environmental Industrial Catalysis), 17 (1–390). Yahiro, H. and Iwamoto, M. (2001) Appl. Catal. A: Gen., 222, 163. Iwamoto, M., Furukawa, H., Mine, Y., Uemura, F., Mikuriya, S., and Kagawa, S. (1986) J. Chem. Soc., Chem. Commun., 1272. Iwamoto, M., Yahiro, H., Tanda, K., Mizuno, N., Mine, Y., and Kagawa, S. (1991) J. Phys. Chem., 95, 3727. Kustova, M., Kustov, A., Christiansen, S.E., Leth, K.T., Rasmussen, S.B., and Christensen, C.H. (2006) Catal. Commun., 7 (9), 705. Kustova, M., Rasmussen, S.B., Kustov, A.L., and Christensen, C.H. (2006) Appl. Catal. B: Environ., 67, 60.

References 12. Price, G.L. and Kanazirev, V. (1996) US 13. 14.

15. 16. 17.

18. 19.

20.

21. 22. 23. 24. 25. 26.

27. 28.

29.

Patent 5583081. Kagawa, S. and Teraoka, Y. (1992)US Patent 5078981. Bell, A., Centi, G., and Wichterlowa, B. (2001) Catalysis by Unique Ion Structures in Solid Matrices, NATO Science Series II: Mathematics, Physics and Chemistry, Vol. 13, Kluwer/Academic Press (Springer) Publishers, New York. Hidenori, Y. and Iwamoto, M. (2003) Catal. Catal., 45 (1), 26. Shelef, M. (1992) Catal. Lett., 15 (3), 305. Wichtelov´a, B., Dˇedˇecek, J., and Sobal`ık, Z. (2001) Single metal ions is Host zeolite Matricles structure-Activity-selectivity in Catalysis by Unique Ion Structures in Solid Matrices, Chapter 1 (eds A. Bell, G. Centi, and B. Wichterlowa), Kluwer/Academic Press (Springer) Publishers, New York, p. 31. Dˇedˇecek, J. and Wichterlov´a, B. (1999) J. Phys. Chem. B, 103, 1462. Wichtelov´a, B., Sobal`ık, Z., and Dˇedˇecek, J. (2003) Appl. Catal. B: Environ., 41, 97. Beutel, T., S´ark´any, J., Lei, G.-D., Yan, J.Y., and Sachtler, W.M.H. (1996) J. Phys. Chem., 100 (2), 845. Ali, I.O. (2007) Mater. Sci. Eng., 459 (1-2), 294. Moretti, G. (1994) Catal. Lett., 28 (2-4), 143. Centi, G., Nigro, C., and Perathoner, S. (1994) React. Kinet. Catal. Lett., 53, 79. Centi, G., Nigro, C., and Perathoner, S. (1994) Mater. Eng., 5, 223. Moretti, G. (1994) Catal. Lett., 23 (1-2), 135. Teraishi, K., Ishida, M., Irisawa, J., Kume, M., Takahashi, Y., Nakano, T., Nakamura, H., and Miyamoto, A. (1997) J. Phys. Chem. B, 101 (41), 8079. Takaishi, T. and Kato, M. (1995) Zeolites, 15, 689. Sayle, D.C., Catlow, C.R.A., Gale, J.D., Perrin, M.A., and Nortier, P. (1997) J. Phys. Chem. A, 101 (18), 3331. Sayle, D.C., Catlow, C.R.A., Gale, J.D., Perrin, M.A., and Nortier, P. (1997) J. Mater. Chem., 7 (8), 1635.

30. Wichterlov´a, B., Dˇedˇecek, J., Sobal`ık,

31. 32.

33.

34.

35.

36.

37. 38. 39. 40.

41.

42.

43.

Z., Vondrov´a, A., and Klier, K. (1997) J. Catal., 169 (1), 194. Centi, G. and Perathoner, S. (1995) Appl. Catal. A: Gen., 132, 179. Pietrzyk, P. and Sojka, Z. (2007) DFT Modeling an Spectroscopic Investigation into Molecular Aspects of DeNOx Catalysis in Past and Present in DeNOx Catalysis, Chapter 2 (eds P. Grange and V.I. Parvulescu), Elsevier Science Publishers, Amsterdam, p. 27. McMillan, S.A., Broadbelt, L.J., and Snurr, R.Q. (2005) Theoretical modeling of zeolite catalysis: Nitrogen oxide catalysis over metal-exchanged zeolites in Environmental Catalysis, Chapter 12 (ed. V.H.Grassian), CRC Press (Taylor & Francis Group), Boca Raton, p. 287. Schneider, W.F. (2005) Fundamental Concepts in Molecular Simulation of NOx Catalysis in Environmental Catalysis, Chapter 10 (ed. V.H.Grassian), CRC Press (Taylor & Francis Group), Boca Raton, p. 233. Busca, G., Larrubia, M.A., Arrighi, L., and Ramis, G. (2005) Catal. Today, 107-108, 139. Centi, G. and Perathoner, S. (2007) Introduction: State of the Art in the Development of Catalytic Reduction of NOx into N2 in Past and Present in DeNOx catalysis, Chapter 1 (eds P. Grange and V.I. Parvulescu), Elsevier Science Publishers, Amsterdam, p. 1. Pulido, A. and Nachtigall, P. (2009) Phys. Chem. Chem. Phys., 11, 1447. Aylor, A.W., Larsen, S.C., Reimer, J.A., and Bell, A.T. (1995) J. Catal., 157, 592. Li, Y. and Hall, W.K. (1991) J. Catal., 129, 202. Lamberti, C., Bordiga, S., Salvalaggio, M., Spoto, G., Zecchina, A., Geobaldo, F., Vlaic, G., and Bellatreccia, M. (1997) J. Phys. Chem. B, 101, 344. Schay, Z., Kn¨ozinger, H., Guczi, L., and Pal-Borbely, G. (1998) Appl. Catal. B: Environ., 18 (3-4), 263. Ramprasad, R., Hass, K.C., Schneider, W.F., and Adams, J.B. (1997) J. Phys. Chem. B, 101, 6903. Beutel, T., Sarkany, J., Lei, G.D., Yan, J.Y., and Sachtler, W.M.H. (1996) J. Phys. Chem., 10, 845.

771

772

24 Environmental Catalysis over Zeolites 44. Kuroda, Y., Kumashiro, R., Yoshimoto,

45.

46.

47.

48.

49.

50.

51.

52. 53.

54. 55.

56. 57. 58.

T., and Nagao, M. (1999) Phys. Chem. Chem. Phys., 1, 649. Kuroda, Y., Mori, T., Yoshikawa, Y., Kittaka, S., Kumashiro, R., and Nagao, M. (1999) Phys. Chem. Chem. Phys., 1, 3807. Lisi, L., Pirono, R., Ruoppolo, G., and Russo, G. (2008) Kinet. Catal., 49 (3), 421. Rejmak, P., Broclawik, E., Gora-Marek, K., Rado, M., and Datka, J. (2008) J. Phys. Chem. C, 112 (46), 17998. Prestipino, C., Berlier, G., Llabres, I.X., Spoto, G., Bordiga, S., Zecchina, A., Turnes Palomino, G., Yamamoto, T., and Lamberti, C. (2002) Chem. Phys. Lett., 363 (3-4), 389. Shi, Y., Pan, H., Li, Z., Zhang, Y., and Li, W. (2008) Catal. Commun., 9 (6), 1356. Larsen, S.C. (2005) Application of Zeolites in Environmental Catalysis in Environmental Catalysis, Chapter 11 (ed. V.H.Grassian), CRC Press (Taylor & Francis Group), Boca Raton, p. 269. Delahay, G. and Coq, B. (2002) Zeolites for Cleaner Technologies, Catalytic Science Series, Imperial college Press, London, Vol. 3, p. 345. Howe, R.F. (2004) Appl. Catal. A: Gen., 271 (1-2), 3. Choi, B.-C. and Foster, D.E. (2005) J. Ind. Eng. Chem. (Seoul, Republic of Korea), 11 (1), 1. Brosius, R. and Martens, J.A. (2004) Top. Catal., 28 (1-4), 119. Iwamoto, M. and Yahiro, H. (2003) Zeolites in the Science and Technology of Nitrogen Monoxide Removal in Handbook of Zeolite Science and Technology (eds S.M. Auerbach, K.A. Carrado, and P.K. Dutta), Marcel Dekker, Inc., New York, p. 951. Satsuma, A., Shichi, A., and Hattori, T. (2003) CATTECH, 7 (2), 42. Goursot, A., Coq, B., and Fajula, F. (2003) J. Catal., 216 (1-2), 324. Delahay, G., Berthomieu, D., Goursot, A., and Coq, B. (2003) Interfacial Applications in Environmental Engineering, Surfactant Science Series, Vol. 108, Marcel Dekker Inc., New York, p. 1.

59. Berthomieu, D. and Delahay, G. (2006)

Catal. Rev. Sci. Eng., 48 (3), 269. 60. Brandenberger, S., Krocher, O., Tissler,

61.

62.

63.

64. 65.

66.

67.

68.

69. 70. 71. 72.

73.

74.

75.

A., and Althoff, R. (2008) Catal. Rev. Sci. Eng., 50 (4), 492. Matsuoka, M. and Anpo, M. (2003) J. Photochem. Photobiol. C: Photochem. Rev., 3 (3), 225. Calzaferri, G., Leiggener, C., Glaus, S., Schuerch, D., and Kuge, K. (2003) Chem. Soc. Rev., 32 (1), 29. Anpo, M. and Matsuoka, M. (2008) Recent Advances in Single-site Photocatalysts Constructed within Microporous and Mesoporous Materials in Turning Points in Solid-State, Materials and Surface Science (eds K.D.M. Harris and P.P. Edwards), Royal Society of Chemistry Publisher, Cambridge, p. 492. Yamashita, H. and Mori, K. (2007) Chem. Lett., 36 (3), 348. Aprile, C., Corma, A., and Garcia, H. (2008) Phys. Chem. Chem. Phys., 10 (6), 769. Hernandez-Ramirez, O. and Holmes, S.M. (2008) J. Mater. Chem., 18 (24), 2751. Apreutesei, R.E., Catrinescu, C., and Teodosiu, C. (2008) Environ. Eng. Manage. J., 7 (2), 149. Widiastuti, N., Wu, H., Ang, M., and Zhang, D.-k. (2008) Desalination, 218 (1-3), 271. Perathoner, S. and Centi, G. (2005) Top. Catal., 33 (1-4), 207. Coronas, J. and Santamaria, J. (2004) Chem. Eng. Sci., 59 (22-23), 4879. Unger, B., Brandt, A., and Tschritter, H. (2001) Chem. Ing. Tech., 73 (6), 723. Hevia, M.A.G. and Perez-Ramirez, J. (2008) Appl. Catal. B: Environ., 77 (3-4), 248. Smeets, P.J., Groothaert, M.H., van Teeffelen, R.M., Leeman, H., Hensen, E.J.M., and Schoonheydt, R.A. (2007) J. Catal., 245 (2), 358. Tabata, T., Kokitsu, M., Othsuka, H., Okada, T., Sabatino, L.F.M., and Bellussi, G. (1996) Catal. Today, 27, 91. Othsuka, H., Tabata, T., Okada, T., Sabatino, L.F.M., and Bellussi, G. (1997) Catal. Lett., 44, 265.

References 76. Dedecek, J., Capek, L., and Wichterlova,

77.

78.

79.

80. 81.

82.

83.

84.

85.

86. 87.

88.

89. 90.

B. (2006) Appl. Catal. A: Gen., 307 (1), 156. Krocher, O. (2007) Past and Present in DeNOx Catalysis, Studies in Surface Science and Catalysis, Vol. 171, Elsevier, Amsterdam, p. 261. Shibata, J., Takada, Y., Shichi, A., Satokawa, S., Satsuma, A., and Hattori, T. (2004) Appl. Catal. B: Environ., 54 (3), 137. Yeom, Y., Li, M., Savara, A., Sachtler, W., and Weitz, E. (2008) Catal. Today, 136 (1-2), 55. Zhu, R., Guo, M., and Ouyang, F. (2008) Catal. Today, 139 (1-2), 146. Nakatsuji, T., Matsubara, M., Rouistenmaeki, J., Sato, N., and Ohno, H. (2007) Appl. Catal. B: Environ., 77 (1-2), 190. Ciambelli, P., Sannino, D., Palo, E., and Ruggiero, A. (2007) Top. Catal., 42/43, 177. Liu, S., Kong, L., Yan, X., Li, Q., and He, A. (2005) Nanoporous Materials IV, Studies in Surface Science and Catalysis, Vol. 156, Elsevier, Amsterdam, p. 379. Silva, E., Catalao, R., Silva, J., Vaz, F., Oliveira, F., Ribeiro, F.R., Magnoux, P., Belin, T., and Ribeiro, F. (2008) Zeolites and Related Materials, Studies in Surface Science and Catalysis, Elsevier, Amsterdam Vol. 174B, p. 1195. Zhou, J., Zhao, L., Huang, Q., Zhou, R., and Li, X. (2009) Catal. Lett., 127 (3-4), 277. Li, D., Zheng, Y., and Wang, X. (2007) Catal. Commun., 8 (11), 1583. Kwong, C.W., Chao, C.Y.H., Hui, K.S., and Wan, M.P. (2008) Environ. Sci. Technol., 42 (22), 8504. Tidahy, H.L., Siffert, S., Lamonier, J.-F., Cousin, R., Zhilinskaya, E.A., Aboukais, A., Su, B.-L., Canet, X., De Weireld, G., Frere, M., Giraudon, J.-M., and Leclercq, G. (2007) Appl. Catal. B: Environ., 70 (1-4), 377. Zrncevic, S. and Gomzi, Z. (2005) Ind. Eng. Chem. Res., 44 (16), 6110. Centi, G., and Perathoner, S. (2005) Use of solid catalysts in promoting water treatment and remediation technologies, Catalysis, Vol. 18, J.J. Spirey Ed., Royal

91.

92.

93.

94. 95.

96. 97.

98.

99.

100.

101.

102.

103.

104.

Society of chemistry pub., Cambridge U.K., 46–71. Giordano, G., Perathoner, S., Centi, G., De Rosa, S., Granato, T., Katovic, A., Siciliano, A., Tagarelli, A., and Tripicchio, F. (2007) Catal. Today, 124 (3-4), 240. Kasiri, M.B., Aleboyeh, H., and Aleboyeh, A. (2008) Appl. Catal. B: Environ., 84 (1-2), 9. Colella, C. (1999) Porous Materials in Environmentally Friendly Processes, Studies in Surface Science and Catalysis, Vol. 125, Elsevier, Amsterdam, p. 641. Larsen, S.C. (2007) J. Phys. Chem. C, 111 (50), 18464. Anpo, M. and Thomas, J.M. (2006) Chem. Commun. (Cambridge, UK), (31), 3273. Corma, A. and Garcia, H. (2004) Chem. Commun. (Cambridge, UK), (13), 1443. Wu, Z.Y., Wang, H.J., Ma, L.L., Xue, J., and Zhu, J.H. (2008) Microporous Mesoporous Mater., 109 (1-3), 436. Yang, J., Zhou, Y., Wang, H.J., Zhuang, T.T., Cao, Y., Yun, Z.Y., Yu, Q., and Zhu, H. (2008) J. Phys. Chem. C, 112 (17), 6740. Makhotkina, O.A., Preis, S.V., and Parkhomchuk, E.V. (2008) Appl. Catal. B: Environ., 84 (3-4), 821. De la Rosa-Gomez, I., Olguin, M.T., and Alcantara, D. (2008) J. Environ. Manage., 88 (4), 853. Girard, J., Cavataio, G., Snow, R., and Lambert, C. (2008)SP-2154(Diesel Exhaust Emission Control), Society of Automotive Engineers Publisher, p. 523. Montreuil, C. and Lambert, C. (2008)SP-2154(Diesel Exhaust Emission Control), Society of Automotive Engineers Publisher, p. 497. Cavataio, G., Jen, H.-W., Warner, J.R., Girard, J.W., Kim, J.Y., and Lambert, C.K. (2008)SP-2154(Diesel Exhaust Emission Control), Society of Automotive Engineers Publisher, p. 451. Cheng, Y., Montreuil, C., Cavataio, G., and Lambert, C. (2008) SP-2154(Diesel Exhaust Emission Control), Society of Automotive Engineers Publisher, p. 437.

773

774

24 Environmental Catalysis over Zeolites 105. Muench, J., Leppelt, R., and Dotzel, R.

106.

107.

108.

109. 110.

111.

112.

113.

114.

115.

116.

117. 118. 119.

120.

(2008)SP-2154(Diesel Exhaust Emission Control), Society of Automotive Engineers Publisher, p. 443. Centi, G., Perathoner, S., and Vazzana, F. (1999) Chem. Tech., 29 (12), 48. Schwefer, M., Groves, M., and Maurer, R. (2001) Chem. Ing. Tech., 73 (6), 603. Groves, M., Maurer, R., Schwefer, M., and Siefert, R. (2006) Abatement of N2O and NOX emissions from nitric acid plants with the Uhde EnviNOx process. presented at NITROGEN 2006 International Conference Vienna, Austria – March 14th, 2006. Centi, G. and Vazzana, F. (1999) Catal. Today, 53, 683. Hevia, M.A.G. and P´erez-Ram´ırez, J. (2008) Appl. Catal. B: Environ., 77 (3-4), 24. Morrill, M. (2007) New s¨ud-chemie off-gas catalysts for the reduction of NO and N2O from nitric acid plants. presented at Chem Show, New York City, October 31, 2007. Hevia, M.A.G. and Perez-Ramirez, J. (2008) Environ. Sci. Technol., 42 (23), 8896. J´ısˇ a, K., Nov´akov´a, J., Schwarze, M., Vondrov´a, A., Sklen´ak, S., and Sobalik, Z. (2009) J. Catal., 262, 27. Kauck´y, D., Dˇedˇecek, J., and Wichterlov´a, B. (1999) Microporous Mesoporous Mater., 31, 75. P´erez-Ram´ırez, J., Groen, J.C., Br¨uckner, A., Kumar, M.S., Bentrup, U., Debbagh, M.N., and Villaescusa, L.A. (2005) J. Catal., 232, 318. Parmon, V.N., Panov, G.I., Uriarte, A., and Noskov, S. (2005) Catal. Today, 100, 115. Hansen, N., Heyden, A., Bell, A.T., and Keil, F.J. (2007) J. Catal., 248, 213. Pirngruber, G.D., Roy, P.K., and Prins, R. (2007) J. Catal., 246, 147. Roy, P.K., Prins, R., and Pirngruber, G.D. (2008) Appl. Catal. B: Environ., 80, 226. Meli´an-Cabrera, I., Mentruit, C., Pieterse, J.A.Z., van den Brink, R.W., Mul, G., Kapteijn, F., and Moulijn, J.A. (2005) Catal. Commun., 6, 301.

121. Oygarden, A.H. and P´erez-Ram´ırez, J.P.

(2006) Appl. Catal. B: Environ., 65, 163. 122. Nov´akov´a, J. and Sobal´ık, Z. (2007)

Curr. Top. Catal., 6, 55. 123. Pirngruber, G.D. (2009) The fascination

124.

125.

126.

127.

128. 129.

130. 131. 132.

133.

134.

135.

136.

chemistry of iron- and coppercontaining zeolites in Ordered Porous Solids: Recent Advances and Prospects (eds V.Valtchev, S. Mintova, and M. Tsapatsis), Elsevier S&T, Oxford, p. 749. Centi, G., Perathoner, S., Pino, F., Arrigo, R., Giordano, G., Katovic, A., and Pedul`a, V. (2005) Catal. Today, 110, 211. Caudo, S., Centi, G., Genovese, C., and Perathoner, S. (2006) Top. Catal., 40 (1-4), 207. Yang, K., Wang, B., Chen, L., and Wang, X. (2008) Catal. Commun., 9 (3), 431. Kanyi, C.W., Doetschman, D.C., Yang, S.-W., Schulte, J., and Jones, B.R. (2008) Microporous Mesoporous Mater., 108 (1-3), 103. Chretien, M.N. (2007) Pure Appl. Chem., 79 (1), 1. Yamashita, H. and Anpo, M. (2003) Curr. Opin. Solid State Mater. Sci., 7 (6), 471. Corma, A. and Garcia, H. (2004) Eur. J. Inorg. Chem., Vol. 2004 (6), 1143. Shiraishi, Y., Tsukamoto, D., and Hirai, T. (2008) Langmuir, 24 (21), 12658. Llabr´es i Xamena, F., Calza, P., Lamberti, C., Prestipino, C., Damin, A., Bordiga, S., Pelizzetti, E., and Zecchina, A. (2003) J. Am. Chem. Soc., 125 (8), 2264. Usseglio, S., Calza, P., Damin, A., Minero, C., Bordiga, S., Lamberti, C., Pelizzetti, E., and Zecchina, A. (2006) Chem. Mater., 18 (15), 3412. Niu, G., Huang, Y., Chen, X., He, J., Liu, Y., and He, A. (1999) Appl. Catal. B: Environ., 21, 63. Aranzabal, A., Gonz´alez-Marcos, J.A., Romero-S´aez, M., Gonz´alez-Velasco, J.R., Guillemot, M., and Magnoux, P. (2009) Appl. Catal. B: Environ., 88 (3-4), 53. L´opez-Fonseca, R., de Rivas, B., Guti´errez-Ortiz, J.I., Aranzabal, A., and Gonz´alez-Velasco, J.R. (2003) Appl. Catal. B: Environ., 41 (1-2), 31.

775

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals Maria J. Climent, Avelino Corma, and Sara Iborra

25.1 Introduction

Zeolites are crystalline microporous materials that are widely used in reﬁning and for the production of chemicals and petrochemicals [1]. The merits of these materials as solid acid or basic catalysts have been extensively discussed in the literature. Particularly, acid zeolites have been widely used as catalysts in reﬁning and petrochemical industries because of their excellent catalytic behavior, allowing the replacement of hazardous acids, the reduction of salts and other waste products, and prevention of plant corrosion. In the past years, the amount of work on the use of acidic zeolites as catalysts for the synthesis of chemical intermediates and ﬁne chemicals has increased considerably [2–6]. The variety of pore topologies and pore dimensions within the zeolite materials along with the possibility to tune their acidity (or basicity) and regenerate them make these materials attractive heterogeneous catalysts for the synthesis of ﬁne chemicals. In this chapter, we present a number of relevant examples in the ﬁelds of application where the use of zeolites as solid catalysts has been shown to be particularly useful. 25.2 Acid-Catalyzed Reactions 25.2.1 Friedel–Crafts Acylation

Aromatic hydrocarbons play an important role as intermediates and products in the ﬁne chemicals industry. It should be considered that although the aromatic hydrocarbons used in this ﬁeld represent only 2–3% of the overall molecules produced, they include very important substances widely applied in the pharmaceutical, agriculture, and ﬂavor and fragrance industries. In this section, we discuss Friedel–Crafts acylation reactions, in which zeolites have been extensively used as acid catalysts to synthesize aromatic hydrocarbons. Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

776

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

COCH3 + CH3–COCl

Acid

COCH3 OCH3

OCH3 Scheme 25.1

+

Catalyst

OCH3

Anisole acetylation.

The ﬁrst example of aromatic acylation reaction using zeolites was reported by Venuto and Landis [2] in 1968, when acidic faujasites were used as acid catalysts with moderate success. Twenty years later, Chiche et al. [7] reported the acylation of toluene with linear carboxylic acids (C2 –C22 ) using exchanged Ce+3 NaY zeolite. The most remarkable results were related to the high p-selectivity (94%) and the increase in the yield of arylketone when increasing the chain length. Since then, zeolites have been used as heterogeneous catalysts for a wide variety of substrates using carboxylic acids, anhydrides, and acid chlorides as acylating agents [8–10]. Anisole acetylation has been one of the most studied acid-catalyzed reactions owing to the importance of p-methoxyacetophenone as intermediate in the manufacture of a variety of ﬁne chemicals [11–24] (Scheme 25.1). Among the different solid acids tested for this reaction including sulfonic resins, amorphous oxides, clays and zeolites, the order of activity found for this Friedel–Crafts acylation is amorphous oxides < clays < zeolites ≤ acid resins, and within the zeolite catalysts, the order of activity reported has been HZSM5 ≈ HMor > HY > H-Beta, HY, and particularly H-Beta being the most studied catalysts from the point of view of their interesting shape selectivity. Concerning the acetylating agents, acetyl chloride and acetic anhydride emerged as the most effective, giving the highest conversions and p-selectivities. Total conversion of anisole with 98% selectivity to the p-methoxyacetophenone was reported by Smith et al. [21] using H-Beta zeolite working with a moderate excess of acetic anhydride at 120 ◦ C. In this case, the high selectivity to the p-isomer was probably due to the effect of zeolite shape selectivity [25]. Formation of phenol and cresol from demethylation and rearrangement of the methoxy group, respectively, was not obtained in any case [16, 25]. It was observed that, using acetic anhydride as acylating agent, the Si/Al ratio of Beta and Y zeolites have a little inﬂuence on catalytic activity [19, 21]. However, when acetyl and phenylacetyl chloride were used as acylating agents, the rate of reaction decreased when increasing the Al content, which was attributed to the predominance of sorption effects due to the density of acid sites [17, 25]. Moreover, it was found that during the acylation of anisole with phenylacetyl chloride, the rate of formation of the arylketone correlated linearly with the amount of Na+ exchanged in different HNaY samples, indicating that all acid sites, regardless of their acid strength, are active in the reaction and, therefore, an acid catalyst does not need to have very strong acidity to exhibit high activity for this reaction when using acyl chlorides as the acylating agent [25]. With respect to catalyst life, deactivation studies of the zeolite during the acylation of anisole [15, 20] revealed that the main deactivation process, which is reversible, is the

25.2 Acid-Catalyzed Reactions

777

strong adsorption of p-methoxyacetophenone due to its larger molecular size and high polarity. Then, adsorption of p-methoxyacetophenone on the catalyst surface could be minimized by operating with a ﬂow reactor by using anisole-rich feeds. Product desorption can easily be carried out by washing the exhausted catalyst by a solvent. The Rhodia Company has developed an industrial process for acylation of anisole with acetic anhydride using HBEA zeolite in a ﬁxed-bed reactor. This process enables considerable simpliﬁcation, and it is not only environmentally friendly but also economically more sustainable than the older process based on the use of AlCl3 as Lewis acid catalyst [22]. Rhodia has extended the process to the synthesis of acetoveratrole, (3,4-dimethoxyacetophenone) (Scheme 25.2), which is an intermediate for the synthesis of papaverine, using HFAU zeolite as catalyst [18, 22]. Recently, the synthesis of acetoveratrole by acetylation of veratrole with acetic anhydride over three-dimensional aluminosilicate of the Al-KIT-5 type as well as hexagonal Al-SBA-15 has been reported [26, 27]. Conversion of acetic anhydride of 92% with 100% selectivity to acetoveratrole was achieved over Al-SBA-15, working at 60 ◦ C. Another acylation reaction extensively studied with zeolites is the acetylation of 2-methoxynaphthalene (2-MN) [18] (Scheme 25.3). This acylation reaction can produce different positional isomers in different ratios depending on the reaction conditions. 1-Acetyl-2-methoxynaphthalene is kinetically preferred, whereas 2,6and 2,8-isomers are favored by thermodynamic factors and constrained environments. Selective formation of 2-acetyl-6-methoxynaphathalene (2-AMN) is desired because this compound is an intermediate for the synthesis of an important anti-inﬂammatory drug naproxen (2-(6-methoxy-2-naphthyl) propionic acid). In contrast to the anisole acetylation, the selectivity of 2-MN acylation is strongly dependent on the zeolite or mesoporous molecular sieve employed as catalyst. COCH3 + (CH3CO)2O

HY

OCH3

OCH3

OCH3

OCH3 Acetoveratrole

Scheme 25.2

Synthesis of acetoveratrole.

COCH3

OCH3

OCH3

OCH3

+ CH3CO 2-MN

1-AMN

Scheme 25.3 Different possible isomers obtained in the acetylation of 2-methoxynaphthalene (2-MN).

COCH3 OCH3

+ 2-AMN

8-AMN

778

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

Because of the high melting point of the substrate and products, the acetylation of 2-MN is carried out in the presence of solvents, acetic anhydride being the preferred acylating agent. Because of the higher molecular size of 2-MN and 2-AMN, the reaction has been preferably performed over large pore zeolites: HY, HMor, H-Beta, and ITQ-7 [28–35], Beta zeolite being the most investigated. With this molecular sieve, the selectivity to 2-AMN is in general higher than with other large pore zeolites due to the pore dimensions. Botella et al. [36] showed that when using the polymorph C of Beta (ITQ-17) zeolite, which has a slightly smaller diameter than the Beta counterpart, a decrease in activity occurs when working in either batch or ﬁxed-bed continuous reactors. However, the selectivity to 2-AMN was higher with ITQ-17. As expected, the selectivity to 2-AMN could be increased by passivation of the external surface of zeolite Beta by a silylation process [28]. 2,6-Naphthalenedicarboxylic acid (2,6-NDCA) is an important intermediate for the synthesis of pharmaceuticals, agricultural chemicals, dyes, liquid crystals, and ﬂuorescents, and it is also used as the monomer for the preparation of the polyester polyethylene naphthalate (PEN). The preferred method for preparing 2,6-NDCA is by the oxidation of 2,6-dialkylnaphthalene (2,6-DAN) or 2-methyl-6-acylnaphthalene (2,6-AMN). Friedel–Crafts acylation of 2-methylnaphthalene (2-MeN) is considered the most promising way for preparing 2,6-NDCA. Recently, Yuan et al. [37] reported on the acylation of 2-MeN using various acylating agents over different zeolites (HY, Beta, HZSM-5, HMor, HUSY) in the absence of the solvent. While HY, HZSM-5, HMor, and HUSY showed no catalytic activity for the acetylation of 2-MeN with acetic anhydride at 140 ◦ C, 2-methyl-6-butyrylnaphthalene could be obtained by the acylation of 2-MeN with butyric anhydride in the presence of H-Beta. Under optimized reaction conditions, conversion of butyric anhydride of 78.3% with 53.1% selectivity to 2-methyl-6-butyrylnaphthalene was obtained at 190 ◦ C. Zeolites have also been used for the selective acylation of heterocyclic compounds. Pyrrole and thiophene are selectively acylated with acetic anhydride at the 2-position under gas-phase conditions with the mildly acidic B-ZSM-5 [38]. Also, 2-methylfuran and benzofuran are selectively acetylated with acetic anhydride at the 3- and 2-positions using HY zeolite [39]. Working in a ﬁxed-bed reactor, fast deactivation of the catalyst was observed in the case of benzofuran, whose conversion fell from 95 to 20%. On the other hand, faster kinetics and slower deactivation were observed when reacting 2-methylfuran and yields >95% were achieved in 2 hours. The small tendency of the 2-methylfuran to polymerize could be the reason for the lower catalyst decay observed. Thiophenes, pyrroles, and furans have also been acylated with acetic anhydride in the liquid phase under mild reaction conditions using modiﬁed and unmodiﬁed zeolite H-Beta. High yields, sometimes in very short reaction times (10 minutes), were achieved. Indium-modiﬁed catalysts have provided a way for attaining good yields during the acylation of pyrrole [40]. Acylation of thiophene with acyl chlorides has been described in the gas phase [38, 41] and in the liquid phase [42, 43] producing selectively 2-acetylthiophene.

25.2 Acid-Catalyzed Reactions

COCH3 + Scheme 25.4

(CH3CO)2O

Acetylation of isobutylbenzene.

p-Acetylisobutylbenzene (Scheme 25.4), an intermediate for the synthesis of the anti-inﬂammatory drug ibuprofen, can be prepared by acetylation of isobutylbenzene with acetic anhydride as acylating agent using H+ , Ce+3 , Zn+2 , and La+3 exchanged zeolites with good p-selectivity. For instance, acylation of isobutylbenzene with Ce+3 -Beta exchanged zeolite at 130 ◦ C gave 30% p-acetylisobutylbenzene, while better results (80% yield) were achieved with H-Beta zeolite [24, 44] working at 140 ◦ C after 1 hour reaction time. Phenol acylation is another interesting reaction, as p-hydroxyacetophenol is used for the synthesis of paracetamol (para-acetaminofenol) while ortho-hydroxy acetophenol is a key intermediate for the synthesis of p-hydroxycumarin and warfarin, which are used as anticoagulant drugs. Gas-phase acylation of phenol with acetic anhydride was performed on different mesoporous molecular sieves and HY and HZSM-5 zeolites [45, 46]. Both zeolites were active for the acylation, giving mainly the ortho-hydroxyacetophenone via the direct C-acylation of phenol as well as by the acylation of phenylacetate intermediate formed from O-acylation of phenol. HZSM-5 showed excellent stability against deactivation by coke formation, due to the limitations imposed by the pore dimensions to the formation of coke precursors. The synthesis of o- and p-hydroxyacetophenones involving the esteriﬁcation of phenol with acetic acid to give phenylacetate which subsequently undergoes the Fries rearrangement to give the o- and p-hydroxyacetophenones has been studied with Y, Beta, and ZSM5 zeolites [47] (Scheme 25.5). Reactions were performed using phenol and acetic acid or acetic anhydride as reagents in liquid and gas phases. The authors found that in the liquid phase (under reﬂuxing conditions) the reaction yields exclusively phenyl acetate (26–80%). However, in gas phase when acetic acid is used as acylating agent, zeolite catalysts form phenyl acetate followed by the Fries rearrangement to yield selectively o-hydroxyacetophenone (40%), the o-hydroxyacetophone/phenyl acetate molar ratio being very high with ZSM5 catalyst due to a shape-selectivity effect. Benzoylation of resorcinol leads to 2,4-dihydroxybenzophenone (Scheme 25.6), which is a precursor of a class of UV-absorbers obtained by the 4-OH etheriﬁcation. van Bekkum et al. [48, 49] performed the catalytic acylation of resorcinol with OCOCH3

OH +

OH

OH

CH3COOH or CH3COCl

COCH3 + COCH3

Scheme 25.5

Products obtained in the acetylation of phenol.

779

780

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

OH

COOH

OH HO

+ OH

OC OH

O

O

2,4-Dihydroxybenzophenone Scheme 25.6

Synthesis of 2,4-dihydroxybenzophenone by benzoylation of resorcinol.

benzoic acid in liquid phase on strong solid catalysts, such as sulfonic resins and zeolites. The reaction appears to proceed via ester formation and subsequent Fries rearrangement, although the direct C-acylation could not be excluded. High yield and selectivity to of 2,4-dihydroxybenzophenone were obtained on Beta zeolite using high boiling point solvents (n-butylbenzene, 4-chlorotoluene), with azeotropic removal of water and relatively large amounts of catalyst (about 30%). 25.2.2 Hydroxyalkylation of Aromatic Compounds

Hydroxyalkylation of aromatics with carbonyl compounds is an important reaction because the functionalization of the aromatic ring gives a wide range of compounds of great interest as ﬁne chemicals (Scheme 25.7). However, this in not an easy reaction to perform since the mild electrophilic character of aldehydes and ketones requires the aromatic ring to be activated; moreover, the hydroxyalkylation leads to a variety of positional isomers (ortho and para) and, ﬁnally, undesirable consecutive reactions are often observed (Scheme 25.8). Pioneering work on heterogeneous aromatic hydroxyalkylation was reported by Venuto and Landis in 1966 [50]. They performed the hydroxyalkylation of phenol with different carbonyl compounds on HY zeolite at 180 ◦ C. The main product obtained was the corresponding bis-arylmethane adduct, and the yields and regioselectivity depended on the type of carbonyl compound involved in the reaction. In the 1980s, Climent et al. [51–53] and then Burgers et al. [54] reported the hydroyalkylation of phenol, toluene, and anisole with formaldehyde [51], acetaldehyde [52], benzaldehyde, and acetophenone [53] using a variety of zeolites as catalysts. In all cases, the main product obtained was the bis-arylmethane adduct. In addition, for these reactions the hydrophobicity of the catalyst surface R

R

OH

O +

R2

R

R1

R1

R2

+

R2 Scheme 25.7

OH

R1

Hydroxyalkylation of aromatics with carbonyl compounds.

25.2 Acid-Catalyzed Reactions

R R +

O R1

O R1 OH R1 R 2 R2

R2

R

R2 OH

R

R R

R

R2

R1

OH

R2

R1

R=Donating group

R1

R1

OH

R2

Scheme 25.8 Different undesirable consecutive reactions taking place in aromatic hydroxyalkylations.

(working with hydrophobic HY zeolite) on the sorption/desorption of substrates and consequently on the kinetic and catalyst decay [51] plays a determinant role. Another interesting example is the hydroxymethylation of guaiacol with formaldehyde to obtain para-hydroxymethylguaiacol, the precursor for vanillin, which is an important food additive (Scheme 25.9). Moreau et al. [55] performed this reaction using H-mordenite. The best selectivity to the para isomer (83%) at 33% conversion was achieved on well-balanced acidic and hydrophobic properties (Si/Al = 18) of the catalyst working at 40 ◦ C. Recently, Cavani et al. [56–58] have reported the synthesis p-vanillin alcohol by hydroxymethylation of guaiacol with an aqueous solution of formaldehyde at 80 ◦ C in the presence of mordenite as catalyst. The results evidence that the methanol present in the commercial aqueous solution of formaldehyde not only decreases the conversion of guaiacol but also favors the formation of mono-aryl by-products with respect to the diaryl compounds. On the other hand, a study over dealuminated zeolites with Si/Al ratios between 10 and 58 showed that the zeolite property mainly affecting catalytic performance is hydrophobicity. Thus, it was observed that an increase in the concentration of acid sites corresponds to a decrease in catalytic activity. However, the results for the hydroxymethylation of guaiacol in CH2OH OH OH

OR

+ HCOH OCH3 Guaiacol

OH OHCH2

OCH3 CHO

p -Vanillin alcohol Scheme 25.9

OH

ox

OCH3 Vanillin

Hydroxymethylation of guaiacol with formaldehyde.

781

782

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

the presence of mordenite with different Si/Al ratios show that the formation of mesoporous secondary structure during dealumination overlapped with the effect of the hydrophobicity in the more dealuminated catalyst, increasing the selectivity to diaryl by-products in comparison to mordenites having intermediate Si/Al ratios. Thus, it is possible to obtain conversions approximately of 20 and 60% selectivity to p-vanillin alcohol for samples with Si/Al ratios between 12 and 25, whereas for samples with higher Si/Al ratios (30–35) the selectivity to p-vanillin alcohol is less than 40% and the diaryl by-products account for 60%. Allylation of guaiacol [59] is also a commercially important process that gives products such as eugenol, o-eugenol, and chavitol used in perfumeries, ﬂavorings, essential oils, insect attractants, and in medicines as local antiseptic, analgesic, and UV light absorbers. The allylation of guaiacol with allyl alcohol using zeolites (HY, H-Beta, ZSM-5, mordenite) as catalysts yields ﬁve isomers of monoallylated products: allyl-2-methoxyphenylether, eugenol (4-allyl-2-methoxyphenol), o-eugenol (2-allyl-6-methoxyphenol), clavibetol (5-allyl-2-metoxyphenol), 3-allyl-2-metoxyphenol, and a small amount of diallyl guaiacol (Scheme 25.10). Conversions of guaiacol between 4 and 61% and selectivities to monoallyl guaiacol in the range of 47–94% after 2 hours reaction time were obtained. K-10 clay presented higher activity compared to zeolites (61% conversion); however, the selectivity to monoallyl guaiacol was poor (47%). Among zeolites, HY gave maximum guaiacol conversion (46%) with selectivity to monoallyl guaiacol (85%), diallyl guaiacol (9%), and others (6%). Only a small decrease in guaiacol conversion was detected after three cycles of reaction. OH

OCH2CH=CH2

OH OCH3

OCH3

OCH3 CH2=CHCH2OH

+

Catalyst

CH2CH=CH2 Eugenol OH

Allyl-2-methoxyphenylether OH H2C=HCH2C

OCH3

o-Eugenol

OCH3

+

H2C=HCH2C

Chavibetol OH

OH OCH3 CH2CH=CH2 3-Allyl-2-methoxyphenolD Scheme 25.10

Allylation of guaiacol.

OCH3

+ H2C=HCH2C

CH2CH=CH2

Diallyl guaiacol

25.2 Acid-Catalyzed Reactions

OH O Furfuryl alcohol Scheme 25.11

HCOH H+

OH

HO O

2,5-Bis-(hydroxymethyl)-furan Synthesis of 2,5-bis(hydroxymethyl) furan.

2,5-Bis(hydroxymethyl) furan is a valuable product in the furan family useful as intermediate in the synthesis of drugs, crown ethers, and polymers. This compound can be synthesized through the hydroxymethylation of furfuryl alcohol with formaldehyde (Scheme 25.11). Lecomte et al. [60–63] performed the hydroxymethylation of furfuryl alcohol with aqueous formaldehyde using a hydrophobic mordenite with a Si/Al ratio of 100 as acid catalyst at low temperatures (40–50 ◦ C). Selectivities in 2,5-bis(hydroxymethyl) furan equal to or higher than 95% for conversions of furfuryl alcohol of 80–90% were achieved. The same authors reported later the hydroxyethylation of furfuryl alcohol with acetaldehyde under similar reaction conditions, although the yield and selectivity in 1-(5-hydroxymethyl-2-furyl)-ethanol were substantially lower [64]. 25.2.3 Diels–Alder Reactions

Diels–Alder reaction between an oleﬁn (a dienophile) and a conjugated 1,3-diene has been widely utilized for the synthesis of six-membered carbon skeletons of ﬁne chemicals and pharmaceuticals [65], and is an ideal reaction from the point of view of atom economy. The reaction can be highly regio- and stereoselective and, when catalyzed by chiral Lewis acids, can, moreover, proceed with high enantioselectivity [66]. Although many Diels–Alder reactions do not require the use of catalysts and can be performed at low-to-moderate temperatures, when less reactive or thermally unstable reactants or products are involved, the use of catalytic methods allow carrying out the reaction under mild conditions. Traditional catalysts for Diels–Alder reactions are Lewis acids such as AlCl3 and BF3 etherate, which are often used in stoichiometric amounts, and result in environmentally hazardous waste-streams. Zeolites have been used for catalyzing Diels–Alder reactions for the synthesis of ﬁne chemicals, and already in 1968 Landis [67] reported the reaction between butadiene and maleic anhydride using X zeolite containing rare-earth-metal cations and/or protonic acid sites. Excellent yield (93%) of the Diels–Alder adduct is achieved after 3 hours at 60 ◦ C. The use of furane derivatives as dienophiles is important, as they are interesting building blocks in the synthesis of biologically active compounds; however, furan is acid-sensitive and the Diels–Alder reaction of this diene using the conventional homogeneous acid catalyst leads generally to low yields. Ipaktschi [68] reported that Cu+ -exchanged Y zeolite was very effective in catalyzing the Diels–Alder reaction of furan with dienophiles such as acrolein or methyl vinyl ketone at 0 ◦ C, achieving adduct yields of 31 and 73%,

783

784

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

O

O

O

CuY 0 °C

+

73%, Exo/endo = 2.5 Scheme 25.12

Diels–Alder reaction of furan with methyl vinyl ketone.

respectively (Scheme 25.12). Similar results were reported by Narayana et al. [69] using Ce-exchanged Y zeolite, while Zn–Y zeolite is a less active and stereoselective catalyst. Pindur et al. [70] reported that highly activated 4-A molecular sieves catalyze with moderate activity the Diels–Alder cycloaddition of N-benzoyl-2,3-dihydro-2,3-bis (methylene)indole with different carbodienophiles at 50–55 ◦ C (Scheme 25.13). Although these adducts were difﬁcult to obtain by thermal reaction, the molecular sieve gives yields between 10 and 63%. Zeolites in the protonic form have been used as catalyst for the hetero Diels–Alder reaction of dihydropyran and acrolein (Scheme 25.14) at 0 ◦ C in the absence of a solvent [71]. Among the different zeolites tested, the best catalysts were an H-Beta sample (Si/Al = 25) and a dealuminated HY (Si/Al = 15) sample, yielding the desired adduct in 65 and 62%, respectively. The results are interesting when comparing with the thermal reaction (at 150 ◦ C) where only a 5% of the Diels–Alder adduct was obtained.

R1

R1

R2

R2

4A molecular sieves

+

50−55 °C

N

R3

N

R4

COPh

R3

R4

COPh

Scheme 25.13 Diels–Alder reaction of N-benzoyl-2,3-dihydro-2,3-bis(methylene)indole with different dienophiles.

H Trans-isomer + regioisomer dimers and trimers

H-Zeolite 0 °C O

Scheme 25.14

O

O

O H cis-1,8-Dioxaoctahydronaphtalene

Diels–Alder reaction of dihydropyran and acrolein.

25.2 Acid-Catalyzed Reactions

O H

+

CHO +

Myrcene Scheme 25.15

NaY/Cl2Zn

“para” Myrac aldehyde

CHO “meta” Myrac aldehyde

Synthesis of Myrac aldehyde.

Yin et al. [72] reported that zeolite HY modiﬁed with anhydrous ZnCl2 can be used as catalyst in the Diels–Alder reaction of myrcene and acrolein, yielding para and meta myrac aldehydes (Scheme 25.15). These compounds are used in perfumery, and particularly the para cycloadduct (4-(4 -methyl-3 -pentyl)-3 -cyclohexenecarboxaldehyde) is the most important, since their subsequent cyclization gives products with a pleasant woody and fruity ambergris smell widely utilized in perfume and cosmetic compositions. Therefore, improving the regioselectivity of the para cycloadduct is of considerable practical interest. Thus, recently the same authors [73] reported that the use of ZnCl2 supported on NaY zeolite prepared by solid-state interaction under microwave irradiation gives a highly active and regioselective catalyst for the Diels–Alder reaction between myrcene and acrolein. Reactions were performed at 30 ◦ C using dichloromethane as a solvent. The authors showed that ZnCl2 loading and the preparation method are important parameters controlling the catalytic properties (Table 25.1). The high regioselectivity of the microwave-prepared ZnCl2 /NaY catalyst is attributed to the formation of new Lewis acid sites of –O–Zn–Cl with medium acid strength. Comparison of different catalysts for the Diels–Alder reaction between myrcene and acrolein.

Table 25.1

Catalyst

ZnCl2 (mmol g –1 )

Preparation method

Conversion (%)

Selectivity (%)

Regioselectivity (p/m)a

None NaY ZnNaY ZnCl2 ZnCl2 /NaY ZnCl2 /NaY ZnCl2 /NaY ZnCl2 /NaY ZnCl2 /NaY ZnCl2 /HY

– 0 0.62 2.75 2.75 2.75 2.75 1.84 0.92 2.75

– – Ion exchange – Physical mixture Heated 200 ◦ C, 1 h Microwave Microwave Microwave Microwave

6.6 6.7 13.8 89.9 90.2 – 90.0 64 15 95.8

98.8 97.8 96.2 76.7 77.2 – 92.6 93 96 50.1

72 : 28 72 : 28 77 : 23 92 : 8 92 : 8 – 93 : 7 92 : 8 83 : 17 92 : 8

Reactions performed in CH2 Cl2 solvent at 30 ◦ C for 6 hours. ‘‘para’’/‘‘meta’’ ratio.

a

785

786

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

O O O Scheme 25.16

O

H-Zeolite

+

O O

Diels–Alder reaction between cyclopentadiene and maleic anhydride.

Onaka et al. [74, 75] also found excellent selectivity to the para isomer (p/m ratio 98 : 2) in the Diels–Alder condensation between 2-methyl-1,3-butadiene and α,β-unsaturated esters performed at −1 ◦ C in hexane with dealuminated HY and H-Beta zeolites as catalysts. However, yields of the Diels–Alder adduct were low (11–62%), whereas an aluminum-rich mesoporous aluminosilicate (AlHMS) showed higher activity per aluminum atom than these conventional zeolites, achieving yields between 87 and 94% with similar regioselectivity. Additionally, AlHMS was also recyclable, and its activity and selectivity were comparable or superior to homogeneous Lewis acid AlCl3 . Transition-metal/Y zeolites were prepared by microwave solid-state and aqueous ion-exchange methods and tested for the Diels–Alder reaction of cyclopentadiene and maleic anhydride, and anthracene and maleic anhydride [76]. Whatever be the preparation method, the authors found that the yield of the Diels–Alder adduct increases in the order NaY < Fe+2 Y < Fe+3 Y < MnY < CuY ≈ NiY < CoY ≈ CrY < ZnY. The yield of these reactions was higher when ion-exchanged metal zeolite was used, which was attributed to the formation of Br¨onsted acid sites by solvation of the metal cations by water molecules. Thus, for small cations such as Zn+2 , a high degree of ion exchange was observed, which generates a high concentration of Br¨onsted acid sites. In this case, maximum yields, about 65–70%, were obtained. In addition, endo-selectivity was observed for these reactions (Scheme 25.16). Mesoporous MCM-41 molecular sieves have also been used as catalysts for the Diels–Alder reaction of different dienes and dienophiles. For instance, Kugita et al. [77] found that, for the Diels–Alder reaction between cyclopentadiene and α,β-unsaturated aldehydes, the activity of Al-MCM-41 catalyst depends on the synthesis process. Thus, an Al-MCM-41 catalyst prepared by a grafting method becomes the most active when compared with samples obtained by other preparation methods such as sol–gel, hydrothermal, and template cation exchange. This behavior is attributed to the higher acidity of the former sample. For instance, for the cycloaddition of cyclopentadiene with crotonaldehyde performed at 37 ◦ C, it was found that the material prepared by grafting shows higher activity (maximum crotonaldehyde conversion of 76.3%) than using mesoporous SBA-15, microporous HZSM5 and HY zeolites, amorphous silica-alumina, amberlysts, and the homogeneous p-toluenesulfonic acid and AlCl3 . However, all catalysts tested (heterogeneous and homogeneous) gave similar stereoselectivity, in the ratio endo/ exo ∼ 1. Satsuma et al. [78] performed the Diels–Alder reaction of 2,3-dimethylbutadiene with p-benzoquinone using Al- and Ga-doped MCM-41, amorphous silica, and zeolites (HY and H-Mordenite). All catalysts exhibited excellent selectivity to the

25.2 Acid-Catalyzed Reactions

H

O H

H O

O

H H+-Zeo−

+

H O

Scheme 25.17

Endo

O

O

H endo−anti−endo + H O H

H O H endo−anti−exo

Diels–Alder reaction between cyclopentadiene and p-benzoquinone.

Diels–Alder adduct, Al- and Ga-doped MCM-41 being most active catalysts (80 and 63% of the Diels–Alder adduct were obtained, respectively) and zeolites much less active. Cycloaddition of cyclopentadiene and p-benzoquinone was studied using amorphous silica and different ITQ-2 and MCM-41 pure silica and metal-containing materials as catalysts [79] (Scheme 25.17). The results showed that while amorphous silica does not have any effect on the Diels–Alder reaction, pure-silica-delaminated ITQ-2 and mesoporous MCM-41 strongly increase the reaction rate. Since silanol groups are not expected to play any signiﬁcant catalytic role, the result was attributed to an adsorption-conﬁnement effect of the reactants. In fact, when decreasing the pore diameter of pure silica MCM-41, an increase in the conversion is observed, achieving 90% conversion with 91% selectivity to endo-anti-endo adducts. When tetrahedrally coordinated Ti or Sn is introduced into the framework of MCM-41, the reaction rate strongly increased. In addition, it was observed that the Sn-MCM-41 catalyzes the retro-Diels–Alder condensation, resulting in an increase in the selectivity to the endo-anti-exo isomer. This effect is attributed to the presence of more acidic Sn-OH sites. When Al-MCM-41 with stronger acid sites was used, the retro-Diels–Alder condensation was much faster and the selectivity to the endo-anti-exo stereoisomer increased four times. 25.2.4 Acetalization of Carbonyl Compounds

The acetalization reaction is a selective protection route of carbonyl groups in the presence of other functionalities during the manipulation of multifunctional organic molecules [80]. Besides the interest of acetals as protecting groups, many of them have applications as ﬂavors or fragrances in cosmetics, food, beverage additives, pharmaceuticals, and detergents, as well as in lacquer industries [81]. The most general method for acetalization is to react a carbonyl compound with an alcohol or an orthoester in the presence of an acid catalyst. Acetalization of aldehydes can be performed in the presence of weak acids, while ketones

787

788

O

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

O O

+

OH

O

O

O

HO

O Fructone

Scheme 25.18

Synthesis of fructone.

generally need stronger ones such as sulfuric, hydrochloric, or p-toluenesulfonic acid. Acid zeolites and mesoporous aluminosilicates have been used as catalysts for the synthesis of dimethylacetals of different aldehydes and ketones using trimethylorthoformate or methanol as reactants, giving in general good yields [82–88]. Mesoporous Al-MCM-41, zeolites, and 12-phosphotungstic acid supported on MCM-41 have been used recently with good success as acid catalysts for the synthesis of pentaerythritol acetals of different carbonyl compounds, which have application as plasticiers and vulcanizers [89, 90]. Here, we present some examples of acetals widely used as ﬂavor and fragrances synthesized using zeolites as acid catalysts. Fructone (ethyl 3,3-ethylendioxybutyrate) is a ﬂavoring material with apple scent. It is prepared by acetalization of ethyl acetoacetate with ethylene glycol using conventional strong acid homogeneous catalysts (Scheme 25.18). The drawback of the homogeneous process is that the acid can cause the hydrolysis of the ester functionality producing the corresponding 3,3-ethylendioxybutanoic acid, which, besides reducing the yield of fructone, can alter the organoleptic characteristic of the ﬁnal product when it is present in amounts >3%. Climent et al. [91] reported the synthesis of fructone fragrance using different acid zeolites (H-Beta, HY, HZSM-5, HMor) and the mesoporous aluminosilicate MCM-41. Reactions were performed both in toluene (at reﬂux temperature) and in the absence of a solvent (at 40 ◦ C) while removing the water formed by azeotropic distillation or working in a vacuum of 8 torr, respectively. The results showed that tridirectional zeolites (Beta and Y) were the most active. By changing the Si/Al ratio of both zeolites, the sorption properties could be optimized, and Beta zeolite with Si/Al ratio of 25 and HY with Si/Al ratio of 20 gave conversion and selectivity to fructone close to 100% within 1 hour reaction time. Recently, the synthesis of fructone has been reported using 12-phosphotungstic acid and cesium salt of the 12-phosphotungstic acid supported over various porous carriers such as silica gel, mesoporous silica SBA-15, ultrastable Y (USY), and dealuminated ultrastable Y zeolite [92, 93]. While supported 12-phosphotungstic acid over dealuminated USY showed poor stability as a result of the leaching of heteropolyanions, the supported cesium salt proved very active and stable catalyst. Working at 90 ◦ C, conversion and selectivity to fructone near 100% in 90 minutes were achieved; however, it was necessary to add a water-removal agent in the reaction media (40% concentration of cyclohexane) [93].

25.2 Acid-Catalyzed Reactions

O H

(a)

OH +

H+

O

OH

OH

O

O

+

O

HO HO CH3O

(b)

OH

H + O

Scheme 25.19

HO + H2O

CH3O

OH

O O

O (c)

H+

OH + H2O

+

OH OH

O H+

O

+ H2O

Synthesis of hyacinth (a), vanilla (b), and blossom orange (c) scents.

Synthesis of acetals with hyacinth, vanilla, and blossom orange scents was performed using zeolites as catalyst [94, 95] (Scheme 25.19). The hyacinth scent is obtained by acetalization of phenyl acetaldehyde with glycerol, which leads to a mixture of the corresponding dioxolan and dioxan isomers (Scheme 25.19a). Acetalization of vanillin with propylene glycol gives an acetal with vanilla scent (Scheme 25.19b). Finally, the acetalization of methyl naphthyl ketone with propylene glycol gives an acetal with blossom orange scent (Scheme 25.19c). Reactions were performed with zeolites of different structures (HY, HB, HMor, HZSM-5) and with the delaminated ITQ-2 zeolite [96, 97] by reﬂuxing with toluene and removing the water by azeotropic distillation. It was shown that for the acetalization of small carbonyl compounds such as phenyl acetaldehyde and vanillin, tridirectional zeolites are intrinsically more active than delaminated zeolite ITQ-2, achieving high yields (88–96%) of the corresponding acetals in short reaction times. However, when the size of the reactants increases, as in the case of the propylenglycol acetal of 2-acetonaphtone, geometrical constraints in tridirectional zeolites decrease the diffusion rate of the reactants, with a corresponding negative effect on reaction rate. In this case, the delaminated ITQ-2 becomes the most active catalyst. As in the case of fructone, in these three examples, reactants with very different polarities are involved (the carbonyl compound and the glycol). Therefore in all cases, the important role played by Si/Al ratio of the catalyst as a parameter controlling both acid site density and sorption (polarity) properties was demonstrated. 25.2.5 Fischer Glycosidation Reactions

An important acetalization reaction is the acid-catalyzed Fischer glycosidation that involves the acetalization of a sugar with an alcohol giving the corresponding acetal or alkyl glycoside. The alkyl glycosides derived from fatty acid alcohols are particularly interesting, as they possess a hydrophilic part (the sugar moiety) and a hydrophobic part (the fatty alcohol moiety) giving the molecule surfactant

789

790

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

CH2OH

HOH2C

OH

O

CH2OH O OR

ROH/H+

OH

O OH

OH

OH OH a-D-Glucose

Scheme 25.20

+

OR

OH OH

OH Alkyl a,b-D-glucofuranoside

OH Alkyl a,b-D-glucofuranoside

Fischer glucosidation reaction.

properties. Nowadays, based on their annual production, alkyl glycosides can be considered as the most important sugar-based surfactants. The Fischer glycosidation is an equilibrium reaction that produces a mixture of alkyl d-glycofuranosides and alkyl d-glycopyranosides (isomers α and β) (Scheme 25.20). The furanosides are the kinetically favored primary products and they isomerize to give the pyranosides. In the case of glucose, if the reaction is taken to the thermodynamic equilibrium, then the glucopyranoside to glucofuranoside ratio is ∼ =95 : 5 while the ratio of α/β is 65 : 35 mol mol−1 . A side reaction is the oligomerization of the sugar with the formation of alkyl polyglycosides, leading to a complex product mixture. Acid zeolites have been used for catalyzing the synthesis of alkyl d-glucosides. Corma et al. [98] showed that three-dimensional 12-ring zeolites such as Beta and Y showed reasonable activity and selectivity for the synthesis of alkyl d-glucosides. Thus, yields of butyl glucosides (α,β-butylglucopyranosides plus α,β-butylglucopyranosides) between 70 and 98% were achieved over Y and Beta zeolites, respectively, working at 110 ◦ C. Interestingly, when zeolites are used as acid catalysts for the glycosidation reaction, the formation of oligomers is signiﬁcantly reduced compared to the homogeneous reaction, because their bulkier transition states are restricted by shape-selective effects. This fact has also been observed by other authors in the glucosidation with 1-butanol in the presence of HY zeolites [99]. Optimization of Beta zeolite crystal size showed that the effectiveness is maximum for samples with a crystallite size ≤0.35 µm, indicating that in this range of crystal sizes the diffusion of the reactants through the micropores is not controlling the reaction [100]. Owing to the different polarities of both reactants (glucose and butanol), the adsorption/desorption properties of the Beta zeolite play an important role on the catalytic activity. Thus, an optimum in the Si/Al ratio of the zeolite was found that matches two key properties, namely, the number of active sites and the adsorption characteristics. Furthermore, the deactivation rate is lower when the hydrophobicity of the zeolite is higher. Similar results have been found later by Chapat et al. [101] for the glucosidation of 1-butanol over a series of dealuminated HY zeolites. Moreover, these authors found [99, 101] that the use of microporous catalysts can increase the stereoselectivity to butyl β-d-glucopyranosides in signiﬁcant manner.

25.2 Acid-Catalyzed Reactions

791

The preparation of long-chain alkyl glycosides can be performed by a two-stage transacetalization process that involves ﬁrst the acetalization of the sugar with a short-chain alcohol (usually butanol) giving the corresponding butyl glycoside and a subsequent reaction with a fatty alcohol, while removing the short alcohol by distillation, forming the long-chain alkyl glycoside through the transacetalization process. Beta zeolite turned out to be an efﬁcient catalyst [102] for the transacetalization of a mixture of butyl d-glucosides with n-octanol at 120 ◦ C and 400 torr. The octyl glucoside isomers α,β-octyl d-glucofuranoside and α,β-octyl d-glucopyranoside were obtained in yields of 10 and 80%, respectively. In general, ratios of fatty alcohol to butyl d-glucosides of 4 and above were found to give a good initial reaction rate, minimizing side reactions. The direct reaction of a long-chain alcohol with d-glucose was also performed using H-Beta zeolite as catalyst [102]. Owing to the low solubility of the d-glucose in long-chain alcohols (octanol and dodecanol), the reactions were performed over Beta zeolite using an excess of alcohol at 120 ◦ C, while glucose was added in incremental amounts. In the case of 1-octanol, a total conversion of d-glucose with 99% yield of octyl d-glucosides was obtained. With 1-dodecanol, the yield of dodecyl glucosides was slightly lower (80%) but signiﬁcantly higher than those obtained by the transglucosidation process. The ordered mesoporous aluminosilicate MCM-41, possessing an acid strength weaker than that of zeolites but with signiﬁcantly larger pore sizes, is also able to catalyze glycosidation reactions with excellent yields [103, 104]. Particularly, MCM-41 with mild acidity proved to be an excellent catalyst for 2-O-alkylation of ketohexoses (fructose and sorbose) [105]. Glycosidation of fructose yields three main products: β-d-fructopyranoside and α- and β-d-fructofuranosides. It was shown that the furanose/pyranose ratio can be modiﬁed by varying the Si/Al ratio in the MCM-41 catalyst (Scheme 25.21). Fructose-containing disaccharides have also been glycosilated using MCM-41 as a mild acid catalyst. For instance, butyl leucroside, butyl isomaltuloside, and butyl lactuloside were obtained in yields of nearly 100% by reﬂuxing the corresponding sugar with 1-butanol during 24 hours in the presence of an MCM-41 [106]. O O HO OH

CH2OH

ROH H+

O

CH2OH

HOH2C

OH

OH

OR

OH

+

O

+

OR

b-D-fructofuranoside

CH2OH

HO OH b-D-fructopyranoside Fischer glycosidation of the fructose.

CH2OH

OH

OH a-D-fructofuranoside

Scheme 25.21

OR

HOH2C

792

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

25.2.6 Isomerization Reactions: Isomerization of α-Pinene and α-Pinene Oxide

The isomerizations of terpenes such as α-pinene and its corresponding oxide are important processes in the ﬁne chemicals industry as they are the source of some high-value compounds (such as camphene, limonene, campholenic aldehyde (CPA)) extensively used as ﬂavor and fragrance additives in cleaning agents and cosmetics, perfumery, food, and pharmaceuticals. Industrially, the isomerization of pinene is performed over TiO2 catalysts under normal pressure at temperatures above 200 ◦ C. The overall yield of camphene, limonene, tricyclene, and small amounts of ﬂenchenes and bornylene is between 75 and 80%. The isomerization rate is low and the oxide is treated with an acid in order to form a layer of titanic acid on the catalyst surface [107, 108]. Owing to these drawbacks, there is great interest in ﬁnding new catalysts that could exhibit higher activity and selectivity than camphene and/or limonene (Scheme 25.22). Solid acids such as zeolites and modiﬁed clays have been largely used as acid catalysts for the isomerization of pinene because of their suitable acid sites and shape selectivity. Recently, Yilmaz et al. [109] used Beta zeolites with different Si/Al ratios incorporating B, Ti, or V atoms as catalysts for the liquid-phase isomerization of α-pinene at 273 ◦ C. H-Beta zeolites (Si/Al = 27–33) exhibited high catalytic activity, with selectivity to camphene and terpinenes close to 27.5 and 13%, respectively, while samples with boron, titanium, or vanadium presented insigniﬁcant catalytic activity. Gunduz et al. [110, 111] working at the same temperature also reported that the maximum yield of camphene (25–27%), at high level of conversion, was obtained in the presence of Beta zeolites with Br¨onsted acid sites. Terpinenes were also produced with 8–20% selectivity. Other products such as terpinolenes and

a-Fenchene

Isomerisation

Isomerisation

+ Camphene

a-Pinene

b-Pinene Limonene Isomerisation

p -Cymene Scheme 25.22

g -Terpinene a-Terpinene Dipentene Terpinolene

Possible products resulting of the isomerization of pinene.

25.2 Acid-Catalyzed Reactions

heavy products were observed. Recently, the same group [112] reported a comparative study of the pinene isomerization using zeolites such as ZSM-5, Beta, and mordenite, and the mesoporous material MCM-41. They observed that in liquid phase at 200 ◦ C, the Beta zeolite with a Si/Al ratio of 55 presented a good combination of acidity and pore size giving 99% conversion with selectivity to camphene of 27%. Over dealuminated mordenites and Y zeolites at 120 ◦ C, Lopez et al. [113] reported that main products formed on dealuminated mordenite were camphene and limonene with a maximum yield of 68%, while over Y zeolite the formation of undesired compounds was favored but the selectivity camphene/camphene + limonene was similar to those obtained with mordenite. The authors concluded that selectivity was not inﬂuenced by the conversion level when below 90% and when the pore diameter of the zeolite was larger, the production of undesired products was higher. Akpolat et al. [114] reported the inﬂuence of the calcination temperature of a natural zeolite (clinoptilolite) on its catalytic activity for the isomerization of pinene at 155 ◦ C. The study revealed that the activity decreased when the calcination temperature increased, as a result of the disappearance of Br¨onsted acid sites. In addition, they observed that selectivity to camphene was constant (30%) independent of the conversion level, whereas the selectivity to limonene decreased at conversions higher than 80–85% from 20 to 5% approximately. Allahverdiev et al. [115–117] performed the pinene isomerization over clinoptilolite in liquid phase at 120–160 ◦ C and 1–10 bars. Camphene and limonene were obtained as main products, achieving 70% selectivity to camphene plus limonene at 80–85% conversion. Findik and Gunduz [118] performed the isomerization reaction on clinoptilolite at 155 ◦ C, and camphene was obtained with a yield of 43%. Ozkan et al. [119] reported the isomerization of pinene over exchanged-clinoptilolite with NH4 + , Ba2+ , and Pb2+ . Untreated zeolite gave high selectivity to limonene because of the low transformation rate, while exchanged zeolites gave more secondary products. The isomerization of α-pinene oxide in the presence of Lewis or Br¨onsted acid catalysts [120] results in the formation of CPA and many others compounds such as pinocarveol (PCV), trans-carveol (TCV), trans-sorbrerol, isopinocamphone (IPC), and p-cymene (Scheme 25.23). Lewis acids favor the formation of CPA and PCV, whereas Br¨onsted acids favor the formation of TCV, trans-sorbrerol, IPC, and p-cymene. From an industrial point of view, the most important product coming from the pinene epoxide is CPA. This compound is used as an intermediate for the preparation of sandalwood-like fragrances [121] and is also used in perfumery and pharmaceuticals as an odorant additive (herbal green woody amber leafy). The maximum selectivity reported to CPA using homogeneous Lewis acid catalysts, such as ZnCl2 and ZnBr2 , is 85% [122, 123]. However, there are numerous drawbacks associated to the use of Zn halides as catalysts, such as the fast deactivation of the catalysts, requirement for a high catalyst/substrate ratio due to their low reaction rates, the turnover numbers lower than 20 [124], and ﬁnally the need for aqueous extraction of the catalyst producing a large amount of wastes

793

794

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

OH

a-Pinene oxide O Lewis acid Lewis acid

Pinocarveol Brönsted acid H 2O

O Campholenic aldehyde (CPA)

Le

wis

ac

id

OH

OH

O

OH Isopinocampheol Isopinocamphone OH

Trans-carveol

Trans-sobrerol

p-cymene Scheme 25.23

Different compounds obtained from the isomerization of α-pinene oxide.

contaminated with heavy metal. Among others solid acid catalysts, zeolites appear as a good alternative to overcome some of these drawbacks. Holderich et al. [123] used acid-treated HUSY zeolites for the isomerization of α-pinene oxide. The authors obtained a selectivity of 75% to CPA at 100% conversion within 24 hour s of reaction time at 0 ◦ C, using toluene as a solvent. The high selectivity was attributed to well-dispersed Lewis acid sites in a nearly all silica matrix obtained by the previous acid treatment. At a reaction temperature −15 ◦ C, the selectivity to CPA of 78% at 90% conversion in 24 hours was achieved by Kunkeler et al. [125] performing the isomerization of α-pinene oxide to CPA in both liquid and gas phase using titanium-Beta (Ti-beta) zeolite. In a ﬂow reactor, 94% selectivity toward CPA was achieved at 95% conversion using 1,2 dichloroethane as solvent and 90 ◦ C, while using alkanes as solvent they obtained a selectivity of 89% with a conversion of 100%, but in this case the catalysts suffered a drastic deactivation after 5–6 hours of reaction. Ravindra et al. [126] reported the catalytic activity of B2 O3 supported on SiO2 and compared the activity with those exhibited by HY, Al- and Zn-MCM-41, and Al-MSU (mesoporous structure with microporous walls (nanoparticles of Faujasite as seeds)). Working at 25 ◦ C, they observed that the most active sample was SiO2 loaded with 15 wt% B2 O3 . However, the selectivity to CPA was constant (about 70%) independent of the B2 O3 loading. HY zeolites and Al-MCM-41 showed high conversion and selectivity (<66%), whereas Al-MSU presented 86% selectivity to CPA at 54% of conversion. The authors attributed the good activity and selectivity of this catalyst to the presence of both micro- and mesopores in its framework.

25.2 Acid-Catalyzed Reactions

25.2.7 Oxidation and Reduction Reactions 25.2.7.1 Epoxidation Reactions Epoxides are generally obtained by a catalyzed oxidation of C=C bonds and they are the key products for the production of a wide variety of alcohols, carbonyl compounds, ethers, aminoalcohols, and so on. The conventional method to synthesize epoxides has been the oxidation with organic peracids. However, owing to safety problems related with the use of peracids, epoxidation procedures using hydrogen peroxide, tert-butyl hydroperoxide (TBHP), and cumyl hydroperoxide (CHP) have been developed. They produce as by-products water, tert-butanol, and cumyl alcohol, respectively, which are easily recyclable products. Different examples of heterogeneous catalysts for epoxidation have been reported in the literature, and among them we can highlight complexes of metals supported or enclosed in microporous or mesoporous materials and titanium silicates [127]. From an environmental point of view, the most convenient peroxide used in epoxidation reactions is hydrogen peroxide. The discovery of the remarkable activity of titanium silicalite-1 (TS-1) as a catalyst for a variety of oxidations, including epoxidation, with aqueous hydrogen peroxide, has constituted an important advance in oxidation catalysis [128–132]. However, the pore size restrictions associated with titanium silicalites was a limitation for oxidation of bulkier molecules, and so the incorporation of titanium in large pore molecular sieves was investigated by several groups. The preparation of Ti-Beta for the ﬁrst time by Corma et al. [133, 134] offered the possibility to perform the oxidation of some larger molecule than with the TS-1. In 1994, the Ti incorporation in the mesoporous material MCM-41 was performed successfully by Corma et al. [133, 135], and epoxidation of still larger molecules could be done selectively with organic peroxides when the hydrophilic character of the mesoporous solid was controlled by silylation [136–138]. A catalyst jointly developed by the ITQ of Valencia and Sumitomo has resulted in the new Sumitomo process for epoxidation of propylene to propylene oxide. In this section, we present some examples of epoxidation of terpenes as high-value compounds for the ﬁne chemical industry, which are performed using green oxidants and molecular sieves as catalysts. The epoxidation of α-pinene with heterogenous catalysts has been extensively studied over a wide variety of heterogeneous catalysts [139]. The reaction usually affords a complex mixture of different compounds, sometimes unidentiﬁable because of the occurrence of simultaneous competitive reaction of isomerization, hydrogenation, and dehydrogenation, hydration, and so on (Scheme 25.24). Thus, Chiker et al. [140, 141] prepared mesoporous silica SBA-15 samples functionalized with titanium as catalysts for the epoxidation of pinene with TBHP as oxidant on reﬂuxing from acetonitrile. For instance, using a Ti-SBA15 with pore diameter of 61 A˚ and a Si/Ti ratio of 17.8, the epoxide was produced with 100% selectivity at 91% conversion, with an efﬁciency of the oxidant of 93%. Over several reaction cycles, the authors observed leaching of Ti species in the liquid phase. With H2 O2 at 70 ◦ C, the selectivity was constant but the conversion decreased to 35%

795

796

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

O [O] O Campholenic aldehyde (CPA)

a-Pinene oxide

a-Pinene

OH

OH O

1,2 Pinanediol Scheme 25.24

D-verbenone

Different compounds obtained in the epoxidation of α-pinene.

and the efﬁciency of the oxidant to 19%. Kapoor et al. [142] reported the catalytic activity of titanium hexagonal mesoporous aluminophosphate (TAP) molecular sieves prepared by hydrothermal crystallization in a ﬂuoride medium. The catalysts were active for the epoxidation of α-pinene, and the oxide was obtained with 67% yield at 72% conversion. However, the reuse of the catalyst showed a decrease in activity. On et al. [143] prepared bifunctional titanium-mesoporous molecular sieves, MCM-41, with different trivalent ions such as B3+ , Al3+ , or Fe3 . The bifunctional catalysts produced the α-pinene epoxide along with a large amount the corresponding diol due to the presence of both acid and oxidizing sites, whereas the Ti materials were 100% selective to the epoxide. However, all catalysts presented low conversion of pinene (4–10%) using H2 O2 or TBHP as oxidant. Suh et al. [144, 145] attempted the one-pot synthesis of CPA from α-pinene over Ti-substituted mesoporous molecular sieves at 75 ◦ C. A 35% conversion of α-pinene with very low selectivity to CPA and pinene oxide was obtained after 24 hours, while the maximum selectivity was toward verbenone (50%). Selectivity to CPA could be increased to 55% at 35% conversion of α-pinene when the water contained in the oxidant (TBHP) was previously removed. Limonene oxide is found in natural sources and is used in fragrances. Since a variety of reactions can occur during the epoxidation (Scheme 25.25), there is interest to ﬁnd a selective epoxidation catalyst that is able to work at high levels of conversion. Prada et al. [146] reported the oxidation of limonene with oxygen using NaY zeolite exchanged with Fe2+ , Co2+ , Mn2+ , or Mo2+ . The best results were obtained with NaCoMoY zeolite, with 53% selectivity to 1,2 epoxylimonene at 57% conversion. Because of the success of Ti-zeolites and Ti-mesoporous materials for epoxidation of oleﬁns [147, 148], these materials have been used for epoxidation of limonene. Hutter et al. [149] performed a comparative study of the epoxidation of several cyclic oleﬁns, including limonene, over structurally different titania–silica catalysts. The authors found that a supercritically dried silica–titania gel showed better catalytic behavior for these reactions than Ti-zeolites and silica-supported titania. Using

25.2 Acid-Catalyzed Reactions

Limonene

OH H2O

[O]

Carveol

OH

O

OH

1,2-Epoxide

Diol

[O]

[O] OH

O O [O] O

Carvacrol O

Diepoxide 8,9-Epoxide Scheme 25.25

Carvone

Different compounds obtained in the epoxidation of limonene.

cumyl hydroperoxide as oxidant, 87% yield of limonene oxide was achieved. Two strategies have been presented to improve the catalytic activity and selectivity of Ti-MCM-41 [150] for the epoxidation of oleﬁns. The ﬁrst approach involves silylation of the surface of Ti-MCM-41 that produces a very hydrophobic catalyst, whereas the second approach is based on removal of water from the reaction media. The increase in catalytic activity is not due to a change in the intrinsic activity of the Ti sites, but rather due to a decrease in the catalyst deactivation by reducing the formation of diols produced by ring-opening of the epoxide. Thus, for instance, when silylated Ti-MCM-41 materials were used for epoxidation of limonene with TBHP, yields of 1,2 limonene epoxide of 80% could be achieved. Camphene can be obtained by isomerization of pinene and can be functionalized in order to enlarge its possible applications. One of the possible functionalizations is the epoxidation that will generate the oxide as well as the alcohols and carbonyl derivatives. van der Waal et al. [151] carried out the oxidation with the aluminum-free zeolite Ti-Beta with H2 O2 . The camphene epoxide is unstable to the weakly acidic TiOOH and rearranges to CPA. Thus, low conversions (4%) with selectivities of 1% into the corresponding epoxide and of 92% into the CPA were obtained. In any case, this is an interesting result since it shows that the catalyst was able to epoxidize and isomerizes the epoxide to the corresponding aldehyde (Scheme 25.26). Methyltrioxorhenium (MTO) was a better catalyst for camphene epoxidation either in the homogeneous phase or when heterogenized on NaY zeolite [152]. When the epoxidation of camphene was performed with NaY/MTO and 85% H2 O2 in water, excellent selectivity >95% to camphene epoxide was obtained at 89% conversion, the ﬁnal mass balance being 79%.

797

798

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

Isomerization

Oxidation O Camphene Scheme 25.26

O Campholenic aldehyde (CPA)

Camphene oxide

Epoxidation of camphene and rearrangement of camphene oxide.

6,7-Epoxylinalool is assumed to be the natural precursor of the furan and pyran hydroxyl ethers presented in Scheme 25.27. They possess an intense rose odor and are selectively formed from linalool by the epoxidase enzyme. Corma et al. [153] reported that bifunctional catalysts such as Ti-Al-Beta and Ti-Al-MCM-41, containing Ti+4 as oxidation sites and Al+3 which has an associated H+ as the complementary cation, are able to transform linalool to the hydroxyl ethers in one-pot reaction. Using TBHP as oxidant at 80 ◦ C, conversions of linalool of 73 and 80% with 100% selectivity to the hydroxyl ethers were achieved with Ti-Al-Beta and Ti- Al-MCM-41, respectively. OH

OH Ti-Al-Beta/TBHP 80 °C

OH

+

O

O

Linalool Scheme 25.27

Furan and pyran hydroxyl ethers obtained by epoxidation of linalool.

Cis- and trans-geraniols were epoxidized selectively using TS-1/H2 O2 to the corresponding cis and trans-epoxides at 2,3-position, while no reaction took place to the most substituted double bond [154]. This stereo- and regioselectivity observed for allyl alcohol epoxidations on Ti-zeolites is attributed to the implication of alcohol function in the oxygen transfer step (Scheme 25.28). O

OH

cis -Geraniol

Scheme 25.28

OH +

cis -1,2-epoxide 90%

Selective epoxidation of cis-geraniol.

CHO

Geranial 10%

25.2 Acid-Catalyzed Reactions

O O

O

RCOOOH

Scheme 25.29

Baeyer–Villiger oxidation.

25.2.7.2 Baeyer–Villiger Oxidations The Baeyer–Villiger reaction is a widely used synthetic method to oxidize ketones to lactones or esters [155] (Scheme 25.29). This transformation has a large variety of applications both in the area of bulk chemicals and ﬁne chemicals and is usually performed using the less desired peroxocarboxylic acids or m-chloroperbenzoic acid (mCPBA) [155]. An additional drawback of the conventional oxidation process is the concomitant formation of, at least, one molecule of the acid that has to be separated and reoxidized to produce the peracid. Furthermore, peracids are explosive and an alternative route that could avoid the use of organic peracids would be highly desirable. This alternative route can be the use of hydrogen peroxide as oxidant and zeolites or other solids as heterogeneous catalysts [156]. Several acid zeolites [157, 158] and TS-1 [159] have been used as catalysts for the Baeyer–Villiger oxidation of cyclic ketones, achieving modest conversions and selectivities to the corresponding lactones. Thus, Wang et al. [160] reported the oxidation of cyclopentanone with H2 O2 using HZSM-5 zeolite. The authors found that the hydrophobicity plays an important role in the process, as this occurred largely on Br¨onsted acid sites. A comparison with the TS-1 showed that HZSM-5 is a better catalyst than TS-1 for the Baeyer–Villiger oxidation in aqueous media since TS-1 provides lower conversion and selectivity to δ-valerolactone (<65%) [159]. The inﬂuence of the thermal treatment of a dealuminated Beta zeolite on their activity for the Baeyer–Villiger reaction of cyclohexanone with hydrogen peroxide was studied by Lenarda et al. [158]. It was found that the conversion and selectivity to ε-caprolactone were very poor, but increased with increasing reaction temperature. An important achievement in the synthesis of zeolites has been the incorporation of metals such as titanium [161, 162] and tin [163] into the framework of larger pore 12-ring zeolites and into mesoporous molecular sieves to overcome the pore restrictions of TS-1 [164, 165]. Particularly, the synthesis of Sn-Beta zeolite with isolated tetrahedral tin sites presented singular activity for the Baeyer–Villiger oxidation. It was found that the active sites for the Baeyer–Villiger oxidation on Sn-Beta were the partially hydrolyzed Sn–OH groups. A broad study of the characterization of active sites and reaction mechanism has been performed by spectroscopic measurements, isotopic labeling, and theoretical methods [166–172]. Sn-Beta is an active and selective catalyst for the oxidation of cyclic ketones using H2 O2 as oxidant [163, 172]. Thus, cyclic ketones such adamantanone and cyclohexanone were oxidized to the corresponding lactones with 35% aqueous H2 O2 at moderate temperatures with high yield and selectivity (>98%). The

799

800

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

O

O

O

MTBE 56 °C 6 hours

O

Table 25.2

O

+

O

O 1

Scheme 25.30 vone.

+

O

2

3

Possible products formed in the Baeyer–Villiger oxidation of dihydrocar-

Baeyer–Villiger oxidation of dihydrocarvone with different oxidants [163].

Oxidant

Reactant conversion (%)

Product selectivity (%)

O

O

O

O

O

O O Sn-Beta/H2 O2 MCPBA Ti-Beta/H2 O2

68 85 48

100 11 0

0 71 79

O 0 18 0

catalyst was reused with virtually constant activity in up to four catalytic cycles. The catalyst was also proved to be highly chemoselective in the Baeyer–Villiger oxidation of unsaturated ketones (Scheme 25.30). In Table 25.2, the results of the oxidation of dihydrocarvone with different catalysts at 56 ◦ C using methyl tert-butyl ether as a solvent are compared. As can be seen, Sn-Beta exhibits 100% selectivity to the lactone, while Ti-Beta is selective for the epoxidation of the double bond. Finally, with mCPBA as oxidant, the selectivity to the lactone was very low. 4-Alkoxy phenols are important compounds in organic chemistry because they are intermediates for drugs, agrochemicals, and dyes. The system H2 O2 /Sn-Beta has also been used for the synthesis of 4-substituted alkoxyphenols starting from the corresponding benzaldehyde derivatives [173]. This synthetic route for 4-alkoxyphenols involves two consecutive steps, that is, the Baeyer–Villiger oxidation of the benzaldehyde derivative giving the corresponding formate ester followed by hydrolysis of the formate to the corresponding phenol (Scheme 25.31). It was found that by working under selected and optimized reaction conditions the main reaction product can be preselected. Thus, 4-alkoxyphenol derivative can be selectively obtained when ethanol or aqueous acetonitrile is used as a solvent (Table 25.3). However, on working with dioxane as a solvent and in hydrogen peroxide deﬁcit, the formate ester is preferentially formed. Importantly, the catalyst could be regenerated by calcination in air and the initial activity and selectivity were restored.

25.2 Acid-Catalyzed Reactions

O H

O

H

OH

O

B−V

H 2O

OR

OR

OR

4-Alkoxybenzaldehyde

Formate ester

4-Alkoxyphenol

Scheme 25.31 Synthesis of 4-alkoxyphenols through Baeyer–Villiger oxidation of 4-alkoxybenzaldehydes. Table 25.3

Inﬂuence of the solvent on the hydrolysis of the ester to the alcohol.a Water

Conversion

Product distribution

Entry

(mg)

Solvent

TON

(%)

Ester 2

Alcohol 3

Other

1 2 3 4 5 6

– 500 – 500 – 500

Acetonitrile Acetonitrile Dioxane Dioxane Ethanol Ethanol

317 327 259 106 317 244

56 59 46 19 57 44

54 4 77 54 1 5

46 96 23 46 99 95

0 0 0 0 0 0

a Reaction conditions: 0.5 g of p-methoxybenzaldehyde, 0.3 g of 50% aqueous hydrogen peroxide, 3.0 g of solvent, water (when indicated), and 50 mg of catalyst (Sn-Beta) were stirred for 7 hours at 80 ◦ C.

Al-Beta zeolite samples bearing Br¨onsted acid sites, although less selective than Sn-Beta, are more efﬁcient as catalyst for both Baeyer–Villiger oxidation and ester hydrolysis provided the molecule does not contain oleﬁnic groups. Thus, when the aldehyde derivative possesses an oleﬁnic substituent (Scheme 25.32), Al-Beta gives no reactivity whereas Sn-Beta promotes the formation of the corresponding unsaturated phenols with high chemoselectivity. A conventional oxidant such as mCPBA gives very poor selectivity. H

R

O

O

Sn-Beta/H2O2 80 °C, acetonitrile O

O

R=CHO, R = H 85%, selectivity

Scheme 25.32

Selective Baeyer–Villiger oxidation catalyzed by Sn-Beta zeolite.

801

802

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

CHO

CHO

O

Oxidation

Citral

Formate ester

Scheme 25.33

O

Hydrolysis

Melonal

Synthesis of melonal through the Baeyer–Villiger oxidation of citral.

Melonal (2,6-dimethyl-5-hepten-1-al), which possess green melon- and cucumber-like scent, is used in fragrances to introduce melon and cucumber notes [174]. Their conventional synthesis involves a Darzens reaction using halogenated esters, followed by saponiﬁcation and decarboxylation [175]. Using the Sn-Beta/H2 O2 system, a novel halogen-free synthesis strategy has been developed that involves the Baeyer–Villiger oxidation of citral (3,7-dimethyl-6-octen-1-al) to the formate ester, which can be subsequently hydrolyzed to melonal [176] (Scheme 25.33). The performance of the Sn-Beta zeolite has been compared with that of other potential heterogeneous catalysts such as Al-Beta, Ti-Beta, and Zr-Beta as well as the mesoporous Sn-MCM-41. Reactions performed at 100 ◦ C using tert-amylalcohol as a solvent showed that Sn-Beta and Sn-MCM-41 were highly chemoselective for the Baeyer–Villiger oxidation (95% selectivity), whereas Al-Beta, Ti-Beta, and Zr-Beta were less efﬁcient and selective catalysts (66−48%). In the case of Ti-Beta, epoxydation by-products were also detected in considerable amounts. The system Sn-Beta/H2 O2 was also applied to the synthesis of (R)-δ-decalactone through the Baeyer–Villiger oxidation of (R)-delfone (Scheme 25.34) [177]. (R)-δ-Decalactone is an important industrial fragrance that possesses a creamy-coconut and peach-like aroma. Starting with an enantiomerically enriched (R)-delfone, it was shown that the rearrangement occurs with retention of the conﬁguration giving the enantiomerically enriched (R)-δ-decalactone. The oxidation was carried out at 60 ◦ C in the absence of solvent and using 0.5 wt% of catalyst. Under these optimized reaction conditions, the lactone was obtained in 86% yield. The catalyst remains active for a long time and turnover numbers (moles of substrate converted per mole of tin sites) of 10 000 were achieved. Finally, it is interesting to note that when the Baeyer–Villiger oxidation involves large substrates, Sn-Beta zeolite can become limited by the dimensions of their microporous structure. In these cases, it has been shown that the use of tin-mesoporous materials such as Sn-MCM-41 and Sn-MCM-48 can overcome this O

O H

Delfone Scheme 25.34

Sn-Beta/H2O2

O H d-Decalactone

Synthesis of δ-decalactone by Baeyer–Villiger oxidation of delfone.

25.2 Acid-Catalyzed Reactions

limitation [165, 178–180]. Other oxidizing metals such as Fe(III) [181, 182] and Nb [183] have also been incorporated into mesoporous solids such as MCM-41 and MCM-48. It was found that Fe-MCM-41 was active for the Baeyer–Villiger oxidation of cyclic ketones, but their activity was lower than that of Sn-MCM-41, whereas Nb-MCM-41 was not chemoselective when a double bond is present in the molecule, promoting exclusively double bond oxidation. 25.2.7.3 Meerwein–Ponndorf Verley Reduction and Oppenauer Oxidation (MPVO) The Meerwein–Ponndorf–Verley (MPV) reduction of aldehydes and ketones and its reverse Oppenauer’s oxidation (O) of alcohols are considered to be highly selective methods to avoid the reaction of other reducible groups such as C=C and C–halogen bonds [184]. MPV reduction and Oppenauer oxidation (MPVO) reactions are hydrogen-transfer processes in which it is generally accepted that they proceed via a transition-state complex where the carbonyl group and the alcohol are both coordinated to a Lewis acid metal center and a hydride transfer from the alcohol to the carbonyl group occurs (Scheme 25.35). Aluminum alkoxide catalysts such as aluminum isopropoxide and tert-butoxide are commonly used, though stoichiometric amount are often required. Over the past 15 years, an increasing number of reports on heterogeneous catalysts, which include zeolites, for the MPV reduction have been published [185–187]. Al- and Ti-zeolites have been reported as excellent catalysts for MPVO reactions, and it was shown that Lewis acid sites were the active sites of the catalyst. Thus, Creyghton et al. [188, 189] reported the stereoselective reduction of 4-tert-butylcyclohexanone to cis-4-tert-butylcyclohexanol (>95%) (which is a fragrance chemical intermediate) in liquid phase using Beta zeolite as catalyst (Scheme 25.36). It was found that the catalytic activity of the Beta zeolite increased with the calcination temperature, which was attributed to the higher framework dealumination degree of the catalyst [190]. In fact, the infrared study of the samples showed the direct relationship between catalytic activity and the amount of partially hydrolyzed framework aluminum. A mechanism was proposed in which both the alcohol and ketone are coordinated to the same Lewis acid metal site (Scheme 25.37). The high selectivity to the cis-alcohol was attributed to a shape-selectivity effect of the OH

Lewis acid catalyst

R−CO−R1 + R2CH−R3

C R1

R−CH−R1 + R2−CO−R3

OiPr

PriO

O R

Al O

+ H

Scheme 25.35

OH

R3

R1 R

R2

H

R3 O

R2

O Al

PrOi

OiPr

Hydrogen-transfer process in the MPVO reactions.

803

804

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

OH O

H

H H

82 °C

OH

+ H

OH 4-tert -butylcyclohexanone Scheme 25.36

cis -4-tert -butylcyclohexanol

trans -4-tert-butylcyclohexanol

MPV reduction of 4-tert-butylcyclohexanone.

zeolite, which favors mainly the formation of the transition state that leads to the cis-isomer. This transition state is smaller in size and can better accommodate in the pores of the zeolite Beta (Scheme 25.37).

H H

CH3

H

CH3

H

CH3

CH3

O Si

O

O

O Al

Al

Si

Si

O

O

O

O

O Si

Si Transition state for the cis -isomer

Si

O

Transition state for the trans -isomer

Scheme 25.37 Proposed transition states for the MPV reduction of 4-tert-butylcyclohexanone on the Lewis metal sites.

Al-free Ti-Beta zeolite not only presented lower activity than their Al analog for the same reaction [190, 191] but also presented excellent selectivity for the cis-isomer (>98%), indicating that the reaction involves a similar mechanism. It was suggested that tetrahedrally incorporated Ti atoms with Lewis acid properties were involved in the alkoxide formation. The gas-phase reduction of 4-methylcyclohexanone with isopropanol at 100 ◦ C showed that Al-Beta was less selective since the presence of Br¨onsted acid sites caused the dehydration of methylcyclohexanol, methyl cyclohexene being the main product obtained, whereas Ti-Beta was found to be selective to cis- and trans-alcohols [192], showing in addition a lower rate of deactivation. Al- and Ti-Beta zeolites were active for the reduction of prochiral phenylacetone with (S)-2-butanol [193], giving the corresponding alcohol in an enantiomeric excess of 34%, the preferred enantiomer formed being the (S)-(+)-1-phenyl-2-propanol. The formation of the (S) enantiomer was again attributed to a shape-selectivity effect of the zeolite micropores on the transition state. Recently, Akata et al. [194] performed the MPV reduction of 4-tert-butylcyclohexanone with isopropanol using a titanosilicalite ETS-10. Reactions were

25.2 Acid-Catalyzed Reactions

performed in a ﬁxed-bed reactor at 100 ◦ C, and it was found that ETS-10 showed higher selectivity to the more thermodynamically favored trans-4-tert-butylcyclohexanol (76%) than Al-Beta zeolite, which was attributed to the existence of a larger volume available in the pores of the former. Corma et al. reported that Sn-Beta zeolite shows excellent activity and selectivity in the MPV reduction of several ketones and in the reverse Oppenauer oxidation of different alcohols [195, 196]. Comparatively, the catalyst showed superior activity and selectivity than Al- and Ti-Beta zeolites, maintaining their activity even after four recycles. Excellent stereoselectivity (99%) to cis-4-tert-butylcyclohexanol at 97.3% conversion was achieved for the reduction of 4-tert-butylcyclohexanone with isopropanol at 100 ◦ C. Also, a reasonable enantioselectivity (50%) for reducing a prochiral ketone with a chiral alcohol was observed. Working with ketones and alcohols of different structures showed that the reaction transition states that can be ﬁtted by size within the micropores of Beta may have problems forming on the Lewis acid sites due to the shielding effect of the neighboring framework oxygen atoms. The higher activity and selectivity of Sn-Beta zeolite with respect to its homologs bearing Ti or Al can be attributed to the stronger and more selective interaction of the carbonyl group with Sn than to Ti and Al centers. This was demonstrated by IR studies using cyclohexanone as probe molecule over Beta zeolites. In addition, the Sn-Beta was found more resistant to the presence of water in the reaction media than Ti- or Al-zeolite Beta (Table 25.4), thereby offering a real alternative to the highly water-sensitive aluminum alkoxide catalysts. Zhu et al. [197] compared the activity and selectivity of aluminum-free Zr-zeolite Beta with its analogs containing Ti-, Al-, and Sn- for the MPV reduction of 4-tert-butylcyclohexanone using isopropanol at 82 ◦ C. Zr-Beta was much more active than Al- and Ti-Beta and even performed slightly better than Sn-Beta, achieving maximum conversion of 97.3% and >99% selectivity to the cis-isomer when working with a Si/Zr = 75 Beta sample. It is interesting to note the high tolerance of Zr-Beta toward moisture (Table 25.5). Table 25.4 Effect of the H2 O addition to the MPV activity of different zeolites for the reduction of cyclohexanone.a

Catalyst

Ti-Beta Al-Beta Sn-Beta Sn-Beta-Silb

TON (mol mol –1 h–1 ) water added 0g

0.2 g

0.5 g

1.2 7.0 109.0 108.0

0.7 0.1 17.8 56.7

0.7 0.1 3.8 48.0

conditions: 100 ◦ C, 1 hour reaction, 1 mmol cyclohexanone, 60 mmol of 2-butanol, and 75 mg of catalyst. b Silylated sample. a Reaction

805

806

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals Table 25.5

Activity over Zr- and Sn-Beta in the presence of water [197].

Water added (wt%)

0 0.6 2.9 9.1

Zr-Beta

Sn-Beta

TOFa

Conv.(%)b

TOFa

Conv.(%)b

632 447 410 304

99 97 95 84

425 251 86 21

81 60 29 7

Reaction conditions: 1.3 mmol of 4-tert-butylcyclohexanone in 83 mmol of 2-propanol, 100 mg catalyst, under reﬂux and stirring 82 ◦ C. a Turnover frequency after initial 5 minute reaction: mol mol. b Conversion after 30 minutes.

A computational study (DFT, density functional theory) of the MPV reduction of cyclohexanone with 2-butanol catalyzed by Sn- and Zr-Beta zeolites showed that the mechanism of Sn-Beta and Zr-Beta catalysis is similar and involves as ﬁrst step the adsorption of both the ketone and the alcohol on the Lewis acid center. This is followed by the deprotonation of the alcohol, carbon-to-carbon hydride transfer (through a six-membered transition state), proton transfer from the catalyst, and ﬁnally product exchange [198]. The catalyst performs the MPV reduction of several substrates with isopropanol, such as aryl- and cyclic ketones, with excellent yield and selectivity to the corresponding alcohols [199]. Also, a variety of α,β-unsaturated aldehydes were reduced with high selectivity to the corresponding alcohol using Zr-Beta catalysts [200, 201]; however, the conversion was dependent on the molecular structure of the aldehyde. Recently, a stereoselective cascade hydrogenation of 4-tert-butylphenol and p-cresol over Zr-zeolite beta-supported rhodium has been reported by van Bekkum et al. [202]. Thus, over 0.5% Rh/Zr-Beta, 4-tert-butylphenol and p-cresol were transformed into the corresponding intermediate 4-alkylketones by metal-catalyzed hydrogenation, which were subsequently stereoselectively reduced with isopropanol via MPV reduction over Zr Lewis acid sites to the cis-alcohols. Under optimized reaction conditions, 100% conversion of 4-tert-butylphenol with 95% selectivity to cis-4-tert-butylcyclohexanol was achieved within 4 hours (Scheme 25.38). One example of a cascade process using Sn- and Zr-Beta zeolites as Lewis acid catalysts has been reported recently by Corma and Renz [203]. The 4-methoxybenzyl 1-methylpropyl ether has a fragrance with fruity pear odor whose commercial preparation involves a ﬁrst-step reduction of 4-methoxybenzaldehyde to the corresponding alcohol, which is separated and puriﬁed before going into a second process (etheriﬁcation). Then the reduction (hydrogenation)–etheriﬁcation steps have been integrated into one cascade process, which is an interesting, alternative preparation procedure. The process involves the reduction of the 4-methoxybenzaldehyde to the corresponding alcohol through a MPV reaction with 2-butanol, followed by

25.2 Acid-Catalyzed Reactions OH

O

OH 2H2

OH

iPrOH Zr

Rh

R

R

R

R cis + trans

R=tert -butyl or methyl

Scheme 25.38 Cascade reaction for the conversion of 4-alkylphenols to 4-alkylcyclohexanols.

etheriﬁcation of the benzyl alcohol intermediate with 2-butanol which is in excess (Scheme 25.39).

O

HO O

MeO 4-Methoxybenzaldehyde

HO

H2O

OH

MeO

O

MeO 4-Methoxybenzyl 1-methylpropyl ether

Scheme 25.39 Synthesis of 4-methoxybenzyl 1-methylpropyl ether through an MPV-etheriﬁcation cascade process.

Sn- and Zr-Beta zeolites were tested as catalysts for the proposed cascade process. Results in Table 25.5 show that both catalysts are active, giving the desired fragrance with high yield, Zr-Beta being more active for the global process. The water-resistant character of both catalysts is of special relevance for the success of the global process, as water is produced during etheriﬁcation and this limits the use of conventional Lewis acid catalysts. The high selectivity of the process is an interesting point to consider for the synthesis of fragrances, where the formation of by-products (even in small quantities) can change considerably the organoleptic characteristics of the compound. Recently, Corma et al. [204] reported the synthesis of Nb- and Ta-Beta zeolites. These catalysts were also tested in the one-pot process described above for the synthesis of 4-methoxybenzyl 1-methylpropyl ether (Table 25.6). Ta-Beta exhibited similar activity and selectivity to that of Sn-Beta, while Nb-Beta resulted in a considerable lower selectivity to the target molecule. Sn-MCM-41, Sn-MCM-48, and Sn-SBA-15 [196, 205] containing mesoporous silica materials have been reported as active catalysts for the MPV reduction. Reduction of different carbonyl compounds with Sn-MCM-41 showed that this catalyst exhibits lower activity than Sn-Beta zeolite [196] and, particularly for the

807

808

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals Table 25.6 Results for the synthesis of 4-methoxybenzyl 1-methylpropyl ether by a tandem hydrogenation/etheriﬁcation sequence using solid Lewis acid catalystsa [203].

Catalyst (mg)

t (h)

Total conversion (%)

Overall selectivity to the ether (%)

Sn-Beta (50) Sn-Beta (100) Zr-Beta (50)

8 24 8

71 99 100

100 99 100

a Reaction

conditions: p-methoxybenzaldehyde (1.1 mmol), 2-butanol (3 g) at 100 ◦ C.

CHO

CH2OH

Zr(OR)x-MCM-41

+ O

O

CH2OH

Citral 79.3% conversion 87.8% selectivity

Geraniol

Scheme 25.40

+

CHO Toluene,110 °C

Oppenauer oxidation of geraniol with furfural.

reduction of substituted cyclic ketones, the selectivity to the thermodynamically favored trans-isomer was very high [205]. Also, grafting of zirconium isopropoxide on SBA-15 resulted in a highly active and selective catalyst for the MPV reduction. It was observed that the activity increases with Zr loading up to a monolayer coverage, while further loading did not lead to a signiﬁcant increase in activity. Additionally, the authors found that the catalyst was not deactivated by the presence of water, as occurs with the zirconium 1-propoxide as homogeneous catalysts [206]. De Bruyn et al. [207] reported the use of Zr and Hf catalysts immobilized in mesoporous materials such as MCM-41 and MCM-48 for the chemoselective MPV reduction of unsaturated ketones with alcohols. Liu et al. [208] found that grafted zirconium 1-propoxide on MCM-41 and silica gel were active catalysts in the Oppenauer oxidation of geraniol with furfural with high selectivity to the desired citral. However, when an acidic support such as Al-MCM-41 was used, the selectivity was low, since the catalyst promotes the dehydration and isomerization of geraniol (Scheme 25.40).

25.3 Base-Catalyzed Reactions

Basic zeolites can be synthesized through two main approaches: ion exchange with alkali-metal cations [209, 210] and generation of small clusters of alkali metals or

25.3 Base-Catalyzed Reactions

their oxides and alkaline-earth oxides within the pores of the zeolites. While the ﬁrst approach generates relatively weak basic sites, the second approach results in strong basic sites. In the case of ion-exchanged zeolites, the basicity is associated with Lewis sites, which corresponds to the framework oxygen atoms. Decreasing the electronegativity character of the nonframework compensating cations the negative charge over oxygen is enhanced, that is, the basicity of alkali-exchanged zeolites increases with the compensating cation size in the order Li < Na < K < Rb < Cs [210, 211]. According to Barthomeuf [212], the electron density on the oxygen atom also depends on other factors such as zeolite structure and chemical composition. It can generally be said that the lower the average Sanderson electronegativity of the system, the higher the basicity of the zeolite [213–215]. To enhance the basic properties of the zeolites, two types of materials have been synthesized: those containing metal clusters entrapped within the void spaces of the zeolites and those containing highly dispersed basic oxides encapsulated in the channels and cavities. This is accomplished by overexchanging the zeolite with alkali [216–219] or alkaline-earth metals salts [220, 221] followed by thermal decomposition. However, in spite of the higher basicity of these materials with respect to ion-exchanged zeolites, they suffer from important disadvantages such as their extreme sensitivity to be exposed to carbon dioxide and water, which are strongly adsorbed on the active sites deactivating the catalyst. These drawbacks have limited much of their use as catalysts for organic reactions. In contrast, exchanged zeolites can be handled in ambient atmosphere, since the absorption of carbon dioxide or water is not too strong and can be removed by high-temperature treatment. However, in spite of their advantages, the use of exchanged zeolites as basic catalysts in the synthesis of ﬁne chemicals is rather limited. This can be attributed to their weak basic character and the fact that their microporous structure limits their use in many chemical processes where bulky reactants are involved. However, their basic strength is in some cases adequate to catalyze reactions such as Knoevenagel and aldol condensations and Michael additions. In this section, we present some examples of industrial interest catalyzed by cation-exchanged and overexchanged zeolitic materials. 25.3.1 The Knoevenagel Condensation

The Knoevenagel reaction [222] is an important C–C bond-forming reaction that has been widely used in synthesis of important intermediate or end products in the ﬁne chemicals industry. The condensation involves the reaction between methylene-active compounds of the form Z–CH2 –Z or Z–CHR–Z with aldehydes or ketones (Scheme 25.41) and it is commonly promoted by bases such as ammonia, primary and secondary amines, and their salts [223]. Knoevenagel condensation between benzaldehyde and molecules containing activated methylene groups with different pKa values, such as ethyl cyanoacetate (pKa 9.0), ethyl acetoacetate (pKa 10.7), and ethyl malonate (pKa 13.3) has been used as a test reaction to measure not only the total number of basic sites but also

809

810

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

CN

R1 C

O + CH2

Base catalyst

C

Z

R2

R2

Z = −CN, −COOEt Scheme 25.41

CN

R1

+ H2O

C Z

Knoevenagel condensation.

their strength distribution [214, 224]. Thus, it was established that alkali-exchanged zeolites possess a large number of basic sites that are able to abstract a proton in organic molecules with pKa in the range 9 < pKa < 13.3 [224]. Also it was found that under heterogeneous catalysis the reaction mechanism is the same as under homogeneous catalysis, the controlling step of the condensation reaction on the basic zeolites being the attack by the carbanionic intermediate of the carbonyl group and not the proton abstraction, as was ﬁrstly proposed for other basic catalysts [175, 223]. The substitution of Si with Ge in the framework of faujasite signiﬁcantly increases the catalytic activity in Knoevenagel condensation [214]. NaGeX faujasite has been used as catalyst in the preparation of intermediate products for the synthesis of dihydropiridines. The Knoevenagel condensation between benzaldehyde derivatives and ethyl acetoacetate leads to 2-acetyl-3-phenylacrylic acid ester derivatives, which can then be converted into dihydropyridine derivatives in a second step by reaction with an aminocrotonic derivative (Scheme 25.42). These compounds are of interest clinically as potent calcium channel blockers and because of their antihypertensive activity. The Knoevenagel condensation step between 2-nitrobenzaldehyde and 2-triﬂuromethyl benzaldehyde with ethyl acetoacetate was performed using NaX zeolite and NaGeX as base catalysts, which was compared with those obtained using homogeneous bases such as pyridine or piperidine. Results showed that NaXGe faujasite is about three times more active than NaX zeolite, the selectivity being the same in both cases. However, the maximum yields achieved of the Knoevenagel adduct were between 32 and 37%. On the other hand, the activity of the NaGeX faujasite is higher than that of pyridine and lower than that of piperidine, whereas the selectivity was the same with homogeneous and heterogeneous catalysts. O COOEt R

CHO R

COCH3 + COOEt

COOR1 R OOC 1 Basic

NH2

catalyst R

R = NO2, CF3 Scheme 25.42

COOEt

Synthesis of dihydropyridine derivatives.

N H

R

25.3 Base-Catalyzed Reactions

The Knoevenagel condensation between benzaldehyde and ethyl cyanoacetate has also been investigated in the microreactor and a membrane microreactor using a Cs-exchanged zeolite X catalyst coated on the microchannel [225–227]. Both types of microreactors achieved better conversion than the conventional packed-bed reactor and packed-bed membrane. Recently, the same authors [228] have performed the Knoevenagel condensation between benzaldehyde and ethyl acetoacetate in a microreactor using Cs-exchanged NaX catalyst and NaA membrane. The authors showed that when using catalyst powder the laminar ﬂow and slow diffusion of the product molecules in the microchannel resulted in poor selectivity since the Knoevenagel adduct was further converted to different by-products. However, when using a catalyst ﬁlm instead of powder, the undesirable reactions were suppressed owing to the smaller external catalyst surfaces in contact with the solution. Water removal by membrane pervaporation increased the reaction conversion. Recently, the use of alkali metal (from Li to Cs) supported on Nb-MCM-41 in the synthesis of 1,4-dihydropyridine intermediates [229] has been reported. The authors found that niobium located in the mesoporous MCM-41 plays a role of structural promoter preventing the solid from the disordering caused by the treatment with alkali-metal solution. In addition, a change in the catalytic properties of the basic centers with respect to alkali metals supported on silicate MCM-41 was observed. For instance, using Rb/Nb-MCM41 catalysts, a conversion close to 85% (100% selectivity) was obtained after 4 hours in the Knoevenagel condensation of benzaldehyde with ethyl acetoacetate working at 140 ◦ C, while the Rb-loaded MCM-41 reached only 50% conversion. Preparation of prepolymers has been performed by Knoevenagel condensation of malononitrile and different ketones, namely, benzophenone, cyclohexanone, and p-aminoacetophenone, using alkali-exchanged zeolites. The reactivity depends on both the catalyst and the ketone structure. Thus, when malononitrile is condensed with cyclohexanone in the presence of a CsY zeolite, a very low level of conversion was achieved. However, CsX zeolite, which already has substantial number of basic sites, was able to abstract hydrogen atoms with pKa in the range 9 < pKa < 10.7, and some with 10.7 < pKa < 13.3 were able to perform the reaction under the same reaction conditions with high yield (82%) [230]. Citronitrile (5-pheyl-3-methyl 2-pentenenitrile) is a compound with citrus-like odor that has commercial interest for the cosmetic and fragrance industries. The synthesis involves in the ﬁrst step the Knoevenagel condensation between benzylacetone and ethyl cyanoatetate. In the second step, the Knoevenagel condensation adduct undergoes hydrolysis, followed by decarboxylation (Scheme 25.43). The Knoevenagel condensation between benzylacetone and ethyl cyanoatetate has been carried out with Cs-exchanged X zeolite and sepiolite and compared with metal oxides such as MgO and Al/Mg mixed oxide. The yields of the Knoevenagel adduct over zeolite and sepiolite were 39 and 49%, respectively, while in the case of the oxides the yield was much higher (70–75%), indicating that this reaction requires basic sites stronger than those present in cesium-exchanged zeolites and sepiolites [231]. Re-exchanged NaY zeolite has also been used as catalysts for performing the Knoevenagel condensation between a variety of substituted aromatic aldehydes such

811

812

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

O PhCH2CH2−C−CH3 + NC−CH2−COOEt

Catalyst

CH3 CN + H 2O

Ph−CH2CH2-C=C

COOEt

CH3

CH3 CN Hydrolysis

Ph−CH2CH2−C=C COOEt Scheme 25.43

−CO2

Ph−CH2CH2−C=CH−CN Citronitrile

Synthesis of citronitrile.

as p-nitrobenzaldehyde, bromobenzaldehyde, and vanillin and active methylene compounds such as malonitrile, ethylcyanoacetate, and 2-cyanoacetamide [232]. These reactions lead to completion in 8–12 hours at 20–60 ◦ C in acetonitrile as solvent, giving the Knoevenagel adduct in 62–87% yield. The role of multicatalytic active sites present inside the supercages is speculated to be the abstraction of the proton from the methylene-active compound by the basic sites, whereas the Lewis acidic site could be responsible for generating the partial positive charge on the carbonyl compound by coordinating with its oxygen, and thereby facilitating C–C bond formation. Recently, Martins et al. [233, 234] have prepared new basic zeolite catalysts, comprising ion-exchanged methylammonium cations in Y and X zeolite [235]. The authors found that the ion exchange of sodium by methylammonium cations reduces the micropore volume of the catalyst but enhances the strength of basic sites. The catalysts showed higher speciﬁc activity than Cs-faujasite zeolites for Knoevenagel condensations. For instance, during the Knoevenagel condensation of benzaldehyde with ethyl cyanoacetate using methylammonium-exchanged X zeolite, 83% conversion with 99% selectivity to the Knoevenagel adduct was achieved after 3 hours of reaction at 60 ◦ C, while only 50% conversion (100% selectivity) was achieved with Cs-exchanged X zeolite working under the same reaction conditions. Related to this study, more recently the authors have reported [236] the catalytic activity for the Knoevenagel condensation of a series of as-synthesized molecular sieves in which the organic structure-directing agents remain occluded in the pores of the molecular sieve. Conversion levels of 97% in the Knoevenagel condensation of butyraldehyde and ethyl cyanoacetate were achieved with aluminum-free large pore molecular sieves (MCM-41, MCM-48, and MCM-50), while aluminum zeolites showed lower activity. It is suggested that pure siliceous molecular sieves contain the highest number of siloxy anions that are the strong basic sites. Concerning the Knoevenagel condensation catalyzed by overexchanged CsY zeolites, Lasperas et al. [237–239] studied the condensation between benzaldehyde and ethyl cyanoacetate. The reaction was carried out under nitrogen atmosphere using dimethyl sulfoxide as solvent, achieving high selectivity to the Knoevenagel product (95% at 90% conversion). The results of activity were in good agreement with the TPD (temperature-programmed desorption) results, and Cs-overexchanged X

25.3 Base-Catalyzed Reactions

zeolites proved to be much more active than Cs-overexchanged Y zeolites as a result of the different cesium oxide species generated [237, 238]. 25.3.2 Michael Addition

Michael addition is another important reaction in synthetic chemistry leading to C–C bond formation. The reaction involves a conjugate or nucleophilic 1,4-addition of carbanions to α,β-unsaturated aldehydes, ketones, esters, nitriles, and sulfones [240] (Scheme 25.44). Base catalysts are used to form the carbanionic species by abstracting a proton from activated methylene precursors (donors) which attack the oleﬁn (acceptor). Strong bases are usually used for these reactions, which then also lead to the formation of by-products coming from side reactions such as condensations, dimerizations, or rearrangements. R1 + O Acceptor Scheme 25.44

Y

Y Base catalyst

CH2 Z Donor

R1 Z O Michael adduct

The Michael addition reaction.

Na-exchanged forms of Y and Beta zeolites performed efﬁciently the Michael addition of diketones (acetylacetone, methyl 2-oxocyclohexane carboxylate) and methylene-active compounds (diethylmalonate, ethylacetoacetate) and thiols (methyl thioglycolate) with methyl vinyl ketone, acrolein, and methyl acrylate giving the Michael adduct with high yield (70–80%). In contrast, when HY was used instead of NaY zeolite, the formation of the Michael adduct was low and the polymerization of the Michael acceptor was the main reaction [241]. Thia–Michael addition of thiols to α,β-unsaturated ketones leads to β-sulﬁdocarbonyl compounds that are important intermediates in the synthesis of many bioactive molecules [242]. Thia–Michael additions of benzenethiol to cyclic alkenones have been recently performed using cation-exchanged faujasites with different Si/Al ratios (Scheme 25.45). Reactions were performed in hexane solution at 0 ◦ C, and the results showed that NaY and KY zeolites were the best catalysts in terms of activity and selectivity to the Michael adduct (90–93% conversion, 100% selectivity) [243]. However, as with most basic X zeolites, although they also exhibit very good activity for the Michael addition, signiﬁcant side reactions take place in which tiols are oxidized to disulﬁde. The authors suggest that a bifunctional catalysis involving simultaneously the acid–base sites of zeolites with well-balanced acid–base strength is taking place in this reaction, since an excess of basicity or acidity in zeolites is detrimental for the reaction. Li-exchanged X zeolite has been used as basic catalysts for the preparation of 13-thiaprostaglandins, which are biologically active compounds [244]. These compounds were synthesized through the Michael addition of substituted 3-aryloxy-2-

813

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

814

O

O Ph−SH

n (H2C)

Zeolite

+ Ph−S−S−Ph

n (H2C) S−Ph

Scheme 25.45

Thia–Michael additions of benzenethiol to cyclic alkenones.

hydroxy-1-propanethiols to 2-substituted-4-hydroxycyclopentenone (Scheme 25.46). Reactions performed in dry chloroform at 0 ◦ C yielded the corresponding 13-thiaprostaglandins (60–80%). O

O

+ HO

(CH2)6COOMe

Li−X

(CH2)6COOMe OR

HS OH

HO

OR

S OH

Scheme 25.46

Synthesis of 13-thiaprostaglandins via Michael addition.

Baba et al. [245] prepared basic zeolites consisting of low-valent Yb or Eu species introduced into alkali-ion-exchanged Y zeolites by impregnation with Yb or Eu metal dissolved in liquid ammonia followed by heating under vacuum at about 197 ◦ C. The zeolites thus loaded with Yb or Eu species showed high catalytic activity for the Michael addition between cyclopent-2-enone with dimethyl malonate at 303 K without solvent, yielding after 20 hours the Michael adduct with 81 and 100% selectivity [246, 247]. Overexchanged Cs-Beta zeolite has been used as basic catalyst for the Michael additions of terpenoids with methylene-active compounds [248, 249]. The reaction products were the typical products of the Michael and Knoevenagel reactions or tandem transformations, which depend on the structure of the terpenoid. For instance, when the ketone 5,5,8-trimethylnona-3,7-dien-2-one (Scheme 25.47) reacts with malononitrile in the presence of overexchanged Cs-Beta zeolite, the Knoevenagel reaction competes with the Michael addition, in which Michael adduct undergoes a tandem transformation leading to a polyfunctional compound that is a mixture of two diastereoisomers. In addition, the authors reported the reaction presented in Scheme 25.47 using a chiral-modiﬁed Cs-Beta zeolite with l-methionine; although the terpenoid conversion was high, the enantiomeric excess of the ﬁnal compound was very low (ee. 6%) [250]. Metal-overexchanged MCM-41 has also been used as basic catalyst for Michael additions. Kloestra et al. [251] reported that strong basic sites are produced by impregnation of MCM-4 with cesium acetate, followed by a thermal treatment that produces highly dispersed Cs2 O particles. This material was tested in the Michael addition of neopentyl glycol diacrylate with diethyl malonate. The reactant

25.3 Base-Catalyzed Reactions NC

815

CN

O CN

+

Knoevenagel

CN Michael O

CN

CN CN CN

CN

CN − OH

CN CN

NH

NH CN

CN

Scheme 25.47

NC OH

NC CN

Michael addition of terpenoids with methylene-active compounds.

possessing two isolated activated double bonds can lead to a monoadduct and a bis-adduct (Scheme 25.48). The inﬂuence of the mesoporous framework of MCM-41 support was illustrated by the high regioselectivity for the Michael-type addition. Selectivities for the monoadduct of up to 98% were achieved at room temperature, while bulk Cs2 O shows a higher preference for the bis-adduct (100% selectivity). CH3 CH2=CHCOOC-C-CH2OCOCH=CH2

+ 2

CH3

COCH3

CH2-CH2COOC-C-CH2OCOCH=CH2 COOEt

CH3

EtOOC

neopentyl glycol diacrylate

COOEt

CH3

mono-adduct

CH3 CH-CH2COOC-C-CH2OCOCH-CH2 EtOOC

COOEt

CH3

EtOOC

COOEt

bis-adduct

Scheme 25.48

Michael addition of neopentyl glycol diacrylate with diethyl malonate.

The activity of cesium-overexchanged MCM-41 also has been tested for the base-catalyzed Michael addition of diethyl malonate to chalcone in a solvent-free system at 150 ◦ C [252]. Conversion of 87% and selectivity of 91% to the Michael adduct were obtained within 30 minutes. A disadvantage is the poor regeneration ability of the material, leading to a drastic decrease in the surface area and pore volume. The same authors have reported the catalytic activity of binary cesium–lanthanum oxides supported on MCM-41 for the Knoevenagel and Michael reactions [253]. CsLa-MCM-41 catalyzes the Michael addition of ethyl cyanoacetate to ethyl acrylate (Scheme 25.49) giving the corresponding monoadduct, which subsequently reacts to form a bis-adduct by a double Michael addition, decreasing the selectivity to

816

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals OEt OEt

+

NC

O

OEt

O O

OEt

O

OEt

CN

CN COOEt

O

EtOOC

Scheme 25.49

COOEt

Michael addition of ethyl cyanoacetate to ethyl acrylate.

monoadduct. It was found that basicity of CsLa-MCM-41 is quite mild, and the catalyst was unable to perform the Michael addition of diethyl malonate under the same conditions. 25.3.3 Aldol Condensations

Natural clinoptilolite zeolite and modiﬁed samples by either acid treatment or ion exchange have been successfully used in aldol condensation between acetaldehyde and formaldehyde [254]. The aldol condensation of aqueous formaldehyde and acetaldehyde in gas phase leads to acrolein (an important intermediate in organic chemistry), crotonaldehyde (from the self-condensation of acetaldehyde), and a variety of by-products such as light hydrocarbons, ethanol, and carbon monoxide. The results indicate that a direct relationship between the acid–base characteristic and catalytic activity exists and, although both acid and basic sites activate the two carbonyl groups, basic sites govern the reaction network. The most basic Ca-exchanged zeolite was the most active, selective, and stable catalyst for this condensation Thus, when the reaction was performed at 350 ◦ C, the conversion of acetaldehyde was 75%, the selectivity to acrolein being 70%, crotonaldehyde 25%, and light aromatics 5%. Glyceraldehyde acetonide (2,3-isopropylidenglyceraldehyde), a natural compound with a great applicability in the ﬁeld of organic chemistry, was condensed with acetone in the presence of an overexchanged CsX zeolite, giving the corresponding α,β-unsaturated carbonyl compound [255] (Scheme 25.50). When the reaction was carried in liquid phase with a molar ratio acetone/glyceraldehyde acetonide equal to 30 at 50 ◦ C, 50.7% of conversion with 93.9% selectivity to the corresponding aldol condensation product was obtained after 4 hours reaction time. Similar conversion was achieved using an Mg–Al mixed oxide as base catalyst; however, lower selectivity (65.8%) to the condensation product was achieved while a signiﬁcant amount diacetone alcohol was also formed. Flavonoids are an important class of natural compounds that have a variety of pharmaceutical applications such as anti-inﬂammatory, antibacterial, anticancer, and anti-AIDS. Among them, ﬂavanone, which is an intermediate in the synthesis

25.3 Base-Catalyzed Reactions O

O CHO

O

O

Cs−X

O

O

Glyceraldehyde acetonide

Scheme 25.50

Aldol condensation of glyceraldehyde acetonide with acetone.

of many pharmaceuticals, is obtained by the Claisen–Schmidt condensation between 2-hydroxyacetophenone and benzaldehyde to give 2-hydroxychalcone, which in a second step can undergo intramolecular cyclization to yield the ﬂavanone (Scheme 25.51). Since both reactions can be catalyzed by acid and base catalysts, Saravanamurugan et al. [256] have been studied the reaction of 2-hydroxyacetophenone and benzaldehyde in liquid phase in the presence of acid zeolite (HZSM-5) as well as ion-exchanged Mg-ZSM-5 and Ba-ZSM-5 as base catalysts. The order of catalyst activity was Mg-ZSM-5 > Ba-ZSM-5 > HZSM-5. Working at 140 ◦ C in dimethyl sulfoxide (DMSO) as a solvent, of all the catalysts tested the conversion of 2-hydroxyacetophenone attained a maximum (40–50%) achieving similar selectivities to 2 -hydroxychalcone plus ﬂavanone. On increasing the reaction temperature up to 160 ◦ C, 85% conversion of 2-hydroxyacetophenone with selectivities of 65 and 30% to 2-hydroxychalcone and ﬂavanone, respectively, were obtained. Kloestra et al. [252] reported that Na-exchanged MCM-41 was also active to perform the aldol condensation between benzaldehyde and different bulky ketones including the Claisen–Schmitd condensation of benzaldehyde and 2’-hydroxyacetophenone to give the chalcone plus the corresponding ﬂavanone. Recently, the Claisen–Schmidt condensation between benzaldehyde and acetophenone has been performed using ion-exchanged methylammonium cations in Y and X faujasites. The methylammonium-exchanged X zeolite exhibited modest activity, achieving 32% benzaldehyde conversion (100% selectivity) after 6 hours at 140 ◦ C; however, it was superior to the of Cs-exchanged homolog [235]. 2-Methylpentenal is an important chemical with application in fragrances, ﬂavors, and cosmetics and as intermediate for the synthesis of various pharmacologically active compounds. Nowadays, it is synthesized by self-condensation of propanal in the presence of liquid bases such as KOH or NaOH in stoichiometric amounts. Recently, research efforts were directed to develop a catalytic process that can produce 2-methylpentenal from propanal by employing heterogeneous reusable base catalysts. Sharma et al. [257] reported the synthesis CHO

OH +

+ H2O

CH3 O

O

OH Base catalyst

O 2′-hydroxychalcone

Scheme 25.51 Claisen–Schmidt condensation of benzaldehyde and 2 -hydroxyacetophenone.

O Flavanone

817

818

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals

of 2-methylpentenal by aldol condensation of propanal (Scheme 25.52) using various alkali-ion-exchanged zeolites (CsX, RbX, KX, NaX), alkali-treated alumina, and hydrotalcites under solvent-free conditions. In all cases, 2-methylpentenal and 3-hydroxy-2-methylpentenal were the products detected. The conversion of propanal varied from 22 to 42% with 92–94% selectivity to 2-methylpentenal using different alkali-ion-exchanged zeolite without any thermal treatment or activation at 450 ◦ C. Catalytic activities obtained follow the order CsX > RbX > KX > NaX, which is in agreement with the order of basicity reported in the literature. With the less polarized cation such as CsX zeolite activated at 450 ◦ C, 99% selectivity to 2-methylpentenal at 38% conversion was obtained. CHO

2

Scheme 25.52

Base catalyst 100 °C

CHO

Aldol condensation of propanal.

Jasminaldehyde, a compound with violet scent widely used in cosmetics and fragrances, has been obtained by the cross-aldol condensation of benzaldehyde with n-heptaldehyde (Scheme 25.53) using a mesoporous molecular sieve Al-MCM-41-supported MgO. The reactions were carried out in a stirred autoclave reactor using a molar ratio benzaldehyde/heptanal of 10 at 100–175 ◦ C [258]. The results showed that although Al-MCM-41 exhibits catalytic activity, its activity is signiﬁcantly increased by the deposition of MgO. Increasing the amount of MgO deposited on Al-MCM-41 enhances the catalytic activity, although the selectivity to jasminaldehyde remains constant. CHO Ph−CHO + CH2− (CH2)4−CH3

Base

Ph−CH=C−CHO

+ CH3−(CH2)5−CH=C−CHO

CH2 (CH2)3CH3 Jasminaldehyde Scheme 25.53

CH2 (CH2)3CH3 2-n-pentyl-2-n -nonenal

Cross-aldol condensation of benzaldehyde with n-heptaldehyde.

For the same reaction, Jaenicke et al. [259] tested the activity of K2 O, BaO, and K2 O/La2 O3 incorporated into MCM-41 working at 140 ◦ C and using a benzaldehyde/heptanal molar ratio of 1.5. The best results were obtained with the most basic catalyst, that is, the binary inorganic catalysts K2 O/La2 O3 -MCM-41, achieving heptanal conversion of 61.9% with 53% selectivity to jasminaldehyde after 2 hours. The authors suggested that the presence of the rare earth not only improves the thermal stability of the material but also has a positive effect on the product selectivity. It is interesting to mention that much better results were obtained for the above reaction when Al-MCM-41 was used as catalyst in a one-pot reaction in which an intermediate dimethylacetal of heptanal was prepared and hydrolyzed in situ [260]. Thus, reactions performed using a benzaldehyde/heptanal molar ratio of 1.5 at 100 ◦ C give selectivity to jasminaldehyde higher than 90% for conversions larger than 80%.

References

25.4 Summary and Outlook

We have seen that acid zeolites (either with Br¨onsted or Lewis acid sites) have been largely applied for carbonium ion reactions as well as for oxidation reactions in the ﬁeld of organic synthesis for ﬁne chemicals. The fact that well-deﬁned, single isolated sites with tunable acidities can be generated offers unique possibilities for reactant activation. Furthermore, the possibility to prepare zeolite samples with controlled polarity can be determinant in some cases to avoid the use of solvents and to facilitate product desorption. The zeolite pores with well-deﬁned dimensions and topologies can be, at the same time, a blessing and a punishment. Indeed, the well-deﬁned pore shape can stabilize certain transition states, thereby increasing the reactivity and introducing shape selectivity. At the same time, the formation of relatively bulky products with slow diffusion rate out of the pores, especially at the lower reaction temperatures normally used, can lead to occlusion of the micropores and rapid catalyst deactivation. This necessitates frequent regenerations and sometimes preclude the use of processes using ﬁxed-bed, continuous reactors. In these cases, the use of nanocrystalline and delaminated zeolites can be of interest. It can also be of interest to synthesize extra-large pore zeolites that should facilitate the diffusion of larger reactants and products. In the case of basic catalysis with zeolites, the problem arises from the fact that the competing catalysts are NaOH and KOH. These are inexpensive and the residues (salts) are easily handleable. Opportunity for zeolites will arise from the special selectivity effects. Finally, we believe that zeolites and, in general, structured molecular sieves offer the still little explored possibility to design multisite catalysts that can lead to process intensiﬁcation by selectively performing one-pot or cascade reactions.

References 1. Corma, A. (1995) Chem. Rev., 95, 559. 2. Venuto, P.B. and Landis, P.S. (1968)

Adv. Catal., 8, 259. 3. Venuto, P.B. (1994) Microporous Mater., 4. 5. 6.

7.

2, 297. Corma, A. and Garcia, H. (1997) Catal. Today, 38, 257. Sheldon, R.A. and Downing, R.S. (1999) Appl. Catal., 189, 163. Derouan, E.G. (ed.) (2006) Catalysis for Fine Chemicals: Microporous and Mesoporous Solid Catalysts, John Wiley & Sons, Ltd, Chichester. Chiche, B., Finiels, A., Gauthier, C., Geneste, P., Graille, J., and Pioch, D. (1986) J. Org. Chem., 51, 2128.

8. Bejblova, M., Prochazkova, D., and

ˇ Cejka, J. (2009) ChemSusChem, 2, 486. 9. Sartori, G. and Maggi, R. (2006) Chem.

Rev., 106, 1077. 10. Sun, Y. and Prins, R. (2008) Appl.

Catal., A, 336, 11. 11. Ji, X., Qin, Z., Dong, M., Wang, G.,

Dou, T., and Wang, J. (2007) Catal. Lett., 117, 171. 12. Wine, G., Pham-Huu, C., and Ledoux, M.J. (2006) Catal. Commun., 7, 768. 13. Zhao, D., Wang, J., and Zhang, J. (2008) Catal. Lett., 126, 188. 14. Choudary, B.M., Sateesh, M., Kantam, M.I., and Prasad, K.V.R. (1998) Appl. Catal., A, 171, 155.

819

820

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals 15. Derouane, E.G., Dillon, C.J., Bethell,

16. 17. 18.

19.

20.

21. 22.

23.

24.

25.

26.

27.

28.

D., and Derouane-Abd Hamid, S.B. (1999) J. Catal., 187, 209. Freese, U., Heinrich, F., and Roessner, F. (1999) Catal. Today, 49, 237. Gaare, K. and Akporiaye, D. (1996) J. Mol. Catal. A, 109, 177. Guisnet, M. and Guidotti, G. (2006) Aromatic Acetylation in Catalysis for Fine Chemicals Synthesis: Microporous and Mesoporous Solid Catalysts, Vol. 4 (ed. E.G. Derouan), John Wiley & Sons, Inc., pp. 69–94. Harvey, G., Vogt, A., Kouwenhoven, H.W., and Prins, R. (1993) Performance of zeolite Beta in Friedel-Crafts reactions of functionalized aromatics in Proceedings of the 9th International Zeolite Conference (eds R. Von Ballmoos, J.B. Higgins, and M.M.J. Treacy), Butterworth-Heinemann, Boston. Rohan, D., Canaff, C., Formentin, E., and Guisnet, M. (1998) J. Catal., 177, 296. Smith, K., Zhenhua, Z., and Hodgson, P.K.G. (1998) J. Mol. Catal., 134, 121. Spagnol, M., Benazzi, E., and Marcilly, C. (1998) US Patent 5817878 (Rhone-Polenc). Wang, Q.L., Ma, Y.D., Ji, X.D., Yan, H., and Qiu, Q. (1995) Chem. Commum., 2307. Kantam, M.L., Ranganath, K.V.S., Sateesh, M., Kumar, K.B.S., and Choudary, B.M. (2005) J. Mol. Catal. A: Chem., 225, 15. Corma, A., Climent, M.J., Garcia, H., and Primo, J. (1989) Appl. Catal. A, 49, 109. Balasubramanian, V.V., Srinivasu, P., Anand, C., Pal, R.R., Ariga, K., Velmathi, S., Alam, S., and Vinu, A. (2008) Microporous Mesoporous Mater., 114, 303. Vinu, A., Justus, J., Anand, C., Sawant, D.P., Ariga, K., Mori, T., Srinivasu, P., Balasubramanian, V.V., Velmathi, S., and Alam, S. (2008) Microporous Mesoporous Mater., 116, 115. Andy, P., Garcia-Martinez, J., Lee, G., Gonzalez, H., Jones, C.W., and Davis, M.E. (2000) J. Catal., 192, 215.

29. Botella, P., Corma, A., and Sastre, G.

(2001) J. Catal., 197, 81. 30. Fromentin, E., Coustard, J.M., and

Guisnet, M. (2000) J. Catal., 190, 433. 31. Fromentin, E., Coustard, J.M., and

32. 33. 34. 35.

36.

37.

38.

39. 40.

41. 42.

43. 44. 45. 46. 47.

48. 49.

Guisnet, M. (2000) J. Mol. Catal. A, 159, 377. Heinichen, H.K. and Holderich, W.F. (1999) J. Catal., 185, 408. Hwang, K.Y. and Rhu, H.K. (2003) React. Kinet. Catal. Lett., 79, 189. Moreau, P., Finiels, A., Meric, P., and Fajula, F. (2003) Catal. Lett., 85, 199. Guidotti, M., Canaff, C., Coustard, J.M., Magnoux, P., and Guisnet, M. (2005) J. Catal., 230, 375. Botella, P., Corma, A., Navarro, M.T., Rey, F., and Sastre, G. (2003) J. Catal., 217, 406. Yuan, B., Li, Z., Liu, Y., and Zhang, S. (2008) J. Mol. Catal. A: Chem., 280, 210. Holderich, W.F., Hesse, M., and Naumann, F. (1988) Angew. Chem. Int., 27, 226. Richard, F., Carreyre, H., and Perot, G. (1996) J. Catal., 159, 427. Alvaro, V.F.D., Brigas, A.F., Derouane, E.G., Lourenco, J.P., and Santos, B.S. (2009) J. Mol. Catal. A: Chem., 305, 100. Lerner, H., Holderich, W.F., and Schwarzmann, M. (1987) DE 3618964. Finiels, A., Calmette, A., Geneste, P., and Moreau, P. (1993) Stud. Sci. Catal., 83, 379. Isaev, Y. and Fripiat, J.J. (1999) J. Catal., 182, 257. Vogt, A. and Pfenninger, A. (1996) EP 0701987A1. Padro, C.L. and Apesteguia, C.R. (2004) J. Catal., 226, 308. Padro, C.L. and Apesteguia, C.R. (2005) Catal. Today, 107-108, 258. Kuriakose, G., Nagy, J.B., and Nagaraju, N. (2004) Stud. Surf. Sci. Catal., 154C, 2803. Hoefnagel, A.J. and van Bekkum, H. (1993) Appl. Catal., A, 97, 87. van Bekkum, H., Hoefnagel, A.J., Vankoten, M.A., Gunnewegh, E.A., Vogt, A.H.G., and Kouwenohoven, H.W. (1994) in Zeolites and Microporous Crystals, Studies in Surface Science

References

50. 51.

52.

53.

54. 55.

56.

57.

58.

59.

60.

61.

62.

63.

64. 65.

Catalysis, Vol. 83 (eds T. Hattori and T. Yashima), Elsevier, Amsterdam, p. 379. Venuto, P.B. and Landis, P.S. (1966) J. Catal., 6, 237. Climent, M.J., Corma, A., Garcia, H., and Primo, J. (1989) Appl. Catal., A, 51, 113. Climent, M.J., Corma, A., Garcia, H., and Primo, J. (1991) J. Catal., 130, 138. Climent, M.J., Corma, A., Garcia, H., Iborra, S., and Primo, J. (1995) Appl. Catal., 130, 5. Burgers, M.H.V. and van Bekkum, H. (1993) Stud. Surf. Sci. Catal., 78, 567. Moreau, C., Fajula, F., Finiels, A., Razigade, S., Gilbert, L., Jacquor, R., and Spagnol, M. (1998) Chemical Industries, 75, 51. Armandi, M., Bonelli, B., Garrone, E., Ardizzi, M., Cavani, F., Dal Pozzo, L., Maselli, L., Mezzogori, R., and Calestani, G. (2007) Appl. Catal., B, 70, 585. Bolognini, M., Cavani, F., Dal Pozzo, L., Maselli, L., Zaccarelli, F., Bonelli, B., Armandi, M., and Garrone, E. (2004) Appl. Catal., A, 272, 115. Cavani, F., Corrado, M., and Mezzogori, R. (2002) J. Mol. Catal. A: Chem., 182-183, 447. Kumbar, S.M., Shanbhag, G.V., and Halligudi, S.B. (2006) J. Mol. Catal. A: Chem., 244, 278. Boulet, O., Emo, R., Faugeras, P., Jobelin, I., Laport, F., Lecomte, J., Moreau, C., Roux, M.C., Roux, G., and Simminger J. (1995) FR, 9513829. Lecomte, J., Finiels, A., Geneste, P., and Moreau, C. (1998) J. Mol. Catal. A, 133, 283. Lecomte, J., Finiels, A., Geneste, P., and Moreau, C. (1998) Appl. Catal. A: Gen., 168, 235. Lecomte, J., Finiels, A., Geneste, P., and Moreau, C. (1999) J. Mol. Catal. A, 140, 157. Finiels, A., Balmer, W., and Moreau, C. (2001) Stud. Surf. Sci. Catal., 135, 230. March, J. and Smith, K. (2001) Advanced Organic Chemistry, John Wiley & Sons, Inc., New York.

66. Ishihara, K. and Yamamoto, H. (1997)

Cattech, 51. 67. Landis, P.S. (1968) US Patent 3359285

(Mobil Oil Corp). 68. Ipaktschi, J. (1986) Z. Naturforsch. B,

41, 496. 69. Narayana Murthy, Y.V.S. and Pillai,

C.N. (1991) Synth. Commun., 21, 783. 70. Pindur, U. and Haber, M. (1991)

Heterocycles, 32, 1463. 71. Durand, R., Geneste, P., Joffre, J.,

72. 73.

74.

75.

76.

77.

78.

79.

80.

81.

82.

83.

84.

and Moreau, C. (1993) Stud. Surf. Sci. Catal., 78, 647. Yin, D.H., Yin, D.L., and Li, Q.H. (1996) Chin. Chem. Lett., 7, 697. Liu, J., Yin, D., Yin, D., Fu, Z., Li, Q., and Lu, G. (2004) J. Mol. Catal. A: Chem., 209, 171. Onaka, M., Hashimoto, N., Yamasaki, R., and Kitabata, Y. (2002) Chem. Lett., 166. Onaka, M., Hashimoto, N., Kitabata, Y., and Yamasaki, R. (2003) Appl. Catal., A, 241, 307. Zendehdel, M., Far, N.F., and Gaykani, Z. (2005) J. Inclusion Phenom. Macrocyclic Chem., 53, 47. Kugita, T., Jana, S.K., Owada, T., Hashimoto, N., Onaka, M., and Namba, S. (2003) Appl. Catal., A, 245, 353. Satsuma, A., Segawa, Y., Yoshida, H., and Hattori, T. (2004) Appl. Catal., A, 264, 229. Gomez, M.V., Cantin, A., Corma, A., and de la Hoz, A. (2005) J. Mol. Catal. A: Chem., 240, 16. Green, T.W. and Wuts, P.G.M. (1991) Protective Groups in Organic Synthesis, John Wiley & Sons, Inc., New York. Bauer, K., Garbe, D., and Surburg, H. (1990) Common Fragrances and Flavors Materials, Wiley-VCH Verlag GmbH, New York. Climent, M.J., Corma, A., Garcia, H., and Primo, J. (1990) Appl. Catal., 59, 333. Climent, M.J., Corma, A., Iborra, S., Navarro, M.C., and Primo, J. (1996) J. Catal., 161, 783. Rodriguez, I., Climent, M.J., Iborra, S., Fornes, V., and Corma, A. (2000) J. Catal., 192, 441.

821

822

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals 85. Thomas, B., Prathapan, S., and

86.

87.

88. 89. 90.

91.

92. 93.

94. 95. 96.

97.

98. 99. 100.

101. 102. 103.

104.

Sugunan, S. (2004) Appl. Catal., A, 277, 247. Thomas, B. and Sugunan, S. (2005) Indian J. Chem. A: Inorg., Bio-inorg., Phys., Theor. Anal. Chem., 44A, 1345. Thomas, B., Prathapan, S., and Sugunan, S. (2005) Microporous Mesoporous Mater., 80, 65. Thomas, B. and Sugunan, S. (2006) J. Porous Mater., 13, 99. Jermy, B.R. and Pandurangan, A. (2005) Appl. Catal., A, 295, 185. Jermy, B.R. and Pandurangan, A. (2006) J. Mol. Catal. A: Chem., 256, 184. Climent, M.J., Corma, A., Velty, A., and Susarte, M. (2000) J. Catal., 196, 345. Yuan, C., Zhang, F., Wang, J., and Ren, X. (2005) Catal. Commun., 6, 721. Zhang, F., Yuan, C., Wang, J., Kong, Y., Zhu, H., and Wang, C. (2006) J. Mol. Catal. A: Chem., 247, 130. Climent, M.J., Velty, A., and Corma, A. (2002) Green Chem., 4, 565. Climent, M.J., Corma, A., and Velty, A. (2004) Appl. Catal., A, 263, 155. Corma, A., Fornes, V., Pergher, S.B., Maesen, T.L.M., and Bugglas, J.G. (1998) Nature, 376, 353. Corma, A., Diaz, U., Fornes, V., Guil, J.M., Martinez-Triguero, J., and Creyghton, E.J. (2000) J. Catal., 191, 218. Corma, A., Iborra, S., Miquel, S., and Primo, J. (1996) J. Catal., 161, 713. Chapat, J.F. and Moreau, C. (1998) Carbohydr. Lett., 3, 25. Camblor, M.A., Corma, A., Iborra, S., Miquel, S., Primo, J., and Valencia, S. (1997) J. Catal., 172, 76. Chapat, J.F., Finiels, A., Joffre, J., and Moreau, C. (1999) J. Catal., 185, 445. Corma, A., Iborra, S., Miquel, S., and Primo, J. (1998) J. Catal., 180, 218. Climent, M.J., Corma, A., Iborra, S., Miquel, S., Primo, J., and Rey, F. (1999) J. Catal., 183, 76. de Goede, T.J.W., van der Leij, Y.G., van der Heijden, A.M., van Rantwijk, F., and van Bekkum, H. (1996) WO, 9636640.

105. van der Heijden, A.M., van Rantwijk,

106.

107.

108. 109.

110.

111.

112.

113.

114.

115.

116. 117.

118. 119.

120.

121. 122. 123.

F., and van Bekkum, H. (1999) J. Carbohydr. Chem., 18, 131. van der Heijden, A.M., Lee, T.C., van Rantwijk, F., and van Bekkum, H. (2002) Carbohydr. Res., 337, 1993. Rudakov, G.A., Ivanova, L.S., Pisareva, T.N., and Borovskaya, A.G. (1975) Gidroliz. Lesokhim. Prom-st, 4, 7. Severino, A., Vital, J., and Lobo, L.S. (1993) Stud. Surf. Sci. Catal., 78, 685. Yilmaz, S., Ucar, S., Artok, L., and Gulec, H. (2005) Appl. Catal. A: Gen., 287, 261. Gunduz, G., Dimitrova, R., Yilmaz, S., and Dimitrov, L. (2005) Appl. Catal. A: Gen., 282, 61. Gunduz, G., Dimitrova, R., Yilmaz, S., Dimitrov, L., and Spassova, M. (2005) J. Mol. Catal. A, 225, 253. Dimitrova, R., Gunduz, G., and Spassova, M. (2006) J. Mol. Catal. A, 243, 17. Lopez, C.M., Machado, F.J., Rodriguez, K., Mendez, B., Hasegawa, M., and Pekerar, S. (1998) Appl. Catal. A: Gen., 173, 75. Akpolat, O., Gunduz, G., Ozkan, F., and Besun, N. (2004) Appl. Catal. A: Gen., 265, 11. Allahverdiev, A.I., Gunduz, G., and Murzin, D.Y. (1998) Ind. Eng. Chem. Res., 37, 2373. Allahverdiev, A.I., Irandoust, S., and Murzin, D.Y. (1999) J. Catal., 185, 352. Allahverdiev, A.I., Irandoust, S., Andersson, B., and Murzin, D.Y. (2000) Appl. Catal. A: Gen., 198, 197. Findik, S. and Gunduz, G. (1997) J. Am. Oil Chem. Soc., 74, 1145. Ozkan, F., Gunduz, G., Akpolat, O., Besun, N., and Murzin, D.Y. (2003) Chem. Eng. J., 91, 257. Kaminska, J., Schwegler, M.A., Hoefnagel, A.J., and Vanbekkum, H. (1992) Rec. Trav. Chim. Pays-Bas, 111, 432. Schulte, E., Karl, H.Dr., and Muller, B. (1988) EP, 155591. Arbusow, B. (1935) Chem. Ber., 68, 1430. Holderich, W.F., Roseler, J., Heitmann, G., and Liebens, A.T. (1997) Catal. Today, 37, 353.

References 124. Lewis, J.B. and Hedrick, G.W. (1965) J. 125.

126.

127.

128. 129. 130. 131. 132. 133.

134.

135.

136.

137.

138.

139. 140.

141.

142.

Org. Chem., 30, 4271. Kunkeler, P.J., van der Waal, J.C., Bremmer, J., Zuurdeeg, B.J., Downing, R.S., and van Bekkum, H. (1998) Catal. Lett., 53, 135. Ravindra, D.B., Nie, Y.T., Jaenicke, S., and Chuah, G.K. (2004) Catal. Today, 96, 147. Sheldon, R.A. and Van Vliet, M.C.A. (2001) Oxidation in Fine Chemicals through Heterogeneous Catalysis (eds R.A. Sheldon and H. van Bekkum), Wiley-VCH Verlag GmbH, New York, p. 473. Clerici, M.G. and Ingallina, P. (1993) J. Catal., 140, 71. Clerici, M.G. and Ingallina, P. (1998) Catal. Today, 41, 351. Notari, B. (1993) Catal. Today, 18, 163. Notari, B. (1996) Adv. Catal., 41, 253. Tarramasso, M., Perego, G., and Notari, B. (1983) US Patent 4410501. Corma, A., Navarro, M.T., and Pariente, J.P. (1994) J. Chem. Soc., Chem. Commun., 147. Blasco, T., Camblor, M.A., Corma, A., Esteve, P., Guil, J.M., Martinez, A., Perdigon-Melon, J.A., and Valencia, S. (1998) J. Phys. Chem. B, 102, 75. Blasco, T., Corma, A., Navarro, M.T., and Pariente, J.P. (1995) J. Catal., 156, 65. Tatsumi, T., Koyano, K.A., and Igarashi, N. (1998) Chem. Commun., 325. Pena, M.L., Dellarocca, V., Rey, F., Corma, A., Coluccia, S., and Marchese, L. (2001) Microporous Mesoporous Mater., 44, 345. Corma, A., Jord´a, J.L., Navarro, M.T., Perez-Pariente, J., Rey, F., and Tsuji, J. (2000) Stud. Surf. Sci. Catal., 129, 169. Corma, A., Iborra, S., and Velty, A. (2007) Chem. Rev., 107, 2411. Chiker, F., Nogier, J.P., Launay, F., and Bonardet, J.L. (2003) Appl. Catal. A: Gen., 243, 309. Chiker, F., Launay, F., Nogier, J.P., and Bonardet, J.L. (2003) Green Chem., 5, 318. Kapoor, M.P. and Raj, A. (2000) Appl. Catal. A: Gen., 203, 311.

143. On, D.T., Kapoor, M.P., Joshi, P.N.,

144.

145.

146.

147.

148.

149. 150.

151.

152.

153.

154.

155. 156. 157. 158.

159. 160.

161. 162.

Bonneviot, L., and Kaliaguine, S. (1997) Catal. Lett., 44, 171. Suh, Y.W., Kim, N.K., Ahn, W.S., and Rhee, H.K. (2001) J. Mol. Catal. A, 174, 249. Suh, Y.W., Kim, N.K., Ahn, W.S., and Rhee, H.K. (2003) J. Mol. Catal. A, 198, 309. Prada, N.Q., Stashenko, E., Paez, E.A., and Martinez, J.R. (1999) Rev. Colomb. Quim., 28, 45. Adam, W., Corma, A., Reddy, T.I., and Renz, M. (1997) J. Org. Chem., 62, 3631. Clerici, M.G., Bellussi, G., and Romano, U. (1991) J. Catal., 129, 159. Hutter, R., Mallat, T., and Baiker, A. (1995) J. Catal., 153, 177. Corma, A., Domine, M., Gaona, J.A., Jorda, J.L., Navarro, M.T., Rey, F., Perez-Pariente, J., Tsuji, J., McCulloch, B., and Nemeth, L.T. (1998) Chem. Commun., 2211. van der Waal, J.C., Rigutto, M.S., and van Bekkum, H. (1998) App. Catal. A: Gen., 167, 331. Adam, W., Saha-Moller, C.R., and Weichold, O. (2000) J. Org. Chem., 65, 2897. Corma, A., Iglesias, M., and Sanchez, F. (1995) J. Chem. Soc., Chem. Commun., 1653. Kumar, R., Pais, G.C.G., Pandey, B., and Kumar, P. (1995) J. Chem. Soc., Chem. Commun., 1315. Krow, G.C. (1993) Org. React., 43, 251. Jimenez-Sanchidrian, C. and Ruiz, J.R. (2008) Tetrahedron, 64, 2011. Fischer, J. and Holderich, W.F. (1999) Appl. Catal., A, 180, 435. Lenarda, M., Da Ros, M., Casagrande, M., Storaro, L., and Ganzerla, R. (2003) Inorg. Chim. Acta, 349, 195. Bhaumik, A., Kumar, P., and Kumar, R. (1996) Catal. Lett., 40, 47. Wang, Z.B., Mizusaki, T., Sano, T., and Kawakami, Y. (1997) Bull. Chem. Soc. Jpn., 70, 2567. Corma, A. and Garcia, H. (2002) Chem. Rev., 102, 3837. Wu, P., Liu, Y., He, M., and Tatsumi, T. (2004) J. Catal., 228, 183.

823

824

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals 163. Corma, A., Nemeth, L.T., Renz, M.,

164.

165.

166.

167.

168.

169.

170.

171. 172.

173.

174.

175.

176. 177.

178. 179.

180.

and Valencia, S. (2001) Nature, 412, 423. Corma, A., Jorda, J.L., Navarro, M.T., and Rey, F. (1998) Chem. Commun., 1899. Corma, A., Navarro, M.T., Nemeth, L., and Renz, M. (2001) Chem. Commun., 2190. Bare, S.R., Kelly, S.D., Sinkler, W., Low, J.J., Modica, F.S., Valencia, S., Corma, A., and Nemeth, L.T. (2005) J. Am. Chem. Soc., 127, 12924. Boronat, M., Corma, A., Renz, M., Sastre, G., and Viruela, P.M. (2005) Chem. Eur. J., 11, 6905. Boronat, M., Concepcion, P., Corma, A., Renz, M., and Valencia, S. (2005) J. Catal., 234, 111. Boronat, M., Corma, A., Renz, M., and Viruela, P.M. (2006) Chem. Eur. J., 12, 7067. Boronat, M., Concepcion, P., Corma, A., Navarro, M.T., Renz, M., and Valencia, S. (2009) Phys. Chem. Chem. Phys., 11, 2876. Corma, A. and Renz, M. (2005) Collect. Czech. Chem. Commun., 70, 1727. Renz, M., Blasco, T., Corma, A., Fornes, V., Jensen, R., and Nemeth, L. (2002) Chem. Eur. J., 8, 4708. Corma, A., Fornes, V., Iborra, S., Mifsud, M., and Renz, M. (2004) J. Catal., 221, 67. Bauer, K., Garbe, D., and Surburg, H. (1997) Common Fragrances and Flavors Materials, Wiley-VCH Verlag GmbH, New York. March, J. (1992) Advanced Organic Chemistry, John Wiley & Sons, Inc., New York. Corma, A., Iborra, S., Mifsud, M., and Renz, M. (2005) J. Catal., 234, 96. Corma, A., Iborra, S., Mifsud, M., Renz, M., and Susarte, M. (2004) Adv. Synth. Catal., 346, 257. Corma, A., Navarro, M.T., and Renz, M. (2003) J. Catal., 219, 242. Corma, A., Iborra, S., Mifsud, M., and Renz, M. (2005) ARKIVOC, Vol. 9, 124. Nekoksova, I., Zilkova, N., Zukal, A., ˇ and Cejka, J. (2005) Stud. Surf. Sci. Catal., 156, 779.

181. Kawabata, T., Ohishi, Y., Itsuki, S.,

182.

183.

184.

185.

186.

187. 188.

189.

190.

191.

192.

193.

194. 195.

Fujusaki, N., Shisido, T., Takaki, K., Zhang, Q., Wang, Y., and Takejira, K. (2005) J. Mol. Catal. A, 236, 99. Subramanian, H. and Koodali, R.T. (2008) React. Kinet. Catal. Lett., 95, 239. Nowak, I., Feliczak, A., Nekoksova, I., ˇ and Cejka, J. (2007) Appl. Catal., A, 321, 40. Graauw, C.F., Peters, J.A., van Bekkum, H., and Huskens, J. (1994) Synthesis, 10, 1007. Chuah, G.K., Jaenicke, S., Zhu, Y.Z., and Liu, S.H. (2006) Curr. Org. Chem., 10, 1639. Creyghton, E.J. and van der Waal, J.C. (2001) Meerwein-Ponndorf-Verley Reduction, Oppenauer Oxidation, and related reactions, in Fine Chemicals through Heterogenous Catalysis (eds R.A. Sheldon, and H. van Bekkum, Wiley-VCH Verlag GmbH, New York. 438. Ruiz, J.R. and Jimenez-Sanchidrian, C. (2007) Curr. Org. Chem., 11, 1113. Creyghton, E.J., Ganeshie, S.D., Downing, R.S., and van Bekkum, H. (1995) J. Chem. Soc., Chem. Commun. 1859–1860. Creyghton, E.J., Ganeshie, S.D., Downing, R.S., and van Bekkum, H. (1997) J. Mol. Catal. A: Chem., 115, 457. Kunkeler, P.J., Zuurdeeg, B.J., van der Waal, J.C., van Bokhoven, J.A., Koningsberger, D.C., and van Bekkum, H. (1998) J. Catal., 180, 234. van der Waal, J.C., Tan, K., and van Bekkum, H. (1996) Catal. Lett., 41, 63. van der Waal, J.C., Kunkeler, P.J., Tan, K., and van Bekkum, H. (1998) J. Catal., 173, 74. van der Waal, J.C., Creyghton, E.J., Kunkeler, P.J., Tan, K., and van Bekkum, H. (1998) Top. Catal., 4, 261. Akata, B., Yilmaz, B., and Sacco, A. Jr. (2008) J. Porous Mater., 15, 351. Corma, A., Domine, M.E., Nemeth, L., and Valencia, S. (2002) J. Am. Chem. Soc., 124, 3194.

References 196. Corma, A., Domine, M.E., and 197. 198. 199. 200.

201. 202.

203. 204.

205.

206. 207.

208.

209. 210.

211. 212. 213. 214.

215. 216. 217.

Valencia, S. (2003) J. Catal., 215, 294. Zhu, Y., Chuah, G., and Jaenicke, S. (2003) Chem. Commun., 2734. Boronat, M., Corma, A., and Renz, M. (2006) J. Phys. Chem. B, 110, 21168. Zhu, Y., Chuah, G., and Jaenicke, S. (2004) J. Catal., 227, 1. Zhu, Y., Liu, S., Jaenicke, S., and Chuah, G. (2004) Catal. Today, 97, 249. Zhu, Y., Chuah, G.K., and Jaenicke, S. (2006) J. Catal., 241, 25. Nie, Y., Jaenicke, S., van Bekkum, H., and Chuah, G.K. (2007) J. Catal., 246, 223. Corma, A. and Renz, M. (2007) Angew. Chem. Int. Ed., 46, 298. Corma, A., Xamena, F.X., Prestipino, C., Renz, M., and Valencia, S. (2009) J. Phys. Chem. C, 113, 11306. Samuel, P.P., Shylesh, S., and Singh, A.P. (2007) J. Mol. Catal. A: Chem., 266, 11. Zhu, Y., Jaenicke, S., and Chuah, G.K. (2003) J. Catal., 218, 396. De Bruyn, M., Limbourg, M., Denayer, J., Baron, G.V., Parvulescu, V., Grobet, P.J., De Vos, D.E., and Jacobs, P.A. (2003) Appl. Catal., A, 254, 189. Liu, S.H., Chuah, G.K., and Jaenicke, S. (2004) J. Mol. Catal. A: Chem., 220, 267. Barthomeuf, D. (1996) Catal. Rev., 38, 521. Joshi, U.D., Joshi, P.N., Tamhankar, V.V., Joshi, C.V., Rode, V., and Shiralkar, P. (2003) Appl. Catal., A, 239, 209. Barthomeuf, D. (1984) J. Phys. Chem., 88, 42. Barthomeuf, D. (2003) Microporous Mesoporous Mater., 66, 1. Barthomeuf, D. (1991) Stud. Surf. Sci. Catal., 65, 157. Corma, A., Fornes, V., Martin-Aranda, R.M., Garcia H., and Primo, J. (1990) Appl. Catal., 59, 237. Mortier, W.J. (1978) J. Catal., 55, 138. Hathaway, P.A. and Davis, M.E. (1989) J. Catal., 116, 263. Hathaway, P.A. and Davis, M.E. (1989) J. Catal., 116, 279.

218. Lasperas, M., Cambon, H., Brunel, D.,

219. 220. 221. 222. 223. 224. 225.

226.

227.

228.

229.

230. 231.

232. 233.

234.

235. 236.

237.

238.

Rodriguez, I., and Geneste, P. (1995) Microporous Mater., 1, 343. Yagi, F. and Hattori, H. (1997) Microporous Mater., 9, 237. Brownscombe, T.F. and Slaugh, I.H. (1989) EU Patent 370553. Brownscombe, T.F. (1991) US Patent 5053372. Knoevenagel, L. (1898) Ber. 31, 258. Jones, G. (1967) Organic Reactions, 15, 204. Corma, A. (1991) Mat. Res. Soc. Symp. Proc., 233, 17. Lai, S.M., Martin-Aranda, R., and Yeung, K.L. (2003) Chem. Commum., 218. Lai, S.M., Ng, C.P., Martin-Aranda, R., and Yeung, K.L. (2003) Microporous Mesoporous Mater., 66, 239. Zhang, X.F., Lai, S.M., Martin-Aranda, R., and Yeung, K.L. (2004) Appl. Catal., A, 261, 109. Lau, W.N., Yeung, K.L., and Martin-Aranda, R. (2008) Microporous Mesoporous Mater., 115, 156. Zienkiewicz, Z., Calvino-Casilda, V., Sobczak, I., Ziolek, M., Martin-Aranda, R., and Lopez-Peinado, A.J. (2009) Catal. Today, 142, 303. Corma, A. and Martin-Aranda, R.N. (1993) Appl. Catal., A, 105, 271. Corma, A., Iborra, S., Primo, J., and Rey, F. (1994) Appl. Catal., A, 114, 215. Reddy, T. and Varma, R.S. (1997) Tetrahedron Lett., 38, 1721. Martins, L., Boldo, R.T., and Cardoso, D. (2007) Microporous Mesoporous Mater., 98, 166. Martins, L., Vieira, K.M., Rios, D., and Cardoso, D. (2008) Catal. Today, 706, 133. Martins, L., Holderich, W.F., and Cardoso, D. (2008) J. Catal., 258, 14. Oliveira, A.C., Martins, L., and Cardoso, D. (2009) Microporous Mesoporous Mater., 120, 206. Lasperas, M., Rodriguez, I., Brunel, D., Cambon, H., and Geneste, P. (1995) Stud. Surf. Sci. Catal., 97, 319. Rodriguez, I., Cambon, H., Brunel, D., Lasperas, M., and Geneste, P. (1993) Stud. Surf. Sci. Catal., 78, 623.

825

826

25 Zeolites as Catalysts for the Synthesis of Fine Chemicals 239. Rodriguez, I., Cambon, H., Brunel, D.,

240. 241.

242. 243.

244.

245.

246.

247.

248.

249.

and Lasperas, M. (1998) J. Mol. Catal. A, 130, 95. Bergmann, E.D., Ginsburg, D., and Pappo, R. (1959) Org. React., 10, 179. Sreekumar, R., Rugmini, P., and Padmakumar, R. (1997) Tetrahedron Lett., 38, 6557. Trost, B.M. and Keeley, D.E. (1975) J. Org. Chem., 40, 2013. Kumarraja, M. and Pitchumani, K. (2006) J. Mol. Catal. A: Chem., 256, 138. Shinde, P.D., Mahajan, V.A., Borate, H.B., Tillu, V.H., Bal, R., Chandwadkar, A., and Wakharkar, R.D. (2004) J. Mol. Catal. A: Chem., 216, 115. Baba, T., Kim, G.C., and Ono, Y. (1992) J. Chem. Soc. Faraday Trans., 88, 891. Baba, T., Satoru, H., Ryutaro, K., and Yoshio, O. (1993) J. Chem. Soc., Faraday Trans., 89, 3177. Baba, T., Satoru, H., Yishio, O., Tomoko, Y., Tsunehiro, T., and Satohiro, Y. (1995) J. Mol. Catal., 98, 49. Volcho, K.P., Kurbakova, S.Yu., Korchagina, D.V., Suslov, E.V., Salakhutdinov, N.F., Toktarev, A.V., Echevskii, G.V., and Barkhash, V.A. (2003) J. Mol. Catal. A: Chem., 195, 263. Volcho, K.P., Suslov, E.V., Kurbakova, S.Yu., Korchagina, D.V., Salakhutdinov, N.F., and Barkhash, V.A. (2004) Russ. J. Org. Chem., 40, 659.

250. Suslov, E.V., Korchagina, D.V.,

251. 252.

253.

254.

255.

256.

257.

258.

259.

260.

Komarova, N.I., Volcho, K.P., and Salakhutdinov, N.F. (2006) Mendeleev Commun., Vol. 4, 202. Kloestra, K.R. and van Bekkum, H. (1997) Stud. Surf. Sci. Catal., 105, 431. Kloestra, K.R. and van Bekkum, H. (1995) J. Chem. Soc., Chem. Commun., 1005. Kloestra, K.R., Van Laren, M., and van Bekkum, H. (1997) J. Chem. Soc., Faraday Trans., 93, 1211. Cobzaru, C., Oprea, S., Dumitriu, E., and Hulea, V. (2008) Appl. Catal., A, 351, 253. Veloso, C.O., Henriques, C.A., Dias, A.G., and Monteiro, J.L. (2005) Catal. Today, 107-108, 294. Saravanamurugan, S., Palanichamy, M., Arabindoo, B., and Murugesan, V. (2004) J. Mol. Catal. A: Chem., 218, 101. Sharma, S.K., Parikh, P.A., and Jasra, R.V. (2007) J. Mol. Catal. A: Chem., 278, 135. Yu, J.I., Shiau, S.Y., and Ko, A.N. (2001) React. Kinet. Catal. Lett., 72, 365. Jaenicke, S., Chuah, G.K., Lin, X.H., and Hu, X.C. (2000) Microporous Mesoporous Mater., 35-36, 143. Climent, M.J., Corma, A., Guil-Lopez, R., Iborra, S., and Primo, J. (1998) J. Catal., 175, 70.

827

26 Zeolites and Molecular Sieves in Fuel Cell Applications King Lun Yeung and Wei Han

26.1 Introduction

Zeolites and molecular sieves belong to an important class of natural and synthetic porous materials [1] that ﬁnd important applications in many industrial processes including chemical separation and puriﬁcation, chemical synthesis and fuel conversion, and pollution treatment and abatement [2–4]. The discovery of MCM-41 in 1992 [5, 6] has attracted research in the synthesis and applications of large pore molecular sieves including ordered mesoporous materials. The mesoporous materials display a greater range of pore sizes and compositions than typical zeolites and molecular sieves, and researches show that they are excellent catalyst, adsorbent, and membrane materials that have potential applications not only in traditional chemical processes but also in microelectronics, sensors, and agriculture [7–10]. Fuel cell promises clean and efﬁcient energy generation for stationary, mobile, and portable applications [11]. Striving for enhanced performance, zeolites and mesoporous materials are increasingly used in fuel cells. They were used to increase proton transport, decrease fuel crossover, and improve water management in the electrolyte membrane. They serve as electrode and electrocatalyst in the fuel cell and are also employed in fuel conversion, reforming, and storage. There are a number of excellent reviews on fuel cell electrolyte membranes [12–17] and electrode materials [18, 19] in the literature, but none includes the recent uses of zeolites and molecular sieves. This chapter reviews the contributions of zeolites and molecular sieves in fuel cell research and is divided into three sections: (i) zeolites in electrolyte membrane, (ii) zeolites in fuel cell electrocatalysis, and (iii) zeolites in fuel processing for fuel cells.

26.2 Zeolites in Electrolyte Membrane

The electrolyte membrane in a fuel cell separates the anode and cathode reactants and mediates the electrochemical reactions at the electrodes by rapid and selective Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

828

26 Zeolites and Molecular Sieves in Fuel Cell Applications

conduction of ions [20]. The perﬂuorosulfonic acid (PFSA) polymers that include Naﬁon from Du Pont are used in the proton-exchange membrane fuel cell (PEMFC). They display high proton conductivity (about 0.1S·cm−1 ) under fully hydrated conditions and excellent long-term stability (about 6000 hours) under fuel cell operation. Table 26.1 lists some of the commercial PFSA polymers produced by Du Pont, Asahi, and Dow. PFSA polymers restrict the operating temperature of PEMFC below 353 K. PFSA loses its mechanical and dimensional stability above this temperature as a result of its low glass transition temperature (i.e., 353 K ≤ Tg ≤ 393 K) [21]. A sharp drop in proton conductivity is also observed as a result of poor membrane hydration at low humidity. The low operating temperature also brings other problems to PEMFC operation including greater sensitivity to fuel impurities (e.g., CO, H2 S) and complicates the heat and water management in the fuel cell [11]. The PFSA membrane also limits the type of fuels that can be used in the fuel cell. In direct methanol fuel cell (DMFC) using PFSA membrane, fuel crossover is a problem as methanol is easily transported together with the solvated protons across the membrane resulting in poor fuel cell performance [22, 23]. These concerns motivate research into new proton conducting membranes that can maintain high proton conductivity and low fuel permeability at elevated operating temperatures. Three general approaches have been reported in the literature Table 26.1

(CF2

Commercial PFSA membranes [13].

CF2)x

(CF2

CF)y (O

CF2

CF)m

O

(CF2)n

SO3H

CF3 Structure parameter

Trade name and type

m = 1; x = 5–13.5; n = 2; y = 1

DuPont Naﬁon 120 Naﬁon 117 Naﬁon 115 Naﬁon 112 Asahi Glass Flemion-T Flemion-S Flemion-R Asahi Chemicals Aciplex-S Dow Chemical Dow

m = 0, 1; n = 1–5

m = 0; x = 1.5–14; n = 2–5 m = 0; x = 3.6–10; n = 2

a The

Equivalent weighta

Thickness (µm)

1200 1100 1100 1100

260 175 125 80

1000 1000 1000

120 80 50

1000–1200

25–100

800

125

equivalent weight is deﬁned as the number of grams of dry polymer per mole SO3 .

26.2 Zeolites in Electrolyte Membrane

to achieve these goals: (i) the development of new proton conducting polymer membranes based on polybenzimidazole (PBI), poly(phenylquinoxalines) (PPQ), and polyetheretherketone (PEEK) [12–17, 21]; (ii) proton conducting inorganic membranes that include CsHSO4 [24], mesoporous zirconium, and titanium phosphates [25, 26] and mesoporous heteropolyacids supported on SiO2 [27]; and (iii) organic/inorganic composite membranes that contain inorganic solid acids including various oxides, Zr-based solid acids, heteropolyacids, acid-modiﬁed clays, and zeolites. Zeolites and molecular sieves are among the most common materials used for improving the membrane performance in PEMFC and DMFC. 26.2.1 Zeolite Conductivities

Zeolite is a crystalline microporous inorganic solid made of a three-dimensional framework of tetrahedral SiO4 and AlO4 building units. The insertion of [AlO4 ]5− into the [SiO4 ]4+ framework creates an excess negative charge that has to be counterbalanced by a proton, an alkaline cation, or half an alkaline-earth cation or organic cation. The zeolite channels and cages are often occupied by the cations, water, or other solvent molecules. The cations located at the extra-framework sites are commonly exchangeable and their migration is responsible for the ionic conductivity of the zeolite. Experiments and calculations have been used to determine the ionic conductivity of zeolites under different conditions. Table 26.2 lists some of the ionic conductivity data for different zeolites determined by experiments and theoretical calculations. The ionic conductivities of typical zeolites (e.g., HZSM-5, HY, mordenite) are rarely higher than 10−4 S·cm−1 and are considerably lower than those of PFSA polymers (i.e., 0.1 S·cm−1 ). However, some zeolites including hydrated tin-mordenite can have very high ionic conductivities (e.g., 0.1 S·cm−1 ) as reported by Knudsen and coworkers [36, 37]. Saad and coworkers [28] observed using complex impedance spectroscopy that the ionic conductivities of Y zeolite and mordenite containing various cations (i.e., Co, Ni, Cu) were enhanced in with the addition of a second lithium cation. The calculations based on semi-classical transition state theory (SC-TST) [32], QM-Pot [33], and density functinal theory (DFT) [34] methods indicated that among the different zeolites (i.e., HY, H-chabazite, HZSM-5), the HZSM-5 had the highest proton mobility due to the greatest ﬂexibility of its framework lattice. The measured and calculated activation energies of ion conduction in zeolites are reported to be between 40 and 130 KJ·mol−1 (cf. Table 26.2), which are much lower than the deprotonation energy of 1300 KJ·mol−1 . This implies that the long-range transport of protons in zeolites should not be possible at these moderate temperatures. Various researchers showed that ammonium-exchanged zeolites displayed ionic conductivity much higher than 10−4 S·cm−1 at room temperature. Moroz and coworkers [29] employed wideline 1 H NMR to investigate the proton transfer rate between extra-framework species in hydrated NH4 -chabazite and NH4 -clinoptilolite. They reported that less than 2% of the conductivity observed

829

126.4 39 (77)

10 –7 –10 –6 N 10 –8 –10 –7 10 –6 –10 –5 (10−5 –10−4 )

15 25 40 75 140 500

HZSM-5

N

(1.12–4.25) × 10 –8

hydrated samples, dry N2 , 393–473 K (613–773 K)

101.4

N

4.5

NH4 -clinoptilolite

15

100.4

10 –7 –10 –5

2.2

NH4 -chabazite

HZSM-5

96.4

N

5

Cu-mordenite

(0.89–2.77) × 10 –8

NMR

Impedance spectroscopy

Impedance spectroscopy

1H

89.9

89.8

60

40

Na

84.8

96.3

2.3

Impedance spectroscopy

Cu-Y

Dehydrated samples, 803–873 K Dehydrated samples, 718–873 K Dehydrated samples, 823–923 K Hydrated samples, 300–400 K Hydrated samples, 200–300 K Dehydrated samples, dry N2 , 423–773 K

10 –9 –10 –8

Ea (kJ·mol –1 )

2.3

Test method

Ni-Y

Test conditions

Si/Al

Conductivity (S·cm –1 )

Conductivity and activation energy data of zeolites.

Zeolite type

Table 26.2

[31]

[30]

[29]

[28]

References

830

26 Zeolites and Molecular Sieves in Fuel Cell Applications

11 47 95 95 25

H-chabazite

HY

HZSM-5

HZSM-5

HBeta

b RT

N Water-soaked samples, RTb

1.47 × 10 –4

N

N

N

N

N

means that the information is not available from the reference. means room temperature.

a ‘‘N’’

2

Impedance spectroscopy

DFT method

QM-Pot method

SC-TST

N

40

N

127

52–98

68–106

70–102

97.1

53 (89)

59 (85)

N

25

Impedance spectroscopy

49 (74)

NH3 -adsorbed samples, dry N2 , 393–473 K (613–773 K)

10 –6 –10 –5 (10−5 –10−4 )

15

HY

HZSM-5

40 (92)

N

40

49 (87)

N

25

[35]

[34]

[33]

[32]

[31]

26.2 Zeolites in Electrolyte Membrane 831

832

26 Zeolites and Molecular Sieves in Fuel Cell Applications

in the NH4 -zeolites could be attributed to proton transport and that the hydrated NH4 -zeolites were mainly NH4 + conductors. Freud and coworkers [38] found that H-zeolites were also ammonium conductors in the presence of NH4 + at concentrations greater than 0.1 NH4 + per cage. This suggests that there are several possible transport mechanisms for ions in zeolites. Simon and coworkers [30, 31] compared the proton transport in solvate-free and solvated HZSM-5 at different temperatures and observed that the proton conductivity was enhanced in the present of NH3 and H2 O. They reported that the activation energies of proton transport in the solvated zeolites were signiﬁcantly lower. This suggests that the solvent plays a role in the proton transport in zeolites at low to moderate temperatures. The Grotthuss transport mechanism [40–42], popular for explaining proton transport in liquid electrolyte, could explain the rapid proton transport observed in solvated zeolites at low temperature. The protons move in along a chain of solvated molecules as shown in Figure 26.1a. In hydrated zeolites, the proton is transferred from an oxonium ion to a neighboring water molecule by tunneling through a hydrogen bond, followed by the rapid reorientation of the water molecule (i.e., former oxonium ion) to accommodate the next proton. The vehicle transport mechanism (Figure 26.1b) was proposed by Rabenau and coworkers [43] to explain the relative small contribution of proton transport in the overall conductivity of NH4 -zeolites [29, 39]. This mechanism proposes that proton does not migrate as H+ but as H3 O+ or NH4 + , bonded to a ‘‘vehicle’’ such as H2 O or NH3 . The ‘‘unladen’’ vehicles move in the opposite direction and the proton conduction is determined by the diffusion rate of the vehicle. Simon and coworkers [31] observed that proton transport in solvated HZSM-5 (Si/Al ≤ 40) displayed both Grotthuss and vehicle transport mechanisms. They reported that the Grotthuss transport mechanism was dominant at temperatures below 393 K, while the vehicle mechanism appeared to be important at the higher temperatures (i.e., 393–473 K). Freude and coworkers [38] observed that the formation of proton vehicle depended on the water and ammonia content of the zeolite and also on its dehydration temperature. Zeolites are promising candidates for proton conducting membranes capable of operating at high temperatures and preventing fuel crossover [50]. Their main drawback is their low proton conductivity compared to PFSA. Zeolites are often

(a)

(b) Figure 26.1 (a) Grotthuss and (b) vehicle transport mechanisms for protons in zeolites [43].

26.2 Zeolites in Electrolyte Membrane

833

used as inorganic ﬁllers added to proton conducting membranes to enhance proton conduction at high temperatures and to decrease methanol crossover. Recent advances show that high conductivity can be obtained by loading the zeolites and molecular sieves with heteropolyacids and by functionalizing the initial materials with sulfonic acid groups (Table 26.3). New microporous materials, such as ICF-5 and ICF-21, display proton conductivities approaching those of PFSA materials. 26.2.2 Zeolite/Polymer Composite Membranes

Figure 26.2 illustrates the most common method of preparing zeolite/polymer composite membranes. The zeolite suspension or powder is mixed with the polymer solution and homogenized by rapid stirring or ultrasonication to obtain a uniform mixture of zeolite and polymer. High boiling point solvents such as DMF and DMSO are used, as the membrane recast from these solvents exhibits better mechanical properties. Also, their high viscosity prevents the aggregation and precipitation of the zeolites and therefore ensures their uniform dispersion in the membrane. The zeolite–polymer suspension is then cast on a clean surface and the solvents are allowed to evaporate at low temperature and in an inert atmosphere. This prevents the oxidation of the solvents by the highly acidic zeolites or polymers. This method has been used to prepare various zeolite/polymer composite membranes listed in Table 26.4. The table summarizes the proton conductivity and methanol permeability of these composite membranes as well as the maximum power density (MPD) of the corresponding fuel cell. It can be seen from Table 26.4 that the zeolite/polymer composite membranes generally have lower methanol permeability and better proton conductivity at high temperatures. This is attributed to the zeolites’ excellent chemical, thermal, and mechanical stability, high water retention, molecular-sized pores, and adjustable adsorption property. Table 26.4

Polymer solution or suspension

Mixing

Vacuum drying

Homogenizing via ultrasonic treatment

Glass dish

Zeolite/polymer composite membrane

Zeolite suspension

Preparation of precursor

Figure 26.2 Schematic diagram of the preparation of zeolite/polymer composite membranes.

Casting

Formation of membrane

1.15 × 10 –2

12.5

2.4 × 10 –3 3.9 × 10 –3 (1–16.7) × 10 –3 2 × 10 –1

∞ ∞ ∞ ∞ – –

Sulfonic acid functionalized MCM-41

Sulfonic acid functionalized MCM-48

SBA-15 containing polyaniline

Sulfonic acid functionalized MCM-41

ICF-5 CuInS-Na

ICF-21 InSe-Na

b RT

means that the information is not available from the reference. means room temperature. c There are no Si and Al elements in materials. d The data are conductivity.

a ‘‘N’’

4.5 × 10 –3

500

10 –10

–3

–2

10 –4 –10 –2 d

6.8 × 10 –3

c

6.7 × 10

50

–3

4.9 × 10 –3

200

25

2.96 ×

25

Sulfonic acid functionalized Beta

1.17 × 10 –3

50

Sulfonic acid functionalized Beta

7–100 % RH, 294 K

0.2–90% RH, 291 K

100% RH, RT

Water-soaked samples, RT

Water-soaked samples, RT

Water-soaked samples, RT

Water-soaked samples, RT

Y containing 40 wt% water, RT

10 –2

10 –3

Y containing 10 wt% water, RTb

5.1 × 10 –5

Na

Molybdophosphoric acid loaded Y zeolite 1.1 ×

Test conditions

Conductivity (S·cm –1 )

Si/Al

Zeolites and mesoporous materials with high proton conductivities.

Sample name

Table 26.3

[49]

[48]

[47]

[46]

[45]

[35]

[44]

References

834

26 Zeolites and Molecular Sieves in Fuel Cell Applications

Thickness (µm)

35–45

175

131–150

50–90

N

Beta/chitosan

Beta/Naﬁon

Beta/SPEEK

(A, mordenite, X, ZSM-5)/chitosan

Y/chitosan

RT

RT

353, 393

(1.7–2.2) × 10 –2 (after immersion in H2 SO4 for 24 h) (1.51–2.58) × 10 –2 (after immersion in H2 SO4 for 24 h)

(1.10–1.55) × 10 –2 (10–50% RH) (4.8–6.7) × 10 –2 (100% RH) N

RTa

353

Conductivity (S·cm –1 )

Operation temperature (K)

5 mol·l –1 methanol, RT

5 mol·l –1 methanol, RT

(6.24–12.2) × 10 –7 (chitosan: 1.15 × 10−6 )

9.04 × 10 –7 (Naﬁon 117 : 2.71 × 10−6 )

–

8 vol% methanol, 353 K

(3.63–5.86) × 10 –6 (Naﬁon 117 : 5.07 × 10−6 ) c

2 mol·l –1 methanol, RT

(5.80–9.55) × 10 –7 (chitosan: 1.17 × 10−6 )

–

Test conditions of methanol permeation

Methanol permeability (cm2 s –1 )

Performances of composite membranes and the corresponding fuel cells.

Membrane

Table 26.4

N

[55]

[54]

[53]

[52]

[51]

References

(continued overleaf)

285 (SPEEK: 180) (wet H2 , O2 at 393 K) N

N

Nb

MPD (mW·cm –2 )

26.2 Zeolites in Electrolyte Membrane 835

ZSM-5/Naﬁon

100–150

RT

10 –3

(1–6) × 10 –5 (43–100% RH) (7–12) × 10 –2

293–423

(1.0–1.6) × 10 –6 (Naﬁon 115: 1.4 × 10−6 )

–

16 wt% methanol, RT

–

1 mol·l –1 methanol, 323 K

(0.04–7.3) × 10 –7 (Naﬁon 115 : 2.5 × 10−6 )

Mordenite/PVDF

N

ETS-10/PVDF

323

(1.03–1.12) × 10 –6 (Naﬁon 117 : 7.4 × 10−6 )

(1.4–3.5) × 10 –3 (in boric acid solution at pH 4.7) (0.13–34) × 10 –3 (after immersion in H2 O for 48 h) (0.8–10) × 10 –3

(0.2–2) × 10 –4

N

Fe-silicalite-1/Naﬁon

308

12 mol·l –1 methanol, RT 1 mol·l –1 methanol, 308 K

4.9 × 10 –7

N

Umbite/PVDF

N

A/Naﬁon

RT

Test conditions of methanol permeation

Methanol permeability (cm2 s –1 )

Conductivity (S·cm –1 )

(0.2–2) ×

50–80

Mordenite/chitosan

Operation temperature (K)

NaA/PVDF

Thickness (µm)

(continued )

Membrane

Table 26.4

N

N

N

N

N

MPD (mW·cm –2 )

[60]

[59]

[58]

[57]

[56]

References

836

26 Zeolites and Molecular Sieves in Fuel Cell Applications

151

100–220 70

156 N 100–300

Beta/Naﬁon

Y/SPEEK Chabazite (Clinoptilolite)/Naﬁon

Chabazite/Naﬁon

Clinoptilolite/Naﬁon

Mordenite/PVA

RT

295, 333

293–413 363–413

343 K

5.93 × 10 –6 (333 K)

4.89 × 10 –2 (295 K)(in 10−5 mol·l –1 H2 SO4 ) (1.3–3.0) × 10 –2 (after immersion in H2 SO4 )

1 mol·l –1 methanol, RT

8 vol% methanol, 333 K

5.75 × 10 –6 (333 K)

4.08 × 10 –2 (295 K)

(3.13–11.3) × 10 –8 (PVA: 3 × 10−7 )

– N

– N

(7.2–9.6) × 10 –3 N

8 vol% methanol, RT

1.40 × 10 –6 (Naﬁon 115 : 2.36 × 10−6 )

8.8 × 10 –2

N

[65]

[64]

[62] [63]

[61]

(continued overleaf)

120 (Naﬁon 115: 62) (5 mol·l –1 methanol, O2 at 343 K) N 350–370 (Naﬁon: 270) (1 mol·l –1 methanol, O2 at 413 K) N

26.2 Zeolites in Electrolyte Membrane 837

29 24–73 N N N

SiO2 /SPEEK

ZrO2 /SPEEK

HPA/Naﬁon SiO2 /PEG

TiO2 /Naﬁon

418

393 308

328

403

1.5 × 10 –2 (35% RH) 10 –3 –10 –2 (after immersion in H2 O) N

N

N

1.3 × 10 –7 −2 × 10−3 (80% RH) 4 × 10 –4 –10 –2 (RT)

Conductivity (S·cm –1 )

b ‘‘N’’

RT means room temperature. means that the information is not available from the reference. c ‘‘ – ’’ means that a column entry is not applicable.

a

165, 90

293–373

300–400

Zirconium phosphate/Naﬁon

348–423

196–860

(X,Y, ZSM-5, mordenite, A)/PTFE Mordenite/PTFE

Operation temperature (K)

Thickness (µm)

(continued )

Membrane

Table 26.4

– 3 wt% methanol, 308 K –

20 wt% methanol, 328 K

610 g·h –1 m –2 70–480 g·h –1 m –2 (SPEEK: 1 300) – (0.78–2.1) × 10 –8 (Naﬁon 117: (7.5−2.3) × 10−6 ) N

–

N

N –

–

Test conditions of methanol permeation

–

Methanol permeability (cm2 s –1 )

[70] [71] [72]

350 (2 mol·l –1 methanol, O2 at 418 K)

[69]

[68]

[67]

[66]

References

N N

4 (methanol, O2 at 343 K) 450 (165 µm), 680 (90 µm) (Naﬁon 115: 110) (wet H2 , O2 at 403 K) N

N

MPD (mW·cm –2 )

838

26 Zeolites and Molecular Sieves in Fuel Cell Applications

26.2 Zeolites in Electrolyte Membrane

Case 1: s(Continuum) >> s (Spheres)

MeOH

H+

Figure 26.3

Case 2: s (Spheres) >> s(Continuum)

MeOH

H+

Transports in composite membranes [65].

also includes composite membranes that contain solid acids, heteropolyacids, and zirconium phosphate for comparison. The zeolites in the composite membrane hinder the transport of methanol in the membrane and prevent crossover. However, if the proton conductivity of the zeolites is much less than the polymer in the continuum matrix (Figure 26.3 case 1), the overall conductivity of the composite membrane will be lower than that of the pure polymer membrane. On the other hand, if the proton conductivity of the zeolites is higher or comparable to the polymer matrix (Figure 26.3 case 2), it is possible to prevent methanol crossover while maintaining and even enhancing the overall proton conductivity of the composite membrane. Although the proton conductivity of most zeolites is lower than that of PFSA, recent advances have been successful in preparing zeolites and molecular sieves with proton conductivities approaching that of PFSA (cf. Table 26.3). 26.2.2.1 Zeolite/PTFE Composite Membranes Early researchers prepared zeolite/polytetraﬂuoroethylene (PTFE) composite membranes [66, 67]. PTFE is inert and nonconducting and therefore the measured conductivities could be attributed mainly to zeolites. Yahiro et al. [66] and Poltarzewshi et al. [67] investigated the effects of zeolite-type, zeolite counterions, zeolite content, and temperature on the conductivity of the composite membrane. PTFE composite membranes containing X and Y zeolites displayed higher conductivity compared to composite membranes of NaA, NaZSM-5, and Na-mordenite. The conductivity of LiY/PTFE composite membrane reached 2 × 10−3 S·cm−1 at 348 K, while the mordenite/PTFE membrane was stable up to a temperature of 623 K. A DMFC unit assembled using the mordenite/PTFE membrane (i.e., 80 wt% mordenite) delivered current and power densities of 50 mA·cm−2 and 4 mW·cm−2 at 343 K, respectively. 26.2.2.2 Zeolite/PFSA Composite Membranes The zeolite/PFSA membranes are the most studied zeolite composite membranes in the literature. Kim and coworkers [60] prepared HZSM-5–Naﬁon composite membranes from 100-nm sized HZSM-5 crystals. The membrane displayed excellent selectivity in terms of proton conductivity to methanol permeability ratio. The optimized membrane displayed better proton conductivity than Naﬁon 115

839

840

26 Zeolites and Molecular Sieves in Fuel Cell Applications

and also had signiﬁcantly lower methanol permeability. Baglio and coworkers [63] used chabazite and clinoptilolite to prepare their Naﬁon composite membranes for high-temperature DMFCs. They obtained a MPD of 370 mW·cm−2 at 413 K using pure oxygen feed from a composite membrane with 6 vol% zeolite. This value was higher than that of recasted Naﬁon membrane (about 270 mW·cm−2 ) of similar thickness. The poor surface compatibility between zeolite particles and Naﬁon polymer can result in pinholes and poor proton conductivity. Roberts and coworkers [57] improved the interfacial bonding between the zeolites and polymer by grafting organic linkers (e.g., 3-aminopropyltrimethoxysilane) on zeolites (i.e., NaA) before preparing the zeolite/Naﬁon composite. The NaA/Naﬁon composite membrane exhibited lower methanol permeability compared to a similar NaA/Naﬁon membrane prepared from unmodiﬁed NaA, which indicated that a better interfacial compatibility was obtained after zeolite functionalization. The methanol permeability was reduced by as much as 86% using functionalized NaA as ﬁllers compared to commercial Naﬁon 117, but with considerable loss of proton conductivity. The group of Yushan Yan [52] remedied this problem by using sulfonic acid functionalized Beta zeolite (AFB) in their Naﬁon composite membranes. A composite membrane with 5 wt% AFB had similar proton conductivity as that of Naﬁon 117 but signiﬁcantly higher selectivity (i.e., proton conductivity/methanol permeability ratio). The composite membrane was better than Naﬁon 117 by 93 and 63% at 294 and 353 K, respectively. The DMFC using this composite membrane had greater power output than a fuel cell assembled from Naﬁon 117. Zeolites/PFSA membranes had been prepared by other methods besides the solution casting method shown in Figure 26.2. Gribov and coworkers [58] prepared Fe-silicalite-1/Naﬁon composite membranes by impregnating a supercritical CO2 pretreated Naﬁon membrane with colloidal suspension of the zeolite crystals. The supercritical CO2 pretreatment altered the Naﬁon pore structure, allowing the zeolite crystals to diffuse into the pores. They also demonstrated that it was possible to crystallize the zeolites in situ by ﬁrst impregnating the altered membrane with the zeolite synthesis gel followed by hydrothermal crystallization. These composite membranes exhibited 6–19 times higher selectivities than a Naﬁon membrane of comparable thickness (i.e., Naﬁon 115). Yan and coworkers [61] also employed in situ crystallization method to prepare their AFB/Naﬁon membranes. The composite membrane had a room temperature proton conductivity of 8.8 × 10−2 S·cm−1 , similar to that of Naﬁon. The DMFC using the composite membrane had a much higher MPD of 120 mW·cm−2 than a DMFC assembled from Naﬁon 115 (i.e., 62 mW·cm−2 ) at the operating temperature of 343 K. 26.2.2.3 Zeolite/Chitosan Composite Membranes and Others There is a growing interest in zeolite/chitosan composite membranes for fuel cell application. Chitosan is derived from shells of crustaceans and considered a green, eco-friendly, and biocompatible material. It displays low alcohol permeability and good mechanical properties and is inexpensive and easy to fabricate. However, chitosan suffers from a high degree of swelling and low proton conductivity, which

26.2 Zeolites in Electrolyte Membrane

can be remedied by using appropriate inorganic ﬁllers and through the use of sulfuric acid as a cross-linker [55]. Jiang and coworkers [51, 54–56] carried out a systematic study on zeolite/chitosan composite membranes. They reported that adding of mordenite created more rigid membranes (i.e., higher Tg ), but it also generated pinholes and defects due to poor compatibility between zeolites and chitosan. This was ameliorated by adding sorbitol as plasticizer, and the optimized mordenite/chitosan composite membrane had a much lower methanol permeability of 4.9 × 10−7 cm2 s−1 in 12 mol·l−1 methanol at room temperature compared to both chitosan and Naﬁon 117 membranes [56]. They also prepared various zeolite/chitosan composite membranes from zeolite 3A, 4A, 13X, and HZSM-5 and concluded that hydrophobic zeolites with high Si/Al ratio (i.e., mordenite and HZSM-5) were more effective in slowing methanol transport [54]. Jiang and coworkers [55] also reported that membrane swelling could be remedied by using Y zeolite modiﬁed with organic amine, thiol, and sulfonic acid surface groups. They observed that using sulfonic acid-modiﬁed Y zeolite as ﬁller improved the membrane proton conductivity to 3 × 10−2 S · cm−1 . Other zeolite/polymer composite membranes have been prepared mostly motivated by the lower polymer cost compared to PFSA. These include Beta/sulfonated polyetheretherketone (SPEEK) [53], heteropolyacid-loaded Y/SPEEK [62], (ETS-10, NaA, umbite, and mordenite)/polyvinylidiene ﬂuoride (PVDF) [59], mordenite/polyvinylalcohol (PVA) [65], and CaY/hydrated styrene butadiene rubber (HSBR) [73] composite membranes, which are summarized in Table 26.4. 26.2.2.4 Self-Humidifying Composite Membranes The proton conductivity of PFSA membrane is sensitive to hydration, and performance at high temperature suffers from low humidity. The need for external humidiﬁcation equipment complicates the system design and operation and lowers the overall energy efﬁciency. There is therefore active research in developing membranes that could retain and manage water through a self-humidiﬁcation process. Kim and coworkers [74] employed PtY zeolite catalysts to prepare their zeolite/PFSA membrane for a self-humidifying PEMFC. Pt provided catalytic sites for water generation while the zeolite absorbed the water and thus helped to keep the membrane hydrated during high-temperature fuel cell operation. 26.2.3 Zeolite and Mesoporous Inorganic Membranes

There are a large number of publications on the fabrication of various zeolite/polymer composite membranes and reports on their proton conductivity and methanol permeability, but only a few of them give an account of their performance in high-temperature fuel cells [53, 63]. Zeolites and polymers have considerably different mechanical and thermal properties, and the membrane durability for high-temperature operation has not been not well studied. Also, the nonuniform

841

842

26 Zeolites and Molecular Sieves in Fuel Cell Applications

distribution of zeolites and polymer in the composite membrane is unavoidable and imposes a practical limit on the extent of zeolite content. Yan and coworkers [61] observed that the Si/F ratio for AFB/Naﬁon membrane varies from 0.01 to 0.10 across the surface and thickness of the membrane. This could signiﬁcantly affect the membrane performance during high-temperature operations. There is therefore a considerable advantage in exploring the use of pure inorganic zeolite membranes. Although most zeolites have lower proton conductivity than the commercial PFSA, they have considerably better mechanical strength and could be fabricated into ultrathin membranes. Recent works by the group of Yeung investigated HZSM-5 as proton conducting membrane for microfabricated miniature hydrogen PEMFCs [75] and DMFCs [76]. They chose HZSM-5 as an experiments model, and calculations indicated that the proton mobility in ZSM-5 was among the highest in zeolites [33, 38, 77]. ZSM-5 ﬁlms and membranes are also easier to prepare and are amenable to miniaturization and microfabrication [78–93]. The freestanding zeolite micromembranes shown in Figure 26.4a were prepared by standard microfabrication technique, and each micromembrane (Figure 26.4b) measured 250 µm × 250 µm in area and 6 µm in thickness (Figure 26.4c). The 6-µm HZSM-5 membrane was sufﬁciently thick to prevent fuel crossover. The membrane was impermeable to H2 even at 473 K and 0.3 MPa after it had been hydrated by a stream of humid air, but was sufﬁciently thin to have comparable proton transport with the much thicker Naﬁon 117 membrane. Thus, the fuel cells assembled using the zeolite micromembranes performed as well as those with Naﬁon 117 MEA (Figure 26.4d). At the Electrochemical Society meeting in 2008, Jiang and coworkers [27] reported on the performance of a mesoporous heteropolyacid (HPW)/silica electrolyte membrane. The HPW is a superacid that has a Keggin-type structure and can retain hydration up to 573 K. This allowed the use of nonprecious metals as catalysts. The researchers built a cell using this membrane and GaFeCu cathode capable of 105 mW·cm−2 power output at 573 K.

26.3 Zeolite Electrocatalysts

Platinum and platinum alloys supported on carbons are typical electrocatalysts for PEMFC and DMFC. A patent ﬁled by Yasumoto and coworkers [94] claimed that the use of a zeolite as support of a metal catalyst in a fuel cell electrode resulted in a lower resistance and less ohmic power losses. Samant and Fernandes [95] investigated the use of Pt(HY) and Pt–Ru(HY) catalysts for electroxidation of methanol in fuel cells. They reported that catalysts prepared in an HY zeolite displayed signiﬁcantly enhanced electrocatalytic activity even compared to Pt supported on carbon (i.e., Vulcan XC-72). The enhancement was believed to be solely derived from the preferential formation of sterically constrained CO clusters that were readily oxidized to CO2 by interaction with the zeolite pore. A similar work was carried out by Pang and coworkers [96] for ethanol electroxidation using

26.3 Zeolite Electrocatalysts

(a)

(b)

P (mW cm−2)

15

10

5 Zeolite Nafion MEA

0 0 (c)

(d)

Figure 26.4 SEM images of (a) the zeolite micromembrane array showing (b) the individual freestanding micromembranes of 250 × 250 µm2 , and (c) the cross section of the micromembrane with the thickness of 6 µm. The inset in (b) is the higher

20

Nafion Si 40

60

j (A cm−2) magniﬁcation image of the zeolite membrane surface. (d) Performance comparison between Naﬁon MEA prepared by the hot-press method, Naﬁon/Si, and HZSM-5 micro fuel cell. (2.5 cm3 min−1 H2 with 100% RH, 2.5 cm3 min−1 dry O2 , T = 294 K).

the Pt/ZSM-5-C catalyst. They used a commercial ZSM-5 of high Si/Al ratio greater than 300 and were able to show higher activity and better stability compared to Pt/C. In spite of these promising results, zeolites are not the well suited for electrocatalysis owing to their poor electrical conductivity [97]. Coker and coworkers [98, 99] of Sandia National Laboratories described a zeolite-templating method for preparing Pt nanoparticles on carbon. Well-deﬁned nanometer-sized Pt clusters were formed within the zeolite host. The pores of zeolite containing the Pt clusters were then ﬁlled with electrically conductive carbon. Following the removal of the zeolite host, a micro-/mesoporous Pt/C electrocatalyst having Pt clusters of controlled size was obtained. Pt/C electrocatalysts with electroactive surface area of 112 m2 g−1 were obtained with better performance than the commercial Pt/C. Recently, two new methanol-tolerant cathode catalysts were prepared via a similar approach but using mesoporous molecular sieves SBA-15 as template. Woo and

843

844

26 Zeolites and Molecular Sieves in Fuel Cell Applications

coworkers [100] inﬁltrated SBA-15 with Pt and carbon precursors to obtain a Pt/C nanocomposite with mesoporous (2.9 nm) and microporous (0.5–1.3 nm) structures. The formation of Pt/C within the conﬁned space prevented metal aggregation and sintering during high-temperature carbonization. The DMFCs using the Pt/C nanocomposite as cathode exhibited higher open circuit voltages, current density, and methanol tolerance than the commercial electrocatalyst for the temperature range of 313–353 K. Li and coworkers [101] prepared mesoporous carbon containing core/shell Pt/C nanoparticles that displayed high tolerance to methanol. They deposited the Pt nanoparticles in SBA-15 channels and catalyzed the polymerization of glucose under hydrothermal condition to create the core/shell structure. This material had a speciﬁc surface area of 633 m2 g−1 and pore size of 3.5 nm and contained 3-nm Pt nanoparticles.

26.4 Zeolites and Molecular Sieves in Fuel Processing

Zeolites and molecular sieves are used as catalysts and adsorbents for fuel processing for PEMFCs shown in Figure 26.5. Zeolites have been used in nearly all processes except for electrolysis and fermentation. This section will attempt to present recent applications of zeolites and molecular sieves in fuel reﬁning and conditioning, fuel conversion and reforming, as well as hydrogen production, puriﬁcation, and storage.

Ferment

Biomass

Alcohol

Solid fuel Reforming

CTL Fossil fuel

Liquid fuel Desulfurization Reforming H + CO 2 Natural gas

Water gas shift

Reforming

H2 + CO2 + CO Syngas (H2 /CO~2) Decomposition or dehydro aromatisation

Methanol/DME

H2 + CO2

Separation Separation Methanol

Selectively CO removal

H2 storage PEMFC

Separation Electrolysis

H2

H2O Photocatalysis Decomposition NH3

Figure 26.5

Schematic diagram of various fuel processes involved in PEMFCs [102].

26.4 Zeolites and Molecular Sieves in Fuel Processing

26.4.1 Removal of Sulfur Compounds in Fuel

Organic sulfur compounds are present in natural gas and liquid fuels and are poison to both the reforming catalyst and fuel cell electrocatalyst. Sulfur removal by adsorption has obvious advantages over hydrodesulfurization. The process can be carried out at room temperature and does not require hydrogen. Table 26.5 lists some zeolite and molecular sieve adsorbents used for the removal of sulfur compounds. Yang and coworkers [118, 119] prepared Ag and Cu ion-exchanged NaY zeolites that selectively adsorbed sulfur compounds via π-complexation between the π orbital of the thiophene and the vacant s orbital of the ion-exchanged metals. They reported on ultra-deep desulfurization at room temperature with 1 g of adsorbent producing more than 30 g of clean diesel with less than 0.2 ppm sulfur from original sulfur content of 430 ppm. The adsorbents were regenerated by calcination and could be reused with less than 5% loss in adsorption capacity. Other researchers investigated Ni/KY [114], AgNO3 /Beta [108], and CuZn/Y [105] adsorbents for desulfurization of various fuels. Feng and coworkers [112] reported the use of Ce(IV)Y zeolite to obtain diesel with unprecedented low levels of sulfur (i.e., <0.01 ppm S). Metal oxides and salts supported on mesoporous materials [103, 107, 109] were also examined for sulfur removal and were reported to be less effective for ultra-deep desulfurization. 26.4.2 Hydrogen Production and Puriﬁcation 26.4.2.1 Reforming of Hydrocarbons To date, 90% of the hydrogen is produced by reforming natural gas or light oil fraction by high-temperature steam [120]. The basic steam reforming reactions of hydrocarbons are as follows:

CH4 + H2 O −→ CO + 3H2 H = 206kJ · mol−1 CnHm + nH2 O −→ nCO + (m/2 + n)H2 The reaction temperature is often above 773 K, and a mixture of hydrogen, carbon monoxide, and carbon dioxide is obtained from the reaction. Noble-metal catalysts are used to reduce metal dusting caused by CO coproduction at the extreme reaction temperatures [121]. Although coking is a serious problem on the less expensive nickel catalyst, it is the preferred catalyst for the reforming reaction. Alumina remains the most widely used support for the nickel catalyst, but a recent report by Wang and coworkers [122] suggested that Ni, Ni–Co, Ni–Mo, and Ni–Re supported on ZSM-5 had better activity and sulfur tolerance for steam-reforming methylcyclohexane. The other major routes for hydrogen production from fossil fuels are partial oxidation and autothermal reforming. Lee [123] used Rh supported rare-earth (La, Ce, Sm, Gd, Dy, and Er) ion-exchanged NaY catalysts for autothermal

845

Sulfur compounds or fuel

JP-5 light fraction

Commercial diesel

Thiophene, dibenzothiophene, 4,6-dimethyldibenzothiophene

Thiophene, tetrahydrothiophene, 4,6-dimethyldibenzothiophene

Cu2 O/MCM-41

Ni/SBA-15

AgY CuZnY

Ga/AlY

293–333

293–353

700→22 (Ag-Y) 700→36 (CuZn-Y) 500→15

240→10 11.7→0.1

841→50

RTa

RT-473

Sulfur content (ppm)

Temperature (K)

Results

1.7 (at 240 ppm) 0.47 (at 11.7 ppm) 44.9 (at 15 00 ppm) 17.5 (at 500 ppm) 7.0, 17.4, 14.5

12.8

Adsorption capacity (mg S · g−1 )

Some adsorbents involving zeolites and mesoporous materials for removal of sulfur compounds.

Material

Table 26.5

Air-calcination at 573 K N2 - calcination at 623 K

Air-calcination at 723 K

Air-calcination at 723 K He-calcination at 973 K Nb

Regeneration method

[106]

[105]

[104]

[103]

References

846

26 Zeolites and Molecular Sieves in Fuel Cell Applications

Ce(IV)Y Ni/KY

HDS-treated diesel Benzothiophene, 2-methylbenzothiophene, 5-methylbenzothiophene

RT

Benzothiophene, dibenzothiophene, 4,6-dimethyldibenzothiophene Benzothiophene

353 RT and 353

303

303

Diesel

CuY, NiY, NaY, and USY

RT-353

Tetrahydrothiophene, tert-butylmercaptane

AgNO3 /BEA, AgNO3 /MCM-41, AgNO3 /SBA-15 Cu(I)/Mesoporous aluminosilicate AgNO3 /mesoporous silica

RT

JP-5 light fraction

CuCl(PdCl2 )/MCM41(SBA-15)

1.87→<0.01 510→<1

N

N

315→54

80→0.1

841→50 (PdCl2 /SBA-15)

54.1 (NaY), 53.8 (USY), 57.6 (NiY), 63 (CuY) N 5

20.5

N

41.1 (AgNO3 /BEA)

38.4 (PdCl2 /SBA-15)

[112] [114]

[111]

[110]

[109]

[108]

[107]

(continued overleaf)

N Air-calcination at 573 K

N

Purge with diethyl ether

N

Purge with benzene at 343 K N

26.4 Zeolites and Molecular Sieves in Fuel Processing 847

Commercial diesel

Commercial gasoline

Cu(I)Y, AgY

Cu(I)Y

b ‘‘N’’

room temperature. means data is not available from the reference.

RT and 353

Commercial diesel

a RT:

RT

Commercial diesel

Cu(I)Y Cu(I)/ZSM-5 Ni(II)X, Ni(II)Y

RT

RT

RT

Commercial jet fuel

Cu(I)Y

Temperature (K)

Sulfur compounds or fuel

(continued )

Material

Table 26.5

335→<0.28

430→<0.2

297→0.06 (AC/Cu(I)Y) 297→0.22

364→0.071

Sulfur content (ppm)

Results

12.5

N

12.2 (Cu(I)Y) 2.6 (Cu(I)/ZSM-5) 10.6

25.5

Adsorption capacity (mg S · g−1 ) Air-calcination at 623K calcination in 5 vol% H2 /He at 504 K Air-calcination at 700–923 K Air-calcination at 623 K Air-calcination at 623 K and at 723 K, washing with dimethylformamide or carbon tetrachloride N

Regeneration method

[119]

[118]

[117]

[116]

[115]

References

848

26 Zeolites and Molecular Sieves in Fuel Cell Applications

26.4 Zeolites and Molecular Sieves in Fuel Processing

reforming of military jet fuel JP8 containing 1096 ppm sulfur to produce hydrogen and carbon monoxide at 1173 K. 26.4.2.2 Steam Reforming of Alcohols Hydrogen production from hydrocarbons consumes valuable fossil fuels and contributes to greenhouse gas emission, whereas alcohols that can be produced from biomass are considered to be more eco-friendly and sustainable sources of clean hydrogen. Until recently, the most common catalysts for ethanol steam reforming were noble metal supported on oxide catalysts. Cantao and coworkers [124] developed a more active catalyst based on Rh and Rh–K supported on NaY, which allowed the steam-reforming of ethanol at lower temperature (i.e., 573 K vs 773 K) with higher conversion (i.e., 99%) and hydrogen yield (i.e., 70%). 26.4.2.3 Decomposition of CH4 and NH3 Methane and ammonia have the highest hydrogen content of all compounds. Their decompositions are COx -free routes to hydrogen.

CH4 −→ C + 2H2 H = 75.6kJ · mol−1 NH3 −→ 1/2N2 + 3/2H2 H = 46.4kJ · mol−1 The catalytic decomposition of methane produces valuable coproducts such as carbon nanotubes or nanoﬁbers. However, carbon formation also results in catalyst deactivation. Goodman and coworkers [125] observed a rapid deactivation (about 1 hour) of Ni/HZSM-5 and Ni/C catalysts during methane decomposition reaction, whereas the Ni/HY catalyst was more stable (i.e., up to several hours). Analysis showed that the Ni/HZSM-5 catalyst was encapsulated with carbon while ﬁlamentous carbons formed on the Ni/HY catalyst. Subrahmanyam and coworkers [126, 127] investigated the decomposition of methane over Ni/HY, Ni/USY, Ni/SiO2 , and Ni/SBA-15 catalysts at 823 K. They found that the Ni/HY catalyst showed the highest activity of 955 mol H2 (mol Ni)−1 . Although ammonia is toxic, it has been considered as a promising carrier for hydrogen because of its convenient supply, storage, and transport. Ammonia is liquid at room temperature under a pressure of 0.8 MPa, and many metals including Fe, Ni, Ru, Ir, Pt, Co, and Rh, alloys (Zr1−x Tix M1 M2 ; M1 , M2 = Cr, Mn, Fe, Co, Ni; x = 0–1) and compounds of nitrides and carbides (MoNx , VNx ,VCx , and MoCx ) can catalyze ammonia decomposition [128]. The Ru–CeO2 /Y catalyst reported by Hashimoto and coworkers [129] displayed high activity for the decomposition of ammonia at 573 K. The reaction rate was 100 times higher than the conventional vanadium nitride catalyst. The reaction over Ru–CeO2 /Y catalyst was found to be of ﬁrst order for ammonia with the activation energy of 67 KJ·mol−1 , which is lower than those over Ru/Al2 O3 , Ru/SiO2 , and Ru/C catalysts (79–96 KJ·mol−1 ) [125]. 26.4.2.4 CO Removal from H2 -Rich Gas Hydrogen is produced from fuels through combined reforming and water gas shift reactions. A high carbon monoxide conversion is not possible during the reforming reaction as high temperature is not favorable to the water gas shift

849

850

26 Zeolites and Molecular Sieves in Fuel Cell Applications

reaction. Typically, about 1 vol% of carbon monoxide is present in hydrogen leaving the water gas shift reactor [130]. However, the acceptable CO concentration for PEMFC is below 10 ppm at Pt anode and about 100 ppm for CO-tolerant alloy anodes. Therefore, the remaining carbon monoxide must be removed from the hydrogen-rich gas. Currently, three methods, H2 separation, methanation, and preferential oxidation of CO, have been successfully used to lower the CO concentration of the hydrogen fuel. However, considering the low efﬁciency of hydrogen separation by palladium membrane and hydrogen consumption by CO methanation [102, 130], the selective oxidation of CO is the current preferred method for CO removal from hydrogen gas. Table 26.6 lists some of the zeolite catalysts tested for selective CO oxidation and their operating conditions and performances. The Pt/NaA catalyst prepared by Watanabe and coworkers [147] displayed comparable activity to the Pt/Al2 O3 catalyst, but was 10-fold more selective for CO oxidation. Improvements over the years have led to higher CO conversion and selectivity, exempliﬁed by Pt/mordenite [146] and Pt–Fe/mordenite [141, 144] catalysts. The latter exhibited remarkable preferential CO oxidation activity even at extremely high space velocity (about 100% selectivity, space velocity = ∼105 hour−1 ) and low temperature (i.e., 323 K). It was speculated that the high activity was due to the Pt acting as a CO adsorption site and Fe acting as an O2 dissociative-adsorption site. Luengnaruemitchai and coworkers [131, 138] prepared Au–Pt/NaA catalyst with comparable performance as the Pt–Fe/mordenite. The addition of gold improved CO selectivity at low temperatures (i.e., 100% over Au–Pt/NaA and ∼65% over Pt/NaA at 373 K).

26.4.3 Hydrogen Storage

Hydrogen storage is an important and urgent issue for practical application of PEMFCs in vehicles and portable devices. The U.S. Department of Energy has set an objective for hydrogen storage capacity of 62 Kg·m−3 and 6.5 wt%. Table 26.7 lists some of the reported uses of zeolites and molecular sieves for hydrogen storage. Typically, the hydrogen storage capacity is in direct proportion to the speciﬁc surface area of the materials. However, the mesoporous silicas and aluminosilicates [156, 157] had very low hydrogen storage capacities of less than 0.5 wt%. Zeolites [148–151] and mesoporous carbon materials [158, 160] had moderate hydrogen storage capacities but were still less than the target of 6.5 wt%. Mokaya and coworkers [152, 154, 155] prepared zeolite-templated carbons that possessed exceptional capacity to store hydrogen. Carbons prepared using Beta zeolite and acetonitrile displayed a speciﬁc surface area of 3200 m2 g−1 and a pore volume of 2.41 cm3 g−1 and reached a hydrogen storage of 6.9 wt% at 77 K and 2 MPa exceeding those of the best carbon nanotubes and metal-organic framework (MOF) materials [161–163].

1% CO, 1% O2 , 40% H2 , 0–10% CO2 , 0–10% H2 O, and He balance 1% CO, 1 (1.5)% O2 , 50 (60)% H2 , 0 (15)% CO2 , and N2 balance 1% CO, 2% O2 , 37% H2 , 18% CO2 , 5% H2 O, and He balance 1% CO, 1% O2 , and H2 balance 1% CO, 1% O2 , 40% H2 , 0–10% CO2 , 0–10% H2 O, and He balance 0.5% CO, 1% O2 , 37% H2 , 18% CO2 , 5% H2 O, and He balance 1% CO, 1% O2 , 40% H2 , and He balance

Au-Pt/NaA

Au–Pt/NaA

Pt/KA

Pt/NaA

Pt-Fe/mordenite

Rh/KA, NaA, CaA

Pt/NaY membrane

Feed composition

100

100

393–573

323–583

95

100

373–403 373–593

100

100

473–493

353–413

99.8

CO conversion (%)a

323–573

Temperature (K)

100 (<373 K)

N

(continued overleaf)

[138]

[137]

<10

N

[136]

[135]

[133, 134]

N

<10

∼60 80

<10

∼50

[132]

[131]

Nb

10–50

References

Exit cO concentration (ppm)

50

58.0

CO selectivity (%)a

Comparison of catalytic performance over various zeolite catalysts for preferential CO oxidation.

Catalyst

Table 26.6

26.4 Zeolites and Molecular Sieves in Fuel Processing 851

10.5% CO, 8.7% O2 , and H2 balance 1% CO, 1% O2 , 20% CO2 , 20% H2 O, and H2 balance 1% CO, 0.5% O2 , and H2 balance 1% CO, 0.5% O2 , and H2 balance 0.5% CO, 1% O2 , 37% H2 , 18% CO2 , 5% H2 O, and He balance 1% CO, 0.5% O2 , and H2 balance 0.5% CO, 0.6% O2 , 0.6% CO2 , 48.6% H2 , and He balance 1% CO, 0.5–2.0% O2 , and H2 balance

Pt/NaY membrane

b

423–623

523

88 (Pt/NaA), 66 (Pt/NaX), 79 (Pt/mordenite)

N

100

353–573

39 (Pt/NaA), 64 (Pt/NaX), 84 (Pt/mordenite)

N

100

50

∼70

∼60 100

100

75

62

CO selectivity (%)a

100

373–573

473

323–573

100

98

433–513 373–573

CO conversion (%)a

Temperature (K)

data are for the highest conversion unless otherwise indicated. ‘‘N’’ means that data are not available from the reference.

a The

Pt/NaA, NaX, and mordenite

Pt/NaY membrane

Pt–Fe/mordenite

Ru/KA, NaA, CaA

Pt/ZSM-5

Pt–Fe/mordenite

Pt–Fe/mordenite

Feed composition

(continued )

Catalyst

Table 26.6

[146, 147]

[145]

<8

N

[144]

[143]

[142]

[141]

N

10

N

N

[140]

[139]

<50 N

References

Exit cO concentration (ppm)

852

26 Zeolites and Molecular Sieves in Fuel Cell Applications

77

Na, Mg, K, Ca, Rb, Sr, CsY

77

1.5

NaX (hydrothermal synthesis) others (ion-exchanged NaX) NaY (commercial zeolite) others (ion-exchanged NaX)

Na, Mg, K, Ca, Rb, Sr, CsX

447–669 (Cs
1.6 0.1

[151]

[150]

[149]

[148]

References

(continued overleaf)

1.33–1.87 (Cs
1.32–2.19 (Cs
1.50, 1.46, 1.33

1.02

717, 642, 570

Ion exchange

Li-ABW

LiX, NaX, KX

1.64

1.6

Na-MAZ

1.75

2.07

1.6

1.6

Hydrothermal synthesis

Na-LEV

1.74

H-OFF

1.49

1.51

NaX

77

1.42 (77 K), 0.55 (195 K) N.A.

1.0 (77 K), 0.50 (195 K)

1.54

Commercial zeolite

1.7

H2 storage capacity (wt%)

NaA

77,195

Pressure (MPa)

SAPO-34

N.A.

Temperature (K)

1.30 (77 K), 0.48 (195 K)

Hydrothermal synthesis

AlPO4-5

SBET (m2 g –1 )

AlPO4-53

Preparation method

Comparison of properties and performances of zeolites and mesoporous materials as hydrogen storage media.

Material

Table 26.7

26.4 Zeolites and Molecular Sieves in Fuel Processing 853

Beta as template, acetonitrile as carbon precursor, CVD method NaY as template, propylene/butylene as carbon precursor, CVD method Beta as template, acetonitrile as carbon precursor, CVD method 13X or Y as template, acetonitrile as carbon precursor, CVD method Microwave synthesis

Hydrothermal synthesis

Zeolite-template carbon

Al-MCM-41 Ni/Al-MCM-41

Ni/MCM-41

Zeolite-template carbon

Zeolite-template carbon

Zeolite-template carbon

Preparation method

(continued )

Material

Table 26.7

1 324 (Ni/Si = 0),1188 (0.02), 1060 (0.06), 858 (0.1) 986 N

3 189 2 611 3 150 1 589 (13X as template) 1 825 (Y as template)

2 535 (CVD at 1 073 K) 2 470 (CVD at 1 123 K) 1 721 (CVD at 1 173 K) 2 117

SBET (m2 g –1 )

0.1

15

RTb

2

2

0.1

0.1, 2

Pressure (MPa)

77

77

77

77

77

Temperature (K)

0.07 0.09–0.19

0.44 (0), 0.53 (0.02), 0.36 (0.06), 0.33 (0.1)

6.0 5.5 6.9 3.4 4.5

2.3 (0.1 MPa), 5.3 (2 MPa) 2.0 (0.1 MPa), 5.2 (2 MPa) 1.2 (0.1 MPa), 3.3 (2 Mpa) 2

H2 storage capacity (wt%)

[157]

[156]

[155]

[154]

[153]

[152]

References

854

26 Zeolites and Molecular Sieves in Fuel Cell Applications

Solvothermal synthesis

Mesoporous silica monolith as template, acetonitrile as carbon precursor, CVD method Sol-gel method Acid-treatment commercial carbon nanotubes Carbonized phenolic resin Solvothermal synthesis

‘‘N’’ means that data are not available from the reference. room temperature.

b RT:

a

IRMOF-8 MOF-505

Activated carbon foam MOF-5

Pd/silica nanotube Single-walled carbon nanotube

Mesoporous carbon monolith

2 1 0.1

298 298 77

N 1 830

0.07

78

2 500−3 000

1.0 2.0 2.47

4.5

2.3

1.85 1.8

∼3.5 0.1

–

3.4

2

–

RT 77

77

2 320

264.3 1 300

1 090

[163]

[161, 162]

–

[159] [160]

[158]

26.4 Zeolites and Molecular Sieves in Fuel Processing 855

856

26 Zeolites and Molecular Sieves in Fuel Cell Applications

26.5 Summary and Outlook

This review shows some of the exciting research in the application of zeolites and molecular sieves in fuel cells for clean energy generation. Zeolites can play an important role in fuel reﬁning and reforming, as well as in hydrogen production, puriﬁcation, conditioning, and storage. Researches also show that the use of zeolites as electrolyte has advantages over PFSA membrane for high-temperature PEMFC applications. Furthermore, zeolites are capable of ameliorating fuel crossover problem in DMFCs. Challenges remain in preparing new zeolites and molecular sieves with better proton conductivity. Although the application of zeolites in electrocatalysts is hindered by their poor electrical conductivity, several exciting works have shown that zeolite electrocatalysts are efﬁcient for methanol and ethanol electroxidation and electrocatalysts derived from zeolite-templated carbons provide superb results. Zeolites will continue to play an important role in fuel conversion and conditioning for stationary, mobile, and portable fuel cell devices. Although zeolite research into proton conducting membrane in the near future is expected to focus primarily on polymer composite membranes, there is expected to be increasing research into zeolite membrane proton conductor. The advances in zeolite synthesis and their assembly and fabrication into complex forms and structures are important for their application as proton conducting electrolyte membrane for fuel cell devices. Ultrathin zeolite membrane ﬁlms and layers could improve the proton conductivity beyond today’s thick PFSA membranes and microfabrication would enable their miniaturization for portable energy generation for electronic devices. Membrane design needs to be optimized for surface area, strength, and portability. Hybrid membrane consisting of different zeolite layers optimized for catalyst support, fuel barrier, and proton conduction is possible and advantageous for fuel cell operation.

Acknowledgment

The authors gratefully acknowledge ﬁnancial supports from the Hong Kong Research Grant Council. References 1. Sing, K.S.W., Everett, D.H., Haul,

R.A.W., Moscou, L., Pierotti, R.A., Rouquerol, J., and Siemieniewska, T. (1985) Pure Appl. Chem., 57, 603–619. 2. Beck, J.S., Vartuli, J.C., Roth, W.J., Leonowicz, M.E., Kresge, C.T., Schmitt, K.D., Chu, C.T.W., Olson, D.H., Sheppard, E.W., Mccullen, S.B., Higgins, J.B., and Schlenker,

J.L. (1992) J. Am. Chem. Soc., 114, 10834–10843. 3. Kresge, C.T., Leonowicz, M.E., Roth, W.J., Vartuli, J.C., and Beck, J.S. (1992) Nature, 359, 710–712. 4. Cundy, C.S. and Cox, P.A. (2003) Chem. Rev., 103, 663–701. 5. Tosheva, L. and Valtchev, V.P. (2005) Chem. Mater., 17, 2494–2513.

References 6. Cundy, C.S. and Cox, P.A. (2005)

7. 8. 9.

10. 11.

12. 13.

14.

15.

16.

17. 18. 19. 20.

21.

22. 23.

24.

25.

Microporous Mesoporous Mater., 82, 1–78. Corma, A. (1997) Chem. Rev., 97, 2373–2419. Davis, M.E. (2002) Nature, 417, 813–821. Taguchi, A. and Schuth, F. (2005) Microporous Mesoporous Mater., 77, 1–45. ˇ Cejka, J. (2003) Appl. Catal. A, 254, 327–338. Shao, Y.Y., Yin, G.P., Wang, Z.B., and Gao, Y.Z. (2007) J. Power Sources, 167, 235–242. Roziere, J. and Jones, D.J. (2003) Ann. Rev. Mater. Res., 33, 503–555. Li, Q.F., He, R.H., Jensen, J.O., and Bjerrum, N.J. (2003) Chem. Mater., 15, 4896–4915. Zhang, J.L., Xie, Z., Zhang, J.J., Tanga, Y.H., Song, C.J., Navessin, T., Shi, Z.Q., Song, D.T., Wang, H.J., Wilkinson, D.P., Liu, Z.S., and Holdcroft, S. (2006) J. Power Sources, 160, 872–891. Hogarth, W.H.J., da Costa, J.C.D., and Lu, G.Q. (2005) J. Power Sources, 142, 223–237. Curtin, D.E., Lousenberg, R.D., Henry, T.J., Tangeman, P.C., and Tisack, M.E. (2004) J. Power Sources, 131, 41–48. Alberti, G. and Casciola, M. (2003) Ann. Rev. Mater. Res., 33, 129–154. Litster, S. and McLean, G. (2004) J. Power Sources, 130, 61–76. Antolini, E. (2004) J. Appl. Electrochem., 34, 563–576. Kreuer, K.D., Paddison, S.J., Spohr, E., and Schuster, M. (2004) Chem. Rev., 104, 4637–4678. Hickner, M.A., Ghassemi, H., Kim, Y.S., Einsla, B.R., and McGrath, J.E. (2004) Chem. Rev., 104, 4587–4611. Heinzel, A. and Barragan, V.M. (1999) J. Power Sources, 84, 70–74. Ravikumar, M.K. and Shukla, A.K. (1996) J. Electrochem. Soc., 143, 2601–2606. Haile, S.M., Boysen, D.A., Chisholm, C.R.I., and Merle, R.B. (2001) Nature, 410, 910–913. Hogarth, W.H.J. (2006) PEM fuel cells: from inorganic proton conducting

26.

27.

28.

29. 30. 31. 32. 33. 34.

35.

36.

37.

38.

39.

40.

41. 42. 43.

membranes to process design. Ph.D. thesis, University of Queensland. Hogarth, W.H.J., da Costa, J.C.D., Drennan, J., and Lu, G.Q. (2005) J. Mater. Chem., 15, 754–758. Tang, H.L. and Jiang, S.P. (2008) Inorganic proton conducting membranes for high temperature PEM fuel cell: concept, synthesis and performance. 214th ECS Meeting. Ben Saad, K., Hamzaoui, H., and Mohamed, M.M. (2007) Mater. Sci. Eng. B, 139, 226–231. Afanassyev, I.S. and Moroz, N.K. (2003) Solid State Ionics, 160, 125–129. Franke, M.E. and Simon, U. (1999) Solid State Ionics, 118, 311–316. Franke, M.E. and Simon, U. (2004) Chemphyschem, 5, 465–472. Fermann, J.T. and Auerbach, S. (2000) J. Chem. Phys., 112, 6787–6794. Sierka, M. and Sauer, J. (2001) J. Phys. Chem. B, 105, 1603–1613. Franke, M.E., Sierka, M., Simon, U., and Sauer, J. (2002) Phys. Chem. Chem. Phys., 4, 5207–5216. Holmberg, B.A., Hwang, S.J., Davis, M.E., and Yan, Y.S. (2005) Microporous Mesoporous Mater., 80, 347–356. Knudsen, N., Andersen, E.K., Andersen, I.G.K., and Skou, E. (1989) Solid State Ionics, 35, 51–55. Knudsen, N., Andersen, E.K., Andersen, I.G.K., and Skou, E. (1988) Solid State Ionics, 28, 627–631. Kanellopoulos, J., Gottert, C., Schneider, D., Knorr, B., Prager, D., Ernst, H., and Freude, D. (2008) J. Catal., 255, 68–78. Simon, U. and Franke, M.E. (2005) Host-guest-systems Based on Nanoporous Crystals, Chapter 19, Wiley-VCH Verlag GmbH & Co. KgaA, pp. 364–378. Kreuer, K.D., Rabenau, A., and Weppner, W. (1982) Angew. Chem. Int. Ed., 21, 208–209. Howe, A.T. and Shilton, M.G. (1979) J. Solid State Chem., 23, 345. Howe, A.T. and Shilton, M.G. (1980) J. Solid State Chem., 34, 149. Bernard, L., Fitch, A., Wright, Af., Fender, B.E.F., and Howe, A.T. (1981) Solid State Ionics, 5, 459.

857

858

26 Zeolites and Molecular Sieves in Fuel Cell Applications 44. Sancho, T., Lemus, J., Urbiztondo,

45.

46.

47.

48.

49.

50. 51.

52. 53.

54.

55.

56.

57.

58.

59. 60.

M., Soler, J., and Pina, M.P. (2008) Microporous Mesoporous Mater., 115, 206–213. Ahmad, M.I., Zaidi, S.M.J., Rahman, S.U., and Ahmed, S. (2006) Microporous Mesoporous Mater., 91, 296–304. Mckeen, J.C., Yan, Y.S., and Davis, M.E. (2008) Chem. Mater., 20, 3791–3793. Mckeen, J.C., Yan, Y.S., and Davis, M.E. (2008) Chem. Mater., 20, 5122–5124. Coutinho, D., Yang, Z.W., Ferraris, J.P., and Balkus, K.J. (2005) Microporous Mesoporous Mater., 81, 321–332. Marschall, R., Rathousky, J., and Wark, M. (2007) Chem. Mater., 19, 6401–6407. Zheng, N.F., Bu, X.H., and Feng, P.Y. (2003) Nature, 426, 428–432. Wang, Y.B., Yang, D., Zheng, X.H., Jiang, Z.Y., and Li, J. (2008) J. Power Sources, 183, 454–463. Holmberg, B.A., Wang, X., and Yan, Y.S. (2008) J. Membr. Sci., 320, 86–92. Carbone, A., Sacca, A., Gatto, I., Pedicini, R., and Passalacqua, E. (2008) Int. J. Hydrogen Energy, 33, 3153–3158. Wang, J.T., Zheng, X.H., Wu, H., Zheng, B., Jiang, Z.Y., Hao, X.P., and Wang, B.Y. (2008) J. Power Sources, 178, 9–19. Wu, H., Zheng, B., Zheng, X.H., Wang, J.T., Yuan, W.K., and Jiang, Z.Y. (2007) J. Power Sources, 173, 842–852. Yuan, W.K., Wu, H., Zheng, D., Zheng, X.H., Jiang, Z.Y., Hao, X.P., and Wang, B.Y. (2007) J. Power Sources, 172, 604–612. Li, X., Roberts, E.P.L., Holmes, S.M., and Zholobenko, V. (2007) Solid State Ionics, 178, 1248–1255. Gribov, E.N., Parkhomchuk, E.V., Krivobokov, I.M., Darr, J.A., and Okunev, A.G. (2007) J. Membr. Sci., 297, 1–4. Sancho, T., Soler, J., and Pina, M.P. (2007) J. Power Sources, 169, 92–97. Byun, S.C., Jeong, Y.J., Park, J.W., Kim, S.D., Ha, H.Y., and Kim, W.J. (2006) Solid State Ionics, 177, 3233–3243.

61. Chen, Z.W., Holmberg, B., Li, W.Z.,

62.

63.

64. 65. 66.

67.

68.

69.

70. 71. 72.

73.

74.

75. 76. 77.

78.

Wang, X., Deng, W.Q., Munoz, R., and Yan, Y.S. (2006) Chem. Mater., 18, 5669–5675. Ahmad, M.I., Zaidi, S.M.J., and Rahman, S.U. (2006) Desalination, 193, 387–397. Baglio, V., Di Blasi, A., Arico, A.S., Antonucci, V., Antonucci, P.L., Nannetti, F., and Tricoli, V. (2005) Electrochim. Acta, 50, 5181–5188. Tricoli, V. and Nannetti, F. (2003) Electrochim. Acta, 48, 2625–2633. Libby, B., Smyrl, W.H., and Cussler, E.L. (2003) AIChE J., 49, 991–1001. Yahiro, H., Konda, Y., and Okada, G. (2003) Phys. Chem. Chem. Phys., 5, 620–623. Poltarzewski, Z., Wieczorek, W., Przyluski, J., and Antonucci, V. (1999) Solid State Ionics, 119, 301–304. Costamagna, P., Yang, C., Bocarsly, A.B., and Srinivasan, S. (2002) Electrochim. Acta, 47, 1023–1033. Nunes, S.P., Ruffmann, B., Rikowski, E., Vetter, S., and Richau, K. (2002) J. Membr. Sci., 203, 215–225. Ramani, V., Kunz, H.R., and Fenton, J.M. (2004) J. Membr. Sci., 232, 31–44. Chang, H.Y. and Lin, C.W. (2003) J. Membr. Sci., 218, 295–306. Baglio, V., Arico, A.S., Di Blasi, A., Antonucci, V., Antonucci, P.L., Licoccia, S., Traversa, E., and Fiory, F.S. (2005) Electrochim. Acta, 50, 1241–1246. Takami, M., Yamazaki, Y., and Hamada, H. (2001) Electrochemistry, 69, 98–103. Son, D.H., Sharma, R.K., Shul, Y.G., and Kim, H. (2007) J. Power Sources, 165, 733–738. Kwan, S.M. and Yeung, K.L. (2008) Chem. Commun., 3631–3633. Yeung, K.L., Kwan, S.M., and Lau, W.N. (2009) Topic Catal., 52, 101–110. Sarv, P., Tuherm, T., Lippmaa, E., Keskinen, K., and Root, A. (1995) J. Phys. Chem. B, 99, 13763. Caro, J. and Noack, M. (2008) Microporous Mesoporous Mater., 115, 215–233.

References 79. Lai, S.M., Au, L.T.Y., and Yeung, K.L.

80.

81.

82. 83.

84.

85.

86.

87. 88.

89.

90.

91. 92. 93.

94.

95. 96.

(2002) Microporous Mesoporous Mater., 54, 63–77. Wong, W.C., Au, L.T.Y., Tellez, C., and Yeung, K.L. (2001) J. Membr. Sci., 191, 143–163. Wong, W.C., Au, L.T.Y., Lau, P.S., and Tellez, C. (2001) J. Membr. Sci., 193, 141–161. Au, L.T.Y. and Yeung, K.L. (2001) J. Membr. Sci., 194, 33–55. Au, L.T.Y., Mui, W.Y., Lau, P.S., Tellez, C., and Yeung, K.L. (2001) Microporous Mesoporous Mater., 47, 203–216. Wan, Y.S.S., Chau, J.L.H., Gavriilidis, A., and Yeung, K.L. (2001) Microporous Mesoporous Mater., 42, 157–175. Chau, J.L.H., Wan, Y.S.S., Gavriilidis, A., and Yeung, K.L. (2002) Chem. Eng. J., 88, 187–200. Wan, Y.S.S., Chau, J.L.H., Gavriilidis, A., and Yeung, K.L. (2002) Chem. Commun., 878–879. Chau, J.L.H. and Yeung, K.L. (2002) Chem. Commun., 960–961. Lai, S.M., Martin-Aranda, R., and Yeung, K.L. (2003) Chem. Commun., 218–219. Lai, S.M., Ng, C.P., Martin-Aranda, R., and Yeung, K.L. (2003) Microporous Mesoporous Mater., 66, 239–252. Zhang, X.F., Lai, S.M., Martin-Aranda, R., and Yeung, K.L. (2004) Appl. Catal. A, 261, 109–118. Chau, J.L.H., Leung, Y.L.A., and Yeung, K.L. (2003) Lab Chip, 3, 53–55. Yeung, Y.L.A. and Yeung, K.L. (2004) Chem. Eng. Sci., 59, 4809–4817. Chau, J.L.H., Leung, A.Y.L., Shing, M.B., Yeung, K.L., and Chan, C.M. (2003) Hydrogen and proton transport properties of nanoporous zeolite micromembranes, in Nano Science and Technology: Novel Structure and Phenomena (eds Z. Tang and P. Sheng), Taylor and Francis, London, pp. 228–232. Yasumoto, E., Hatoh, K., and Gamou, T., (1997) US Patent No. 5,702,838, December 30, 1997. Samant, P. and Fernades, J.B. (2004) J. Power Sources, 125, 172–177. Pang, H., Chen, J., Yang, L., Liu, B., Zhong, X., and Wei, X. (2008) J. Solid State Electrochem., 12, 237–243.

97. Alvaro, M., Cabeza, J.F., Fabuel,

98.

99.

100.

101. 102. 103.

104.

105.

106.

107.

108.

109.

110.

111. 112.

113. 114.

D., Garcia, H., Guijarro, E., and Martinez de Juan, J.L. (2006) Chem. Mater., 18, 26–33. Coker, E.N., Steen, W.A., and Miller, J.E. (2007) Microporous Mesoporous Mater., 104, 236–247. Coker, E.N., Steen, W.A., Miller, J.T., Kropf, A.J., and Miller, J.E. (2007) J. Mater. Chem., 17, 3330–3340. Choi, W.C., Woo, S.I., Jeon, M.K., Sohn, J.M., Kim, M.R., and Jeon, H.J. (2005) Adv. Mater., 17, 446–451. Wen, Z.H., Liu, J., and Li, J.H. (2008) Adv. Mater., 20, 743–747. Song, C.S. (2002) Catal. Today, 77, 17–49. Wang, Y.H., Yang, R.T., and Heinzel, J.M. (2009) Ind. Eng. Chem. Res., 48, 142–147. Park, J.G., Ko, C.H., Yi, K.B., Park, J.H., Han, S.S., Cho, S.H., and Kim, J.N. (2008) Appl. Catal. B, 81, 244–250. Zhang, Z.Y., Shi, T.B., Jia, C.Z., Ji, W.J., Chen, Y., and He, M.Y. (2008) Appl. Catal. B, 82, 1–10. Tang, K., Song, L.J., Duan, L.H., Li, X.Q., Gui, J.Z., and Sun, Z.L. (2008) Fuel Process. Technol., 89, 1–6. Wang, Y., Yang, R.T., and Heinzel, J.M. (2008) Chem. Eng. Sci., 63, 356–365. Ko, C.H., Song, H.I., Park, J.H., Han, S.S., and Kim, J.N. (2007) Korean J. Chem. Eng., 24, 1124–1127. Li, W.L., Liu, Q.F., Xing, J.M., Gao, H.S., Xiong, X.C., Li, Y.G., Li, X., and Liu, H.Z. (2007) AIChE J., 53, 3263–3268. Yang, L.M., Wang, Y.J., Huang, D., Luo, G.S., and Dai, Y.Y. (2007) Ind. Eng. Chem. Res., 46, 579–583. Jiang, M. and Ng, F.T.T. (2006) Catal. Today, 116, 530–536. Xue, M., Chitrakar, R., Sakane, K., Hirotsu, T., Ooi, K., Yoshimura, Y., Toba, M., and Feng, Q. (2006) J. Colloid Interface Sci., 298, 535–542. Jayaraman, A., Yang, F.H., and Yang, R.T. (2006) Energy Fuels, 20, 909–914. Velu, S., Song, C.S., Engelhard, M.H., and Chin, Y.H. (2005) Ind. Eng. Chem. Res., 44, 5740–5749.

859

860

26 Zeolites and Molecular Sieves in Fuel Cell Applications 115. Hernandez-Maldonado, A.J., Yang,

116. 117.

118.

119.

120.

121. 122. 123. 124.

125. 126.

127.

128.

129. 130. 131.

132.

133.

R.T., and Cannella, W. (2004) Ind. Eng. Chem. Res., 43, 6142–6149. Hernandez-Maldonado, A.J. and Yang, R.T. (2004) AIChE J., 50, 791–801. Hernandez-Maldonado, A.J. and Yang, R.T. (2004) Ind. Eng. Chem. Res., 43, 1081–1089. Yang, R.T., Hernandez-Maldonado, A.J., and Yang, F.H. (2003) Science, 301, 79–81. Hernandez-Maldonado, A.J. and Yang, R.T. (2003) Ind. Eng. Chem. Res., 42, 3103–3110. Haryanto, A., Fernando, S., Murali, N., and Adhikari, S. (2005) Energy Fuels, 19, 2098–2106. Armor, J.N. (1999) Appl. Catal. A, 176, 159–176. Wang, L.S., Murata, K., and Inaba, M. (2004) Appl. Catal. A, 257, 43–47. Lee, I.C. (2008) Catal. Today, 136, 258–265. Campos-Skrobot, F.C., Rizzo-Domingues, R.C.P., Fernandes-Machado, N.R.C., and Cantao, M.P. (2008) J. Power Sources, 183, 713–716. Choudhary, T.V. and Goodman, D.W. (2002) Catal. Today, 77, 65–78. Ashok, J., Kumar, S.N., Venugopal, A., Kumari, V.D., and Subrahmanyam, M. (2007) J. Power Sources, 164, 809–814. Ashok, J., Subrahmanyam, M., and Venugopal, A. (2008) Catal. Surv. Asia, 12, 229–237. Li, L., Zhu, Z.H., Yan, Z.F., Lu, G.Q., and Rintoul, L. (2007) Appl. Catal. A, 320, 166–172. Hashimoto, K. and Toukai, N. (2000) J. Mol. Catal. A., 161, 171–178. Park, E.D., Lee, D., and Lee, H.C. (2009) Catal. Today, 130, 280–290. Luengnaruemitchai, A., Naknam, P., and Wongkasemjit, S. (2008) Ind. Eng. Chem. Res., 47, 8160–8165. Bernardo, P., Algieri, C., Barbieri, G., and Drioli, E. (2008) Sep. Purif. Technol., 62, 629–635. Galletti, C., Specchia, S., Saracco, G., and Specchia, V. (2008) Ind. Eng. Chem. Res., 47, 5304–5312.

134. Galletti, C., Fiorot, S., Specchia, S.,

135.

136.

137.

138.

139.

140.

141.

142.

143.

144.

145.

146.

147.

148.

149.

150. 151.

Saracco, G., and Specchia, V. (2007) Top. Catal., 45, 15–19. Maeda, N., Matsushima, T., Uchida, H., Yamashita, H., and Watanabe, M. (2008) Appl. Catal. A, 341, 93–97. Luengnaruemitchai, A., Nimsuk, M., Naknam, P., Wongkasemjit, S., and Osuwan, S. (2008) Int. J. Hydrogen Energy, 33, 206–213. Rosso, I., Galletti, C., Fiorot, S., Saracco, G., Garrone, E., and Specchia, V. (2007) J. Porous Mater., 14, 245–250. Naknam, P., Luengnaruemitchai, A., Wongkasemjit, S., and Osuwan, S. (2007) J. Power Sources, 165, 353–358. Bernardo, P., Algieri, C., Barbieri, G., and Drioli, E. (2006) Catal. Today, 118, 90–97. Kotobuki, M., Watanabe, A., Uchida, H., Yamashita, H., and Watanabe, M. (2006) Appl. Catal. A, 307, 275–283. Kotobuki, M., Watanabe, A., Uchida, H., Yamashita, H., and Watanabe, M. (2005) J. Catal., 236, 262–269. Kotobuki, M., Watanabe, A., Uchida, H., Yamashita, H., and Watanabe, M. (2005) Chem. Lett., 34, 866–867. Rosso, I., Antonini, M., Galletti, C., Saracco, G., and Specchia, V. (2004) Top. Catal., 30-31, 475–480. Watanabe, M., Uchida, H., Ohkubo, K., and Igarashi, H. (2003) Appl. Catal. B, 46, 595–600. Sotowa, K.I., Hasegawa, Y., Kusakabe, K., and Morooka, S. (2002) Int. J. Hydrogen Energy, 27, 339–346. Igarashi, H., Uchida, H., Suzuki, M., Sasaki, Y., and Watanabe, M. (1997) Appl. Catal. A, 159, 159–169. Watanabe, M., Uchida, H., Igarashi, H., and Suzuki, M. (1995) Chem. Lett., 21–22. Dong, J.X., Ban, G.Z., Zhao, Q., Liu, L., and Li, J.P. (2008) AIChE J., 54, 3017–3025. Dong, J.X., Wang, X.Y., Xu, H., Zhao, Q., and Li, J.P. (2007) Int. J. Hydrogen Energy, 32, 4998–5004. Li, Y.W. and Yang, R.T. (2006) J. Phys. Chem. B, 110, 17175–17181. Langmi, H.W., Book, D., Walton, A., Johnson, S.R., Al-Mamouri, M.M.,

References

152. 153.

154.

155.

156.

157.

Speight, J.D., Edwards, P.P., Harris, I.R., and Anderson, P.A. (2005) J. Alloys Compd., 404, 637–642. Pacula, A. and Mokaya, R. (2008) J. Phys. Chem. C, 112, 2764–2769. Chen, L., Singh, R.K., and Webley, P. (2007) Microporous Mesoporous Mater., 102, 159–170. Yang, Z.X., Xia, Y.D., and Mokaya, R. (2007) J. Am. Chem. Soc., 129, 1673–1679. Yang, Z.X., Xia, Y.D., Sun, X.Z., and Mokaya, R. (2006) J. Phys. Chem. B, 110, 18424–18431. Wu, C.D., Gao, Q.M., Hu, J., Chen, Z., and Shi, W. (2009) Microporous Mesoporous Mater., 117, 165–169. Ramachandran, S., Ha, J.H., and Kim, D.K. (2007) Catal. Commun., 8, 1934–1938.

158. Xia, Y. and Mokaya, R. (2007) J. Phys.

Chem. C, 111, 10035–10039. 159. Jung, J.H., Rim, J.A., Lee, S.J., Cho,

160.

161.

162.

163.

S.J., Kim, S.Y., Kang, J.K., Kim, Y.M., and Kim, Y.J. (2007) J. Phys. Chem. C, 111, 2679–2682. Takagi, H., Hatori, H., Soneda, Y., Yoshizawa, N., and Yamada, Y. (2004) Mater. Sci. Eng. B, 108, 143–147. Li, H., Eddaoudi, M., O’Keeffe, M., and Yaghi, O.M. (1999) Nature, 402, 276–279. Rosi, N.L., Eckert, J., Eddaoudi, M., Vodak, D.T., Kim, J., O’Keeffe, M., and Yaghi, O.M. (2003) Science, 300, 1127–1129. Chen, B.L., Ockwig, N.W., Millward, A.R., Contreras, D.S., and Yaghi, O.M. (2005) Angew. Chem. Int. Ed., 44, 4745–4749.

861

863

Index

a Ab initio wavefunction methods 303 ABC pattern, in FTIR spectrum 511–512 absolute temperature 3 accessibility index (ACI) 217 P-acetylisobutylbenzene 779f acid–base reactions, within zeolite-based materials 211–217 acid-catalyzed reactions – acetalization reaction of carbonyl groups 787–789 – Baeyer–Villiger reaction 799–803 – Diels–Alder reaction 783–787 – epoxidation reactions 795–798 – Fischer glycosidation reactions 789–791 – Friedel–Crafts acylation reactions 775–780 – hydroxyalkylation of aromatics 780–783 – isomerizations 792–794 – Meerwein–Ponndorf–Verley (MPV) reduction and Oppenauer’s oxidation (O) of alcohols 803–808 acid sites, in zeolites – acetonitrile adsorptions 511–512 – acid strength 212 – 27 Al/31 P insensitive nuclei, characterization of 519 – bridging of OH groups 494 – catalytic activity 498–499 – characterization of 498–521 – chemical equilibrium 500 – dealumination of 495–497, 510 – differential heat curves plotted versus the acid site coverage 505 – differential heat of n-hexane adsorption 507 – differential heats of ammonia adsorption 506f

– disproportionation of ethylbenzene 498–499 – electronegativities, effect of 495 – formation of Bronsted and Lewis 496–498, 499 – framework aluminum content, linearity of 498 – FTIR spectroscopic characterization 508–514 – gas-phase adsorption microcalorimetric characterization 504–507 – gas-phase adsorption of reactants 507 – H-FER zeolite spectra 512, 513f – 1 H MAS NMR spectrum 514–517, 519, 520f – heat of adsorption of probe molecules and reactants on 504–507 – heat treatment/steaming, impact 497 – hydroxyl groups at lanthanum cations (La(OH)n ) 520–521 – medium-pore zeolites, characterization of 501 – nature of 494–496 – NMR spectroscopic characterization 514–521 – NNN concept 495 – 31 P MAS NMR spectrum 517–519 – pivalonitrile-loaded zeolite SSZ-33 512 – preparation of the H-form 497 – preparation of the Na-form 497 – proton transfers 500 – reactions corresponding to Hirschler–Plank mechanism 496 – Si–O–Al arrangement 494–496 – Si–O–T arrangements 495 – SiOHAl groups 519–520 – strength of Brønsted acid sites 513

Zeolites and Catalysis, Synthesis, Reactions and Applications. Vol. 2. ˇ Edited by Jiˇr´ı Cejka, Avelino Corma, and Stacey Zones Copyright  2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim ISBN: 978-3-527-32514-6

864

Index acid sites, in zeolites (contd.) – stretching vibrations of SiOHAl groups 509 – temperature-programmed desorption (TPD) 501–504 – terminal OH groups 495 – titration with bases and indicators 500–501 – tri-coordinated frameworks 497–498 – vibrations (vOH )of hydroxyl groups 508–510 acidity index 166 adhesive-type growth 11 adsorption 339–341 – acetonitrile 511–512 – of benzene 350–351 – carbon dioxide 351–352 – gas-phase metal-carbonyl 661 – H2 534 – importance of 364 – isosteric heats of 340 – of liquid-phase isotherms of water 339f – of N2 O4 complex on alkali-exchanged faujasite 322–325 – physical adsorption of the microparticles 414 – of probe molecules and reactants on acidic zeolites 504–507 – of pyrrole at alkali metal ions 525–526 – of water in MFI 347f – on zeolites in reﬁneries 464–466 adsorption constant (K) 367 AFI structure – effect of triethylamine and water molecules in the synthesis of 120f – SDAs 119 aﬂatoxicosis 401 aﬂatoxins 400 Akzo-Fina–CFI Dewaxing 613 Al-Beta zeolite 801 AlCl3 635 AlCl3 -based catalysts 549 alcohols 89, 93 aldol condensations 816–818 ALF-n 96 Al-free Ti-Beta zeolite 804 alkali-exchanged zeolite Y 522 alkali-ion-exchanged Y zeolites 814 alkali metal hydroxide molten salts 88 alkali oxide impregnated zeolites 524, 526f alkoxides 396 4-alkoxy phenols 800 4-alkoxybenzaldehydes 801f alkyl amines 412

alkylations of aromatics – ethylation of benzene 635–636 – methylation of toluene 636–639 – of toluene and ethylbenzene 640–641 Allred–Rochow scheme 145 all-silica EUO structure 117 Al-MCM-41 794, 818 Al-MCM-58 156 Al-MFI 136 – molecular sieves 142f – morphology of 141 – (NH4)- 141 – pore structure of 139f – synthesis in glycerol solvent 141 Al–NH2 –Si–NH2 –Al 201 Al–NH2 –Si species 201 Al–O–P alternation 90 Al8 (PO4 )10 H3 ·E3C6 H11 N2 90 Al16 P12 Si6 O72 59 Al-SSZ-33 zeolites 159–161 Al-zeolite Beta 804–805 Al-ZSM-11 161–162 aluminium reinsertion process, using aqueous Al(NO3)3 solution 158–162 aluminophosphate HSAPO-34 (CHA topology) 319 aluminophosphates (AlPOs) 58–59, 100, 472 – AlPO-4 392 – AlPO-5 120–121, 142 – frameworks, synthesis of 59, 78–79, 90, 92 aluminosilicate zeolites 43, 494 aluminum, in a zeolite framework 361 – ab initio calculations of 296 – 27 Al MAS NMR spectra 283, 293f , 296 – ‘‘Al pairs’’ 294 – aluminum-rich cores 290 – aluminum species (AlOH) 514 – at 150 ◦ C 286 – catalytic activity in HY 287–288 – desilication of the crystals, impact 291 – distribution factor, impacts 283–284 – distribution over the crystallographic T sites 283–284, 292–296 – on the effect of SDA–zeolite interactions 70 – ﬂuorescence radiation emission 295 – formation of framework-associated octahedrally coordinated 286–289 – framework-associated aluminum 285–286 – Henry coefﬁcients of linear alkanes 296 – non-random aluminum distribution 294 – octahedrally coordinated aluminum 285 – reversible versus irreversible structural changes 285

Index – – – – – –

role of steaming 288 spatial distribution 290 structure 284–285 in structures of AlPOs and SAPOs 58–59 Td1 and Td2 resonances 292–294 tetrahedrally coordinated framework aluminums 284–285 – tetrapropylammonium 290 – under various conditions 289f – upon heating to 400◦ C 285 – zoning phenomenon 289–291 aluminum-phosphorus tetrahedra 58 aluminum-rich mesoporous aluminosilicate (AlHMS) 786 amber force ﬁeld 336 amine 1-methyl-imidazole (MIA) 116 aminoproply/PSS− /aminopropyl linkages 414 aminosilanes 395 amorphous reactants 11 amorphous silica–aluminas (ASAs) 566 amphiphilic organosilanes 479 Anderson–Schulz–Flory (ASF) distribution 660 anisotropic growth 41 anti-l¨owenstein Al–O–Al bond 59 AP-tethering microcrystals 427, 428f aperiodic interruptions 1 aromatic hydrocarbons – alkylations of 633–635 – ethylbenzene disproportionation 630–633 – structural characteristics of zeolites 625 – toluene disproportionation 625–630 – trimethylbenzenes via transalkylation with toluene or disproportionation reactions 633–635 – zeolites related to transformations of 623–625 Arrhenius plots 34 Arrhenius-type equation 5 atomic force microscopy (AFM) studies 2, 14 – growing facet of a HKUST-1 crystal 48f – growth of two interlaced spirals 45f – image of a face of an STA-7 crystal 39f – images of an elliptical spiral with cross section 40f – interlaced spiral growth on a ZnPO-SOD crystal 44f – silicate 30f – zeolite A crystal under static solution of 0.5M NaOH 29f – zeolite A (LTA) 23–25, 26f , 28f – of zeolite L 33f , 35f

Auger electron spectroscopy 290–291 azaoxacryptand template 72

b B88 functionals 304 11 B magic angle spinning nuclear magnetic resonance (MAS NMR) 159 B-(Al)-ZSM-11 161–162 Baeyer–Villiger oxidation 737, 799–803 bare zeolite microcrystals (BZ-G) 422 barite 10 base catalysis, in zeolites – analytical and spectroscopic methods for characterization of 525–529 – characterization of 523–529 – double bond isomerization of 2,3-dimethylbut-1-ene and 1-butene 523–524 – formation of 522–523 – hydrogen bonds, characterization of 525–526 – Knoevenagel condensation 525 – nature of 521–522 – NMR spectroscopic studies of 527–529 – Sanderson principle 521 base-catalyzed reactions – aldol condensations 816–818 – Knoevenagel reaction 809–813 – Michael addition 813–816 basic building units (BBUs) 132 basis sets 304–305, 339 BCF theory 10 BdB-FHH (Frenkel-Halsey-Hill) method 251 benzene alkylation 635–636 benzene, toluene, and xylene (BTXs) 585 benzene-1,3,5-tricarboxylate (BTC) groups 47 1-benzyl-1-methylpyrrolidinium (bmp) 110 (S)-N-benzyl-pyrrolidine-2-methanol 121 beryllium 66 BET (Brunauer–Emmett–Teller) theory 247 beta zeolite (BEA) 31, 62, 182, 195, 290, 293, 319, 381, 454, 481, 503, 608, 624, 628, 631, 634, 642, 791 bifunctional hydrocracking catalysts 566 bifunctional medium-pore zeolites 533 bimetallic zeolites 531 biot mass number (Bi) 373 birth-and-spreadmechanism 29 1,2-bis(triethoxysilyl)ethane (BTEE) 481 1,4-bis(N-methylpyrrolidinium) butane 60 2,5-bis(hydroxymethyl) furan 783 bis(2-hydroxyethyl)dimethylammonium chloride (TCl) 122

865

866

Index bis(triethoxysilyl)methane (BTEM) 481 BJH (Barret–Joyner–Halenda) method 249, 251 BJH-BdB (Brockhoff-de Boer) method 251, 256 BKS force ﬁeld 336 B3LYP functional 304 B3LYP/6–31G(d) level of theory 317 BMIM bromide 101 Boltzmann constant 3 boralites 66 boron 66 boron–oxygen tetrahedral bond 66 borosilicate ERS-12 110 borosilicate zeolites 66, 158, 67t B–O–Si bond angle 66 B3PW functional 304 bromide 97 Brønsted acid sites 60, 124, 212, 215, 216f Brownian motion 364 B-SSZ-33 160, 164–166 bulk crystal 3 Burchart force ﬁeld 336 Burgers vector 38 butane, π -complex of 309, 310

c C6 –C9 alkenes 559 C9 + aromatic hydrocarbons 638 C10 -cyclopentanes 533 C10 -naphthenes 533 C-centered orthorhombic structure 74 C–C bond formation 317, 321 C–O stretching band 214 Ca-exchanged aluminophosphates 400 camphene epoxide 797 cancrinite columns 34 CAN-type zeolite 399 carbine–carbenoid mechanism 692 carbocationicmechanism 693 carbon templating 259 catalyst effectiveness 367 catalytic dewaxing – deﬁnition 602 – diffusion-limited situations in 606 – double-stage conﬁguration 611f – of gasoil via selective cracking of n-parafﬁns 605–606 – single-stage conﬁguration 611f – via isomorphous substitution 607–609 – via shape selective cracking 605–609 catalytic reactions – acid–base reactions within zeolite-based materials 211–217

– Al–OH groups 217 – of Al–Si ordering 219 – application of bulky pyridine derivatives 215 – band positions of SSZ-33 zeolite 212–213 – conversion of methanol to hydrocarbons (MTH) 225–229 – in crystallization of zeolite X 220–221 – decomposition of nitric oxides 221–225 – effect of CO and collidine (2,4,6-trimethylpyridine) 216–217 – effect of water on nitrite formation 225 – ferrierite spectrum 217–218 – formation of the Al–O–P bonds 220 – hydroxyl groups 212–213 – in situ Raman spectra for synthesis gels 220, 221f – interaction of carbon monoxide 210 – interaction of NO with CuII centers 222 – IR microspectroscopic studies 231–232 – IR spectroscopy 210–211 – mesoporosity systems 216 – methanol-to-oleﬁn conversion 225–230 – origin of in situ spectroscopy in 210 – with probe molecules 214–215 – in SAPO-34 crystallization 219, 220f , 227–229 – of silicoaluminophosphates 218 – synthesis process of aluminophosphates (AlPOs) 219–220 – Topsoe integrated gasoline synthesis (TIGAS) 225 – during zeolite A synthesis 219 – zeolite synthesis processes 218–221 catalytic reforming – catalysts for 595–600 – chemistry 591–594 – process 587–591 cation-exchange properties, of zeolites 401 cavitation effect 250 cesium-exchanged zeolites X and Y 522 cesium-overexchanged MCM-41 815 CGF framework 78 CGS framework 78 chabazite zeolite 180–181 charge density mismatch (CDM) 73 CHARMM force ﬁeld 336 chemical potential of a molecule 3 Chevron-ICR 410 process 615–616 chiral zeolites 195–197 chiral zincophosphate (CZP) framework 43 choline chloride/urea-based DES ILs 96 chromium-containing silicalite-2 (CrS-2) 737 CI triples 303

Index Cis- and trans-geraniols 798 CIT-1 (CON) 68 citronitrile (5-pheyl-3-methyl 2-pentenenitrile) 811 Claisen–Schmidt condensation 817 Clausius–Clapeyron equation 340, 504 clear solution synthesis 12–13 closed-cage structures 21 closed hydrologic conditions 132 cloverite (-CLO) 60, 78 cluster 3 CoGaPhosphate-6 65 CO hydrogenation catalysts 663 CO2 /CH4 selectivities 123 CO2 emission 449 cobalt-containing AlPO4–5 (CoAPO-5) 219–220 collidine 216 colloid-sized gel particles 13 commercial adsorption separations 463 composite building units (CBUs) 132 concomitant nucleation 5 conﬁguration interaction doubles (CID) 303 conﬁguration interaction singles (CIS) 303 conﬁgurational diffusion 237 conﬁgurational-bias Monte Carlo (CBMC) technique 340–341, 349, 351 confocal ﬂuorescence microscopy 272 Conradson Carbon 558 consecutive-type mechanism 693 Co/OMS samples 666 coordination polymers 92 copper–cyclam complex, of LTA 114 copper-zeolite (Cu/zeolite) SCR catalysts 757–758 cotemplating, idea of 72–73 – for control of size and morphology 122 coupled cluster (CC) methods 303, 306 covalent linkages, between zeolite microcrystals 412–413 coverage, deﬁned 416 CrAPO-5 271, 376, 377f Cresex separation technologies 465 critical radius 5, 8 crown ethers 122 crystal growths 6 – c-direction 34 – energy levels 21, 22f – at equilibrium 21–22 – MOF 49 – Monte Carlo modeling of 21–22, 27 – for nanoporous materials 1 – on zeolite and zeotypes 14–15 – zeolite L 33–35

crystallization curve, for a zeolite synthesis 12–13 crystallization of frameworks 87 crystal program 305 crystal surface structure 6–8 Cs-Beta zeolite 814 CsLa-MCM-41 815–816 CsX zeolite 816, 818 Cu-based MSC 672 Cu3 (C9 H3 O6 )2 (H2 O)3 (HKUST-1) 47 CuHZSM-5 222 cumyl hydroperoxide (CHP) 795 CuZSM-5 222–223 Cu/ZSM-5 746 Cu/ZSM-11 746 Cu/ZSM-12 746 3-cyanopropyl groups 413 cycloaddition of cyclopentadiene and p-benzoquinone 787 Cymex separation technologies 465

d D-biotin-tethered zeolite microcrystals 435 2D-connected 10-ring channel system 194 2D crystal growth mechanism 49 2D EXSY spectrum 267–268 2D nucleation 11 2D nucleation energetics 8–10 2D nuclei 7 2D TEM images 262–263, 265f 3D TEM 261–262 D4R-containing structures 64, 73, 114–115, 117, 472 D111 crystal spacing 49 D222 crystal spacings 49 De novo strategy 72 dealumination, of zeolite 627–628 deep eutectic solvents (DESs) 88 degree of close packing (DCP) 412, 417 delaminated or delayered zeolites 668 DeNOx catalysis 221–222, 225 density functional theory plus damped dispersion (DFT+D) approach 314–315 DFT calculations, of zeolites 303–304, 307–308, 311, 316, 322 DFT-predicted adsorption energies 314 di-alkylated imidazolium cations 96 dialkylimidazolium salts 79 1,4-diazabicyclo [2.2.2]octane 137 dibenzyldimethylammonium (DBDMA) SDAs 70 Diels–Alder chemistry 69 Diels–Alder reaction 783–787

867

868

Index diesel fuel 601 – cold-ﬂow improvement 612f diethanolamine 114 N, N-diethyl 3,5-dimethyl piperidinium iodide (DEDMPI) 31, 33 diethylene glycol (DEG) 142 diffuse reﬂectance infrared Fourier transform (DRIFT) spectroscopy 211, 508 diffusion, in zeolite catalysts – acid-catalyzed oligomerization of furfuryl alcohol 378 – of alkanes in zeolite MFI 370 – basic concepts 362–363 – catalyst effectiveness factor, concept of 366–368 – concentration dependence 368–370 – corrected or Maxwell–Stefan (MS) diffusivity 364 – difference between gas phase and adsorbed phase 364 – diffusion anisotropy 363 – importance 361 – intermolecular repulsion effects 369 – linear hydrocarbons in zeolite MFI 370 – loading dependencies for self-diffusivities 368–369 – measured by macroscopic techniques 366 – measured by microscopic techniques 366 – measurement techniques 365–366 – in mixtures 365 – para-selectivity 379 – pore orientation 378 – of a porous material 363 – reactant pathway from bulk to active site 362f – saturation loading 369 – segregated adsorption 370 – self- or tracer diffusivity (Dself ) 364 – single-ﬁle diffusion phenomenon 370–372 – slurry phase conditions 450 – ‘‘strong conﬁnement’’ scenario 369 – at the subcrystal level 375–379 – surface barriers 372–374 – Thiele concept 374–375 – and transport of molecules 379–382 dimethyl formamide 93 2,3-dimethyl-2-butene (DMB) 325 N, N-dimethyl-3-azoniabicyclo 116 dimethylcyclopentenyl cation 317 2,6-dimethylpyridine spectrum 215

dimethyl sulfoxide (DMSO) 817 dinitrogen 214 direct synthesis, of zeolites 155–156, 157 discover (CFF) 336 dispersion, of zeolite microcrystals 417–418 disproportionation reactions on zeolites – ethylbenzene 630–633 – toulene 625–630 – and transalkylation of trimethylbenzene 633–635 2,6-di-tert-butylpyridine 216–217 dodecasil-3C (MTN structure type) 108 double oriented rotation (DOR) NMR 527 double 6-rings (D6Rs) 35 double six rings 90 Dreiding force ﬁeld 336 D6R orientation on STA-7 37–38, 41 dry gel conversion (DGC) syntheses 61–62 dusty gas model 363 DW-10 615

e

Ebex separation technologies 464 ECEPP/2 force ﬁeld 336 ECR-1 (EON) 70 ECR-34 472 ECR-34 (ETR) 65 ECR-40 59 ECR-42 78 edge-to-edge (ETE) contact mode 431, 433 edingtonite-type (EDI topology) zeolite 122 effective diffusivity (Deff ) 367 eight-ring windows 91 electrolyte membrane, in fuel cell applications – electrocatalysts 842–844 – hydrogen production and puriﬁcation 845–850 – hydrogen storage 850, 853–855 – inorganic membranes 841–842 – ionic conductivities 829–833 – PFSA membranes 828 – self-humidifying composite membranes 841 – zeolite/chitosan composite membranes 840–841 – zeolite/PFSA membranes composite membranes 839–840 – zeolite/polymer composite membranes 833–839 – zeolite/polytetraﬂuoroethylene (PTFE) composite membranes 839 – zeolites and molecular sieves 844–855 EMC-2 (EMT) zeolites 70 EMIM Br IL 95

Index EMIM Br solvents 97 EMIM NTf2 97, 98f EMT 122 EMT/FAU intergrowth zeolites 261 energy-dispersive X-ray mapping 291 Eni Carbon Silicate (ECS) 482, 483f environmental catalysis – catalysts 758 – cation introduction into the zeolite 747, 748f – Cu/zeolites 746 – decomposition of NO 746–747, 749f – dinitrosyl complexes 750–751 – emission controls 758–760 – ﬁelds of application 756–769 – isolated copper ions 751–755 – issues 746–756 epoxidation reactions 795–798 epoxide-coated zeolite (EP-Z) 427 6,7-epoxylinalool 798 ERI-type zeolites 342, 348 1-ethyl-3-methyl imidazolium bromide (EMIM Br) 90 1-ethyl-3-methylimidazolium (EMIM) cation 92 N-ethyl-hexamethylenetetrammonium bromide 123 ethylbenzene, dehydrogenation of 390 ethylbenzene disproportionation 630–633 ethylene glycol (EG) 78, 142 ethylene production, worldwide 659 P-ethyltoluene 640–641 EtPhSiNH2 626 ETS-10 31 EU-1 (EUO) 58, 67 EU-20b 74 EUO framework 70 eutectic mixtures 88–89 ex situ operation 2, 14, 27 – characterization 209 – zeolite A (LTA) 27 exchange–correlation functional 315 excited-state wavefunctions 303

f fabrication of nanoporous materials 143–146 Far infrared (FIR) 211 FAU-type zeolite 31, 57, 122, 398, 495 Fe-MCM-58 156 FER layers 74 FeZSM-5 223, 224f Fick’s law 363 Fickian diffusivity 364

Fickian-type diffusion 369 Fischer glycosidation reactions 789–791 Fischer–Tropsch synthesis (FTS) 651, 660, 663–670, 687 ﬂavonoids 816 ﬂuid catalytic cracking (FCC) 455–459, 493, 551–561, 660 – application of zeolite Beta 561 – catalysts 555–556 – coke combustion 554 – diesel fractions from 586 – elementary reactions in 556f – endothermic cracking reactions 554 – product slate 552f – residue cracking 558–559 – riser reactor temperature proﬁle 554, 554f – role of zeolite acidity 556–558 – shell resid 553f – unit operator 551–553 – use 549 – VGO cracking activity 558f – ZSM-5 application 559–560 ﬂuid-to-crystalline-like phase transitions 255 ﬂuoride 100–101 ﬂuoride, in SAPO synthesis 60–61 ﬂuoride mineralizers 87 force ﬁelds, for morphology predictions of zeolites 336–337 Fourier transform IR (FTIR) instruments 211 framework structure codes 155 free-energy landscape approach 604 free energy of a system 4 free radical mechanism 693 Friedel–Crafts acylation reactions 775–780 front factor 394 fructone (ethyl 3,3-ethylendioxybutyrate) 788 FTIR spectroscopy 376 FU-9 109 functional group-coated zeolite (FZ) 422 furanosides 790

g gallium 65 gallium phosphate system 60 gallosilicate TsG-1 (CGS) 65 GaN 10 GaPO4 structures 60 gasoils, cold-ﬂow properties of 601–602 gas-phase metal-carbonyl adsorption 661 gas sensors 392–394 gas-to-liquids (GTLs) processes 660 gauge-independent atomic orbital (GIAO) methods 327

869

870

Index Gaussian-type functions (GTO) 305 Gd-zeolite 400 Ge-containing materials 107 Gedanken experiment 71 General Utility Lattice Program (GULP) 70–71 germanium-free ITQ-24 64 germanium zeolites 63–65 germanosilicate ITQs 187–190 Gibbs–Thomson relation 252 d-glucose-tethered zeolite microcrystals 435 glyceraldehyde acetonide 816 Grand canonical Monte Carlo (GCMC) simulations 366 granular activated carbon (GAC) 467 ‘‘Green’’ chemistry 88 GROMOS force ﬁeld 336 ground-state wavefunctions 303 g-surface 196 guaiacol 782 guest molecules, modeling of 337–338

h 1

H NMR spectra of water 270 H-BEA 695 H-beta zeolite 652 H-bonding 414 3-halopropyl-tethering glass (XP-G) plate 429–430 3-halopropyl-tethering zeolite (XP-Z) crystals 430 Hammett indicators 500 Hartree-Fock (HF) theory 302–303 HC-80 615 HC-coated cubic zeolite microcrystals 431 heavy cycle oil (HCO) 456 Henry coefﬁcient 340 Henry coefﬁcients 348–349 heterogeneous nucleation 5 hexagonal crystal systems 1 hexamethonium (HM) 68 hexamethylene-1,6-bis(N-methyl-Npyrrolidinium) 69 H-ferrierite 309 hierarchical MFI-type zeolites 479 hierarchical zeolites 237, 473–479 high-resolution scanning electron microscopy (HRSEM) 14 high-resolution transmission electron microscopy (HRTEM) 14, 260–261 Hirschler indicators 500 HMVUSY (high-meso very ultra stable Y) 247, 250, 252, 254 Hofmann elimination reaction 503

host–guest interactions, in a zeolite 396–399 H2 O/EtOH separation factor 391 H2 PO4 groups 92 H-SAPO-5 502 H-SAPO-11 502 H-SSZ-13 218 H-USY zeolite 507 HUSY zeolites 794 hybrid density functionals 307 hybrid quantum chemical embedding schemes 307 hydrocarbon pool 317 hydrocarbon-pool mechanism 693 hydrocarbons 89 hydrocracking – catalyst systems, and catalytic chemistry 566–569, 575–576 – feedstocks and products 563–566 – of a gas oil/deasphalted oil blend 565 – heavy feeds conversions 576 – of heavy virgin gas oil 564 – hydrogenation and 568–569 – NiMo/zeolite catalyst 566, 567 – process 561–563 – zeolite Y in 570–574 hydrogen economy 403 hydrogen peroxide 795 hydrogen sorption 403–404 hydrophilic AlPO frameworks 120 hydrophobic pure-silica zeolites, synthesis of – crystalline 163 – effects of type of acid, pH, temperature, and other factors 163–164 – experimental procedures 162–163 – SSZ-33, physicochemical and catalytic properties 164 hydrothermal synthesis 89–90, 96 hydrothermal synthesis of zeolites 94–95 hydrotreating 549, 614 hydroxide concentration, importance of 59–60 hydroxyl groups, in zeolite structures 213 hydroxyalkylation, of aromatics 780–783 2-hydroxymethyl-1-benzyl-1-methylpyrrolidinium (bmpm) 110, 111f 3-hydroxy-2-methylpentenal 818 hypothetical structures 71 hysteresis loops 247 H-ZSM5 628 HZSM-5 226, 311, 319, 312f , 320f , 632, 656, 661, 668, 673, 817 H-ZSM-5 375, 379, 496, 501–503, 631, 641, 653, 692, 695

Index

i I-C4/n-C4 ratio 604 I molecule, in the crystal 4 ideal adsorbed solution theory (IAST) 365 IFR-type zeolites 156 IM-5 193–195, 624 IM-16 (UOS) 69 induction time 6 industrial zeolite catalysts – anisole acylation 454–455 – conversion of bottom Fluid Catalytic Cracking (FCC) 457–459 – groundwater remediation 467–471 – hierarchical zeolites 473–479 – liquid phase oxidation processes 451–453 – parafﬁn alkylation 454 – preparation of silica-based crystalline organic-inorganic hybrid materials 479–483 – in reﬁneries 455–462 – in separation technology 462–467 in situ AFM 2, 14 in situ conditions, zeolite A (LTA) 27 in situ ﬂuorescence microscopy 271–273 in situ growth experiments – MOF 47 – ZnPO-FAU 47 – ZnPO-SOD 43–47 in situ measurements 13, 27 in situ M¨ossbauer spectroscopy 224 in situ spectroscopic techniques 210 – in slurry processes 450–455 – synthesis of materials with novel framework topologies 471–473 inorganic–organic hybrid materials 88–89 inorganic zeolites 71–72 intercrystalline pores 251 interference microscopy(IM) 373 interlaced spirals 10 interlacing-inducing growth anisotropy 43 intermediate Al-species 160 International Zeolite Association (IZA) 91 intracrystalline diffusivity 373 ionic linkages, between zeolite microcrystals 413–414 ionic liquids (ILs), 78–79, 87–89, See also ionothermal synthesis – unstable 101

ionothermal synthesis – AlPOs 92 – aluminophosphate 90–92 – ambient pressure 93–95 – anion, role of 97–98, 99f – under autogenous pressure 93–94 – from butyl methyl imidazolium (BMIM) cation 98, 99f – in coatings technology 94 – of metal organic frameworks and coordination polymers 92–93 – silicon-based zeolites 92 – templating processes 95–97 – water, role of 99–101 IR microspectroscopy 231–232 IR spectra of zeolites beta 503, 504f iron 559 iron zeolite (Fe/zeolite) SCR catalysts 758 isocontour pictures, of zeolites 343f isomorphous substitution, of aluminum 65–67 isopropylamine 503 ITQ-2 668 ITQ-4 156 ITQ-6 76 ITQ-7 63, 561 ITQ-12 63, 69 ITQ-13 63 ITQ-18 76 ITQ-20 76 ITQ-21 481, 561 ITQ-22 668 ITQ-33 472, 561 ITQ-37 196

j jasminaldehyde JIS-1 116

818

k K–Al–Si system 71 Keggin-like chlorohydrol cation 74 Kelvin equation 249 Kiselev-type model 337 Kiselev-type potentials 337 KJS (Kruk-Jaroniec) method 251 Knoevenagel condensation reaction 391 Knoevenagel reaction 809–813 Knudsen diffusion 237, 365 Knudsen diffusion mechanism 363 Kohn–Sham orbitals 304 K2 O/La2 O3 -MCM-41 818 Kossel model, of a crystal surface 7f kryptoﬁx-n macrocycles 114

871

872

Index

l Langmuir isotherm 364, 369 lanthanum-based MOFs 96 large zeolite single crystals 133f – Al-MFI zeolite (ZSM-5) 136 – benzene-1,2-diol, role of 135 – ﬂuoride anion, role of 134–135 – LTA and FAU zeolites 134, 135f – methods of synthesis 133–135 – natural 132–133, 134t – rings and cages 132f – of Si-MFI (all-silica MFI) zeolite 135–137, 136f , 140f – synthesis of 133–138 laser-hyperpolarized (HP) 129 Xe NMR 2D exchange spectroscopy (EXSY) 267 layer growth 7, 8f layered precursors, of zeolites 73–76, 77t layered silicate precursor zeolites 198–199 Lennard–Jones potential 314, 337–338 Lewis acid sites 216 Li-exchanged X zeolite 813 liftpot 554 light cycle oil (LCO) 456 limonene oxide 796–797, 797f Linde type A (LTA) 57, 340, 363, 393 lithium 66 lithosilicate structure 67 local density approximation (LDA) 304 long-chain alkyl glycosides 791 ludox AS-40 silica 639 lutidine 216 Lyondell direct-oxidation propylene oxide technology 453

m macro-defects 34 macroscopic crystal 13 magnetic resonance imaging (MRI) 400 magnetic zeolites 402 mass transfer coefﬁcient 373 matrix rigidiﬁcation 395 Maxwell’s equation, to model gas permeability across MMMs 395 Maxwell–Stefan (MS) diffusivity 364, 395 Maxwell–Stefan (MS) equation 363 MAZ zeolite 70 MCM-22 75–76, 507, 561, 608, 624, 626–628, 631, 636, 637, 639, 656, 668 MCM-40 631 MCM-41 252, 254, 501, 522–523, 653, 665, 791, 793, 796, 808, 817 MCM-47 74, 76, 110 MCM-48 665, 808

MCM-49 656 MCM-56 76, 656 MCM-58 624 MCM-65 74, 111 MCM-68 624 MCM-68 (MSE) 61, 68 MDDW process 613–615 mean Sanderson electronegativity 495, 509 medical applications, of zeolite 399–400 medium/mid infrared (MIR) 211 medium pore zeolites 161–162 Meerwein–Ponndorf–Verley (MPV) reduction and Oppenauer’s oxidation (O) of alcohols 803–808 MEI framework 59 melonal (2,6-dimethyl-5-hepten-1-al) 802 MeNH2 SiOH 626 MePhSiNH2 626 MePhSiOH 626 mercury porosimetry (MP) 255–256 mesoporous Al-MCM-41 788 mesoporous MCM-41 molecular sieves 786 mesoporous silica MCM-41 270 mesoporous Y zeolites 255–256 mesoporous zeolites – crystals 380 – electron microscopic (EM) characterizations of 256–266 – gas physisorption of 246–251 – In situ ﬂuorescence microscopy characterizations of 271–273 – mercury porosimetry (MP) for characterizations of 255–256 – NMR characterizations of 266–270 – with dealumination process 239–241 – with desilication process 241 – with detitanation process 242 – by stacking nanosized zeolites as intercrystalline pores 245 – with templating method 243–245 – textual characterization 246–273 – thermoporometry for characterizations of 251–255 metal clusters, in zeolites 654 – characterization of 532–535 – formation of 530–532 – nature of 529–530 metal deposition, for zeolite modiﬁcation 628 metal dispersion D 534 metal-organic frameworks (MOFs) 1, 47–49, 79, 96–97 metal-overexchanged MCM-41 814 metastable zone 5

Index methane activation – alkylation of aromatics with methane 653 – autothermal reforming 651 – CH4 conversion 654–655 – dimethyl ether (DME)-to-oleﬁns(DMO) process 660 – direct oxidative conversion routes 649–650 – Fischer–Tropsch synthesis (FTS) 651, 663–670 – high-pressure acid-catalyzed oxidative alkylation of toluene with methane 653 – methane dehydroaromatization (MDA) 649 – methane partial oxidation (MPO) 650 – methanol carbonylation using homogeneous Rh or Ir complexes 676–678 – methanol-to-gasoline (MTG) reactions 673 – methanol-to-oleﬁns (MTO) process 660 – NOCM reaction 654 – nonoxidative coupling of methane (NOCM) 649 – nonoxidative methane dehydroaromatization (MDA) 655–659 – nonoxidative methane homologation and alkylation processes 654–655 – oxidative alkylation of benzene with methane 654 – oxidative conversion method 651–654 – oxidative coupling of methane (OCM) 650 – oxidative self-coupling 652 – oxygenates, synthesis of 670–678 – routes 649–651 – selective synthesis of short-chain (C2–C4) oleﬁns 659–663 – steam-reforming process 650–651 – syngas conversion processes 659–678 – syngas-to-dimethyl ether (STD) 670–671 – synthesis of higher (C2+ ) oxygenates 674–676 methane-derived aromatics 653 methanol-to-gasoline (MTG) process 703f – catalysts and reaction conditions 702 – commercial 688 – deactivation mechanism 702–703 – development 689 – mechanisms and kinetics 690–697 – technology 703 methanol-to-oleﬁn (MTO) process 312, 316–321

methanol-to-oleﬁns (MTO) process – catalysts and reaction conditions 697 – commercial aspects/economic impact 700–701 – deactivation mechanism 698–699 – development 688 – future prospects 701–702 – mechanisms and kinetics 690–697 – technology and process design 699–700 – UOP/Norsk Hydro MTO process 700–701 methanol-to-propene (MTP) process 703–705 1-methyl-3-ethyl imidazolium templates 91 1-methyl-3-ethylimidazolium bromide (EMIMBr) 116 m-methyl tropane 69 N-methyl homopipendine 69 methylimidazolium (MIA) 95 2-methylpentenal 817–818 methylpyridines 215 methyl-tert-butyl ether (MTBE) 390, 467 methyltrioxorhenium (MTO) 797 MFI/MEL intergrowths 28–33 MFI zeolite particles, crystal morphology of – of AL-MFI 141 – structure-directing agent (SDA) for 139–141 MgO modiﬁcation 631 Michael addition 813–816 microporous materials, hydrothermal synthesis of 119 microreactors 391, 392 microwave synthesis, of zeolites – of Al- and Ti-bimetal incorporated MFI zeolite ((Al, Ti)-MFI) 146–147 – of bimetal-incorporated systems 147 – examples of dependency 142 – formation of the stacked morphology 146–149 – of metal (Ti, Sn) incorporated systems 144 – morphological fabrication 143–146 – Si-MFI crystals 142, 143f, 147, 148f – TEM image and ED patterns 145 – of Ti-MFI zeolite 144–146, 147f MILOS process 560 mineral zeolite structures 184–185 mineralizers 59–61 mixed-matrix membranes (MMMs) 394–396 mixing time 267 mixture diffusion 365 Mn–Fe bimetallic catalyst 662 MMFF force ﬁeld 336 MM2 force ﬁeld 336 MM3 force ﬁeld 336

873

874

Index MM4 force ﬁeld 336 Mo/HZSM-5 catalysts 655–658 Mo/ZSM-5 catalyst 659 Mobil (now ExxonMobil) 609 Mobil isomerization dewaxing (MIDW) process 610, 614–615 Mobil’s oleﬁn-to-gasoline and distillate process (MOGD) 688 – catalyst and process operation 706 – technical aspects 707 – thermodynamic considerations 706–707 modeling zeolites 306–307 modiﬁed constraint index (CI) 533 modiﬁed FTS 668–670 molar ratio, in synthesis mixture 157 molecular crystals 10 molecular dynamics (MDs) simulations techniques 340, 366, 749 molecular Gate technology 466 molecular linkages 413f molecular recognition 322 molecular simulations in zeolites – adsorption 339–341 – benzene, adsorption of 350–351 – carbon dioxide adsorption 351–352 – of conﬁned water 346–348 – diffusion 344–346 – enthalpic effects 350 – force ﬁelds for morphology predictions 336–337 – free energy barriers 341–342 – free energy proﬁles of propane and propylene 349, 350f – green chemistry, applications in 351–353 – guest molecules 337–338 – of hydrocarbons 348–349 – isosteric heats of adsorption 340 – liquid-phase adsorption isotherms of water 339f – natural gas composition 352–353 – nonframework protons or cations 336–337 – pore volumes obtained from 345 – pure silica zeolites 346 – self-diffusivities 344 – of separation of mixtures 349–351 – shape-selective catalysis 348 – with a Si/Al ratio 336 – of sodium cations in MFI- and MOR-type structures 349 – technological processes, applications in 346–351

– vapor–liquid equilibrium curves 337, 338f – volume-rendered pictures, zeolite surface areas, and zeolite pore volumes 343–344 – of water adsorption in MFI 347f Molex separation technologies 464 Moller–Plesset perturbation theory (MP2) 303 Mo/MCM-22 catalyst 656 monoalkylated imidazole species 96 monolayer assembly of zeolite microcrystals, See Zeolite microcrystals, organization of Monte Carlo modeling, of crystal growth 21–22, 23f Monte Carlo simulated annealing (MCSA) 42 Monte Carlo simulations 344, 346 – in grand-canonical ensemble (GCMC) 339–340 mordenite 179, 626, 631, 634 morphology of zeolites – of large single crystals 132–138 – of MFI zeolite particles 138–141 – by MW 142–149 MP2:DFT hybrid method 309, 311, 313–315 MP2 method 308, 311 MP2 theory 308 MPW1PW91 functional 304 MQD Uninonﬁning process 615–617f M41S family 478 multinucleation multilayer growth 8 multiple-quantum magic-angle spinning (MQMAS) NMR spectroscopy 527 MWW framework 74–76, 112 m-xylene 625–626

n N2 adsorption–desorption isotherms 452 Na-exchanged MCM-41 817 nanocrystalline zeolites 454 nanoporous MOF crystals 47 nanozeolites 475 natural zeolites 468 NaX zeolite 398 NaY zeolite 398 Nb-MCM-41 811 near room temperature ionic liquids (nRTILs) 88 needle-shaped crystals 1 Nelson Complexity Index (NCI) 455 next nearest neighbors (NNNs) 495 N-glide plane 41 NH4 -ZSM-5 655 nickel 559 nitrex (UOP) 466

Index nitrogen physisorption, of zeolite NaY 247–251, 248f NiZSM-5 zeolites 225 NLDFT (Non-Local Density Functional Theorie) 251 NMR-invisible 284 N-octane hydrocracking 166 NO2 /NO3 species 223 N-particle system 304 NSI framework 74–75 NU-87 633 nuclear magnetic resonance (NMR) 3, 14 – characterizations, of mesoporous zeolites 266–270 nucleation 3 – rate 5 – zeolite 13–14 nucleus size 5 Nu-6(2)-polysulfone hybrid membranes 396 NU-87 (NES) 58 NVT ensemble 340

o octadecasil (AST) 63 O-ﬂuorobenzyl-benzyl-dimethylammonium 117, 118f O–H stretching band 214 o-xylene 625 Olex separation technologies 465 oligomerization 3 open-cage structures 21 OPLS force ﬁeld 336 ordered mesoporous materials (OMMs) 379–380 organosilicon compounds 626 oxonium ylide mechanism 691

p p tetrahedra 46 parafﬁns 601 Parex separation technologies 464 PBE functionals 304 PBE0 functional 304 pentamethylbenzenium cation 317 pentasil chain 31 perforated zeolite microballs 434–435 permeable reactive barriers (PRBs) 467–468, 470 petroleum derived products 601 phenylaminopropyl-trimethoxysilane (PHAPTMS) 476 phenylethyltrimethoxysilane (PETMS) 480 PHI/MER products 78 α-pinene oxide 793–794

polybed PSA System 465 polyelectrolytes 414 polyhedral morphologies 11 polymeric amines 412 polymeric membranes 394–395 polymer–zeolite adhesion 395 polystyrene 414 pore-ﬁlling agents 116–117 post-synthetic treatment and modiﬁcations – aluminium reinsertion using aqueous Al(NO3)3 solution 158–162 – of hydrophobic compounds 162–166 – into lattice sites 157 precoking, of zeolite-based catalyst 627 PREFER layer 74, 76 PREITQ-19 zeolite 76 pressure swing adsorption (PSA) processes 400, 465 primary amorphous phase 12 primary nucleation 3 prismatic Si-MFI zeolite crystals 137 product shape selectivity (RSS) 603 projection–slice theorem 262 propylamine 110 PtNaY-covered capacitor 393 pulsed ﬁeld gradient nuclear magnetic resonance (PFG-NMR) 366 pulsed ﬁeld gradient (PFG) NMR 269–270 PW91 functionals 304 P-xylene 626–628, 637, 639–640, 642 pyridine 89, 110, 124, 214–215, 509, 214f

q QM:force ﬁeld (QM:MM) approach 315–316 quantum chemical methods (QM) 307 quartz crystal microbalances (QCMs) 392–393 quasi elastic neutron scattering (QENS) 366 quasi-steady-state distribution, of molecular clusters 6 Questair technology 466 quinuclidine 69, 110

r 3R-containing zeolites 67 racemic separations, using zeolites 401–402 Rb-containing zeolites Y and X 322 Rb/Nb-MCM41 catalysts 811 reactant shape selectivity (RSS) 603 reaction engineering 450 reconstruction 262 reﬁneries, use of zeolites – adsorption of nitrogen from natural gas 466

875

876

Index self- or tracer diffusivity (Dself ) 364 SEM–EDX (energy dispersive X-ray spectroscopy) analysis 257 shape selectivity behavior, of zeolitic materials 602–609, 624 – commercial applications 609–618 SHARP force ﬁeld 336 sheet silicates 10 ship-in-a-bottle complexes 713–714 ship-in-the-bottle method 398 Si/Al ratio 160, 178–179, 183, 219, 228, 257, 263, 267, 322, 336, 352, 463, 471, 627, 658, 781–782, 792 29 Si MAS NMR 61–62, 201 Si(OH)Al bridged group 212 SiOH-SSZ-33 164–166 Si-SSZ-33 164–166 Si-SSZ-42 156 Si/Ti ratio 451 Si/X ratio 156 sieve-in-a-cage morphology 395 silanol groups 212–213 silica–alumina gels 78 silica-based crystalline organic–inorganic hybrid materials 479–483 silicalite 28–33, 60 silicalite-1 661 siliceous zeolites 60, 62 silicoaluminophosphates (SAPOs) 58–59, 90, 523, 660 – SAPO-5 124, 142, 271, 376, 377f – SAPO-11 124, 608, 610, 639 – SAPO-18 697 – SAPO-20 61 – SAPO-34 123–124, 218, 271, 376, 377f , 639, 689, 694–695, 697–699 – SAPO-37 (FAU) 72 – SAPO-41 125 – SAPO STA-7 structure 35–36 s silicon, in AIPO structure 59 SAPO-34-based MTO process 689 silicon carbide 10 SBA-15 665–667, 789, 795, 808 scanning electron microscopy (SEM), analysis single-crystal nanoporous ﬁlms 1 single four rings 90 of mesoporous crystals 256–266 single nucleation growth 7 B-scission mechanism 608–609, 610 – SIZ-1 90, 96, 101 scolecite structure 294, 295f – SIZ-2 101 SCR catalysts 759 – SIZ-3 101 SDD-800 612 – SIZ-4 96, 101 SDD-801 612–613 – SIZ-6 92 SDD-821 613 – SIZ-7 79 secondary amorphous phase 12 – SIZ-13 100 secondary building units (SBUs) 63 SIZ-n(St. Andrews Ionothermal Zeolite) secondary nucleation 3, 5 materials 90, 91f seed crystals 13, 29 reﬁneries, use of zeolites (contd.) – adsorption on zeolites 464 – alkylbenzenes, hydrogenation of 460f – in ﬂuid catalytic cracking (FCC) 455–459 – global oil demands and distribution 550f – hydrocracking and cracking capacity in oil 547–551 – LCO upgrading processes 456–457 – natural gas, separation of 465–466 – oleﬁn oligomerization 461–462 – permeable reactive barriers (PRBs) 467–468, 470 – petroleum products 456f – production of diesel fuel 461 – schematization of 548f – sulfur speciﬁcations for automotive fuels 551f – synthesis of a crystalline Y zeolite 458–459 – thermofor catalytic cracking (TCC) 458 – trends 549–550 – worldwide reﬁning capacities 585, 587 – ZSM-5 zeolite-based technology 462 relaxation time (tr ) 12 replacement, of surface-attached zeolite microcrystals 422–423 resolution of identity (RI) approximation 303 room temperature ionic liquids (RTILs) 88 RRO framework 75 RUB-39 75 RWR framework 75

Index slurry phase reaction, use of zeolites in 450–455 Sn-Beta zeolites 802, 806, 807 Sn-MCM-41 802, 807 Sn-MCM-48 807 SnO2 sensing surface 393 Sn-SBA-15 807 sodium aluminate 158 sodium butyrate (Na+ Bu− ) 413 sodium dodecylsulfate (SDS) 434 sol–gel process 396, 398, 452 solid-state reactions 89 solution–gel interface 13 solvothermal synthetic methods 89–90 sonication 415–416, 424 Sorbead Quick-Cycle Process 465 Sorbex separation technologies 464 sorption processes 403 spaciousness index SI 533 spiral dislocations 9–10 spiral growth, formation and development of 9–10 SSZ-13 697–698 SSZ-23 (SST) 62 SSZ-25 113, 116 SSZ-33 156, 642 SSZ-33 zeolite 212 SSZ-35 642 SSZ-47 73, 116 SSZ-53 380 SSZ-74 193, 195, 198, 634, 199f SSZ-75 634 STA-7 43, 72 – AFM image of a face 38, 39f , 40–42 – orientation of the D6R 36–38, 41 – SAPO structure 35–37 – scanning electron micrograph of 37f STA-14 (KFI) 72 steady-state isotopic transient kinetic analysis (SSITKA) 664 steam-assisted conversion (SAC) method 61 steamed zeolite Y (USY) 285–287 stereo-TEM 262 sticking coefﬁcients 373 structural chemistry – Al–O–Al linkages 173 – arrangement of sodalite cages 177–178 – based on the sodalite cage 172–185 – bonding to ﬂuoride during synthesis 191–192 – α-cages 181–182 – β-cages 177–178 – chirality and mesoporosity 195–197

– coordination sphere of framework atoms 191–192 – crankshaft chains 176f – 3D 12-ring channel systems 187–188, 189 – effect of variation of Si/Al and Si/B ratios in the gel 186–187 – F− ion balances 192 – FAU structure type 178 – faujasitic zeolites X and Y 178–179 – formation of D4Rs 173, 176, 178, 187, 195–196 – formation of D6Rs 176, 178, 180–181 – gmelinite cages 182, 183f – inorganic cation-only zeolites pre-1990 179–182 – ITQ zeolites with ﬂuoride ions 186 – K+ cations 180 – limits of structural complexity 193–195 – L¨owenstein’s rule 173 – natural mineral sodalites 184–185 – ordered vacancies and growth defects 197–198 – pore geometries 187–191f – pores 174–175 – rings 174, 179, 182 – secondary building units (SBUs) 175–176 – substituting framework oxygen 199–201 – tetrahedral positions 173–174 – TOT bond angles 173 – using TPA+ cations as SDAs 183 – of zeolites from layered silicate precursors 198–199 Structure Commission of the International Zeolite Association 155–156, 174 structure-directing agents (SDAs) 58, 60, 67–70, 472 – AFI structure 119 – degradable 117 – F–F repulsion 118 – modeling 70–72 – ortho-ﬂuorinated organic 119f SU-15 195 SU-32 195f s¨ud FChemie-hydex-g process 617–618 supersaturation 3–5 – intergrowths in zeolites 31 – low 11 – mechanisms of growth as a function of 11f – Monte Carlo treatment, impact 21 – silicalite, effect on 29–31 surface attachment, of zeolite microcrystals 418

877

878

Index surface-controlled mechanism 14 surface free energy 4 surface migration, of zeolite microcrystals 418–420 surface processes 6 surfactant modiﬁed zeolite (SMZ) 468–469 SUZ-4 zeolite 217–218 syn technologies 612 syngas-to-dimethyl ether (STD) 670–671 synthetic zeolites 58, 547 – AIPOs and SAPOs 58–59 – cotemplating, idea of 72–73 – dry gel conversion (DGC) syntheses 61–62 – germanium zeolites 63–65 – isomorphous substitution, of aluminum 65–67 – layered precursors 73–76, 77 – low water syntheses 62–63 – mineralizers 59–61 – silica frameworks 63 – structure-directing agents (SDA) 67–72 – using nonaqueous solvents 77–79

– crystallization of silicoaluminophosphate (SAPO)-37 113–114 – crystallization of SSZ-25 113 – of dual-void structures 108–113 – of FER and MWW-type materials 109, 112 – membrane systems 123 – of pure silica zeolite ITQ-1 using N, N, Ntrimethyl-1-adamantanammonium (TMAda+) 112 – for tailoring the catalytic activity of microporous materials 123–125 – use for pore-ﬁlling and basicity capacities 116–117 – using tetralkylammonium cations 113 – variants of the LTA structure 114–115 – XRD techniques 108–110 terminal OH groups 495 terpenic resin 395 terraces, on the crystal surface 6, 15, 33 tert-butyl hydroperoxide (TBHP) 737, 795 tertiary alkanolamine 134 tetra butylammonium (TBA) cations 31 tetraalkylammonium cations 69 tetraalkylammonium hydroxide 452 t tetraethylammonium (TEA) 110 4T cluster model 312 tetraethylene glycol (tEG) 142 T-plot analysis 248–249 tetraethylortosilicate (TEOS) 76, 481 T-site vacancies 197–198 tetrahedrons, See basic building units (BBUs) tablet-shaped crystals 1 tetramethylammonium (TMA) cations 137 TAPSO- 5, 732 1,2,3,5-tetramethylbenzene 226 tatoraySM process 633 1,2,4,5-tetramethylbenzene 226 TEAOH 123–124 tetramethylene-1,4-bis-(N-methylpyrroliditechniques, on solid crystal nium) di-quarternary cations 634 – AFM 15–16 tetrapropyl ammonium cation 29 – confocal microscopy 16–17 thermal swing adsorption (TSA) 465 – HRSEM 16 thermodynamic correction factor 364 techniques, on solutions thermofor catalytic cracking (TCC) 458 – cryo-TEM 20 thermoporometry 251–255 – mass spectrometry 19 Thia–Michael addition of thiols 813, 814f – nuclear magnetic resonance 17–18, 19f Thiele concept 362, 374–375 teﬂon-lined autoclaves 89 Thiele modulus 367, 368f – steel 93 thymine-tethering zeolite crystals 414 TEM analysis of mesoporous crystals Tian–Calvet type calorimeter 504 256–266 Ti-Beta 722–724 temperature-programmed desorption (TPD), – solvent effects 732–736 of basic molecules 501–504 Ti-Beta zeolites 804–805 templates 90 Ti-coated MCF materials 381 – carbon 259 Ti-ITQ-7 732 templating-based synthesis 95 Ti-MCM-41 797 – addition of TEA cations 115 Ti-MWW 724–731 – of aluminophosphate structures 113–116 – solvent effects 732–736 – control of crystal size and morphology 122 TIP4P model, for water 338 – cooperative structure-directing effects titanium-silicalite-1 (TS-1) 451–453, 117–121 715–722, 795

Index – catalytic chemistry 716 – crystallization process of 717 – route to the synthesis of 718 – solvent effects 732–736 – in the YNU system 718 Ti-UTD-1 731 Ti-YNU-2 732 Ti–zeolites 454t Ti-ZSM-12 731 Ti-ZSM-48 732 TMA + bmp mixed-template system 123–124 TMA + trimethylamine (Si/Al = 10.5) the C/N ratio 109 TNU-9 60, 193–195, 624, 627 toluene alkylation 636–638 toluene disproportionation 625–630 Topsoe integrated gasoline synthesis (TIGAS) process 705 TPA/TBA preparations 31 ‘‘Trafﬁc junction’’ effects 350 transalkylation processes 633–635 transition state shape selectivity (RSS) 603–604 transmission electron microscopy (TEM) 1 transport process 6 trial wavefunction 303 triethylene glycol (TEG) 142 triﬂimide NTf2 97, 98f 1,3,5-triisopropylbenzene 269 N, N, N-trimethyladamantyl ammonium 116 N, N, N-trimethyl-8-tricyclo-[5.2.1.02.6 ]decane ammonium 156 trimethyladamantammonium SDA 62 1,2,4-trimethylbenzene 633 trimethylbenzene disproportionation 634 trimethylpropyl ammonium iodide 413 triple-point temperature 251 tripos force ﬁeld 336 Type I isotherms 247 Type IV isotherms 247

u ultralarge-pore zeolites 380 ultrastabilization 178 ultrastable Y (USY) zeolite 237, 269–270, 555, 626, 633, 788 unit-cell height steps 10 universal force ﬁeld (UFF) 336 UOP catalytic dewaxing process 615 UOP/Norsk Hydro MTO process 700–701 UTD-1 731

UZM-5 (UFI) 73 UZM series of materials

110

v vacuum gas oil (VGO) 561 VALBON force ﬁeld 336 van der Waals density functional (vdW-DF) 316 van der Waals interactions 14 – dispersive 314 – of guest molecules 337–338 van der Waals spheres 72 vanadium 559, 736 vapor–liquid equilibrium curves 337, 338f vapor phase transport (VPT) method 61 veterinary applications, of zeolite 400–401 volcanogenic sedimentary rocks, crystals in 132 VS-2–H2 O2 –hexane system 736

w Wacker-type reactions 713 Washburn equation 255 water deactivation, with ILs 99–101 wide-pore zeolite 34

x X-ray diffraction (XRD) studies 108–109 X-ray powder diffractometry 133 X-ray reﬂectivity 412 129 Xe NMR spectroscopy 266–268 Xe–Xe collisions 267 XVUSY (eXtra very ultrastable Y) 247, 252, 254, 263 Xylene-plusSM process 633 xylenes 625

y YNU-2 62

z ZEBEDDE algorithm 42, 72 zeolite A (LTA) 62, 65 – AFM studies 23–25, 26f , 28f – aluminosilicate version 115 – crystallization process 24–25 – diethanolamine (DEA), effects 24, 26 – HRSEM resolution 24, 25f – Petranovskii synthesis 26–29 – Thompson synthesis 24–26 – SOD units 23, 27 – STA-7 possesses 115 – templates 114–115

879

880

Index zeolite A (LTA) (contd.) – under ex situ conditions 27 – under in situ conditions 27 zeolite-bound aminopropyl groups 414 zeolite ﬁlms 391 zeolite frameworks 65–67 zeolite–guest composites 398 zeolite H-beta 507 zeolite L (LTL) 33–35, 180, 395–396, 427, 429f zeolite membranes – coatings 390 – gas sensors 392–394 – MFI-type 389 – mixed-matrix membranes (MMMs) 394–396 – orientation of 390 – reactors and microreactors 390–391, 392 zeolite merlinoite (MER) 115 zeolite microcrystals, organization of – of 2D and 3D arrays of silicalite-1 crystals 437–438 – a and b orientations 427 – applications 441 – coverage and coverage-t plot 416–429 – degree of close packing (DCP) 417 – degrees of uniform orientation and close packing 427–428 – dispersion of 417–418 – by dry manual assembly 438–440 – factors that affect binding strengths 426–427 – in situ self-organization method 437 – into functional materials by self-assembly 411–412 – linkages, types of 412–414 – morphology of the ﬁbrils 435–437 – multilayer assembly on substrates 431 – patterned substrates 429–431 – prepared on glass plates 419, 422–423, 432–433f – replacement of surface-attached 422–423 – residual amounts (%R) 417, 420 – self-assembly of substrate-tethering zeolite crystals with proteins 435–437 – silicalite-1 416f – sonication impact 415–416 – substrates, types of 415 – surface-aligned zeolite microballs 434–435 – surface attachment 418 – surface migration 418–420 – surface-tethered hydrophobic hydrocarbon (HC) chains 431–434 – Teﬂon support for 415–416

– – – – –

thymine-tethering 414 TMPA+ I – tethering 414 use of polymeric linkages 426 using RO and RS methods 415, 425 using SO and SWS methods 415–416, 419, 424 – wet self-assembly 438 zeolite microcrystals, use of 411 zeolite N-A (Si/Al = 1.2) 58 zeolite Nu-1 66 zeolite nucleation 13–14 zeolite P 122 zeolite reactivity, theoretical concepts – Al distribution and structure and charge compensation of extra-framework cations 326–330 – Brønsted acid-catalyzed transformations of hydrocarbons 308 – cation-exchanged Al-rich zeolites 322 – conﬁnement-induced reactivity 325 – dehydrogenation catalytic activity of ZSM-5 328 – embedding schemes 308 – Ga-exchanged zeolites 328–329 – methanol-to-oleﬁn (MTO) process 312, 316–321 – methodological aspects of quantum chemical modeling of zeolites 302–307 – molecular recognition and conﬁnement-driven reactivity 321–326 – N2 O4 adsorption complex on alkali-exchanged faujasite 322–325 – nonlocalized charge compensation, concept 329 – photochemical activation of 2,3-dimethyl-2-butene (DMB) and O2 325, 326f – protonation of isobutene 308–309 – quantum chemical calculations of zeolite-catalyzed transformations of hydrocarbons 307–316 – reaction kinetics 318 – relative stabilities of different reaction intermediates 321 zeolite Y 213f , 252, 570–574, 604, 624, 627, 631, 634 zeolite/zeotype synthesis 58 – AFM studies 14–15 – channels 624 – crystal growth 14–15 – formation of 623 – parameters 11 – rate of nucleation 12–13 – schematic representation of 12f

Index – steps in 11–12 – structural motifs of 624f – supersaturation impact 13 – UTD-1 70 – zeolite Rho 179, 181–182 – zeolite X 323 zeozyme 396 zinc 66 zincophosphate 43–47 ZK-5 179, 181 ZnPO4 -Faujasite 47 ZnPO4 -Sodalite 43–47 ZnPO-SOD crystal 43 Zn-SSZ-42 156 Zn tetrahedra 46 Zr-Beta zeolites 806–807 ZSM-2 218

ZSM-4 182 ZSM-5 58–59, 136, 183, 193–194, 201, 215–217, 222, 226, 228–229, 231–232, 255, 257–259, 264, 266–268, 271–272, 289–291, 295, 373, 376, 379, 381–382, 396, 399, 428–429, 461–462, 470, 476–479, 481, 525, 604–605, 607, 610, 615, 624, 626–628, 631, 633, 635–637, 640–642, 653, 656, 662, 688–689, 696–697, 702–703, 746, 793 ZSM-11 59, 183–184, 477, 604, 746 ZSM-12 58, 60, 182, 477, 636, 746 ZSM-18 71 ZSM-22 58, 604 ZSM-23 58 ZSM-39 108, 110 ZSM-48 58, 67, 108 ZSM-57 58

881