Polymer Brushes: Synthesis, Characterization, Applications

Polymer Brushes Synthesis, Characterization, Applications Edited by Rigoberto C. Advincula, William J. Brittain, Kenne...

Author: Rigoberto C. Advincula | William J. Brittain | Kenneth C. Caster | Jurgen Ruhe

189 downloads 1912 Views 16MB 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

Polymer Brushes Synthesis, Characterization, Applications

Edited by Rigoberto C. Advincula, William J. Brittain, Kenneth C. Caster, Jrgen Rhe

Polymer Brushes Edited by Rigoberto C. Advincula, William J. Brittain, Kenneth C. Caster, Jrgen Rhe

Related Titles: P. Ayajan, L. S. Schadler, P. V. Braun

Nanocomposite Science and Technology 2003 ISBN: 3-527-30359-6

G. Decher, J. B. Schlenoff (Eds.)

Multilayer Thin Filmes Sequential Assembly of Nanocomposite Materials 2003 ISBN: 3-527-30440-1

P. Go´mez-Romero, C. Sanchez (Eds.)

Functional Hybrid Materials 2003 ISBN: 3-527-30484-3

F. Caruso (Ed.)

Colloids and Colloid Assemblies Synthesis, Modification, Organization and Utilization of Colloid Particles 2004 ISBN: 3-527-30660-9

W. Krenkel (Ed.)

High Temperature Ceramic Matrix Composites 2001 ISBN: 3-527-30320-0

I. Manners

Synthetic Metal-containing Polymers 2003 ISBN: 3-527-29463-5

Polymer Brushes Synthesis, Characterization, Applications

Edited by Rigoberto C. Advincula, William J. Brittain, Kenneth C. Caster, Jrgen Rhe

Editors Rigoberto C. Advincula Department of Chemistry University of Houston 136 Fleming Building Houston, TX 77204 USA William J. Brittain Department of Polymer Science University of Akron Akron, OH 44325-3909 USA Kenneth C. Caster Center for Biologically Inspired Materials and Material Systems Pratt School of Engineering Box 90303 Duke University Durham, NC, 27708 USA Jrgen Rhe Institute for Microsystem Technology (IMTEK) University of Freiburg Georges-K@hler-Allee 103 79110 Freiburg Germany

&

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. Library of Congress Card No. applied for British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. Bibliographic information published by Die Deutsche Bibliothek Die Deutsche Bibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data is available in the Internet at . 2004 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 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.

Printed in the Federal Republic of Germany. Printed on acid-free paper. Typesetting K+hn & Weyh, Satz und Medien, Freiburg Printing betz-druck GmbH, Darmstadt Bookbinding Großbuchbinderei J. Sch4ffer GmbH & Co. KG, Gr+nstadt ISBN

3-527-31033-9

V

Contents Preface

XV

List of Contributors

XVII

Polymer Brushes: On the Way to Tailor-Made Surfaces 1 Jrgen Rhe 1 Growth of Polymer Molecules at Surfaces: Introductory Remarks 2 Coatings: From First Principles to High-Tech Applications 3 3 Surface-Coating Techniques 6 4 Surface-Attached Polymers 10 5 Polymer Brushes: General Features 13 6 Theory of Polymer Brushes 15 7 Synthesis of Polymer Brushes 18 8 Polymer Brushes as Functional Materials 22 9 Microstructured Polymer Brushes 24 10 Surface-Initiated Polymerization: The Overall Picture 28

Part I

Synthesis

1

Recent Advances in Polymer Brush Synthesis Anthony M. Granville and William J. Brittain Introduction 35 “Grafting To” Synthesis Technique 37 “Grafting From” Synthesis Technique 41

1.1 1.2 1.3 2

2.1 2.2 2.2.1 2.2.2 2.2.3

1

33 35

Polymer Brushes by Atom Transfer Radical Polymerization Jeffrey Pyun, Tomasz Kowalewski, and Krzysztof Matyjaszewski Introduction 51 Polymer Brushes on Flat Surfaces 52 Controlled ATRP from Flat Surfaces 53 Block Copolymer Brushes on Flat Surfaces 54

51

Stimuli-Responsive Ultrathin Films from “Grafting To” Approach

55

VI

Contents

2.3 2.3.1 2.3.2 2.3.3 2.3.4 2.4 2.4.1 2.4.2

Polymer Brushes from Particles 57 Spherical Brushes from Inorganic Colloids 58 Multilayered Core-Shell Colloids 58 Imaging of Individual Spherical Brushes 61 Modification of Carbon Black Fillers 63 Molecular Brushes 63 Synthesis of Molecular Brushes from Linear Polymeric Macroinitiators 64 Molecular Brushes from Dendritic Macroinitiators 65

3

Polymer Brushes by Atom Transfer Radical Polymerization Initiated from Macroinitiator Synthesized on the Surface 69 Viktor Klep, Bogdan Zdyrko, Yong Liu, and Igor Luzinov Introduction 69 Experimental 72 Results and Discussion 73 Synthesis of Macroinitiator for ATRP 73 ATRP from Macroinitiator 77

3.1 3.2 3.3 3.3.1 3.3.2 4

4.1 4.2 4.3 4.3.1 4.3.2 4.3.3 4.4 4.4.1 4.4.2 5

5.1 5.2 5.3 5.4 5.4.1 5.4.2 5.4.3 5.4.4

Synthesis of Polypeptide Brushes Henning Menzel and Peter Witte Introduction 87

87

Preparation of Peptide Brushes by “Grafting To” 88 Preparation of Peptide Brushes by Grafting From Polymerization 90 Mechanisms of NCA Polymerization 90 Amine-Initiated Grafting From Polymerizations in Solution 93 Other Techniques for Amine-Initiated Grafting From Polymerizations Preparation of Peptide Brushes by Living Grafting From Polymerization 95 Copolymerization Approach 95 Alloc-Amide Approach 99 Bottle Brush Brushes: Ring-Opening Polymerization of Lactide from Poly(hydroxyethyl methacrylate) Surfaces 105 Jong-Bum Kim, Wenxi Huang, Chun Wang, Merlin Bruening, and Gregory L. Baker Introduction 105 Synthesis of PHEMA-g-PLA 109 Conclusions and Implications for Future Studies 114 Experimental Section 115 Materials 115 Preparation of Monomer Solution and Substrates 115 Ring-Opening Polymerization from PHEMA Surface 115 Analytical Methods 115

94

Contents

6

6.1 6.2 6.2.1 6.2.2 6.2.3 6.3 6.3.1 6.3.2

7

7.1 7.2 7.2.1 7.2.2 7.2.3 7.3 7.3.1 7.3.2 7.3.3 7.4 7.4.1 7.4.2 7.4.3 7.5 7.6 7.6.1 7.6.2 7.6.3 7.6.4 7.6.5 7.6.6 7.6.7

Preparation of Well-Defined Organic-Inorganic Hybrid Nanostructures using Living Cationic Surface-Initiated Polymerization from Silica Nanoparticles 119 Il-Jin Kim, Su Chen, and Rudolf Faust Introduction 119 Experimental Section 120 Materials 120 Characterization 120 Synthesis of Immobilized Macroinitiators 121 Results and Discussion 122

Living Cationic Surface-Initiated Polymerization of IB from Silica Nanoparticles in the Presence of Sacrificial Free Initiator 122 Living Cationic Surface-Initiated Polymerization of IB from Silica Macroinitiators 125 Photoinitiated Polymerization from Self-Assembled Monolayers 129 Daniel J. Dyer, Jianxin Feng, Charles Fivelson, Rituparna Paul, Rolf Schmidt, and Tongfeng Zhao Introduction 129 Substrates 131 Silicon, Silica and Glass 131 Planar Gold 131 Nanoparticles 133 Photoinitiated Radical Polymerization Mechanisms 133 Free Radicals 133 Photosensitizers 134 Photo-Iniferters 135 Polymerization from AIBN-type SAMs 135 Design and Synthesis 135 Monolayer Characterization 137 Polymerization of Styrene 138 Conclusions and Future Studies 143 Experimental 144 Initiator Synthesis 144 Polymerizations 145

Reflection Absorption Infrared Spectroscopy (FT-RAIRS) Measurements 146 Ellipsometry 146 X-Ray Photoelectron Spectroscopy (XPS) 146 Molecular Weight Measurements 147 Molecular Modeling 147

VII

VIII

Contents

8

8.1 8.2 8.2.1 8.3 8.3.1 8.3.2 8.4 8.4.1 9

9.1 9.2 9.2.1 9.2.2 9.3 9.3.1 9.3.2 9.4 9.4.1

Recent Advances in the Synthesis and Rearrangement of Block Copolymer Brushes 151 Stephen G. Boyes, Anthony M. Granville, Marina Baum, Bulent Akgun, Brian K. Mirous, and William J. Brittain Introduction and Background 151 Controlled/“Living” Free Radical Polymerization 153 Atom Transfer Radical Polymerization (ATRP) 153 Synthesis of Block Copolymer Brushes 154 Diblock Copolymer Brushes 154 Triblock Copolymer Brushes 158 Rearrangement of Block Copolymer Brushes 160 Rearrangement of Diblock Copolymer Brushes 160 Surface-Grafted Hyperbranched Polymers Hideharu Mori and Axel H. E. Mller Introduction 167 “Grafting To” Approach 170

167

9.4.3 9.4.4

Synthesis of 2D Hybrids by “Grafting To” 170 Synthesis of 3D Hybrids by “Grafting To” 171 Multi-Step Grafting Approach 172 “Grafting To-Grafts” 172 “Grafting From-Grafts” 173 “Grafting From” Approach 173 Synthesis of 2D Hybrids by Surface-Initiated, Self-Condensing Vinyl (Co)polymerization 175 Synthesis of 3D Hybrids by Surface-Initiated Self-Condensing Vinyl (Co)polymerization 178 Theoretical Considerations 181 Other Systems 182

Part II

Characterization

9.4.2

187

10

The Analysis and Characterization of Polymer Brushes: From Flat Surfaces to Nanoparticles 189 Rigoberto C. Advincula 10.1 Introduction 190 10.1.1 Polymer Brushes 190 10.1.2 SIP on Flat Surfaces and Particle Substrates 192 10.2 Characterization of Ultrathin Polymer Films and Polymer Brushes 10.2.1 Spectroscopy and Optical Techniques 194 10.2.2 Microscopy 195 10.2.3 Other Methods 196 10.3 Investigating Polymer Brush Systems 198 10.3.1 Characterization of the Step-by-Step Procedure 198

193

Contents

10.3.2 Investigating the Different Regimes of Polymer Brush Conformation on Surfaces 199 10.3.3 Investigating Phase Segregation and Formation of Patterns 200 10.3.4 Polymerization Mechanism 201 10.3.5 Patterning Using Nonlithographic Methods 204 10.4 The Importance of Characterizing Particles and Nanoparticles 204 10.5 Characterization and Analysis Methods for Polymer Brushes on Particles 205 10.5.1 In-Situ Investigations on Particles 206 10.5.2 Degrafted Polymers from Particles 208 11

11.1 11.2 11.2.1 11.2.2 11.2.3 11.2.4 11.2.5 11.2.6 11.2.7 11.2.8 11.2.9 11.2.10 11.2.11 11.2.12 11.3 12

Characterization of Polymer Brushes on Nanoparticle Surfaces 213 Thomas A. P. Seery, Mark Jordi, Rosette Guino, and Dale Huber Introduction 213 Experimental 215 Materials 215 Instrumentation 215 Pyrolysis GC-MS 216 Infrared Monitoring of Polymer Formation 216 Synthesis of Alkanethiol-Stabilized Gold Nanoparticles 217 Synthesis of Stober Silica Nanoparticles 218 Synthesis of NCSEOS 218

Synthesis of BCH, NCSEOS, and TMEOS-Coated Nanoparticles Synthesis of TMEOS Silica-Polymer Mixture 219 Synthesis of Silica-Poly(norbornene) Nanocomposites 219 Isolation of Grafted Polymer Chains 220 Polymer Stability Test 220 Results and Discussion 221 Spherical Polyelectrolyte Brushes Matthias Ballauff Introduction 231

231

12.1 12.2 Synthesis and Characterization 234 12.2.1 Determination of Core Radius R, Contour Length LC, and Grafting Density r 234 12.2.2 Titration Curve 235 12.3 Experimental Verification of Theoretical Predictions 236 12.3.1 Confinement of the Counterions 237 12.3.2 Correlation of the Counterions to the Macroion 238 12.4 Flow Behavior 240 12.5 Applications 242 12.5.1 Interaction with Charged Surfaces 242 12.5.2 Interaction with Proteins in Solution 243

218

IX

X

Contents

13

Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up with Oppositely Charged Polyelectrolytes 249 Rupert Konradi, Haining Zhang, Markus Biesalski, and Jrgen Rhe 13.1 Introduction 249 13.2 Synthesis and Data Evaluation 251 13.2.1 Synthesis 251 13.2.2 Multiple-Angle Nulling Ellipsometry 252 13.3 Swelling Behavior of Weak Polyelectrolyte Brushes in Aqueous Environments 253 13.3.1 The Influence of pH Value 254 13.3.2 Interaction with Monovalent Cations 254 13.3.3 Interaction with Divalent Cations 257 13.3.4 Interaction with a Trivalent Cation: Aluminum 263 13.3.5 A Classification 264 13.4 Interaction Between Polyelectrolyte Brushes and Oppositely Charged Polyelectrolytes in Solution 265 13.4.1 The Formation of Surface-Attached PEL-PEL Complexes 265 13.4.2 The Formation of PEL Multilayer Assemblies 268 14

Structure and Properties of High-Density Polymer Brushes 273 Yoshinobu Tsujii, Muhammad Ejaz, Shinpei Yamamoto, Kohji Ohno, Kenji Urayama, and Takeshi Fukuda 14.1 Introduction 273 14.2 Controlled Synthesis of High-Density Polymer Brush by ATRP 274 14.3 Structure and Properties of High-Density PMMA Brushes 277 14.3.1 Swollen Brushes 277 14.3.2 Dry Brushes 279 14.4 Application of High-Density Polymer Brushes 282 15

15.1 15.2 15.2.1 15.2.2 15.2.3 15.3 15.4 15.5 15.5.1 15.5.2 15.5.3 15.5.4

Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients on Solid Substrates 287 Tao Wu, Jan Genzer, Peng Gong, Igal Szleifer, Petr Vlcˇek, and Vladim6r 7ubr Glossary 287 Introduction 288 Experimental Section 292 Formation of the Gradient of the Polymerization Initiator 292 Preparation of PtBA and Hydrolysis into PAA 293 Polymer Characterization 294 Theory Section 296 Experimental Results 300 Discussion 304 Surface Hydrolysis of PtBA 304 Dependence of H on Ionic Strength 306 Dependence of H on the PAA Grafting Density 308 Molecular Insight from Calculations 310

Contents

16

Kinetics of Polymer Brush Formation With and Without Segmental Adsorption 317 Lynn S. Penn, Heqing Huang, Roderic P. Quirk, and Tae H. Cheong Introduction 317 Experimental 320

16.1 16.2 16.2.1 Synthesis and Characterization of Amine Chain-End Functionalized Polystyrene 320 16.2.2 Introduction of Active Sites to Surface of Solid 320 16.2.3 Tethering Reactions in Good Solvent 320 16.2.4 Tethering Reactions in Poor Solvent 321 16.2.5 Monitoring the Tethering Reactions 321 16.3 Results and Discussion 323 16.3.1 Results in Absence of Segmental Adsorption 323 16.3.2 Results in the Presence of Segmental Adsorption 325

Part III Applications 17

17.1 17.2 17.2.1 17.2.2 17.2.3 17.3 17.3.1 17.3.2 17.3.3 17.3.4 17.3.5 17.3.6 17.3.7 17.3.8 17.3.9 17.4

18

329

Applications of Polymer Brushes and Other Surface-Attached Polymers 331 Kenneth C. Caster Introduction 331 Surface Modification and Functionalization 332 Polymerization Methodologies for Surface-Attached Polymers 332 Property Control 336 Impact on Types of Materials 336 Applications 336 Adhesion 337 Tribology 341 Stabilization and Compatiblization 341 Surface Coatings 343 Stimuli-Responsive and Switchable Surfaces 345 Separations 346 Nanofabrication 349 Surfaces for Electronics 350 Other Uses 351 Future Prospects 351 Appendix 353 Polymer Brushes: Towards Applications 371 Gregory L. Whiting, Tamer Farhan, and Wilhelm T. S. Huck Introduction 371 Experimental 372 Materials 372 Characterization 373

18.1 18.2 18.2.1 18.2.2 18.2.3 Synthesis of Triphenylamine Acrylate (TPAA) Monomer

373

XI

XII

Contents

18.2.4 18.2.5 18.3 18.3.1 18.3.2 18.3.3

Synthesis and Deposition of Trichlorosilane ATRP Initiator 373 Surface-Initiated Polymerizations 373 Results and Discussion 374 Kinetics of Surface-Initiated ATRP of MMA from Silicon 374 Surface-Initiated ATRP from Polymeric Substrates 375 Synthesis of Conjugated Polymer Brushes from ITO 377

19

Polymerization, Nanopatterning and Characterization of Surface-Confined, Stimulus-Responsive Polymer Brushes 381 Marian Kaholek, Woo-Kyung Lee, Bruce LaMattina, Kenneth C. Caster, and Stefan Zauscher Introduction 381 Experimental 382 Materials 382 Substrates 383 Preparation of Initiator Monolayers 383 Nanopatterning of Initiator 383 NIPAAM Polymerization 384 Polymer Characterization 385 Results and Discussion 386 Surface-Initiated Bulk Polymerization 386 Phase Behavior and Mechanical Characterization 390 Surface Force Measurements 393 Nano-Patterning 396

19.1 19.2 19.2.1 19.2.2 19.2.3 19.2.4 19.2.5 19.2.6 19.3 19.3.1 19.3.2 19.3.3 19.3.4 20

Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale 403 Sergiy Minko, Marcus Mller, Valeriy Luchnikov, Mikhail Motornov, Denys Usov, Leonid Ionov, and Manfred Stamm 20.1 Introduction 403 20.2 Theory of Mixed Polymer Brushes 404 20.3 Synthesis of Mixed Brushes 409 20.3.1 The “Grafting To” Method 409 20.3.2 The “Grafting From” Method 411 20.4 Experimental Study of Phase Segregation in Mixed Brushes 412 20.5 Adaptive Responsive Behavior: Regulation of Wetting and Adhesion 417 20.6 Patterning of Mixed Brushes 420 21

21.1 21.1.1 21.1.2 21.1.3

Local Chain Organization of Switchable Binary Polymer Brushes in Selective Solvents 427 Melbs C. LeMieux, Denys Usov, Sergiy Minko, Manfred Stamm, and Vladimir V. Tsukruk Introduction 427 Polymer Surface Modification 427 Polymer Brushes 428 Binary Polymer Brushes 429

Contents

21.2 21.2.1 21.2.2 21.3 21.3.1 21.3.2 21.3.3

Experimental 431 Materials and Synthesis 431 Methods 432 Results and Discussion 433 Dry State Analysis 433 Morphology in Solvent 434 Mechanical Response in Solvent

22

Motion of Nano-Objects Induced by a Switchable Polymer Carpet 441 Svetlana Prokhorova, Alexey Kopyshev, Ayothi Ramakrishnan, and Jrgen Rhe Introduction 441 Materials 443 Results and Discussion 444

22.1 22.2 22.3 23

437

Photochemical Strategies for the Preparation and Microstructuring of Densely Grafted Polymer Brushes on Planar Surfaces 449 Oswald Prucker, Rupert Konradi, Martin Schimmel, J:rg Habicht, In-Jun Park, and Jrgen Rhe 23.1 Introduction 449 23.1.1 Topological and Chemical Patterning of Surfaces 449 23.1.2 Photochemical Pathways for Grafting Polymers onto Surfaces: A Literature Survey 451 23.2 General Features of Surface-Initiated Polymerization from Monolayers of Azo Initiators 453 23.3 Photolithographic Procedures for the Generation of Microstructured Polymer Brushes on Planar Surfaces 455 23.3.1 Photoablation of Polymer Brushes 455 23.3.2 Photoablation/Photodecomposition of the Initiator Layer followed by Thermal Polymerization 460 23.3.3 Patterning by Photopolymerization 462 23.4 Multifunctional Patterns 465 23.5 Applications of Photostructured Polymer Brushes 466 Index

471

XIII

XV

Preface Amongst others, notably industrial coatings, barriers, packaging, laminates, and lubricants, have been associated with the modification of the surface of an object so as to have a specific interaction or non-interaction towards an external environment. These terms have mostly implied the use of organic polymer, surfactant, and resin type materials. While a large body of academic and technical literature associated with bulk films and coatings already exists, development of ultra thin coatings of the sub-micron order has recently become of great interest, spanning a new field of surface engineers in the last decade. In the area of organic and polymer thin films, one may be familiar with the more common industrial techniques like spin-coating, dipcoating, doctor blade coating, and roll-to-roll coating processes. At the few nm thickness regimes, the application of both self-assembly and directed-assembly methods becomes fascinating. Some of the more “nanostructured” assemblies include: Langmuir-Blodgett films, self-assembled monolayers (multilayers), alternate polyelectrolyte (sequential) deposition, and thermal and molecular beam epitaxy methods of evaporated organic molecules. To these examples, one can now add the method of thin films by polymer brushes. The study of polymer brushes has long been dominated by physicists because of their interest in investigating macromolecule phenomena in confined environments. From the theoretical standpoint, end-tethering of polymers reduces the degrees of freedom for different macromolecule conformations such that it is possible to define a “stretch” conformation for a neutral polymer. From the experimental side numerous innovative surface sensitive spectroscopic and microscopic methods have been developed and applied as a result of this interest. Grafting density, surface concentration, osmotic pressure, solvent quality, interaction parameters, etc. are important factors to consider in experimental methods. Due to the predominance of “physisorption” models in the formation of these confined polymers, not much focus has been given yet on chemically grafted polymers, which are more thermodynamically robust. For a while, chemisorption methods were popular, but very soon, it was realized that significantly higher grafting densities are not achievable with this method. In the area of polymer grafting, a lot of previous work can be cited on particle modification and the use of plasma or irradiation initiated polymerizations. With the recent advances in polymer synthetic methodologies and their adaptation to surface chemistry, it has become possible for synthetic chemists to reclaim

XVI

Preface

this field. The contribution of late is evident. In a technique called “grafting from”, a highly cited term from the first papers by R?he and Pr?cker, it is possible to associate the formation of polymer brushes as a type of surface initiated polymerization. Contributions to this field are numerous and we have tried to include in this book the works by Brittain, Huck, Menzel, Fukuda, Matyjaszewski, Bruening, Minko, Stamm, M?ller, Luzinov, Dyer, Advincula, Seery, Quirk, Stamm, Tsukruk, Zauscher, Boyes, Baker, Ballauf, Caster, Genzer, Faust, their co-authors and many others (see list of contributors). However, this field is growing and is indeed very interdisciplinary. The book starts with an introduction and overview of the field by R?he. The subsequent chapters are then grouped into three major parts: Synthesis, Characterization, and Applications. Each division begins with a review chapter by the editors. This is followed by individual contributions and reviews from invited authors. In Synthesis, an overview in Chapter 1 gives the highlights of recent advances in synthetic methodologies. Efforts have been made to include living free-radical polymerization (Chapters 2 and 3), ring-opening polymerizations (Chapter 4 and 5), cationic polymerization (Chapter 6), and hyper branched polymer synthesis (Chapter 9). Other polymerization mechanisms are reviewed in Chapter 1. In Characterization, it was helpful to outline the different methods for polymer brush analysis on both flat film substrates and particles (Chapter 10). The characterization of particles and flat surfaces is exemplified in Chapters 11, 12 and Chapters 13, 14, 15, 16, respectively. Lastly in the application part, the review in Chapter 17 gives an excellent overview of the myriads of possibilities in applications of polymer brushes: from microelectronics to bio-applications. The contributions include patterning (Chapter 18, 19), mixed polymer brushes (Chapter 20, 21), nano-object movement (Chapter 22), and photochemical strategies in applications (Chapter 23). More chapters could have been included but this collection should well suffice to whet the appetite of the readers. A number of reviews have been written but this work should be the most comprehensive yet. It is hoped that this book will be a valuable reference and resource to scientists, engineers, and technologists in this rapidly evolving field. June 2004

Rigoberto C. Advincula, Houston William J. Brittain, Akon Kenneth C. Caster, Durham Jrgen Rhe, Freiburg

XVII

List of Contributors Rigoberto C. Advincula Department of Chemistry and Center for Materials Chemistry University of Houston Houston, TX 77204-5003 USA

Markus Biesalski Institute for Microsystems Technology University of Freiburg Georges-K2hler-Allee 103 79110 Freiburg Germany

Bulent Akgun Department of Polymer Science University of Akron Akron, OH 44325-3909 USA

Stephen G. Boyes Department of Polymer Science University of Akron Akron, OH 44325-3909 USA

Gregory L. Baker Department of Chemistry and Center for Sensor Materials Michigan State University East Lansing, MI 48824 USA

William J. Brittain Department of Polymer Science University of Akron 170 University Ave. Akron, OH 44325 USA

Matthias Ballauff Physikalische Chemie I University of Bayreuth 95440 Bayreuth Germany

Merlin Bruening Department of Chemistry and Center for Sensor Materials Michigan State University East Lansing, MI 48824 USA

Marina Baum Department of Polymer Science University of Akron Akron, OH 44325-3909 USA

XVIII

List of Contributors

Kenneth C. Caster Center for Biologically Inspired Materials and Material Systems Pratt School of Engineering Box 90303 Duke University, Durham, NC 27708 USA Su Chen Department of Chemistry University of Massachusetts Lowell One University Avenue Lowell, MA 01854 USA Tae H. Cheong Department of Polymer Science University of Akron Akron, OH 44325-3909 USA Daniel J. Dyer Department of Chemistry & Biochemistry Southern Illinois University Carbondale, IL 62901-4409 USA Muhammad Ejaz Institute for Chemical Research Kyoto University Uji, Kyoto 611-0011 Japan Tamer Farhan Melville Laboratory for Polymer Synthesis Department of Chemistry University of Cambridge Lensfield Road Cambridge, CB2 1EW UK

Rudolf Faust Department of Chemistry University of Massachusetts Lowell One University Avenue Lowell, MA 01854 USA Jianxin Feng Department of Chemistry & Biochemistry Southern Illinois University Carbondale, IL 62901-4409 USA Charles Fivelson Department of Chemistry & Biochemistry Southern Illinois University Carbondale, IL 62901-4409 USA Takeshi Fukuda Institute for Chemical Research Kyoto University Uji, Kyoto 611-0011 Japan Jan Genzer Department of Chemical Engineering North Carolina State University Riddick Laboratories PO Box 7905 Raleigh, NC 27695-7905 USA Peng Gong Department of Chemistry Purdue University West Lafayette, IN 47907-1393 USA

List of Contributors

Anthony M. Granville Department of Polymer Science University of Akron 170 University Ave. Akron, OH 44325 USA Rosette Guino Department of Chemistry and Polymer Program University of Connecticut Storrs, CT 06269 USA J+rg Habicht Institute for Microsystem Technology (IMTEK) University of Freiburg Georges-K2hler-Allee 103 79110 Freiburg Germany Heqing Huang Department of Chemical and Materials Engineering University of Kentucky Lexington, KY 40506-0046 USA Wenxi Huang Department of Chemistry and Center for Sensor Materials Michigan State University East Lansing, MI 48824 USA Dale Huber Sandia National Laboratories Nanostructures & Advanced Materials Chemistry Dept. P.O. Box 5800 MS 1421 Albuquerque, NM 87185-1421 USA

Wilhelm T. S. Huck Melville Laboratory for Polymer Synthesis Department of Chemistry University of Cambridge Lensfield Road Cambridge, CB2 1EW UK Leonid Ionov Institut fEr Polymerforschung Dresden Hohe Str. 6 01069 Dresden Germany Mark Jordi Department of Chemistry and Polymer Program University of Connecticut Storrs, CT 06269 USA Marian Kaholek Department of Mechanical Engineering and Materials Science School of Engineering Duke University Box 90300 Durham, NC 27708-300 USA Il-Jin Kim Department of Chemistry University of Massachusetts Lowell One University Avenue Lowell, MA 01854 USA Jong-Bum Kim Department of Chemistry and Center for Sensor Materials Michigan State University East Lansing, MI 48824 USA

XIX

XX

List of Contributors

Viktor Klep School of Materials Science and Engineering Clemson University Clemson, SC 29634 USA

Melbs C. LeMieux Materials Science & Engineering Department Iowa State University Ames, IA 50011 USA

Rupert Konradi Institute for Microsystem Technology (IMTEK) University of Freiburg Georges-K2hler-Allee 103 79110 Freiburg Germany

Yong Liu School of Materials Science and Engineering Clemson University Clemson, SC 29634 USA

Alexey Kopyshev Freiburger Materialforschungszentrum (FMF) University of Freiburg Stefan-Meier-Str. 21 79104 Freiburg Germany Tomasz Kowalewski Department of Chemistry Carnegie Mellon University 4400 Fifth Avenue Pittsburgh, PA 15213 USA Bruce LaMattina Army Research Office PO Box 12211 Research Triangle Park NC 27709-2211 USA Woo-Kyung Lee Department of Mechanical Engineering and Materials Science School of Engineering Duke University Box 90300 Durham, NC 27708-300 USA

Valeriy Luchnikov Institut fEr Polymerforschung Dresden Hohe Str. 6 01069 Dresden Germany Igor Luzinov School of Materials Science and Engineering 161 Sirrine Hall Clemson University Clemson, SC 29634-097 USA Krzysztof Matyjaszewski Department of Chemistry Carnegie Mellon University 4400 Fifth Avenue Pittsburgh, PA 15213 USA Henning Menzel Institute for Technical Chemistry Technical University Braunschweig Hans-Sommer-Str. 10 38106 Braunschweig Germany

List of Contributors

Sergiy Minko Department of Chemistry Clarkson University Potsdam, NY 13699-5810 USA Brian K. Mirous Department of Polymer Science University of Akron Akron, OH 44325-3909 USA Hideharu Mori Makromolekulare Chemie II UniversitGt Bayreuth UniversitGtsstrasse 30 95440 Bayreuth Germany Mikhail Motornov Institut fEr Polymerforschung Dresden Hohe Str. 6 01069 Dresden Germany Axel H. E. M2ller Makromolekulare Chemie II UniversitGt Bayreuth UniversitGtsstrasse 30 95440 Bayreuth Germany

In-Jun Park Institute for Microsystem Technology (IMTEK) University of Freiburg Georges-K2hler-Allee 103 79110 Freiburg Germany Rituparna Paul Department of Chemistry & Biochemistry Southern Illinois University Carbondale, IL 62901-4409 USA Lynn S. Penn Dept. Chemical and Materials Engineering 177 Anderson Hall University of Kentucky Lexington, KY 40506-0046 USA Svetlana Prokhorova Freiburger Materialforschungszentrum (FMF) University of Freiburg Stefan-Meier-Str. 21 79104 Freiburg Germany

Marcus M2ller Institut fEr Physik WA 331 Johannes Gutenberg UniversitGt 55099 Mainz Germany

Oswald Prucker Institute for Microsystem Technology (IMTEK) University of Freiburg Georges-K2hler-Allee 103 79110 Freiburg Germany

Kohji Ohno Institute for Chemical Research Kyoto University Uji, Kyoto 611-0011 Japan

Jeffrey Pyun University of California Department of Chemistry Berkeley, CA 94720-1460 USA

XXI

XXII

List of Contributors

Roderic P. Quirk Department of Polymer Science University of Akron Akron, OH 44325-3909 USA

Manfred Stamm Institut fEr Polymerforschung Dresden Hohe Str. 6 01069 Dresden Germany

Ayothi Ramakrishnan Indian Institute of Technology Madras Chennai 600 036 India

Vladim5r 6ubr Department of Biomedicinal Polymers Institute of Macromolecular Chemistry Academy of Sciences of the Czech Republic 162 06 Prague Czech Republic

J2rgen R2he Institute for Microsystem Technology (IMTEK) University of Freiburg Georges-K2hler-Allee 103 79110 Freiburg Germany Martin Schimmel Institute for Microsystem Technology (IMTEK) University of Freiburg Georges-K2hler-Allee 103 79110 Freiburg Germany Rolf Schmidt Department of Chemistry & Biochemistry Southern Illinois University Carbondale, IL 62901-4409 USA Thomas A. P. Seery Department of Chemistry and Polymer Program University of Connecticut Storrs, CT 06269 USA

Igal Szleifer Department of Chemistry Purdue University West Lafayette, IN 47907-1393 USA Yoshinobu Tsujii Institute for Chemical Research Kyoto University Uji, Kyoto 611-0011 Japan Vladimir V. Tsukruk Materials Science & Engineering 3053 Gilman Hall Iowa State University Ames, IA 50011 USA Kenji Urayama Institute for Chemical Research Kyoto University Uji, Kyoto 611-0011 Japan Denys Usov Institut fEr Polymerforschung Dresden Hohe Strasse 6 01069 Dresden Germany

1

Polymer Brushes: On the Way to Tailor-Made Surfaces Jrgen Rhe

Abstract

In recent years, the synthesis of polymer brushes through surface-initiated polymerization reactions has received significant attention. In this overview, several different synthetic strategies for the generation of polymer brushes are reviewed. The unique physical properties of polymer brushes that arise from the covalent anchoring of the polymer chains to the solid substrate are discussed and compared to the properties of polymer layers deposited by other techniques of thin film generation. Finally, examples are provided that highlight some recent developments aimed at strategies for the functionalization of surfaces with polymer brushes, at ways of realizing smart surfaces with switchable properties, and at the generation of micro- and nanostructured polymer monolayers.

1

Growth of Polymer Molecules at Surfaces: Introductory Remarks

Thin coatings applied to the surface of materials can improve the properties of objects dramatically as they allow control of the interaction of a material with its environment. This has been known more or less empirically to man for several thousand years. Lacquer generated from tree sap was used in China some 7000 years ago as a protective coating for wooden objects. Cold process coatings were also used around 3000 bc, where Egyptian ship builders used beeswax, gelatin and clay to produce varnishes and enamels and (later) coatings from pitch and balsam to waterproof their ships. The early Greeks and Romans, as well as the ancient Asian cultures in China, Japan and Korea, used lacquers and varnishes applied to homes and ships for decoration and as protective measures against adverse environmental conditions. In modern times, the coatings industry is a multi-billion dollar business and – especially if the value of the protected objects is considered – a very important contribution to the world economy. Today, however, the application range of coatings extends much beyond the simple decoration and protection aspects, and functional coatings have become an enabling technology in a vast variety of different high-tech

2

Polymer Brushes: On the Way to Tailor-Made Surfaces

Schematic depiction of the growth of polymer molecules at a surface of a solid substrate through surface-initiated polymerization.

Figure 1

areas. Fields in which such high-tech coatings are applied range from computer chips [1] and hard disk manufacturing [2] to the use of special coatings in biomedical and aviation applications [3,4]. Accordingly, many different techniques have been developed for the generation of protective coatings, and these will be discussed further below. Surface-initiated polymerization reactions as a new pathway for the preparation of functional, high-tech coatings have recently received much attention [5,6]. This technique is based on the growth of polymer molecules at the surface of a substrate in situ from surface-bound initiators, which results in the attachment of polymer molecules through covalent bonds to this substrate (Figure 1). Polymer layers in which the polymer chains are irreversibly attached to the substrate are especially attractive for a variety of applications, as such layers can have a good long-term stability, even in rather adverse environments. For example, it poses no problem to expose surfaces with such surface-attached coatings to good solvents for the polymers without being concerned that the polymer will be either dissolved or displaced, and that the coating is more or less rapidly removed from the surface. In addition to the issue of stability, the number of functional groups present at a surface can also be greatly enhanced by connecting large polymer molecules with functional groups to the surface instead of binding the functional groups directly to that surface. Such a “skyscraper” approach allows high densities of functional groups to be obtained at the surface of the substrate through moving from the strictly two-dimensional arrangement of these groups present in typical surfaces to a more three-dimensional situation. An example, which illustrates such a behavior is the attachment of DNA probe molecules to surface-attached polymer chains, which can significantly enhance the sensitivity of a DNA-chip (Figure 2). Systems in which the polymer chains are attached with one end to a solid substrate are very interesting, not only from a chemical but also from a physical point of view. If the grafting density of the polymer molecules is very high, the polymer chains adopt a rather unusual conformation wherein the individual coils overlap [7–9]. Under these conditions, the polymer molecules are strongly stretched away

2 Coatings: From First Principles to High-Tech Applications

Fluorescence image obtained from a DNA chip based on a oligonucleotide functionalized polymer brush. The pattern and the intensity of the spots allows for the determination of the sequence of the unknown analyte-DNA.

Figure 2

from the surface and achieve a molecular shape which is far from the typical random coil conformation that polymer molecules assume in solution. Such surfaceattached films with strongly stretched chains are usually referred to as “polymer brushes” [10]. Polymer brushes are very interesting systems, as the strong stretching of the polymer chains leads to concurrent drastic changes in the physical properties of the systems. For unstretched polymer chains, a slight molecular deformation leads to a moderate increase of the energy stored in the system (entropy elasticity). However, when the molecules are already strongly stretched – as is in the case of a polymer brush – the energy penalty for the same small deformation is large. Accordingly, in all situations where the stretching of the polymer chains is of concern – for example, during the shearing of such surfaces or when the film is penetrated by other polymer chains from solution – very strong differences can be observed to the behavior of free coils [11–13]. Whilst systems in which polymer chains have one end tethered to a substrate appeared some years ago to be quite exotic, and significant doubts persisted that such brushes with high grafting densities could be obtained in practice, the development of methods where polymers are grown directly on the surface of a substrate by using surface-initiated polymerization has led to a large number of such systems becoming available. However, before describing more detailed aspects of surface-initiated polymerization, more general aspects of coatings will be briefly discussed.

2

Coatings: From First Principles to High-Tech Applications

For a large number of chemical and physical processes – both in daily life and in technical applications – the bulk properties of a material as well as the structure and composition of its surfaces determine the performance of the entire system. In order to control the interaction of a material with its environment, coatings consisting of thin organic films are frequently applied to the surfaces of these solids (Figure 3). In many cases, the coating serves simply as a barrier against a hostile envi-

3

4

Polymer Brushes: On the Way to Tailor-Made Surfaces

Schematic depiction of use of thin polymer coatings to control the interaction of a material with the surrounding environment.

Figure 3

ronment and allows for protection against corrosion or other chemical or photochemical degradation. Although corrosion protection is certainly the most prominent aspect of surface coatings as far as market and materials volumes are concerned, thin organic coatings are also applied in a large number of high-tech applications, ranging form microelectronics [1] to biomedical devices [3,4]. When considering such applications, thin organic coatings are applied to control the interactions between the material and its environment. Examples of interface properties which can be controlled by deposition of a thin organic film onto a surface include friction [11,13–17], adhesion, adsorption of molecules from the surrounding environment, or wetting with water or other liquids. In medical applications, coatings allow control of the interaction of biological cells and biomolecules with artificial materials in order to enhance the biocompatibility of an implant, or to avoid the nonspecific adsorption of proteins onto the active surfaces of an analytical device [18]. It is well known that coatings, even when only a few Angstroms thick, can influence the surface properties of a material so strongly that the chemical nature of the underlying material becomes completely hidden and the interaction of the whole system with the surrounding environment is governed by these extremely thin coatings (“stealth effect”). This is an advantageous situation for materials engineering as it allows optimization of the bulk and surface properties of a material separately from each other. In addition, the application of functional coatings allows the coverage of a surface with groups which interact with other molecules in their environment through specific molecular recognition processes. Such a strategy is, for example, very important for the control of the adhesion of biological cells to artificial substrates. In such a case, thin layers containing cell recognition peptide sequences can induce strong adhesion of the cells to the substrate surfaces, to which they otherwise would show only a very unfavorable adhesion behavior [19]. One example of a system where the covering of a surface with an ultrathin coating is a prerequisite for that system to function is a computer hard disk [2] (Figure 4). If the uncoated surface of a thin film magnetic disk is subjected to strong shear, such as the sliding of a read/write head on the disk surface, then almost instantaneous damage can be observed. The disk shows, even upon the first contact with the head, a strong stick-slip behavior and a high friction coefficient, while the debris

2 Coatings: From First Principles to High-Tech Applications

a)

(a) Computer hard disks are protected against mechanical wear by ultrathin layers of perfluorinated polymers. (b) Hard disk in an accelerated wear test: (i) unlubricated and (ii) after application of 1.5 nm chemisorbed and 1 nm physisorbed lubricant; the high friction coefficient and

Figure 4

5

b)

the strong noise indicate a strong stick-slip behavior, which is the beginning of a catastrophic failure of the system. (Reprinted with kind permission from Ref. [2]; . American Society of Mechanical Engineering, 1996.)

generated by this damage leads to rapid failure of the disk. However, if a film of a perfluoropolyether of typically only 2–4 nm thickness is attached to the disk, the tribological properties are greatly improved, the wear is reduced, and the mean time to failure of the disk is greatly prolonged (Figure 4). A second example where ultrathin organic coatings control the performance of the whole system is the control of interface properties of materials in contact with blood. If artificial materials are brought into contact with blood, then blood proteins such as fibrinogen adsorb very rapidly to the surfaces of the implant or sensor surface, followed by the adhesion of blood cells to these protein layers. This reaction cascade leads almost immediately to strong changes of the surface composition of the active surfaces of the sensor or implant. After a short period of time, blood clots

SEM image of a fibrin network and thrombocytes on the surface of an artifical heart value (picture courtesy of Dr. A. Schlitt, University Hospital Mainz, Germany).

Figure 5

6

Polymer Brushes: On the Way to Tailor-Made Surfaces

become attached to the surfaces of the blood-exposed materials (Figure 5). Eventually, the blood clots can break off from the surface into the blood stream, where they pose a life-threatening situation for the patient [18]. It has been shown, that the application of just a polymer layer which is just a couple of nanometers thick can dramatically reduce the adhesion of the blood proteins and thereby greatly improve the blood compatibility of the material [20–22].

3

Surface-Coating Techniques

Depending on the type of interaction between the molecules which are constituents of the coating and the substrate which is to be modified, two classes of strategies for the deposition of thin organic coatings can be distinguished. In one of these, the molecules interact with the substrate by physical forces [7–9], whilst the other class consists of molecules which are attached to the surfaces through chemical bonds. In the latter case, a monomolecular layer or a surface-attached network is very strongly (“irreversibly”) attached to the surface. This classification is not simply a formality, but the type – and accordingly the strength of interaction – also has a very strong influence on the physical properties of the coating, the film thicknesses which can be obtained through such a method, and the long-term stability of the coating in problematic environments. A number of technologically important coating techniques rely on physical interactions between the deposited molecules and the substrate, including: . . . . .

painting/droplet evaporation spray coating spin coating dip coating doctor blading

Although being quite different in detail, a common feature of all of these processes is that the molecules are deposited from solution and the solvent evaporates during the coating process (Figure 6). The techniques described above are somewhat empirical in nature, as certain parameters such as the rate of evaporation of the solvent depend on specific details of the individual process and are accordingly difficult to predict a priori, but in many cases are simple to reproduce. Accordingly if the deposition conditions are properly controlled, layers with well-defined thickness and good homogeneity can be generated without major effort. Several of these processes, such as dip- and spin-coating, allow the deposition of extremely thin film coatings (starting from just a few nanometers thickness), but essentially no upper limit to film thickness exists, if appropriate conditions are applied. In contrast to these rather empirical processes, more sophisticated coating techniques have been developed, including the Langmuir-Blodgett technique [23], the adsorption of monomolecular layers of homo- and block copolymers [7] from solution, and the Layer-by-Layer (LbL) [24] technique in which multilayer stacks of oppo-

3 Surface-Coating Techniques

a)

b)

c)

Schematic illustration of different processes used for the deposition of organic molecules and/or polymers on surfaces: (a) spin-coating. (b) Langmuir-Blodgett-Kuhn (LBK) technique; (c) adsorption from solution.

Figure 6

sitely charged polyelectrolytes are deposited onto a (charged) substrate. These techniques allow for much better control of the internal structure of the deposited layers, and also for extremely high precision with regard to the thickness of the coatings. All of the coating techniques, except perhaps for the Langmuir-Blodgett technique, are – from a technological viewpoint – rather simple, and the generation of layers typically requires no complicated set-ups to generate the coatings. The molecules are attached to their substrates by physical interactions, and consequently the forces holding them at the surface are rather weak. In some cases this situation is desirable, but in others it becomes problematic as it is more likely to lead to adhesive failure of the system. Under unfavorable conditions, the films can be subject to destruction by the “Big Four Ds”:

7

8

Polymer Brushes: On the Way to Tailor-Made Surfaces . . . .

desorption during solvent exposure; displacement by molecules which have stronger interaction with the surface; dewetting (for films above the glass transition temperature, Tg); and delamination (for films below Tg).

Desorption and displacement are especially important, as coatings are usually not prepared and kept under ideal (i.e., ultrahigh vacuum) conditions, but rather are exposed to environments containing all sorts of contaminants. In these “real-life” environments, contaminants are present on every surface, and/or competing adsorbates will fight for surface sites during or after the coating process. Examples of molecules which are present in many different environments, and which typically compete quite efficiently with coating materials for surface sites, include water, ions, polyelectrolyte molecules, or oils. The contaminants or displacing agents might have such a strong interaction with the surface that the molecules of the coating can no longer remain in contact with the substrate, but eventually will instead be located on top of a thin layer consisting of contaminant/displacer molecules. In this respect, polar surfaces which absorb ambient water are especially problematic, as water has a strong affinity for such surfaces and exhibits a very high adsorption enthalpy. Accordingly, water functions as a very efficient competitor for surface sites and easily displaces adsorbed molecules from such high-energy surfaces. What renders the situation even worse is that under these conditions the surface properties of the material become strongly dependent on the history of the sample – that is, which environment the sample has been exposed before use – and this is, potentially, a very problematic situation. Dewetting occurs in all systems, where the surface tension of the substrate is lower than that of the coating material if the molecules are allowed to reach an equilibrium. This may be achieved either by heating the film above Tg, or by exposure to molecules which can act as a plasticizer for the polymers of the coating. In contrast, delamination occurs if the films are in the glassy state and subjected to wide temperature swings, or if the coating swells in the environment to which it had been exposed, while the substrate does not swell. In such cases, strong mechanical stress develops at the interface, and this may cause the entire film to peel off, leading to large-scale adhesive failure. An alternative to the above-mentioned procedures which allows improvement in the long-term stability of coatings even in very adverse environments, is to attach the molecules of the coating to the surface of the substrate through chemical bonds. The price which must be paid for an enhanced stability of the system is a more complicated coating procedure and/or the requirement to choose the coating conditions more carefully, so that the surface reaction proceeds in high yield and with limited side reactions. A current, very frequently employed strategy for the preparation of well-controlled surface layers is the use of small molecules with a reactive head group that is amenable to form a covalent bond with a corresponding chemical moiety on the surface of the substrate, which is to be modified. As this process is selflimiting – that is, the surface-attachment reaction stops when all the reactive surface groups have been consumed or are no longer accessible – such layers are commonly

3 Surface-Coating Techniques

Schematic of the self-assembly process and examples of anchor groups used for the modification of surfaces with selfassembled monolayers (SAMs) of organic molecules.

Figure 7

called self-assembled monolayers (SAM) [25]. Examples are silanes on oxide surfaces, phosphates or phosphonate on metal(oxide)s, and thiols or disulfides on noble metal surfaces (Figure 7). In this way, surface coatings can be obtained which are very stable and may even have a strong degree of positional and orientational order. In some cases, even crystalline packing of the surface-attached molecules has been observed. If molecules are assembled that carry at their tail end a specific chemical moiety or a biochemically active group, it is possible to obtain a more or less strict 2D arrangement of these functionalities (Figure 8) [26]. Examples are molecules which contain fluorocarbon segments in the assembling units [27–29], and can convert a hydrophilic surface into a highly water-repellent hydrophobic one, or the introduction of “ligands” as recognition sites in bio-affinity assays. In this way, surfaces can be generated – for example, on top of the transducer of a biosensor – that very specifically bind proteins from solution [30,31].

Example of a structure prepared via soft lithography from acid and methyl-terminated thiols. First, the Me-terminated thiol was stamped onto a gold surface; second, the

Figure 8

unmodified areas were backfilled from a solution containing the acid-terminated thiol. (Reprinted with kind permission from Ref. [26]; .Wiley-VCH, 1998.)

9

10

Polymer Brushes: On the Way to Tailor-Made Surfaces

In some of these applications the intrinsic limitations of this strictly 2D arrangement of the functional groups are evident: the maximal surface density of the functional moieties is limited by the surface area cross-section of the assembled unit. In some cases it is even lower than the arrangement of the individual functional units at such high packing densities in some cases leads to a mutual blocking or, at least, to a limited accessibility. One obvious solution to the above problem is the extension into the third dimension – that is, the use of polymers carrying the functional groups along the chain, thus generating higher cross-sectional densities of these groups and simultaneously guaranteeing good accessibility.

4

Surface-Attached Polymers

Most approaches which aim at attaching polymers to a surface use a system where the polymer carries an “anchor” group either as an end group or in a side chain. This anchor group can be reacted with appropriate sites at the substrate surface, thus yielding surface-attached monolayers of polymer molecules (termed “grafting to”) (Figure 9) [32–37]. While the attachment of terminally functionalized polymers to the surface leads to layers, where one group is connected to the surface, side chain attachment usually leads to multiple attachment points and, accordingly, a rather flat conformation of the polymer molecules. In the latter case, the functional groups of different molecules compete for reactive sites on the surface, and accordingly the amount of polymer which can be immobilized depends strongly on the reaction conditions, and especially on the concentration of the polymer in solution. This chemical linking of polymers to a substrate surface is, in principle, closely related to the formation of self-assembled monolayers of low molecular-weight compounds described above. Accordingly, if such (end)functionalized polymers are availa)

b)

Schematic illustration of different processes used for the attachment of polymers to surfaces: (a) “grafting to”; (b) grafting via incorporation of surface-bound monomeric units; (c) “grafting from/surface-initiated polymerization”.

Figure 9

c)

4 Surface-Attached Polymers

able (which is a nontrivial condition, as the synthesis of polymers with reactive end groups is far from being trivial), the attachment of the polymers is, from a chemical point of view, rather simple. Another straightforward technique for the attachment of polymers to surfaces which allows the generation of a great variety of functional surfaces is to carry out a polymerization reaction in the presence of a substrate onto which monomers had been attached [32,38–41]. In such a polymerization reaction, the surface-attached monomers are incorporated into growing polymer chains in the very same way as their peers in solution (Figure 9). However, once one or more surface-attached monomers are incorporated, the polymer is “glued” firmly to the surface. During the process, a macroradical initially attacks the monomers on the surface, while in a second step further monomers units are added, so that the chain grows again, away from the surface. However, careful studies of the polymerization mechanism have shown that the “grafting to” step represents the bottle-neck of the reaction and thus limits the polymer immobilization [42,43]. Accordingly, very similar layers are obtained by using immobilized monomers as by the chemisorption of preformed chains. A general limitation of the technique is that the substrate must be immersed in a polymerization solution, but if this poses no problem it is one of the simplest techniques to generate surface-attached layers, especially as there is no need to synthesize a polymer functionalized with an anchor group. Although “grafting to” reactions are easy to perform, it should be noted that certain rather strict limitations apply to the structures which can be realized by use of a “grafting to” strategy. First, the use of reactive anchor groups for the surface-attachment of polymers imposes some rather strict limitations on the choice of functional groups available for incorporation into the polymer. One of the reasons for this is that the functional groups on the polymer can compete with the anchor moieties for surface sites. Especially if the aim is to immobilize functional polymers containing highly polar or charged groups onto polar surfaces, the adsorption of functional groups to the surface can be very strong and compete very effectively with the chemisorption process. Such competition between anchor and functional groups has been observed, for example, in the case of the attachment of a low molecular-weight alkoxysilane containing amine groups to a silicon oxide surface [44–46]. In such a system, interactions between the basic amine groups of the SAM-forming silane and the rather acidic silanol groups of the silicon (oxide) substrate can strongly compete with the condensation reaction of the alkoxysilyl moiety with the substrate silanol groups. As result, layers are obtained which contain both physisorbed molecules due to acid–base interactions and chemically attached molecules. Second, in order to obtain a fast and complete surface attachment reaction with a high surface density of chains covalently bound to the substrate, rather reactive anchor groups are required. These groups, however, tend not to tolerate the simultaneous presence of a large variety of functional groups in the polymer. For example, if a chlorosilyl group is chosen as an anchor group for the attachment of the polymer to a silicon oxide surface, this choice excludes the incorporation of many functional groups into the polymer, including amine-, hydroxyl- or carboxylic acid moieties, as these would react with the chlorosilyl groups.

11

12

Polymer Brushes: On the Way to Tailor-Made Surfaces

At first view, there is a tendency to consider that resorting to less reactive anchor groups – for example, using a less reactive alkoxysilane instead of a highly reactive chlorosilane – would solve the problem. This, however, is incorrect as a more indepth analysis shows. If the nucleophilicity of the anchor group is reduced, this affects both the undesired side reaction, which is the reaction of the functional group with the anchor group, and which leads to loss of anchor moieties, and the desired reaction of the anchor group with a group at the surface of the substrate, which results in a successful chemisorption reaction. Accordingly, both reactions are slowed down at the same time, and the ratio between the rates of the two reactions remain the same in all cases. Another complication inherent to “grafting to” processes is an intrinsic limitation of the film thickness, and accordingly the number of functional groups per surface area which can be obtained by using such an approach. Films generated by chemisorption from solution are limited to (dry) thicknesses of typically 1 to 5 nm. This limitation has both kinetic and thermodynamic origins. With increasing coverage of the surface with attached chains, the polymer concentration at the interface quickly becomes larger than the concentration of polymers in solution. Additional chains, which are to become attached to the surface, must diffuse against this concentration gradient that ever increases with increasing grafting density of the attached polymer (Figure 10). This diffusion slows down the immobilization reaction at the surface further and further as the reaction proceeds. Thus, the rate of the attachment reaction levels off rather quickly and further polymer is linked to the substrate only at an extremely slow rate due to this kinetic hindrance. Indeed, it has been shown [42,43] both theoretically and experimentally that once the surface-attached coils overlap, the attachment of further polymer molecules takes place on a logarithmic time scale, and already at rather low graft density time frames of thousands or even millions of years would be required to add a few more nanometers of polymer to the layer. Accordingly, as far as practical reaction times are concerned, films generated by this technique are intrinsically limited with regard to the film thickness. Furthera)

Figure 10 Schematic illustration of the “grafting to” process. (a) Chains that are to be attached to the surface can easily reach the surface at low graft densities. (b) The attachment process comes to a virtual halt as soon as the

b)

surface is covered with polymers, as the already attached chains form a kinetic barrier against which incoming chains have to diffuse to reach the surface.

5 Polymer Brushes: General Features

more it should be noted that, even if this kinetic limitation is somehow circumvented, the attachment of chains to a strongly covered surface becomes unfavorable also for thermodynamic reasons. At high grafting densities the surface-attached polymer chains are in a rather stretched conformation due to the presence of strong segment–segment interactions, as will be discussed in more detail below. A chain, which is now becoming attached to the surface, must change from a coil conformation in solution to a stretched (“brush-like”) conformation at the surface. The entropy loss during this process, however, is only compensated by the establishment of one chemical bond, namely the one connecting the polymer to the surface. Hence, the higher the graft density of the chains at the surface, the stronger will be the entropy penalty, and this rapidly precludes the attachment of further chains.

5

Polymer Brushes: General Features

As mentioned briefly above, the term “polymer brush” refers to a system in which chains of polymer molecules are attached with one or with a few anchor points to a surface in such a way that the graft density of the polymers is high enough that the a)

b)

c)

Figure 11 Artist’s perception of the terms (a) “mushroom”, (b) “pancake” and (c) “brush” used for the different possible conformations of surface-attached polymers.

13

14

Polymer Brushes: On the Way to Tailor-Made Surfaces

surface-attached chains become crowded and are stretched away from the surface (Figure 11). From the stretching of the polymer chains perpendicular to the surface, several new physical phenomena arise. Examples are ultralow friction surfaces [11,12] obtained through coating of two surfaces that slide against each other with polymer brushes, or the so-called autophobic behavior [47–50], in which materials coated with surface-attached polymer chains do not become wetted by free polymer, even if the surface-attached and the free chains are chemically identical (Figure 12).

Figure 12 Optical micrograph of a dewetted polystyrene layer (initial thickness 60 nm) on top of a polystyrene brush (6 nm; prepared via grafting from). This picture was taken after annealing the sample for 40 h at 180 CC (scale bar = 200 lm). (Reprinted from Ref. [50], with kind permission; . American Chemical Society, 1996.)

In the following discussion, the focus will be placed on polymer brushes at solid surfaces, although brush-like chain conformations can also be obtained at the boundary between phases in block copolymers [8] or in so-called molecular “bottlebrushes” [51–53]. In the latter system, polymers are attached as side chains to the backbone of a polymer molecule, so that every segment of the backbone carries such a polymeric side chain. Although the overall physical picture for the different systems is very similar, here only chains attached to solid surfaces at one end will be described and discussed. When polymer molecules are tethered to a surface, two basic cases must be distinguished depending on the graft density of the attached chains [8–10]: 1.

2.

If the distance between two anchoring sites is larger than the size of the surface-attached polymers, the segments of the individual chains do not “feel” each other and behave more or less like single chains “nailed” down onto the surface by one end. Depending on the strength of interaction of the polymer segments with the surface, again two cases must be distinguished [10]. If the interaction between the polymer and the surface is weak (or even repulsive), the chains form a typical random coil that is linked to the surface through a “stem” of varying size. For such a situation, the term “mushroom” conformation has been coined (Figure 11). However, if the segments of the surfaceattached chains adsorb strongly to the underlying surface, the polymer molecules obtain a flat, “pancake”-like conformation (Figure 11). A completely different picture is obtained if the chains are attached to the surface at such short distances between the anchor points that the polymer molecules overlap. In this case, the segments of the chains try to avoid each other as much as possible and minimize segment–segment interactions by

6 Theory of Polymer Brushes

stretching away from the surface (Figure 11). This chain stretching, however, reduces the number of possible polymer conformations, which is equivalent to a reduction in the entropy of the chains. This loss of entropy gives rise to a retracting force trying to keep the chains coiled, as occurs in a stretched piece of rubber. Thus, a new equilibrium at a higher energy level is obtained in which the chains are stretched perpendicular to the surface. 6

Theory of Polymer Brushes

The theoretical description of polymer brushes attached to surfaces of different topologies – that is, planar and curved surfaces – is well developed [7–9]. However, as in this book the main focus is set on new developments concerning the chemical methodology, only a very brief outline of the theory of brushes is provided here. For a more detailed discussion, the reader is referred to reviews recently published on this subject [7–9]. The key idea behind the theoretical description of polymer brushes is that the free energy F of the chains is obtained from a balance between the interaction energy between the statistical segments Fint and energy difference between stretched and unstretched polymer chains Fel (elastic free energy) caused by the entropy loss of the chains: F = Fint + Fel

(1)

The most important parameters, which are of interest for a description of brush systems, are the segment density profile (u(z)) of the surface-attached chains and/or the brush height h as a function of the graft density r, the molecular weight (/degree of polymerization) of the surface-attached chains, and the solvent quality of the contacting medium (Figure 13). The first description of such a brush system has been attempted by Alexander [54] for monodisperse chains consisting of N segments, which are attached to a flat, non-adsorbing surface with an average distance of the anchor points d much smaller

Figure 13 Two hundred chains of a polymer brush (chain length N = 100) under good solvent conditions. (Reproduced with kind permission from Ref. [11]; . Springer, 1998.)

15

16

Polymer Brushes: On the Way to Tailor-Made Surfaces

Figure 14 Schematic illustration of the Alexander model for the theoretic description of polymer brushes. The chain segments with the “blobs” (indicated by the circles) behave as random (“Gaussian”) coils. (d represents the average distance between anchor points.)

than the radius of gyration of the same unperturbed chains not in contact with the surface (Figure 14). If both the interaction energy resulting from binary monomer– monomer interactions and the elastic energy of a Gaussian chain are calculated and minimized in respect to the brush height h, the following equation is obtained for brushes in a good solvent: h ~ N N r1/3

(2)

In a poor solvent – that is, close to H conditions – the exponent describing the influence of the grafting density is slightly different and h ~ N N r1/2

(3)

is obtained. It should be noted, that in both cases the brush height scales linearly with the degree of polymerization/molecular weight of the polymer molecules, which is a much stronger dependency than that of the size of a polymer coil in solu-

Figure 15 Schematic illustration of segment density profiles for surface-attached polymers in different regimes. For details, see the text (adopted from [9]).

6 Theory of Polymer Brushes

tion on the molecular weight, where the radius of gyration Rg, scales with Rg ~ N0.59 for a polymer in a good solvent and Rg ~ N0.50 for solutions close to H conditions. Although the Alexander model is very simple, it predicts the experimentally observed scaling behavior more or less correctly and allows an understanding of some of the most striking properties of polymer brushes, such as lubrication and the wetting behavior. More sophisticated models have been developed to describe the segment density profile of the brushes (Figure 15). To this numerical and analytical self-consistent field (SCF), theories [55–57] for such systems have been proposed based on the assumptions that, for strong stretching and high molecular weights of the brushes, fluctuations around the most favorable configuration of the polymer chain diminish. A general result of the SCF calculations is, that the segment density profile is more or less parabolic as long as the grafting density is moderate and the molecular weight of the brush chains is high. At very high grafting densities the SCF assumptions are no longer valid, as three body interactions between the polymer segments become significant. The results of the SCF calculations have been verified both experimentally and in simulations. For the latter, molecular dynamics and Monte Carlo methods have been employed [58]. If smaller differences in the numerical coefficient are neglected, then the SCF results are in good agreement with the results from simple scaling arguments. In addition to these somewhat straightforward calculations, more complicated situations have also been tackled where the polymer chains have a distinct polydispersity [59], are in specific topologies such as attached to small particles [60], which exhibit a significant curvature also on the molecular scale, and to brushes which carry charges along the polymer chain [61] (Figure 16). In particular, the latter case can become very complicated if the polymer chains interact specifically with ions in the surrounding medium, as under these circumstances the situation can no longer be described by simple mean field approaches, but specific complex formation and (local) changes in the solubility of the polymer play a key role in describing the swelling behavior of such brushes.

Figure 16

Schematic illustration of a polyelectrolyte brush (PEL brush).

17

18

Polymer Brushes: On the Way to Tailor-Made Surfaces

7

Synthesis of Polymer Brushes

An obvious requirement for forcing polymer molecules into brush-like conformations is that the strength of anchoring of the molecules to the interface is sufficiently high that the molecules are connected irreversibly to the surface of the substrate. A second requirement is that the synthetic strategy allows for the generation of grafting densities high enough to cause sufficient repulsive segment–segment interactions within the surface-attached chains to induce significant chain stretching. In particular, the latter condition imposes some strict limitations onto the appropriate synthetic strategy for brush formation as the chains lose a considerable amount of entropy when stretched into an elongated form. In the following section, four different approaches to reach these goals will be briefly discussed. A complete review of the published literature on this subject would clearly be beyond the limits of this introductory chapter.

Approach 1

In the first approach, amphiphilic block copolymers consisting of a water-soluble block and a water-insoluble block are spread at the air-water interface [62,63]. The water-soluble block attempts to dissolve into the aqueous subphase, but is anchored to the air-water interface by the hydrophobic block. Upon compression of the thus obtained Langmuir monolayer, the distance between the anchor points of the polymer chains decreases and the hydrophilic block is stretched away from the surface into the aqueous subphase. A prerequisite for this is that the hydrophobic block is in the molten state, because only is it possible for a rearrangement of the chains within the film to occur upon compression. Furthermore, it is important that the hydrophilic balance is chosen in such a way that the loss of chains to the subphase and the formation of micelles can be avoided. The thus obtained films can be crosslinked through photochemical reactions and transferred to a solid substrate.

Approach 2

In the second case, block copolymers or end-functionalized polymers are physisorbed to a solid surface [7]. End-functionalized polymers can be discussed together with block copolymers as they are structurally very similar to such systems in terms of their essential physics of adsorption to a solid surface. In some ways they can be viewed as block copolymers with a very short block, consisting only of one unit. In the block copolymer concept, one block adsorbs strongly at the surface and acts as an anchor for the polymer chains. The other block adsorbs only weakly at the surface – that is, the interactions of the polymer with the solvent are stronger than those with the surface – and so the block floats in the solvent like a buoy. Although during the past, many different polymer layers have been prepared by this route, the

7 Synthesis of Polymer Brushes

chemical variability of these systems is somewhat limited as a solvent must be available in which the block copolymer adsorbs to the surface without formation of micelles either in solution or at the surface. Furthermore, as the layer formation requires the diffusion of polymer molecules through the layer of already attached chains, this limits the range of graft densities that can be obtained using this technique. In addition, as the interaction with the surface is based simply on physical interactions, anchoring of the molecules to the substrate surface is relatively weak, and this further limits the graft densities available and decreases the stability of the films.

Approach 3

As has been discussed above, the chemisorption of polymer molecules leads to chains which are covalently attached to surfaces [32–37]. Although situations can be envisioned in which the polymer chains are slightly stretched, such processes are strongly limited in terms of the obtainable graft density, especially for high molecular-weight polymers, and this results in only relatively weak stretched polymer chains.

Approach 4

Much higher graft densities can be obtained when the polymer chains are grown at the surface of the substrate in situ (Figure 17) [5,6]. To this initiator, species are either generated or self-assembled at the surface of the substrate, followed by initiation of chain growth from these surface-attached initiators, for example by controlled or free radical chain polymerization. The surface-polymerization can be started thermally either through a chemical process or photochemically. In this way, polymer monolayers with film thicknesses of more than 2000 nm in the dry state have been obtained (Figure 18). In this case, polymer molecules with number aver-

Figure 17 Common synthetic strategy for the generation of polymer brushes via surface-initiated polymerization. An initiator molecule is deposited on a surface by means of a self-assembly process via the reaction of an anchor group to suitable surface sites and, subsequently, chains are grown on the surface from the initiating sites.

19

20

Polymer Brushes: On the Way to Tailor-Made Surfaces

a)

b)

Figure 18 (a) Optical waveguide spectrum (symbols) obtained from a PMMA brush deposited on an evaporated SiO2 layer. The solid line was obtained from model calculations based on a Fresnel formalism assuming a 2200 nm-thick polymer layer. The sample was

prepared in neat MMA at 50 CC, polymerization time: 96 h. (b) Thickness of PMMA brushes as a function of monomer concentration; polymerizations were carried out at 60 CC for 18 h in toluene as a solvent (if required).

age molecular weights of several 106 g mol–1 are attached at distances of anchor points of less than 3 nm. Surface-initiated polymerization reactions work for any polymer which can be obtained by a chain growth reaction such as free and controlled radical polymerization, carbocationic polymerization, anionic polymerization, and ring-opening metathesis polymerization (Table 1) [63–97]. The different polymerization reactions can be carried out on surfaces of very different topologies (planar, curved, and irregular surfaces), and allow for the generation of polymers from a wide spectrum of different monomers. It would be far beyond the scope of this overview to try to review all recent developments on the synthesis of such systems, and a large variety of different synthetic routes for the generation of polymer brushes through surface-initiated polymerizations will be detailed in the following chapters. However, at this point some comments should be made on controlled or living polymerization reactions for the growth of polymer molecules through surface-attached initiators. In this respect, liv-

7 Synthesis of Polymer Brushes Selected systems for the generation of polymer brushes via surface-initiated polymerization. The list is by no means exhaustive, and is only meant to demonstrate the wide variety of synthetic strategies that have been developed over the past decade.

Table 1

Mechanism

Initiator/initiating species

H N

Me O Si Me

Free radical

Maximum thickness Reference(s) (nm)

CN N Me

O

N

O

N Me

O

N

N Me

O

N

N

Me Me

73,74

120 nm

75,76

150 nm; 700 nm (water accelerated)

77–83

OH

Si (CH2)n

O

Br

O O

O

N

O Si (CH2)2

SO2Cl

100 nm

84–86

Me O Si (CH2)2 Me

CH2Cl

n.a.

87–89

< 60 nm

82,90

< 40 nm

91–97

Me O Si Me

Me O

Br O

Various systems for cationic and anionic polymerizations, RAFT and reverse ATRP

NA = not applicable.

n.a.

CN Me

Me O Si Me

Others

65–72

CN

TEMPO

ATRP

up to 2200 nm

Me Me

N

O

O

O

63,64

CN

Me O Si Me

O

~ 100 nm

CN

CN S (CH2)11 O

CO2H

CN

CN

Me O Si Me

Me

21

22

Polymer Brushes: On the Way to Tailor-Made Surfaces

ing systems with rapid initiation are of major interest as they allow, in principle, surface-attached polymer chains with relatively narrow molecular weight distributions to be obtained. This facilitates comparison with theoretical models developed for surface-attached polymer brushes, provided that the initiation process is sufficiently efficient to allow high graft densities and that the molecular weight of the surfaceattached chains is high enough to allow such a discussion. Indeed, controlled polymerization approaches are expected to become even more interesting for the synthesis of surface-attached polymer brushes, as a large variety of functional brushes can also be obtained by using these methods. At present, major efforts are made – especially in the area of controlled radical polymerization – to polymerize functionalized monomers to create high molecular-weight compounds with low polydispersity.

8

Polymer Brushes as Functional Materials

For many applications of polymer brushes, it is not simply protection against mechanical or chemical damage that is important. Rather, where the polymer layer acts

Figure 19

Examples of functional groups incorporated into polymer brushes.

8 Polymer Brushes as Functional Materials

as a barrier against contact with the environment, a more specific chemical response to the surrounding medium is desirable. Examples of this situation include layers into which DNA, protein molecules or complexing agents – each of which shows a specific reaction towards certain metals – are chemically incorporated [99]. To this end, polymers with desired functional groups can be formed directly from the corresponding monomers (Figure 19). For example, brushes carrying either charges (“polyelectrolyte brushes”) [71,74,75,100–102] or pendant mesogenic units (“LCbrushes”) [103,104] have been prepared using this direct route. An alternative would be first to generate a brush from a simple and inexpensive precursor monomer containing a reactive group, and this can then be transformed into the final moiety through a polymer analogous reaction. Examples of such compounds are monomers carrying an active ester, epoxide, azalactone or amine groups [99]. It is quite evident that, in principle, the direct approach is much simpler as the desired brush can be prepared in a one-step reaction. However, this places some rather stringent requirements on the availability of the monomer, because if an incorporation of repeat units with especially valuable groups into the polymer is desired, then the amount of the valuable monomer needed for the brush generation is rather large. The reason for this is that the molecular weight of the brushes is, for most polymerization mechanisms, directly connected to the monomer concentration; consequently, if high molecular-weight polymers are desired, then relatively large amounts of monomer are required. A second requirement is that the functional group is compatible with the polymerization process used for brush formation. Monomers containing moieties that show excessive transfer properties such as sulfur groups cannot be used in direct polymerization processes as they would lead to side reaction and/or only low molecular-weight brushes. This is especially important, as for a surface-initiated polymerization reaction any chain transfer is equivalent to a termination reaction, because after the transfer further polymer is only generated in solution and removed in a subsequent extraction of the film. In addition to this, a two-step pathway for the generation of functional brushes has the advantage that it is not necessary to study the polymerization behavior of each new monomer with a new functional group “from scratch” because a number of different functionalities can be incorporated using the same precursor monomer. Examples are brushes of homo- or copolymers with N-hydroxysuccinimide ester or epoxide groups through which a large variety of different functionalities can be introduced by aminolysis. For example, the preparation of brushes that carry thiol, pyrene, oligoethyleneoxide or bioactive groups such as peptides or oligonucleotide units have been reported using the same precursor monomer. If the direct polymerization procedure is applied, then each and every one of these monomers must be studied with regard to the polymerization kinetics in order to obtain an in-depth understanding of the brush-forming properties. The use of “living” polymerization reactions – that is, reactions where the number of active or dormant and thus potentially active species remains more or less constant on the time scale of the polymerization reaction – allows the generation of brushes which carry at the end pointing away from the surface a functional group, or brushes which consist of a copolymer [96,98,105,106]. The latter constitute a very

23

24

Polymer Brushes: On the Way to Tailor-Made Surfaces

interesting system, as all polymer molecules are surface-attached and accordingly large-scale, irreversible reorganizations of the chains are prohibited, and the morphology of the polymer film is directly coupled to the composition of the copolymer brush. Thus, upon exposure to an environment – which is selective for one of the two components – the morphology of the layers of such copolymer brushes can be easily switched from one morphology to the other, and monolayers with very unusual topographies can be obtained. Another interesting system is generated if not all of the initiator is used up during the polymerization reaction, or if two different initiators are co-immobilized on the substrate surface. In such a case, after completion of the growth of one polymer species, some initiator is still present which can be used to kick-off the polymerization of another monomer [105,107,108]. This then results in the growth of a second type of polymer in direct neighborhood to the chains already attached to the surface. Such systems – which commonly are called “mixed brushes” – seem especially attractive as the two polymers can have very different interaction strengths with the surrounding of the film. This situation is very similar to that of block copolymer brushes described above. If one environment strongly prefers one polymer over the other, whilst a second environment favors the reverse situation, surfaces with switchable surface chemistries are generated. When the polymer layer is alternately exposed to one or the other environment, the internal structure of the polymer changes accordingly and a system that can adapt to the substrate environment (“smart surfaces”) is obtained.

9

Microstructured Polymer Brushes

The (micro-)patterning of polymer brushes is especially interesting as all the polymer molecules are permanently attached to the surface [73,109]. This is an important aspect, both for the generation of the patterns as well as for applications of the microstructured surfaces, as it allows exposure of the microstructures to good solvents for the polymers. The latter aspect is especially important for biological applications, as it allows strong swelling of the brush and provides a soft “cushion” for the biological system at the surface of the substrate. This is of special significance as proteins tend to denature in contact with hard, solid surfaces. Also, from the viewpoint of preparing microstructured systems, the generation of thick, surfaceattached monolayers is rather attractive as it allows the washing away of reagents after completion of a chemical reaction in the patterned structures, and hence the generation of multifunctional chemical patterns with high resolution (Figures 20 and 21). Indeed, in addition to simple chemical structures being “written” into the film, the use of step-and-repeat procedures allows the generation of very complicated chemical surfaces and structures. This contrasts strongly with the conventional lithographic procedures used in the semiconductor industry where, upon irradiation and solvent exposure, a relief is generated and hence topological rather than chemical structures are generated on the surface of the substrate.

9 Microstructured Polymer Brushes

a)

b) Figure 20 (a) Process used by Hawker et al. for the generation of polymer brushes with spatially resolved properties. A poly(t-butyl methacrylate) brush is covered with a photoresist containing a photoacid generator. Upon illumination of the sample through a mask, protons are generated in the illuminated areas. The protons diffuse into the underlying brush and hydrolyze the ester groups. (b) Illustration of the different wetting properties of a sample prepared as described in (a). The water on the sample only wets the illuminated areas – that is, the areas in which the chains were transformed to a poly-(methacrylic acid). (Reprinted with kind permission from Ref. [77]; . American Chemical Society, 2000.)

25

26

Polymer Brushes: On the Way to Tailor-Made Surfaces

a)

b)

c)

Figure 21 System used by Carter, Hawker et al. for the tuning of the feature size on nanostructures via a combined process consisting of (a) nanoimprinting and (b) surface-initiated polymerization from initiator sites (“inimers”) embedded into the mold; the AFM and SEM

images shown in (c) demonstrate the feature size of lines after nonoimprinting (A,B) and after the subsequent surface-initiated polymerization (C,D). (Reprinted with kind permission from Ref. [111]; . American Chemical Society, 2003.)

In principle, three different strategies can be followed for the generation of chemically micropatterned brushes, besides the trivial photoablation of the polymers by deep UV-irradiation: 1. 2. 3.

Deposition of the initiator in a patterned fashion and/or spatially addressed deactivation of a complete initiator monolayer. Spatially controlled growth of the polymer molecules through local addressing of the initiator and/or confinement of the monomer access. Spatially addressed chemical transformations of precursor brushes.

9 Microstructured Polymer Brushes

In the first case, the initiator is deposited (by inkjet printing or stamping) in certain areas of the substrate [110,111]. In a subsequent reaction step, the polymer is generated through growth of the polymer chains. In further reaction steps, initiator can be deposited in other, still uncovered areas of the substrate. Alternatively, a complete initiator monolayer can be formed and in selected areas the initiator deactivated or photochemically destroyed and evaporated, followed by growth of the brushes. In the latter case (photoablation of the initiator), new initiator can be attached to the substrate in the thus obtained blank areas, either directly or after a short etching process. In the second case, the surface-attached polymer chains can be generated through photopolymerization reactions or other means, spatially to kick-off the polymerization reaction. An alternative, in which many different polymers can be formed in a)

b)

Figure 22 (a) Schematic description of the l-stamping process used for the spatially resolved deposition of laminin to a brush containing active ester groups. (b) Neuronal cells aligning along the laminin grid deposited via this process.

27

28

Polymer Brushes: On the Way to Tailor-Made Surfaces

one polymerization step, would be to supply the monomer only locally, followed by simultaneous induction of the polymerization process for all monomers in the different locations– for example, through flood exposure with UV irradiation or through thermal initiation (depending on the monomer). In the third case, a precursor polymer brush would first be formed in pretty much the same way as has been described above for the generation of functional brushes. The only difference here is that the transformation reactions into the final functional brush are carried out locally through administering the reagents [77] (Figure 22).

10

Surface-Initiated Polymerization: The Overall Picture

Without attempting to gather all the information available for the synthesis of polymer monolayers by growing chains away from the surface, it can be safely stated that surface-initiated polymerization to generate tailor-made surfaces is becoming increasingly accepted. The number of systems for which the surface-attached layers of the initiator and the polymers have been well characterized, and the mechanism of the growth of surface-attached layers is well understood, has grown considerably during the past few years. Although only a short time ago it was doubtful that systems with a high graft density and strong stretching of the polymer chains could be synthesized at all, a variety of approaches is now available that allows to study the physics of densely grafted polymer brushes at ease. As most systems described in the literature have been extensively extracted with good solvents for the polymer, it is also clear that these systems show a high stability, even under rather adverse conditions. This is a clear distinction from other techniques of deposition of thin polymer films, where only weaker interactions to the substrate are employed. By using surface-initiated polymerization reactions, a wide range of monolayers containing functional groups has been synthesized over a wide spectrum of substrates. Indeed, one of the most attractive features of surface-initiated polymerization reactions is that they lead to highly swellable polymer layers attached to spherical, tubular, planar, and even very irregular surfaces such as those of components of complex microsystems. Another key feature of surface-initiated polymerization is that the application of a local stimulus allows local initiation of the polymerization reaction which, in turn, yields a spatially addressed growth of polymer chains. The performance of such locally addressable attachment of polymer molecules to generate chemically l-structured surfaces is particularly of interest, as this cannot be achieved by other techniques of thin layer deposition. Although at present it seems premature to outline the practical implications of surface-initiated polymerization reactions in depth, the future will undoubtedly reveal a wide variety of applications, most notably in the areas of microsystems technology and biomedical devices. On this basis, aspects of the practical applications of systems generated by surface-initiated polymerization will be discussed in Chapter 17.

References

References 1 Introduction to Microlithography, 2nd Ed.

2 3 4

5 6 7 8

9

10 11 12

13 14 15 16 17

18

19 20

(Eds.: L. F. Thompson, C. G. Willson, M. J. Bowden) American Chemical Society, Washington DC, 1994. J. RPhe, V. Novotny, T. Clarke, G. B. Street, J. Tribol. Trans. ASME 1996, 118, 663. M. Tirrell, E. Kokkoli, M. Biesalski, Surf. Sci. 2002, 500, 61. Surface and Interfacial Aspects of Biomedical Polymers (Ed.: J. P. Andrade), Plenum Press, New York, 1985. B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677. J. Pyun, T. Kowalewski, K. Matyjaszewski, Macromol. Rapid Commun. 2003, 24, 1043. A. Halperin, M. Tirrell, T. P. Lodge, Adv. Polym. Sci. 1992, 100, 31. R. A. L. Jones, R. W. Richards, Polymers at Surfaces and Interfaces, Cambridge University Press, Cambridge, 1999. G. J. Fleer, M. A. Cohen Stuart, J. M. H. M. Scheutjens, T. Cosgrove, B. Vincent, Polymers at Interfaces, Chapman & Hall, London, 1993. P. G. deGennes, J. Physique 1976, 37, 1445. G. S. Grest, in: Polymers in Confined Environments, Vol. 138, 1999, pp. 149. J. Klein, Y. Kamiyama, H. Yoshizawa, J. N. Israelachvili, G. H. Fredrickson, P. Pincus, L. J. Fetters, Macromolecules 1993, 26, 5552. J. Klein, E. Kumacheva, D. Mahalu, D. Perahia, L. J. Fetters, Nature 1994, 370, 634. G. S. Grest, Curr. Opin. Colloid Interface Sci. 1997, 2, 271. G. S. Grest, Phys. Rev. Lett. 1996, 76, 4979. J. Klein, Colloid Surfaces – Physicochem. Eng. Aspects 1994, 86, 63. J. Klein, E. Kumacheva, D. Perahia, D. Mahalu, S. Warburg, Faraday Disc. 1994, 173. B. D. Ratner, A. S. Hoffmann, F. J. Schoen, J. E. Lemons, Biomaterials Science, An Introduction to Materials in Medicine, Academic Press, San Diego, 1996. E. Ruoslahti, Annu. Rev. Cell Dev. Biol. 1996, 12, 697. K. L. Prime, G. M. Whitesides, J. Am. Chem. Soc. 1993, 115, 10714.

21 J. M. Harris, S. Zalipsky, Poly(ethylene glycol):

Chemistry and Biological Applications, American Chemical Society, Washington DC, 1997. 22 T. Lehmann, J. RPhe, Macromol. Symp. 1999, 142, 1. 23 K. Mathauer, F. Embs, G. Wegner, in: Comprehensive Polymer Science, Vol. 1. Suppl. (Ed.: G. Allen), Pergamon Press, Oxford, 1992, p. 449. 24 Multilayer Thin Films. Sequential Assembly of Nanocomposite Materials (Ed.: G. Decher), Wiley-VCH, Weinheim, 2002. 25 A. Ulman, An Introduction to Ultrathin Organic Films, Academic Press, New York, 1991. 26 Y. N. Xia, G. M. Whitesides, Angew. Chem. Int. Ed. 1998, 37, 551. 27 M. J. Pellerite, E. J. Wood, V. W. Jones, J. Phys. Chem. B 2002, 106, 4746. 28 A. Hozumi, K. Ushiyama, H. Sugimura, O. Takai, Langmuir 1999, 15, 7600. 29 P. W. Hoffmann, M. Stelzle, J. F. Rabolt, Langmuir 1997, 13, 1877. 30 D. Kambhampati, P. E. Nielsen, W. Knoll, Biosensors Bioelectronics 2001, 16, 1109. 31 W. Knoll, M. Zizlsperger, T. Liebermann, S. Arnold, A. Badia, M. Liley, D. Piscevic, F. J. Schmitt, J. Spinke, Colloids Surfaces – Physicochem. Eng. Aspects 2000, 161, 115. 32 K. P. Krenkler, R. Laible, K. Hamann, Angew. Makromol. Chem. 1953, 53, 101. 33 N. Tsubokawa, M. Hosoya, K. Yanadori, Y. Sone, J. Macromol. Sci. – Chem 1990, A27, 445. 34 N. Tsubokawa, A. Kuroda, Y. Sone, J. Polym. Sci. 1989, A27, 1701. 35 H. Ben Ouada, H. Hommel, A. P. Legrand, H. Balard, E. Papirer, J. Colloid Interface Sci. 1988, 122, 441. 36 K. Bridger, B. Vincent, Eur. Polym. J. 1980, 16, 1017. 37 R. Laible, K. Hamann, Adv. Colloid Interface Sci 1980, 13, 65. 38 A. V. Dmitrenko, N. E. Shadrina, S. S. Ivanchev, N. N. Ulinskaya, A. M. Volkov, J. Chromatogr. 1990, 520, 21. 39 K. Hashimoto, T. Fujisawa, M. Kobayashi, R. Yosomiya, J. Macromol. Sci. – Chem 1982, A18, 173.

29

30

Polymer Brushes: On the Way to Tailor-Made Surfaces 40 K. Hashimoto, T. Fujisawa, M. Kobayashi,

41

42 43 44 45 46 47 48 49 50 51

52

53 54 55

56 57 58 59 60 61 62 63 64

R. Yosomiya, J. Appl. Polym. Sci. 1982, 27, 4529. V. I. Trachenko, Y. N. Zil’berman, T. F. Shatskaya, E. G. Pomerantseva, Polym. Sci. USSR 1986, 28, 646. A. Kopf, J. Baschnagel, J. Wittmer, K. Binder, Macromolecules 1996, 29, 1433. R. Zajac, A. Chakrabarti, Physical Rev. E 1995, 52, 6536. H. Ishida, C. H. Chiang, J. L. Koenig, Polymer 1982, 23, 251. S. Naviroj, J. L. Koenig, H. Ishida, J. Macromol. Sci. Physics 1983, B22, 291. A. M. Zaper, J. L. Koenig, Polymer Composites 1985, 6, 156. G. Reiter, Actualite Chimique 2003, 12. G. Reiter, R. Khanna, Phys. Rev. Lett. 2000, 85, 2753. G. Reiter, R. Khanna, Langmuir 2000, 16, 6351. G. Reiter, P. Auroy, L. Auvray, Macromolecules 1996, 29, 2150. K. L. Beers, S. G. Gaynor, K. Matyjaszewski, S. S. Sheiko, M. Moller, Macromolecules 1998, 31, 9413. M. Wintermantel, M. Gerle, K. Fischer, M. Schmidt, I. Wataoka, H. Urakawa, K. Kajiwara, Y. Tsukahara, Macromolecules 1996, 29, 978. Y. Tsukahara, K. Mizuno, A. Segawa, Y. Yamashita, Macromolecules 1989, 22, 1546. S. Alexander, J. Physique 1977, 38, 977. E. B. Zhulina, O. V. Borisov, V. A. Pryamitsyn, T. M. Birshtein, Macromolecules 1991, 24, 140. C. M. Wijmans, J. Scheutjens, E. B. Zhulina, Macromolecules 1992, 25, 2657. S. T. Milner, T. A. Witten, M. E. Cates, Macromolecules 1988, 21, 2610. P. Y. Lai, K. Binder, J. Chem. Physics 1992, 97, 586. S. T. Milner, T. A. Witten, M. E. Cates, Macromolecules 1989, 22, 853. R. Hariharan, C. Biver, J. Mays, W. B. Russel, Macromolecules 1998, 31, 7506. P. Pincus, Macromolecules 1991, 24, 2912. H. Ahrens, S. Forster, C. A. Helm, Phys. Rev. Lett. 1998, 81, 4172. H. Ahrens, S. Forster, C. A. Helm, Macromolecules 1997, 30, 8447. G. Boven, M. Oosterling, G. Challa, A. J. Schouten, Polymer 1990, 31, 2377.

65 G. Boven, R. Folkersma, G. Challa,

A. J. Schouten, Polym. Commun. 1991, 32, 50. 66 T. Stohr, J. Heinz, J. RPhe, Mater. Sci. Eng. C-Biomimetic and Supramolecular Systems 1999, 8-9, 225. 67 T. Stohr, J. RPhe, Macromolecules 2000, 33, 4501. 68 O. Prucker, J. RPhe, Macromolecules 1998, 31, 592. 69 O. Prucker, J. RPhe, Macromolecules 1998, 31, 602. 70 J. RPhe, Macromol. Symp. 1998, 126, 215. 71 M. Biesalski, D. Johannsmann, J. RPhe, Macromol. Symp. 1999, 145, 113. 72 O. Prucker, J. RPhe, Langmuir 1998, 14, 6893. 73 O. Prucker, J. Habicht, I. J. Park, J. RPhe, Mater. Sci. Eng. C-Biomimetic and Supramolecular Systems 1999, 8-9, 291. 74 X. Guo, M. Ballauff, Langmuir 2000, 16, 8719. 75 X. Guo, A. Weiss, M. Ballauff, Macromolecules 1999, 32, 6043. 76 M. Husseman, E. E. Malmstrom, M. McNamara, M. Mate, D. Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang, T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424. 77 M. Husemann, M. Morrison, D. Benoit, K. J. Frommer, C. M. Mate, W. D. Hinsberg, J. L. Hedrick, C. J. Hawker, J. Am. Chem. Soc. 2000, 122, 1844. 78 X. Y. Chen, D. P. Randall, C. Perruchot, J. F. Watts, T. E. Patten, T. von Werne, S. P. Armes, J. Colloid Interface Sci. 2003, 257, 56. 79 C. Perruchot, M. A. Khan, A. Kamitsi, S. P. Armes, T. von Werne, T. E. Patten, Langmuir 2001, 17, 4479. 80 J. B. Kim, W. X. Huang, M. L. Bruening, G. L. Baker, Macromolecules 2002, 35, 5410. 81 W. X. Huang, J. B. Kim, M. L. Bruening, G. L. Baker, Macromolecules 2002, 35, 1175. 82 A. Ramakrishnan, R. Dhamodharan, J. RPhe, Macromol. Rapid Commun. 2002, 23, 612. 83 T. von Werne, T. E. Patten, J. Am. Chem. Soc. 2001, 123, 7497. 84 K. Matyjaszewski, P. J. Miller, N. Shukla, B. Immaraporn, A. Gelman, B. B. Luokala, T. M. Siclovan, G. Kickelbick, T. Vallant,

References H. Hoffmann, T. Pakula, Macromolecules 1999, 32, 8716. 85 M. Ejaz, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 2870. 86 M. Ejaz, Y. Tsujii, T. Fukuda, Polymer 2001, 42, 6811. 87 Y. Tsujii, M. Ejaz, S. Yamamoto, T. Fukuda, K. Shigeto, K. Mibu, T. Shinjo, Polymer 2002, 43, 3837. 88 T. von Werne, T. E. Patten, J. Am. Chem. Soc. 1999, 121, 7409. 89 X. Huang, M. J. Wirth, Macromolecules 1999, 32, 1694. 90 X. Y. Huang, M. J. Wirth, Anal. Chem. 1997, 69, 4577. 91 J. D. Jeyaprakash, S. Samuel, R. Dhamodharan, J. RPhe, Macromol. Rapid Commun. 2002, 23, 277. 92 R. A. Sedjo, B. K. Mirous, W. J. Brittain, Macromolecules 2000, 33, 1492. 93 M. Baum, W. J. Brittain, Macromolecules 2002, 35, 610. 94 M. D. K. Ingall, C. H. Honeyman, J. V. Mercure, P. A. Bianconi, R. R. Kunz, J. Am. Chem. Soc. 1999, 121, 3607. 95 R. Jordan, A. Ulman, J. F. Kang, M. H. Rafailovich, J. Sokolov, J. Am. Chem. Soc. 1999, 121, 1016. 96 B. Zhao, W. J. Brittain, J. Am. Chem. Soc. 1999, 121, 3557. 97 B. Zhao, W. J. Brittain, Macromolecules 2000, 33, 342.

98 B. Zhao, W. J. Brittain, Macromolecules 2000,

33, 8813. 99 J. RPhe, W. Knoll, J. Macromol. Sci. -Polym.

Rev. 2002, C42, 91. 100 M. Biesalski, J. RPhe, Macromolecules 2003,

36, 1222. 101 M. Biesalski, D. Johannsmann, J. RPhe,

J. Chem. Phys. 2002, 117, 4988. 102 M. Biesalski, J. RPhe, Macromolecules 1999,

32, 2309. 103 B. Peng, J. RPhe, D. Johannsmann, Adv.

Mater. 2000, 12, 821. 104 B. Peng, D. Johannsmann, J. RPhe, Macro-

molecules 1999, 32, 6759. 105 S. A. Prokhorova, A. Kopyshev, A. Ramak-

rishnan, H. Zhang, J. RPhe, Nanotechnology 2003, 14, 1098. 106 B. Zhao, W. J. Brittain, W. S. Zhou, S. Z. D. Cheng, J. Am. Chem. Soc. 2000, 122, 2407. 107 S. Minko, D. Usov, E. Goreshnik, M. Stamm, Macromol. Rapid Commun. 2001, 22, 206. 108 A. Sidorenko, S. Minko, K. Schenk-Meuser, H. Duschner, M. Stamm, Langmuir 1999, 15, 8349. 109 O. Prucker, M. Schimmel, G. Tovar, W. Knoll, J. RPhe, Adv. Mater. 1998, 10, 1073. 110 T. A. von Werne, D. S. Germack, E. C. Hagberg, V. V. Sheares, C. J. Hawker, K. R. Carter, J. Am. Chem. Soc. 2003, 125, 3831. 111 R. R. Shah, D. Merreceyes, M. Husemann, I. Rees, N. L. Abbott, C. J. Hawker, J. L. Hedrick, Macromolecules 2000, 33, 597.

31

Part I

Synthesis

35

1

Recent Advances in Polymer Brush Synthesis Anthony M. Granville and William J. Brittain

1.1

Introduction

Recently, the synthesis of polymer brushes has garnered attention, due in part to their unique properties and applications [1–4]. Colloidal stabilization [5], surface property tailoring, “chemical gates” [3], and microlithographic patterning [2] have all been discussed as potential applications for these materials. Polymer brushes can be described as polymer chains tethered to a surface or interface with a sufficiently high grafting density such that the chains are forced to stretch away from the tethering site [6]. The purpose of this introductory chapter is to provide a brief overview of the recent literature on the synthesis of polymer brushes. The chapter is intended to augment previous reviews on polymer brushes [7].

Figure 1.1

Examples of physically attached polymer brushes.

36

1 Recent Advances in Polymer Brush Synthesis

The formation of these tethered chains is generally carried out using either of two techniques, namely physisorption or chemical bonding of chains to an interface. Typically, polymer brushes synthesized using a physisorption approach consists of two-component polymer chains, where one part strongly adheres to the interface and the second part extends to generate the polymer layer [8]. This tethering point can be a single point, in the case of a functionalized polymer chain, or a diblock copolymer chain. Figure 1.1 depicts several examples of physically attached diblock copolymer brushes where the interface tethering varies. Due to the physical nature of the tethering points, the brush layers are rendered thermally and solvolytically unstable, and there is poor control of grafting density. Covalently grafting chains, either by the “grafting to” or “grafting from” technique, to the surface can overcome these shortcomings. In the “grafting to” method (Scheme 1.1), preformed polymer chains containing a suitable end-functionalized group are reacted with a surface to obtain the desired brush [9]. Although the brush layer exhibits thermal and solvolytic stability, it inherently possesses a low grafting density and film thickness on the surface. This observation is due to the inability of large polymer chains to diffuse to the reactive surface sites that are sterically hindered by surrounding bonded chains.

Scheme 1.1 Idealized representation of the “grafting to” (left) and “grafting from” (right) approaches.

Figure 1.2

Examples of chemically attached polymer brushes.

1.2 “Grafting To” Synthesis Technique

In order to circumvent these disadvantages, a “grafting from” technique (Scheme 1.1) is utilized in which the polymer brush layer is generated in situ from a suitable surface immobilized initiator [9]. Due to the small molecule nature of immobilized initiators, which are predominantly formed via a self-assembled monolayer (SAM) containing the desired initiating functionality [10], high grafting densities are easily obtained. Since brush length is directly proportional to polymer brush molecular weight, control over the polymerization would lead to uniform, tunable brush thickness. For the purposes of this review, the scope will be limited to the recent work in the field of covalently bound polymer brush research. Furthermore, the review will focus primarily on polymer brush synthesis from flat and particulate surfaces. The studies into the area of “bottle-brushes” or “molecular brushes” (see Figure 1.2, Polymers from Macromonomers) conducted by several researchers [11–14] will be omitted.

1.2

“Grafting To” Synthesis Technique

As previously described, a preformed macromolecule possessing a suitable endfunctionality is reacted with a reactive substrate to generate the polymer brush in the “grafting to” technique. The synthesis of the polymer chains has been performed using several techniques such as anionic, cationic, living free radical, and ring-opening metathesis polymerization. These techniques in particular allow for the facile conversion of the chain ends to any number of desired functionalities (hydroxyl, carboxyl, amino, thiol, etc.). The polymerization methods also present the added bonus of synthesizing polymer chains possessing narrow molecular weight distributions, which allows for a uniform brush layer thickness. The substrate surface also plays an important role in the synthesis of the polymer brush layers. Silica and gold surfaces possess surface functionalities that can undergo condensation reactions with polymer chains containing thiol, hydroxyl, and carboxyl functionalities. Furthermore, these and other surfaces can be modified by the use of SAMs and other coupling agents to introduce various other surface functionalities. Tran and Auroy [15] anionically polymerized styrene, using sec-butyllithium in benzene, followed by termination with trichlorosilane to generate polystyryl trichlorosilane chains. Once a silicon wafer was coated with the polymer, the film was annealed on the surface at 160 FC to allow the chlorosilane chain ends to react with the hydroxyl surface functionality. The polystyrene brushes were converted to poly(styrene sulfonate) via a soft sulfonation reaction using acetyl sulfate. However, the brush was observed to degraft from the surface either immediately after sulfonation or steadily after the polyelectrolyte was subjected to water treatment. The “grafting to” method has a limited surface grafting density due to the increased steric hindrance of grafted chains. This steric hindrance inhibits diffusion of large free polymer chains to diffuse to open-surface reactive sites and creation of a dense polymer brush layer. It is this low surface density that allowed for small

37

1 Recent Advances in Polymer Brush Synthesis

molecules to migrate to the polymer brush anchoring sites and resulted in brush degrafting. Therefore, these authors backfilled the surface prior to sulfonation with deuterated polystyrene oligomers as a means of consolidating the polymer brush surface and protecting the anchoring sites. Sirard et al. [16] took a similar approach to polymer brush synthesis by spin-coating deuterated poly(dimethylsiloxane) (PDMS) onto silicon wafers containing surface hydroxyl functional groups. The PDMS was obtained commercially, and contained a monofunctional silanol group, which can also undergo condensation with the hydroxyl surface to generate a covalently bound brush. The spin-coated samples were annealed in order to allow for condensation reaction between the chain end and the surface functionalities to occur. As with the studies conducted by Auroy and Tran [15], neutron reflectivity of the polymer brush layer was performed to characterize the brush layer. Minko et al. [17] took a slightly different approach to the synthesis of polymer brushes using the “grafting to” technique first utilized by Luzinov [18]. Both groups first grafted 3-glycidoxypropyl trimethoxysilane (GPS) to silicon wafer substrates. A condensation reaction took place between the alkoxysilane end group and the hydroxyl surface functionality to allow for surface modification of the substrates. However, the nature of the GPS molecules gave rise to several types of surface functionalities (hydroxyl, carboxyl, and epoxide rings; Scheme 1.2).

H3C

OH OH

+ H3C

OH

O Si

O

O

O

O H3C

silica substrate

silica substrate

OH

OH

Si

O

O

O

Si

+

O O

O O

HOOC

OH

O

Si

CH3

O O

OH

O OH Si

O

OH O

O

O Si

O OH

O O

+

OH

O

Si

CH3

O O

HOOC

O

silica substrate

OH

silica substrate

38

Si

O

O O OH

O O

O Si

O

Si

O

OH

O O

O

CH3 O

OH

OH

Scheme 1.2 Schematic representation of the synthesis of binary brushes. PS chains, gray; PVP chains, black [17].

From this modified silica surface, carboxyl-terminated polystyrene (PS) was grafted via a spin-coating/thermal annealing methodology. However, Minko and coworkers then removed ungrafted PS chains using a THF Soxhlet extraction step in order to allow for the subsequent grafting of carboxyl-terminated poly(2-vinylpyridine) (PVP) to the surface. In this manner, they were able to synthesize a binary polymer brush layer where the composition can be controlled through the time and temperature of the grafting at each step. Furthermore, they were able to show through the use of atomic force microscopy (AFM) and X-ray reflectivity that the

1.2 “Grafting To” Synthesis Technique

surface was macroscopically homogeneous yet nanoscopically able to phase segregate. Minko et al. [19] were able to expand on this approach in later investigations by grafting onto other types of substrates. Rather than using silicon wafers with a GPSmodified surface, poly(tetrafluoroethylene) (PTFE) foils were oxygen- and ammonia plasma-treated to yield hydroxy and amino functional groups, respectively. Sequential polymer grafting was carried out with carboxyl-terminated poly(styrene-co-pentafluorostyrene) (PSF) and PVP. The PSF copolymer was synthesized using a 4,4¢-azobis(4-cyanopentanoic acid) free radical initiator, whereas the PVP was anionically polymerized. Control experiments conducted under the same sequential grafting conditions on silicon wafers yielded 3.5 mg m–2 for each polymer, and a total thickness of 7 nm. Surface testing under various solvolytic conditions yielded a tunable surface that was capable of controlled wettability, adhesion, and chemical composition. Finally, a vinyl-terminated polymer chain can be grafted to a surface via a simple addition reaction. Maas et al. [20] studied the effect of hydroxyl and hydride surface functionality on the grafting behavior of two different molecular weight, vinyl-terminated PS chains. When the two different PS chains (20 and 200 monomeric units)

R1

CH2 CH

R1

R2

CH2 CH

O

CH2 H3C

+

N

CH3

S O

CH3 Cl

-

R2

O -

+

Na

poly(sodium 4-styrenesulfonate) [NaPSS]

R1

CH2 CH

R2

CH2 CH

H3C

N

O

CH3

poly(N,N-dimethylacrylamide)

+

N

S O

R2 O

N

CH2 O

C

N

O H3C

[PDMA]

CH2 CH

CH2

C H3C

O CH2 C

CN

C R1

CH3

R1 = C

poly((ar-vinylbenzyl) trimethylammonium chloride) [PVBTAC]

H3C

CH3

S

2

CH3

R2 = S

3

O -

poly(3-[2-(N-methylacrylamido)-ethyldimethylammonio] propane sulfonate)-b-poly(N,N-dimethylacrylamide) [P(MAEDAPS-b-DMA)]

Figure 1.3 Polymer chains synthesized by McCormick [21] and coworkers utilizing aqueous reversible addition-fragmentation chain transfer (RAFT) techniques.

-

O Na+

39

40

1 Recent Advances in Polymer Brush Synthesis

were solution-grafted on hydride surfaces, a negligible brush thickness was seen. This solution grafting to the surface is performed under reflux conditions to allow for the anchoring reaction to occur. However, when the same chains were annealed to the surface after spin coating, a roughly seven-fold increase in brush thickness was obtained with the PS200 sample as compared to the PS20 sample. This same brush layer increase is observed when the chains are grafted to a hydroxyl-modified silica surface. In a recent study, Sumerlin and coworkers [21] grafted a series of polymers onto gold surfaces. These polymers, the structures of which are shown in Figure 1.3, were all synthesized using aqueous reversible addition-fragmentation chain transfer (RAFT) free radical polymerization techniques. The RAFT technique allows for the controlled synthesis of polymers, copolymers, and block copolymers through the equilibrium between dormant and active species in which a dithioester compound acts as a chain transfer agent (Scheme 1.3). S P

A

+

S

kexAB

S C

S

P

PB

C

A

kexAB

ka kf

P

B

Z

Z Monomer

+

Monomer dA

B k frA

B

rB A

S P

A

k ad

S C

BA

PB

Z Scheme 1.3

General reversible addition-fragmentation chain transfer (RAFT) mechanism.

After the polymerizations were completed, the dithioester end-capped polymers were converted to thiol functionalities by addition of NaBH4 in water. This reaction was performed in the presence of the gold substrates so that conversion and surface grafting via the sulfur linkage was a one-step process. The surfaces were characterized using ATR-FTIR, AFM, and water contact angles measurements, which showed the generation of a more hydrophilic surface when compared to the native gold layer. Despite all of the investigations conducted in the field of “grafting to” brush synthesis, there remain inherent disadvantages in this process. As stated earlier, the grafting density of this process is limited by the diffusion of large polymer chains to the reactive sites on the substrate. The low surface coverage, (C, mg m–2) results in lower brush thicknesses (h) as given by Eq. (1) [18], where is the density of the polymer: C=hL

(1)

r = (6.023C L 100) / Mn

(2)

1.3 “Grafting From” Synthesis Technique

The grafting density (r, chains per nm2) is calculated from the surface coverage according to Eq. (2) [19], thus showing that the grafting density decreases as the Mn of the adsorbed chain increases for a constant C. By utilizing a “grafting from” approach, polymer brushes are generated in situ. Thus, rather than having bulky polymer chains diffuse to a reactive site, monomer units migrate to the growing polymer brush layer, resulting in denser and thicker polymer brush layers.

1.3

“Grafting From” Synthesis Technique

Unlike the situation in the “grafting to” technique, the substrate surface must be modified to generate the initiator functionality suitable for the polymer brush synthesis from a surface. This surface modification can be performed using LangmuirBlodgett techniques or SAM deposition. Furthermore, depending on the polymerization method, the initiator can be a free radical, ionic, ring-opening metathesis, or controlled radical polymerization type. By varying the substrate (gold, silicon, nanoparticles, etc.), initiator deposition technique, and polymer synthesis route, virtually limitless possibilities present themselves for brush formation. Biesalski and coworkers [22] used a free radical polymerization approach to synthesize a weak polyacid brush based on methacrylic acid. A SAM containing a monochlorosilane functional group and an azo initiator group (referred to as RNhe’s initiator) was first formed on silica wafers. The polymerizations were carried out in bulk and in 50 vol.% in water at 60 FC. The reaction times, which correlate to polymer conversion, were varied such that the maximum dry polymer brush thickness obtained was 400 nm. Biesalski and RNhe [23] used this same free radical polymerization approach to synthesize a poly(p-styrenesulfonate) brush using a p-styrenesulfonate ethyl ester. Rather than performing a sulfonation reaction to generate the desired polyelectrolyte like Tran and Auroy [15], Biesalski and RNhe performed a saponification. Furthermore, no degrafting was seen when forming the polyelectrolyte, and the brush thickness was as high as 35 nm, compared to roughly 15 nm for the “grafting to” technique. A similar type of azo free radical initiator was employed by Fan and coworkers [24] to synthesize polystyrene brushes from clay nanoparticles. Gold and silicon flat substrates were first cleaned and modified with a SAM to generate a positively charged surface. Montmorillonite clay particles were then deposited onto the surfaces prior to deposition of the azo initiator containing a quarternized amine head group. Upon 8 h polymerization in THF at 60 FC, an 8-nm polystyrene layer on the functionalized clay nanoparticles was observed. “Grafting from” the surface can also be performed using controlled polymerization techniques. By depositing the appropriate initiators, controlled polymerization can lead to uniform brush layers, tunable brush thicknesses via molecular weight control, and the ability to perform sequential polymerization steps to yield either thicker homopolymer layers or diblock copolymer layers. Several groups have performed ring-opening polymerizations (ROP) to generate various polymers from

41

42

1 Recent Advances in Polymer Brush Synthesis

both silica and gold substrates. Zhou [25], Advincula [26], and Quirk [27] have, separately, utilized this approach to synthesize polymer brushes from a 1,1-diphenylethylene (DPE) derivative. Detrembleur and coworkers [28], as well as Moon and Swager [29] and Harada et al. [30], employed a Grubbs catalyst for ring-opening metathesis polymerizations. By depositing a triethoxysilane containing an amine surface functionality, Wieringa et al. [31] were able to synthesize poly(L-glutamate) brushes from silica surfaces. ROP of the N-carboxyanhydrides of c-benzyl and c-methyl L-glutamates were performed in anhydrous DMF. MSller and coworkers [32] deposited a thiol SAM containing a hydroxyl-initiating end to conduct ROP of L-, D-, and L,D-lactides using a tin catalyst system. Unlike the glutamates, these polymerizations were performed in bulk. The work of both Wieringa and MSller consisted of brush re-initiation studies to prove the “living” characteristics of the process. Using a quarternized amine-anchoring group for the initiator, Zhou and coworkers [25] anionically polymerized styrene brushes from clay particles. Rather than first depositing the clay particles onto silica or gold, as did Fan [24], the initiator was directly deposited in an aqueous dispersion. The lithium-derivatized DPE initiator head group was then reacted with styrene, in benzene, to yield the desired brush. However, under similar conditions, polymer that was formed in solution had a lower PDI (polydispersity index) and larger Mn than that cleaved from the clay surfaces. Advincula et al. [26] used a similar DPE derivative initiator, but functionalized it with monochlorosilane and thiol so as to deposit on silica and gold wafers, respectively. From these surfaces, polystyrene brushes were synthesized in benzene with various additives in the reaction [THF, BuOLi, or tetramethylethylenediamine (TMEDA)]. The greatest water contact angle (94F) and brush thickness (23.4 nm) were obtained using s-BuLi to activate the DPE initiator and TMEDA as an additive. From this homopolymer surface, isoprene was polymerized to generate a PS-b-PI diblock copolymer. Typically, homopolymer and diblock copolymer thicknesses were no greater than 25 nm. Perhaps one of the most comprehensive “living” anionic brush synthesis reports was by Quirk and coworkers [27]. A series of homopolymers and diblock copolymers were synthesized via both “grafting from” and “grafting to” using a monochlorosilane DPE derivative. Homopolymer brushes of polystyrene were generated by reacting the DPE surface with poly(styryl)lithium via a “grafting to” technique, or they reacted the DPE surface with butyllithium followed by styrene for the “grafting from” brush. Diblock copolymers of PI-b-PEO were also generated using “grafting from” synthesis of isoprene followed by ethylene oxide when using n-butyllithium to activate the DPE monolayer. This diblock was also synthesized by grafting telechelic PI chains to the surface followed by in-situ polymerization of ethylene oxide from the hydroxyl-terminated PI brush. Grafting density and brush layer thickness was higher for the “grafting from” technique compared to the “grafting to” technique. Ring-opening metathesis polymerization (ROMP) has also been used to generate polymer brushes. Detrembleur and coworkers [28] first polymerized norbornenylmethylene acrylate (NBE-A) from a steel electrode surface using an electrografting process. The norbornenyl functionality was then reacted with a Grubbs’ initiator cat-

1.3 “Grafting From” Synthesis Technique

alyst to generate the ROMP initiator. Norbornene was then polymerized from the acrylate brush to form a thick polynorbornene layer. A similar ROMP of a norbornene-capped poly(p-phenylene ethynylene) macromonomer using a surface-bound Grubbs’ catalyst was performed by Moon and Swager [29] from silica wafers. Harada et al. [30] took an interesting approach to synthesizing polymer brushes of a norbornene derivative from the surface of silica wafers using a ROMP methodology. 7-Octenyltrichlorosilane was deposited on the wafer using a microcontact printing (lCP) approach to generate a SAM surface. This linear vinyl group was then reacted with a Grubbs’ catalyst to afford the appropriate ROMP surface catalyst. However, when subsequent polymerization of 2,2,2-trifluoroethyl bicyclo[2.2.1]hept2-ene-5-carboxylate was performed, low brush thickness was obtained. By lCP a 40:60 (mol%:mol%) 7-octenyltrichlorosilane to octyltrichlorosilane “primer” solution, the vinyl groups were sufficiently dilute along the surface that reactions of the neighboring ROMP surface catalyst was avoided. This primer solution resulted in the thickest polymer brush films in the study. By utilizing a “living” free radical approach, control over brush molecular weight and thickness can be obtained as well as relatively low polydispersity (brush thickness uniformity) and the ability to form diblocks. However, the conditions are more tolerant to impurities than the ionic systems, and some methods are capable of aqueous solution polymerizations. Since the polymerizations utilize a radical approach, the techniques have a wider range of polymerizable monomers available than the anionic or ROMP methods. Controlled radical polymerizations (CRP) can be conducted using iniferters, RAFT, nitroxide-mediated polymerization (NMP), and atom transfer radical polymerization (ATRP), as well as others. An iniferter molecule works on the premise that the molecule decomposes into a highly reactive radical capable of initiation as well as a rather stable counter radical, which predominantly acts as a chain transfer agent and terminating species. In the research performed by de Boer et al. [33], a photoiniferter (decomposition stimulated by UV irradiation) first used by Otsu [34] was deposited on silica wafers via a trimethoxy anchoring group. Synthesis of PS and PMMA homopolymer brushes was successful in addition to diblock copolymer synthesis of PS-b-PMMA. By etching the silica surfaces with chromium prior to initiator deposition, surface patterning was accomplished. Baum and Brittain [35] were able to synthesize both PMMA and PS brushes, in addition to PDMA brushes, utilizing a RAFT synthesis scheme from silica substrates. RNhe’s azo initiator was first deposited on silica surfaces, and 2-phenylprop2-yl dithiobenzoate was used as the RAFT chain transfer agent. 2,2¢-azobis isobutyronitrile (AIBN) was used in the polymerizations to act as free initiator for the polymerizations. This polymer generated in solution was found to correlate well with the degrafted brush chains when analyzed using gel-permeation chromatography. Along with the homopolymer brushes, diblock copolymer brushes of PS-b-PDMA and PDMA-b-PMMA were synthesized. Rather than synthesizing PDMA brushes, Hu and coworkers [36] polymerized N-isopropylacrylamide (NIPAM) using RAFT. Also, the brush formation was on the surface of spherical poly[NIPAM-co-acrylic acid(2-hydroxyethyl ester)] microgel particles. The surface-bound initiator used was a-butyl acid dithiobenzoate along with AIBN free initiator in THF solution.

43

44

1 Recent Advances in Polymer Brush Synthesis

kact PA

O

NR2

PA

kdeact

+

O

NR2

kp n M Scheme 1.4

General nitroxide-mediated polymerization (NMP) reaction scheme.

Rather than utilizing a chain transfer agent, nitroxide-mediated polymerization (NMP) decreases the propagating radical lifetime by reversibly coupling and decoupling from the chain end (Scheme 1.4). Parvole et al. [37] used this methodology to generate polyacrylate brushes on silica particles. First, an azo initiator was deposited on the silica surfaces using a triethoxy silane-anchoring site. From the surfacebound initiator, n-butylacrylate and ethylacrylate were separately polymerized from the surface in the presence of a monofunctional alkoxyamine or counter-radical/ AIBN (see Figure 1.4). AIBN was used to build up the concentration of initiator with respect to monomer to allow for a well-controlled system. CH3

CH3 CH3

O

O

N

CH3 CH3

CH3 O

CH3

CH3

O

H3C

H3C

H3C H3C

O O

O

P

O

H3C

CH3

P

N

O Et

Et

O Et

Et

Parvole's alkoxyamine initiator

Parvole's counter-radical control agent used with AIBN

H3C

CH3

O OEt O EtO

Si

N

O H3C

OEt

Devaux's surface NMP initiator

Figure 1.4 Examples of nitroxide-mediated polymerization (NMP) counter-radical chemical structures [37–39].

CH3

1.3 “Grafting From” Synthesis Technique

Devaux and coworkers [38,39] utilized a surface-bound TEMPO-type NMP initiator to graft styrene from the surface of both flat silica and AFM tips (Figure 1.4). For both studies, the initiator was deposited using a Langmuir-Blodgett technique in order to control the initiator layer density. After deposition, the surfaces were thermally treated to promote initiator self-condensation and chemisorption of the chains to the surface. Polymerizations of styrene were conducted from both surfaces using a 50/50 (v/v) solution of styrene in xylene at 130 FC. Free initiator, TEMPO, was used to establish the monomer/initiator reaction equilibrium ratio and effectively control the surface polymerization. A unique method of generating patterned polymer brushes was employed by von Werne et al. [40] and utilizes several different controlled radical polymerization techniques. In the first step, nanocontact printing and UV curing of a spin-coated network photopolymer resin is performed to generate a pattern on the substrate. The photopolymer contains an imbedded inimer (initiator/monomer single molecule), which is later used to graft brushes from the surface thereby increasing the feature size of the pattern. The two inimers used contained functionalities capable of NMP and ATRP controlled reactions (Figure 1.5). The NMP inimer was used to synthesize polystyrene brushes from the surface, while the ATRP inimer was applied in the synthesis of PS, PMMA, and poly(2-hydroxyethyl)methacrylate (PHEMA) brushes. Similar to NMP, atom transfer radical polymerization (ATRP) works on the premise of reversibly terminating growing chain ends. Unlike NMP, the capping group is not a stable free radical but rather a halogen group (labeled X in Scheme 1.5). The halogen is reversibly transferred from the deactivated chain end to a metal halide

H3C Br

CH3 CH3

CH3 N

H3C

H3C

O

O O

CH3

H3C

O O

O CH2 H3C

CH2 O CH3

ATRP inimer

NMP inimer

Figure 1.5 Controlled radical polymerization (CRP) inimers utilized by von Werne and coworkers [40].

45

46

1 Recent Advances in Polymer Brush Synthesis

+

Polymer-X

kact

Mtn-X/Ligand

+

Polymer kdeact

Mtn+1-X2/Ligand kterm dead polymer

kp Monomer General atom transfer radical polymerization (ATRP) reaction mechanism.

Scheme 1.5

catalyst. Typically, the metal catalyst is copper-based, but can also be nickel, ruthenium, and iron to name a few. As with NMP, the equilibrium for this reaction lies strongly towards the dormant/deactivated chain end, thereby lowering the radical concentration and minimizing termination. Mori and coworkers [41] used an a-bromoester ATRP surface-bound initiator to polymerize brushes from silica wafers. The monomers used were methyl methacrylate and tert-butylacrylate. Two inimers were also polymerized (BPEA and BIEM; Figure 1.6) in order to synthesize hyperbranched polymer brushes in a controlled manner. Also, as a means for decreasing the branching, the inimers were copolymerized with their linear analogues, acrylic or methacrylic monomers. O

CH3

O

O

H2C

CH3

O

Br O

O

Br

CH3

BPEA Figure 1.6

CH3

O

H2C

O

BIEM

Inimers utilized for hyperbranched polymer brush synthesis.

In the above investigations, sacrificial free initiator was used to set up the ATRP equilibrium and thereby lead to the controlled polymerization. In surface-bound ATRP, the reaction equilibrium can also be established by the addition of deactivator species (Mtn+1-X2 species in Scheme 1.5). The effects of such changes to the brush synthesis were studied by Jeyaprakash and coworkers [42]. Polystyrene brushes were grown from silicon surfaces using various [styrene]0/[CuBr]0 ratios and using CuBr2. The thickest polymer brushes, when comparing the same polymerization times, were obtained at lower ratios. These results were then compared to those from experiments using sacrificial initiator instead of CuBr2 in the reaction mixture. For each trial, the corresponding deactivator system led to higher film thicknesses than the sacrificial initiator analogue. However, the use of a sacrificial initiator leads to the generation of free polymer, which has been shown to correlate well with the molecular weight of the polymer brush [43]. Using a sacrificial initiator, Ohno et al. [44] synthesized PMMA brushes from gold nanoparticles. The sacrificial initiator used was ethyl 2-bromoisobutyrate

1.3 “Grafting From” Synthesis Technique

Cl OH

+

Cl

H3C Si

CH3

dry toluene

O

60 oC, 4 h

Br

CH2

Cl

silica substrate

silica substrate

(E2BriB), which was of the same initiating species as the surface-bound initiator. Since the anchoring group to the gold nanoparticle consisted of thiol groups, which are known to be thermally unstable, the polymerization reaction was performed at 40 FC. Jones and Huck [45] utilized a similar ATRP initiator that was deposited on gold through a microcontact printing process to generate a patterned surface. Polymerization of various acrylate monomers (MMA, glycidyl methacrylate, n-butylacrylate, and HEMA) was conducted under aqueous (water/methanol) ATRP conditions. Diblock copolymer brushes of PMMA-b-PHEMA were also synthesized. ATRP, being a radical process, is tolerant to impurities and solvents; aqueous polymerizations are also possible and have even been shown to be faster than their organic solvent counterparts. This group has extended this aqueous ATRP synthesis process to generate polyelectrolyte [1] and thermoresponsive poly(N-isopropylacrylamide) [46] brushes. More recently, Kizhakkedathu and Brooks [47] performed aqueous ATRP to form poly(N,N-dimethyacrylamide) (PDMA) from PS latex particles. The particles contained a chlorinated inimer (chlorinated version of BPEA; Figure 1.6) to initiate the CuCl-catalyzed ATRP reaction. Boyes and coworkers [4] synthesized ABA-type triblock copolymers from silicon wafers. Using a bromoisobutyrate initiator bound to the surface as a SAM, PS-bPMA-b-PS and PMA-b-PS-b-PMA triblock copolymer brushes were synthesized.

11

O

O O

H3C

Si

CH3

O CH2

O

Br 11

O

styrene, anisole CuBr, PMDETA

O O

H3C

Si

CH3

methyl acrylate, anisole

O CH2

O

CH2 CH

Br

CuBr, PMDETA

n

11

O

90-100 oC, 24h

silica substrate

silica substrate

90-100 oC, 24h

O O

H3C

Si

CH2 CH

CH2 O

CH3

O CH2 CH n

11

O

styrene, anisole CuBr, PMDETA silica substrate

90-100 oC, 24h O O

H3C

CH2 CH

CH2 O

CH3

O

Si

m

p

O O

Scheme 1.6

Br

CH2 CH

CH2 CH n

11

Br m

O

OMe

Synthesis of ABA triblock copolymer brush, PS-b-PMA-b-PS.

OMe

47

48

1 Recent Advances in Polymer Brush Synthesis

Scheme 1.6 depicts the general triblock copolymer synthesis reactions to form the PS-b-PMA-b-PS brush. The synthesis was characterized using ellipsometry to measure the sequential brush thickness and ATR-FTIR spectroscopy. Finally, Balamurugan et al. [48] synthesized poly(N-isopropylacrylamide) brushes on gold substrates using ATRP in DMF. The metal halide was a CuBr system, and no sacrificial initiator or deactivator metal was added to the reaction solution. Furthermore, the ATRP reaction was carried out at room temperature and resulted in thick (~51 nm) polymer brushes. The surfaces were further characterized using contact angle measurements and surface plasmon resonance in addition to the thermal response studies on the lower critical solution temperature transitions of the polymer.

Summary

In summary, a number of recent advances in the field of polymer brush synthesis have been described in this chapter. Whether utilizing a “grafting to”, “grafting from”, or a combination of the two approaches, polymer brushes are being formed of various sizes, structures, and compositions. Although they are formed predominantly on either gold or silicon substrates, this field is constantly expanding to introduce such brushes onto latex particles, PDMS films, carbon nanotubes, and a wealth of other materials. With the ability to polymerize these materials by using any process, the possible combinations of substrate, synthesis approach, and polymerization route are providing numerous opportunities.

References

References 1 V. L. Osborne, D. M. Jones, W. T. S. Huck, 2 3 4

5 6 7 8 9 10 11 12 13 14

15 16

17

18

19

20

21

Chem. Commun. 2002, 1838–1839. X. Kong, T. Kawai, J. Abe, T. Iyoda, Macromolecules 2001, 34, 1837–1844. Y. Ito, S. Nishi, Y. S. Park, Y. Imanishi, Macromolecules 1997, 30, 5856–5859. S. G. Boyes, W. J. Brittain, X. Weng, S. Z. D. Cheng, Macromolecules 2002, 35, 4960–4967. M. Mayes, S. K. Kumar, MRS Bull. 1997, 22 (1), 43–47. S. T. Milner, Science 1991, 251, 905–914. B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677–710. G. F. Belder, G. ten Brinke, G. Hadziioannou, Langmuir 1997, 13, 4102–4105. P. Mansky, Y. Liu, E. Huang, T. P. Russell, C. Hawker, Science 1997, 275, 1458–1460. O. Prucker, J. RNhe, Macromolecules 1998, 31, 592–601. I. Potemkin, A. R. Khokhlov, P. Reineker, Eur. Phys. J. E: Soft Matter 2001, 4, 93–101. R. Choi, H. K. Cho, H. A. Chun, S. T. Noht, Polymer(Korea), 2000, 24, 488–498. S. Qin, K. Matyjaszewski, H. Xu, S. S. Sheiko, Macromolecules 2003, 36, 605–612. Matyjaszewski, S. Qin, J. R. Boyce, D. Shirvanyants, S. S. Sheiko, Macromolecules 2003, 36, 1843–1849. Y. Tran, P. Auroy, J. Am. Chem. Soc. 2001, 123, 3644–3654. S. M. Sirard, R. R. Gupta, T. P. Russell, J. J. Watkins, P. F. Green, K. P. Johnston, Macromolecules 2003, 36, 3365–3373. S. Minko, S. Patil, V. Datsyuk, F. Simon, K.-J. Eichhorn, M. Motornov, D. Usov, I. Tokarev, M. Stamm, Langmuir 2002, 18, 289–296. Luzinov, D. Julthongpiput, H. Malz, J. Pionteck, V. V. Tsukruk, Macromolecules 2000, 33, 1043–1048. S. Minko, M. MNller, M. Motornov, M. Nitschke, K. Grundke, M. Stamm, J. Am. Chem. Soc. 2003, 125, 3896–3900. H. Maas, M. A. Cohen Stuart, A. B. Sieval, H. Zuilhof, E. J. R. Sudholter, Thin Solid Films 2003, 426, 135–139. B. S. Sumerlin, A. B. Lowe, P. A. Stroud, P. Zhang, M. W. Urban, C. L. McCormick, Langmuir 2003, 19, 5559–5562.

22 Biesalski, D. Johannsmann, J. RNhe, J. Chem.

Phys. 2002, 117, 4988–4994. 23 Biesalski, J. RNhe, Macromolecules 2003, 36,

1222–1227. 24 X. Fan, C. Xia, T. Fulghum, M.-K. Park,

J. Locklin, R. C. Advincula, Langmuir 2003, 19, 916–923. 25 Q. Zhou, X. Fan, C. Xia, J. Mays, R. Advincula, Chem. Mater. 2001, 13, 2465–2467. 26 R. Advincula, Q. Zhou, M. Park, S. Wang, J. Mays, G. Sakellariou, S. Pispas, N. Hadjichristidis, Langmuir 2002, 18, 8672–8684. 27 R. P. Quirk, R. T. Mathers, T. Cregger, M. D. Foster, Macromolecules 2002, 35, 9964– 9974. 28 C. Detrembleur, C. JWrXme, M. Claes, P. Louette, R. JWrXme, Angew. Chem. Int. Ed. 2001, 40, 1268–1271. 29 H. Moon, T. M. Swager, Macromolecules 2002, 35, 6086–6089. 30 Y. Harada, G. S. Girolami, R. G. Nuzzo, Langmuir 2003, 19, 5104–5114. 31 R. H. Wieringa, E. A. Siesling, P. F. M. Geurts, P. J. Werkman, E. J. Vorenkamp, V. Erb, M. Stamm, A. J. Schouten, Langmuir 2001, 17, 6477–6484. 32 MSller, F. Nederberg, L. S. Lim, R. KYnge, C. J. Hawker, J. L. Hedrick, Y. Gu, R. Shah, N. L. Abbot, J. Polym. Sci., Polym. Chem. Ed. 2001, 39, 3529–3538. 33 B. de Boer, H. K. Simon, M. P. L. Werts, E. W. van der Vegte, G. Hadziioannou, Macromolecules 2000, 33, 349–356. 34 T. Otsu, M. Yoshida, Makromol. Chem. Rapid Commun. 1982, 3, 127–140. 35 Baum, W. J. Brittain, Macromolecules 2002, 35, 610–615. 36 T. Hu, Y. You, C. Pan, C. Wu, J. Phys. Chem. B 2002, 106, 6659–6662. 37 J. Parvole, L. Billon, J. P. Montfort, Polym. Int. 2002, 51, 1111–1116. 38 C. Devaux, J. P. Chapel, E. Beyou, Ph. Chaumont, Eur. Phys. J. E 2002, 7, 345–352. 39 C. Devaux, J. P. Chapel, Eur. Phys. J. E 2002, 10, 77–81. 40 T. A. von Werne, D. S. Germack, E. C. Hagberg, V. V. Sheares, C. J. Hawker, K. R. Carter, J. Am. Chem. Soc. 2003, 125, 3831–3838. 41 H. Mori, A. BSker, G. Krausch, A. H. E. MNller, Macromolecules 2001, 34, 6871–6882.

49

50

1 Recent Advances in Polymer Brush Synthesis

42 J. D. Jeyaprakash, S. Samuel, R. Dhamod-

46 D. M. Jones, J. R. Smith, W. T. S. Huck,

haran, J. RNhe, Macromol. Rapid Commun. 2002, 23, 277–281. 43 Ejaz, Y. Tsujii, T. Fukuda, Polymer 2001, 42, 6811–6815. 44 K. Ohno, K. Koh, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 8989–8993. 45 D. M. Jones, W. T. S. Huck, Adv. Mater. 2001, 13, 1256–1259.

C. Alexander, Adv. Mater. 2002, 14, 1130– 1134. 47 J. N. Kizhakkedathu, D. E. Brooks, Macromolecules 2003, 36, 591–598. 48 S. Balamurugan, S. Mendez, S. S. Balamurugan, M. J. O’Brien II, G. P. LZpez, Langmuir 2003, 19, 2545–2549.

51

2

Polymer Brushes by Atom Transfer Radical Polymerization Jeffrey Pyun, Tomasz Kowalewski, and Krzysztof Matyjaszewski

Abstract

Atom transfer radical polymerization (ATRP) is a robust method for the preparation of well-defined (co)polymers. This process has also enabled the preparation of a wide range of molecular brushes where (co)polymers of precise molar mass, composition, and architecture are covalently attached to either curved or flat surfaces. In this chapter, the general methodology for the synthesis of polymer brushes from flat surfaces, polymers and colloids is summarized, focusing on reports using ATRP. Additionally, the nanoscale structure and organization of ultrathin films formed by polymer brushes is discussed based on their characterization with atomic force microscopy (AFM) and other techniques.

2.1

Introduction

Polymer brushes are defined as dense layers of chains confined to a surface or interface where the distance between grafts is much less than the unperturbed dimensions of the tethered polymer. Due to the high steric crowding, grafted chains extend from the surface, thus residing in an entropically unfavorable conformation. Polymer brushes have been prepared by the end-grafting of chains to flat, or curved surfaces that are either organic, or inorganic in nature. These include functional colloids, highly branched polymers and block copolymer aggregates, such as micelles or phase-separated nanostructures [1,2]. This chapter focuses on polymer brushes that have been synthesized using atom transfer radical polymerization (ATRP). Since initial reports of this work in 1998 from flat silicon wafers [3], surface-initiated ATRP has also been performed from flat gold surfaces [4–8], inorganic particles/colloids, organic latexes [9–16], nanopatterned networks [17], dendrimers [18–22], and highly functional linear polymers [23–31] (Scheme 2.1). The characterization of nanoscale features in these materials was performed using various techniques. Atomic force microscopy (AFM) proved to be a particularly useful tool, owing to its ability to visualize individual macromole-

52

2 Polymer Brushes by Atom Transfer Radical Polymerization

cules [32,33] and phase-separated domains. Here, we focus on developments from our own investigations to discuss key steps in preparing brushes on a wide range of substrates and dimensions. Complementary general information on polymer brushes can be found in recent reviews, which cover also the use of other controlled/living radical processes, such as nitroxide mediated polymerizations (NMPs) [34] and reversible addition-fragmentation transfer polymerization (RAFT) [35].

Scheme 2.1 Examples of polymer brushes synthesized by ATRP using the “grafting from” approach from various functional substrates such as flat wafers, particles, colloids, and polymers (X = halogen).

2.2

Polymer Brushes on Flat Surfaces

Dense brush layers on flat wafers and surfaces are among the most extensively investigated systems, owing to potential applications in advanced microelectronics and biotechnology. Brushes on flat surfaces provide an excellent system to combine both “top-down” and “bottom-up” approaches, since the nanoscale organization and functionality of tethered polymers can be directed by techniques such as photolithography [36], micro- [37] and nano-contact [17] printing. Fundamental investigations into the parameters affecting surface-initiated ATRP have been focused on exploring the conditions for controlling film thickness, functionality and properties. Recent studies have also demonstrated the ability to control grafted polymer architecture ranging from tethered linear and hyperbranched [38] polymers to network crosslinked films [5]. In the characterization of these systems, techniques such as ellipsometry, contact angle, X-ray photoelectron spectroscopy (XPS) and AFM are

2.2 Polymer Brushes on Flat Surfaces

central in assessing whether tethered (co)polymers obtained from surface-initiated ATRP possess precise molar mass and composition. 2.2.1

Controlled ATRP from Flat Surfaces

As pointed out previously, the main challenge in ATRP from flat surfaces with very low concentrations of initiating groups stems from the fact that after halogen atom transfer to the transition metal catalyst, the concentration of persistent radical (deactivator) may be too low to reversibly trap the propagating radicals, leading to uncontrolled chain growth. This challenge can be effectively addressed through the addition of persistent radical (deactivator), or “sacrificial initiator” at the beginning of the reaction. Addition of Persistent Radical (Deactivator) As predicted from the Persistent Radical Effect [39–41], the addition of radical deactivating complexes (e.g., Cu(II), Fe(III), etc. halides) at the beginning of the reaction facilitates exchange reactions between active radicals and dormant oligo/polymeric halides. The ATRP of styrene (S) and methyl acrylate (MA) in the presence of externally added Cu(II) complexes resulted in a progressive increase in brush film thickness with time, as determined using ellipsometry [42]. Systematic studies of the effect of both Cu(I) and Cu(II) concentrations on surface-initiated ATRP from flat wafers have been conducted. In a study conducted by Baker and Bruening et al. [8], the kinetics of MA ATRP from Au-coated wafers was investigated by precise variation of the concentration of Cu(I) complexes from 40 mM to 2.5 G 10–4 mM and determining its effect on the controlled brush growth. Optimal conditions for sustained increases in brush thickness were achieved at a Cu(I) concentration of 0.1 mM. In these experiments, the MA concentration was held at 2 M (solution in acetonitrile), and all catalyst systems used tris[2-(dimethylamino)ethyl]amine as the ligand, with the Cu(II) concentration kept at 30 mol% relative to complexes [43]. These results indicate that excessive termination reactions of surface-bound radicals can still occur in the presence of deactivator, if the overall Cu(I) concentration is too high. Thus, controlled growth can also be achieved by dilution of the catalyst, where the concentration of the monomer is varied, as shown by Wirth et al. [44], for the ATRP of acrylamides from Si wafers. 2.2.1.1

2.2.1.2 Addition of “Sacrificial Initiator” The addition of untethered small molecule initiators to ATRP mixtures with functional flat substrates serves a number of beneficial purposes in both synthesis and characterization of polymer brushes, as demonstrated by Fukuda et al. [3], and Hawker et al. [45]. In systems with added free initiator, sufficient concentrations of persistent radical (deactivator) are generated by the termination of radicals formed in solution. Furthermore, the final degree of polymerization (DP) of tethered chains on surfaces can be dictated by the concentration of sacrificial initiator added at the initial stages of the polymerization. The determination of both monomer conversion

53

54

2 Polymer Brushes by Atom Transfer Radical Polymerization

and molar mass of polymers in the system is also greatly facilitated, as an analysis of free polymers formed in solution can be performed by using standard techniques, such as 1H NMR, gas chromatography (GC), and size-exclusion chromatography (SEC). An important fundamental question in these systems is whether tethered polymers on surfaces possess similar molar masses and polydispersity to polymers formed from sacrificial initiators. Although the molar masses of tethered polymers from brushes and free polymers prepared by ATRP in the presence of sacrificial initiator have not yet been compared, SEC analysis of cleaved polymers from brushes prepared using NMPs showed that cleaved chains from surfaces had similar molar masses and polydispersities to polymers formed in solution [46]. 2.2.1.3 Effect of Initiator Coverage The conformation of tethered chains and film thickness in polymer brush layers are directly affected by surface coverage of initiating groups on the flat surface. Ellipsometry measurements in the dry state of polymer brushes on Si wafers revealed that silane coupling agents with long spacers yielded brushes with a lower grafting density of tethered chains. The use of Langmuir-Blodgett techniques enabled the deposition of tightly packed monolayers of initiators onto native oxide of Si wafers which, after ATRP of methyl methacrylate (MMA), formed brushes with a very high grafting density and caused tethered chains to reside in a highly extended conformation [3]. Controlled variation of initiator coverage of flat surfaces has been conducted by photodecomposition of a dense functional monolayer [47] or by dilution of grafted initiators via blending with coupling agents inactive in ATRP [48,49]. From these approaches, polymer brushes with tunable film thickness were synthesized. The effect of varying initiator concentrations and grafting density on the conformation of tethered polymers has been investigated using a number of characterization techniques to study the conformation and properties of brushes, both in solvent-swollen and dry states [47,50–52]. As discussed in Section 2.2.3, effects similar to those associated with a low concentration of tethered initiator can be achieved by grafting polymers onto various substrates, to provides a lower density of surface coverage. 2.2.2

Block Copolymer Brushes on Flat Surfaces

The preparation of block copolymer brushes using a “grafting from” ATRP approach was first reported by tethering polystyrene-block-poly(t-butyl acrylate) (pS-b-ptBA) to Si wafers. Hydrolysis of the t-butyl groups yielded a polystyrene-block-poly(acrylic acid) brush, and demonstrated a versatile approach to tune film properties and wettability [42]. Novel block copolymer brushes have been synthesized using sequential ATRP from wafers possessing amphiphilic di- and triblock copolymers tethered to various substrates (e.g., Si, Au) [6,7,53]. The synthesis of stimulus-responsive brushes has also been achieved by the immobilization of di- and triblock copolymers via sequential surface-initiated polymerizations [54–59]. The ability to control conformation and rearrangement of teth-

2.2 Polymer Brushes on Flat Surfaces

ered block copolymers has been shown directly to affect both the morphology and properties of the ultrathin film. As first demonstrated by Brittain and Cheng et al. [56], the morphology of pS-b-pMMA block copolymer brushes can be reversibly controlled by treatment with selective solvents, such as dichloromethane, or dichloromethane-cyclohexane mixtures. This work has also been extended to triblock copolymer brushes composed of tethered pS-b-pMA-b-pS and pMA-b-pS-b-pMA copolymers [59]. A key lesson learned from these solvent-selective ultrathin films is that in order to exhibit distinct morphological differences due to stimulus-driven rearrangement, in addition to appropriate compositions, copolymer brushes must possess sufficient conformational freedom. This important observation prompted us to design a stimulus-responsive brush utilizing a “grafting to” approach of a functional ABC triblock copolymer. Due to the lower grafting densities inherent to this synthetic method, tethered copolymers retained adequate degrees of freedom enabling segment selective reactions to various solvents. 2.2.3

Stimuli-Responsive Ultrathin Films from “Grafting To” Approach

The synthesis of a reactive ABC triblock copolymer precursor was conducted with a combination of living anionic ring-opening and atom transfer radical polymerizations [60,61]. In the first step, a polydimethylsiloxane (pDMS) macroinitiator (Mn SEC = 6200; Mw/Mn = 1.19) was prepared by the living anionic ring-opening polymerization and hydrosilation of the silane end-group with an alkyl halide functional alkene. Chain extension from the pDMS macroinitiator using ATRP yielded a pDMS-b-pS diblock copolymer (Mn = 66 730; Mw/Mn = 1.38). A final ATRP step using the diblock copolymer macroinitiator with 1-(dimethoxymethylsilyl)propyl acrylate (DMSA) yielded a triblock copolymer (Mn NMR pDMS-b-pS-b-pDMSA = 156 700) capable of covalent bonding to a silicon wafer while also containing rubbery and glassy segments (Scheme 2.2). Surface properties of the copolymer ultrathin films could be reversibly controlled to present either pS or pDMS segments by treatment with toluene, or a mixture of toluene/hexane. Tapping-mode AFM observations revealed that the surfaces of brush films immersed in a good solvent for both segments (toluene) and subsequently dried under a stream of nitrogen exhibited fractal morphology characteristic for glassy polymers. This result indicated that pS segments were predominantly presented at the surface. In contrast, the surfaces of brush films exposed to toluene/ hexane mixtures and dried under nitrogen, were completely featureless under ultralight tapping conditions (A/Ao ! 1, where A and Ao denote, respectively the setpoint and “free” cantilever oscillation amplitude). This suggested that, following treatment with the solvent of lower affinity towards pS, “soft” pDMS segments were preferentially segregating to the surface. Upon switching to “normal” tapping conditions (A/Ao = 0.9), the globular morphology, consistent with the presence of collapsed glassy pS domains under the PDMS layer, was observed (Figure 2.1).

55

56

2 Polymer Brushes by Atom Transfer Radical Polymerization

Tapping-mode AFM height images of pDMS-b-pS-bpDMSA brushes after the following treatments. (left) After immersion in toluene and drying with nitrogen; (right) after immersion in toluene, gradual addition of hexane and drying with nitrogen [61].

Figure 2.1

2.3 Polymer Brushes from Particles

Scheme 2.2 Synthesis of ABC triblock copolymer of pDMS-b-pS-b-pDMSA using living anionic polymerization and ATRP. PDMS macroinitiator (Mn SEC = 6200; MW/Mn = 1.19) chain

extended with Sty using ATRP. PDMS-b-pS (Mn SEC = 66 730; MW/Mn = 1.38) chain extended with DMSA to prepare surface reactive triblock copolymer (Mn NMR = 156 700) [61].

2.3

Polymer Brushes from Particles

The synthesis of brushes on particle surfaces has been widely conducted to prepare solid supports [62,63], chromatographic stationary phases [64–67], or high surface area models for brushes on flat surfaces [45,68–70]. Brushes were grown by surfaceinitiated ATRP, from organic latex colloids emulsion [9,12–14,71–73] shells of shellcrosslinked micelles [74,75], and from functionalized inorganic particles, such as silica [38,44,45,64,65,68–70,76–86], gold [87,88], alumina [89], polysilsesquioxane [90],

57

58

2 Polymer Brushes by Atom Transfer Radical Polymerization

titanium oxide clusters [91,92], iron oxides [93] and germanium [94]. Herein, we focus on the ATRP synthesis of spherical brushes from functionalized nanoparticles serving as colloidal initiators. 2.3.1

Spherical Brushes from Inorganic Colloids

Organic/inorganic hybrid nanoparticles containing an inorganic core and tethered glassy or rubbery homopolymers have been prepared by the ATRP of styrene and (meth)acrylates from colloidal initiators. One particularly interesting aspect of such spherical brushes is the possibility to achieve a high grafting density of polymer chains owing to the high curvature of nanoparticle surfaces. In addition, the use of preformed, colloidal initiators opens the way to the systematic control of the brush length by changing the ratio of monomer to initiating sites ([M]o/[I]o), while maintaining a core of static dimensions. In the synthesis of silica-graft-polystyrene (SiO2g-pS) spherical brushes, the ATRP of S was performed using 2-bromoisobutyrate functional silica colloids targeting varying DP of tethered chains. SiO2-g-pS hybrid nanoparticles possessing molar masses of tethered pS in the range of Mn = 5000 to 33 000 g mol–1 were prepared and characterized both in the solid state and in solution using transmission electron microscopy (TEM) and dynamic light scattering (DLS), respectively. TEM images of SiO2-g-pS colloids revealed the formation of (sub)monolayer patches with interparticle spacing increasing with the increase of tethered pS molar mass (Figure 2.2). Comparison of hydrodynamic radii (Rh) for hybrid nanoparticles of varying size determined by DLS in toluene, versus molar masses (Mn) of pS chains cleaved from colloids determined by SEC, revealed a linear relationship (Figure 2.2). Such linear dependence of Rh versus Mn is a strong indication that when the particles are dispersed in toluene, the tethered chains adopt highly extended conformations, presumably due to steric interactions caused by the high grafting density [85]. 2.3.2

Multilayered Core-Shell Colloids

Block copolymers containing both glassy and rubbery segments grafted to inorganic core by consecutive ATRP of styrene and (meth)acrylates (Scheme 2.3) provide an example of multilayered core-shell colloids. Polysilsesquioxane-graft-(polystyreneblock-poly(benzyl acrylate)) (SiO1.5-g-(pS-b-pBzA)) hybrid nanoparticles with approximately 1000 grafted chains were prepared by ATRP [90]. Spherical block copolymer brushes were also prepared from 2-bromoisobutyrate functional silica colloids (Deff = 20 nm) using various combinations of S, MMA and n-butylacrylate (BA) [95]. For block copolymer brushes from SiO1.5 or SiO2 cores, destruction of the interior and recovery of cleaved copolymers enabled evaluation of molar mass and polydispersity using SEC. Cleaved blocks from both SiO1.5-g-(pS-b-pBzA) and SiO2-g-(pBAb-pMMA) had higher polydispersities than macroinitiator precursors (Mn pS-b-pBzA = 26 760, Mw/Mn = 1.48; Mn pBA-b-pMMA = 14 190, Mw/Mn = 3.10). Higher polydispersi-

2.3 Polymer Brushes from Particles

Figure 2.2 Left: Transmission electron microscopy (TEM) images of SiO2 colloidal initiator and SiO2-g-PS hybrid nanoparticles (Mn, tethered pS = 5230, 14 960, 32 670 g mol–1), black bar =

100 nm. Right: Plot of hydrodynamic radius (Rh in toluene, 25 AC) of SiO2-g-PS versus molar mass of cleaved pS determined by size-exclusion chromatography [85].

ties obtained in sequential surface-initiated ATRP originate from intramolecular termination of neighboring tethered (oligo)polymeric radicals. However, due to the large number of tethered chains per particle (approximately 1000 chains), a significant retention of active macroinitiator sites enabled incorporation of well-defined

59

60

2 Polymer Brushes by Atom Transfer Radical Polymerization

Scheme 2.3 General methodology for the preparation of core-shell colloids. Sequential ATRP of selected monomers from functional particles, or colloids.

outer blocks to colloidal initiators possessing tethered homopolymers. In SEC chromatograms, terminated tethered chains formed low molar mass tails, which comprised less than 15 wt.% of the entire molar mass distribution. SEC peaks and corresponding molar masses of the main product were as follows: Mn peak pS-b-pBzA = 34 810, Mw/Mn = 1.22; Mn peak pBA-b-pMMA = 39 130, Mw/Mn = 1.29. a)

b)

Figure 2.3 (a) Tapping-mode AFM height image of SiO2-g-(pBA94-bpMMA352) core-shell colloid with rubbery inner segment and glassy outer shell. (b) Height image of same region as for (a) SiO2-g-(pBA94-bpMMA352) ultrathin

films imaged with increasing applied tapping force. Regions of different contrast assigned to SiO2 core (bright spots), pBA inner segment (dark halo around SiO2) and pMMA segment (continuous matrix).

2.3 Polymer Brushes from Particles

The core shell structure of these colloids was reflected in AFM images, owing to different mechanical compliances of blocks. Variable-force tapping-mode AFM images of monolayer patches formed by SiO1.5-g-(pS-b-pBzA) nanoparticles revealed the presence of rigid protrusions (inorganic cores surrounded by pS) dispersed uniformly in a soft matrix formed by outermost pBzA blocks [90]. In addition, phase images revealed that the innermost regions of rigid protrusions were surrounded by “halos”, which were attributed to the pS phase. Phase contrast between rigid inorganic cores and glassy pS was most likely facilitated by the higher mechanical lossiness of the latter. Further studies were conducted with SiO2-g-(pBA-b-pMMA) systems where a rubbery pBA segment (DPn = 94) was inserted between the rigid core (Deff = 20 nm) and glassy pMMA segment (DPn = 352) [95]. Height and phase tapping-mode AFM images of SiO2-g-(pBA94-b-pMMA352) ultra-thin films revealed the presence of distinct spherical protrusions with center-to-center spacing of about 70 nm, embedded in a continuous matrix (Figure 2.3(a)). Change of appearance of height images upon switching from light tapping (A/Ao ! 1) to harder tapping (A/Ao = 0.78), indicated that the protrusions consisted of rigid cores (SiO2 particles), surrounded by the well-defined zone of more deformable material (pBA), and embedded in a less deformable matrix (pMMA) (Figure 2.3(b)). 2.3.3

Imaging of Individual Spherical Brushes

Upon casting from sufficiently dilute solutions, spherical brushes could be deposited on surfaces as individual nano-objects, which could be further studied in detail using tapping-mode AFM. In general, the AFM images revealed that upon deposition of individual particles on the surface, the tethered chains collapsed around the core forming distinct “coronas”. In some cases, the peripheral chains of these coronas could be imaged with molecular resolution, reminiscent of molecularly resolved images of molecular brushes. 2.3.3.1 Spherical Brushes with Polar Homopolymer Shells Cast onto Mica Tapping-mode AFM images of individual spherical brushes composed of SiO2-gpMMA175 (Mn cleaved pMMA = 17 560; Mw/Mn = 1.30) with silica cores of average diameter Deff = 20 nm are shown in Figure 2.4. Samples for imaging were prepared by spin-coating a dilute polymer solution in chloroform onto mica. Spherical cores with diameters ranging from 20 to 40 nm are surrounded by well-defined coronas of pMMA. Although individual pMMA chains were not clearly resolved in this case, the frayed edges of the coronas corresponded most likely to bundles of glassy pMMA. Figure 2.4 illustrates also an important aspect related to the synthesis of spherical brushes from colloidal inorganic particles: the proximity of the cores strongly suggests that the particles shown in this image are linked together. Such inter-particle crosslinking is generally undesirable, and most likely occurs when the synthesis is carried out at too high concentration of colloidal inorganic precursors. SiO2-g-pBA140 spherical brushes (Mn tethered pBA = 17 990; Mw/Mn = 1.28) of comparable size of molar mass were spin-coated onto a mica surface and also imaged using

61

62

2 Polymer Brushes by Atom Transfer Radical Polymerization

Figure 2.4 Tapping-mode AFM height (left) and phase (right) images of SiO2-gpMMA175 spherical brushes.

Figure 2.5 Tapping-mode AFM height (left) and phase (right) images of SiO2-g-pBA hybrid nanoparticle. SEC of cleaved pBA indicated Mn = 17 990; Mw/Mn = 1.28.

tapping-mode AFM. Similar to images of pBA molecular brushes, individual colloids with a corona of tethered pBA extending on the mica substrate were observed in both height and phase AFM images (Figure 2.5). In AFM height images, silica cores were discernable as tall features (white spots) surrounded by tethered pBA. Strikingly different features were observed in phase images, where silica cores (bright central spot) were embedded in a dense shell of collapsed pBA (dark corona), followed by extension of single chains from the outer edge of the pBA corona (dark outer shell). Spherical Brushes with Polar Block Copolymer Shells Cast onto Mica Block copolymer brushes of SiO2-g-(pBA94-b-pMMA352) ultra-thin films which were described in the previous section were also visualized at the level of individual isolated particles. As shown previously in Figure 2.3, AFM images of SiO2-g-(pBA94-bpMMA352) ultrathin films revealed clearly discernable domains of more compliant pBA sandwiched between the rigid SiO2 core and pMMA shell. Three distinct zones were also observed in AFM images of individual SiO2-g-(pBA94-b-pMMA352) particles 2.3.3.2

2.4 Molecular Brushes

Figure 2.6 Tapping-mode AFM height (left) and phase (right) image of an individual SiO2-g-(pBA94-b-pMMA352) core-shell nanoparticle.

deposited on mica by spin-coating from dilute chloroform solution (Figure 2.6). An interesting feature observed for this system was the spreading of outer pMMA chains on the mica surface. In contrast to both SiO2-g-pMMA (Figure 2.4) and SiO2g-pBA (Figure 2.5) brushes, the outer pMMA segments had a distinct “blobby” appearance, which could be related to local collapse of pMMA segments. 2.3.4

Modification of Carbon Black Fillers

Carbon black is an industrially important filler which is composed of primary particles with diameters of 10 to 75 nm that are fused into aggregates that range from 50 to 500 nm in size [76,96,97]. The synthesis of carbon black grafted with pBA chains was conducted via surface-initiated ATRP from functional carbon black surfaces [98]. An analysis of pBA chains cleaved from the cores indicated that polymers with tunable molar mass and low polydispersity (Mn = 48 450; Mw/Mn = 1.33) were obtained. AFM of pBA-coated carbon black particles cast onto mica confirmed the efficient grafting of pBA to the central core, with morphologies consistent with those observed for other pBA spherical brushes.

2.4

Molecular Brushes

For the purpose of this review, the definition of molecular brushes is limited to densely grafted copolymers bound to either a linear or dendritic backbone. Due to the high local concentration of tethered chains, grafted chains extend away from the polymeric backbone, despite the conformation of flexibility that it may exhibit. ATRP has proved to be a valuable polymerization technique for the preparation of molecular brushes, as both the composition and DP of the backbone and side chains can be precisely controlled (Scheme 2.4).

63

64

2 Polymer Brushes by Atom Transfer Radical Polymerization

Scheme 2.4

General approach to densely grafted copolymers by ATRP.

2.4.1

Synthesis of Molecular Brushes from Linear Polymeric Macroinitiators

Current efforts towards the synthesis of molecular brushes from densely grafted copolymers are focused on the ATRP of various (meth)acrylates, acrylamides and styrene from linear macroinitiators that carry an alkyl halide moiety at each repeat unit. Due to the high degree of functionality and local concentration of initiating groups on the macroinitiators, ATRP in these systems usually requires the use of low catalyst concentrations, the external addition of deactivator at the initial stages, and long reaction times in order to avoid termination reactions (as discussed in Section 2.2).

2.4 Molecular Brushes

Systematic investigations of structural variations in molecular brushes have been pursued to prepare a wide range of nanostructured materials. Variation of DP in either backbone (DP = 100 to 4000) and side chains (DP = 10 to 100) has been conducted to prepare either very long brushes spanning several hundreds of nanometers, or very “hairy” structures with long extensions from a dense core. The availability of a range of monomers (e.g., S, acrylamides, alkyl (meth)acrylates, poly(ethylene oxide) macromonomers) enabled the preparation of hydrophobic or hydrophilic brushes which possessed either rubbery or glassy bulk properties. Additionally, the modification of backbone and side chain composition allowed the preparation of a wide range of complex copolymers with varying topology, composition and density of grafting of brush-like segments present in the material [23–31]. When densely grafted brushes are deposited on surfaces which facilitate favorable interactions, the side chains tend to spread on the surface, enabling AFM imaging with single-chain resolution [32,33]. Imaging of brushes with block copolymer side chains [24,25] has also been conducted, in addition to brushes with block, statistical and gradient copolymer segments along a linear backbone [26,27]. Recently, molecular brushes with branched architectures have also been synthesized, and these serve as interesting probes to examine the initiation efficiencies of ATRP using multifunctional initiators [31]. 2.4.2

Molecular Brushes from Dendritic Macroinitiators

The synthesis of various hybrid dendritic-linear copolymers has been performed using dendritic initiators [18,22,99]. Additionally, densely grafted “dendri-graft” structures have been prepared by alternation of nitroxide-mediated chain extensions and ATRP [100]. Molecular brushes that are spherical in nature have been synthesized by ATRP of various methacrylates from a multifunctional dendrimer core. The synthesis and application of these dense spherical molecular brushes as nanoporogens toward the preparation of low k dielectric materials for microelectronic applications have been extensively studied by Hedrick and Miller et al. [20,21,101,102].

Summary

Atom transfer radical polymerization has emerged in recent years as a versatile and powerful tool for the synthesis of organic (co)polymers of precise molar mass, composition, and topology. Due to the high functional group tolerance of this process, both polymerizable and initiating groups have been incorporated to a variety of substrates, thereby enabling the modification of copolymer or surface properties. As demonstrated by results presented in this chapter, ATRP is particularly suitable for the preparation of polymer brushes from well-defined copolymers, nanocomposites, or covalently bound ultrathin films.

65

66

2 Polymer Brushes by Atom Transfer Radical Polymerization

Acknowledgments

Support from the National Science Foundation through grants DMR 00-90409 (K. M.), DMR 0110247 (T. K.) and ECS 01-03307 (K. M.) and CRP Consortium at Carnegie Mellon University is gratefully acknowledged. The text of this Chapter corresponds closely to a review article published recently by the same authors [103].

References 1 A. Halperin, M. Tirrell, T. P. Lodge, Adv.

Polym. Sci. 1992, 100, 31. 2 T. P. Lodge, M. Muthukumar, J. Phys. Chem. 1996, 100, 13275. 3 M. Ejaz, S. Yamamoto, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules 1998, 31, 5934. 4 J.-B. Kim, M. L. Bruening, G. L. Baker, J. Am. Chem. Soc. 2000, 122, 7616. 5 W. Huang, G. L. Baker, M. L. Bruening, Angew. Chem. Int. Ed. 2001, 40, 1510. 6 W. Huang, J.-B. Kim, M. L. Bruening, G. L. Baker, Macromolecules 2002, 35, 1175. 7 J.-B. Kim, W. Huang, M. L. Bruening, G. L. Baker, Macromolecules 2002, 35, 5410. 8 J.-B. Kim, W. Huang, M. D. Miller, G. L. Baker, M. L. Bruening, J. Polym. Sci., Polym. Chem. Ed. 2003, 41, 386. 9 M. M. Guerrini, B. Charleux, J.-P. Vairon, Macromol. Rapid Commun. 2000, 21, 669. 10 K. N. Jayachandran, A. Takacs-Cox, D. E. Brooks, Macromolecules 2002, 35, 4247. 11 D. Bontempo, N. Tirelli, G. Masci, V. Crescenzi, J. A. Hubbell, Macromol. Rapid Commun. 2002, 23, 417. 12 D. Bontempo, N. Tirelli, K. Feldman, G. Masci, V. Crescenzi, J. A. Hubbell, Adv. Mater. 2002, 14, 1239. 13 K. Min, J. Hu, C. Wang, A. Elaissari, J. Polym. Sci., Polym. Chem. Ed. 2002, 40, 892. 14 G. Zheng, H. D. H. Stoever, Macromolecules 2002, 35, 6828. 15 M. Ejaz, Y. Tsujii, T. Fukuda, Polymer 2001, 42, 6811. 16 A. Carlmark, E. Malmstroem, J. Am. Chem. Soc. 2002, 124, 900.

17 T. A. von Werne, D. S. Germack, E. C. Hag-

18 19

20

21

22 23

24

25

26

27 28

29

30

berg, V. V. Sheares, C. J. Hawker, K. R. Carter, J. Am. Chem. Soc. 2003, 125, 3831. M. R. Leduc, C. J. Hawker, J. Dao, J. M. J. Frechet, J. Am. Chem. Soc. 1996, 118, 11111. M. R. Leduc, W. Hayes, J. M. J. Frechet, J. Polym. Sci., Part A: Polym. Chem. 1998, 36, 1. A. Heise, J. L. Hedrick, C. W. Frank, R. D. Miller, J. Am. Chem. Soc. 1999, 121, 8647. A. Heise, S. Diamanti, J. L. Hedrick, C. W. Frank, R. D. Miller, Macromolecules 2001, 34, 3798. K. Inoue, Prog. Polym. Sci. 2000, 25, 453. K. L. Beers, S. G. Gaynor, K. Matyjaszewski, S. S. Sheiko, M. Moeller, Macromolecules 1998, 31, 9413. H. G. Boerner, K. Beers, K. Matyjaszewski, S. S. Sheiko, M. Moeller, Macromolecules 2001, 34, 4375. G. Cheng, A. Boekerm M. Zhang, G. Krausch, A. H. E. Mueller, Macromolecules 2001, 34, 6883. H. G. Boerner, D. Duran, K. Matyjaszewski, M. da Silva, S. S. Sheiko, Macromolecules 2002, 35, 3387. S. Qin, K. Matyjaszewski, H. Xu, S. S. Sheiko, Macromolecules 2003, 36, 605. S. Qin, H. G. Borner, K. Matyjaszewski, S. S. Sheiko, Polym. Prep. (Am. Chem. Soc., Div. Polym. Chem.) 2002, 43, 237. D. Neugebauer, K. Matyjaszewski, M. da Silva, S. S. Sheiko, Polym. Prep. (Am. Chem. Soc., Div. Polym. Chem.) 2002, 43, 239. D. Neugebauer, K. Matyjaszewski, Poly. Prep. (Am. Chem. Soc., Div. Polym. Chem.) 2002, 43, 241.

References 31 K. Matyjaszewski, S. Qin, J. R. Boyce,

50 S. Yamamoto, M. Ejaz, Y. Tsujii, M. Matsu-

D. Shirvanyants, S. S. Sheiko, Macromolecules 2003, 36, 1843. 32 S. S. Sheiko, S. A. Prokhorova, K. L. Beers, K. Matyjaszewski, I. I. Potemkin, A. R. Khokhlov, M. Moeller, Macromolecules 2001, 34, 8354. 33 S. S. Sheiko, M. Moeller, Chem. Rev. 2001, 101, 4099. 34 C. J. Hawker, A. W. Bosman, E. Harth, Chem. Rev. 2001, 101, 3661. 35 J. Chiefari, E. Rizzardo, in: Handbook of Radical Polymerization, K. Matyjaszewski and T. P. Davis, Ed.; Wiley-Interscience: Hoboken, 2002; p. 629. 36 M. Husemann, M. Morrison, D. Benoit, J. Frommer, C. M. Mate, W. D. Hinsberg, J. L. Hedrick, C. J. Hawker, J. Am. Chem. Soc. 2000, 122, 1844. 37 R. R. Shah, D. Merreceyes, M. Husemann, I. Rees, N. L. Abbott, C. J. Hawker, J. L. Hedrick, Macromolecules 2000, 33, 597. 38 H. Mori, D. C. Seng, M. Zhang, A. H. E. Mueller, Langmuir 2002, 18, 3682. 39 H. Fischer, Macromolecules 1997, 30, 5666. 40 H. Fischer, Chem. Rev. 2001, 101, 3581. 41 C. Yoshikawa, A. Goto, T. Fukuda, Macromolecules 2002, 35, 5801. 42 K. Matyjaszewski, P. J. Miller, N. Shukla, B. Immaraporn, A. Gelman, B. B. Luokalam T. M. Siclovan, G. Kickelbick, T. Vallant, H. Hoffmann, T. Pakula, Macromolecules 1999, 32, 8716. 43 J. Xia, S. G. Gaynor, K. Matyjaszewski, Macromolecules 1998, 31, 5958. 44 D. Xiao, M. J. Wirth, Macromolecules 2002, 35, 2919. 45 M. Husseman, E. E. Malmstroem, M. McNamara, M. Mate, D. Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang, T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424. 46 C. Devaux, J. P. Chapel, E. Beyou, P. Chaumont, Eur. Phys. J. E: Soft Matter 2002, 7, 345. 47 S. Yamamoto, M. Ejaz, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 5608. 48 M. Ejaz, S. Yamamoto, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 1412. 49 D. M. Jones, A. A. Brown, W. T. S. Huck, Langmuir 2002, 18, 1265.

moto, T. Fukuda, Macromolecules 2000, 33, 5602. 51 S. Yamamoto, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 5995. 52 S. Yamamoto, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 6077. 53 X. Kong, T. Kawai, J. Abe, T. Iyoda, Macromolecules 2001, 34, 1837. 54 B. Zhao, W. J. Brittain, J. Am. Chem. Soc. 1999, 121, 3557. 55 B. Zhao, W. J. Brittain, Macromolecules 2000, 33, 8813. 56 B. Zhao, W. J. Brittain, W. Zhou, S. Z. D. Cheng, J. Am. Chem. Soc. 2000, 122, 2407. 57 B. Zhao, W. J. Brittain, W. Zhou, S. Z. D. Cheng, Macromolecules 2000, 33, 8821. 58 R. A. Sedjo, B. K. Mirous, W. J. Brittain, Macromolecules 2000, 33, 1492. 59 S. G. Boyes, W. J. Brittain, X. Weng, S. Z. D. Cheng, Polym. Prep. (Am. Chem. Soc., Div. Polym. Chem.) 2002, 43, 549. 60 P. J. Miller, K. Matyjaszewski, Macromolecules 1999, 32, 8760. 61 J. Pyun, S. Jia, T. Kowalewski, K. Matyjaszewski, Macromol. Chem. Phys. 2004, 205, 411. 62 M. R. Buchmeiser, R. Kroll, K. Wurst, T. Schareina, R. Kempe, C. Eschbaumer, U. S. Schubert, Macromol. Symp. 2001, 164, 187. 63 N. Ayres, D. M. Haddleton, A. J. Shooter, D. A. Pears, Macromolecules 2002, 35, 3849. 64 X. Huang, L. J. Doneski, M. J. Wirth, Anal. Chem. 1998, 70, 4023. 65 F. C. Leinweber, J. Stein, M. Otto, Fresenius’ J. Anal. Chem. 2001, 370, 781. 66 U. Meyer, F. Svec, J. M. J. Frechet, C. J. Hawker, K. Irgum, Macromolecules 2000, 33, 7769. 67 C. Viklund, A. Nordstroem, K. Irgum, F. Svec, J. M. J. Frechet, Macromolecules 2001, 34, 4361. 68 O. Prucker, J. Ruehe, Macromolecules 1998, 31, 602. 69 O. Prucker, J. Ruehe, Macromolecules 1998, 31, 592. 70 X. Huang, M. J. Wirth, Macromolecules 1999, 32, 1694. 71 J. Qiu, B. Charleux, K. Matyjaszewski, Prog. Polym. Sci. 2001, 26, 2083.

67

68

2 Polymer Brushes by Atom Transfer Radical Polymerization 72 M. Antonietti, K. Landfester, Prog. Polym.

Sci. 2002, 27, 689. 73 J. M. Asua, Prog. Polym. Sci. 2002, 27, 1283. 74 Q. Zhang, E. E. Remsen, K. L. Wooley, J. Am. Chem. Soc. 2000, 122, 3642. 75 S. Liu, J. V. M. Weaver, M. Save, S. P. Armes, Langmuir 2002, 18, 8350. 76 N. Tsubokawa, S. Yoshikawa, Recent Res. Dev. Polym. Sci. 1998, 2, 211. 77 T. Von Werne, T. E. Patten, J. Am. Chem. Soc. 1999, 121, 7409. 78 H. Bottcher, M. L. Hallensleben, S. Nuss, H. Wurm, Polym. Bull. (Berlin) 2000, 44, 223. 79 T. K. Mandal, M. S. Fleming, D. R. Walt, Chem. Mater. 2000, 12, 3481. 80 G. Carrot, S. Diamanti, M. Manuszak, B. Charleux, J. P. Vairon, J. Polym. Sci., Part A: Polym. Chem. 2001, 39, 4294. 81 S. C. Farmer, T. E. Patten, Chem. Mater. 2001, 13, 3920. 82 T. von Werne, T. E. Patten, J. Am. Chem. Soc. 2001, 123, 7497. 83 S. Habaue, O. Ikeshima, H. Ajiro, Y. Okamoto, Polym. J. (Tokyo, Jpn) 2001, 33, 902. 84 C. Perruchot, M. A. Khan, A. Kamitsi, S. P. Armes, T. von Werne, T. E. Patten, Langmuir 2001, 17, 4479. 85 D. A. Savin, J. Pyun, G. D. Patterson, T. Kowalewski, K. Matyjaszewski, J. Polym. Sci., Polym. Phys. 2002, 40, 2667. 86 X. Chen, C. Perruchot, S. P. Armes, Polym. Prep. (Am. Chem. Soc., Div. Polym. Chem.) 2002, 43, 265. 87 S. Nuss, H. Bottcher, H. Wurm, M. L. Hallensleben, Angew. Chem. Int. Ed. 2001, 40, 4016. 88 K. Ohno, K.-m. Koh, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 8989.

89 B. Gu, A. Sen, Macromolecules 2002, 35,

8913. 90 J. Pyun, K. Matyjaszewski, T. Kowalewski,

D. Savin, G. Patterson, G. Kickelbick, N. Huesing, J. Am. Chem. Soc. 2001, 123, 9445. 91 G. Kickelbick, U. Schubert, Monatsh. Chem. 2001, 132, 13. 92 G. Kickelbick, D. Holzinger, C. Brick, G. Trimmel, E. Moons, Chem. Mater. 2002, 14, 4382–4389. 93 C. R. Vestal, Z. J. Zhang, J. Am. Chem. Soc. 2002, 124, 14312. 94 R. S. Tanke, S. M. Kauzlarich, T. E. Patten, K. A. Pettigrew, Chem. Mater. 2003, 15, 1682–1689. 95 J. Pyun, S. Jia, T. Kowalewski, G. D. Patterson, K. Matyjaszewski, Macromolecules 2003, 36, 5094. 96 W. Li, Z. Xie, Z. Li, J. Appl. Polym. Sci. 2001, 81, 1100. 97 N. Tsubokawa, T. Satoh, M. Murota, S. Sato, H. Shimizu, Polym. Adv. Technol. 2001, 12, 596. 98 T. Liu, S. Jia, T. Kowalewski, K. Matyjaszewski, R. Casado-Portilla, J. Belmont, Langmuir 2003, 19, 6342. 99 M. Jikei, M.-a. Kakimoto, Prog. Polym. Sci. 2001, 26, 1233. 100 C. J. Hawker, Acc. Chem. Res. 1997, 30, 373. 101 J. L. Hedrick, T. Magbitang, E. F. Connor, T. Glauser, W. Volksen, C. J. Hawker, V. Y. Lee, R. D. Miller, Chem. Eur. J. 2002, 8, 3308. 102 H.-C. Kim, J. B. Wilds, C. R. Kreller, W. Volksen, P. J. Brock, V. Y. Lee, T. Magbitang, J. L. Hedrick, C. J. Hawker, R. D. Miller, Adv. Mater. 2002, 14, 1637. 103 J. Pyun, T. Kowalewski, K. Matyjaszewski, Macromol. Rapid Comm. 2003, 24, 1043.

69

3

Polymer Brushes by Atom Transfer Radical Polymerization Initiated from Macroinitiator Synthesized on the Surface Viktor Klep, Bogdan Zdyrko, Yong Liu, and Igor Luzinov

Abstract

In the present study, a primary polymer layer approach was used for the preparation of an effective macroinitiator for the synthesis of polymer brushes by atom transfer radical polymerization (ATRP) initiated from the surface. For the initial surface modification, poly(glycidyl methacrylate) (PGMA) was used. When deposited on a substrate, the primary PGMA layer first reacted with the surface through formation of covalent bonds. The glycidyl methacrylate units that were not connected to the substrate served as reactive sites for the subsequent attachment of ATRP initiator. Accordingly, ATRP macroinitiator was synthesized on the substrate surface by the reaction between epoxy groups of PGMA and carboxy functionality of bromoacetic acid. Variation of the PGMA layer thickness allowed control over the amount of bromoacetic acid (BAA) attached to the surface. Two different surface concentrations of BAA were used in grafting experiments to investigate the relationship between the amount of initiator anchored to the surface through PGMA and the rate of brush formation. Polystyrene brushes of different thicknesses were synthesized on the PGMA/BAA-modified substrates using ATRP.

3.1

Introduction

The composition and behavior of surfaces and interfaces plays a pivotal role in dictating the overall efficiency of a majority of materials and devices. For instance, the control of surface and interfacial properties is critical in many traditional areas of science and technology such as colloid stabilization, adhesion, lubrication, rheology, immobilization of catalysts, and generation of multiphase materials [1–5]. Besides the traditional fields, surface modification has recently found use in bioengineering, nonlinear optics, (bio)sensors, nanopatterning, molecular recognition, waveguides, and electronic microcircuit processing. A promising means to optimize surface properties is via the deposition of thin polymer layers that are tethered to the surface and possess appropriate physical and chemical properties [6,7]. As a consequence,

70

3 Polymer Brushes by Atom Transfer Radical Polymerization

solid surface modification with thin polymer films has attracted great interest during the past few years. The most commonly used methods of thin film preparation on solid substrates include polymer deposition by spin casting, precipitation, Langmuir-Blodgett technique, polymer adsorption and chemical grafting of polymers. Grafting techniques have certain advantages over others, including the easy and controllable introduction of polymer chains with a high surface density, precise localization of the chain at the surface, and the long-term stability of the grafted layers [8]. Tethered polymer chains that are connected to the solid substrate by one chain end may be distinguished from other grafted polymer layers, as they form polymer brushes if a relatively high grafting density is reached. Brush-like layers are formed due to the excluded volume effect, when the substrate is completely covered with a relatively dense monolayer of grafted chains stretched normal to the support. There are several major parameters that control the brush properties, namely: grafting density, chain length, and chemical composition of the chains. It is possible to regulate the properties of the brush, which can be predicted by simple scaling relationships, by tuning the grafting parameters [6]. The chemical grafting of polymer brushes can be accomplished using either “grafting to” or “grafting from” methods. In the “grafting to” technique, end-functionalized polymer molecules react with complementary functional groups located on the surface to form tethered chains. However, only a limited amount of the polymer can be tethered onto the substrates when using the “grafting to” approach. The attaching polymer chains must overcome an activation barrier, which appears as soon as the earlier attached chains begin to overlap. In the “grafting from” technique, the polymerization is initiated from the substrate surface by attached (usually covalently bonded) initiating groups. Molecules of a monomer penetrate through the already grafted polymer layer easily and significant amounts of grafted polymer can be attained. This technique has been used for the preparation of thick grafted layers of high grafting density on the surface. Anionic [9], cationic [10,11], controlled/living [12,13], and conventional [14,15] free radical polymerizations have been successfully used to synthesize tethered polymer layers on solid substrate surfaces. By making an appropriate choice of initiating system, temperature, monomer, and concentration, it is quite possible to synthesize layers possessing different morphology, thickness and composition [6]. Thus, fine-tuning of the layer properties is possible. Controlled/“living” free radical polymerization has a number of advantages over traditional polymerization procedures. The main advantages that a “living” process provides are reliable control over the polymer molecular weight and narrow polydispersities. Thus, the nature of the polymerization process permits structural characteristics of the grafted polymer brush to be readily varied and controlled. An added benefit is the frontal character of the chain growth on the surface. In this manner, all chains have very similar history that may be translated in more predictable cooperative behavior of the chains and make the brush almost “defect-free”. Numerous effective approaches are reported for the synthesis of homogeneous polymer brushes by “living” free radical polymerization [16–38].

3.1 Introduction

Husseman et al. [12,39] have demonstrated that initiator functionalized surfaces, suitable for both alkoxyamine and atom transfer radical polymerization (ATRP) living free radicals procedures, can be readily prepared. These surfaces are stable to prolonged storage, and can be used for the controlled synthesis of polymeric brushes. The use of “living” free radical chemistry permitted the accurate control of molecular weight or thickness of the brush while maintaining low polydispersity. It was noted that different functional monomers could be polymerized utilizing the same synthetic approach to the brush synthesis. Matyjaszewski et al. [13] reported on the ATRP of styrene and acrylates from a silicon wafer modified with an initiator layer composed of 2-bromoisobutyrate fragments. In the presence of the proper ratio of activating and deactivating transition metal species, controlled radical polymerizations of styrene were observed. The thickness of the layer constituting chains grown from surface increased linearly with the molecular weight of the macromolecules obtained in solution in identical experiment. Boyes et al. [40] successfully synthesized and characterized a tethered triblock copolymer by sequential monomer addition to a self-assembled monolayer of (11-(2-bromo-2-methyl)propynyloxy) undecyltrichlorsilane. The block copolymer brushes were prepared using ATRP. Upon treatment with different solvents, the tethered triblock copolymer brushes exhibited reversible changes in surface properties. De Boer et al. [41] developed a method for chemically modifying a surface with a grafted iniferter monolayer, which can be used for a “living” free radical photopolymerization. By using this process, it was possible to control the length of the grafted polymer chains and therefore the layer thickness up to approximately 100 nm. Single layer grafted block copolymers were also obtained by subsequent polymerization of styrene and methyl methacrylate monomers. Tsujii et al. [42] studied the mechanism and kinetics of reversible addition fragmentation chain transfer technique (RAFT) initiated from a solid surface. It was shown that RAFT could also be used for the synthesis of polymer brushes. RAFT has been successfully applied to the synthesis of various polymer brushes by Baum and Brittain [43]. Styrene, methyl methacrylate, and N,N-dimethylacrylamide brushes were prepared under RAFT conditions using silicate surfaces that were modified with surface-immobilized initiators. There are two common approaches for the attachment of polymerization initiators for the brush synthesis by the controlled/living method. The first approach relies on the reactions between end-functionalized initiator and native functional groups originally present on the substrate surface [12,30]. A different approach involves the formation of a monolayer consisting of functional groups active towards the terminally functionalized (e.g., epoxide, amine, anhydride or hydroxide) initiator [35]. Silane and thiol chemistries have proved to be suitable for the grafting in this case. Usually, the coupling methods are relatively complex and specific for certain substrate/initiator combinations. An alternative method for attachment of initiating functionalities to modified surface involves the deposition of a primary polymer (mono)layer with activity towards both surface and functionalized molecules

71

72

3 Polymer Brushes by Atom Transfer Radical Polymerization

[44–46]. The polymer is used for the initial surface modification as well as for generation of the highly reactive primary layer. When deposited on a substrate, the primary layer first reacts with the surface through formation of covalent bonds. The reactive units located in the “loops” and “tails” sections of the attached macromolecules are not connected to the surface [47]. These free groups offer a synthetic potential for the further chemical modification reactions and serve as reactive sites for the subsequent attachment of the initiators. If the polymer used for building the primary layer contains functional groups, which are highly active in various chemical reactions, the primary layer approach becomes virtually universal towards both surface and end-functionalized initiating species being used for brush formation. In the present study, the primary polymer layer approach was used for the synthesis of an effective macroinitiator for the ATRP. For the initial surface modification poly(glycidyl methacrylate) (PGMA) was used. A polymer with epoxy functionality was chosen since the epoxy groups are quite reactive with carboxyl, hydroxyl, amino, and anhydride functional groups. The glycidyl methacrylate units that were not connected to the substrate served as reactive sites for the subsequent attachment of ATRP initiator. Accordingly, ATRP macroinitiator was synthesized on the substrate surface by the reaction between epoxy groups of PGMA and carboxy functionality of bromoacetic acid (BAA). Variation of the PGMA layer thickness allowed control over the amount of BAA attached to the surface. Polystyrene brushes of different thicknesses were synthesized on the PGMA/BAA modified substrates by ATRP.

3.2

Experimental

Highly polished single-crystal silicon wafers of <100> orientation (Semiconductor Processing Co.) were used as a substrate. The wafers were first cleaned in an ultrasonic bath for 30 min, placed in a hot piranha solution (3:1 concentrated sulfuric acid/30% hydrogen peroxide) for 1 h, and then rinsed several times with high-purity water. Glycidyl methacrylate and styrene (Aldrich) were purified by treatment with an inhibitor-removing adsorbent and vacuum distillation. Glycidyl methacrylate was polymerized via a free radical method to give poly(glycidyl methacrylate) (PGMA), Mn = 84 000, polydispersity (PDI) = 3.4. Copper(I) bromide (CuBr), 4,4¢-dinonyl-2,2¢bipyridyl (dNbP), BAA, and ethylbromoacetate (Et-BA) were used as received from Aldrich. PGMA was deposited on the silicon surface by dip-coating or adsorption from solution. Resulting PGMA films were treated with BAA vapor in argon at 110 LC, and then washed with ethanol. ATRP of styrene was carried out with maximum precaution to avoid oxygen. For this purpose, CuBr (1 mol), dNbP (2 or 2.1 mol), free initiator (Et-BA, 2 mol) or CuBr2 (0.05 mol), and styrene (300– 1000 mol) were sealed in a flask and immediately subjected to several repeating freeze-pump cycles. This reaction mixture was stirred at 50 LC until complete dissolution of solids and formation of a deep-brown homogeneous solution occurred. The final mixture was distributed between test tubes in a glove box (free O2 content < 5 ppm). The silicon wafers were immersed into each test tube in the glove box and

3.3 Results and Discussion

sealed. For polymerization, the test tubes were immersed into a preheated silicon oil bath (110 LC) for different time periods. Ungrafted polystyrene was removed from wafers by multiple washings with toluene, followed by washing in a toluene ultrasonic bath. Samples of the ungrafted polystyrene formed in the bulk were isolated by precipitation with ethanol and purified by multiple precipitations from tetrahydrofuran solution into ethanol, and dried in a vacuum oven. The thickness of the synthesized brushes was determined using ellipsometry, which was performed on a COMPEL Automatic Ellipsometer (InOm Tech, Inc.). Surface characterization of obtained brushes was performed by scanning probe microscopy (SPM) in tapping mode on a Dimension 3100 (Digital Instruments, Veeco) microscope. MW/Mn was determined using gel-permeation chromatography (GPC; Waters). The surface coverage (adsorbed amount), C (mg m–2), was estimated from the ellipsometry thickness of the layer, h (nm) by the following equation [48]: C = h

(1)

where is the bulk density of the attached molecule (g cm–3). The surface density, R (molecules per nm2) – that is, the inverse of the average area per adsorbed chain, was approximated by: R = C NA O 10–21/M = (6.023 C O 100)/Mn

(2)

where NA is the Avogadro’s number and M (g mol–1) is the molar mass of the attached molecule. The density of BAA (1.934 g cm–3) used in the calculations was obtained from the supplier [49].

3.3

Results and Discussion 3.3.1

Synthesis of Macroinitiator for ATRP

Figure 3.1 illustrates schematically the process used to synthesize the macroinitiator. The synthesis included two major steps: (1) deposition of the primary PGMA layer on the surface; and (2) attachment of BAA to the surface modified with the primary layer through the reaction between the carboxyl and epoxy functionalities. In order to obtain a thin layer of the PGMA macromolecular precursor attached to the surface of silicon wafer, two different procedures were employed, namely dip-coating and adsorption from PGMA solution. When the adsorption process was utilized, the thickness of the primary PGMA layer was varied from a few Angstroms to 5 nm by changing the adsorption time, temperature, and/or solvent quality. Generally, the thickness of the primary PGMA layer increased with time, and reached a plateau after 15–20 h. The solvent quality

73

74

3 Polymer Brushes by Atom Transfer Radical Polymerization

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

HO

-C(O)CR 2-Br R=H or CH 3

(

CH2-C(CH3) O

)n

O

O Figure 3.1 Schematic representation of poly(glycidyl methacrylate) (PGMA)/ bromoacetic acid (BAA) macroinitiator synthesis on the surface.

was adjusted by varying the solvent composition. Methylethylketone (MEK) and toluene were employed as a good solvent and a non-solvent, respectively. As the concentration of toluene in the solution increased, the thickness of the PGMA adsorbed layer increased accordingly. It should be mentioned that the adsorption modification gave good quantitative reproducibility only when fresh solutions were prepared from newly synthesized PGMA. It is supposed that some hydrolytic and/or crosslinking processes involving the PGMA epoxy groups might occur during storage of the PGMA samples and solutions. As a result of these side processes (which may be due to trace amounts of water and/or basic/acidic contaminants), small fractions of the hydrolyzed and/or crosslinked macromolecules having different (typically higher) adsorptive ability may be formed and change the course of the PGMA adsorption. However, any changes that may have occurred in the PGMA structure were undetectable with FT-IR and GPC. No significant differences in the FT-IR spectra and GPC data were observed after one year of PGMA storage under ambient conditions. Dip-coating gave excellent quantitative reproducibility, as compared to the adsorption procedure. After the dip-coating, not only smooth and uniform covering (Figure 3.2(a) and (b)) were observed, but also stable reproducibility in the thickness of the

3.3 Results and Discussion

reactive primary layer. Aging of the adsorbed film overnight at ambient conditions or annealing for 20 min at 110 LC led to permanent attachment of the deposited film to the substrate. The annealed PGMA layer was smooth, and uniformly covered the substrate (Figure 3.2(c) and (d)). The silicon wafers covered with the adsorbed PGMA layers were vigorously rinsed with series of highly polar solvents including dimthylsulfoxide (DMSO) and tetrahydrofuran (THF). The layer could not be removed from the substrate after vigorous solvent treatment, which suggested that the epoxy groups of the polymer had chemically anchored the PGMA to the surface [44]. Next, the stability and deactivation of the PGMA at the elevated temperatures was checked, as during the synthesis of ATRP macroinitiators the PGMA layer was kept at elevated temperatures for significant amounts of time. Dodecyl amine (DA) was grafted to the PGMA films annealed at 120 LC for different times to induce self-

a)

b)

c)

d)

Figure 3.2 Scanning probe microscopy (SPM) images of PGMA layer deposited on the surface of the silicon wafer by dip-coating. (a, b) As deposited; (c, d) annealed at 110 %C. Image sizes: (a, c) 10 * 10 lm; (b, d) 1 * 1 lm. Vertical scales: (a, c) 10 nm; (b, d) 2 nm.

75

3 Polymer Brushes by Atom Transfer Radical Polymerization

40 2

Σ (molecules/nm )

76

30 20 10

1

2

3

4

5

PGMA thickness (nm) Figure 3.3

a)

Surface density (R) of attached BAA versus thickness of the annealed PGMA layer.

b)

c)

Figure 3.4 SPM images of PGMA/BAA macroinitiator synthesized using: (a) as deposited PGMA layer; (b, c) PGMA layer annealed at 110 %C. Image sizes: (a, b) 10 * 10 lm; (c) 1 * 1 lm. Vertical scale 10 nm.

3.3 Results and Discussion

crosslinking and possible hydrolysis of the epoxy layer. This low molecular-weight substance was used as a probe for the presence of the accessible epoxy groups. The experiment showed that approximately 40% of the epoxy groups were still available on the surface after the annealing. The drop in the initial activity of the PGMA layer towards DA occurred almost immediately when the epoxy layer was heated. Since dip-coating process gave more reproducible primary PGMA layers, this process of layer deposition was used in further experiments. Attachment of the BAA to the surface covered with the PGMA film was conducted from the gaseous phase in argon at 110 LC. It was observed by ellipsometry that the effective thickness of the PGMA layer increased 1.3- to 1.6-fold after the treatment with BAA; this indicated attachment of a significant amount of the substance. Ellipsometry showed near-linear dependence between thickness of the primary PGMA layer and the amount of BAA attached (Figure 3.3). The reaction between the epoxy groups and the carboxyl functionality of the BAA produces a bromoacetic ester derivative of the PGMA (see Figure 3.1). Such a-bromoesters are known as effective initiators for ATRP of styrene, acrylic and some other vinyl monomers [50]. Therefore, the ATRP macromolecular initiator, covalently anchored to the silicon surface, was obtained. It should be mentioned that, when BAA was attached to the unannealed PGMA layer, the morphology of the PGMA/BAA film became irregular with a significant amount of aggregates scattered on the surface (Figure 3.4(a)). Conversely, the preannealed PGMA film produced macroinitiator possessing smooth and uniform surface morphology (Figure 3.4(b) and (c)). Thereby, for ATRP experiments we used only macroinitiator derived from the pre-annealed PGMA layer. These annealed PGMA films allowed initiator surface densities that were significantly higher than were reported for the self-assembled monolayer of ATRP initiators (e.g., for 6 nmthick primary PGMA film approximately 40 a-bromoester fragments per nm2 versus approximately three initiating sites per nm2 for the typical self-assembled monolayer of the ATRP initiator) [30]. 3.3.2

ATRP from Macroinitiator

The fundamental idea of ATRP is the halogen exchange in the polymerizing system between the halogen-terminated growing polymer chain/Cu(I) dNbP complex and macroradical/Cu(II) dNbP complex (Figure 3.5) [50]. Chain propagation is a firstorder process, while termination is a second-order reaction. For conventional bulk or solution ATRP the equilibrium is strongly shifted to the left. The free radical conR-Br

+

CuBrL2

k1

R

k2

+

CuBr2L2

M k3 Figure 3.5

Idealized mechanism of atom transfer radical polymerization (ATRP).

77

78

3 Polymer Brushes by Atom Transfer Radical Polymerization

centration is as low as 10–7 to 10–8 mol L–1, and the Cu(II) concentration is approximately 5% of that for Cu(I) [50]. Consequently, the termination is diminished and all chains grow simultaneously during the polymerization, without noticeable termination. An adequate concentration of Cu(II) is critical for effective reaction control. However, when the ATRP is initiated from the surface, the amount of the initiator located on the substrate is not sufficient to create the concentration of Cu(II) species required for polymerization control [13,51]. There are two methods developed for maintaining the Cu(II) concentration for ATRP initiated from a surface: (1) Simultaneous initiation of ATRP from the surface and in solution; and (2) Addition of the necessary amount of Cu(II) at the beginning of the process [13]. In the present study, both approaches were tested for the ATRP grafting from the PGMA/BAA macroinitiator adsorbed onto the surface, and the polymer brushes grafted to the silicon surface were successfully obtained using the two methods. Control experiments were conducted under exactly the same polymerization conditions, but with no initiator attached to the PGMA layer. Without the initiator almost all polymer, formed by the free initiator and deposited on the surface, was removed during the rinsing procedure and only film thickness of 1–2 nm remained on the surface. The observation showed that virtually no macroradicals from the bulk could be attached to the surface and that the PGMA/BAA macroinitiator was, indeed, necessary for brush formation. Additional experiments were conducted to confirm the role of other ingredients in the polymerization initiated from the surface. In the absence of both free initiator and CuBr2 in solution, very rapid formation of a polystyrene brush was observed on the surface covered with PGMA/BAA macroinitiator (in the range of 50–100 nm during the first hour). It is believed that for this case the initiation occurred by the reaction between BAA and CuBr, but no control was achieved and conventional free radical polymerization was carried out. When the brush synthesis was attempted without CuBr, dNbP, or PGMA on the surface of the wafer, only traces of polymer tethered to the substrate were detected. Figure 3.6 illustrates how the thickness of the polystyrene brush varies with time throughout ATRP initiated from the surface by the PGMA/BAA macroinitiator. Two different surface concentrations of BAA were used in these grafting experiments in order to acquire knowledge regarding the relationship between the amount of initiator anchored to the surface through PGMA and the rate of brush formation. The wafers used for grafting were covered with PGMA layers of two different thicknesses (0.8 nm and 4 nm). The effective thickness of the BAA attached to the surface (measured by ellipsometry) was 0.5 nm and 2 nm for the thinner and thicker PGMA adsorbed films, respectively. These amounts corresponded to 4 and 17 BAA molecules per nm2, respectively. These values were close to, and higher than, the typical surface density of self-assembled monolayers of ATRP initiators employed (~3 molecules nm–2) [30]. The process was carried out in the presence of an equivalent amount of Cu(I) complex to the free initiator (Et-BA) to generate Cu(II) species. At the beginning of the process, a virtually linear increase of polystyrene layer thickness was observed. Later, the grafting rate decreased and the brush thickness practically leveled off. The deceleration of the grafting could be explained by the monomer consumption for

3.3 Results and Discussion 2

PGMA 0.8 nm; BAA 4 molecules/nm 2 PGMA 4 nm; BAA 17 molecules/nm

35 Thickness (nm)

30 25 20 15 10 5 0 0

1000 2000 3000 4000 5000 Time (min)

Figure 3.6 Thickness of grafted polystyrene layer versus polymerization time. Sacrificial initiator added. Styrene/CuBr/Et-BA/dNbP ratio: 300/1/1/2.

a)

b)

c)

d)

Figure 3.7 Topographic SPM images of polystyrene brushes obtained with sacrificial initiator added. Initiator surface density: (a, b) BAA 4 molecules nm–2; (c, d) BAA 17 molecules nm–2. Brush thickness: (a) 6.5 nm;

(b) 18 nm; (c) 7 nm; (d) 21.5 nm. Vertical scale 10 nm. Image sizes 1 * 1 lm. Roughness: (a) 0.31 nm; (b) 0.56 nm; (c) 0.51 nm; (d) 0.67 nm.

79

3 Polymer Brushes by Atom Transfer Radical Polymerization

-1

Molecular Weight (gmol )

the ATRP process initiated in the bulk. Figure 3.7 shows the surface morphology of the grafted layers of different thickness. The brushes completely cover the substrate surface, and their structure is virtually independent of the surface concentration of the PGMA/BAA macroinitiator used for the synthesis. The roughness of the grafted layers increases with thickness (see legend for Figure 3.7). Since it has been shown [20] that grafted chains have almost the same molecular weight and molecular weight distribution as ungrafted (free) polymer being formed in the bulk or solution, the molecular weight of the free polystyrene synthesized was analyzed at the different stages of the process. The molecular weight increased simultaneously with the thickness of the grafted polymer layer on the surface (Figure 3.8). Polydispersity remained low (Mw/Mn < 1.3) throughout the grafting, and confirmed the controlled/living character of the polymerization in the bulk. Linear dependence between the thickness of the brush and the molecular weight was found (Figure 3.9). The linear relationship indicated that the polymerization processes were of the same nature in the bulk and on the surface. Thereby, providing further 25000 20000 15000 10000 5000 0

500 1000 Time (min)

1500

Figure 3.8 Molecular weight of ungrafted polystyrene versus polymerization time. Sacrificial initiator added. Styrene/CuBr/ Et-BA/dNbP ratio, 300/1/1/2.

2

PGMA 0.8 nm; BAA 4 molecules/nm 2 PGMA 4 nm; BAA 17 molecules/nm 35 30

Thickness (nm)

80

25 20 15 10 5 0

5000 10000 15000 20000 25000 -1 Molecular Weight (gmol ) Figure 3.9 Thickness of grafted polystyrene layer versus molecular weight of ungrafted polymer. Sacrificial initiator added. Styrene/CuBr/Et-BA/dNbP ratio, 300/1/1/2.

3.3 Results and Discussion

evidence that the ATRP of styrene from the PGMA/BAA macroinitiator also occurred in controlled/living manner. In a subsequent set of experiments, the polymerization was conducted without the Et-BA initiator in the bulk. CuBr2 (5% mol to CuBr) was added to the system to provide the necessary amount of the Cu(II) species. As expected, for the polymerization without the unbound initiator only a minute quantity of the free polystyrene was detected for long reaction times. The molecular weight of this polymer increased with time, as expected for typical ATRP reaction, up to 30 000 g mol–1 and the polydispersity index was narrow (Mw/Mn < 1.5). At the same time, significant amounts of the grafted polymer accumulated on the surface (Figure 3.10). During this period, the thickness of the polystyrene brush first increased linearly, after which the grafting process slowed down (Figure 3.10). SPM studies showed that the grafted layers uniformly covered the surface at different stages of the grafting process (Figure 3.11). After 1000–1500 min, a noteworthy amount of unbound polystyrene was detected in the bulk. This free polymer had a Mw > 200 000 and Mw/Mn > 2. Clearly, the free polymer with such characteristics was formed due to self-initiating polymerization of styrene, and not to the controlled/living ATRP process. At 110 LC, self-initiation is typical for the styrene-containing systems [52]. Thus, slow formation of the radicals in the bulk was present throughout the experiment. Interaction of these radicals with the CuBr/CuBr2 system resulted in slow ATRP polymerization in the bulk. With time, the amount of newly formed radicals became comparable with the amount of CuBr2 initially introduced in the reaction mixture, and the polymerization went out of control. The rate of grafting increased at this stage of the process, the brush formation proceeded in an uncontrolled manner, and very thick brushes were obtained (Figure 3.10). Jeyaprakash et al. [51] also reported self-acceleration of the ATRP grafting of styrene at very long polymeriza2

Thickness (nm)

250

PGMA 0.8 nm; BAA 4 molecules/nm 2 PGMA 4 nm; BAA 17 molecules/nm

100

50

0 0

500

2000 4000

Time (min) Thickness of grafted polystyrene layer versus polymerization time. No sacrificial initiator added. Styrene/CuBr/ CuBr2/dNbP ratio, 300/1/0.05/2.1

Figure 3.10

81

82

3 Polymer Brushes by Atom Transfer Radical Polymerization

tion times, and attributed this to the following possible pathways: (1) still active ends polymerize uncontrolled due to Trommsdorf conditions; or/and (2) chains growing in solution abstract hydrogen atoms from grafted chains, and further uncontrolled growth starts from the site where the hydrogen was lost. A recent investigation [30] has identified a strong relationship between the initiator concentration on the surface and the amount of ATRP-grafted poly(methyl methacrylate). These authors estimated that only one of ten initiating groups in the selfassembled monolayer produced a grafted polymer chain. The surface concentration of initiator was varied from the maximum value observed for the monolayer to concentration that was 10-fold lower. Surprisingly, an almost linear relationship was found between the initiator density and the grafting rate. These data, which were obtained at room temperature in aqueous media, could not be directly compared

a)

b)

c)

d)

Topographic SPM images of polystyrene brushes obtained without sacrificial initiator added. Initiator surface density: (a, b) BAA 4 molecules nm–2; (c, d) BAA 17 molecules nm–2. Brush thickness: (a) 4.3 nm; (b) 46 nm; (c) 14 nm; (d) 24.5 nm. Vertical scale 10 nm. Image size 1 * 1 lm. Figure 3.11

3.3 Results and Discussion

with the present results due to differences in monomers and conditions. However, the same tendency was found in the present study where sufficiently higher concentrations of the initiator were used for the brush synthesis. Figure 3.12 shows how grafting density, R, changes during the course of the brush synthesis. (The Mn values obtained for the ungrafted polymer formed by the added sacrificial initiator were used for the R determination.) The grafting density was approximately two-fold higher for the samples that possessed higher amounts of the PGMA/BAA initiator located on the surface. From a comparison between the surface densities of the initiator and the attached polymer, it is clear that the efficiency of the initiation was in the region of 7–15%, but this efficiency was halved when the concentration of the initiator was quadrupled. Initially, the grafting density increased and then remained virtually constant. The existence of an initial stage showed that there was a period at the beginning of the grafting when new grafted chains constituting the brush originated on the surface. The duration of this period was similar for the higher and lower initiator concentrations. The following constancy of the grafting density at longer times indicated that the rate of exchange between the active and dormant chain was sufficiently fast, and that all chains grew slowly, at an almost identical rate [53]. It is clear that there is some controversy between tendencies observed for the ATRP polymerization from the surface and bulk/solution. According to the classical controlled/living ATRP mechanism, the rate constant for initiation must be in the same range as that for polymer chain propagation [50]. The initiation is rapid, and occurs at the very beginning of the polymerization process. All chains grow at the same time/rate, and the number of polymer chains obtained is almost equal to the number of initiator molecules introduced into the system. Hence, the efficiency of the initiator is high. The simultaneous growth of the polymer chains leads to a low polydispersity of the final polymer. The molecular weight can be forecast with high 2

PGMA 0.8 nm; BAA 4 molecules/nm 2 PGMA 4 nm; BAA 17 molecules/nm

1.2

2

Σ (chains/nm )

1.6

0.8 0.4 0.0

0

1000 2000 Time (min)

3000

Grafting density (R) of polystyrene (PS) chains attached to the surface versus polymerization time. Molecular weights of PS grafted and ungrafted were assumed to be the same.

Figure 3.12

83

84

3 Polymer Brushes by Atom Transfer Radical Polymerization

accuracy from the monomer conversion and the amount of initiator introduced. Alternatively, for the ATRP from initiator attached to the surface, the initiating efficiency is unexpectedly low. Additionally, there was a certain period at the beginning of the grafting when new grafted chains constituting the brush originated on the surface. This period may be much longer than the initiation stage for the ATRP in bulk/solution, and lead to somewhat higher polydispersity for the grafted chains. Of interest was the finding that the other experimental observations were in good agreement with the presumption that the surface grafting process was close to the controlled/living ATRP process proceeding at the same time in bulk/solution. For instance, a linear dependence was found between the thickness of the brush and the molecular weight of polymer formed in the bulk/solution. One probable reason for this might be the different rates of initiation and propagation on the surface and in bulk/solution during the initial period. In fact, diffusion limitations may significantly affect the initiation and dormant species formation at the surface. The diffusion control may lead to a lower rate of initiation but a higher rate of initial propagation. Clearly, further investigations are warranted to understand the ATRP grafting initiated at the phase boundary.

Summary

The primary polymer layer approach was used successfully to prepare an effective macroinitiator for the synthesis of polymer brushes by ATRP initiated from the surface. PGMA was used for the initial surface modification. The ATRP macroinitiator was synthesized on the substrate surface by reaction between the epoxy groups of PGMA and the carboxy function of BAA. Variation in the PGMA layer thickness allowed control to be exerted over the amount of BAA attached to the surface. When BAA was attached to an unannealed PGMA layer, the morphology of the PGMA/ BAA film became irregular, with a significant amount of aggregates scattered on the surface. Conversely, the pre-annealed PGMA film produced a macroinitiator layer that possessed a smooth and uniform surface morphology. The annealed PGMA film allowed the achievement of an initiator surface density which was significantly higher than that reported for self-assembled monolayer of ATRP initiators. Polystyrene brushes of different thicknesses were synthesized on the PGMA/ BAA modified substrates using ATRP. At the start of the polymerization process, a linear increase in polystyrene layer thickness was observed, but later on the rate of grafting decreased and the brush thickness practically leveled off. Two different surface concentrations of BAA were used in these grafting experiments in order to obtain information regarding the relationship between the amount of initiator anchored to the surface through PGMA and the rate of brush formation. The grafting density was higher for samples which possessed a higher amount of the PGMA/ BAA initiator located on the surface. From a comparison between the surface densities of the initiator and the attached polymer, it was concluded that the efficiency of initiation was in the region of 7–15%, but this efficiency was halved when the concentration of the initiator was quadrupled.

References

Acknowledgments

These investigations were supported by the Department of Commerce through National Textile Center, M01-C03 Grant, and in part by the ERC Program of National Science Foundation under Award Number EEC-9731680. The authors thank Dr. Gary Lickfield for helpful discussions, and Ms. Kim Ivey for the GPC measurements.

References 1 R. P. Wool, Polymer Interfaces: Structure and

14 Luzinov, S. Minko, V. Senkovsky, A. Voronov,

Strength, Hunser/Gardner, New York, 1995. 2 J. T. Koberstein, Polymer Surfaces and Interfaces, Guest Editor, MRS Bulletin, January 1996, p. 16. 3 J. Israelachvili, Intermolecular and Surface Forces, Academic Press, New York, 1992. 4 B. Ratner, J. Biomed. Mater. Res. 1993, 27, 837–850. 5 D. H. Napper, Polymeric Stabilization of Colloidal Dispersions, Academic Press, London, 1983. 6 B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677–710. 7 A. Halperin, M. Tirrell, T. P. Lodge, Adv. Polym. Sci. 1992, 100, 31–71. 8 Y. Uyama, K. Kat, Y. Ikada, Adv. Polym. Sci. 1998, 137, 1–39. 9 R. Jordan, A. Ulman, J. F. Kang, M. H. Rafailovich, J. Sokolov, J. Am. Chem. Soc. 1999, 121, 1016–1022. 10 B. Zhao, W. J. Brittain, Macromolecules 2000, 33, 342–348. 11 R. Jordan, N. West, A. Ulman, Y.-M. Chou, O. Nuyken, Macromolecules 2001, 34 1606– 1611. 12 M. Husseman, E. E. Malmstrom, M. McNamara, M. Mate, D. Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang, T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424–1431. 13 K. Matyjaszewski, P. J. Miller, N. Shukla, B. Immaraporn, A. Gelman, B. B. Luokala, T. M. Siclovan, G. Kickelbick, T. Vallant, H. Hoffmann, T. Pakula, Macromolecules 1999, 32, 8716–8724.

S. Hild, O. Marti, W. Wilke, Macromolecules 1998, 31, 3945–3952. 15 O. Prucker, J. RThe, Langmuir 1998, 14, 6893–6898. 16 J.-B. Kim, M. L. Bruening, G. L. Baker, J. Am. Chem. Soc. 2000, 122, 7616–7617. 17 R. R. Shah, D. Merreceyes, M. Husemann, I. Rees, N. L. Abbott, C. J. Hawker, J. L. Hedrick, Macromolecules 2000, 33, 597–605. 18 M. Ejaz, S. Yamamoto, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 1412–1418. 19 M. Ejaz, S. Yamamoto, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules 1998, 31, 5934– 5936. 20 M. Ejaz, Y. Tsujii and T. Fukuda, Polymer 2001, 42, 6811–6815. 21 T. Yoshinobu, M. Ejaz, S. Yamamoto, T. Fukuda, K. Shigeto, K. Mibu, T. Shinjo, Polymer 2002, 43, 3837–3841. 22 S. Yamamoto, M. Ejaz, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 5608–5612. 23 U. Meyer, F. Svec, J. M. J. Frechet, C. J. Hawker, K. Irgum, Macromolecules 2000, 33, 7769–7775. 24 M. Ejaz, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 2870–2874. 25 D. M. Jones, W. T. S. Huck, Adv. Mater. 2001, 13, 1256–1259. 26 D. Xiao, M. J. Wirth, Macromolecules 2002, 35, 2919–2925. 27 N. Jayachandran, A. Takacs-Cox, D. E. Brooks, Macromolecules 2002, 35, 4247–4257. 28 N. Luo, J. B. Hutchison, K. S. Anseth, C. N. Bowman, Macromolecules 2002, 35, 2487–2493.

85

86

3 Polymer Brushes by Atom Transfer Radical Polymerization 29 H. Mori, A. Boker, G. Krausch, A. H. E. Mul30 31

32

33

34

35 36

37 38 39

40

41

42

ler, Macromolecules 2001, 34, 6871–6882. D. M. Jones, A. A. Brown, W. T. S. Huck, Langmuir 2002, 18, 1265–1269. L. Robinson, M. V. de Paz-Banez, X. S. Wang, S. P. Armes, Macromolecules, 2001, 34, 5799–5805. C. Perruchot, M. A. Khan, A. Kamitsi, S. P. Armes, T. von Werne, T. E. Patten, Langmuir 2001, 17, 4479–4481. G. Cheng, A. Boker, M. Zhang, G. Krausch, A. H. E. Muller, Macromolecules 2001, 34, 6883–6888. W. Huang, J.-B. Kim, M. L. Bruening, G. L. Baker, Macromolecules 2002, 35, 1175– 1179. X. Kong, T. Kawai, J. Abe, T. Iyoda, Macromolecules 2001, 34, 1837–1844. G. Carrot, S. Diamanti, M. Manuszak, B. Charleux, J.-P. Vairon, J. Polym. Sci. A: Polym. Chemistry 2001, 39, 4294–4301. H. BVttcher, M. L. Hallensleben, S. Nuß, H. Wurm, Polym. Bull. 2000, 44, 223–229. Min, J. Hu, Ch. Wang, A. Elaissari, J. Polym. Sci. A: Polym. Chemistry 2002, 40, 892–900. Husemann, M. Morrison, D. Benoit, J. Frommer, C. M. Mate, W. D. Hinsberg, J. L. Hedrick, C. J. Hawker, J. Am. Chem. Soc. 2000, 122, 1844–1845. S. G. Boyes, W. J. Brittain, X. Weng, S. Z. D. Cheng, Macromolecules 2002, 35, 4960–4967. B. de Boer, H. K. Simon, M. P. L. Werts, E. W. van der Vegte, G. Hadziioannou, Macromolecules 2000, 33, 349–356. Y. Tsujii, M. Ejaz, K. Sato, A. Goto, T. Fukuda, Macromolecules 2001, 34, 8872–8878.

43 Baum, W. J. Brittain, Macromolecules 2002,

35,610–615. 44 M. Kothe, M. Muller, F. Simon, H. Komber,

45

46

47

48

49

50

51

52 53

H.-J. Jacobasch, H.-J. Adler, Colloid and Surfaces A 1999, 154, 75–85. V. Klep, I. Luzinov, Polymer Preprints 2002, 43(2), 706–707; K. S. Iyer, B. Zdyrko, H. Malz, J. Pionteck, I. Luzinov, Macromolecules 2003, 36, 6519–6526; Y. Liu, V. Klep, I. Luzinov, Polymer Preprints 2003, 44(1), 564–565. A. Sidorenko, X. W. Zhai, F. Simon, D. Pleul, V. V. Tsukruk, Macromolecules 2002, 35, 5131–5139. G. J. Fleer, M. A. Cohen Stuart, J. M. H. M. Scheutjens, T. Cosgrove, B. Vincent, Polymers at Interfaces, Chapman & Hall, New York, 1993. G. Henn, D. G. Bucknall, M. Stamm, P. Vanhoorne, R. Jerome, Macromolecules 1996, 29, 4305–4313. Aldrich: Handbook of Fine Chemicals and Laboratory Equipment, Milwaukee 2003–2004, p. 268. K. Matyjashewski, in: Controlled Radical Polymerization. ACS Symposium Series 685, American Chemical Society, Washington DC, 1998, Chapter 16, pp. 258–283. J. D. Jeyaprakash, S. Samuel, R. Dhamodharan, J. RThe, Macromol. Rapid Commun. 2002, 23, 277–281. G. Odian, Principles of Polymerization, WileyInterscience, New York, 1991, p. 231. A. Ramakrishnan, R. Dhamodharan, J. RThe, Macromol. Rapid Commun. 2002, 23, 612– 616.

87

4

Synthesis of Polypeptide Brushes Henning Menzel and Peter Witte

Abstract

Tethering rigid rod-like polypeptides covalently to a surface provides brushes with interesting properties. The dipole moment of the a-helical peptides can result in a net dipole moment of the film and in piezoelectricity. The dipole moment of the macromolecules on the other hand causes also problems in the preparation of brushes. Several approaches for peptide brush synthesis by using “grafting to” and “grafting from” methods are discussed, and a new method based on a living nickelmediated N-carboxyanhydride polymerization is presented.

4.1

Introduction

Polypeptide brushes can be formed by tethering polymers such as poly(benzyl-l-glutamate) (PBLG) to surfaces. Peptides such as PBLG form an a-helix as a stable secondary structure, and therefore are rigid and rod-like. In this respect, they differ from polymers with a flexible secondary structure, and this results in some unique characteristics. For instance, with its anisotropic molecular shape arising from its ahelical structure, PBLG shows a large dipole moment of 8000 D along the molecular axis, as well as a high hyperpolarization b of 5 - 10–28 esu for Mn = 500 000 g mol–1 [1]. Therefore, a non-linear optical (NLO) response (e.g., second harmonic generation, SHG) can be expected if a monomolecular polyglutamate film has an unidirectional orientation of the polymer backbones [2,3]. The permanent dipole moment of the film also results in piezoelectric effects [3], while polypeptide brush films also show promise in applications as a novel orienting layer in liquid crystal displays [4–6]. Because of their chirality and their helical structure, the amino acid building blocks might find applications in coatings of capillaries for chiral separations. Last but not least, polypeptide brushes may be useful for testing the theories for rigid rod-like polymer brushes.

88

4 Synthesis of Polypeptide Brushes

4.2

Preparation of Peptide Brushes by “Grafting To”

One approach to unidirectional orientation of the polypeptide helices is covalent attachment of one of the end groups to the substrate with a high grafting density (Figure 4.1). The grafting can be achieved by coupling of preformed polypeptides to the surface (“grafting to”) or by polymerization of the monomer N-carboxyanhydride (NCA) employing surface-bound initiators (“grafting from”).

Figure 4.1

Schematic drawing of an aligned polypeptide brush.

The problems associated with the grafting to approach for the preparation of peptide brushes are, in principle, well known and are the same as for all other polymers. The grafting density remains rather low due to the coverage of adsorption sites by the first molecules, the formation of diffusion barriers, and so forth. In the case of the helical polypeptides, the situation is even more complicated due to a strong tendency to form aggregates [7–10]. In such aggregates, the peptide rods are in an antiparallel arrangement in order to compensate the dipole moments. Therefore, it is expected that only half of the molecules are tethered to the surface, while the other half can be removed [11]. Despite these problems, Samulski et al. reported a nonplanar arrangement of PBLG end-functionalized with disulfide groups on gold substrates – that is, a relatively high surface coverage was achieved [12,13]. The orientation of the rods, on the other hand, is not perpendicular but almost isotropic [13], and these authors obtained the thickest layers upon adsorbing the polypeptide from solvents in which strong aggregation occurs [12]. An improvement in layer thickness and orientation of the peptide rods was possible by employing electrical fields to pre-orient the peptides before adsorbing them to the surface [14]. However, both layer thickness and orientational order in the film were still low. Coupling of PBLG to the silicon substrate modified with (N-(2-aminoethyl)-3-aminopropyl)methyldimethoxysilane was carried out by Machida et al. with dicyclohexylcarbodiimide [15]. In this case the surface coverage was relatively low, and the rods were oriented more or less parallel to the surface. These authors used the PBLG films to induce spiral textures in nonchiral nematic liquid crystalline 4-n-pentyl-4¢-cyano-biphenyl molecules [4–6,16]. Self-assembled layers of short-chain polyglutamic acid end-functionalized with a disulfide moiety were prepared on gold by Niwa et al. [17,18]. The density of the peptides on the gold surface can be adjusted via the pH of the aqueous solutions used for the adsorption experiments, which affects the helical content of peptides (the

4.2 Preparation of Peptide Brushes by “Grafting To”

higher the content of helical material, the higher the surface coverage) [17]. The low coverage films adsorb guest peptides that are not functionalized with an anchor group, which was shown by quartz microbalance measurements. By using electrochemical measurements and guest peptides which have electrochemically active groups as probes either at the C-terminus or the N-terminus of the peptide chain, it was possible to show that adsorption of the guest peptides takes place in such a way that the peptide rods are anti-parallel. This result is a strong indication that the interaction of peptide macrodipoles is the driving force for adsorption of the guest peptides [17,18]. This adsorption induced by macrodipole interaction can be reversed by using strongly helix-breaking solvents, and this principle has been suggested for the immobilization of DNA via a peptide functionalized with an intercalator. This peptide binds to the DNA via the intercalator, and to the peptide SAM layer at the surface via macrodipole interaction [11]. The importance of the macrodipole interaction and aggregation for the formation of better-ordered peptide SAMs was also emphasized by Kimura and coworkers [19– 21], who have prepared well-ordered peptide monolayers with almost perpendicular orientation of the helices by complex formation between a surface-bound ammonium group and a crown ether end group. The driving force here is the complexation and macrodipole interaction, as only one half of the peptides is bound via the complex formation. Similar results were obtained with disulfide end-functionalized poly(alanin-2-aminobutyric acid). This peptide has a strong helix-formation tendency, and smaller side chains allow stronger intermolecular interaction than in the case of PBLG. In contrast to PBLG, poly(alanin-2-aminobutyric acid) forms SAM on gold in which the orientation of the peptide rods is not random, but almost perpendicular with only a small tilt angle if the correct conditions are used for the adsorption [20]. In particular, the solvent from which the adsorption is carried out plays an important role (ethanol or dimethylformamide, DMF). Better results are obtained when using ethanol as solvent, because the peptides strongly aggregate in this medium, and the aggregation supports the formation and orientation of the helices [20]. By using the Kelvin technique it was possible to show that the SAMs of these peptides have a surface potential. Negative surface potentials of a few hundred millivolts were detected when the helix was immobilized via the N-terminus, but positive surface potentials were observed when the peptide was bound via the C-terminus. Furthermore, the longer the helix peptide, the larger was the surface potential. Both experimental results indicate that the surface potential is generated by the macrodipole of the peptide helix [21]. These authors also showed that the macrodipole orientation influences electron transfer in photocurrent generation [21]. Niwa et al. and Kimura et al. have shown that the problems associated with the grafting to approach can be overcome – at least partially – by taking advantage of the macrodipole interaction of the peptide helices.

89

90

4 Synthesis of Polypeptide Brushes

4.3

Preparation of Peptide Brushes by Grafting From Polymerization

Poly-l-glutamates can be prepared by polymerizing the corresponding N-carboxyanhydride 1 (Figure 4.2). With regard to initiation from the surface, it must be considered that at least two mechanisms of NCA-polymerization are competing, depending on the initiator used, namely: (1) the activated monomer (a.m.) mechanism, which is initiated by bases; and (2) the amine- or protic-mechanism, which is initiated by primary amines [22]. In order to understand the requirements and limitations of a grafting from polymerization, it is worthwhile examining more closely the mechanism of NCA polymerization.

Figure 4.2 Initiation and propagation in the N-carboxyanhydride (NCA)-polymerization via the amine-mechanism according to Ref. [22].

4.3.1

Mechanisms of NCA Polymerization Amine-Initiated Polymerization If the polymerization is initiated by primary or secondary amines, without any steric hindrance, then the process proceeds via the “amine mechanism” (Figure 4.2). The amine attacks the NCA-ring, and ring opening occurs. The intermediate 3 eliminates carbon dioxide, and a dimer with an amino end-group is generated, which can attack the next monomer. The new amino group is less reactive than a primary amine. Therefore, initiation is much faster than propagation, and the degree of polymerization can be adjusted to some extent by varying the monomerto-initiator ratio [22–24]. In this respect, the NCA polymerization has a kind of “living character”. Another mechanism which has been discussed in the literature is 4.3.1.1

4.3 Preparation of Peptide Brushes by Grafting From Polymerization

the “carbamate mechanism” (see, for example, Ref. [22]), but this does not play an important role in grafting from polymerizations. Activated Monomer Mechanism Strong bases can deprotonate the N-carboxyanhydride, whereupon the deprotonated anionic monomer is a strong nucleophilic agent that can attack another NCA molecule and, therefore, is called an “activated monomer”. The attack of a NCA ring results in the intermediate 6 (Figure 4.3). The intermediate 6 has two reactive sites: (1) the electrophilic N-acyl group; and (2) the nucleophilic carbamate group. The latter group can react according to the carbamate mechanism (pathway A in Figure 4.3). Furthermore, 6 can decarboxylate and the generated amine end-group in 7 can react according to the amine mechanism shown in Figure 4.2 (pathway B in Figure 4.3). Since decarboxylation of the intermediate 7’ gives an amide anion, the decarboxylation also results in deprotonation (activation) of a monomer molecule. This activated monomer can now attack a NCA ring, either of a monomer or at the end of a growing chain (pathway C in Figure 4.3). 4.3.1.2

Figure 4.3 Initiation and the various propagation reactions for the NCA-polymerization via the activated monomer mechanism.

91

92

4 Synthesis of Polypeptide Brushes

The polymerization occurs much more rapidly if strong bases or sterically hindered amines are used as initiators. Therefore, it was concluded that the attack of the activated monomer (pathway C in Figure 4.3) is much faster than the other possible polymerization reactions [22]. The diverse propagation steps encountered with this mechanism can explain the following results. First, the molecular weight is not determined by the ratio of monomer to initiator. Second, the molecular weight distribution is broad. Since the polymer chains are bifunctional, intermolecular coupling can occur, which increases the molecular weight [22–24]. If polymerization of the NCA proceeds via the “activated monomer mechanism”, then the initiator is not incorporated into the polymer. Thus, this mechanism must be avoided if the polymer is to be attached to a surface in a grafting from experiment. The “activated monomer mechanism” is preferred in the case of strong bases as initiators. Unfortunately, amines are also strong bases, and so can initiate NCA polymerization via the “activated monomer mechanism”, especially when sterically hindered. Therefore, for an efficient grafting from polymerization of NCA the surface should have primary amino groups with low steric hindrance.

Figure 4.4 Mechanism of the nickel-mediated polymerization of a-amino acid NCA, according to Deming [27,28].

4.3 Preparation of Peptide Brushes by Grafting From Polymerization

Living Nickel-Mediated NCA Polymerization According to Deming Deming published a new type of NCA polymerization in which nickel complexes are employed as initiators. This new NCA polymerization is a living polymerization – that is, the degree of polymerization can be adjusted via the initiator to monomer ratio, block copolymerization is possible, and there are no side reactions [25,26]. The mechanism of nickel-mediated polymerization of NCA as suggested by Deming is depicted in Figure 4.4 [27,28]. The first step is an oxidative addition of the NCA to the Ni0 complex, followed by decarbonylation. Subsequently, another NCA molecule is added and a Ni-amido amidate complex 12 is formed. This complex does not initiate NCA polymerization, and can be isolated in a dimeric form. It can also be activated by adding a ligand for the Ni in the complex (i.e., bipyridine). The activated complex 14 polymerizes NCA by insertion of a monomer unit into the Ni-amidate bond. 4.3.1.3

4.3.2

Amine-Initiated Grafting From Polymerizations in Solution

The first reports describing the grafting of polyglutamates from a surface outlined the use of spherical substrates (Aerosil A200V [29] and carbon black [30]) modified with aminopropyltrimethoxysilane (APS) as initiator on the surface. Schouten and coworkers used essentially the same chemistry to graft polyglutamates and polyaspartates on glass and silicon slides [29]. The layer thickness of the grafted polymer was estimated from infra-red (IR) spectroscopy data to be 5–10 nm [29]. These authors were also able to show that the preferred secondary structure of the peptide is not influenced by grafting the polymer to a substrate [31]. Whitesell et al. used indium-tin oxide modified with APS or a gold surface with a specially designed initiator with thiol anchoring groups to polymerize alanine NCA [32,33]. The thiol initiator was designed to fit the surface requirements of the polypeptide helix and, therefore, to reduce steric hindrance. These authors claimed that their surface-grafted polyalanine had a perpendicular helical structure with a film thickness of up to 300 nm. However, this layer thickness was estimated from the intensity of an IR absorption band, whilst ellipsometry as a direct method produced highly variable data. It is most likely that the layer thickness of 300 nm was overestimated because of physisorbed polymers which could not be removed by washing the substrate with ethanol. According to Whitesell et al., it is favorable to adjust the surface density of initiating groups in order to reduce the steric hindrance for the polymerization reaction [32]. This can be carried out by using specially designed initiator molecules. Heise et al. also employed self-assembled monolayers as initiating layers, and adjusted the surface density of initiating sites by diluting through use of binary mixtures [34,35]. The thickness of PBLG films grafted from those substrates after washing with dichloracetic acid/chloroform mixtures was up to 12 nm, as measured by ellipsometry and X-ray reflectometry [34,36]. The layer thickness is comparable to those measured by Schouten and coworkers [29], and there was also no significant dependence of film thickness on the monolayer composition, which determines the surface den-

93

94

4 Synthesis of Polypeptide Brushes

sity of initiators. It was concluded that limitations inherent to the polymerization mechanism [22,36] restricts the chain length and layer thickness and not steric hindrance due to a high surface density of initiators. The Schouten group recently published the results of their solution-based grafting polymerizations with improved techniques. These authors used DMF as solvent (as this reduces the tendency to aggregate) and also adjusted the monomer concentration. In this way they were able to show that a layer thickness of 24 nm [37] and a tilt angle of 32–48M (depending on the side chain of the polyglutamate, whether methyl or benzyl ester) can be achieved [38]. However, the most important result was that the polymer layer could be used for a renewed polymerization after the polypeptide which was formed as a free polymer in solution and physisorbed to the graft polymer layer had been removed by careful washing steps. This result indicates that the degree of polymerization in amine-initiated grafting from polymerization is not limited by side reactions but rather by a type of physical death. That is, the adsorption of polymers restricts the accessibility of the growing chain end such that almost no more growth occurs. These renewed polymerizations enabled the authors also to prepare block copolymers by simple amine-initiated grafting from polymerization [39]. This method was also used for the preparation of PBLG surface-grafted films on aluminum substrates, which were used to measure the electromechanical properties of polar films [3]. The electric field-induced change in film thickness, which was dominated by a large inverse-piezoelectric effect, showed that polypeptide layers with a large, persistent polarization can be fabricated employing the grafting from approach. The molecular dynamics of PBLG surface grafted from aluminum substrates modified with 1-phosphoric acid-12-N-ethylaminododecane was investigated by Hartmann et al. [40]. KratzmNller et al. prepared microstructured polypetide brushes by using microcontact printing of 12-mercaptododecane amine on gold surfaces and subsequent amine-initiated NCA graft polymerization in solution. The thickness of the film was shown by ellipsometry (homogeneous film) and by AFM (microstructured film) to be ~30 nm [41]. 4.3.3

Other Techniques for Amine-Initiated Grafting From Polymerizations

Wieringa and Schouten also developed a solvent-free technique to graft PBLG and poly(methyl-l-glutamate) from silicon wafer modified with 3-aminopropyltriethoxysilane (APS) or (4-aminobutyl)dimethylmethoxysilane [42]. These authors spin-coated a NCA film onto the amino functionalized wafer, and heated the film above the melting point of the NCA. The polymerization then took place in the melt, and therefore at much higher monomer concentrations. Beside the tethered polymer, large amounts of unbound PBLG are also formed when using this technique, and consequently the films must be washed with chloroform/dichloroacetic acid mixtures. The layer thickness of the films after the washing step was approximately 20 nm[42].

4.4 Preparation of Peptide Brushes by Living Grafting From Polymerization

Another solvent-free method is the vapor deposition method developed by Chang and Frank [43]. By adjustment of parameters such as monomer concentration, reaction time, reaction temperature and pressure, the layer thickness can be increased and fine-tuned. By using specially designed chambers and optimized reaction parameters, layers of up to 180 nm thickness can be created using this technique. Vapor deposition polymerization also allows a renewed polymerization. After washing away any physisorbed material, the grafted polymer can be used as an initiator for new monomer, and the preparation of surface-grafted block copolypeptides is possible [44]. Polar and quadrupolar anisotropy has been established for these films by electro-optic measurements [45]. By employing both a good solvent to stretch out and orient the molecular chains, and a nonsolvent to “freeze” their upright orientation, the order of the peptide rods could be improved significantly. The better orientation after solvent treatment has been proven by grazing incidence reflection FT-IR spectroscopy [46].

4.4

Preparation of Peptide Brushes by Living Grafting From Polymerization 4.4.1

Copolymerization Approach

Although some the problems of surface grafting by use of amine-initiated polymerization have been solved (see previous sections), there remains a need for controlled graft polymerization in order to improve the layer thickness, to achieve block copolymerizations, and to graft to nonflat surfaces, for example the inside of a capillary. Use of the living nickel-mediated polymerization as described by Deming (see Section 4.3.1.3) appears to be highly advantageous in this respect. In order to use nickel-mediated NCA polymerization for a grafting from polymerization, it is necessary to bind the growing chain to the surface. This binding cannot be carried out via the bipyridine ligand, as the initiating complex is destroyed upon work-up. The tethering, therefore, must be made via the amido amidate part of the complex, which is formed by the reaction of NCA molecules with Ni0 species in three steps. In the case of amino acids with a second functional group such as l-glutamic acid, this can be used to link the NCA to a surface. The question arising is in which step of the initiator formation the tethered NCA molecule should be used to obtain an optimized system for grafting from polymerization. Tethering the first NCA molecule would result in an active chain end which remains at the surface while the chain grows. This would result in an increasing diffusion barrier, and limitations in the degree of polymerization and layer thickness are likely. The second NCA molecule which is added in the decarboxylation step cannot be used to tether the growing chain to the surface, because the five-membered metallacycle 11 (see Figure 4.4), as it is produced in the decarbonylation step and which is necessary as reagent, can not be isolated.

95

96

4 Synthesis of Polypeptide Brushes

Figure 4.5 Strategy for tethering a growing polymer chain to a surface employing a block copolymerization approach: the surface-bound NCA (X = surface) is subjected to an excess of

initiator A to create a growing chain end at the surface. After removing the excess of A, free NCA is added to grow the polymer chain from the surface.

Additional NCA molecules are incorporated at the reaction center during the course of the polymerization reaction. Therefore, one possible way to tether the growing polymer chain to a surface is a process similar to block copolymerization. The procedure comprises the addition of an activated complex to a surface bearing covalently bound NCA molecules to create a growing chain end. Subsequent

Figure 4.6

Preparation of NCA tethered to a polystyrene resin via a photolabile linker [47].

4.4 Preparation of Peptide Brushes by Living Grafting From Polymerization

removal of the excess of initiating complex and addition of free NCA to the active chain ends at the surface allow for further growth of the polymer chain (Figure 4.5). The proof of principle for this approach was presented only recently [47,48], whereby polystyrene beads were used as substrates after functionalization with a photo-labile linker. Glutamic acid was bound to this substrate and subsequently converted into surface-bound NCA (Figure 4.6). A bromo-a-methylphenacyl group was chosen as linker that can be introduced by Friedel-Crafts acylation of the polystyrene beads with a-bromopropionic acid chloride [49]. The a-methylphenacyl ester can be cleaved by UV light, but is otherwise stable to the reaction conditions. The selective esterification resulting exclusively in the c-l-glutamic acid ester can be performed by a substitution reaction between a copper complex of the glutamic acid salt and a halogen [50] species. Treatment with either phosgene or triphosgene yields the surface-bound NCA. Resin-3 with the surface-bound NCA was then reacted with activated Ni amido amidate complex A according to the scheme presented in Figure 4.5. After the reaction, the excess of A was removed by several careful washing steps, and a solution of c-benzyl-l-glutamic acid NCA (BLG-NCA) in DMF was added. After 16 h the solution was removed and the resin carefully washed and extracted in order to remove any physisorbed polymer. The resin showed a significant mass increase due to surface-grafted polyglutamate. The polymer at the surface can be established from FT-IR spectra, which show all the bands typical for PBLG, in addition to the bands stemming from the polystyrene resin itself (Figure 4.7). The polymer can be cleaved from the resin by photolysis of the methylphenacyl group with UV light at 365 nm. The cleaved polymer can be isolated and investigated by size-exclusion chromatography (SEC; Figure 4.8). The peak shape in the SEC trace is almost Gaussian; however, a second peak in the SEC trace can be detected by light scattering. This second peak occurs at very high molecular weights,

FT-IR spectra of poly(benzyl-l-glutamate) (PBLG) grafted from a polystyrene resin, and of PBLG prepared in solution.

Figure 4.7

97

4 Synthesis of Polypeptide Brushes 0.25 LS detector 90º (x 10) RI detector

0.20

intensity

98

0.15

0.10

0.05

0.00 8

9

10

11

12

13

14

15

16

17

volume

Gel permeation chromatography (GPC) traces [refractive index (RI) and light scattering (LS) detectors) of the cleaved poly(benzyl-l-glutamate). Figure 4.8

and is caused by debris formed by the prolonged stirring of the resin that could not be removed by membrane filtration. From the SEC traces and the light scattering data, a molecular weight of Mn = 22 000 g mol–1 and a polydispersity (PD) of 1.52 could be determined for the cleaved graft polymer. Beside the graft polymer at the surface, there is polymer present in the solution after the reaction, and this can be easily isolated by precipitation. The molecular weight and polydispersity of this polymer suggests that it has been formed by some physisorbed initiator, which starts polymerization while still physisorbed at the surface. With increasing length of the polymer chain, it becomes more soluble in the solution and eventually desorbs. The results of the polymerization experiments – and in particular the polydispersity of the cleaved PBLG – provide evidence that a living polymerization of NCA started by a nickel initiator tethered to a surface is possible. However, the protocol used to prepare the substrate with the initiator at the surface is complicated, and it was not possible to remove all of the physisorbed initiator. Therefore, PBLG was formed not only at the surface but also in solution. In the preparation of polar films by using a grafting from polymerization approach, the presence of free polypeptides in solution is unfavorable, because of possible aggregates with anti-parallel arrangement of the dipoles, which can in turn result in strong physisorption [17,19–21,51] and physical “death” of the growing polymer chain [39].

4.4 Preparation of Peptide Brushes by Living Grafting From Polymerization

4.4.2

Alloc-Amide Approach

Simplification of the synthetic protocol and the exclusion of physisorbed initiator requires that the initiator is built up at the surface in more defined way. This is possible when employing allyoxycarbonyl (alloc) amides as precursors (Figure 4.9) [52]. The R*-group in the alloc amide can be varied widely [52], and it should be possible to use this position in the precursor to tether the complex to the surface.

Figure 4.9 Schematic drawing of the preparation of nickel amido amidate complexes from alloc amides [52].

We have prepared surfaces equipped with alloc amide functionalities by reacting the N-hydroxysuccinimide ester of the leucine alloc amide [52] with commercially available amino-functionalized polystyrene beads (Figure 4.10). The functionalized beads were then activated with an excess of Ni(COD)2 and a ligand (e.g., phenanthroline) according to the scheme depicted in Figure 4.9.

Figure 4.10

Alloc amide functionalization of polystyrene resin with amino groups.

The functionalization and activation can be monitored using FT-IR spectroscopy (Figure 4.11). In order to verify the active species, a model complex was synthesized and compared to the complex synthesized with the alloc amide resin. As can be seen in Figure 4.11, the bands typical for N-H bonds and the amide group for the surface-bound species and the model complex are identical. Thus, initiator produced at the resin surface can be used to polymerize BLG-NCA in DMF solution. The polymerization results in a significant mass increase of the resin, even after careful washing with dichloroacetic acid/chloroform, and corresponds to approximately 18% of the monomer used in the reaction. The presence of poly-c-benzyl-l-glutamate at the surface of the resin can also be established in the

99

100

4 Synthesis of Polypeptide Brushes

FT-IR spectra of polystyrene beads functionalized with alloc amide groups, polystyrene beads functionalized with nickel amido amidate initiating sites, and the corresponding model initiator, respectively.

Figure 4.11

FT-IR spectra, which show all peptide bands, as well as bands from the polystyrene substrate (Figure 4.12). In contrast to the graft polymerization employing the “copolymerization approach” (see Section 4.4.1), a substantially lower amount of free polymer in solution has been isolated and the amount of monomer grafted to the surface is threefold higher. The amount of free polymer seems to depend on the experimental conditions for the preparation of the surface-bound initiator, and is expected to be lowered significantly by further optimization, for example by a reduction of the excess of Ni(COD)2 and phenanthroline in the activation step. Graft polymerization employing surfaces on which initiators have been prepared via the alloc amide precursors show great promise for the preparation of peptide brushes. The synthetic protocol is relatively simple, and in particular the number of steps requiring handling of the resin in the glove box are reduced. Moreover, the formation of free polymer in solution is also reduced.

4.4 Preparation of Peptide Brushes by Living Grafting From Polymerization

Comparative FT-IR spectra of polyglutamate grafted from polystyrene beads and polyglutamate.

Figure 4.12

Summary

Polypeptide brushes can be prepared by using both grafting to and grafting from approaches, though in both cases the strong interaction of the macrodipoles must be taken into account. These interactions are helpful and facilitate the fabrication of relatively thick films by grafting to methods, because larger aggregates present in the solution are adsorbed to the surface. The macrodipole interaction also has potential for applications of the grafted films in diagnostics. Homogeneous thick polypepide layers, however, are only accessible by using grafting from methods. The amine-initiated NCA polymerizations from solution yield peptide brushes of up to 40 nm thickness. The possibility of carrying out a renewed polymerization after the removal of physisorbed material shows that the adsorption of free polymer due to macrodipole interaction can cause physical “death” of the polymerization. The renewed polymerization allows the preparation of block copolymers. The occurrence of an inverse-piezoelectric effect showed that this grafting from technique does indeed force the a-helical chains into a parallel arrangement with a persistent polarization. Vapor deposition polymerization of NCA is

101

102

4 Synthesis of Polypeptide Brushes

another promising method, by which higher film thickness can be achieved (up to 70 nm) and block copolymerization is possible. The grafting from polymerization making use of nickel amido amidate complexes as initiators represents another very promising approach. Side reactions of the NCA can be minimized, giving a living character. The applicability of this method was established by using a polystyrene resin and a block copolymerization approach, but its major drawback is the large amount of free polymer that is produced in solution due to unbound, physisorbed initiator. The free polymer in solution can cause physical “death” of the living chain ends at the surface because of the macrodipole interaction, though improvements have been made in this respect by synthesizing the initiator directly at the surface from alloc amide precursors.

References 1 B. F. Levine, L. G. Bethea, J. Chem. Phys.

1976, 65, 1989. 2 H. Block, C. P. Shaw, Polymer 1992, 33, 2459. 3 T. Jaworek, D. Neher, G. Wegner, R. H. Wieringa, A. J. Schouten, Science 1998, 279, 57. 4 S. Machida, T. I. Urano, K. Sano, Y. Kawata, K. Sunohara, H. Sasaki, M, Yoshiki, Y. Mori, Langmuir 1995, 11, 4838. 5 S. Machida, K. Sano, K. Sunohara, Y. Kawata, Y. Mori, J. Chem. Soc., Chem. Commun. 1992, 1628. 6 T. I. Urano, S. Machida, K. Sano, Chem. Phys. Lett. 1995, 242, 471. 7 P. Doty, J. H. Bradbury, A. M. Holtzer, J. Am. Chem. Soc. 1956, 78, 947. 8 M. Goodman, E. E. Schmitt, D. Yphantis, J. Am. Chem. Soc. 1962, 84, 1288. 9 A. K. Gupta, C. Dufour, E. Marchal, Biopolymers 1974, 13, 1293. 10 A. Wada, J. Polym. Sci. 1960, 45, 145. 11 M. Niwa, M. Morikawa, T. Nabeta, N. Higashi, Macromolecules 2002, 35, 2769. 12 E. P. Enriquez, K. H. Gray, V. F. Guarisco, R. W. Linton, K. D. Mar, E. T. Samulski, J. Vac. Sci. Technol. A 1992, A10, 2775. 13 E. P. Enriquez, N. Y. Yin, R. C. Jarnagin, E. T. Samulski, Mater. Res. Soc. Symp. Proc. 1993, 292, 163. 14 C. G. Worley, R. W. Linton, E. T. Samulski, Langmuir 1995, 11, 3805.

15 S. Machida, K. Sano, H. Sasaki, M. Yoshiki,

Y. Mori, J. Chem. Soc., Chem. Commun. 1992, 1626. 16 T. I. Urano, S. Machida, K. Sano, J. Chem. Soc., Chem. Commun. 1994, 231. 17 M. Niwa, M. Morikawa, N. Higashi, Langmuir 1999, 15, 5088. 18 M. Niwa, M. Morikawa, N. Higashi, Angew. Chem. Int. Ed. Engl. 2000, 39, 960. 19 Y. Miura, S. Kimura, Y. Imanishi, J. Umemura, Langmuir 1998, 14, 2761. 20 Y. Miura, S. Kimura, Y. Imanishi, J. Umemura, Langmuir 1998, 14, 6935. 21 S. Kimura, Y. Miura, T. Morita, S. Kobayashi, Y. Imanishi, J. Polym. Sci. A., Polym. Chem. Ed. 2000, 38, 4826. 22 H. R. Kricheldorf, alpha-Aminoacid-N-Carboxy-Anhydrides and Related Heteocycles, (Ed. H. R. Kricheldorf) Springer-Verlag, Berlin, 1987, p. 170. 23 H.R. Kricheldorf, in: Models of Biopolymers by Ring-Opening Polymerization (Ed.: S. Penczek), CRC-Press, Boca Raton, Fl, 1990, 1. 24 H. Block, Poly-gamma-benzyl-L-Glutamate and other Glutamic Acid Containing Polymers, Gordon and Breach, New York, 1983, 131. 25 T. J. Deming, Nature 1997, 390, 386. 26 T. J. Deming, J. Am. Chem. Soc. 1997, 119, 2759. 27 T. J. Deming, J. Am. Chem. Soc. 1998, 120, 4240.

References 28 T. J. Deming, S. A. Curtin, J. Am. Chem. Soc. 29 30 31 32 33

34

35

36

37

38

2000, 122, 5710. M. L. C. M. Oosterling, E. Willems, A. J. Schouten, Polymer 1995, 36, 4463. E. Dietz, N. Fery, K. Haman, Angew. Makromol. Chem. 1974, 35, 115. M. L. C. M. Oosterling, E. Willems, A. J. Schouten, Polymer 1995, 36, 4485. J. K. Whitesell, H. K. Chang, Science 1993, 261, 73. J. K. Whitesell, Hye Kyung Chang, C. S. Whitesell, Angew. Chem. Int. Ed. Engl. 1994, 33, 871. A. Heise, H. Menzel, Hyun Kim, M.D. Foster, R. H. Wieringa, A. J. Schouten, V. Erb, M. Stamm, Langmuir 1997, 13, 723. A. Heise, M. Stamm, M. Rauscher, H. Duschner, H. Menzel, Thin Solid Films 1998, 327-, 199. H. Menzel, A. Heise, H. Yim, M.D. Foster, R. H., Schouten, A. J. Wieringa, in: Organic thin Films, Structure and Application (Ed.: C. W. Frank), Oxford University Press, New York, 1998, 131. R. H. Wieringa, E. A. Siesling, P. F. M. Geurts, P. J. Werkman, E. J. Vorenkamp, V. Erb, M. Stamm, A. J. Schouten, Langmuir 2001, 17, 6477. R. H. Wieringa, E. A. Siesling, P. J. Werkman, H. J. Angerman, E. J. Vorenkamp, A. J. Schouten, Langmuir 2001, 17, 6485.

39 R. H. Wieringa, E. A. Siesling, P. J. Werk-

man, E. J. Vorenkamp, A. J. Schouten, Langmuir 2001, 17, 6491. 40 L. Hartmann, T. KratzmNller, H.-G. Braun, F. Kremer, Macromol. Rapid Commun. 2000, 21, 814. 41 T. KratzmNller, D. Appelhans, H. G. Braun, Adv. Mater. 1999, 11, 555. 42 R. H. Wieringa, A. J. Schouten, Macromolecules 1996, 29, 3032. 43 Y. C. Chang, C. W. Frank, Langmuir 1998, 14, 326. 44 Y. Wang, Y. C. Chang, Langmuir 2002, 18, 9859. 45 Y. C. Chang, C. W. Frank, G. G. Forstmann, D. Johannsmann, J. Chem. Phys. 1999, 111, 6136. 46 Y. Wang, Y. C. Chang, J. Am. Chem. Soc. 2003, 125, 6376. 47 P. Witte, H. Menzel, Macromol. Chem. Phys. in press. 48 P. Witte, H. Menzel, Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 2003, 44(1), 423. 49 S. S. Wang, J. Org. Chem. 1976, 41, 3258. 50 W. A. R. Van Heeswijk, M. J. D. Eenink, J. Feijen, Synthesis Commun. 1982, 744. 51 S. Kimura, Y. Miura, T. Morita, S. Kobayashi, Y. Imanishi, J. Polym. Sci. A., Polym. Chem. Ed. 2000, 38, 4826 52 S. A. Curtin, T. J. Deming, J. Am. Chem. Soc. 1999, 121, 7427.

103

105

5

Bottle Brush Brushes: Ring-Opening Polymerization of Lactide from Poly(hydroxyethyl methacrylate) Surfaces Jong-Bum Kim, Wenxi Huang, Chun Wang, Merlin Bruening, and Gregory L. Baker

Abstract

Ring-opening polymerization of lactide was initiated from poly(hydroxyethyl methacrylate) anchored to gold surfaces using Sn(Oct)2 as the catalyst. Control experiments showed that polymerization was initiated from the hydroxy groups of poly(hydroxyethyl methacrylate) and followed first-order kinetics. The polylactide growth rate was constant and proportional to the thickness of the poly(hydroxyethyl methacrylate) substrate, consistent with initiation of lactide polymerization from the bulk of the film. Thus, the overall structure of the polymer is a surface-anchored poly(hydroxyethyl methacrylate)-g-polylactide, a “bottle-brush brush”. During lactide polymerization, the film thickness increased by 250%, which corresponds to a substantial extension of the poly(hydroxyethyl methacrylate) backbone. A side reaction involving transesterification of polylactide side chains limits the growth in film thickness at long reaction times.

5.1

Introduction

The preparation of polymer brushes on solid surfaces has been of great interest because of their potential as sensing layers, anti-corrosion layers, for controlling the wetting of surfaces, and for preparing nanostructured surfaces [1]. More than two decades ago, DeGennes developed a model for understanding polymer brush systems and suggested the synthesis of polymer brushes from surface-immobilized initiators [2]. According to Milner’s definition, polymer brushes are long-chain polymer molecules attached by one end to a surface or interface by some means, with a density of attachment points high enough so that the chains are obliged to stretch away from the interface, sometimes much farther than the typical unstretched size of a chain [3]. Initially, the preparation of polymer brushes relied on the strong, selective adsorption of one block of a block copolymer from a polymer solution. Desorption of polymer from the surface and the low areal density of adsorbed polymer chains on sur-

106

5 Bottle Brush Brushes: Ring-Opening Polymerization

faces spurred synthetic chemists to devise new methods for the synthesis of dense, well-defined ensembles of polymers covalently bound to surfaces. In order to obtain homogeneous polymer brushes, the grafting density must be high and uniform, the polydispersity (PDI) of the polymers should be near 1 [4], and all chains must be terminally grafted onto surfaces. The two principal strategies for attaching polymer brushes to surfaces are the “grafting to” technique [5], which involves the covalent tethering of preformed polymer chains from solution to a surface, and the “grafting from” technique pioneered by Sogah [6], in which polymerization occurs from surface-anchored initiators. The two methods have both advantages and disadvantages. In the “grafting to” approach, the polymer can be fully characterized in terms of its chemical structure and size prior to immobilization on the surface, but polymer– polymer interactions hinder the formation of high-density brushes. The growth of polymers from surface-bound initiators avoids limitations on the density of chains anchored to surfaces, but characterization of the chemical structure and molecular weight are more difficult. Relatively dense brushes have been prepared using free radical [7–10], cationic [11,12], anionic [13,14], ring-opening [15], and ring-opening metathesis [16–18] polymerization. With the adaptation of atom transfer radical polymerization (ATRP) to surface polymerizations [19–24], both the breadth of monomers that can be used to prepare brushes and control over the polymer architecture improved. However, one limitation of all surface growth processes is that while the number of initiating sites on surfaces can be well-characterized, their efficiency is difficult to assay. We have adopted ATRP for the growth of polymer brushes from surfaces. In an early example [25], we initiated polymerization of methyl methacrylate from initiators anchored to a layer of sputtered Au deposited on silicon wafers. After polymerization, the surface-anchored chains and initiators were cleaved from the gold surface with I2, and the cleaved products were directly analyzed by gel permeation chromatography (GPC). From the known surface area and film thickness, and by assuming that the dried film has the same density as poly(methyl methacrylate) (PMMA) homopolymer, it was found that ~1 in 10 initiator sites led to high molecular-weight polymer (>30 000 g mol–1). The fate of the remaining initiator sites was unclear, and it was surmised that they were either buried in the polymer and inaccessible to catalyst and monomer, or lost through termination and chain transfer reactions, both of which are well-known limitations of conventional free radical polymerizations. The results of further studies by the present authors [26] and others [27] point to termination as being the primary loss mechanism. This was not unexpected, as the initial state of the system is a dense, two-dimensional array of initiator sites exposed to monomer and catalyst – conditions that favor bimolecular termination via coupling and disproportionation. The implication of these results for the conformation of polymer brushes is that it will be difficult to obtain dense arrays of fully extended of chains on flat surfaces using ATRP or free radical polymerizations. Another strategy for generating extended polymer chains is to adopt a two-step approach inspired by the properties of poly(hydroxyethyl methacrylate) (PHEMA) grown from gold surfaces [28]. Aqueous polymerization of hydroxyethyl methacrylate (HEMA) is rapid, and leads to polymer brush layers approaching 1 lm in thick-

5.1 Introduction

ness. The resulting PHEMA brushes are particularly interesting as the pendent hydroxy groups are easily functionalized using standard organic chemistry. Furthermore, infra-red (IR) characterization of PHEMA films before and after functionalization shows that nearly all of the alcohols are chemically accessible and converted to products. Functionalization of the hydroxy groups of PHEMA films also increased the film thickness. As shown in Scheme 5.1, a 176 nm-thick PHEMA brush increased to 358 nm after esterification with cinnamoyl chloride.

Scheme 5.1 Extension of PHEMA brushes via functionalization. The filled squares represent groups added to the polymer using simple chemical reactions, causing a change in film thickness of Dt.

The increase in thickness is comparable to the increase in mass per repeat unit, assuming complete conversion of the hydroxy groups to cinnamate esters. Since the hydroxy groups are regularly distributed along the PHEMA chain and the cinnamate groups are relatively small, the increase in thickness must also represent an extension of the main chain of the polymer brush. Further extension of the brush could be achieved by increasing the size of the added group, and for the case reported here, by using the hydroxy group of the HEMA repeat unit to initiate a ring-opening polymerization (ROP) of a cyclic ester such as lactide (Scheme 5.2). The sequential combination of two controlled polymerization techniques, ATRP of HEMA and ROP of lactide, should lead to a bottle-brush brush – a polymer brush where each anchored polymer chain has a bottle brush architecture, with the length of the PHEMA backbone and polylactide (PLA) bristles controlled by the polymerization time. Polymerization of macromonomers [29] and ATRP from a PHEMA derivative [30] have been used previously to prepare nontethered bottle brush copolymers. There are a limited number of reports devoted to surface-initiated ROP of cyclic esters. Husemann et al. used this technique to graft poly(e-caprolactone) from hydroxy-terminated self-assembled monolayers (SAMs) as a step in a soft lithography patterning scheme [15]. Two recent examples of the surface-initiated ROP of lactides

107

108

5 Bottle Brush Brushes: Ring-Opening Polymerization

Scheme 5.2

The synthetic route to bottle-brush

brushes.

were reported by Choi [31] and MJller [32]. In both cases, PLA was initiated from -OH or -NH2 groups at the termini of long-chain alkyl thiolates on gold or silicate surfaces. These examples are comparable to polymerization from a dense layer of initiators, and should lead to linear polymers. If PHEMA films functioned similarly and initiated solely from a thin surface layer, one should expect formation of a linear block copolymer with a relatively low grafting density. A comparison of the contact angles for water on PHEMA [33], 63K, and on an x-hydroxyalkyl thiolate SAM [34], 34K, suggests that the surface density of PHEMA -OH groups is less than full surface coverage. However, based on our earlier results on the functionalization of PHEMA films, we expected that subsurface -OH groups would increase the grafting efficiency and produce a comb-like structure. To prepare the surface-anchored PLA film, a layer of PHEMA on gold was synthesized by ATRP, and then ROP was used to grow PLA from the PHEMA layer. We were able to effect controlled increases in film thickness from several nanometers up to 450 nm The PHEMA-g-PLA system on gold has several important features. Polymer layers on gold are chemically homogeneous and are compatible with a wide variety of thin film characterization techniques. Compared to thiolate on gold SAMs, PHEMA films grown from initiators anchored on gold are more thermally robust due to a small amount of transesterification during HEMA polymerization. The resulting crosslinks enhance the mechanical stability of PHEMA films, and provide the thermal stability needed to support lactide polymerization. (The Au–S bond is unstable above ~60 KC [17,35,36], and organic monolayers often detach from surfaces.) Finally, the PHEMA-g-PLA system is a biocompatible hydrogel [37] that can be used as a drug delivery system [38–41].

5.2 Synthesis of PHEMA-g-PLA

5.2

Synthesis of PHEMA-g-PLA

The synthesis of PHEMA-g-PLA tethered to a gold surface is outlined in Scheme 5.3. In a typical procedure ATRP of HEMA was initiated from the surface-bound initiator layer using a catalyst system prepared from a mixture of Cu(I)Br/tris-[2(dimethylamino)ethyl]amine (Me6TREN) (0.1 mol% based on monomer) and Cu(II)Br2/2 equiv. of 4,4¢-di-n-nonyl-2,2¢-bipyridine (dnNbpy). The Cu(II) complex (30 mol%, relative to Cu(I)) ensures the deactivation of active radicals, and also provides some control over the polymerization. The dormant a-bromocarbonyl initiator was activated by immersing substrate 1 in a 5:1 (v/v) CH3CN-THF solution of 3.5 m HEMA and the copper catalyst. Polymerizations of HEMA were run at room temperature,

Scheme 5.3

Preparation of PHEMA-g-PLA.

with the thickness of the PHEMA film controlled by the reaction time. After HEMA polymerization, the substrate was washed sequentially with DMF, EtOH, EtOAc, EtOH and deionized water, and then dried under a stream of nitrogen. The formation of PHEMA was apparent from the appearance of a carbonyl peak at 1733 cm–1 and a broad hydroxy peak at 3200–3600 cm–1 in the reflectance FT-IR spectrum (Figure 5.1, spectrum (a)). A number of PHEMA substrates were prepared to sup-

109

110

5 Bottle Brush Brushes: Ring-Opening Polymerization

Figure 5.1 Reflectance FT-IR spectra of (a) 2: a grafted PHEMA layer (174 nm) on gold; (b) 3: a grafted PHEMA (170 nm)-g-PLA (29 nm) layer prepared using 90 min of lactide polymerization; (c) 3: a grafted PHEMA (173 nm)-gPLA (127 nm) layer prepared using 180 min of lactide polymerization; (d) 3: a grafted PHEMA

(178 nm)-g-PLA (179 nm) layer prepared using 240 min of lactide polymerization; and (e) 3: a grafted PHEMA (178 nm)-g-PLA (409 nm) layer prepared with 480 min of lactide polymerization. For clarity, the 2700–3700 cm–1 region in each spectrum is magnified five-fold relative to the remainder of the spectrum.

port kinetic experiments of the ROP of lactide from PHEMA. The PHEMA films on these substrates all had similar ellipsometric thicknesses (174 € 2 nm). Using the hydroxy groups of PHEMA side chains as initiators, rac-lactide was polymerized in toluene at 95 KC using Sn(2-ethylhexanoate)2 (Sn(Oct)2) as the catalyst. The polymerizations were run in a drybox in order to avoid contamination from water, which can act as a competing initiator. At regular intervals, samples were removed for IR and film thickness measurements. The IR spectra in Figure 5.1 show the growth of PLA on PHEMA. The initial spectrum (a) showed a single carbonyl peak at 1733 cm–1 from PHEMA, but after 90 min of lactide polymerization, the carbonyl peak broadened (Figure 5.1, spectrum (b)) and shifted to higher wavenumbers. Eventually, the PLA carbonyl peak dominated the spectrum, and only a single peak at 1767 cm–1 was observed (Figure 5.1, spectrum (e)). Parallel growth in the methyl stretching peak at 2993 cm–1 and a slight decline of the hydroxy peak at 3200–3600 cm–1 relative to the methyl stretching peak also confirmed PLA formation. Note that since PLA chains prepared by the ROP of lactide are terminated by hydroxy groups, the absolute number of hydroxy groups in the brush should be conserved. The growth of PLA chains was monitored by following the change in film thickness with time. Each sample was characterized by reflectance FT-IR and ellipsometric data collected at three different spots on the sample. The net increase in thickness due to the PLA, calculated by subtracting the thickness of the PHEMA film from the total film thickness, is plotted in Figure 5.2 (line (a)). The homoge-

5.2 Synthesis of PHEMA-g-PLA

Figure 5.2 Plot of PLA thickness versus polymerization time using: (a) 174 € 2 nm-thick PHEMA substrates; and (b) 49 € 1 nm-thick PHEMA substrates. Ellipsometric thicknesses were measured at three different spots on a sample, and the error bars are smaller than the symbols.

neous ROP of lactide follows a “coordination-insertion” mechanism [42], and the data of Figure 5.2 (line (a)) show the expected linear increase in film thickness with polymerization time, thus confirming that the polymerization of lactide from PHEMA is a living process and follows first-order kinetics. The data also show the presence of a short induction time before the onset of PLA polymerization. Similar induction times, which were observed for solution polymerizations of lactide [43], may be related to the time needed to form the alkoxy-Sn complex from Sn(Oct)2 and the hydroxy group of PHEMA. Two control experiments showed that the PLA only initiates from the -OH group of the HEMA repeat unit. In the first experiment, the hydroxy groups of PHEMA were blocked by acetylation with acetyl chloride. Substrates with PHEMA and acetylated PHEMA films were placed in the same vial with rac-lactide and Sn(Oct)2 at 95 KC for 12 h. As shown in Figure 5.3, the IR spectrum of the PHEMA substrate (spectrum (b)) shows the expected changes after ROP of lactide, an increase in the carbonyl peak intensity, its shift from 1733 cm–1 to 1767 cm–1, and a slight decrease in the hydroxy peak at 3200–3600 cm–1. However, the IR spectrum of the substrate with acetylated PHEMA shows no sign of PLA growth (spectrum (d)). Similarly, the behavior of substrates with PHEMA and PMMA [17] films were compared using the same conditions as before. As seen in Figure 5.4, the polylactide grew from the PHEMA layer, while the PMMA control showed no PLA growth. Instead of growing PLA, a decrease in IR intensity suggests that some of PMMA chains desorbed from surface under the polymerization conditions. PLA was grown from PHEMA layers of different thicknesses, with the growth rate being extracted from the linear portion of thickness versus time curves (Figure 5.2). PLA grew at a rate of 9.8 O min–1 from a 174 nm-thick PHEMA film, and at 2.4 O min–1 from a 49 nm-thick PHEMA film. The ratio of the rates (4.1) is similar to the ratio of the film thicknesses (3.6), which suggests that the -OH groups of

111

112

5 Bottle Brush Brushes: Ring-Opening Polymerization

Figure 5.3 Reflectance FT-IR spectra of: (a) a surface-grafted PHEMA layer on gold; (b) after lactide polymerization from PHEMA; (c) a surface-grafted acetylated-PHEMA layer; and (d) after attempted lactide polymerization from acetylatedPHEMA.

Figure 5.4 Reflectance FT-IR spectra of: (a) a surface-grafted PMMA layer on gold; (b) after attempted lactide polymerization from PMMA; (c) a surface-grafted PHEMA layer; and (d) after lactide polymerization initiated from PHEMA.

PHEMA are accessible, and that ROP of lactide was initiated from most -OH groups. If initiation were restricted to a thin surface layer of -OH groups, then the rate of film growth should be independent of the PHEMA thickness. Acetylation studies also are consistent with the view that most of the -OH groups of PHEMA are accessible to reagents. As shown in Figure 5.3, spectrum (c), the absence of a signifi-

5.2 Synthesis of PHEMA-g-PLA

cant -OH peak in the FT-IR spectrum of PHEMA films after reacting with acetyl chloride indicates near-complete acylation of PHEMA. Thus, the structure of PHEMA-g-PLA should be viewed as a graft copolymer with comb-like character. As ROP of lactide continues, the tethered PHEMA chains must extend to accommodate the volume occupied by the PLA. Eventually, PHEMA chains would fully extend and, for steric reasons, PLA growth in the interior of the film would cease. Kinetically, this transition should be manifested as a sharp decrease in the rate of ROP of lactide to that which corresponds to polymer growth from a surface layer. Using PHEMA substrates of 49 € 1 nm thickness, we extended the polymerization time to test for such a change in growth rate at long times. As seen in Figure 5.2(b), PLA growth eventually stopped at ~10 h; however, the 130-nm increase in thickness seemed too small to be ascribed to full extension of the PHEMA chain, given our estimate of ~10% initiator efficiency as in the related PMMA system. By assuming a cross-sectional area for PHEMA which was twice that of PMMA, we expected to realize a fully extended PHEMA chain when the increase in thickness was >250 nm. This corresponds to a degree of polymerization for lactide of at least five. Further investigation showed that the termination of growth seen in Figure 5.2(b) is consistent with equilibrium polymerization, where the propagation and depropagation rates are identical [44–46]. The polymerization solution was slightly viscous, and characterization of the solution by 1H NMR revealed a 79:21 mixture of soluble PLA homopolymer and monomer. The concentration of monomer calculated from the NMR data (~0.03 M) is in good agreement with literature values for the equilibrium monomer concentration in lactide polymerizations: for polymerization of l-lactide in 1,4-dioxane, [M]eq ~0.15 M at 406 KK and 0.06 M at 353 KK [44]. To prove equilibrium control, we isolated a sample that had grown to a limiting PLA film thickness (131 nm), washed it with solvent, and then transferred it to a vial contain-

Scheme 5.4

Intermolecular transesterification of PHEMA-g-PLA.

113

114

5 Bottle Brush Brushes: Ring-Opening Polymerization

ing fresh monomer solution with catalyst. As expected, PLA growth resumed, and an additional 53 nm of PLA was added over a 5-h period. The growth of polymer in solution requires free initiator, but the control experiments described earlier showed no evidence for PLA in solution. The initiation of ROP of lactide by adventitious H2O in the solvent or on the surface of the glassware seems unlikely, given the control experiments and our efforts to exclude H2O by working in a drybox, using anhydrous solvents, and silanizing reaction vials – all conditions that lead to successful living anionic polymerizations. Since Sn(Oct)2 also catalyzes transesterification reactions, a more plausible explanation is that transesterification involving neighboring chains forms crosslinks and releases the tin alkoxide species into solution where it can further initiate polymerization of lactide (Scheme 5.4).

5.3

Conclusions and Implications for Future Studies

We successfully initiated ROP of lactide from PHEMA brushes and observed an apparent 250% elongation of the PHEMA chain. The observed growth in thickness is consistent with our earlier estimates that ~1 in 10 initiators in SAMs lead to high molecular-weight polymer, and suggests that the conformation of methacrylate brushes grown by ATRP falls between that of random coils and fully elongated chains. At 250% elongation, the kinetic data indicate that the PHEMA chains are not yet fully extended. Establishment of an equilibrium between monomer and polymer in the polymerizing solution is consistent with the film thickness being limited by transesterification reactions involving the PHEMA side chains. Preliminary data show that isolation of the bottle brush brushes and reinitiating polymerization with fresh monomer will enable further extension of the chains. The full characterization of PHEMA-based bottle brush brushes is hindered by crosslinking, presumably via transesterification, during the growth of PHEMA films. When treated with I2, PHEMA films detach from the surface as coherent films that swell, but do not dissolve. The use of a HEMA monomer with a masked hydroxy group should preclude crosslinking, but the monomer must be chosen carefully in order to retain the potential advantages of aqueous polymerization [47]. The choice of polymer used for the bristles of the bottle brush brushes is also important. For polymerization from gold substrates, low- temperature polymerizations are needed to avoid thermally induced desorption of polymers from the surface. In addition, the high concentration of growing chains in bottle brush schemes precludes the use of free radical and other polymerization techniques that have substantial contributions from bimolecular termination and chain transfer reactions.

5.4 Experimental Section

5.4

Experimental Section 5.4.1

Materials

THF (Aldrich, 99%) and toluene (Aldrich, 99.5%) were distilled from calcium hydride followed by a second distillation from sodium/benzophenone ketyl. 3,6-Dimethyl-1,4-dioxane-2,5-dione (lactide; Aldrich, reagent grade) was recrystallized three times in ethyl acetate. Other ACS reagent grade solvents were used as received without further purification. Gold-coated Si wafers (200 nm of gold sputtered on 20 nm of Cr on Si(100) wafers) were cleaned in a UV/O3 chamber for 15 min, immersed in deionized water for 15 min, and dried under a flow of N2 just before use. 5.4.2

Preparation of Monomer Solution and Substrates

The monomer solution was prepared in a drybox as follows. Lactide was added to toluene (120 mL) and vigorously stirred at room temperature for 1 h using a magnetic stirrer. Stirring was stopped and the clear solution (determined to be ~0.13 M by gravimetry) was decanted from undissolved lactide. Sn(Oct)2 (0.0844 g) was added to the lactide solution (100 mL) to give a final catalyst concentration of ~2.1 mM. The initial organic monolayer, 1, was prepared using a literature procedure [25]. PHEMA substrate, 2, was synthesized by surface-initiated atom transfer radical polymerization. 5.4.3

Ring-Opening Polymerization from PHEMA Surface

ROP of lactide was carried out in a drybox filled with helium. The PHEMA substrates, 2, were immersed in the solution of lactide and Sn(Oct)2 catalyst. Silanized glass vials (15 mL) were used as reaction vessels. Polymerizations were run at 95 KC in an oil-bath, and samples were removed at intervals for kinetic studies. As the samples were taken out from the solution, they were sequentially immersed in a series of vials containing toluene, ethyl acetate, ethanol, and deionized water to clean the surface. The surface was dried under a flow of N2 and characterized by reflectance FT-IR and ellipsometry. 5.4.4

Analytical Methods

Reflectance FT-IR spectroscopy was performed using a Nicolet Magna-IR 560 spectrometer containing a PIKE grazing angle (80K) attachment. Ellipsometric measurements on polymer films were obtained with a rotating analyzer ellipsometer (model M-44; J. A. Woollam) at a 75K angle of incidence.

115

5 Bottle Brush Brushes: Ring-Opening Polymerization

116

Acknowledgments

The authors thank the NSF Center for Sensor Materials at Michigan State University for financial support of this research.

References 1 B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2 3 4 5 6 7

8 9 10

11 12 13 14

15

16

17

2000, 25, 677. P. G. De Gennes, Macromolecules 1980, 13, 1069. S. T. Milner, Science 1991, 251, 905. N. Dan, M. Tirrell, Macromolecules 1993, 26, 6467. K. Bridger, B. Vincent, Eur. Polym. J. 1980, 16, 1017. W. R. Hertler, D. Y. Sogah, F. P. Boettcher, Macromolecules 1990, 23, 1264. O. Prucker, C. A. Naumann, J. RThe, W. Knoll, C. W. Frank, J. Am. Chem. Soc. 1999, 121, 8766. B. Peng, D. Johannsmann, J. RThe, Macromolecules 1999, 32, 6759. M. Niwa, M. Date, N. Higashi, Macromolecules 1996, 29, 3681. B. de Boer, H. K. Simon, M. P. L. Werts, E. W. van der Vegte, G. Hadziioannou, Macromolecules 2000, 33, 349. R. Jordan, A. Ulman, J. Am. Chem. Soc. 1998, 120, 243. B. Zhao, W. J. Brittain, J. Am. Chem. Soc. 1999, 121, 3557. N. Tsubokawa, H. Ueno, J. Appl. Polym. Sci. 1995, 58, 1221. R. Jordan, A. Ulman, J. F. Kang, M. H. Rafailovich, J. Sokolov, J. Am. Chem. Soc. 1999, 121, 1016. M. Husemann, D. Mecerreyes, C. J. Hawker, J. L. Hedrick, R. Shah, N. L. Abbott, Angew. Chem., Int. Ed. 1999, 38, 647. W. A. Herrmann, A. W. Stumpt, T. Priermeier, S. Bogdanovic, V. Dufaud, J.-M. Basset, Angew. Chem., Int. Ed. Engl. 1997, 35, 2803. N. L. Jeon, I. S. Choi, G. M. Whitesides, N. Y. Kim, P. E. Laibinis, Y. Harada, K. R. Finnie, G. S. Girolami, R. G. Nuzzo, Appl. Phys. Lett. 1999, 75, 4201.

18 M. R. Buchmeiser, F. Sinner, M. Mupa,

K. Wurst, Macromolecules 2000, 33, 32. 19 M. Husseman, E. E. Malmstroem, M. McNa-

mara, M. Mate, D. Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang, T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424. 20 X. Huang, M. J. Wirth, Macromolecules 1999, 32, 1694. 21 M. Ejaz, S. Yamamoto, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules 1998, 31, 5934. 22 X. Huang, L. J. Doneski, M. J. Wirth, Anal. Chem. 1998, 70, 4023. 23 K. Matyjaszewski, P. J. Miller, N. Shukla, B. Immaraporn, A. Gelman, B. B. Luokala, T. M. Siclovan, G. Kickelbick, T. Vallant, H. Hoffmann, T. Pakula, Macromolecules 1999, 32, 8716. 24 R. R. Shah, D. Merreceyes, M. Husemann, I. Rees, N. L. Abbott, C. J. Hawker, J. L. Hedrick, Macromolecules 2000, 33, 597. 25 J.-B. Kim, M. L. Bruening, G. L. Baker, J. Am. Chem. Soc. 2000, 122, 7616. 26 J. B. Kim, W. X. Huang, M. D. Miller, G. L. Baker, M. L. Bruening, J. Polym. Sci., Part A: Polym. Chem. 2003, 41, 386. 27 D. Q. Xiao, M. J. Wirth, Macromolecules 2002, 35, 2919. 28 W. Huang, J.-B. Kim, M. L. Bruening, G. L. Baker, Macromolecules 2002, 35, 1175. 29 H. G. Borner, K. Beers, K. Matyjaszewski, S. S. Sheiko, M. Moller, Macromolecules 2001, 34, 4375. 30 S. C. Hong, D. Neugebauer, Y. Inoue, J. F. Lutz, K. Matyjaszewski, Macromolecules 2003, 36, 27. 31 I. S. Choi, R. Langer, Macromolecules 2001, 34, 5361. 32 M. MJller, F. Nederberg, L. S. Lim, R. KWnge, C. J. Hawker, J. L. Hedrick, Y. Gu, R. Shah,

References

33 34 35

36 37 38

39

N. L. Abbott, J. Polym. Sci,. Part A: Polym. Chem. 2001, 39, 3529. H. Lim, Y. Lee, S. Han, J. Cho, K.-J. Kim, J. Vac. Sci. Technol., A 2001, 19, 1490. A. G. Frutos, J. M. Brockman, R. M. Corn, Langmuir 2000, 16, 2192. C. D. Bain, E. B. Troughton, Y. T. Tao, J. Evall, G. M. Whitesides, R. G. Nuzzo, J. Am. Chem. Soc. 1989, 111, 321. J. B. Schlenoff, M. Li, H. Ly, J. Am. Chem. Soc. 1995, 117, 12528. D. W. Lim, S. H. Choi, T. G. Park, Macromol. Rapid Commun. 2000, 21, 464. I. Barakat, P. Dubois, C. Grandfils, R. Jerome, J. Polym. Sci., Part A: Polym. Chem. 1999, 37, 2401. M. Iijima, Y. Nagasaki, T. Okada, M. Kato, K. Kataoka, Macromolecules 1999, 32, 1140.

40 T. Riley, S. Stolnik, C. R. Heald, C. D. Xiong,

41

42 43 44 45 46 47

M. C. Garnett, L. Illum, S. S. Davis, S. C. Purkiss, R. J. Barlow, P. R. Gellert, Langmuir 2001, 17, 3168. I. Barakat, P. Dubois, R. Jerome, P. Teyssie, E. Goethals, J. Polym. Sci., Part A: Polym. Chem. 1994, 32, 2099. K. Stridsberg, M. Ryner, A.-C. Albertsson, Macromolecules 2000, 33, 2862. M. Yin, G. L. Baker, Macromolecules 1999, 32, 7711. A. Duda, S. Penczek, Macromolecules 1990, 23, 1636. K. Majerska, A. Duda, S. Penczek, Macromol. Rapid Commun. 2000, 21, 1327. H. Nishida, M. Yamashita, T. Endo, Y. Tokiwa, Macromolecules 2000, 33, 6982. D. M. Jones, W. T. Huck, Adv. Mater. 2001, 13, 1256.

117

119

6

Preparation of Well-Defined Organic-Inorganic Hybrid Nanostructures using Living Cationic Surface-Initiated Polymerization from Silica Nanoparticles Il-Jin Kim, Su Chen, and Rudolf Faust

Abstract

Structurally well-defined polymer/inorganic nanocomposites were prepared by surface-initiated living cationic polymerization of isobutylene (IB). The living cationic polymerization of IB was initiated from initiators self-assembled on the surface of silica nanoparticles in the presence of additional soluble “free initiator” with TiCl4 in methylcyclohexane:CH3Cl (60:40, v/v) at –80 -C. The polymerization displayed the diagnostic criteria for living cationic polymerization, and provided densely grafted polymers of controlled molecular weight with an approximate graft density of 3.3 chains per nm2. The surface-initiated polymerization of IB without added “free initiator” also yielded grafted polymer chains with good molecular weight control and narrow molecular weight distribution (Mw/Mn). The polymer-nanoparticle hybrids were characterized by thermogravimetry, gel permeation chromatography, and dynamic light scattering measurements.

6.1

Introduction

Organic-inorganic nanocomposites have been attracting much attention due to their potential applications, such as colloid stabilizers, electro-optical devices, and nanocomposite materials [1–4]. There are several techniques for attaching polymer chains to nanoparticle surfaces, including chemisorption [5], covalent attachment of end-functionalized polymers to a reactive surface (“grafting to”) [6], and in-situ monomer-by-monomer growth of polymer chains from immobilized initiators (“grafting from”) [7,8]. Among these methods, the “grafting from” approach offers the most promising route to achieve maximum structural control. For the preparation of polymer brushes, a living polymerization technique is the most desirable, due to its feasibility to control the molecular weight, molecular weight distribution and composition of surface-initiated polymers. Recently, several research groups have reported the use of controlled/living polymerizations to fabricate well-defined polymer-nanoparticle hybrid materials by anionic [9], radical [10,11], and ring-open-

120

6 Preparation of Well-Defined Organic-Inorganic Hybrid Nanostructures

ing polymerization (ROP) [12]. Living cationic surface-initiated polymerization (LCSIP), however, remains relatively unexplored for the preparation of nanocomposite materials. Recently, we have reported that polymer brushes with high graft density could be prepared by employing self-assembled covalently attached initiators on flat silicon substrates in the presence of added free initiator for the living cationic polymerization of isobutylene [13]. Since a flat silica surface has a low surface area, corresponding to low concentration of initiating sites on the substrates, it was necessary to add a predetermined amount of “free initiator” to the reaction mixture to maintain controlled growth of polymer from the flat surface. The direct synthesis of polymer brush on the surface might be interesting by surface-initiated living polymerization without using sacrificial free initiator if we choose high surface area substrates. In this chapter, we report preliminary results on the synthesis of well-defined polymernanoparticle hybrid materials by living cationic surface-initiated polymerization, and the effects of free initiator on the polymerization kinetics and molecular control.

6.2

Experimental Section 6.2.1

Materials

Methyl chloride (MeCl) and isobutylene (IB) were dried in the gaseous state by passing them through in-line gas-purifier columns packed with BaO/Drierite. They were condensed in the cold bath of a glovebox prior to polymerization. Methylcyclohexane (MeChx, 99% anhydrous; Aldrich), 2,6-di-tert-butylpyridine (DTBP, 99.4% by GC; Aldrich), and titanium(IV) chloride (TiCl4, 99.9%; Aldrich) were used as received. Spherical silica nanoparticle sol (organo silicasol MEK-ST; Nissan Chemical Industries) dispersed in methylethyl ketone, MEK (30 wt.%, diameter = 21.8 nm, specific surface area = 202 m2 g–1) was dried over CaH2 in a Soxhlet extractor for 48 h at room temperature under a vacuum providing gentle reflux. The average particle diameter in the anhydrous silica MEK sol (37.15 wt.%) was 24.6 nm, determined by Microtrac UPA. 6.2.2

Characterization

Molecular weights were measured using a Waters HPLC system equipped with a Model 510 HPLC pump, a Model 250 dual refractometer/viscometer detector (Viscotek), a Model 486 UV/Vis detector, a Model 712 sample processor, and five ultra-Styragel GPC columns connected in the following series: 500, 103, 104, 105, and 100 L. Tetrahydrofuran (THF) was used as eluent at a flow rate of 1.0 mL min–1. The particle sizes and distributions were measured by dynamic light scattering (DLS) using a

6.2 Experimental Section

Microtrac UPA analyzer. The samples of PIB nanocomposites were dissolved in THF to adjust the solid content to around 0.5 wt.% and placed directly in the cell. The temperature of the cell was kept at around 25 -C, and the measuring time was in the range of 60 to 180 s. 6.2.3

Synthesis of Immobilized Macroinitiators

Spherical silica nanoparticles with a narrow size distribution and an average diameter of 24.6 nm, determined by DLS measurement, were used to prepare polymernanoparticle hybrids. The immobilization of chlorosilyl functional initiator, 3-(1chlorodimethyl-silylmethyl)ethyl-1-(1-chloro-1-methyl)ethyl-benzene (CECE) on the particle surface was accomplished by adding CECE (6 mM) dissolved in anhydrous CH2Cl2 containing DTBP (6 mM) into a 250-mL flask, followed by addition of silica sol in MEK. The reaction mixture was stirred vigorously at room temperature overnight in a glovebox. After the condensation reaction of CECE with silanol groups on the surface of the silica nanoparticle, dried MeChx was added to the mixture and the solvent system of CECE-modified nanoparticles was changed into good solvent (MeChx) for the living cationic polymerization via vacuum distillation. CECE-modified nanoparticles were fully resuspended in MeChx without losing some “clumped” materials. Unreacted residual free initiator was removed by repeated suspension and ultracentrifugation (six times). To the resulting surface-modified silica nanoparticles additional “free initiator” (0.83 mM) was added. DLS measurement of CECEmodified particles showed an average diameter of 26.7 nm; thus, the particles remained nonagglomerated. A slight increase in the average diameter with respect to the starting silica nanoparticles was within the measurement error.

Scheme 6.1

Synthetic scheme of structurally well-defined PIB-silica nanoparticle hybrids.

121

122

6 Preparation of Well-Defined Organic-Inorganic Hybrid Nanostructures

6.3

Results and Discussion 6.3.1

Living Cationic Surface-Initiated Polymerization of IB from Silica Nanoparticles in the Presence of Sacrificial Free Initiator

Our initial studies to control polymer chain growth from the silica macroinitiators without added “free initiator” were successful due to the sufficiently high concentration of initiating sites. When the particle size is much larger, however, the addition of a predetermined amount of “free initiator” to the reaction mixture is desirable [10,11], to ensure molecular weight control. It is generally assumed that the molecular weights of the surface-bound and free polymer are identical. To determine if this is also valid in our case, the surface-initiated polymerization of IB was carried out using silica nanoparticle macroinitiators (4.46 g L–1) in conjunction with “free initiator” CECE (0.85 mM), TiCl4 (0.072 M) and DTBP (3 mM) in MeChx:CH3Cl (60:40, v/v) at –80 -C (Scheme 6.1). The polymerization was then started by injection of IB (1.43 M). After a predetermined reaction time, samples were quenched with anhydrous and prechilled methanol. Since the free PIB formed in the solution is extremely sensitive to moisture, the chloro head-group of polymer chain was replaced with stable butyl-group by addition of nBuLi. Isolation of free PIB (60 wt.%) from surface-bound PIB (40 wt.%) was achieved by a series of extraction and ultracentrifugation steps in hexanes (10 times). Thermogravimetric analysis (TGA) of PIB hybrid nanoparticles revealed that the amount of grafted polymer on the silica particle surface was ca. 89.5% by mass. In contrast to most published examples, where the nanoparticles are the major component (ca. 80~96 wt.%), the Results for the polymerization of IB with chlorosilyl functional initiator (CECE)-modified silica nanoparticlesa).

Table 6.1

Reaction time (min)

Conversion (%)

Mnb)

Mw/Mnb)

Mnc)

Mw/Mnc)

10 20 30 45 60 75

47.14 77.83 88.01 96.01 100 100

32 48 54 56 56 56

1.69 1.42 1.40 1.41 1.40 1.39

36 56 58 62 66 65

1.68 1.41 1.34 1.36 1.32 1.29

a)

b) c)

300 400 800 500 700 800

Reaction conditions: [TiCl4] = 72 mM; [DTBP] = 3 mM; [CECE] = 0.85 mM; [Silica Macroinitiators] = 0.48 mM; [IB] = 1.43 M; MeChx:CH3Cl (60:40, v/v) at –80 -C. Polymer formed from the additional “free initiator” in the solution, obtained by GPC with standard polystyrene as reference. Surface-bound polymer cleaved from the silica nanoparticle surface by etching with aqueous HF (5%) for 6 h at room temperature.

300 400 300 900 000 200

6.3 Results and Discussion

PIB-silica nanoparticle hybrids in our system consisted of 89.5 wt.% PIB and 10.5 wt.% silica particle. These hybrids were inorganic nanoparticle-modified polymers rather than polymer-modified nanoparticles. In order to gain insight into the polymer growth characteristics of the surface-initiated polymerization, the surface-bound PIB was cleaved from the silica nanoparticle surface by etching with aqueous HF (5%) for 6 h at room temperature. The results of characterization for both the cleaved graft polymer and the free polymer simultaneously formed from the additional “free initiator” are summarized in Table 6.1. It can be seen that the molecular weights (Mn) of the cleaved graft polymer and the free polymer are nearly identical, both increasing with increasing monomer conversion. The molecular weight distributions (Mw/Mn) remain fairly low, with increasing monomer conversion. All these results are in good agreement with previous reports [10,11]. Comparison of GPC traces of cleaved PIB and free PIB is shown in Figure 6.1. On the basis of weight percent of surface-bound PIB (40 wt.%) and PIB molecular weight, we calculated a grafting density of 3.3 PIB chains per nm2. This grafting density is very high, and similar to those reported from Patten et al. where grafting densities of approximately 2–5 initiators per nm2 were obtained [10]. This result can be compared to the generally accepted hydroxyl group density of five hydroxyl groups per nm2 on the silica surface [14]. From the grafting density, the concentration of silica-bound initiator (that actually initiates polymerization) of 0.48 mM can be calculated. To examine the living characteristics of the polymerization, ln([M]o/[M]) versus time and Mn versus conversion plots were constructed. The first-order kinetic plot of ln([M]o/[M]) versus reaction time for the polymerization of IB in the presence of silica macroinitiators and free initiator is depicted in Figure 6.2. The semilogarithmic dependence ln([M]o/[M]) on time is straight-line from almost the very beginning of the polymerization, which indicates that the concentration of the propagating spe-

Figure 6.1 Comparison of GPC traces of cleaved PIB and free PIB. Polymerization conditions: [Silica macroinitiators] = 0.48 mM; [CECE] = 0.85 mM; [TiCl4] = 0.072 M; [DTBP] = 3 mM; [IB] = 1.43 M in MeChx:CH3Cl (60:40, v/v) at –80 8C.

123

124

6 Preparation of Well-Defined Organic-Inorganic Hybrid Nanostructures

First-order kinetic plot of ln([M]o/ [M]) versus time for the polymerization of IB with added “free initiator” from the surface of 20-nm CECE-functionalized silica. Conditions:

Figure 6.2

[Silica macroinitiators] = 0.48 mM; [CECE] = 0.85 mM; [TiCl4] = 72 mM; [DTBP] = 3 mM; [IB] = 1.43 M in MeChx:CH3Cl (60:40, v/v) at –80 8C.

cies is constant (absence of irreversible termination) throughout the course of both solution and surface-initiated polymerization. The number-average molecular weight (Mn) of free PIB and silica-bound PIB versus conversion plots are shown in Figures 6.3 and 6.4, respectively. The linear plot of Mn versus conversion plot demonstrates the absence of chain transfer. The molecular weight (Mn) of the grafted polymer and free PIB, as determined using GPC, are close to the theoretical line calculated from the total initiator concentration (0.48 mM + 0.86 mM) (the solid lines in Figures 6.3 and 6.4). The corresponding Mw/Mn values decrease with increased conversion, as is generally encountered in controlled/living cationic polymerization. Figure 6.5 summarizes an ability to control the hydrodynamic diameter of the polymer chains (as measured by DLS) by varying the molecular weight (Mn). In agreement with kinetic studies, the thickness of the PIB chains is a linear function of the degree of the polymerization of the polymer formed on the silica surface, corresponding to the growth of a layer of densely grafted polymer chains.

Mn (j) and Mw/Mn (s) of PIB obtained from free initiator versus conversion plots for the polymerization of IB with added “free initiator” from the surface of 20-nm CECE-functionalized silica.

Figure 6.3

6.3 Results and Discussion

Figure 6.4 Mn (d) and Mw/Mn (s) of cleaved PIB from silica surface versus conversion plot for the polymerization of IB with added “free initiator” from the surface of 20-nm CECE-functionalized silica.

Figure 6.5 Variation in hydrodynamic diameter of PIB-nanoparticle hybrid with molecular weight of PIB formed from the silica surface.

6.3.2

Living Cationic Surface-Initiated Polymerization of IB from Silica Macroinitiators

Based on the polymer growth characteristics of the surface-initiated polymerization and the graft density of polymer chain from the silica nanoparticles in the presence of free initiator described above, it is possible to design well-defined polymerization system without “sacrificial” free initiator. Contrary to flat substrates and the large spherical particles, silica particles with a sufficiently small diameter can provide a sufficiently high concentration of initiating site in polymerizations and allow the polymerization to proceed in a controlled fashion. First, the surface-initiated polymerization of IB from the 20-nm silica nanoparticle macroinitiators (4.46 g L–1) was carried out using the same procedure as described above, except that the surfacebound initiators were only used as initiator. The results of the surface-initiated polymerization of IB are summarized in Table 6.2. Polymerization of IB from silica nanoparticles exhibited good molecular control

125

126

6 Preparation of Well-Defined Organic-Inorganic Hybrid Nanostructures Summary of molecular weight and molecular weight distribution for the surface-initiated polymersa).

Table 6.2

Sample

Reaction time (min)

Conversion (%)

Mnb)

Mw/Mnb)

1 2 3 4 5

5 10 20 30 45

65.18 92.10 100 100 100

25 38 39 42 47

1.56 1.39 1.41 1.38 1.28

a)

b)

000 800 500 800 900

Reaction conditions: [TiCl4] = 72 mM; [DTBP] = 3 mM; [IB] = 1.43 M; [Silica Macroinitiators] = 2 mM; MeChx:CH3Cl (60:40, v/v) at –80 -C. Polymer formed from silica nanoparticle macroinitiators.

and narrow molecular weight distribution, with increased conversion for the grafted polymer samples. The first-order kinetic plot of monomer conversion for the polymerization of IB with immobilized initiators under the specified condition is shown in Figure 6.6. The linear relationship reveals that the number of active chains is constant (absence of irreversible termination). It can be also observed that the rate of polymerization, as calculated from the slope of the figure, in the absence of free initiator is slightly higher than that in the presence of the free initiator. Figure 6.7 shows Mn and Mw/Mn of the cleaved graft PIB from the silica nanoparticle surface based on monomer conversion. The linear plot of Mn versus conversion plot demonstrates the absence of chain transfer. The experimental molecular weights of the cleaved graft polymer exhibited close to theoretical Mn (Mn theoretical = 40 000, Mn experimental = 43 000) indicating close to 90% initiator efficiency (Ieff ) and narrow molecular weight distributions reaching 1.3~1.4 at complete IB conversion. This implies that most of the initiator sites on the nanoparticle surface initiated the growth of polymer chains. Since polymer growth from the high surface area substrates (i.e., small-diameter spheres and highly porous materials) might be quite dif-

First-order kinetic plot of ln([M]o/[M]) versus time for the polymerization of IB with silica-bound macroinitiators in MeChx:CH3Cl (60:40, v/v) solvent mixture at –80 8C.

Figure 6.6

6.3 Results and Discussion

Mn (d) and Mw/Mn (s) versus conversion plots in the polymerization of IB with silica-bound macroinitiators in MeChx:CH3Cl (60:40, v/v) solvent mixture at –80 8C.

Figure 6.7

ferent from growing polymers from low surface area substrates (i.e., flat surface, large-diameter spheres, and low-porous materials), it is possible to reduce the steric hindrance of growing polymer chain on the curved surface of 20-nm diameter particle [15]. All of these results confirm that the polymerization of IB from the 20 nmdiameter particle surface exhibited the characteristics of a living polymerization.

Summary

We presented the first report on surface-initiated cationic polymerization of IB from functionalized silica nanoparticles. Structurally well-defined PIB/nanoparticle hybrid materials can be prepared by using silica macroinitiators in the presence of added “free initiator”. The polymerization displayed the diagnostic criteria for living cationic polymerization and provided densely grafted polymers of controlled molecular weight with an approximate graft density of 3.3 chains per nm2. Due to a high surface area of the 20 nm-diameter particles, the surface-initiated polymerization of IB from the immobilized macroinitiators without added “free initiator” also exhibited good molecular weight control and narrow molecular weight distribution. Efforts are under way to extend the utility of LCSIP to other homo- and copolymers.

127

6 Preparation of Well-Defined Organic-Inorganic Hybrid Nanostructures

128

References 1 E. P. Giannelis, R. Krishnamoorti, E. Man2 3 4

5 6 7 8 9

ias, Adv. Polym. Sci. 1999, 138, 107–147. M. Baum, W. J. Brittain, Macromolecules 2002, 35, 610–615. J. Pyun, K. Matyjaszewski, Chem. Mater. 2001, 13, 3436–3448. S. G. Boyes, W. J. Brittain, X. Weng, S. Z. D. Cheng, Macromolecules 2002, 35, 4960–4967. P. Auroy, L. Auvray, L. Leger, J. Colloid Interface Sci. 1992, 150, 187–194. J. M. Jethmalani, W. T. Ford, Chem. Mater. 1996, 8, 2138–2146. O. Prucker, J. RShe, Macromolecules 1998, 31, 592–601. O. Prucker, J. RShe, Macromolecules 1998, 31, 602–613. Q. Zhou, S. X. Wang, X. W. Fan, R. Advincula, J. Mays, Langmuir 2002, 18, 3324–3331.

10 T. von Werne, T. Patten, J. Am. Chem. Soc.

2001, 123, 7497–7505. 11 M. Husseman, E. E. Malmstroem, M. McNa-

mara, M. Mate, D. Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang, T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424–1431. 12 R. Jordan, N. West, A. Ulman, Y.-M. Chou, O. Nuyken, Macromolecules 2001, 34, 1606– 1611. 13 I.-J. Kim, R. Faust, J. Macromol. Sci., Pure Appl. Chem. 2003, 40(10), 991–1008. 14 S. A. Asher, J. Holtz, J. Weissman, G. Pan, Mater. Res. Soc. Bull. 1998, 23, 44–50. 15 R. Jordan, A. Ulman, J. F. Kang, M. H. Rafailovich, J. Sokolov, J. Am. Chem. Soc. 1999, 121, 1016–1022.

129

7

Photoinitiated Polymerization from Self-Assembled Monolayers Daniel J. Dyer, Jianxin Feng, Charles Fivelson, Rituparna Paul, Rolf Schmidt, and Tongfeng Zhao

Abstract

A review of recent results on photochemical surface-initiated polymerization is presented. The synthesis of polymer brushes by “grafting from” approaches is discussed, with an emphasis on free radical initiating systems that are similar in structure to 2,2¢-azobisisobutyronitrile. Three different thio-based initiators were discussed, and the grafting kinetics from gold substrates examined. It was found that the growth kinetics for polystyrene brushes exhibits a multi-stage behavior that is consistent with the gel, or Trommsdorf effect. Film thicknesses from 40 to 200 nm thickness were observed, and the molecular weight was found to be relatively independent of film thickness. Grafting densities as high as 1.8 nm2 per chain were observed.

7.1

Introduction

The design and synthesis of functional organic thin films is an important goal in modern polymer science. Polymer brushes offer a unique approach to the synthesis of well-defined structures with controlled functionality [1–4]. In particular, “grafting from” (GF) strategies, or surface-initiated polymerization (SIP), offer distinct advantages over alternative modes of deposition such as spin casting or the “grafting to” (GT) approach. For instance, a cast film is merely adsorbed, or physisorbed, to a surface and may delaminate under various conditions, particularly in organic solvents. Polymers that are “grafted to” a substrate are more robust and may stretch away from the substrate when the grafting density is high. However, the GT technique typically yields low-density polymer brushes because, once grafted, the chains inhibit the diffusion of additional reactive polymers to the active functional groups at the surface. In contrast, the GF technique utilizes a polymer initiator that is covalently linked to the surface so that polymer grows out away from the substrate and always remains tethered (Figure 7.1). These inorganic/organic polymer composite films will play an important role in many

130

7 Photoinitiated Polymerization from Self-Assembled Monolayers

Figure 7.1

Grafting from strategy for synthesizing polymer brushes from self-assembled monolayers.

emerging fields, including the design of novel biomaterials [5,6] for tissue engineering [7,8], drug delivery [8–13], implants and cell adhesion [14–16], and protein recognition [17]. Other areas that will be affected by this technology include adhesion and wetting [17–19], microfluidics [20], microfabrication [19,21,22], molecular recognition [17,23], chemical sensing [21,24], and organic synthesis [22,23,25]. Self-assembled monolayers (SAMs) offer a convenient approach to modify inorganic substrates such as gold, glass, or oxidized silicon [26]. In the SAM technique, a pre-cleaned inorganic substrate is immersed into a dilute organic solution of reactive molecules that assemble onto the substrate and form covalent bonds. Typically, these are amphiphilic molecules with long chains and a polar head group that has an affinity for the inorganic substrate. For example, alkyl thiols will form S-Au bonds, yielding densely packed monolayers when the alkyl chains are greater than ten carbons in length; similarly, glass and silicon may be modified with chlorosilanes or alkoxysilanes. The SAM formation is confirmed by a variety of techniques including contact angle measurements [27], ellipsometry [28,29], reflection-absorption infrared spectroscopy (RAIRS) [30], surface plasmon resonance (SPR) [30,31], and X-ray photoelectron spectroscopy (XPS) [30]. SAMs may be designed so that the terminal groups at the air or liquid interface consist of functionalities that may initiate a polymerization. Upon immersion into a suitable monomer solution, the initiators may then be activated either thermally or photochemically to yield a grafted polymer film. After polymerization, a simple rinsing step will remove most of the untethered polymer that is physisorbed to the substrate. However, it is usually necessary to immerse the substrate into a series of organic solvents, or to use a Soxhlet extraction in order to remove the remaining physisorbed polymer. Polymer brushes have been synthesized by a variety of initiating mechanisms including anionic [32,33], cationic [34,35], ring-opening (ROP) [36–39], ring-opening metathesis (ROMP) [40], free radical [41–47], controlled radical [48–58], enzymatic [59], and organometallic [60] catalysts. Radical polymerization, whether from a normal free radical or a controlled process, is preferred for many applications due to a tolerance for moisture, and a wide variety of organic functional groups. Furthermore, photochemistry is a convenient method for the initiation of free radicals. In particular, it allows for the lithographic patterning of planar substrates and in-situ polymerizations within ordered liquid crystalline materials [61–65]. In this chapter, we review some of the previous work on photochemical SIP from various SAM-coated substrates. We exclude much of the work with thermal or chemical initiators, except where it exhibits clear advantages or introduces alternative pho-

7.2 Substrates

tochemical techniques. We also present recent results from our group on photochemical initiators from alkylthiolate SAMs on planar gold.

7.2

Substrates 7.2.1

Silicon, Silica and Glass

Silicon wafers are the most common substrates for SIP experiments as they are commercially available and relatively smooth and reflective. The surface of these substrates consists of a thin layer of silicon oxide with dangling Si-OH bonds. Thus, treatment with chlorosilanes or alkoxysilanes will form polysiloxane networks that are covalently linked to the native SiOx surface via strong Si-O-Si bonds. Chlorosilanes are more reactive, particularly with water, and generally form less-ordered SAMs than alkoxysilanes; thus, alkoxysilanes are utilized more often as they are less reactive and readily stored. Importantly, the synthesis of alkoxysilane derivatives is more straightforward and amenable to various transformations and purification techniques, including chromatography on modified silica [66]. The chemical preparation of SAMs on glass is virtually identical to that of silicon. While glass is optically transparent and is easily characterized by UV-visible spectroscopy, it yields poor IR spectra due to the high number of Si-O vibrations. In contrast, silicon is more amenable to IR and ellipsometry characterization, but is not optically transparent. Siloxane-based SAMs are quite stable, both thermally and chemically [67]. They are difficult to hydrolyze with acid or base due to the strong Si-O-Si bond, but can be removed with HF treatments. Therefore, polymer may be removed for gel permeation chromatography (GPC) analysis, provided that the polymer is also stable to hydrolysis. In addition, the initiators can be designed to include ester or amide linkages that may be hydrolyzed after polymerization. Some of the earliest attempts at SIP were from free radical initiators attached to silica gel [68–76]. This high surface area substrate ensures a high polymer loading compared to planar substrates. Although others have tethered initiators to silica microspheres and glass fibers [77], silicon wafers remain the substrate of choice for photoinitiated polymerizations [78–80].

7.2.2

Planar Gold

Gold is easily deposited onto silicon or glass by evaporation or sputtering. Since gold does not adhere well to these substrates, a thin layer of titanium or chromium is added to promote adhesion. The predominant crystal structure is Au(111), and the typical thickness of the gold layer is ~100 nm. SAMs are easily deposited by immersing the gold substrates into a dilute solution (~0.1–1.0 mM) of thiol or disulfide pre-

131

132

7 Photoinitiated Polymerization from Self-Assembled Monolayers

cursors. The adsorption kinetics depend on the structure of the precursor, but are generally rapid. However, bulky structures may take several hours to yield densely packed SAMS; our depositions are typically performed overnight. Furthermore, it is not clear if densely packed monolayers are necessary for SIP as the surface area of a polymer chain should be much greater than that of a single initiator molecule within the SAM. Gold is an excellent substrate for SIP for several reasons [81]. First, alkylthiolates are readily adsorbed from solution or the vapor phase. Second, gold substrates are easily characterized by optical techniques such as RAIRS or ellipsometry. Third, alkylthiolate SAMs may be removed from gold by treatment with iodine or by electrochemical desorption. Therefore, the detached polymer may be characterized by techniques such as GPC or multi-angle light scattering (MALS) [82,83]. Fourth, planar gold may be patterned by microcontact printing (lCP) [67,84,85]. This technique utilizes an “inked” polydimethylsiloxane (PDMS) stamp to deposit alkylthiols at well-defined positions on the gold surface. Subsequent polymerization yields polymer brushes in regions that are controlled by the lCP process. An approach for patterning planar gold by combining lCP with photolithography is illustrated in Figure 7.2 [81,86]. First, a PDMS stamp is inked with a SAM initiator precursor and stamped onto the gold with a predefined pattern. Second, a lithographic mask is used to create a secondary polymer brush pattern by irradiating in the presence of monomer. Thus, polymer is deposited only in the regions that were irradiated. Third, a second mask could be used with a different monomer to add another polymer to the pattern. Such patterning would be more difficult with thermal initiation strategies. In addition, thermal initiation is limited by the instability of the S-Au bond, which is unstable above 70 KC. Furthermore, the resolution is limited by the lCP technique, which typically yields patterns from 10 to 100 lm. New strategies are under development for the patterning of gold at submicron resolution. In particular, dip-pen nanolithography (DPN) utilizes an AFM tip to deposit thiols at nanometer resolution [87]. Alternatively, nanocontact printing (nCP) has been used to pattern hyperbranched polymers at 50-nm resolution [88]. These techniques could be modified to pattern photoinitiators for brush synthesis. Furthermore, electron beam chemical lithography (EBCL) was recently described, whereby a SAM is modified with a focused beam of electrons [89]. The irradiated SAM is then chemically modified to produce an azo polymer initiator; brush patterns with lateral resolutions approaching 70 nm have been observed with EBCL. These techniques, and others, will play an important role in the development of new thin-film technologies.

Figure 7.2 Photopatterning of substrates with microcontact printing and surface-initiated polymerization.

7.3 Photoinitiated Radical Polymerization Mechanisms

7.2.3

Nanoparticles

The modification of nanoparticles has become an increasingly important goal in polymer science owing to their unique quantum-size effects [90–93]. The synthesis of polymer-modified nanoparticles is complicated by the fact that it is usually desirable to isolate the nanoparticles. Thus, normal free radical polymerization is unsuitable as polymer is also formed in the bulk; therefore, the nanoparticles would be embedded in the bulk polymer and would be difficult to separate. For this reason, a controlled polymerization must be used to obtain well-defined coatings and narrow size distributions. In particular, atom transfer radical polymerization (ATRP) and reversible addition-fragmentation chain transfer (RAFT) polymerization have been used to modify gold [94–96] and silica [97] nanoparticles. In addition, cationic ROP has been used to synthesize 2-oxazoline-substituted gold nanoparticles [35]. While photoinitiated strategies have been lacking, a recent report demonstrated the synthesis of PMMA from a photoinitiator bound to silica nanoparticles [98].

7.3

Photoinitiated Radical Polymerization Mechanisms 7.3.1

Free Radicals

The most widely used polymer initiators are based on free radicals, as they tolerate water as well as a wide variety of organic functional groups. Typically, these polymerizations are very rapid owing to fast propagation kinetics [99]. The reactive monomers usually include a substituted double bond, where acrylates, methacrylates, and vinylarenes are commonly used. The initiating systems for free radical polymerization are diverse as illustrated in Figure 7.3. Among the most common are 2,2¢-azobisisobutyronitrile (AIBN) (1), benzoyl peroxide (2), and benzoin (3) derivatives. Of these three compounds, only AIBN has been utilized in GF strategies for polymer brush synthesis [1,74,76,78,79,100–102], although peroxides have been generated with plasma treatments [103,104]. In particular, AIBN may be activated thermally above 60 KC, or photochemically by irradiating at 300–360 nm. However, in both cases the initiation step is relatively slow; for instance, AIBN has a half-life of 10 h at 65 KC [105], and the photochemical efficiency is less than 0.5 in styrene [106]. Thus, new surface radicals are continuously formed throughout the course of the reaction. A surface-bound AIBN initiator before and after thermal or photochemical activation is depicted in Scheme 7.1. In both cases, homolytic cleavage of the azo bond yields nitrogen as a byproduct and two radicals, both of which may initiate a polymerization. Importantly, only one of the initiating radicals is bound to the surface, and therefore polymer is always formed in the bulk. This bulk polymer will become entangled with the tethered polymer and will also adsorb onto the brush surface. Thus, a rigorous post-polymerization cleaning step is necessary to remove the physi-

133

134

7 Photoinitiated Polymerization from Self-Assembled Monolayers

Figure 7.3 Known free-radical photochemical initiators. (a) AIBN-type initiators (1) have been incorporated into self-assembled monolayers (SAMs); others such as benzoyl peroxide (2) or benzoin (3) could be modified for SAM deposition. (b) A photosensitizer could extract

a proton from a hydroxyl terminated SAM, yielding an initiating radical at the surface and a ketyl radical that prefers self-condensation. (c) Iniferter strategies might yield well-controlled polymerizations and block copolymers.

Scheme 7.1 Mechanism of free radical polymerization from surface-bound 2,2¢-azobisisobutyronitrile (AIBN) initiators.

sorbed polymer from the chemisorbed brush. This extra cleaning step represents a major disadvantage for the free radical initiators described in Figure 7.3(a). 7.3.2

Photosensitizers

One possible strategy to minimize the formation of bulk polymer is to utilize a photosensitizer [107]. Figure 7.3(b) demonstrates how a triplet sensitizer such as benzophenone (4) may be used to abstract a hydrogen atom from an amine or a secondary alcohol (5) to yield two radicals, 6 and 7 [108]. Importantly, the ketyl radical 7 is unreactive towards monomer, and typically dimerizes to form benzopinacole-type structures. In contrast, the hydroxyl radical (6) initiates the polymerization

7.4 Polymerization from AIBN-type SAMs

of vinyl and acrylate radicals. Thus, polymer does not form in the bulk and arises solely from the surface-bound radical 6. This strategy has not been utilized for the synthesis of polymer brushes, although appropriate studies are currently in progress within our group. 7.3.3

Photo-Iniferters

The radicals from photosensitizers or the decomposition of AIBN will yield uncontrolled polymerizations with high polydispersities. Furthermore, these strategies do not allow for the synthesis of block copolymers. Thus, living polymerizations would be ideal as they allow for the synthesis of well-defined polymers with specific end groups [109–111]. Unfortunately, “living” nitroxide-mediated polymerizations [112] require temperatures in excess of 70 KC, and cannot be used for gold due to the thermally labile S-Au bond. While ATRP methods [113–115] may be activated at room temperature, neither of these “living” techniques may be initiated photochemically. An iniferter is an initiating molecule that also acts as a chain transfer agent and a terminating species (iniferter: initiator-transfer-terminator) [108]. Iniferters may be used to initiate well-controlled polymerizations, yielding homopolymers, copolymers, or block copolymers with low polydispersities. Figure 7.3(c) illustrates a dithiocarbamate photo-iniferter (8) that has been used to synthesize block copolymers [116]. Upon irradiation, 8 will dissociate into a highly reactive benzyl radical (9) and a noninitiating thiyl radical (10). Importantly, radical 10 periodically caps the reactive end and terminates the propagation step. The iniferter capped polymer may absorb another photon and dissociate again to allow additional chain growth. Since the dithiocarbamate is relatively stable, the polymer may be isolated and additional monomer may then be added to grow block copolymers. Indeed, iniferters have been used for the synthesis of polymer brushes by the GF strategy [98,117–119]. These types of initiators will undoubtedly play a major role in future polymer brush syntheses.

7.4

Polymerization from AIBN-type SAMs 7.4.1

Design and Synthesis

Our recent investigations have been focused on thiolate SAMs with AIBN-type initiators as described in Scheme 7.2. The synthesis of initiator 12 has been described previously, and was accomplished by coupling 1,10-decanedithiol with the acid chloride of 4,4¢-azo-bis-(cyanopentanoic acid) [101]. Note that 12 has two thiol groups and may bind to Au by a fully extended structure with a terminal thiol at the air interface, or by a bent structure where both of the thiols are bound to Au. The results of our previous studies have suggested that the favored binding is the

135

136

7 Photoinitiated Polymerization from Self-Assembled Monolayers

Scheme 7.2 Self-assembled monolayer (SAM) forming free radical photoinitiators based on 2,2¢-azobisisobutyronitrile (AIBN).

extended structure with a sulfur terminus at the air interface [101]. The terminal thiol group and the internal thioesters could potentially act as chain transfer agents and thereby inhibit polymer growth from the surface. In order to determine the effects of sulfur and chain transfer on these polymerizations, we designed two other SAM initiator systems, 16 and 19 [120]. In particular, initiator 16 does not exhibit a terminal thiol, so the radical that diffuses from the initial cleavage will not act as an efficient chain transfer agent. However, the surface-bound radical includes a thioester and the sulfur bound to gold, both of which could participate in chain transfer. Control experiments with octadecanethiolate monolayers indicate the SAMs are stable under these reaction conditions; thus, it is unlikely that the S-Au bond plays a role. In order to probe the effect of the thioester, we synthesized compound 19. The SAMs resulting from this disulfide should be similar to both 12 and 16, but without the terminal sulfur or the thioester. As Scheme 7.2 illustrates, compound 16 was synthesized from 4,4¢-azo-bis (cyanopentanoic acid) (13) in two steps: first, 13 was treated with DCC/DMAP and butanol to form the monobutyl ester (14) in 42% yield, followed by a second esterification with 1,10-decanedithiol to give 16 in 20% yield. Disulfide 19 was also synthesized in two steps: first, 11-mercaptoundecanol (17) was treated with iodine to yield disulfide (18) in 95% yield. This was then coupled with 14 and DCC/DMAP to give 19 in 20% yield.

7.4 Polymerization from AIBN-type SAMs

7.4.2

Monolayer Characterization

The SAMs were adsorbed onto Au-coated silicon wafers according to standard procedures. The resulting monolayers were characterized by water contact angle, RAIRS, XPS, and ellipsometry. In particular, the SAM of 12 was described previously and exhibits a static water contact angle of ~64K, consistent with a terminal thiol group. Furthermore, the water contact angle for 16 and 19 increased to 77K and 78K, respectively, suggesting a more hydrophobic surface and consistent with terminal butyl esters. The elemental composition of the SAMs was probed by XPS and was also consistent with the expected structures for all three monolayers. The thickness of each SAM was calculated using a semi-empirical PM3 model, and yielded values of 4.0, 3.15, and 3.3 nm for 12, 16, and 19, respectively. These values were close to the optical thickness measured by ellipsometry of 3.2, 2.5, and 2.0 nm for 12, 16, and 19, respectively. In addition, the calculated tilt angle (h) for 12 and 16 was ~37K, suggesting a fairly disordered monolayer; the tilt of a highly crystalline octadecanethiol SAM is ~30K. The calculated tilt for 19 is quite high at 52K, and suggests a very disordered, low-density SAM. One possible reason for this disorder is that the initiator structure is chiral and consists of a mixture of diastereomers. The lowest energy conformations calculated for two diastereomers of model compound (20) are illustrated in Figure 7.4. In particular, the (R,R) diastereomer exhibits a bent shape, whereas the (R,S) diastereomer prefers an extended structure and is about 0.2 kcal mol–1 more stable than the (R,R). Importantly, the (S,R) diastereomer would be expected to from densely packed monolayers with low tilt angles and a relatively high degree of crystallinity compared to the more bulky (R,R) diastereomer. It seems reasonable that a moderate degree of disorder in the SAM probably will not impact the polymerization or the resulting brushes to a great extent, as the surface area of a tethered polymer chain is much larger than that of a single SAM molecule.

Figure 7.4 Conformations of two diastereomers of an AIBN model compound (20). (Semi-empirical PM3 geometry optimization).

137

138

7 Photoinitiated Polymerization from Self-Assembled Monolayers

Figure 7.5 Fourier transform-reflection-absorption infrared spectroscopy (FT-RAIRS) spectra of a SAM of disulfide (19), monothiol (16), and bisthiol (12).

Nevertheless, it would be interesting to separate the diastereomers in order to probe this effect, and this investigation is planned for the future. Infrared spectroscopy was also used to confirm formation of the SAMs. The RAIR spectra of monolayers of disulfide (19), monothiol (16), and bisthiol (12) are shown in Figure 7.5. The carbonyl bands are evident for all three compounds, with 16 exhibiting two bands due to the presence of the ester (1739 cm–1) and thioester (1690 cm–1) moieties. For all three monolayers, the symmetric and asymmetric methylene bands are at approximately 2853 and 2925 cm–1, respectively. This suggests a disordered, liquid-like state for the SAMs and is consistent with the large calculated tilt angles; a highly crystalline SAM exhibits an asymmetric methylene stretch at ~2918 cm–1 [26]. The asymmetric CH3 band is also present in 16 and 19 at 2968 cm–1, which is consistent with a methyl-terminated SAM. 7.4.3

Polymerization of Styrene

Prior to polymerization, the SAM formation was confirmed by contact angle and RAIRS. The polymerizations were carried out in Schlenk tubes with neat styrene after three consecutive freeze-pump-thaw cycles. The tubes were back-filled with argon and irradiated at 300 nm (~650 lW cm–2) for a specified period of time. After polymerization, the substrates were removed and rinsed with tetrahydrofuran (THF) or toluene. This first rinse is sufficient to remove most of the physisorbed polymer, but it is also necessary to perform a Soxhlet extraction or to immerse the substrates into a series of clean solvents in order to remove the remaining physisorbed polymer that might have become entangled within the brush during the polymerization. The presence of a polystyrene (PS) brush is clearly evident in the RAIR spectrum (Figure 7.6) after cleaning. In addition, the static water contact angle of 88 € 2K is

7.4 Polymerization from AIBN-type SAMs

Figure 7.6 Fourier transform-reflection-absorption infrared spectroscopy (FT-RAIRS) spectrum of a polystyrene brush.

consistent with spun films of PS, as is the XPS spectrum. RAIRS may be used to gain a qualitative picture of the film growth, whereas ellipsometry is used to gain a more quantitative value for the film thickness. As illustrated in Figure 7.7, the film

Figure 7.7 Effect of reaction time on brush thickness for initiator SAMs of: (a) disulfide 19; (b) monothiol 16; and (c) bis-thiol 12.

139

140

7 Photoinitiated Polymerization from Self-Assembled Monolayers

growth was monitored as a function of time by ellipsometry. Interestingly, the disulfide initiator (19) yielded a maximum thickness of ~100 nm, whereas 12 and 16 terminated at ~200 nm. The most rapid growth was seen with the bisthiol initiator (12), which was closely matched by monothiol initiator (16). The film growth in all cases is nonlinear, suggesting that the increase in thickness is not directly correlated to an increase in molecular weight, which would be expected to increase linearly. It is postulated that the multistage growth is largely due to an increase in the density of grafted chains rather than to a simple increase in the kinetic chain length of tethered polymer chains. It is interesting to note that polymer growth from the disulfide initiator 19 is much slower than from 12 or 16. In particular, the PS thickness for 19 after 8 h of irradiation was only about 10 nm, whereas the thickness for both 12 and 16 was ~115 nm. In addition, after 16 h initiator 19 had a thickness of only 25 nm as compared to 200 nm for 12, which had already terminated. It is believed that the primary reason for this slow growth is the poor quality of the disulfide SAMs, which limits the number of initiators and reduces the grafting density of tethered polymer. For example, in Figure 7.5 the methylene bands are very weak for 19 compared to the other initiators. Repeated attempts to improve the SAM by varying solvents and deposition conditions proved unsuccessful. One possible reason lies in the rather bulky structure of the disulfide. In particular, it is hypothesized that after the initial adsorption of a low-density monolayer, diffusion of 19 to the gold surface is inhibited due in part to the bulky nature of the diastereomeric disulfide structure, much in the same way that the GT technique is limited for brush synthesis [121]. Compounds 12 and 16 are much smaller, and therefore they may penetrate the low-density SAM to form a more extended, high-density structure. While simple long-chain disulfides, such as C18, may form high-quality monolayers, from our experience more complex disulfides are less likely to form high-quality SAMs than thiol-terminated compounds of similar structure. In order to gain more insight into the growth of these films, we removed the polymer from the gold substrates and used GPC and MALS to determine the Mn. Removal is easily accomplished by treating the substrates with a solution of iodine in THF. Typically we used three to four substrates in separate Schlenk tubes, and then combined them before treating with iodine. It should be noted that this oxidative cleavage forms a disulfide between neighboring S-Au chains; therefore, it is possible that two high-Mn chains were coupled together and the experimental Mn is larger than the true Mn expected for a single chain. However, reduction of the disulfide with sodium borohydride yielded little change in the Mn of the polymer; thus, it is likely that the large chains couple with a neighboring chain that is either an unreacted initiator or a small oligomer. In all cases, the polydispersity for tethered PS and polymer recovered from the bulk was ~2, which is typical for these conditions. Interestingly, the increase in Mn of the detached polymer from 12 increases only slightly from 2 to 18 h, as illustrated in Figure 7.8. In particular, the Mn increases by ~40% over this time frame, whereas the film thickness increases by ~470%. Furthermore, the increase in Mn from 2 to 6 h is ~14%, and from 6 to 18 h is 22%; however,

7.4 Polymerization from AIBN-type SAMs

Figure 7.8

Comparison of polymer brush thickness with molecular weight (Mn) for initiator 12.

the film thickness increases by 85% and 200% over the same time intervals, respectively. Surprisingly, the middle segment from 7 to 10 h exhibits a 130% increase in film thickness, whereas the Mn is nearly constant, increasing by a mere 3%. Clearly, the increase in Mn from 2 to 18 h for the detached polymer does not correlate with the observed increase in film thickness. The film thickness appears to increase much more rapidly than Mn during the first two stages of the polymerization. Thus, the growth of the film is relatively independent of the Mn, which is fairly constant regardless of reaction time. This is not so surprising, as a free-radical chain polymerization is characterized by rapid growth, followed by termination. In the present case, the AIBN-type initiator is not very efficient and has a long half-life. Therefore, chains are initiated at all stages of the polymerization and most likely grow rapidly and then terminate after a relatively short period of time. Based on this behavior, we propose a four-stage growth model for the free radical polymerization of styrene from AIBN-type SAMs; we use initiator 12 as an example. The initial stage is characterized by the rapid growth of a low-density polymer brush. The film growth then slows down during the second stage, from 1 to 7 h. During this period, the grafting density reaches a point where interchain termination becomes more favorable than in the low-density brush. The third stage is characterized by a rapid acceleration in film growth (7–10 h). We hypothesize that at the onset of the third stage, the grafted chains are sufficiently crowded that termination is inhibited due to the so-called gel effect, also known as the Trommsdorf effect [122]. This has two consequences: first, the rate of polymerization is accelerated, leading to faster growth kinetics. Second, the grafting density increases rapidly, and this forces the chains to stretch away from the surface at an accelerated rate. The

141

142

7 Photoinitiated Polymerization from Self-Assembled Monolayers

fourth stage is characterized by a rapid termination of film growth as saturation is reached and monomer no longer diffuses to the surface. This multi-stage growth has precedence in the literature for both GT and GF approaches. In particular, ROhe has described the photopolymerization of PS from siloxane-based SAMs [80]. The film thickness was only monitored for up to 4 h and resulted in a thickness of ~30 nm in neat styrene; this is close to our observations for both 12 and 16, which exhibited a thickness of 45 and 59 nm, respectively. Similarly, polymerizations from gold with a thiolate initiator yielded PS films of ~300 nm after 12 h, compared to 180 nm for 12 over the same time frame [86]. The larger thickness is likely due to the fact that we irradiated at 300 nm with an intensity of ~0.65 mW cm–2 compared to 350 nm and 30 mW cm–2 used by ROhe. Thus, their experiments were at a much higher intensity, which is also closer to the kmax of AIBN. However, it should also be noted that thermal polymerizations with similar initiators do not exhibit an accelerated growth period that is as dramatic as with photoinitiated polymerizations [123]. Interestingly, this multi-stage growth behavior is also fairly consistent with theoretical and experimental investigations on GT systems. In particular, Penn and coworkers described a three-stage growth model for the grafting of amino-terminated PS to an epoxide-terminated siloxane SAM [124]. Their observations and descriptions of the kinetics of film growth are remarkably similar to ours. However, the GF technique that we use yields faster grafting kinetics by an order of magnitude. For instance, the saturation point for a 44-kDa PS brush was reported at ~116 h (at 100 KC) compared to ~12 h for a 200-kDa brush (at 25 KC) with initiator 12. Clearly, further experimental and theoretical investigations must be conducted to carefully examine the kinetics of grafting and the accelerated film growth in polymer brushes. In addition to the film growth, the grafting density in surface area per chain (Sd, nm2 per chain) may be calculated according to Eq. (1) from the molecular weight (Mn, g mol–1), the thickness (Th, nm), density (d, g nm–3), and Avogadro’s number (Na, molecules per mol). Thus, the Sd for a 193-nm film with a Mn of 225 000 Da is calculated to be 1.85 nm2 per chain, assuming a density of 1.047 Q 10–21 g nm–3 for bulk PS. Interestingly, the theoretical length of a fully stretched chain with this Mn is 463 nm, which yields an Sd of 0.77 nm2 per chain. Furthermore, the radius of gyration (Rg) for the tethered PS in toluene is ~25 nm, as measured by MALS. Thus, the grafted PS is highly stretched, and the Sd approaches that of the theoretical limit. Sd ¼

Mn ThdNa

(1)

In general, these films are very smooth, with the surface roughness ranging from € 0.4 to 3.6 nm. Not surprisingly, the brushes from the disulfide initiator 19 were the roughest, with a r-value of € 3.6 nm, as determined by ellipsometry. The other initiators yielded better results, with standard deviations in the range of € 0.6 nm and regions as smooth as € 3.6 R. In particular, Figure 7.9 illustrates a

7.5 Conclusions and Future Studies

Figure 7.9 A topographic thickness map of a polystyrene brush with an average thickness of 140 € 0.65 nm over an area of 100 7 193 nm2. The area within the square is very smooth, with a r-value of € 4: (data acquired using ellipsometry).

100 Q 193 nm2 region of a 140 nm-thick PS brush from initiator 12. The r for this film was € 0.65 nm, with large regions as smooth as € 0.4 nm. These results are similar to initiator 16, where a 119-nm film exhibited a r -value of € 0.59 nm and regions as smooth as € 0.36 nm.

7.5

Conclusions and Future Studies

Polymer brushes are conveniently synthesized from surface-bound photoinitiators. Three AIBN-type initiators were synthesized with various amounts of sulfur, and the effect of chain transfer was found to be minimal. We propose that the disulfide (19) yields poor SAMs, and that this affects the resulting brush density, yielding thinner films than the thiol initiators 12 and 16. These data suggest that the quality of the monolayer plays an important role in the resulting polymerizations. A four-stage model was also proposed to explain the accelerated growth in the middle stages of the polymerization, namely that the gel-effect leads to acceleration in the rate of polymerization. This is concomitant with a rapid increase in grafting density. Since the Mn is fairly constant throughout the course of the polymerization, the overall film thickness is controlled more by the grafting density than by the kinetic chain length of the polymer chains. The PS films synthesized here were very smooth, with an average roughness on the order of € 0.6 nm. Furthermore, the grafting densities from initiator 12 were as high as 1.8 nm2 per chain, and are among the highest reported to date. Future investigations should be directed at the synthesis of multi-component films, particularly mixed brushes with specific response characteristics. These may include homopolymers, copolymers, ionic polymers, and liquid crystalline polymers. Furthermore, the ability to pattern substrates will play an important role in the manufacturing of new thin-film devices. Photoinitiating SAMs represent an enabling technology for the synthesis of these functional substrates. In particular, well-controlled “living”-like polymer initiators will play an increasingly important role in polymer brush synthesis. Thus, photo-iniferters should be fully explored in addition to new photosensitizer strategies. These emerging technologies promise unprecedented control of the surface architecture in the micron and nanometer domains.

143

144

7 Photoinitiated Polymerization from Self-Assembled Monolayers

7.6

Experimental 7.6.1

Initiator Synthesis

Silicon wafers (100) were used to prepare thin gold films by using 99.995+% pure Au and Cr. Styrene (Acros) was purified by passage through a column filled with Alumina A (Fisher Scientific). 11-Mercapto-1-undecanol (17) was purchased from Aldrich Chemical Company. 4-Dimethylaminopyridine (DMAP), dicyclohexylcarbodiimide (DCC), and 4,4¢-azo-bis (4-cyanopentanoic acid) (13) were purchased from Acros chemicals and used without further purification. NMR solvents were purchased from Cambridge Isotopes. Column chromatography was performed with standard grade silica gel (63–200 Mesh) from Sorbent Technologies, and TLC was performed on 250-lm silica gel 60 polyester-backed plates with F254 fluorescent indicator (Whatman). All other reagents were obtained from Acros or Fisher Scientific, and used as received. The synthesis of 4,4¢-azo-bis[(1,10-dimercaptodecyl)-4-cyanopentanoate] (12) and 1,10-dithiol-decane (15) were described previously [101]. 1H and 13C NMR data were collected on a Varian VXR-300 MHz NMR spectrometer and reference to the solvent; coupling constants (J) are in Hertz. 7.6.1.1 4-Cyano-4-(azo-[4¢-cyano-(butyl)pentanoate])-pentanoic acid (14) This compound was recently synthesized using a similar method [102]. 4,4¢-Azo-bis(4-cyanovaleric acid) (13) (5.6 g, 20 mmol), 1-butanol (1.48 g, 20 mmol), and DMAP (488 mg, 4.0 mmol) were dissolved in THF (120 mL). The reaction mixture was cooled to 0 KC in an ice bath for 15 min. DCC (4.13 g, 20 mmol) was then added, and the resulting mixture was stirred for 5 min at 0 KC and then for 15 h at room temperature. The dicyclohexylurea precipitate was filtered off, and the crude filtrate concentrated by rotary evaporation and then purified by column chromatography on silica gel (hexanes:ethyl acetate, 50:50) to yield 2.96 g (42%) of a colorless oil: RF = 0.46 (hexanes:ethyl acetate, 50:50); 1H NMR (300 MHz, CDCl3) d 0.92 (t, J = 7.3, 2H), 1.37 (m, 2H), 1.58–1.72 (m, 8H), 2.38–2.58 (m, 8H), 4.10 (t, J = 6.6, 2H); 13 C NMR (75 MHz, CDCl3) d 13.66, 19.06, 23.72 (2 C), 28.86, 29.17, 30.49, 32.91, 33.16, 65.05, 71.70, 71.85, 117.44, 117.48, 171.53, 176.48. 7.6.1.2 4-Cyano-4-(azo-[4¢-cyano-(butyl)pentanoate])-(1,10-dimercaptodecyl)pentanoate (16) Compound 14 (610 mg, 1.81 mmol), 1,10-dithiodecane (377 mg, 1.80 mmol), and DMAP (44.2 mg, 0.36 mmol) were dissolved in THF (20 mL). The reaction mixture was cooled to 0 KC (ice bath) for 15 min. DCC (370 mg, 1.80 mmol) was then added and the resulting mixture was stirred for 5 min at 0 KC and then for 15 h at room temperature. The crude mixture was concentrated by rotary evaporation and then purified by column chromatography on silica gel (hexanes:ethyl acetate, 50:50) to yield 200 mg (20%) of a colorless oil: RF = 0.66 (hexanes:ethyl acetate, 50:50). 1H NMR (300 MHz, CDCl3) d 0.89 (t, J = 7.3, 3H), 1.22–1.34 (m, 14H), 1.53–1.68 (m,

7.6 Experimental

12H), 2.35–2.71 (m, 10H), 2.84 (t, J = 6.8, 2H), 4.03–4.09 (m, 2H); 13C NMR (75 MHz, CDCl3) d 13.64, 19.03, 23.88, 23.93, 24.58, 28.27, 28.70, 28.95, 29.04, 29.07, 29.13, 29.29, 29.33, 30.48, 33.12, 33.19, 33.96, 38.32,39.08, 64.94, 71.79, 71.86, 117.36, 117.39, 171.31, 196.79; FT-IR (neat): 2936, 2858, 1761, 1731, 1592, 1493, 1456, 1296, 1193 cm–1. 11-(11-Hydroxy-undecyldisulfanyl)-undecanol (18) 11-Mercapto-1-undecanol (17) (500 mg, 2.45 mmol) was dissolved in 50 mL absolute ethanol. The solution was titrated with a saturated solution of iodine in ethanol until the brown color of iodine persisted. The solvent was removed by rotary evaporation. Water (20 mL) was added, and the solution was extracted with diethyl ether (3 Q 100 mL). Evaporation of solvent yielded 470 mg (95%) of 18 as a white solid: RF= 0.53 (EtOAc); 1H NMR (300 MHz, CDCl3) d 1.18–1.39 (m, 24H), 1.48–1.70 (m, 8H), 2.64 (t, J = 7.5, 4H), 3.63 (t, J = 6.8, 4H). 7.6.1.3

1,1¢-Bis-[4-Cyano-4-(azo-[4¢-cyano-(butyl)pentanoate])-(1-mercaptoundecyl)pentanoate]disulfanyl (19) Compound 17 (610 mg, 1.5 mmol), compound 14 (1.008 g, 3.0 mmol), and DMAP (73 mg, 0.6 mmol) were dissolved in 30 mL THF, and the solution was cooled and stirred at 0 KC. DCC (619 mg, 3.0 mmol) was added and the mixture was stirred at 0 KC for 5 min, and then at room temperature for 12 h. The crude mixture was concentrated by rotary evaporation and then purified by column chromatography on silica gel (hexane:ethyl acetate, 50:50) to yield 500 mg (33%) of an impure product as a viscous oil. The product was further purified by column chromatography on silica gel (diethyl ether:petroleum ether, 70:30) and then recrystallized from ether at low temperature (–20 KC) to yield 300 mg (20%) of a white solid: RF = 0.36 (diethyl ether:petroleum ether, 70:30); 1H NMR (300 MHz, CDCl3) d 0.91 (t, J = 7.3, 6H), 1.25–1.37 (m, 32H), 1.60–1.71 (m, 24H), 2.38–2.50 (m, 16H), 2.67 (t, J = 7.5, 4H), 4.08 (m, 8H); 13 C NMR (75 MHz, CDCl3) d 13.93, 19.32, 23.97, 24.21, 26.08, 28.74 (2C), 29.32, 29.40, 29.42, 29.46 (2C), 29.70 (2C), 29.72, 30.76, 33.40, 33.43, 39.35, 65.22, 65.51, 72.06, 72.16, 117.12, 117.80, 171.60, 171.66; FT-IR (neat) 2926, 2858, 1739, 1456, 1395, 1294, 1188 cm–1; HRMS (FAB+): calcd. 1043.5151; found, 1043.6401; Empiric analysis for C54 H90N8O8S2: calcd. C, 62.16%; H, 8.69%; N, 10.74%; S, 6.15%; found: C, 62.44%; H, 8.67%; N, 10.70%, S, 6.19%. 7.6.1.4

7.6.2

Polymerizations

The gold substrates and SAMs were prepared according to previously published procedures [101,125]. The SAM and styrene were placed into a Schlenk tube, which was degassed by using three freeze-pump-thaw cycles, and then backfilled with argon. The polymerization was performed in a Rayonet photochemical chamber at 300 nm (~650 lW cm–2) for a specified time. The chamber was cooled by air through a rubber tube attached to an air compressor. After the reaction, the gold substrates were rinsed with toluene and then extracted with a water-cooled jacketed Soxhlet appara-

145

146

7 Photoinitiated Polymerization from Self-Assembled Monolayers

tus by using toluene as solvent for 15 h. The UV intensity was measured with a Model UV-340 light meter (Digital Technologies) at 2- and 6-cm distances from the lamps, and the results averaged. 7.6.3

Reflection Absorption Infrared Spectroscopy (FT-RAIRS) Measurements

FT-RAIRS was used to generate spectra of monolayers and polymer brushes on gold substrates. This was carried out using a Nicolet-670 FT-IR spectrometer fitted with a VeeMax II variable angle attachment (Pike Technologies) set to 80K and an MCT-B detector. The sample compartment was purged with nitrogen for 20 min (flow rate: 25 ml min–1) prior to data acquisition, and a clean gold substrate was used as the background. All spectra were averaged over 650 scans, and an atmospheric suppression algorithm was utilized. 7.6.4

Ellipsometry

Thickness of the films was monitored using an I-Elli2000 imaging ellipsometer (Nanofilm Technologie, GmbH). The experiments were performed with a 20-mW Nd:YAG laser (532 nm) at an incident angle of 70K. The optical constants n and k (refractive index and extinction coefficient) were measured from bare gold. Refractive indices of 1.46 and 1.59 were used for the calculation of initiator SAMs and PS films, respectively. The films were considered to be optically transparent, and data were collected and averaged over at least five different spots per slide. At least three separate trials were averaged for each data point in Figure 7.7. Figure 7.9 represents a delta map obtained by using the Micro-Mapping feature of the I-Elli2000 software package. This delta map is transformed into a thickness map by applying the optical model used to determine the thickness of the PS layer as mentioned above. The 3-D viewer module of the I-Elli2000 software is used to obtain the final 3-D plot. 7.6.5

X-Ray Photoelectron Spectroscopy (XPS)

XPS measurements were conducted using a Kratos Axis Ultra X-ray photoelectron spectrometer. Analysis was carried out under ultra-high vacuum conditions (10–9 Torr) using monochromatic Al Ka (1486.6 eV) excitation. The hemispherical energy analyzer was operated in the hybrid mode (a combined magnetic and electrostatic lens mode) with the slot (300 lm Q 700 lm) selected area aperture. The sample stage was grounded to the spectrometer and the neutralizer was off. Spectra were collected in the constant pass energy, or fixed analyzer transmission, mode. Survey spectra were collected using a pass energy of 160 eV with a scan step size of 1 eV.

7.6 Experimental

7.6.6

Molecular Weight Measurements

The polymer was isolated from three to four PS-coated substrates (~1.8 Q 8.0 cm2) that were reacted in separate Schlenk tubes. The PS brush thickness was confirmed before the substrates were combined into a solution of iodine (4.0 mM) in dichloromethane for 10 h. The solvent was taken up by syringe and passed through a 0.2-lm Anotop syringe filter into a pear-shaped flask. The solvent was removed by rotary evaporation, and the excess iodine removed by sublimation under high vacuum. The residual polymer was taken up in 200 lL of toluene and transferred to a 300-lL autosampler vial. The molecular weights of PS were measured in toluene with a Waters Alliance 2690 separation module fitted with a Waters 2410 differential refractive index detector and a column heater set to 35 KC. Two 7.5 Q 300 mm PLgel 5-lm MIXED-C columns from Polymer Laboratories were used and calibrated with PS standards. The Rg was measured with a MiniDAWN multi angle light scattering (MALS) detector (Wyatt Technologies). 7.6.7

Molecular Modeling

Structures were determined using a semi-empirical PM3 model and the PolakRibiere optimization algorithm for isolated molecules in vacuo, performed with HyperChem v. 6.03 (Hypercube, Inc.).

Acknowledgments

The authors thank the National Science Foundation under grant CHE-0094195, 3M Corporation, the Materials Technology Center at SIUC, and the donors of the Petroleum Research Fund, administered by the American Chemical Society for partial support of this research. The XPS measurements carried out in the Center for Microanalysis of Materials at the University of Illinois, which is supported by the U.S. Department of Energy under grant DEFG02-96ER45439 and the University of Illinois. The Kratos XPS system was purchased with funds provided by the National Science Foundation under grant DMR-9977482 and by The State of Illinois.

147

7 Photoinitiated Polymerization from Self-Assembled Monolayers

148

References 1 J. ROhe, W. Knoll, J. Macromol. Sci. Polym. 2 3 4 5 6

7

8

9

10 11

12

13

14

15

16

17

18 19

Rev. 2002, C42, 91–138. B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677–710. Y. Nagasaki, K. Kataoka, Trends Polym. Sci. 1996, 4, 59–64. S. T. Milner, Science 1991, 251, 905–914. A.M. Rouhi, Chem. Eng. News 1999, 77, 51– 59. C. Perry, in: Chemistry of Advanced Materials: An Overview, Eds. L. V. Interrante, M. J. Hampden-Smith, Wiley-VCH, Weinheim, 1998, Chapter 11. L. E. Niklason, J. Gao, W. M. Abbott, K. K. Hirschi, S. Houser, R. Marini, R. Langer, Science 1999, 284, 489–493. S. Rimmer, in: Emerging Themes in Polymer Science, Ed. A. J. Ryan, Royal Society of Chemistry, Cambridge, 2001, Chapter 8. C. D. James, R. Davis, M. Meyer, A. Turner, S. Turner, G. Withers, L. Kam, G. Banker, H. Craighead, M. Isaacson, J. Turner, W. Shain, IEEE Trans. Biomed. Eng. 2000, 47, 17–21. R. Langer, Acc. Chem. Res. 2000, 33, 94–101. N. R. Cameron, in: Emerging Themes in Polymer Science, Ed. A. J. Ryan, Royal Society of Chemistry, Cambridge, 2001, Chapter 9. J. T. Santini, A. C. Richards, R. Scheidt, M. J. Cima, R. Langer, Angew. Chem. Int. Ed. Engl. 2000, 39, 2396–2407. B. D. Klugherz, P. L. Jones, X. Cui, W. Chen, N. F. Meneveau, S. DeFelice, J. Connolly, R. L. Wilensky, R.J. Levy, Nature Biotechnol. 2000, 18, 1181–1184. S. L. Ishang-Riley, L. E. Okun, G. Prado, M. A. Applegate, A. Ratcliffe, Biomaterials 1999, 20, 2245–2256. C. D. Tidwell, S. I. Ertel, B. D. Ratner, B. J. Tarasevich, S. Atre, D. L. Allara, Langmuir 1997, 13, 3404–3413. B. D. Ratner, A. B. Johnston, T. J. Lenk, J. Biomed. Mater. Res. Appl. Biomater. 1987, 21, 59–90. F. E. Black, M. Hartshorne, M. C. Davies, C. J. Roberts, S. J. B. Tendler, P. M. Williams, K. M. Shakesheff, Langmuir 1999, 15, 3157– 3161. P. Lenz, Adv. Mater. 1999, 11, 1531–1534. Y. Xia, D. Qin, Y. Yin, Curr. Opin. Colloid Inter. Sci. 2001, 6, 54–64.

20 D. E. Kataoka, S. M. Trolan, Nature 1999,

402, 794–797. 21 L. Kricka, Clin. Chim. Acta 2001, 307, 219–

223. 22 M. Krishnan, V. Namasivayam, R. Lin, R. Pal,

M. A. Burns, Curr. Opin. Biotech. 2001, 12, 92–98. 23 J. Lahiri, L. Isaacs, B. Grzybowski, J. D. Carbeck, G. M. Whitesides, Langmuir 1999, 15, 7186–7198. 24 A. M. Rouhi, Chem. Eng. News 1997, 75, 41– 45. 25 T. P. Sullivan, W. T. S. Huck, Eur. J. Org. Chem. 2003, 17–29. 26 A. Ulman, Chem. Rev. 1996, 96, 1533–1554. 27 J. Israelachvili, Intermolecular and Surface Forces, 2nd Ed., Academic Press, New York, 1992. 28 H. Arwin, Phys. Stat. Sol. (A) 2001, 188, 1331–1338. 29 D. L. Allara, P. Zhang, Mater. Sci. Technol. 1994, 2B, 657–755. 30 G. A. Somorjai, Mater. Res. Soc. Bull. 1998, 23, 11–29. 31 A. G. Frutos, R. M. Corn, Anal. Chem. 1998, 70, 449A–455A 32 R. Jordan, A. Ulman, J. F. Kang, M. H. Rafailovich, J. Sokolov, J. Am. Chem. Soc. 1999, 121, 1016–1022. 33 R. Advincula, Q. Zhou, M. Park, S. Wang, J. Mays, G. Sakellariou, S. Pispas, N. Hadjichristidis, Langmuir 2002, 18, 8672–8684. 34 R. Jordan, A. Ulman, J. Am. Chem. Soc. 1998, 120, 243–247. 35 R. Jordan, N. West, A. Ulman, Y.-M. Chou, O. Nuyken, Macromolecules 2001, 34, 1606– 1611. 36 M. Husemann, D. Mecerreyes, C. J. Hawker, J. L. Hedrick, R. Shah, N. L. Abbott, Angew. Chem. Int. Ed. 1999, 38, 647–649. 37 T. KratzmOller, D. Appelhans, H.-G. Braun, Adv. Mater. 1999, 11, 555–558. 38 M. MXller, F. Nederberg, L. S. Lim, R. KYnge, C. J. Hawker, J. L. Hedrick, Y. Gu, R. Shah, N. L. Abbott, J. Polym. Sci. Part A: Polym. Chem. 2001, 39, 3529–3538. 39 Y. Gu, F. Nederberg, R. KYnge, R. R. Shah, C. J. Hawker, M. MXller, J. L. Hedrick, N. L. Abbott, ChemPhysChem 2002, 448–451.

References 40 M. Weck, J. J. Jackiw, R. R. Rossi,

41 42

43 44 45 46

47 48 49

50 51 52 53

54 55

56 57

58

59

60

P. S. Weiss, R. H. Grubbs, J. Am. Chem. Soc. 1999, 121, 4088–4089. M. Niwa, M. Date, N. Higashi, Macromolecules 1996, 29, 3681–3685. F. Zhang, E. T. Kang, K. G. Neoh, W. Huang, J. Biomater. Sci. Polym. Ed. 2001, 12, 515– 531. J. Hyun, A. Chilkoti, Macromolecules 2001, 34, 5644–5652. L. K. Ista, S. Mendez, V. H. PZrez-Luna, G. P. L[pez, Langmuir 2001, 17, 2552–2555. A. Laschewsky, E. D. Rekai, E. Wischeroff, Macromol. Chem. Phys. 2001, 202, 276–286. A. Laschewsky, O. Ouari, C. Mangeney, L. Jullien, Macromol. Symp. 2001, 164, 323– 340. X. Liu, K. G. Neoh, E. T. Kang, Langmuir 2002, 18, 9041–9047. J.-B. Kim, M. L. Bruening, G. L. Baker, J. Am. Chem. Soc. 2000, 122, 7616–7617. R. R. Shah, D. Merreceyes, M. Husemann, I. Rees, N. L. Abbott, C. J. Hawker, J. L. Hedrick, Macromolecules 2000, 33, 597–605. W. Huang, G. L. Baker, M. L. Bruening, Angew. Chem. Int. Ed. 2001, 40, 1510–1512. S. Choi, R. Langer, Macromolecules 2001, 34, 5361–5363. D. Gopireddy, S. M. Husson, Macromolecules 2002, 35, 4218–4221. D. M. Jones, J. R. Smith, W. T. S. Huck, C. Alexander, Adv. Mater. 2002, 14, 1130– 1134. D. M. Jones, A. A. Brown, W. T. S. Huck, Langmuir 2002, 18, 1265–1269. J.-B. Kim, W. Huang, M. L. Bruening, G. L. Baker, Macromolecules 2002, 35, 5410– 5416. V. L. Osborne, D. M. Jones, W. T. S. Huck, Chem. Commun. 2002, 1838–1839. A. Y. Sankhe, S. M. Husson, S. M. Kilbey, Mater. Res. Soc. Symp. Proc. 2002, 710, 277– 282. S. Balamurugan, S. Mendez, S. S. Balamurugan, M. J. O’Brien, G. P. Lopez, Langmuir 2003, 19, 2545–2549. Y.-R. Kim, H.-J. Paik, C. K. Ober, G. W. Coates, C. A. Batt, Polymer Preprints 2002, 43(1), 706–707. M. D. K. Ingall, S. J. Joray, D. J. Duffy, D. P. Long, P. A. Bianconi, J. Am. Chem. Soc. 2000, 122, 7845–7846.

61 T. J. Bunning, S. H. Chen, L. C. Chien,

T. Kajiyama, N. Koide, S. C. A. Lien (Eds.), Liquid Crystal Materials and Devices, Materials Research Society, Warrendale, PA, Vol. 559, 1999. 62 A. I. Isayev, T. Kyu, S. Z. D. Cheng (Eds.), Liquid Crystalline Polymer Systems: Technological Advances, American Chemical Society, Washington, D.C., Vol. 632, 1996. 63 C. M. Paleos (Ed.), Polymerization in Organized Media, Gordon and Breach, Philadelphia, PA, 1992. 64 G. P. Crawford, S. Zumer (Eds.), Liquid Crystals in Complex Geometries, Taylor & Francis, Bristol, PA, 1996. 65 B. Peng, D. Johannsmann, J. ROhe, Macromolecules 1999, 32, 6759–6766. 66 Y. Barness, O. Gershevitz, M. Sekar, C. N. Sukenik, Langmuir 2000, 16, 247–251. 67 Y. Xia, G. M. Whitesides, Angew. Chem. Int. Ed. 1998, 37, 550–575. 68 N. Fery, R. Laible, K. Hamann, Angew. Makromol. Chem. 1973, 34, 81–109 69 Hamann, R. Laible, J. Horn, Polym. Sci. Technol. 1975, 9A, 93–105 70 M. Koehler, J. Ohngemach, in: Radiation Curing of Polymeric Materials (Eds. C. E. Hoyle, J. F. Kinstle), American Chemical Society, Vol. 417, Washington, D.C., 1990, Chapter 9, pp. 106–124. 71 G. Boven, M. L. C. M. Oosterling, G. Challa, A. J. Schouten, Polymer 1990, 31, 2377–2383. 72 E. Carlier, A. Guyot, A. Revillon, M.-F. Llauro-Darricades, R. Petiaud, React. Polym. 1991, 16, 41–49. 73 E. Carlier, A. Guyot, A. Revillon, React. Polym. 1992, 16, 115–124. 74 N. Tsubokawa, Y. Shirai, K. Hashimoto, Coll. Polym. Sci. 1995, 273, 1049–1054. 75 N. Tsubokawa, Y. Shirai, H. Tsuchida, S. Handa, J. Polym. Sci. Part A: Polym. Chem. 1994, 32, 2327–2332. 76 O. Prucker, J. ROhe, Macromolecules 1998, 31, 592–601. 77 K. Fujiki, M. Sakamoto, Polym. Polym. Comp. 1999, 7, 453–460. 78 G. Tovar, S. Paul, W. Knoll, O. Prucker, J. ROhe, Supramol. Sci. 1995, 2, 89–98. 79 X. Chen, L. M. Tolbert, C. L. Henderson, D. W. Hess, J. ROhe, J. Vac. Sci. Technol. B 2003, 19, 2013–2019. 80 O. Prucker, M. Schimmel, G. Tovar, W. Knoll, J. ROhe, Adv. Mater. 1998, 10, 1073–1077.

149

150

7 Photoinitiated Polymerization from Self-Assembled Monolayers 81 J. Dyer, Adv. Funct. Mater. 2003, 13, 667–670 82 P. J. Wyatt, Anal. Chim. Acta 1993, 272,

1–40. 83 J. A. Bumpus, J. J. Kim, S. R. Coon, J. Chem. Ed. 2000, 77, 1396–1400. 84 R. Crooks, ChemPhysChem 2001, 2, 644–654. 85 N. L. Jeon, I. S. Choi, G. M. Whitesides, N. Y. Kim, P. E. Laibinis, Y. Harada, K. R. Finnie, G. S. Girolami, R. G. Nuzzo, Appl. Phys. Lett. 1999, 75, 4201–4203. 86 O. Prucker, J. Habicht, I.-J. Park, J. ROhe, Polym. Preprints 2003, 44(1), 470–471. 87 R. D. Piner, J. Zhu, F. Xu, S. H. Hong, C. A. Mirkin, Science 1999, 283, 661–663. 88 H.-W. Li, B. V. O. Muir, G. Fichet, W. T. S. Huck, Langmuir 2003, 19, 1963–1965. 89 U. Schmelmer, R. Jordan, W. Geyer, W. Eck, A. GXlzh\user, M. Grunze, A. Ulman, Angew. Chem. Int. Ed. 2003, 42, 559–563. 90 Z. Zhang, W. Sun, P. Cheng, Molecules 2003, 8, 207–222. 91 J.-T. Lue, J. Phys. Chem. Solids 2001, 62, 1599–1612. 92 F. Remacle, R. D. Levine, ChemPhysChem 2001, 2, 20–36. 93 J. Murphy, Anal. Chem. 2002, 74, 520A– 526A. 94 K. Kamata, Y. Lu, Y. Xia, J. Am. Chem. Soc. 2003, 125, 2384–2385. 95 K. Ohno, K. Koh, Y. Tsujii, T. Fakuda, Macromolecules 2002, 35, 8989–8993. 96 J. Raula, J. Shan, M. Nuopponen, A. Niskanen, H. Jiang, E. I. Kauppinen, H. Tenhu, Langmuir 2003, 19, 3499–3505. 97 T. E. Patten, Polym. Preprints 2003, 44(1), 425–426. 98 J. Bai, K.-Y.Qiu, Y. Wei, Polym. Int. 2003, 52, 853–858. 99 G. Moad, D. H. Solomon, The Chemistry of Free Radical Polymerization, Pergamon, Terry Town, New York, 1995. 100 W. Huang, G. Skanth, G. L. Baker, M. L. Bruening, Langmuir 2001, 17, 1731–1736. 101 R. Schmidt, T. Zhao, J.-B. Green, D. J. Dyer, Langmuir 2002, 18, 1281. 102 X. Fan, C. Xia, T. Fulghum, M.-K. Park, J. Locklin, R. C. Advincula, Langmuir 2003, 19, 916–923. 103 J. Zhang, X. Xu, J. Chen, E. Kang, Thin Solid Films 2002, 413, 76–84. 104 J. Zhang, C. Q. Cui, T. B. Lim, E. Kang, Macromol. Chem. Phys. 2000, 201, 1653– 1661.

105 S. P. Vernekar, N. D. Ghatge, P. P. Wadgao-

knar, J. Polym. Sci. Part A: Polym. Chem. 1988, 26, 953–958. 106 F. De Schrijver, G. Smets, J. Polym. Sci. Part A: Polym. Chem. 1966, 4, 2201–2210. 107 J. P. Fouassier, D. Ruhlmann, B. Graff, F. Wieder, Prog. Org. Coatings 1995, 25, 169– 202. 108 Y. Yagci, Macromol. Symp. 2000, 161, 19–35. 109 O. W. Webster, Science 1991, 251(4996), 887– 893. 110 B. Ivan, Macromol. Symp. 1994, 88, 201–215. 111 K. Hatada, T. Kitayama, O. Vogl (Eds.), Macromolecular Design of Polymeric Materials, Marcel Dekker, Inc., New York, 1997. 112 C. J. Hawker, A. W. Bosman, E. Harth, Chem. Rev. 2001, 101, 3661–3688. 113 T. E. Patten, K. Matyjaszewski, Adv. Mater. 1998, 10, 901–915. 114 K. Matyjaszewski (Ed.), Controlled/Living Radical Polymerization: Progress in ATRP, NMP, and RAFT, American Chemical Society, Washington, D.C., Vol. 768, 2000. 115 H. S. Bisht, A. K. Chatterjee, J. Macromol. Sci.: Polym. Rev. 2001, C41, 139–173. 116 A. R. Kannurpatti, S. Lu, G. M. Bunker, C. N. Bowman, Macromolecules 1996, 29, 7310–7315. 117 Y. Nakayama, T. Matsuda, Macromolecules 1996, 29, 8622–8630. 118 H. J. Lee, T. Matsuda, J. Biomed. Mater. Res. 1999, 47, 564–567. 119 B. de Boer, H. K. Simon, M. P. L. Werts, E. W. van der Vegte, G. Hadziioannou, Macromolecules, 2000, 33, 349–356. 120 J. Feng, D. J. Dyer, Polym. Preprint 2003, 44(1), 516–517. 121 G. H\hner, Ch. WXll, M. Buck, M. Grunze, Langmuir 1993, 9, 1955–1958. 122 G. Odian, Principles of Polymerization, 3rd Ed., Wiley, New York, 1991, p. 286. 123 O. Prucker, J. ROhe, Langmuir 1998, 14, 6893–6898. 124 L. S. Penn, H. Huang, M. D. Sindkhedkar, S. E. Rankin, K. Chittenden, R. P. Quirk, R. T. Mathers, Y. Lee, Macromolecules 2002, 35, 7054–7066. 125 R. Paul, R. Schmidt, J. Feng, D. J. Dyer, J. Polym. Sci. Part A: Polym. Chem. 2002, 40, 3284–3291.

151

8

Recent Advances in the Synthesis and Rearrangement of Block Copolymer Brushes Stephen G. Boyes, Anthony M. Granville, Marina Baum, Bulent Akgun, Brian K. Mirous, and William J. Brittain

Abstract

The synthesis of tethered block copolymer brushes by the use of controlled/“living” free radical polymerization techniques presents many significant advantages over traditional free radical polymerization methods. The authors’ group has found the most versatile controlled/“living” free radical polymerization techniques to be atom transfer radical polymerization (ATRP) and reversible addition fragmentation transfer (RAFT) polymerization. Both diblock and ABA-type triblock copolymer brushes have been synthesized using either ATRP or RAFT. Of particular interest in the case of block copolymer brushes is their ability to reversibly rearrange upon treatment with selective solvents. This rearrangement of block copolymer brushes can result in the formation of unusual surface morphologies that have been attributed to the formation of either “pinned micelles” or “folded” structures. Alternatively, thermal rearrangement has also been demonstrated for diblock copolymer brushes containing a fluorinated block.

8.1

Introduction and Background

Polymer brushes refer to an assembly of polymer chains which are tethered by one end to a surface or interface [1,2]. Tethering of the chains in close proximity to each other forces the chains to stretch away from the surface in order to avoid overlapping. Polymer brushes have recently attracted considerable attention, and numerous studies have been conducted to examine their structure and novel properties [3–9]. Polymer brushes are typically synthesized by two different methods, namely physisorption and covalent attachment. Of these methods, covalent attachment is preferred as it overcomes the disadvantages of physisorption, which include thermal and solvolytic instabilities [10]. Covalent attachment of polymer brushes can be achieved by using either “grafting to” or “grafting from” techniques. The “grafting to” technique involves tethering preformed, end-functionalized polymer chains to a suitable substrate under appropriate conditions [11]. This technique often leads to

152

8 Recent Advances in the Synthesis and Rearrangement of Block Copolymer Brushes

low grafting density and low film thickness, as the polymer molecules must diffuse through the existing polymer film to reach the reactive sites on the surface. The steric hindrance for surface attachment increases as the tethered polymer film thickness increases. In order to overcome this problem, the “grafting from” approach can be used, and this has generally become the most attractive way to prepare thick, covalently tethered polymer brushes with a high grafting density [10]. The “grafting from” technique involves the immobilization of initiators onto the substrate, followed by insitu surface-initiated polymerization to generate the tethered polymer brush. Surface-immobilized initiators can be generated by either treating the substrate with plasma or glow discharge in the presence of a gas [12], or forming an initiator which contains self-assembled monolayers (SAMs) on the substrates [13,14]. As the chains are growing from the surface, the only limit to propagation is diffusion of monomer to the chain ends, thus resulting in thick tethered polymer brushes with high grafting density. Recent advances in polymer synthesis techniques have underlined the importance of controlled/“living” free radical polymerization, as it provides a number of advantages over traditional free radical methods [15]. One of the main advantages provided by a controlled/“living” free radical system for polymer brush synthesis is that of control over the brush thickness, this being achieved via the control of molecular weight and narrow polydispersities [13,14]. Another advantage of the controlled/“living” free radical system is the ability to produce polymer brushes of specific architectures. The most interesting architectures produced to date are the block copolymer brushes [7–9,16–27], and this is due mainly to vertical phase separation occurring when the block copolymer chains are tethered by one end to a surface or substrate. By changing the grafting density, chain length, relative block length, block composition, or the interaction energy between the blocks and the surrounding environment, the formation of a variety of novel well-ordered structures has been predicted in theoretical terms [3,4], and in some cases this has been demonstrated experimentally [7,8,16]. This chapter presents a brief overview of the authors’ findings in the synthesis of block copolymer brushes, tethered to flat substrates or surfaces, using controlled/ “living” free radical polymerization techniques. The techniques employed include atom transfer radical polymerization (ATRP) and reversible addition fragmentation transfer (RAFT) polymerization, both of which have been used to synthesize diblock and triblock copolymer brushes.

8.2 Controlled/“Living” Free Radical Polymerization

8.2

Controlled/“Living” Free Radical Polymerization 8.2.1

Atom Transfer Radical Polymerization (ATRP)

The basic mechanism of ATRP involves a reversible switching between two oxidation states of a transition metal complex (Scheme 8.1) [28]. The radicals, or the active species, are generated though a reversible redox process which is catalyzed by a transition metal complex (Mtn-Y/Ligand) that undergoes a one-electron oxidation with simultaneous abstraction of a transferable halogen, X, from a dormant species, R-X. Polymer chains grow by the addition of the intermediate radicals to monomers in a manner similar to conventional radical polymerization [15]. The equilibrium represented in Scheme 8.1 is predominantly shifted to the left (dormant) side so as to suppress termination and transfer reactions. Termination does occur, though in a well-controlled system it is limited to a few percent of the polymer chains. A small amount of termination is important as it generates oxidized metal complexes, XMtn+1, as persistent radicals to reduce the stationary concentration of growing radicals and thereby minimize the contribution of termination [29]. A successful ATRP will have a uniform growth of all the chains which is achieved by fast initiation and rapid, reversible deactivation.

Polymer-X

+

kact

Mtn-X/Ligand

+

Polymer kdeact

Mtn+1-X2/Ligand kterm dead polymer

kp Monomer Scheme 8.1

Basic mechanism of atom transfer radical polymerization.

S P

A

+

S

kexAB

S C

S

P

PB

C

A

kexAB

ka kf

P

B

Z

Z Monomer

+

Monomer dA

B k frA

B

rB A

S P

A

k ad

S C

BA

PB

Z Scheme 8.2

Basic mechanism of reversible addition fragmentation transfer polymerization.

153

154

8 Recent Advances in the Synthesis and Rearrangement of Block Copolymer Brushes

The basic mechanism of RAFT polymerizations involves a reversible addition fragmentation cycle, in which transfer of a dithioester moiety between active and dormant species maintains the controlled character of the polymerization (Scheme 8.2) [30–32]. A RAFT polymerization is usually carried out by the addition of a suitable RAFT agent (dithioester, trithiocarbonates, dithiocarbonates, or dithiocarbamates) to a conventional free radical polymerization mixture. The addition of the RAFT agent typically results in polymers of predetermined chain length and narrow polydispersity. Use of the RAFT process also results in polymer chains terminated by thiocarbonylthio segments which can be chain extended to yield a variety of copolymer structures.

8.3

Synthesis of Block Copolymer Brushes 8.3.1

Diblock Copolymer Brushes

Diblock copolymer brushes tethered to flat silicon substrates have been synthesized using both ATRP and RAFT techniques. The first diblock copolymer brushes synthesized in the authors’ group were made by a combination of carbocationic polymerization and ATRP (Figure 8.1) [25]. Zhao and co-workers [24] synthesized diblock copolymer brushes consisting of a tethered chlorine-terminated polystyrene (PS) block, produced using carbocationic polymerization, on top of which was added a block of either poly(methyl methacrylate) (PMMA), poly(methyl acrylate) (PMA) or poly((N,N¢-dimethylamino)ethyl methacrylate) (PDMAEMA), synthesized using ATRP. The thickness of the outer poly(meth)acrylate block was controlled by adding varying amounts of free initiator to the ATRP media. It has been reported that the addition of free initiator is required to provide a sufficiently high concentration of deactivator, which is necessary for controlled polymerizations from the surface [33]. The properties of some diblock copolymer brushes which have been synthesized are summarized in Table 8.1. The first diblock copolymer brush to be synthesized completely using controlled/ “living” free radical polymerization techniques in our group was by Sedjo and coworkers [23]. In this study, a tethered diblock copolymer of PS and PMMA was synthesized using a combination of reverse atom transfer radical polymerization (RATRP) and standard ATRP techniques (Figure 8.2) [23]. The properties of this diblock copolymer brush are listed in Table 8.1. RATRP involves initiation by conventional radical initiators in the presence of an ATRP deactivator. RATRP has been shown to produce polymers that are end-functionalized with a transferable halogen, thus allowing continued polymerization [34–36]. In order to perform RATRP from the surface, an azo-initiator was first immobilized on the silicon substrate, followed by the polymerization of styrene in the presence of copper(II) bromide and ligand. This resulted in the formation of a tethered block of PS with a terminal bromine group; the latter was subsequently used to initiate MMA under standard ATRP conditions.

8.3 Synthesis of Block Copolymer Brushes Table 8.1 Summary of the properties of diblock copolymer brushes.

Diblock copolymer brush structurea)

Thickness of Thickness of tethered blockb) outer blockb) (nm) (nm)

Polymerization techniquec)

Reference(s)

Si/SiO2//PS-b-PMMA Si/SiO2//PS-b-PMA Si/SiO2//PS-b-PDMAEMA Si/SiO2//PS-b-PMMA Si/SiO2//PS-b-PDMA Si/SiO2//PDMA-b-PMMA Si/SiO2//PS-b-P(t-BA) Si/SiO2//PS-b-PAA Si/SiO2//PMA-b-P(t-BA) Si/SiO2//PMA-b-PAA Si/SiO2//PS-b-PPFS Si/SiO2//PS-b-PHFA Si/SiO2//PMA-b-PPFS Si/SiO2//PMA-b-PHFA

28 24 27 25 11 11 21 21 14 14 16 10 11 15

Cationic/ATRP Cationic ATRP Cationic/ATRP RATRP/ATRP RAFT RAFT ATRP ATRP/Hydrolysis ATRP ATRP/Hydrolysis ATRP ATRP ATRP ATRP

24, 25 25 25 23 19 19 16 16 16 16 38 38 38 38

a)

b) c)

11 9 3 7 12 10 17 8 16 9 5 6 5 5

PS = polystyrene; PMMA = poly(methyl methacrylate); PMA = poly(methyl acrylate); PDMAEMA = poly((N,N-dimethylamino)ethyl methacrylate); PDMA = poly(dimethylacrylamide); P(t-BA) = poly(tert-butyl acrylate); PAA = poly(acrylic acid); PPFS = poly(pentafluorostyrene); PHFA = poly(heptadecafluorodecyl acrylate). Representative structure is Si/SiO2//tethered block-b-outer block. ATRP = atom transfer radical polymerization; RATRP = reverse atom transfer radical polymerization; RAFT = reversible addition fragmentation transfer polymerization.

O O

O

CH3

Si

OCD3 CH3

O

CH3

Styrene O TiCl4, -78 oC, CH2Cl2

Si

CH2 CH n Cl CH3

O

DtBP, 50 min

Methyl Methacrylate

O O

CH3

CH3

Si O

PMDETA CuBr

CH2 CH n

CH2

CH m

CH3

O OCH3

Figure 8.1 Synthesis of surface-immobilized diblock copolymer brush (Si/SiO2//PS-b-PMMA) using a combination of carbocationic polymerization and atom transfer radical polymerization (ATRP) [25].

Br

155

156

8 Recent Advances in the Synthesis and Rearrangement of Block Copolymer Brushes

To make further use of the azo-initiator, tethered diblock copolymers were prepared using RAFT polymerization. Baum and Brittain [19] were able to prepare diblock copolymer brushes of PS, PMMA and poly(dimethylacrylamide) (PDMA) from a surface immobilized azo-initiator in the presence of 2-phenylprop-2-yl dithiobenzoate as a chain transfer agent (Figure 8.3). The properties of the diblock copolymer brushes produced are listed in Table 8.1. The addition of “free” initiator, 2,2¢azobisisobutyronitrile (AIBN), was required in order to obtain a controlled polymerization, and this resulted in the formation of “free” polymer chains in solution. As the immobilized azo-initiator contains a cleavable ester group, Baum and Brittain [19] were able to compare the molecular weight of free polymer chains with that of the tethered polymer chains, by cleaving a homopolymer brush prepared on high surface area, nonporous silica prepared in the presence of “free” initiator. These results indicated that for homopolymer brushes of either PS or PMMA, both the number average molecular weight (Mn) and polydispersity (PDI) were comparable for Ofree’ polymer versus cleaved polymer [19].

O O

Si

CN (CH2)11 OCO(CH2)2

CN N

CH3

N

CH3

O

CH3 S

H3C Styrene

S CH3 CTA

O O

Si

CN (CH2)11 OCO(CH2)2

O

Dimethyl Acrylamide

O O

Si O

CH2

CH

Br

n

CH3

AIBN, CTA

CN (CH2)11 OCO(CH2)2

CH3 CH2

CH

n

CH2

CH3

CH m Br O N

H3C

Figure 8.2 Synthesis of surface-immobilized diblock copolymer brush (Si/SiO2//PS-b-PMMA) using reverse atom transfer radical polymerization [23].

CH3

8.3 Synthesis of Block Copolymer Brushes

In order to produce block copolymer brushes by ATRP directly from the surface, the ATRP initiator, (11-(2-bromo-2-methyl)propionyloxy)undecyltrichlorosilane, was prepared and immobilized on silicon substrates. From this immobilized bromo-isobutyrate-type ATRP initiator both Boyes et al. [37] and Granville et al. [38] were able to synthesize diblock copolymer brushes using ATRP. Boyes et al. [37] synthesized diblock copolymer brushes of either PS or PMA and poly(tert-butyl acrylate) (P(tBA)) using ATRP, with subsequent hydrolysis of the P(t-BA) to poly(acrylic acid) (PAA) (Figure 8.4). The properties of these diblock copolymer brushes are listed in Table 8.1. The diblock copolymer brushes, Si/SiO2//PS-b-PAA and Si/SiO2//PMA-bPAA, were both treated with aqueous silver acetate to produce polyelectrolyte diblock copolymer brushes. The polyelectrolyte brushes were subsequently reduced using H2 resulting in the formation of silver nanoparticles within the diblock copolymer brush [37]. Granville and colleagues [38] used similar ATRP techniques to synthesize diblock copolymer brushes that contained the fluorinated monomers pentafluorostyrene (PFS) and heptadecafluorodecyl acrylate (HFA). The properties of these diblock copolymer brushes are also listed in Table 8.1. The use of fluorinated monomers to

O O

Si

CN (CH2)11 OCO(CH2)2

CN N

CH3

N

CH3

O

CH3 S

H3C Styrene

S CH3 CTA

O O

Si

CN (CH2)11 OCO(CH2)2

O

Dimethyl Acrylamide

O O

Si O

CH2

CH

Br

n

CH3

AIBN, CTA

CN (CH2)11 OCO(CH2)2

CH3 CH2

CH

n

CH2

CH3

CH m Br O N

H3C

Figure 8.3 Synthesis of surface-immobilized diblock copolymer brush (Si/SiO2//PS-b-PDMA) using reverse addition fragmentation transfer polymerization [19].

CH3

157

158

8 Recent Advances in the Synthesis and Rearrangement of Block Copolymer Brushes

O O

Si

O

O

CH3

(CH2)11 O CH3

O

O Anisole, CuBr, PMDETA 90 oC, 24 h

O Si

O

CH3

(CH2) 11 O

CH2

CH n Br

CH3

Acetone, CuBr, PMDETA, 60 oC, 6 h

CH3

(CH2)11 O

CH2

CH n

CH2

CH3

O

Si O

t-Butyl Acrylate

O

O

Styrene

Br

CH m Br O OC(CH3)3

10% aq. HCl, Reflux, 12 h

O O

Si

O

CH3

(CH2) 11 O

CH2

CH n

CH2

CH3

O

CH m Br O OH

10 mM Ag(ac.)aq. 40 oC, 24 h O O

Si

O

CH3

(CH2)11 O

O

CH2

CH

n

CH2

CH3

CH m Br O + O - Ag

Figure 8.4 Synthesis of surface-immobilized polyelectrolyte diblock copolymer brush (Si/SiO2//PS-b-PAA(Ag+)) using atom transfer radical polymerization [37].

produce outer blocks of either poly(pentafluorostyrene) (PPFS) or poly(heptadecafluorodecyl acrylate) (PHFA) resulted in surfaces that were highly hydrophobic [38]. 8.3.2

Triblock Copolymer Brushes

In general, very few reports have been made relating to the synthesis of tethered triblock copolymer brushes [16,27]. The authors’ group has used surface-immobilized ATRP techniques to produce tethered triblock copolymer brushes, while Boyes et al. [16] synthesized ABA-type triblock copolymer brushes of PS and PMA via sequential monomer addition to a self-assembled monolayer (SAM) of a bromo-isobutyrate ATRP initiator (Figure 8.5). The properties of the Si/SiO2//PS-b-PMA-b-PS and Si/ SiO2//PMA-b-PS-b-PMA brushes are listed in Table 8.2; these data indicate that, for the Si/SiO2//PS-b-PMA-b-PS brush, there appears to be incomplete reinitiation in the formation of the third block, as although a PS thickness of approximately 20 nm was targeted, a thickness of only 3 nm was obtained. The incomplete reinitiation was attributed to radical-radical termination occurring in the formation of the previous blocks, resulting in tethered chains that were unable to reinitiate [16]. In the

8.3 Synthesis of Block Copolymer Brushes

case of the Si/SiO2//PMA-b-PS-b-PMA brush, the outer PMA block had a thickness of 15 nm, which is close to the target thickness of 20 nm, indicating that the degree of termination occurring in this system was less.

O O

Si

O

O

CH3

(CH2)11 O

Br

O Anisole, CuBr, PMDETA 90 oC, 24 h

CH3

O

O Si

O

O O

Si

O

CH2

CH n

CH2

CH n Br

CH m Br O OCH3

Anisole, CuBr, PMDETA 90 oC, 24 h

CH3

(CH2) 11 O

CH2

CH

n

CH2

CH3

O

CH2 CH3

Anisole, CuBr, PMDETA, 90 oC, 24 h

CH3

Styrene

CH3

(CH2) 11 O

CH3

(CH2)11 O

O

Si O

Methyl Acrylate

O

O

Styrene

CH m O

CH2

CH p Br

OCH3

Figure 8.5 Synthesis of surface-immobilized ABA type triblock copolymer brush (Si/SiO2//PS-b-PMA-b-PS) using atom transfer radical polymerization [16].

Table 8.2

Summary of the properties of tethered ABA-type triblock copolymer brushesa).

Layer

Water contact angleb) ha hr

Thicknessc) (nm)

Molecular weight (Mn) Expt.e) Calcd.d)

PS PS-b-PMA PS-b-PMA-b-PS PMA PMA-b-PS PMA-b-PS-b-PMA

97 73 88 72 87 73

20 18 3 20 23 15

13.2 P 17.9 P 21.7 P 15.2 P 21.8 P 18.4 P

a) b) c) d) e)

87 63 70 59 72 58

From Ref. [16]. The standard deviation of contact angles was <2R. Thickness determined by ellipsometry. Calculated using the conversion determined from the free polymer. Experimental molecular weight of the free polymer.

103 103 103 103 103 103

10.5 P 17.2 P 18.6 P 13.6 P 18.7 P 15.5 P

103 103 103 103 103 103

PDI

1.10 1.23 1.07 1.50 1.08 1.23

159

160

8 Recent Advances in the Synthesis and Rearrangement of Block Copolymer Brushes

8.4

Rearrangement of Block Copolymer Brushes

The behavior of tethered block copolymer brushes is potentially very interesting due to the fact that the polymer chains are covalently bonded to the surface and are vertically micro-phase separated due to the immiscibility of the different blocks. Theoretical work using self-consistent field calculations, computer simulations and scaling arguments have demonstrated that the behavior of tethered diblock copolymer brushes is complex and can depend on a variety of factors [3,4]. Balazs and co-workers [3] have demonstrated that changes in chain architecture, grafting density, composition, overall molecular weight and the interaction energies between blocks and also between blocks and solvent can result in a variety of novel structures. One of the most interesting structures is the Opinned micelle’ structure, which can be formed when tethered AB diblock copolymer brushes are treated with a solvent which is a poor solvent for the outer B block but a good solvent for the tethered A block [3]. 8.4.1

Rearrangement of Diblock Copolymer Brushes

Experimentally, very few investigations have been conducted into the rearrangement of tethered diblock copolymer brushes. The first report of nanopattern formation from a tethered diblock copolymer brush was by Zhao and co-workers [7,8]. Using the combination of carbocationic and ATRP techniques described above (see Figure 8.1), Zhao et al. [24] produced a tethered Si/SiO2//PS-b-PMMA brush where the PS thickness was 23 nm and the PMMA thickness was 14 nm. When this diblock copolymer brush was treated with dichloromethane (which is a good solvent for both PS and PMMA), water contact results indicated a characteristic advancing contact angle of PMMA, 74R, while atomic force microscopy (AFM) analysis indicated that the surface of the brush was smooth (Figure 8.6(a)) [8]. When the same sample is treated with mixed solvents of dichloromethane and cyclohexane, and the percentage of cyclohexane is gradually increased, the advancing water contact of the brush increased to 120R, and the tethered diblock brush was seen to reorganize, resulting in the formation of a regular nanopattern on the surface (Figure 8.6(b)) [8]. Zhao and colleagues speculated that, with an increasing cyclohexane content, the PMMA blocks would collapse and aggregate to form a core, so as to avoid contact with the cyclohexane (Figure 8.7) [7]. Baum and Brittain [19] also attempted to rearrange their diblock copolymers made using RAFT polymerizations (see Figure 8.3 and Table 8.1). Upon treatment with methylcyclohexane (a good solvent for PS and a non-solvent for PDMA) at 35 RC, the advancing water contact angle of the Si/SiO2//PS-b-PDMA brush increased from 42R to 65R. It appears as though there was incomplete rearrangement of the brush after treatment with methylcyclohexane, as the advancing contact angle did not reach the characteristic contact angle of PS (approx. 100R). Treatment of the same sample with tetrahydrofuran (THF)/water (1/1, v/v) at 35 RC reversed the con-

8.4 Rearrangement of Block Copolymer Brushes

(a)

(b) Figure 8.6 Atomic force microscopy (AFM) images of tethered Si/SiO2//PS-b-PMMA brush after: (a) treatment with dichloromethane; and (b) gradual treatment with cyclohexane [8].

Figure 8.7 Speculative model for nanopattern formation from tethered Si/SiO2//PS-b-PMMA brush [7].

161

162

8 Recent Advances in the Synthesis and Rearrangement of Block Copolymer Brushes Table 8.3

Solvent rearrangement of diblock copolymer brushes containing PPFSa)

Solventb)

Si/SiO2//PS-b-PPFS hr ha

Si/SiO2//PMA-b-PPFS hr ha

1st Fluorobenzene 1st Cyclohexane/acetone 2nd Fluorobenzene 2nd Cyclohexane/acetone

121 101 119 102

115 79 118 82

a) b)

90 85 88 87

88 67 92 68

From Ref. [38,39]. Sample immersed in solvent at 60 RC for 1 h.

tact angle back to the original value. Similar results were seen for the Si/SiO2// PDMA-b-PMMA brush. In this case, the advancing water contact angle decreased from 66R to 58R after the sample was treated with THF/H2O (1/1, v/v) at 35R and returned to the original contact angle value after treatment with dichloromethane. Again, it appears as though this sample has incomplete rearrangement, as the contact angles achieved after solvent treatment do not match the characteristic contact angles of either PDMA (~40R) or PMMA (~74R). No AFM images were taken of these samples. Granville and colleagues [39] investigated the potential to rearrange diblock copolymer brushes consisting of a tethered hydrocarbon block and an outer fluorocarbon block, by using both solvent and thermal treatments. The diblock copolymer brushes with PPFS as the outer block and either PS or PMA as the tethered block were found to undergo near-complete rearrangement upon treatment with a selective solvent for the tethered block (Table 8.3). By contrast, the diblock copolymer brushes with PHFA as the outer block and either PS or PMA as the tethered block did not completely rearrange upon treatment with a selective solvent for the tethered block (Table 8.4). The incomplete rearrangement of the diblock copolymer brushes containing PHFA was attributed to the large interaction parameter between blocks as a result of the long fluorocarbon chain on the HFA monomer. As an alternative switching mechanism, Granville et al. [38] used thermal treatment of samples that had been rearranged using solvent, to reorganize the surface composition (Figure 8.8). Thermal treatment of the solvent-rearranged Si/SiO2//PS-b-PPFS brush at 100 RC for 20 min, after it had been treated with cyclohexane at 35 RC, resulted in an Table 8.4

Solvent rearrangement of diblock copolymer brushes containing PHFAa)

Solventb)

Si/SiO2//PS-b-PPFS hr ha

Si/SiO2//PMA-b-PPFS hr ha

1st Trifluorotoluene 1st Cyclohexane/ethyl acetate 2nd Trifluorotoluene 2nd Cyclohexane/ethyl acetate

127 110 125 111

126 110 125 109

a) b)

From Ref. [38]. Sample immersed in solvent at 60 RC for 1 h.

98 92 97 92

100 90 100 91

8.4 Rearrangement of Block Copolymer Brushes

Figure 8.8

Thermal rearrangement of fluorinated diblock copolymer brushes [38].

increase in the advancing contact angle from 101R to 121R. Similar results were seen for the Si/SiO2//PMA-b-PPFS brush, although thermal treatment at 60 RC for 5 min was all that was required to increase the advancing contact angle from 80R to 118R. The only report on the rearrangement of triblock copolymer brushes has been by Boyes and colleagues [16], who were able to rearrange their ABA-type triblock copolymer brushes by solvent treatment. Each of the triblock copolymer brushes (see Table 8.2) was treated with a solvent which was effective for the middle block but was a nonsolvent for the tethered and outer blocks. After treatment of the Si/SiO2// PMA-b-PS-b-PMA brush with cyclohexane, the advancing water contact angle increased from 74R to 94R, which is close to the characteristic contact angle for PS

(a)

(b)

Figure 8.9 Atomic force microscopy (AFM) images of tethered Si/SiO2//PMAb-PS-b-PMA brush after: (a) treatment with dichloromethane; and (b) gradual treatment with cyclohexane [16].

163

164

8 Recent Advances in the Synthesis and Rearrangement of Block Copolymer Brushes

and AFM analysis, thereby demonstrating that the surface went from smooth (Figure 8.9(a)) to an unusual nanomorphology with greater roughness (Figure 8.9(b)). Similar results were seen for the Si/SiO2//PS-b-PMA-b-PS brush after treatment with acetone; however, in this case the advancing water contact angle increased from 70R to 92R. The formation of this unusual nanomorphology was speculated to be a “folded” brush structure, where the two end blocks aggregate and force the mid-block to the air interface (Figure 8.10) [16].

Cyclohexane or Acetone

CH2Cl2

Proposed response of tethered triblock copolymer brushes to different solvents [16].

Figure 8.10

Summary

The use of controlled/“living” free radical polymerization techniques to synthesize tethered block copolymer brushes offers many advantages over traditional radical polymerization methods. The results of the present studies have shown RAFT and ATRP to be the most versatile polymerization methods to produce block copolymer brushes. Using these techniques, a series of both diblock and ABA-type triblock copolymer brushes has been synthesized. Most of these systems display reversible rearrangement of block segments upon treatment with selective solvents. Diblock copolymer brushes containing fluorinated blocks demonstrate an ability to rearrange by thermal treatment after the surface has been treated with solvent. In some of the

References

rearranged systems, unusual nanomorphologies were observed that were attributed to the formation of either “pinned micelles” or “folded” structures.

References 1 S. T. Milner, Science 1991, 251, 905. 2 A. Halpern, M. Tirrell, T. P. Lodge, Adv.

Polym. Sci. 1992, 100, 31. 3 E. B. Zhulina, C. Singh, A. C. Balazs, Macromolecules 1996, 29, 6338. 4 E. B. Zhulina, C. Singh, A. C. Balazs, Macromolecules 1996, 29, 8254. 5 Y. Ito, S. Nishi, Y. S. Park, Y. Imanishi, Macromolecules 1997, 30, 5856. 6 Y. S. Park, Y. Ito, Y. Imanishi, Macromolecules 1998, 31, 2606. 7 B. Zhao, W. J. Brittain, W. Zhou, S. Z. D. Cheng, Macromolecules 2000, 33, 8821. 8 B. Zhao, W. J. Brittain, W. Zhou, S. Z. D. Cheng, J. Am. Chem. Soc. 2000, 122, 2407. 9 X. Kong, T. Kawai, J. Abe, T. Iyoda, Macromolecules 2001, 34, 1837. 10 B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677. 11 P. Mansky, Y. Liu, E. Huang, T. P. Russell, C. J. Hawker, Science 1997, 275, 1458. 12 Y. Ito, Y. Ochiai, Y. S. Park, Y. Imanishi, J. Am. Chem. Soc. 1997, 119, 1619. 13 O. Prucker, J. RUhe, Macromolecules 1998, 31, 592. 14 O. Prucker, J. RUhe, Macromolecules 1998, 31, 602. 15 K. Matyjaszewski, in: Controlled/Living Radical Polymerization (Ed.: K. Matyjaszewski), ACS Symposium Series 768, American Chemical Society; Washington DC, USA, 2000, p. 2. 16 S. G. Boyes, W. J. Brittain, X. Weng, S. Z. D. Cheng, Macromolecules 2002, 35, 4960. 17 K. Matyjaszewski, P. J. Miller, N. Shukla, B. Immaraporn, A. Gelman, B. B. Luokala, T. M. Siclovan, G. Lickelbick, T. Vallant, H. Hoffmann, T. Pakula, Macromolecules 1999, 32, 8716. 18 M. Ejaz, S. Yamamoto, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules 1998, 31, 5934. 19 M. Baum, W. J. Brittain, Macromolecules 2002, 35, 610. 20 D. M. Jones, W. T. S. Huck, Adv. Mater. 2001, 13, 1256.

21 V. L. Osborne, D. M. Jones, W. T. S. Huck,

Chem. Commun. 2002, 1838. 22 M. R. Tomlinson, J. Genzer, Chem. Commun.

2003, 1350. 23 R. A. Sedjo, B. K. Mirous, W. J. Brittain, Mac-

romolecules 2000, 33, 1492. 24 B. Zhao, W. J. Brittain, Macromolecules 2000,

33, 8813. 25 B. Zhao, W. J. Brittain, J. Am. Chem. Soc.

1999, 121, 3557. 26 D. M. Jones, A. A. Brown, W. T. S. Huck,

Langmuir 2002, 18, 1265. 27 J.-B. Kim, W. Huang, M. L. Bruening,

G. L. Baker, Macromolecules 2002, 35, 5410. 28 K. Matyjaszewski, J. Xia, Chem. Rev. 2001,

101, 2921. 29 H. Fischer, Chem. Rev. 2001, 101, 3581. 30 B. Y. K. Chong, T. P. T. Le, G. Moad, E. Riz-

31 32

33

34

35 36 37 38

39

zardo, S. H. Thang, Macromolecules 1999, 32, 2071. D. G. Hawthorne, G. Moad, E. Rizzardo, S. H. Thang, Macromolecules 1999, 32, 5457. R. T. A. Mayadunne, E. Rizzardo, J. Chiefari, Y. Kwong Chong, G. Moad, S. H. Thang, Macromolecules 1999, 32, 6977. M. Husseman, E. E. Malmstrom, M. McNamara, M. Mate, O. Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang, T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424. G. Moineau, P. Dubois, R. Jerome, T. Senninger, P. Teyssie, Macromolecules 1998, 31, 545. J.-S. Wang, K. Matyjaszewski, Macromolecules 1995, 28, 7572. J. Xia, K. Matyjaszewski, Macromolecules 1997, 30, 7692. S. G. Boyes, B. Akgun, W. J. Brittain, M. D. Foster, Macromolecules 2003, 36, 9539. A. M. Granville, S. G. Boyes, B. Akgun, W. J. Brittain, M. D. Foster, Macromolecules 2004, 37, 2790. A. M. Granville, S. G. Boyes, B. Akgun, M. D. Foster, W. J. Brittain, submitted to Adv. Func. Mater.

165

167

9

Surface-Grafted Hyperbranched Polymers Hideharu Mori and Axel H. E. Mller

Abstract

Surface-grafted hyperbranched polymers are a type of polymer brush, which have characteristic branched architectures and unique properties. Recent advances in this field are summarized in terms of the preparation methods, which can be divided into three categories, namely “grafting to”, “grafting from”, and multi-step grafting approaches. The growth of three-dimensional nanostructures (dendrimers and hyperbranched layers) on surfaces has been used to obtain various types of layered materials on planar or curved surfaces. This chapter focuses on surface-initiated, self-condensing vinyl (co)polymerization, which is a primary example of a one-step “grafting from” approach and an effective method for the synthesis of hyperbranched polymers linked covalently to a surface. Molecular parameters of the surface-grafted hyperbranched polymers obtained by self-condensing vinyl (co)polymerization are discussed from a theoretical point of view.

9.1

Introduction

Highly branched polymers are of considerable scientific and industrial interest, due to their low intrinsic viscosity, high solubility and miscibility, and their potential as polyfunctional carriers. The interest in hyperbranched polymers arises from the fact that they combine some features of dendrimers – for example, an increasing number of end groups and a compact structure in solution – with the ease of preparation of linear polymers by means of a one-pot reaction. However, the polydispersities are usually high and their structures are less regular than those of dendrimers, which are monodisperse molecules with well-defined, perfectly branched architectures. Dendrimers have a highly compact and globular shape, and are produced in a multi-step organic synthesis. During the past decade, the field of arborescent polymers (dendrimers, hyperbranched, and highly branched polymers) has become well established, with a large variety of synthetic approaches, fundamental studies on the structure and properties of these unique materials, and the identification of possible applications for these materials [1–4].

168

9 Surface-Grafted Hyperbranched Polymers

Surface grafting of polymer chains has been investigated as an effective and versatile method for the manipulation and control of surface properties. In recent years, much attention has been paid to the use of controlled/“living” polymerizations from planar and spherical surfaces [5,6], because this allows better control over the molecular weights and molecular weight distribution of the target polymer. By using these techniques, a high grafting density and a controlled film thickness can be obtained, as such brushes consist of end-grafted, strictly linear chains of the same length and the chains are forced to stretch away from the flat surface. Many research groups have recently reported the application of controlled/living polymerization systems for the synthesis of a variety polymer brushes on two-dimensional (2D) and threedimensional (3D) substrates. For recent representations of progress in this field, the reader is referred to other chapters of this book. These techniques have also been successfully applied for the preparation of the linear polymers grafted on a linear polymer chain with high density, resulting in cylindrical polymer brushes [7–11]. The surface chemistry and interfacial properties of hyperbranched polymers have also become a field of growing interest [2,12]. In recent years, much interest has been paid to highly branched polymers grafted chemically onto surfaces, as their distinctive chemical and physical properties can be used advantageously as functional surfaces and as interfacial materials. Such surface-grafted hyperbranched polymers can be regarded as a type of polymer brush, as they refer to an assembly of polymer chains which are tethered by one end to a surface or an interface. The typical structures of hyperbranched, branched, and linear polymers grafted onto surfaces are summarized in Figure 9.1. Depending upon the substrates, they can be divided into 3D, 2D, and one-dimensional (1D) hybrids, which correspond to products grafted onto spherical particles, planar surfaces, and linear polymers, respec-

3D

2D

1D

high

branched

low

Figure 9.1 Surface-grafted hyperbranched, highly branched, and linear polymers: from one-dimensional (1D) to three-dimensional (3D).

9.1 Introduction (a) “Grafting to” approach

grafting to

Y X

X

X

X

(reaction of X with Y)

X

X

X

X

X

I

I

I

I

(b) “Grafting from” approach IM IM I

I

grafting from

IM

IM I

I

Surface-initiated polymerization

I

(c) Multi-step grafting approach Y

Y

Y

grafting to

modifications

X X X

Y X

X

X

X

X

M M

X

X

X

X

X

X

modifications

grafting from

M

X

X

I I I

I I

M

X X X

I I

I

I

I

I

I

I

I

I

I

I

I

I

Scheme 9.1 Synthesis of surface-grafted hyperbranched polymer by: (a) “grafting to”; (b) “grafting from”; and (c) multi-step grafting approaches.

tively. The growth of 3D nanostructures (dendrimers and hyperbranched layers) onto and/or from surfaces has been used to obtain various types of layered materials on planar or curved surfaces. There are basically three routes for preparing 2D and 3D hybrids with surfacegrafted hyperbranched polymers: (1) “grafting to”; (2) “grafting from”; and (3) multistep grafting methods (Scheme 9.1). The concept of “grafting to” and “grafting from” in the synthesis of surface-grafted hyperbranched polymers is basically the same as that for conventional linear polymer brushes. The one-step “grafting to” approach involves a reaction (or interaction) of one or several reactive groups of hyper-

169

170

9 Surface-Grafted Hyperbranched Polymers

branched polymers with functional groups on the substrate. In contrast, the “grafting from” technique is performed by an in-situ surface-initiated polymerization from immobilized initiators. For the preparation of surface-grafted hyperbranched polymers, the branched structures should be formed during the polymerization process, and therefore small multifunctional molecules, which can afford branching points, are required. The multi-step grafting method involves a series of repeated “grafting to” or “grafting from” steps. In this case, the branched architecture is formed during the repeated reactions. The scope of this chapter will cover the surface-grafted hyperbranched polymers that have been prepared using the techniques.

9.2

“Grafting To” Approach

Similar to the synthesis of linear polymer brushes by using the “grafting to” technique, surface-grafted hyperbranched polymers can be prepared by the adsorption of hyperbranched polymers onto surfaces, or by the reaction of one or several group of hyperbranched polymers with functional groups on a substrate. Only the latter case can provide hyperbranched polymers linked covalently to a surface. The “grafting to” technique has been mainly employed for the synthesis of dendrimers that are chemically attached to surfaces (2D and 3D). Several research groups have also focused on investigating the physical properties of dendrimers [12–15] and hyperbranched polymers [2,16,17] adsorbed onto solid surfaces. In general, these systems are limited to monolayers, as further grafting is hindered by the polymer chains already adsorbed onto the surface. Dendrimers grafted on a linear polymer chain (1D hybrid) have been also synthesized using the “grafting to” technique, in which attachment of prefabricated dendron building blocks onto a reactive polymer chain is achieved by a polymer-analogous reaction [18,19]. However, the problem in using this approach is the limited conversion, leading to incomplete coverage of the backbone anchor groups with dendrons. This drawback is more significant if a higher generation of dendrons are employed [18,20]. In order to avoid the problem, the so-called macromonomer route (“grafting through”) has been used preferably, in which monomers already carrying dendrons are subjected either to polymerization or to polycondensation [7,19,21]. 9.2.1

Synthesis of 2D Hybrids by “Grafting To”

The synthesis of dendrimers grafted onto surfaces has been mainly conducted using the “grafting to” technique [12–14,22,23]. Due to the highly compact and globular shape, as well as the monodispersity, dendrimers attached to flat surfaces are useful for many applications, such as data storage or nanolithography systems [12]. Similar to conventional self-assembled monolayers of small organic molecules on a planar surface, there are generally two systems that have undergone extensive study: (1) al-

9.2 “Grafting To” Approach

kanethiols on gold; and (2) alkylsiloxanes on hydroxylated silicon surfaces. For example, Gorman and co-workers [24] have investigated self-assembled monolayers of dendrimers, functionalized at the focal point of poly(ether) dendrons with a thiol group, on gold. On the other hand, FrHchet et al. [25] designed modified poly(benzyl ether) dendrimers covalently tethered to a silicon substrate that could serve as passivation resists in scanning probe lithography. For that purpose, they prepared the dendrimers functionalized at the focal point with a covalent tether, consisting of a long alkyl chain derivatized with a terminal chlorosilane coupling agent. The poly(benzyl ether) dendrons containing a tethered carboxylic acid moiety at their focal point have been used as a monolayer of dendrons which could be assembled onto an aminated silicon wafer surface prepared by pre-treatment of the clean silicon surface with (3-aminopropyl)triethoxysilane [22]. Similarly, Wells and Crooks [26] have reported the preparation of dendrimer selfassembled monolayers by using amine-terminated poly(amidoamine) dendrimers with gold modified with a mercaptoundecanoic acid. This was accomplished by forming amide bonds between the peripheral amino-groups of the dendrimer and the carboxylic acid groups on a gold surface. The chemical and physical properties of amine-terminated [13,27] and thiol-terminated [28,29] poly(amidoamine) dendrimers attached to gold surfaces have been thoroughly examined by Crooks and coworkers. One interesting example of applications involving dendrimers grafted onto a flat surface includes the use of biomolecule-functionalized dendrimer monolayers for the development of affinity biosensors. For example, Niemeyer et al. [30,31] have recently demonstrated a novel method for the surface immobilization of DNA using prefabricated poly(amidoamine) starburst dendrimers as mediator moieties. Dendrimers containing 64 primary amino groups in their outer sphere were covalently attached to silylated glass supports and, subsequently the dendritic macromolecules were modified with glutaric anhydride and activated with N-hydroxysuccinimide. These authors claimed that the resulting surfaces revealed both a very high immobilization efficiency for amino-modified DNA-oligomers, and also a remarkably high stability during repeated regeneration and re-using cycles. 9.2.2

Synthesis of 3D Hybrids by “Grafting To”

The “grafting to” technique has been also employed for the synthesis of 3D hybrid systems. For example, Guo and Yu [32] recently reported the synthesis of FrHchettype polyether dendrons bearing COOH groups at the focal points, which were then grafted onto silica nanoparticles premodified with 3-aminopropyltriethoxysilane by using N,N-dicyclohexylcarbodiimide-mediated amidation. The immobilization of novel amino acid (l-glutamic acid derivatives)-based chiral dendrimers with peptide linkages was accomplished using the “grafting to” method involving the carboxylderivatized dendrimers with an aminopropyl-modified silica [33]. These authors demonstrated that the immobilized chiral dendrimers constitute a new class of chiral stationary phases for use in HPLC. The dendritic branches composed of enantio-

171

172

9 Surface-Grafted Hyperbranched Polymers

merically pure l-lysine building blocks have been immobilized on aminopropylmodified silicas [34]. The reaction was achieved using standard peptide coupling methodology between NH2 groups on the silica surface and the COOH group at the focal point of the dendric branch.

9.3

Multi-Step Grafting Approach

The second method is a repeated combination of “grafting to” and “grafting from“ approaches. The first step is a formation of the linear polymer brushes by a polymerization (“grafting from”) or a reaction (“grafting to”). Modification steps provide new sites on the resulting surface-grafted polymer chains, at which point the next series of the polymerization or the reaction can take place. Repetition of these steps can produce surface-grafted hyperbranched polymers. This multi-step grafting approach can be divided into two categories; “grafting to-grafts” and “grafting from-grafts”. To date, only a few systems have been reported for the preparation of surfacegrafted hyperbranched polymers. In these cases, the branched architecture is formed during the repeated reactions, and the film thickness and grafting density increase with increasing the number of repeats (or generation), whereas the degree of branching is independent of the number of repeated cycles. Theses approaches are useful for the preparation of surface-grafted hyperbranched polymers, particularly when a higher grafting density or tightly packed multilayer is required. However, they have the inherent disadvantage that many tedious synthetic steps are necessary to reach the defined surface structures. 9.3.1

“Grafting To-Grafts”

This approach involves a series of repeated “grafting to” steps. Zhou et al. [35] reported the preparation of a highly branched poly(acrylic acid) film attached to a self-assembled monolayer of mercaptoundecanoic acid on gold using this method. The synthetic procedure involves: (1) activation of the carboxylic acid groups in the monolayer via a mixed anhydride; (2) reaction with a,x-diamino-terminated poly(tert-butyl acrylate) to give a grafted polymer layer; and (3) hydrolysis of the tert-butyl esters to yield a poly(acrylic acid) graft. Repetition of these steps produced additional grafting at multiple sites on each prior graft, leading to hyperbranched polymer films. The advantage of the system is that each new layer contains more polymer branches than the previous one, and thus is more tightly packed. These authors showed that the hyperbranched polymers can be covalently modified with a broad range of functional groups, such as fluorophores, electroactive groups, perfluorinated moieties, dyes, and even with other polymers [36–39]. The surface-confined hyperbranched polymers are suitable for a number of technical applications, including corrosion inhibition, chemical sensing, cellular engineering, and micrometerscale patterning [37,40–43]. The same strategy has been employed for the synthesis

9.4 “Grafting From” Approach

of hyperbranched poly(acrylic acid)s grafted on polyethylene [44,45], polypropylene [46], and porous alumina supports [47]. Another example is the preparation of hyperbranched dendritic poly(amidoamine) using the divergent approach from terminal amino groups grafted onto an ultrafine SiO2 surface [48,49]. The preparation was achieved by repeating two processes: (1) Michael addition of methyl acrylate to surface-bound amino groups; and (2) amidation of the resulting esters with ethylenediamine or hexamethylenediamine. The repeating process can be regarded as a “grafting to-grafts” approach. Because the procedure is also similar to divergent approach for dendrimer synthesis, it can be rather expressed as “grafting from”/divergent synthesis. These authors claimed that both the amount of amino groups and the percentage of the grafted poly(amidoamine) increase with an increase in the number of generations. The poly(amidoamine) grafted on the surface is a hyperbranched polymer rather than a precise dendrimer. A similar methodology has been used to prepare dendric poly(amidoamine) grafted on carbon black containing amino groups on the surface [50]. 9.3.2

“Grafting From-Grafts”

The “grafting from-grafts” method relies on the combination of surface-initiated polymerization from immobilized sites and subsequent modifications of the resulting polymer chain to provide new sites where the next series of the polymerization can be started. Matsuda and colleagues [51,52] have reported the synthesis of surfacegrafted hyperbranched polymers by using iniferter (initiator-transfer agent-terminator)-based “living” radical graft copolymerization via photolysis of the benzyl N,Ndiethyldithiocarbamate group. The principle is based on a sequential reaction of iniferter copolymerization with chloromethylstyrene, and subsequent dithiocarbamylation of the chloromethylstyrene units in the copolymers. A variety of vinyl monomers could be used as a functional comonomer, including N,N-dimethylacrylamide, N,N-dimethylaminoethyl methacrylate, and sodium methacrylate. After a stem (parent) chain was progressively propagated from an iniferter-immobilized surface under UV irradiation, chloromethylstyrene units in graft chains were dithiocarbamylated. Subsequently, a branch (daughter) chain was progressively propagated on multiply derivatized iniferter units in the stem chains. The repeated cycles of photopolymerization/dithiocarbamylation provided successively higher generations of graft architectures. The authors indicated that the chain length for both parent and daughter chains was controlled by photoirradiation time, and the degree of branching was determined by the composition of chloromethylstyrene units.

9.4

“Grafting From” Approach

The one-step “grafting from” approach relies basically on in-situ, surface-initiated polymerization. In order to achieve branched structures, a small multifunctional

173

174

9 Surface-Grafted Hyperbranched Polymers

molecule is required, which includes ABx monomers, AB* initiator-monomers (inimers), and cyclic inimers. These molecules form the branch points in the resulting hyperbranched polymers, which are formed during the growth of polymer chains. For the preparation of conventional hyperbranched polymers, several strategies are currently employed. The polymerization reactions are classified into three categories [1,4,53,54]: (1) Step-growth polycondensation of ABx monomers; (2) chain-growth self-condensing vinyl polymerization (SCVP) of AB* initiator-monomers (“inimers”); and (3) chain-growth self-condensing, ring-opening polymerization of cyclic inimers. The step-growth and chain growth methods are compared in Table 9.1. Although the most common method is the polycondensation of ABx monomers, vinyl monomers cannot be polymerized using that approach. The recent discovery of SCVP made it possible to use vinyl monomers for a convenient, one-pot synthesis of hyperbranched vinyl polymers with degree of a branching (DB) £ 0.5 [55–62]. Initiator-monomers (“inimers”) of the general structure AB* are used, where the double bond or heterocycle is designated A, and B* is a group capable of being activated to initiate the polymerization of vinyl groups or heterocycle. A variety of inimers have been reported, which include (meth)acrylate-type, styrene-type, and vinyl ether-type inimers [54]. Cationic [55], anionic [56], group transfer [57], controlled radical [58–62], and ring-opening mechanisms [63] have been used. In an ideal SCVP process, living polymerization systems are preferred in order to avoid crosslinking reactions and gelation caused by chain transfer or recombination reactions. By copolymerizing AB* inimers with conventional monomers, this technique was extended to self-condensing vinyl copolymerization (SCVCP), leading to highly branched copolymers with DB controlled by the comonomer ratio [64–67]. Depending on the chemical nature of the comonomer, various functional groups can be incorporated in the branched polymer, leading to highly branched polymers. Table 9.1

Classification of different types of monomers for the synthesis of hyperbranched poly-

mersa). [88]

Step growth AB2

A

OH

OSiMe3

B

HOOC

ClOC

B

I H

OH

OSiMe3

I

Self-condensing vinyl polymerization [68]

[68,70] O

B*

Chain growth AB*

Cl

O

O

Br

O

O

Br O

Self-condensing ring-opening polymerization [69] B*

a)

O O

[90]

OH

H

O OH

N

O

The monomers indicated with reference number were employed for the synthesis of surface-grafted hyperbranched polymers.

OH

O

O

9.4 “Grafting From” Approach

By using these techniques in combination with polymerization from initiators immobilized on the surface, a variety of surface-grafted hyperbranched polymers have been synthesized. The procedure is just a one-pot polymerization, which is significantly different from other approaches. As mentioned above, generally many tedious synthetic steps are necessary to reach the defined surface structures. We describe here a convenient synthetic approach for preparing hyperbranched (meth)acrylates on 2D and 3D surfaces in which a silicon wafer or silica nanoparticles grafted with an initiator layer composed of an a-bromoester fragment were used for SCVP via atom transfer radical polymerization (ATRP) (Scheme 9.2) [54]. SCVCP was also applied as a method for the synthesis of highly branched polymers grafted from surfaces. In contrast, surface-initiated ATRP resulted in the preparation of linear polymer brushes.

(a) SCVP I I I I I I

I

I

(b) SCVCP I

I

I I

I

I

I = ATRP initiator

(c) ATRP

Scheme 9.2 Synthesis of hyperbranched, highly branched, and linear polymer brushes from planar surfaces and spherical particles via surface-initiated polymerization. a) self-condensing vinyl polymerization (SCVP), b) self-condensing vinyl copolymerization (SCVCP), c) atom transfer radical polymerization (ATRP).

9.4.1

Synthesis of 2D Hybrids by Surface-Initiated, Self-Condensing Vinyl (Co)polymerization

The synthesis of hyperbranched (meth)acrylates grafted onto a planar surface was conducted by SCVP via ATRP from a silicon substrate [68]. The reaction mechanism is shown in Scheme 9.3. 2-(2-Bromopropionyloxy)ethyl acrylate (BPEA) and 2-(2-bromoisobutyryloxy)ethyl methacrylate (BIEM) were used as (meth)acrylic AB* inimers, where the double bond is designated A, and B* is a group capable of initiating the polymerization of vinyl groups. The formation of a 2-bromoisobutyryl fragment (B*) layer on the surface was conducted by the reaction of a trichlorosilyl derivative

175

176

9 Surface-Grafted Hyperbranched Polymers Functionalized Si wafer

AB* initiator-monomer (inimer)

Br

Br O

O

O

O

A (a vinyl group)

A-b-a

A-B* A-B* A-B* A-B* A-B* A-B*

A*

A-B* A-B*

B* B*

B* B*

A* B* a

B*

b

B* B*

A*B*

B*

A*B*

A*B* A*B*

A*B* A*

B* B*

A*B*

kA

A*

A-b-A* A-B* A-B* A-B* A*

B* B* b

B* B* B*

A*B*

A-b-A*

B*

A-B*

A*B*

A*B*

A

kB

: R1 = H, R 2= H : R1 = Me, R 2 = Me

B* (a group capable of initiating)

B* B*

A-B*

B* B* B*

BPEA BIEM

Br

O

R1

Si O

O

O

O

B* B* B*

= Si O

R2

O

O

b

kB A-B*

B* B*

A* B*

A-B* A* B*

A* b b

A*B*

A*

B* B*

A*B*

A*B* A*B* B* B*

B* B*

Scheme 9.3 Schematic illustration of SCVP of an AB* inimer from a functionalized Si surface. Capital letters indicate vinyl group (A) and active centers (A*, B*), and lower-case letters (a, b) represent reacted ones.

with a silicon wafer. Because both the AB* inimer and the functionalized silicon wafer have groups capable of initiating the polymerization, the chain growth can be started from both B* in the initiators immobilized on the silicon substrate, and a B* group in the inimer. Further addition of AB* inimer or dimer to A* and B* centers results in hyperbranched polymers. Because the polymer formed in solution may also add to active centers of attached polymers, the method described here can be considered as a combination of “grafting from” and “grafting to” approaches. SCVP of AB* inimers with the functionalized silicon wafer was conducted under various conditions. CuBr/N,N,N¢,N†,N†-pentamethyldiethylenetriamine (PMDETA) was used as a catalyst system for the polymerization of the acrylic inimer, whereas bis(triphenylphosphine)nickel(II) bromide ((PPh3)2NiBr2) was employed for the methacrylic inimer. For example, when the polymerization of the acrylic AB* inimer, BPEA, was performed in bulk at a ratio [BPEA]0:[CuBr]0 = 125 at 30 NC for 2 h, a viscous polymer with apparent number-average molecular weight Mn = 4200 and polydispersity index Mw/Mn = 2.28 (as determined by GPC using linear polystyrene standards) was obtained. Because chain growth can be started from both the B* initiators immobilized on the silicon wafer and from B* groups in the inimers, both ungrafted and grafted polymers can be produced. The unattached polymer chains were separated from the covalently bound polymers by Soxhlet extraction in THF, and used to estimate the molecular weight, the molecular weight distribution,

9.4 “Grafting From” Approach

and the degree of branching. Characterization of soluble polymer by 1H NMR suggested the formation of highly branched polymers (degree of branching = 0.47). The mean film thickness of the polymer layer was 7.5 – 1.0 nm.

(a)

Copolymerization (AB* and M)

SCVP of AB* A-B*

A-B*

M A-B* A-B* M M M

B* B*

B* B* B*

M

M

A-B*

A-B*

Homopolymerization (M: tBuA, MMA)

Hyperbranched polymer

M

M B*

M MM

B* B* B* B* B*

B* B* B* B* B*

(Highly) branched polymer

Linear polymer

M

(b) SFM images

Film thickness = ca. 7.5 nm

ca. 23 nm

ca. 9.0 nm

(c) XPS spectra 4200

7000

Br

5500

Intensity

4000

1000

2600

3600

2400

3000 75

600

70

2200 75

65

200

0

1000

600

70

65

200

Binding Energy (eV)

Figure 9.2 (a) Synthetic routes; (b) SFM images; and (c) XPS results of hyperbranched, highly branched, and linear polymers grafted from functionalized silicon wafers. (Reproduced with permission from Ref. [68].)

75

0

1000

600

70

65

200

0

177

178

9 Surface-Grafted Hyperbranched Polymers

Surface-initiated SCVP of BPEA was found to yield polymer films with a high degree of branching, and with a characteristic surface topography. Tapping mode scanning force microscopy (SFM) and X-ray photoelectron spectroscopy (XPS) were used to investigate the surface topography and chemical composition of the grafted hyperbranched polymers. Typical results are shown in Figure 9.2. A characteristic surface topography on a nanoscale is clearly visible by SFM, which may be attributed to branched architectures. The size and density of the nanoscale protrusions obtained on the surface and the film thickness were observed to depend on the polymerization conditions, such as the ratio [BPEA]0:[catalyst]0. Similar results were obtained by SCVP of the methacrylate-type inimer (BIEM). In this way, we have been able to create novel surface architectures, in which characteristic nano-protrusions with different densities and sizes are composed of hyperbranched polymers tethered directly to the surface. The chain architecture and chemical structure could be modified by SCVCP, leading to a facile, one-pot synthesis of surface-grafted branched polymers. The copolymerization gave an intermediate surface topography and film thickness between the polymer protrusions obtained from SCVP of an AB* inimer and the polymer brushes obtained by ATRP of a conventional monomer (Figure 9.2). The difference in the Br content at the surface between hyperbranched, branched, and linear polymers was confirmed by XPS, suggesting the feasibility of controlling the surface chemical functionality. The principal result of this work is a demonstration of utility of the surface-initiated SCV(C)P via ATRP to prepare surface-grafted hyperbranched and branched polymers having characteristic architecture and topography. Another approach is based on chain-growth self-condensing, ring-opening polymerization of cyclic inimers from immobilized initiators. Recently, Khan and Huck [69] demonstrated a new procedure to synthesize covalently linked hyperbranched polyglycidol brushes on Si/SiO2 via anionic self-condensing ring-opening polymerization of glycidol. Optimization of the polymerization experiments by exploiting reinitiation cycles showed that the polyglycidol brushes produced can achieve ellipsometric thickness values of 70 nm. These authors also reported modification of hydroxyl end groups, which allow the polarity of the polymer brushes to be tailored. 13 C NMR spectroscopy analysis of cleaved polymer allowed the elucidation of the structure and degree of branching of the polymer. 9.4.2

Synthesis of 3D Hybrids by Surface-Initiated Self-Condensing Vinyl (Co)polymerization

Synthesis of hyperbranched polymer-silica hybrid nanoparticles was conducted by SCVP via ATRP from silica surfaces (Scheme 9.4) [70]. A typical condition is a bulk polymerization of BPEA with the functionalized silica particles in the presence of CuBr/PMDETA at a ratio of [BPEA]0:[CuBr]0:[PMDETA]0 = 125:1:1. Under these conditions, full conversion was reached after 2 h at 30 NC. The weight fraction of the surface-attached poly(BPEA) chains obtained after Soxhlet extraction was ca. 3.2 g polymer g–1 silica. The number of the inimer (BPEA) units per silica particle is about 33 000, which can be calculated from the weight fraction and the number of

9.4 “Grafting From” Approach

Cl3Si (

)3

O

Br

B* B* B*

B*

O

Br

O

Br CH3

B*

SiO2 B*

O

O

H

H

H

O

O

SiO2

End group

AB* inimer (BPEA)

Surface-active initiator (B*)

B* B*

SiO2

SCVP

(a)

SCVCP (AB* inimer + tBuA)

SiO2

C H2 CH C

Hydrolysis

n

O

O

SiO2

C H2 CH C

(b)

n

O

OH

Scheme 9.4 Synthetic routes for: (a) a silica particle with hyperbranched polymer shell; and (b) a branched polyelectrolyte shell.

the silica particles per gram (2.3 P 1017 g–1 silica). The value corresponds to the number of end-groups per silica particle, which may be distributed on the outermost surface of the hybrid as well as inside the polymer chains. The surface coverage calculated by the elemental analysis and specific surface area of the bare silica substrate (200 m2 g–1) was 10–21 mg m–2, depending upon the polymerization conditions. Calculation from the weight fraction shows that about 10–20% of BPEA polymerizes from the surface of the silica nanoparticles under the conditions used in this study. The successful preparation of hyperbranched polymer-silica hybrid nanoparticles was confirmed by FT-IR measurements (strong C=O vibration at 1740 cm–1 and broad Si-O stretching vibration around 1100 cm–1) and elemental analyses (carbon content = 28.0–32.9% in the products obtained by the bulk polymerization). The bulk polymerization of BPEA with the functionalized silica particles yielded well-defined hybrid nanoparticles comprised of silica cores and hyperbranched polymer shells having multifunctional bromoester end groups. Such surface multifunctionality is ideally independent of the surface curvature of the core particle and the layer thickness of the polymer shell, which could not be achieved by linear polymers. The soluble polymers obtained in the bulk polymerization at a ratio of [BPEA]0/ [CuBr]0 = 50 and 125 showed absolute molecular weights (Mn,GPC-VISCO = 23 800 and 30 800 using universal calibration), which are much higher than the corresponding

179

180

9 Surface-Grafted Hyperbranched Polymers

(b)

(a)

100 nm

200 nm

(c)

400 nm

(d)

Figure 9.3 Representative field emission-scanning electron microscopy (FE-SEM) images of the branched PtBuA-silica hybrid nanoparticles obtained by self-condensing vinyl copolymerization (SCVCP) of 2-(2-bromopropionyloxy)-

400 nm

ethyl acrylate (BPEA) and tert-butyl acrylate (tBuA) at c = [tBuA]0/[BPEA]0 = 6.1 (a, b) and =1.1 (c, d). (Reproduced with permission from Ref. [70].)

apparent values (Mn = 4600 and Mw/Mn = 2.9 and 2.3, as determined by GPC using linear standards). Note that the determination of the molecular weights of branched polymers is complicated because the hydrodynamic volume for a given molecular weight differs significantly from that of a linear sample. Accurate molecular weights can be obtained by the use of mass-sensitive on-line detectors such as a multi-angle light scattering photometer (MALS) [71,72] or a viscosity detector using the universal calibration (UNICAL) principle [73,74]. Higher molecular weights determined by GPC/viscosity using universal calibration than the corresponding apparent values are due to a lower hydrodynamic volume of the hyperbranched polymers, indicating the formation of compact architectures. Low values of the Mark-Houwink exponent (a = 0.36–0.38) showed undoubtedly a densely packed, 3D structure resulting from the hyperbranched topology. The exponent is typically in the region of 0.6–0.8 for linear homopolymers in a good solvent with a random coil conformation. The contraction factors [75], g = branched/linear, g¢ = [g]branched/[g]linear , are another way of expressing the compact structure of branched polymers. Further detailed characterization of hyperbranched polymers, such as the Mark-Houwink plots, log

9.4 “Grafting From” Approach

[g] versus log M, and contraction factor, g¢ = [g]branched/[g]linear , as a function of log M, can be found in a recent review [54] and an article [76]. The degree of branching (DB = 0.43–0.44) determined by 1H NMR indicates the achievement of the highly branched structures. The grafting density, molecular weights and degree of branching of the hyperbranched polymers could be manipulated by simply changing the polymerization conditions. SCVCP of BPEA and tert-butyl acrylate (tBuA) from the functionalized silica nanoparticles created branched PtBuA-silica nanoparticles. The functionality of the end-groups on the surface, and the chemical composition as well as the structure of the branched polymers grafted on the silica nanoparticles, could be controlled by composition in the feed during the SCVCP, as confirmed by elemental analysis and FT-IR measurement. Characterization using field emission-scanning electron microscopy (FE-SEM; Figure 9.3), transmission electron microscopy (TEM), scanning force microscopy (SFM), and dynamic light scattering (DLS) measurements indicated that the hybrid nanoparticles comprising the silica core and the hyperbranched polymer shell exist as isolated and aggregated forms. Novel hybrid nanoparticles with branched polyelectrolytes, poly(acrylic acid) (PAA)-silica, were obtained after hydrolysis of linear segments of the branched PtBuA, as can be seen in Scheme 9.4. The resulting hybrid material can be regarded as a characteristic branched polyelectrolyte [77]. These polymers, when grafted onto the nanoparticles, can be designed to have a fairly open structure, allowing the functional materials (e.g., metal ions) to penetrate the film more easily than in conventional linear polymer layers. SCVCP was also applied for the synthesis of branched PAAs having different molecular weights and degree of branching [76,78]. In order to understand the growth characteristics of the surface-initiated polymerization as well as the molecular parameters of the grafted polymers, it is necessary to cleave these chains from the surface at their points of attachment. However, attempts to achieve this were unsuccessful in the case of the hybrids prepared by the technique described here. For example, treatment of the hyperbranched polymer-silica hybrid particles in toluene suspension with aqueous HF (5%) and a phase-transfer catalyst, which has been used to detach polystyrene or PMMA grafted to a silica surface [79,80], was found to provide only low-molecular-weight products. This may be due to chain scission reactions on ester linkages of the poly(BPEA), which was also confirmed by NMR measurements of the soluble fraction and FT-IR measurement of the remaining insoluble part. Further experimental efforts are required to clarify the relationship between the molecular parameters (molecular weights, molecular weight distribution, and degree of branching) of hyperbranched polymers formed in solution and on the surface. 9.4.3

Theoretical Considerations

A series of theoretical studies of the SCV(C)P have been reported [81–87], which provide valuable information on the kinetics, the molecular weights, the molecular weight distribution (MWD), and the degree of branching (DB) of the polymers

181

182

9 Surface-Grafted Hyperbranched Polymers Molecular parameters of polymers obtained by self-condensing vinyl polymerization and copolymerizationa)

Table 9.2

SCVP of AB*

Copolymerization of AB* + M

Polymerization without initiator

DB » 1/2 [81,82] MW =Mn » 1 + DPn

DB » 2/(c + 1) (c >> 1) [85,86] MW =Mn = 1 + DPn /(c + 1)

Polymerization with multifunctional initiator (batch)

DB » 1/2 [83] 2 MW =Mn = 1 + DPn =f

DB » 2/(c + 1) (c >> 1) [87] MW =Mn = 1 + DPn /(c + 1) f 2

Polymerization with multifunctional initiator (semi-batch)b)

DB » 2/3 [84] MW =Mn » 1 + 1/f

DB » 2/(c + 1) (c >> 1) [87] MW =Mn = 1 + 1 / f

a) b)

DB: degree of branching, f: initiator functionality, c: [M]0/[AB*]0. Semi-batch = slow monomer addition.

obtained. The calculated MWD and DB of hyperbranched polymers obtained by SCVP and SCVCP under various conditions are summarized in Table 9.2. All calculations were conducted assuming an ideal case – that is, no cyclization, no excluded volume effects, and no side reactions. The calculated MWD of polymers formed in SCVP without initiators (conventional SCVP in bulk or solution) is broader than that obtained from SCVP in the presence of multifunctional initiators, B*f [83,84]. The presence of these initiators leads to a considerable narrowing of the polydispersity index, which decreases with increasing initiator functionality, f. Thus, the molecular weights and MWD of the ungrafted polymer obtained in solution might be different from those of the grafted polymer produced by a surface-initiated SCVP. On the other hand, the effect of the f-functional initiators on the DB was calculated to be negligible under batch conditions used here (i.e., inimers and initiators grafted on the surface are mixed instantaneously) [83]. This indicates that the DB does not depend on whether the polymer is formed in solution or on the surfaces. Therefore, it is reasonable to suppose that SCVP of the inimer with functionalized silica particles (or silicon wafers) provides surface-grafted poly(acrylate)s having a highly branched structure, even if the correlation of the molecular parameters of the soluble polymers with the polymers grafted on the surface is not confirmed experimentally. 9.4.4

Other Systems

In addition to SCV(C)P of AB* inimers, several one-step “grafting from” approaches have been recently reported for the synthesis of surface-grafted hyperbranched polymers (Table 9.1). One approach is the polycondensation of ABx monomers. For example, a one-step AB2-type polycondensation, which takes place on an insoluble solid support, was reported by Moore and colleagues [88,89]. These authors showed that hyperbranched polymers with low polydispersity and controlled molecular weights could be produced by this method. This is an approach toward a challenging

9.4 “Grafting From” Approach

goal in this field, which is the development of general polymerization methods to produce hyperbranched polymer with controlled DB and narrow MWD. By contrast, Kim et al. [90] reported the preparation of hyperbranched polymers on solid supports, such as Si wafers and fused SiO2, by using aziridines as cyclic monomers and an aminosilylated substrate as an initiator-modified surface. These authors indicated that the primary amine on the substrate is good enough to initiate the ring-opening polymerization of aziridine, leading to highly branched poly(ethyleneimine), as confirmed by measuring the thickness of the film and the absolute density of the primary amine functionality. In this case, branching is generated by a transfer reaction.

Summary

A variety of surface-grafted hyperbranched polymers have been synthesized using three different methods, namely “grafting to”, “grafting from”, and multi-step grafting approaches. A novel synthetic concept, surface-initiated self-condensing vinyl (co)polymerization was demonstrated, which is a first example of the one-step “grafting from” technique for the synthesis of the surface-grafted hyperbranched polymers. The development of various 2D and 3D hybrid materials with hyperbranched (meth)acrylate polymers could be achieved using this method. This methodology has a high feasibility to manipulate a variety of important parameters, such as grafting density, surface topography, branched architectures, functionality, molecular weights, simply by changing the polymerization conditions. This development substantially broadens and extends the scope of the surface-grafted hyperbranched polymers, which have unique properties and numerous possible practical applications. A well-controlled synthesis for these materials should lead to the creation of an entirely new category of materials which are controllable on the nanoscale, and have chemically sensitive interfaces.

Acknowledgments

The authors would like to thank the Deutsche Forschungsgemeinschaft (DFG) for financial support.

183

9 Surface-Grafted Hyperbranched Polymers

184

References 1 M. Jikei, M. Kakimoto, Prog. Polym. Sci. 2 3 4 5 6 7 8

9

10 11 12 13

14

15 16 17

18 19 20

21 22

2001, 26, 1233–1285. B. Voit, J. Polym. Sci. Part A: Polym. Chem. 2000, 38, 2505–2525. A. V. Ambade, A. Kumar, Prog. Polym. Sci. 2000, 25, 1141–1170. A. Sunder, J. Heinemann, H. Frey, Chem. Eur. J. 2000, 6, 2499–2506. B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677–710. J. Pyun, K. Matyjaszewski, Chem. Mater. 2001, 13, 3436–3448. S. S. Sheiko, M. MVller, Chem. Rev. 2001, 101, 4099–4123. K. L. Beers, S. G. Gaynor, K. Matyjaszewski, S. S. Sheiko, M. MVller, Macromolecules 1998, 31, 9413–9415. H. G. Boerner, K. Beers, K. Matyjaszewski, S. S. Sheiko, M. MVller, Macromolecules 2001, 34, 4375–4383. G. Cheng, A. BVker, G. Krausch, A. H. E. MWller, Macromolecules 2001, 34, 6883–6888. M. Zhang, T. Breiner, H. Mori, A. H. E. MWller, Polymer 2003, 44, 1449–1458. D. C. Tully, J. M. J. FrHchet, Chem. Commun. 2001, 1229–1239. A. Hierlemann, J. K. Campbell, L. A. Baker, R. M. Crooks, A. J. Ricco, J. Am. Chem. Soc. 1998, 120, 5323–5324. J. Li, L. T. Piehler, D. Qin, J. R. Baker, Jr., D. A. Tomalia, D. J. Meier, Langmuir 2000, 16, 5613–5616. G. Sui, M. Micic, Q. Huo, R. M. Leblanc, Langmuir 2000, 16, 7847–7851. M. E. Mackay, G. Carmezini, B. B. Sauer, W. Kampert, Langmuir 2001, 17, 1708–1712. D. Beyerlein, G. Belge, K.-J. Eichhorn, G. Gauglitz, K. Grundke, B. Voit, Macromol. Symp. 2001, 164, 117–131. H. Frey, Angew. Chem. Int. Ed. 1998, 37, 2193–2197. A. D. SchlWter, J. P. Rabe, Angew. Chem. Int. Ed. 2000, 39, 864–883. B. Karakaya, W. Claussen, K. Gessler, W. Saenger, A. D. SchlWter, J. Am. Chem. Soc. 1997, 119, 3296–3301. A. D. SchlWter, Topics in Current Chemistry 1998, 197, 165–191. D. C. Tully, A. R. Trimble, J. M. J. FrHchet, K. Wilder, C. F. Quate, Chem. Mater. 1999, 11, 2892–2898.

23 A. Sidorenko, X. W. Zhai, S. Peleshanko,

24

25

26 27

28

29 30

31 32 33

34

35

36

37

38

39 40

A. Greco, V. V. Shevchenko, V. V. Tsukruk, Langmuir 2001, 17, 5924–5931. C. B. Gorman, R. L. Miller, K.-Y. Chen, A. R. Bishop, R. T. Haasch, R. G. Nuzzo, Langmuir 1998, 14, 3312–3319. D. C. Tully, K. Wilder, J. M. J. FrHchet, A. R. Trimble, C. F. Quate, Adv. Mater. (Weinheim, Germany) 1999, 11, 314–318. M. Wells, R. M. Crooks, J. Am. Chem. Soc. 1996, 118, 3988–3989. H. Tokuhisa, M. Zhao, L. A. Baker, V. T. Phan, D. L. Dermody, M. E. Garcia, R. F. Peez, R. M. Crooks, T. M. Mayer, J. Am. Chem. Soc. 1998, 120, 4492–4501. W. M. Lackowski, J. K. Campbell, G. Edwards, V. Chechik, R. M. Crooks, Langmuir 1999, 15, 7632–7638. V. Chechik, R. M. Crooks, Langmuir 1999, 15, 6364–6369. R. Benters, C. M. Niemeyer, D. Drutschmann, D. Blohm, D. Woehrle, Nucleic Acids Res. 2002, 30, e10/11–e10/17. R. Benters, C. M. Niemeyer, D. Wohrle, ChemBioChem 2001, 2, 686–694. Z.-X. Guo, J. Yu, J. Mater. Chem. 2002, 12, 468–472. B. T. Mathews, A. E. Beezer, M. J. Snowden, M. J. Hardy, J. C. Mitchell, New J. Chem. 2001, 25, 807–818. M. Driffield, D. M. Goodall, A. S. Klute, D. K. Smith, K. Wilson, Langmuir 2002, 18, 8660–8665. Y. Zhou, M. L. Bruening, D. E. Bergbreiter, R. M. Crooks, M. Wells, J. Am. Chem. Soc. 1996, 118, 3773–3774. M. L. Bruening, Y. Zhou, G. Aguilar, R. Agee, D. E. Bergbreiter, R. M. Crooks, Langmuir 1997, 13, 770–778. J. G. Franchina, W. M. Lackowski, D. L. Dermody, R. M. Crooks, D. E. Bergbreiter, K. Sirkar, R. J. Russell, M. V. Pishko, Anal. Chem. 1999, 71, 3133–3139. M. Zhou, Y. Zhou, M. L. Bruening, D. E. Bergbreiter, R. M. Crooks, Langmuir 1997, 13, 1388–1391. D. L. Dermody, R. F. Peez, D. E. Bergbreiter, R. M. Crooks, Langmuir 1999, 15, 885–890. R. F. Peez, L. D. Dermody, J. G. Franchina, S. J. Jones, M. L. Bruening, D. E. Bergbreiter,

References

41 42 43

44

45 46 47 48

49

50

51 52 53 54 55

56 57

58

59

60

R. M. Crooks, Langmuir 1998, 14, 4232– 4237. W. M. Lackowski, P. Ghosh, R. M. Crooks, J. Am. Chem. Soc. 1999, 121, 1419–1420. A. Aoki, P. Ghosh, R. M. Crooks, Langmuir 1999, 15, 7418–7421. P. Ghosh, M. L. Amirpour, W. M. Lackowski, M. V. Pishko, R. M. Crooks, Angew. Chem. Int. Ed. 1999, 38, 1592–1595. D. E. Bergbreiter, G. Tao, J. G. Franchina, L. Sussman, Macromolecules 2001, 34, 3018– 3023. D. E. Bergbreiter, G. Tao, J. Polym. Sci. Part A: Polym. Chem. 2000, 38, 3944–3953. G. Tao, A. Gong, J. Lu, H.-J. Sue, D. E. Bergbreiter, Macromolecules 2001, 34, 7672–7679. M. Nagale, B. Y. Kim, M. L. Bruening, J. Am. Chem. Soc. 2000, 122, 11670–11678. N. Tsubokawa, H. Ichioka, T. Satoh, S. Hayashi, K. Fujiki, React. Func. Polym. 1998, 37, 75–82. K. Fujiki, M. Sakamoto, T. Sato, N. Tsubokawa, J. Macromol. Sci., Pure Appl. Chem. 2000, A37, 357–377. N. Tsubokawa, T. Satoh, M. Murota, S. Sato, H. Shimizu, Polym. Adv. Tech. 2001, 12, 596– 602. H. J. Lee, Y. Nakayama, T. Matsuda, Macromolecules 1999, 32, 6989–6995. Y. Nakayama, M. Sudo, K. Uchida, T. Matsuda, Langmuir 2002, 18, 2601–2606. H. Galina, J. B. Lechowicz, e-Polymers 2002, Paper No. 12. H. Mori, A. H. E. MWller, Topics in Current Chemistry 2003, 228, 1–37. J. M. J. FrHchet, M. Henmi, I. Gitsov, S. Aoshima, M. R. Leduc, R. B. Grubbs, Science (Washington, D. C.) 1995, 269, 1080– 1083. D. Baskaran, Macromol. Chem. Phys. 2001, 202, 1569–1575. P. F. W. Simon, W. Radke, A. H. E. MWller, Macromol. Rapid. Commun. 1997, 18, 865– 873. C. J. Hawker, J. M. J. FrHchet, R. B. Grubbs, J. Dao, J. Am. Chem. Soc. 1995, 117, 10763– 10764. M. W. Weimer, J. M. J. FrHchet, I. Gitsov, J. Polym. Sci. Part A: Polym. Chem. 1998, 36, 955–970. K. Matyjaszewski, S. G. Gaynor, A. Kulfan, M. Podwika, Macromolecules 1997, 30, 5192– 5194.

61 K. Matyjaszewski, S. G. Gaynor, A. H. E.

MWller, Macromolecules 1997, 30, 7034–7041. 62 K. Matyjaszewski, S. G. Gaynor, Macromole-

cules 1997, 30, 7042–7049. 63 A. Sunder, R. Hanselmann, H. Frey, R. MWl-

haupt, Macromolecules 1999, 32, 4240–4246. 64 J. M. J. FrHchet, S. Aoshima, 1996,

WO9614345, WO9614346. 65 S. G. Gaynor, S. Edelman, K. Matyjaszewski,

Macromolecules 1996, 29, 1079–1081. 66 P. F. W. Simon, A. H. E. MWller, Macromole-

cules 2001, 34, 6206–6213. 67 C. Paulo, J. E. Puskas, Macromolecules 2001,

34, 734–739. 68 H. Mori, A. BVker, G. Krausch, A. H. E. MWl-

ler, Macromolecules 2001, 34, 6871–6882. 69 M. Khan, W. T. S. Huck, Macromolecules

2003, 36, 5088–5093. 70 H. Mori, D. Chan Seng, M. Zhang, A. H. E.

MWller, Langmuir 2002, 18, 3682–3693. 71 P. J. Wyatt, Anal. Chim. Acta 1993, 272,

1–40. 72 P. J. Wyatt, J. Chromatogr. 1993, 648, 27–32. 73 H. BenoYt, Z. Grubisic, P. Rempp, D. Decker,

J. G. Zilliox, J. Chem. Phys. 1966, 63, 1507– 1514. 74 R. A. Sanayei, K. G. Suddaby, A. Rudin, Makromol. Chem. 1993, 194, 1953–1963. 75 W. Burchard, Adv. Polym. Sci. 1999, 143, 113–194. 76 H. Mori, D. Chan Seng, H. Lechner, M. Zhang, A. H. E. MWller, Macromolecules 2002, 35, 9270–9281. 77 J. Bohrisch, C. D. Eisenbach, W. Jaeger, H. Mori, A. H. E. MWller, M. Rehahn, C. Schaller, S. Traser, P. Wittmeyer, Adv. Polym. Sci. 2004, 165, 1–41. 78 H. Mori, A. H. E. MWller, Prog. Polym. Sci. 2003, 28, 1403–1439. 79 T. von Werne, T. E. Patten, J. Am. Chem. Soc. 1999, 121, 7409–7410. 80 T. von Werne, T. E. Patten, J. Am. Chem. Soc. 2001, 123, 7497–7505. 81 A. H. E. MWller, D. Yan, M. Wulkow, Macromolecules 1997, 30, 7015–7023. 82 D. Yan, A. H. E. MWller, K. Matyjaszewski, Macromolecules 1997, 30, 7024–7033. 83 W. Radke, G. I. Litvinenko, A. H. E. MWller, Macromolecules 1998, 31, 239–248. 84 D. Yan, Z. Zhou, A. H. E. MWller, Macromolecules 1999, 32, 245–250. 85 G. I. Litvinenko, P. F. W. Simon, A. H. E. MWller, Macromolecules 1999, 32, 2410–2419.

185

186

9 Surface-Grafted Hyperbranched Polymers 86 G. I. Litvinenko, P. W. F. Simon, A. H. E.

MWller, Macromolecules 2001, 34, 2418–2426. 87 G. I. Litvinenko, A. H. E. MWller, Macromolecules 2002, 35, 4577–4583. 88 P. Bharathi, J. S. Moore, J. Am. Chem. Soc. 1997, 119, 3391–3392.

89 P. Bharathi, J. S. Moore, Macromolecules

2000, 33, 3212–3218. 90 H. J. Kim, J. H. Moon, J. W. Park, J. Colloid

Interface Sci. 2000, 227, 247–249.

Part II

Characterization

189

10

The Analysis and Characterization of Polymer Brushes: From Flat Surfaces to Nanoparticles Rigoberto C. Advincula

Abstract

The study of polymer brushes has been of great interest, not only for their synthesis but also for their unique properties and characterization approaches. The analysis of these polymers is a challenge. Because of their reduced dimensionality, assumptions based on analogous solution or bulk systems do not necessarily hold. There is a desire to understand their fundamental properties at interfaces and in developing new applications. The interface is a highly anisotropic environment where a variety of conditions exists (e.g., surface energy, polarity, electrical double layer), not to mention the various geometries, size, shape, and surface properties of the solid-support substrate to which the polymer is bound. This review begins with a short description of the variety of polymer brush systems and solid-support substrates that have been investigated. It introduces the uniqueness of the polymer brush approach and the challenges on their characterization. The importance of using and interpreting the right analysis methods is emphasized. In general, this can be divided into spectroscopic, microscopic, and optical in approach, although other classifications can be mentioned. Both in-situ and ex-situ methods are possible. This chapter is not meant to be comprehensive, but is structured to provide a survey of the various options and possibilities in approaching the analyses of polymer brushes. While a number of analytical methods have focused on “grafted to” systems, an emphasis on this review will be on “grafted from” systems or those prepared by surface-initiated polymerization (SIP). Indeed, other than synthesis, a large part of the success in investigating the phenomena of polymer brushes is proving their dimensionality and that the physical and chemical properties of polymers are unique when tethered to surfaces.

190

10 The Analysis and Characterization of Polymer Brushes

10.1

Introduction 10.1.1

Polymer Brushes

The idea of confining long-chain macromolecules to surfaces has always intrigued polymer scientists. Theories and scaling parameters abound which predict their stretching behavior versus random walk configuration in solutions. An assembly of these polymer chains which are tethered by one end to a surface or an interface can be termed as “polymer brushes” [1]. This tethering is of sufficient density such that the polymer chains are crowded and forced to stretch away from the surface to avoid crowding. They are an essential model for investigating practical polymer systems such as the formation of polymer micelles; block copolymer equilibration at fluid– fluid interfaces, formation of microemulsions and vesicles, physically or chemically grafted polymers on a solid surface, and adsorbed diblock copolymers. In all of these systems, the conditions of the subphase (solvent), temperature, osmotic pressure, will affect the degree to which they reach equilibrium. Within the polymer system itself the presence of noncovalent interactions, grafting density, molecular weight, polydispersity, etc. are factors which also determine their stretch configuration. The deformation of these polymer chains is generally a contribution of both interaction and elastic free energies. The internal structure of polymer brushes has also been investigated by numerical and analytical self-consistent field (SCF) calculations, and by computer simulations [2]. Practically, polymer brushes are of interest for surface modification [3]: adhesives, biosurfaces [4], lubrication [5], separations, compatibilizers [6], and composite material preparation [7]. In the area of coatings, ultrathin and patterned organic films could be prepared which are useful in microelectrics, cell growth control, biomimetic material fabrication, microfluidics, and drug delivery. All the combinations that are possible for polymer brushes in terms of: composition, molecular weight, polydispersity, density, block sequencing, and microstructure are intriguing for each type of application. Functional polymers such as rigid rods, cross-linkable groups, side-chain functional groups, and polyelectrolyte behavior all have unique functions when localized to surfaces. The idea of modifying surfaces of inorganic solids with organic polymers can also be visualized as a type of “shell” or coating that gives the substrate the surface properties of a polymer. A number of model surface grafting techniques have been investigated on planar surfaces and particles. In general, this can be classified as either physical adsorption or “physisorption” and chemical adsorption or “chemisorption”. Physisorption involves the adsorption of functionalized homopolymers or block copolymers with “anchoring groups” from solution or melt [8]. The physical structure and density of such adsorbed layers are difficult to control both from entropic and enthalpic considerations. Nevertheless, these systems have been well investigated and have benefited from many innovative in-situ methods of surface analysis. By chemisorption, adhesion between polymer and the substrate may be greatly

10.1 Introduction

enhanced due to surface covalent attachment. This approach involves reaction of end-functionalized polymers with reactive sites on a substrate surface [9]. However, a disadvantage of this grafting method is that chain-ends have to first find their way to the surface and react. Like physisorption, a major impediment is that polymer chains already attached at the beginning of the reaction sterically shields the remaining surface reactive sites such that brush density becomes self-limiting [10]. While, chemisorption is related to the study of self-assembled monolayers (SAM), it is obvious that tethered polymers are unique systems with very different thermodynamic and kinetic considerations compared to small molecule amphiphiles. Thus, both physisorption and chemisorption have their drawbacks with a “grafting to” surface approach (Figure 10.1). In this respect, “grafting from” surfaces by surface- initiated polymerization (SIP) has important advantages in terms of polymer brush densities and other physical properties. SIP promotes polymerization of monomers directly from initiator sites already attached to surfaces, in which case activation of the initiator, interface properties and diffusion of monomers to the reactive sites are the primary factors [11]. Furthermore, it is possible to prepare grafted polymers where the average distance between grafting points is much smaller than the radius of gyration (Rg). This allows a linear scaling between the degree of polymerization and equilibrium thickness of a film. Thus, this SIP approach appears to be more promising and versatile for preparing tethered polymer brushes [12]. They can be prepared by a number of polymerization mechanisms including; free radical [11], cationic [13], ring-opening metathesis polymerization (ROMP) [14], atom transfer free-radical polymerization (ATRP) [15], polymerizations using 2,2,6,6-tetramethyl1-piperidyloxy (TEMPO) [16], reversible addition-fragmentation transfer (RAFT) polymerization [17], and anionic polymerization [18]. All of these methods are suit-

Surface Initiated Polymerization "Grafting from"

Chemisorption or Physisorption "Grafting onto"

General diagram of the polymer brush approach: “grafting to” versus “grafting from” approaches. The influence of grafting density and the enthalpic and entropic factors are clearly present in the “grafting onto” approach.

Figure 10.1

191

192

10 The Analysis and Characterization of Polymer Brushes

able for polymerizing different types of monomers on a variety of flat surfaces and particles. More, recently a number of innovative analytical tools both for particle and surface analysis have allowed elucidation of polymerization mechanism and physical properties of these tethered polymers. 10.1.2

SIP on Flat Surfaces and Particle Substrates

By focusing on SIP in flat ideal surfaces, it is possible to take advantage of a wide range of surface-sensitive spectroscopic and microscopic analytical techniques. Flat substrates that have been investigated include glass, quartz, Si wafer, Au-coated glass, and aluminum. There is potential for tailored polymer brush applications via surface modification and patterning [19]. Theoretical predictions have been used to calculate the molecular weight (MW) and polydispersity of polymer brushes grown by SIP on flat surfaces [20]. Other theoretical predictions have been made on the conformation and dynamic behavior of tethered polymers at flat surfaces [21]. It is always interesting to observe theoretical predictions on block copolymer brush behavior with respect to the Flory-Huggins (v) interaction parameter, Kuhn length, block volume fraction, and substrate surface energies [22]. Particle surface modification is an important procedure for improving the processing properties of colloidal dispersions and in the preparation of composite materials. This involves preventing flocculation, improving rheological properties, and adding thermodynamic stability to formulations. In the case of composite materials, the preparation of homogeneous compositions with thermodynamic stability, controlled phase separation, delamination and rheology, are important. Again, polymer brushes on particle surfaces can be approached both by grafting to and grafting from procedures. Polymerization from particle surfaces is challenging from the perspective of performing surface chemistry with colloids [23]. The particle size and geometry vary; likewise, the surface energy and solvent polarity changes, which can make a difference in terms of forming stable dispersions in selective solvents at each stage of the polymerization process. There are a number of analytical techniques which have been used to investigate this hybrid systems in situ. One of the compelling reasons to carry out SIP on particles is that it offers the advantage of preparing large quantities of SIP-grown polymers that can be degrafted and extensively analyzed ex situ. This is because of the high surface to volume ratio afforded by these substrates. However, it is also advantageous to characterize these systems with the polymers still grafted onto the particles for the sake of learning more about certain fundamental properties of colloids [23]. With the recent popularity of the SIP protocols, the controlled grafting densities poses unique parameters and conditions towards various polymerization mechanisms (initiation, propagation, termination, etc.) that warrants comparison with solution or bulk polymerization methods [24]. In terms of chemistries involved at the interface, the homogeneity or heterogeneity of the system and the differences between the bulk phase defines the lifetime of the initiator, the flux of the monomer,

10.2 Characterization of Ultrathin Polymer Films and Polymer Brushes

and the rate of termination. In other words, a detailed characterization involving surface-sensitive methods is very important. Again, fortunately, when dealing with flat surfaces, a variety of new and innovative techniques is available and have been applied to both physisorbed and chemisorbed polymer brushes. Direct surface characterization is even more challenging when dealing with particle surfaces. The interface is a highly anisotropic environment where a variety of conditions exists (e.g., surface energy, polarity, electrical double layer), not to mention the various geometries, size, shape, and surface properties of the solid-support substrate to which the polymer is bound. The behavior of polymer brushes under electric field, temperature gradient, solvent polarity, photochemical effects, and flow gradients can lead to novel types of field-responsive polymers. The importance of using the correct analysis methods and interpreting the results is emphasized. In general, this can be divided into spectroscopic, microscopic, and optical in approach. Both in-situ and ex-situ methods are possible. Other methods include: acoustic, mechanical, gravimetric, direct-surface force measurements, and electrochemical. This chapter is not meant to be comprehensive; rather, it is structured to provide a survey of the various options and possibilities in approaching the analyses of polymer brushes. While a number of the spectroscopic and microscopic methods have been reported for physisorbed polymer brushes, the contents of this chapter do not necessarily focus on these systems.

10.2

Characterization of Ultrathin Polymer Films and Polymer Brushes

The study of ultrathin organic and polymer films has been the subject of extensive investigations for the past few decades. They have important applications in microelectronics, electro-optics and biotechnology, and also facilitate the study of the fundamental nature of surfaces [25]. In most cases, they have finite thicknesses of a few to several hundred nanometers, on a variety of solid substrate-supported systems. The quasi two-dimensional ordering within layers can be extended into stacked structures perpendicular to the solid substrate. Patterning by both lithographic and nonlithographic techniques results in films with features ranging from a few nanometers (nanopatterning) to microns (micropatterning) [26]. Spin coating is the most common method for preparing ultrathin films, and is widely used in the microelectronics industry for photolithography. Other methods, based on “molecular or macromolecular assemblies”, have largely driven the forefront of organic ultrathin films research. The interest in highly-ordered ultrathin film fabrication methods such as alternate polyelectrolyte (APD) [27], Langmuir-Blodgett (LB) technique [28], and selfassembled monolayers (SAM) [29], have paved the way for a variety of analytical methods that can be used to probe the thickness, mesophases, microstructure, and optical properties of other ultrathin film systems (Figure 10.2). Thus, the same techniques on the various molecular and macromolecular assemblies mentioned above, can be applied to the study of polymer brushes on a solidsupport substrate. As mentioned, the characterization can primarily be divided into

193

194

10 The Analysis and Characterization of Polymer Brushes

spin-coating

self-assembled monolayer (SAM)

Langmuir-Blodgett-Kuhn (LBK)

alternate polyelectrolyte deposition (APD)

Different types of organic molecular and macromolecular assemblies on flat, ultrathin film substrates. These have finite thicknesses of a few to several hundred nanometers, on a variety of solid substrate-

Figure 10.2

vacuum deposition and OMBE

supported systems. The quasi two-dimensional ordering within layers can be extended into stacked structures perpendicular to the solid substrate.

surface-sensitive spectroscopic, microscopic and optical techniques (Table 10.1). This involves both in-situ and ex-situ techniques – that is, during and after grafting or after the whole film has been formed. As will be described, polymer brushes on flat surfaces offers unique properties that differ from those of other organic and polymer ultrathin films. 10.2.1

Spectroscopy and Optical Techniques

Spectroscopic methods such as ultra-violet-visible (UV-vis) absorbance, fluorescence, and other intensity-sensitive spectroscopic techniques can, in principle, be used to investigate change in optical constants, optical thickness, energy transfer procession, and monitor film growth in situ. UV-vis spectroscopy is a direct method for following the linearity of the polymer film build-up because of the ease in obtaining absorbance spectra either by transmission or reflection from absorbing species. Fluorescence measurements can be used where there are fluorophore probes present on a polymer brush, or for monitoring photochemical conversion and energytransfer properties. The formation of uniform polymer films can also be verified and observed by Xray reflectivity and other scattering methods [30]. X-ray diffraction can reveal Bragg peaks only on systems where there is high electron density contrast between layers or crystallinity. This can be used to determine the d-spacing for distinct layered systems. Neutron diffraction and reflectometry can be used, relying on deuterated spe-

10.2 Characterization of Ultrathin Polymer Films and Polymer Brushes

cies within layers [31]. The importance of these techniques is the fact that the mobility of polymer chains, diffusion within layers, and short-range and long-range order parameters can be correlated with other spectroscopic and microscopic techniques. In a kinematic approximation, the reflectivity R of a laterally homogeneous, surfaceassociated structure as function of the momentum transfer, Qz, is related to the scattering length density (SLD) – that is, the electron density or neutron SLD. The appearance of Kiessig fringes in X-ray and neutron reflectometry can be used to determine the thickness of the films and the relationship between the layers of the substrate, film, and subphase (air) [32]. Techniques for determining the chemical functional group and molecular (elemental) species also include FT-IR and X-ray photoelectron spectroscopy (XPS) [33]. XPS in particular is useful for monitoring the presence of different oxidized states of atoms, relative abundance of atomic species, and the presence of the substrate. FT-IR and IR-RAS has been used to monitor specific IR-sensitive functional groups, even at monolayer thicknesses [34]. This can be used to monitor attachment of polymers on surfaces starting from the initiator fixation stage by SAM. It also allows chemical group identification of polymers that have undergone chemical conversion. Several surface-sensitive FT-IR methods are available, from grazing incidence to polarization-modulated infra-red reflection absorption spectroscopy (PM-IRRAS) [18a]. Optical techniques can be used to monitor optical thickness and dielectric constant parameters. This includes ellipsometry, multiple reflection interferometry [35], and evanescent wave [36] and surface plasmon resonance spectroscopy (SPS) techniques [37]. Ellipsometry has been used widely and routinely to investigate film thickness of polymer brush films [38]. For example, null or spectroscopic ellipsometric data with in-situ solvent swelling experiments, involves exposure of solvent with constant evaporation rate and temperature, and has been used to estimate the MW of grafted polymer brushes [18b]. For the optical properties of films, it is important that the average film roughness and uniformity is specified. Often, sampling is localized by the “spot” size, such that it is necessary to probe and average different areas of a sample. 10.2.2

Microscopy

Microscopy is very useful, especially in monitoring lateral morphologies, layer roughness, domains, or patterns. This includes atomic force microscopy (AFM), scanning electron microscopy (SEM), optical microscopy, and other surface probe microscopy methods. Surface-imaging techniques, known collectively as scanning probe microscopy (SPM), has been widely developed since the invention of the scanning tunneling microscope (STM) [39]. Each technique relies on a scanning probe “tip” positioned within a few nanometers above the surface of interest. Using a sensory feedback with the probe signal to create a three-dimensional map of “surface height”, the probe is scanned horizontally along a series of parallel lines, recording the height at

195

196

10 The Analysis and Characterization of Polymer Brushes

each point along the line. This is done until a square region of the sample has been measured resulting in a two-dimensional point-by-point height or energy profile of the sample, or z (x,y), analogous to a topographic map. However, probe signals that have been used to sense surfaces can include electron tunneling current, interatomic forces, photons, capacitive coupling, electrostatic force, magnetic force, and frictional force. While STM relies on a very sharp tip to tunnel current between the probe and conducting surface, the AFM probe tip is normally integrated into a microfabricated, thin film cantilever. Changes in the cantilever vertical position is usually monitored with an optical lever scheme. SPM has been used to investigate polymer brushes in a number of ways. This includes: 1) General mapping of topology or investigating morphologies [40]; 2) identifying features resulting from phase segregations [41]; 3) measuring the surface forces involved in different brush geometries [42]; 4) estimating the MW of brushes [43]; and 5) investigating patterned polymer brushes [19b]. SEM can be used for relatively thick films combined with microtoming and cryogenic techniques [44]. It is very useful for obtaining three-dimensional morphologies and characterization of features of these films. 10.2.3

Other Methods

The characterization methods are not limited to those mentioned above. It is conceivable that other surface-sensitive methods will be specifically developed to probe these systems. Thus, other methods have been utilized, if not routinely: 1.

2.

3.

4.

5.

Recently, investigators have increasingly used quartz crystal microbalance (QCM) methods to investigate the deposition process, especially in situ [45]. The value of in-situ adsorption monitoring methods is valuable in determining the rate functions of the process. Contact angle measurements or surface tensiometry allows the determination of surface energy or surface tension which is dependent on the hydrophobicity or hydrophilicity of surfaces [46]. This also provides information on differences in morphology and functional group distribution on surfaces. Electrochemical methods involve using the redox behavior of probes or the polymers themselves, and it is possible to utilize this technique to investigate permselectivity [47]. This method can also be used to probe any redox active species in the polymer films. The use of streaming potential measurements with ellipsometry has also been reported, and can possibly be applied to polyelectrolyte polymer brushes [48]. A very important technique is the use of a surface force apparatus (SFA) [49]. This is an instrument which is capable of measuring the surface forces directly between two molecularly smooth surfaces, for example, mica in vapors or liquids with a sensitivity of a few millidynes (10 nN) and a distance resolution of about 0.1 nm. These flat, smooth surfaces of mica can be cov-

10.2 Characterization of Ultrathin Polymer Films and Polymer Brushes Table 10.1

Summary of surface-sensitive spectroscopic, microscopic and optical techniques.

Technique A. Spectroscopy FT-IR (transmission, grazing incidence, IR-RAS, PM-IRRAS, etc.

Function

Probe chemical functional groups sensitive to vibrational spectroscopy Monitor changes from each stage of polymer brush formation Investigate orientation XPS (X-ray photoelectron Probe atomic species; presence, distribution, spectroscopy or ESCA abundance, i.e., surface elemental analysis Probe polymer thickness and presence of substrate surface UV-vis absorbance or Absorbance increase as a function of concentration fluorescence energy transfer mechanisms and as probes X-ray and neutron – investigate order and presence of amorphous or scattering techniques crystallite species (diffraction and – thickness, refractive index reflectometry) – substrate to film and film to subphase interaction – short-range and long-range ordering – diffusion kinetics and chain dynamics B. Optical Ellipsometry Investigate thickness and dielectric constants in-situ swelling experiments Surface plasmon Investigate thickness and dielectric constants resonance in-situ experiments roughness, absorption, in-situ environment changes Interferometry In-situ experiments Polymer brush regimes C. Microscopy SEM Investigate three-dimensional structures and contrast of materials Direct imaging and patterning AFM Surface morphology, phase segregation, ordering, roughness domains, crystallization, patterning Determine surface forces Optical microscopy Optical image and can also use epi-fluorescence microscopy D. Other Methods QCM Mass change, in-situ adsorption and kinetics Electrochemistry Redox activity, diffusion of ions (ion mobility), permeability Contact angle Surface energy measurements, differences in morphology, functional group distribution Wetting behavior Surface force apparatus Direct surface force measurement Changes in grafting density BAM Direct observation of morphology (optical image) based on reflection applicable with imaging ellipsometry Streaming potential Measurement of “charges” on the polymer

References

18a,b, 33, 34

15b, 18a, 37, 17

9b, 18c, 30, 31, 32

38, 46, 48, 54, 55 18a,37

35

44

18a, 40–43, 54 19

45 47 12–18, 46

49 51 48

197

198

10 The Analysis and Characterization of Polymer Brushes

6.

ered to obtain the force between different polymer brush materials. The force is measured by the gap distance difference given by a piezoelectric device and that measured directly by interferometry (attractive forces make the micas closer and repulsive forces try to move away the micas). Direct measurement of surface forces is also possible by AFM [50]. Brewster angle microscopy (BAM) is a technique that is sensitive to the surface density and to the anisotropy of phase domains in films, where the reflectivity of a planar interface between two media depends on the polarization of the incident light and on the angle of incidence [51]. The reflectivity of a real interface at the Brewster angle for the mentioned polarization has three origins: (i) the thickness of the interface; (ii) the roughness of real interfaces; and (iii) the anisotropy of films.

Other methods will definitely be reported over time and find unique utility in the characterization and analysis of polymer brushes. It should also be mentioned that computational and simulation methods are important “tools” for predicting and scaling the observed behavior in polymer brushes, and should go hand-in-hand with experimental results [52]. 10.3

Investigating Polymer Brush Systems 10.3.1

Characterization of the Step-by-Step Procedure

The SIP grafting of polymers involves a step-by-step procedure which begins with the preparation of substrates and ends with characterization of post-polymerized films (Figure 10.3) [18a]. Usually, the chosen substrate is also dependent on the applicability of various analysis methods. For example, Si wafers can be investigated by using ellipsometry, interferometry, AFM, transmission FT-IR, and XPS. However, for SPS, specular reflection FT-IR – and even electrochemistry – it is ideal to utilize Au-coated glass. Primarily, the substrate used is determined by the type of SAM technique which is used to tether initiators onto surfaces. In principle, the characterization protocol is typical for investigating SAMs at surfaces, but is extended towards macromolecular dimensions once a polymer is attached. The polymerization and post-polymerization characterization is equally important. Few methods are available for in-situ characterization during the polymerization itself. It is difficult to monitor the growth of polymer brushes in real time, due to the fact that the mechanism and kinetics of are not necessarily the same as in solution or bulk. The last step – post-polymerization analysis – is a stage where the characterization of terminated brushes can be carried out after several “washing” methods to isolate grafted polymers on surfaces. Any post-polymerization treatments such as cross-linking or functional group conversion of the brush can be examined at this stage. Overall, the step-by-step analysis is useful for characterizing the polymer brush formation and comparing the films “before” and “after” each stage of the grafting or treatment.

10.3 Investigating Polymer Brush Systems CH2

Au

HS-(CH2)11O CH2 n-BuLi -S -(CH2)11O Bu

styrene Li

-S-(CH2)11O

benzene Bu n

isoprene

Li

benzene

-S-(CH2)11O

Bu

PS

n

m

Li

-S-(CH2)11O benzene

PS

PI

The surface-initiated polymerization (SIP) grafting of polymers involves a stepby-step procedure which begins with the preparation of substrates and ends with characterization of post-polymerized films. Shown in

Figure 10.3

this schematic diagram is the step-by-step SIP procedure in anionic polymerization, including block copolymer formation. Note that the termination step is not shown.

10.3.2

Investigating the Different Regimes of Polymer Brush Conformation on Surfaces

The configurational space of the polymer chains is limited by the presence of an interface in polymer brushes. The deformation of densely tethered polymer chains reflects a balance between interaction and elastic free energies. Dense tethering of polymer chains on an interface enforces a strong overlap among the undeformed coils, and increases the polymer-to-polymer contacts and the corresponding interaction energy. The polymer chains are forced to stretch away along the direction normal to the grafting sites. Stretching lowers the interaction energy per chain, at the price of a high elastic free energy. The interplay of these two terms determines the equilibrium thickness and brush regimes of the layer. The most important and distinctive characteristic of polymer brushes is that the equilibrium thickness varies

199

200

10 The Analysis and Characterization of Polymer Brushes

The two limiting cases of a polymer brush regime, showing: (a) tethered chains having the critical grafting density s* = 1/pRe2 and (b) chains having r > r* and forming the stretched brush. Other types

Figure 10.4

of brush regime terminologies have been reported in the literature. In this case, a characteristic length in the uncompressed brush (part b) is given by l = r–1/2. (From Ref. [49b].)

linearly with the degree of polymerization. This is very different from the behavior of the free polymer chains in a theta solvent, where polymer chains possess an unperturbed configuration. Thus, densely tethered polymer chains can be deformed and result in a variety of “brush regimes”. Various terms have been utilized to describe these regimes, including “mushroom”, “pancakes”, “micelle”, and “dimples”. Research studies employing theory, scaling theories, simulation, and surface probe techniques have shown that the different regimes can result from solvent swelling, differences in MW, differences in grafting density as the tethered polymer chains change their conformation (Figure 10.4) [53]. 10.3.3

Investigating Phase Segregation and Formation of Patterns

Multicomponent polymer brushes have also been extensively studied [54]. The SCF theory has been used to examine the equilibrium properties of a binary polymer brush composed of immiscible chains under melt conditions. For two homopolymers with sufficiently high immiscibility, different ordered phases can be described (equivalent to lateral microphase separation), where the composition varies as a statistical mixture of the two components or as blocks. Recent results have shown that a number of interesting morphologies and mesophases are possible [17b,54b]. For example, it has been observed that if two components were sufficiently immiscible, lateral binary microphase separation can occur over a wide range of solvent conditions. The onset of phase separation can be delayed as solvent quality increased. Under poor solvent conditions, interesting structural variations as a result of the combination of phase separation from solvent and phase separation of the two components can be observed (Figure 10.5) [55]. By changing chain architecture, grafting density, whole chain length, relative chain length, interaction energy between different blocks and interaction energies between blocks and solvents, a variety of novel well-ordered structures have been predicted and observed. Indeed, studies have now been extended to ternary polymer systems and different graft architectures. The theoretical results indicate that tethered copolymer brushes on a flat substrate are an excellent candidate for forming nanopatterned polymer films. By varying the sequence distribution of tethered linear AB copolymers, brushes composed of block

10.3 Investigating Polymer Brush Systems

Phase segregation and formation of patterns in multicomponent polymer brushes can be observed by atomic force microscopy (AFM). In this case, glassy and rubbery binary components of grafted polymethacrylate (PMA) and fluorinated

Figure 10.5

polystyrene copolymer (PSF) were sensitively imaged and differentiated between topography (left), phase imaging (right) and glassy (top) and rubbery (bottom) states of the binary brushes. Dimensions at 5 < 5 lm. (From Ref. [54b].)

copolymers can show distinct lateral inhomogeneities, with large domains of A and B units. These predictions can be easily described in the following manner [56]: The interaction parameter between polymer-polymer, vAB, vBA and polymer-solvent, vAS, vBS plays an important role. For brushes in which polymer chains are tethered by the less soluble block, the copolymer chains associates into distinct structures, where the less soluble component can form the inner core and the more soluble component forms the outer layer to shield the former from the unfavorable solvent. Changes in the composition (fraction) obviously will result in a continuum of different phase-separated structures. Also, the polymer brush density profile can provide a picture of local concentration gradients in these mixed systems. 10.3.4

Polymerization Mechanism

Because of the vast number of available polymerization mechanisms, what is sometimes overlooked is the uniqueness of each of the different polymerization mecha-

201

202

10 The Analysis and Characterization of Polymer Brushes

nisms at interfaces [11–18]. Very good characterization methods allow for the monitoring of differences in initiator/monomer composition ratios, time of polymerization, solvent and temperature conditions. Investigating the initiation, propagation, termination in addition to the type of polymerization is important [24]. This includes the effectiveness of initiation, kinetics of monomer diffusion, growing polymer brush density (initiation efficiency), propagating reactive center (e.g., their lifetime, reactivity, and stability), widening polydispersity, and lastly the rate and mode of termination. It is possible that in-situ monitoring probes will afford the greatest contribution in this area in terms of answering some of the most important fundamental questions underlying these differences. Also, the fact that other polymerization mechanisms involving step-by-step condensation, ring-opening polymerization, or even metathesis methods, are widely known to have complex kinetics and thermodynamic requirements in order for polymerization to take place. The assumption in

Figure 10.6 a) Strategy for patterning of a polymer brush using a sacrificial photoresist layer and lithographic imaging. (From Ref. [19a].)

10.3 Investigating Polymer Brush Systems

Figure 10.6 (b) Strategy for patterning of a polymer brush nonlithographic patterning. (A) the AFM image of a patterned brush of PMMA formed by combination of microcontact printing and ATRP. The bright areas of the image correspond to brushes of PMMA, while the dark regions correspond to patterned areas of SAMs formed from HDT. (B) Cross-sectional profile of the patterned PMMA brush shown in (A). (C) Optical image of a patterned brush of PMMA after immersion into aqueous KI/I2 for 60 s. (From Ref. [19b].)

all these measurements is that the systems investigated are related to the classical or fundamental behavior of these polymerization methods as observed in bulk, or in solution. Using those assumptions as a starting point, it is possible to probe the unique behavior at interfaces.

203

204

10 The Analysis and Characterization of Polymer Brushes

10.3.5

Patterning Using Nonlithographic Methods

New and simple strategies to fabricate surface-confined patterns with lithographic and nonlithographic methods have been widely reported (Figure 10.6) [19,57]. Nonlithographic methods using microcontact printing [58] and dip-pen nanolithography (DPN) [59] have recently been shown to be popular. Patterning offers a number of advantages, including: (1) extrinsic, triggered control of interfacial properties at the micrometer or nanometer scale; (2) precisely localized presentation of chemical or topographical features; and (3) controlled surface densities (e.g., as required to achieve higher throughput, as in combinatorial methods) [60]. Various schemes have been reported where microcontact printing and even lithographic methods [61] can be used to prepare microscale patterns. Fabricating nanopatterns by SIP is an interesting goal. One key strategy in the nanopatterning approach is the use of SAM level of initiator self-assembly. In combination with an appropriate nanolithographic technique, such as DPN, this provides the foundation for the fabrication of future polymeric nanostructures.

10.4

The Importance of Characterizing Particles and Nanoparticles

At present, the preparation of polymer brushes grafted onto particles and even nanoparticles is by far the most widely studied system in polymer grafting methods [62]. The ease of preparation and analysis of such systems by simple gravimetric methods is one of the most common reasons. In addition, particles are readily available and their dispersion properties have been widely studied in industry. A relatively large number of reports have been made involving a host of polymerization mechanisms. Again, a major advantage of direct polymerization on particle substrates is that the polymers can then be degrafted and analyzed as normal polymers isolated from solution. In this way it is possible to determine the MW, polydispersity, polymer microstructure, composition, directly, whereas at present reliance must be placed on indirect methods in the case of polymer brushes prepared from flat surfaces. In addition, there is wide interest in much smaller “nanoparticle” systems as host for the polymerization process [63]. The synthesis, characterization, and development of new nanoparticle materials have both scientific and technological significance. Primarily through the quantum size effect, a number of these nanoparticle hosts have interesting electro-optical and magnetic properties [64]. SIP from particle surfaces involves the growth of end-tethered polymer brushes where the length or thickness can be more than twice the radius of gyration (Rg) compared to a free polymer in solution (Figure 10.7). Different mechanisms are possible with a variety of initiators, reaction conditions and monomers, where a continuum of properties may be observed ongoing, from flat surfaces to high surface to volume nanoparticles. Important differences between solution and bulk polymerization can be observed, where the nanoparticles with grafted initiators behave essentially as macroinitiators.

10.5 Characterization and Analysis Methods for Polymer Brushes on Particles

initiation

- reactive functional surface - coating with stable surfaces - size / geometry

polymerization

- controlled SAM initiator formation / coverage - controlled activation and initiation

- polymerizability - end-group reactivity - aggregation / stability

General scheme for surface-initiated polymerization (SIP) on a spherical colloidal particle, showing the attachment of an initiator and the growth of the polymer by SIP.

Figure 10.7

In turn, the development of these hybrid materials will allow the preparation of thermodynamically and kinetically stable nanocomposites and colloids. A number of polymerization mechanisms for nanoparticles have been reported. This includes: freeradical [65], cationic [66], TEMPO [67], anionic [68], ATRP [69], and metathesis [70]. Through the careful use of surface-sensitive spectroscopic and microscopic techniques, much has been gained from the direct and in-situ analysis of grafted polymers on the nanoparticles with regards to the kinetics and mechanism of the polymerization process. Parallels can be drawn to SIP on flat surfaces where surface-sensitive spectroscopic and microscopic measurements are complementary to analysis methods for colloidal particles. Thus, this part of the chapter surveys the different characterization methods and procedures employed when observing the formation of the core-shell type of hybrid inorganic-organic polymer brush systems.

10.5

Characterization and Analysis Methods for Polymer Brushes on Particles

Polymer brushes grafted onto particles can be analyzed in situ. Direct or in-situ methods allow the monitoring of differences in initiator/monomer composition ratios, the time of polymerization, and solvent and temperature conditions. Classical methods for analyzing polymer brush-particle systems includes gravimetric [71], thermal analysis [72], light scattering [73], NMR [74], zeta potential [75], and rheology [76]. Microscopy methods such as TEM, SEM, and AFM can also be employed, thereby providing direct visualization of core-shell composite architectures [77]. An interesting method for investigating polymer-particle dynamics using fluorescence correlation spectroscopy has recently been reported [78]. Once polymers have been detached from the substrate, they can be analyzed like any normal polymer product.

205

206

10 The Analysis and Characterization of Polymer Brushes

10.5.1

In-Situ Investigations on Particles

There are several important reasons for in-situ analysis of the polymers grafted onto particles. These include: monitoring the grafting process of initiator to particle surface; monitoring surface functional group conversion; investigating the polymerization mechanism (initiation, propagation, termination); investigating the formation of core-shell particles; and investigating the properties of a hybrid organic-inorganic particle or nanocomposites. Examples of properties that can be measured include grafting density, polymer shell thickness, viscosity change, swelling and contraction of shell, and polymerization from different particle geometries, sizes, and shapes. The nature of the particle substrate is also important. The analysis of these materials can be further complicated by the fact that different particle and nanoparticle substrates have varying dispersion properties. The particles are governed, for example, by size, shape, size distribution, surface charge, and hydrophilicity. Each must be handled properly such that the material is neither lost nor precipitated with the different polymerization methods. For example, when performing LASIP grafting on particles, a modified reaction vessel was utilized in order to prevent the particles from being removed from the reaction vessel under high vacuum [68]. The methods of analysis for polymer brushes on particles are categorized as follows: .

.

.

Gravimetric methods [71,79]: these can be used to monitor the changes in weight before and after each grafting procedure. Each modification and polymerization step adds mass and changes the volume of the particles. By carefully weighing the addition or removal of mass, one can determine the amount of initiator or polymer added with each step. Key aspects of this procedure include removal of “unattached” material and accurate and precise weighing procedures. The calculations of weight gain and “yield” are straightforward. Thermal analysis [68,72,80]: it is possible to monitor the changes in thermal stability, enthalpy, heat capacity, for example simply by performing various thermal analysis methods such as differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) [81]. TGA is used routinely to monitor the composition of the hybrid material as well as the decomposition kinetics and mechanism. It is even possible to monitor the changes in the viscoelastic properties of a material using methods such as dynamic mechanical analysis, coupled with other rheological measurements [82]. As the polymer is grafted to a particle versus simply mixed (blends), the changes in the thermal behavior will be evident. Light scattering [73,83]: in a light-scattering experiment, a light beam is sent to a temperature-controlled sample solution, which scatters the light in all directions due to the thermal fluctuations. In one case of a specific angle, a detector receives the scattered light. In dynamic light scattering (DLS), which is a quasi-elastic light-scattering experiment, a laser light beam is sent through an inhomogeneous sample. The inhomogeneities are formed by colloidal par-

10.5 Characterization and Analysis Methods for Polymer Brushes on Particles

.

.

.

.

.

ticles, micelles, hydrodynamic modes (acoustic modes), etc., which scatter light in all directions almost without changing the light frequency. The scattered light is received in a detector at a particular direction. If the intensity of light is measured as a function of the scattered direction, this is referred to as a static light-scattering experiment. The different directions for measuring the scattered light are measured by the scattering angle h, which is the angle between the incident beam defined by the laser ray and the scattered light. This angle defines the scattering vector q, which is related to the change in direction of the vector k of scattered light with respect to the incident beam. In this case, light-scattering measurements can be utilized to probe the change in size with the addition of the polymer “shell” brush and probe the hydrodynamic volume properties. NMR [74]: For monitoring the appearance and disappearance of specific resonance peaks (e.g., average polymer microstructure). This technique is primarily important for monitoring the structure of bound polymers on the particle. The absence of substrate effects or the “invisibility” of the particles during the NMR experiment is relevant. Some recent experiments have emphasized the importance of substrate-polymer interactions [84]. Zeta potential [75,85]: These measurements are primarily for determining changes in surface charges, surface charge density, changes in the dispersion properties, surface energy, stability, etc. [86]. This is particularly important in the analysis of polyelectrolyte polymer brushes bound onto particles. Rheology [76]: This is especially important for monitoring changes in melt or solution viscoelastic behavior. Viscosity is the measure of the internal friction of a fluid, and this friction becomes apparent when a layer of fluid is made to move in relation to another layer. The greater the friction, the greater the amount of force required to cause this movement, which is called “shear”. Shearing occurs whenever the fluid is physically moved or distributed, as in pouring, spreading, spraying, or mixing. Highly viscous fluids, therefore, require more force to move than less viscous materials. The presence of a polymer brush on the particle surface not only stabilizes their dispersion but also dramatically changes the viscoelastic behavior. Microscopy [65–70]: The use of microscopy has been reported throughout literature for highlighting the formation of a polymer shell and aggregation properties. Depending on the surface area and geometry of the particle, intercalation and delamination may also be present, for example as in layered silicate clay particles. Nanoscale versus mesoscale dispersion properties are especially observed, and TEM imaging has been particularly effective. It is important to have very good contrast between the particle and the polymer component. AFM studies have also been used to show changes in particle dimension and morphology. It will be interesting to see in the future how scanning near-field optical microscopy (SNOM) techniques can be used to probe specific particle-polymer interactions [87]. Bulk analysis of composite materials: Lastly, these hybrid inorganic particleorganic polymer composites or nanocomposites can be probed just like any

207

208

10 The Analysis and Characterization of Polymer Brushes

bulk polymer system for their physical and mechanical properties. Some of the engineering polymer-related testing methods include: tensile modulus, flexural strength, scratch resistance, viscoelastic behavior, barrier properties, melt processing, dimensional stability with temperature, etc., where structure-property relationships can be developed. This is important in comparing the properties of these material systems with simple blend materials. 10.5.2

Degrafted Polymers from Particles

The biggest advantage of direct formation of polymer brushes on particle materials is that the polymers can be degrafted and analyzed as normal polymers isolated from solution. There are two main methods that can be used to detach polymers from surfaces: 1. 2.

Direct cleavage of the polymer brush from the grafting point. Dissolution or destruction of the particle substrate.

The first method requires a careful design of the initiator such that the polymer, once formed, can be removed from the surface by a simple and effective “cleavage” reaction which does not cause polymer-analogous reactions on the formed polymer, and the yield of which is 100%. Functional groups such as esters are easily removed by acid or base hydrolysis. This means that the initiator must be designed such that the ester group is readily hydrolyzed after the polymer has been formed [11]. Other reactions involve nucleophilic attack or exchange “metathesis” reactions such that the polymer is removed through a more facile reaction. A key parameter is the ability of these reagents to go to the “root” of the polymer. An alternative method is to use a core substrate that can be readily decomposed. This means that inorganic reagents should be able to destroy the substrate without harming the polymers. In both instances, centrifugation, partition, and filtration methods are used to isolate the desired degrafted polymer from the rest of the unwanted constituents and residues. The isolated polymers should be of sufficient purity to apply spectroscopic and macromolecular methods for analyses, and should then be readily dissolved in solution. Once the polymer is readily isolated, one can determine the MW, MW distribution, polymer microstructure, composition, thermal analysis, etc. as typically applied to “normal” polymers [24]. In this manner, the polymer properties are determined directly, whereas reliance must be placed on indirect methods in the case of polymer brushes prepared from flat surfaces. Methods such as size-exclusion chromatography (SEC), light scattering, NMR, viscometry, and DSC can be utilized. Essentially, the same analytical techniques can be utilized to polymers that have been degrafted from flat substrate surfaces – the only difference is that it will take a very large area of flat substrate!

10.5 Characterization and Analysis Methods for Polymer Brushes on Particles

Summary

This chapter has highlighted the importance of using different analytical methods, and the need for several options of optical, spectroscopic, and microscopic methods. While new techniques are constantly being developed and utilized, the correlation of results between these different methods helps to paint a complete picture of polymer brush phenomena at any substrate geometry. Different polymerization mechanisms on flat and particle surfaces have been demonstrated. Comparison can be made to their solution or bulk polymerization counterparts, but this requires careful evaluation of the conditions that are specific at interfaces. Highly surface-sensitive techniques are necessary for in-situ analysis of the grafted polymers on surfaces, especially with regard to understanding the kinetics and the detailed polymerization mechanism. For particles, the most effective analysis methods – especially with molecular weight parameters and macromolecular characteristics – are to degraft the polymers in a post-polymerization analysis. The importance of analyzing the polymers in situ is related to increasing our understanding of colloidal and interfacial phenomena, and unique applications of composites. It is possible that with an increasing understanding of polymer brushes on surfaces, the observed properties between SIP on flat surfaces and particles will appear to be a continuum.

Acknowledgments

The production of this chapter would not have been possible without the support and encouragement from my wife, children, family, friends, colleagues, and the hard-working Advincula Research Group. Special thanks also to my coorganizers Bill Brittain, Ken Caster, and JOrgen ROhe for making this symposium possible.

209

210

10 The Analysis and Characterization of Polymer Brushes

References 1 S. T. Milner Science 1991, 251, 905. 2 (a) P. G. de Gennes, Macromolecules 1980, 13,

1069; (b) R. Cantor, Macromolecules 1981, 14, 1186; (c) R. R. Netz, M. Schick, Macromolecules 1998, 31, 5105; (d) A. Semenov, S. H. Anastasiadis Macromolecules 2000, 33, 613. 3 A. Halperin, M. Tirrell, T. P. Lodge, Adv. Polym. Sci. 1992, 100, 31. 4 M. Amiji, K. Park, J. Biomater. Sci. Polym. Ed. 1993, 4, 234. 5 J. F. Joanny, Langmuir 1992, 8, 989. 6 T. Tadros, The Effect of Polymers on Dispersion Properties, London: Academic Press, 1982. 7 R. Krishnamoorti, R. Vaia, Polymer Nanocomposites, ACS Symposium Series 804, Oxford University Press, North Carolina, 2002. 8 G. J. Fleer, M. A. Cohen-Stuart, J. M. H. M. Scheutjens, T. Cosgrove, B. Vincent, Polymers at Interfaces, Chapman & Hall: London, 1993. 9 (a) R. Jordan, K. Graf, H. Riegler, K. K. Unger, Chem. Commun. 1996, 9, 1025; (b) W. Zhao, G. Krausch, M. H. Rafailovich, J. Sokolov, Macromolecules 1994, 27, 2933. 10 A. Balazs, Y. Lyatskaya, Macromolecules 1998, 31, 6676. 11 O. PrOcker, J. ROhe, Langmuir 1998, 14, 6893. 12 B. Zhao, W. Brittain, Prog. Polym. Sci. 2000, 25, 677. 13 R. Jordan, A. Ulman, J. Am. Chem. Soc. 1998, 120, 243. 14 M. Weck, J. Jackiw, R. Rossi, P. Weiss, R. Grubbs, J. Am. Chem. Soc. 1999, 121, 4088. 15 (a) M. Ejaz, S. Yamamoto, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules 1998, 31, 5934; (b) X. Kong, T. Kawai, J. Abe, T. Iyoda, Macromolecules 2001, 34, 1837; (c) D. Jones, A. Brown, W. Huck, Langmuir 2002, 18, 1265. 16 M. Husseman, E. Malmstrom, M. McNamara, M. Mate, D. Mecereyes, G. Genoit, J. Hedrick, P. Mansky, E. Huang, T. Russell, C. Hawker, Macromolecules 1999, 32, 1424. 17 (a) Y. Tsujii, M. Ejaz, K. Sato, A. Goto, T. Fukuda, Macromolecules 2001, 34, 8872; (b) M. Baum, W. Brittain, Macromolecules 2002, 35, 610. 18 R. Advincula, Q. Zhou, M. Park, S. Wang, J. Mays, G. Sakellariou, S. Pispas, N. Hadji-

christidis, Langmuir, 2002, 18, 8672; (b) R. Jordan, A. Ulman, J. F. Kang, M. H. Rafailovich, J. Sokolov, J. Am. Chem. Soc. 1999, 121, 1016; (c) R. P. Quirk, R. T. Mathers, T. Cregger, M. D. Foster, Macromolecules 2002, 35, 9964. 19 M. Husemann, M. Morrison, D. Benoit, J. Frommer, M. Mate, W. Hinsberg, J. Hedrick, C. Hawker, J. Am. Chem. Soc. 2000, 122, 1844; (b) R. R. Shah, D. Merreceyes, M. Husemann, I. Rees, N. Abbott, C. J. Hawker, J. L. Hedrick, Macromolecules 2000, 33, 597. 20 (a) S. T. Milner, T. A. Witten, M. E. Cates, Macromolecules 1988, 21, 2610; (b) S. Minko, G. Gafijchuk, A. Sidorenko, S. Voronov, Macromolecules 1999, 32, 4525. 21 S. Alexander, J. Phys. 1977, 38, 977. 22 (a) L. Rockford, S. Mochrie, T. Russell, Macromolecules 2001, 34, 1487; (b) T. Russell, T. Thurn-Albrecht, M. Tuominen, E. Huang, C. Hawker, Macromolecular Symp. 2000, 159, 77; (c) L. Rockford, Y. Liu, P. Mansky, T. Russell, M. Yoon, S. Mochrie, J. Phys. Rev. Lett. 1999, 82, 2602; (d) E. B. Zhulina, A. C. Balazs, Macromolecules 1996, 29, 6338. 23 A. W. Adamson, A. P. Gast, Physical Chemistry of Surfaces, Wiley-VCH, 1997. 24 G. Odian, Principles of Polymerization, 3rd Edition, Wiley-VCH, 1991. 25 C. W. Frank, Organic Thin Films: Structure and Applications ACS Symp. Series 695, American Chemical Society, 1998. 26 (a) C. G. Willson, Introduction to Microlithography, 2nd Edn. (Eds. L. F. Thompson, C. G. Willson, M. J. Bowden) ACS, Washington, D.C., 1994, p. 139; (b) J. L. Wilbur, A. Kumar, H. Niebuyck, E. Kim, G. Whitesides, Nanotechnology 1996, 7, 452; (c) S. Clark, M. Montague, P. Hammond, Supramol. Sci. 1997, 4, 141. 27 G. Decher, J. D. Hong, Makromol. Chem. Macromol. Symp. 1991, 46, 321. 28 G. Roberts, Langmuir-Blodgett Films. Plenum Press: New York, 1990. 29 (a) J. Sagiv, J. Am. Chem. Soc. 1980, 102, 92; (b) A. Ulman, Organic Thin Films and Surfaces, San Diego, Academic Press: 1991. 30 R. Mendelsohn, J. W. Brauner, A. Gericke, Annu. Rev. Phys. Chem. 1995, 46, 305.

References 31 R. Levy, N. Koneripalli, M. Tirrell,

S. K. Satija, Macromolecules 1998, 31, 3731. 32 E. P. K. Currie, M. Wagemaker, M. A. Cohen Stuart, A. A. van Well, Physica B 2000, 283, 17, 21. 33 X. Kong, T. Kawai, J. Abe, T. Iyoda, Macromolecules 2001, 34, 1837. 34 (a) M. Kawaguchi, M. Kawarabayashi, N. Nagata, T. Kato, A. Yoshioka, A. Takahashi, Macromolecules 1988, 21,1059. 35 M. R. Munch, A. P. Gast Macromolecules 1990, 23, 2313. 36 C. Allain, D. Ausserr, F. Rondelez, Phys. Rev. Lett. 1981, 49, 1694. 37 X. Fan, C. Xia, T. Fulghum, M. K. Park, J. Locklin, R. C. Advincula, Langmuir 2003, 19, 916. 38 (a) J. Habicht, M. Schmidt, J. Ruehe, D. Johannsmann, Langmuir 1999, 15, 2460. 39 S. N. Magonov, M. Whangbo, Surface Analysis with STM and AFM: Experimental and Theoretical Aspects of Image Analysis. VCH: New York, 1996; (b) Scanning Probe Microscopy and Spectroscopy. Theory, Techniques, and Applications. 2nd Edn. (Ed. D. W. Bonnell, University of Pennsylvania). Wiley-VCH: New York, 2001. 40 (a) H. Iwata, I. Hirata, Y. Ikada, Langmuir 1997, 13, 3063; (b) X. Kong, T. Kawai, J. Abe, T. Iyoda, Macromolecules 2001, 34, 1837. 41 S. G. Boyes, W. J. Brittain, X. Weng, S. Z. D. Cheng, Macromolecules 2002, 35, 4960. 42 S. Yamamoto, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 5995. 43 S. Al-Maawali, J. E. Bemis, B. B. Akhremitchev, R. Leecharoen, B. G. Janesko, G. C. Walker, J. Phys. Chem. B. 2001, 105, 3965. 44 A. Juang, O. A. Scherman, R. H. Grubbs, N. S. Lewis, Langmuir 2001, 17, 1321. 45 (a) M. Ivanchenko, H. Kobayashi, E. Kulik, N. Dobrova, Anal. Chim. Acta 1995, 314, 23; (b) K. A. Marx, Biomacromolecules 2003, 4, 1099. 46 D. Julthongpiput, Y. H. Lin, J. Teng, E. R. Zubarev, V. V. Tsukruk, Langmuir 2003, 19, 7832. 47 A. Anne, J. Moiroux, Macromolecules 1999, 32, 5829. 48 N. Houbenov, S. Minko, M. Stamm, Macromolecules 2003, 36, 5897. 49 (a) S. M. Kilbey, II, H. Watanabe, M. Tirrell, Macromolecules 2001, 34, 5249; (b) H. Wata-

nabe, S. M. Kilbey, II, M. Tirrell, Macromolecules 2000, 33, 9146. 50 T. W. Kelley, P. A. Schorr, K. D. Johnson, M. Tirrell, C. D. Frisbie, Macromolecules 1998, 31, 4297. 51 Y. Seo, K. Paeng, S. Park, Macromolecules 2001, 34, 8735. 52 (a) N. A. Spenley, Macromolecules 1998, 31, 4004; (b) Y. Lyatskaya, A. C. Balazs, Macromolecules 1997, 30, 7588. 53 T. Wu, K. Efimenko, P. Vlcek, V. Subr, J. Genzer, Macromolecules 2003, 36, 2448. 54 (a) I. Luzinov, D. Julthongpiput, V. V. Tsukruk, Macromolecules 2000, 33, 7629; (b) B. Zhao B, W. J. Brittain, J. Am. Chem. Soc. 1999, 121, 3557; (b) M. Lemieux, D. Usov, S. Minko, M. Stamm, H. Shulha, V. V. Tsukruk, Macromolecules 2003, 36, 7244. 55 A. Sidorenko, S. Minko, K. Schenk-Meuser, H. Duschner, M. Stamm, Langmuir 1999, 15, 8349. 56 (a) E. B. Zhulina, C. Singh, A. C. Balazs, Macromolecules 1996, 29, 6338; (b) E. B. Zhulina, A. C. Balazs, Macromolecules 1996, 29, 2667. 57 Y. Tsujii, M. Ejaz, S. Yamamoto, T. Fukuda, K. Shigeto, K. Mibu, T. Shinjo, Polymer 2002, 43, 3837. 58 Y. Xia, G. M. Whitesides, Angew. Chem. Int. Ed. Engl. 1998, 37, 550. 59 (a) S. Hong, C. Mirkin, Science 2000, 288, 1808; (b) S. Hong, C. Mirkin, Science 1999, 286, 523. 60 J. C. Meredith, A. P. Smith, A. Karim, E. J. Amis, Macromolecules 2000, 33, 9747. 61 G. Tovar, S. Paul, W. Knoll, O. Prucker, J. ROhe, Supramolec. Sci. 1995, 2, 89. 62 (a) N. Tsubokawa, Prog. Polym. Sci. 1992, 17, 417; (b) N. Tsubokawa, A. Kogure, Polym. J. 1993, 25, 83; (c) N. Tsubokawa, A. Kogure, J. Polym. Sci. Part A, Polym. Chem. 1991, 29, 697; (d) N. Tsubokawa, H. Ishida, Polym. J. 1992, 24, 809; (e) N. Tsubokawa, A. Kogure, K. Maruyama, Y. Sone, M. Shimomura, Polym. J. 1990, 22, 827. 63 R. Advincula, J. Dispers. Sci. Technol. 2003, 24, 343. 64 WTEC Proceedings, International Technology Institute, Baltimore, Maryland, 1998, pp. 1– 233. 65 O. Prucker, R. Ruhe, Macromolecules 1998, 31, 602.

211

212

10 The Analysis and Characterization of Polymer Brushes 66 R. Jordan, N. West, A. Ulman, Y.-M. Chou,

O. Nuyken, Macromolecules 2001, 34, 1606. 67 S. Yoshikawa, S. Machida, N. Tsubokawa, J. Polym. Sci. Poly. Chem. 1998, 36, 3165. 68 Q. Zhou, S. Wang, X. Fan, R. Advincula, J. Mays, Langmuir 2002, 18, 3324. 69 T. von Werne, T. E. Patten, J. Am. Chem. Soc. 1999, 121, 7409. 70 N. Y. Kim, N. L. Jeon, I. S. Choi, S. Takami, Y. Harada, K. R. Finnie, G. S. Girolami, R. G. Nuzzo, G. M. Whitesides, P. E. Laibinis, Macromolecules 2000, 33, 2793. 71 S. Hayashi, K. Fujiki, N. Tsubokawa, React. Funct. Polym. 2000, 46, 193. 72 M. Z. Rong, Q. L. Ji, M. Q. Zhang, K. Friedrich, Eur. Polym. J. 2002, 38, 1573. 73 T. von Werne, T. E. Patten, J. Am. Chem. Soc. 2001, 123, 7497–7505. 74 K. J. Watson, J. Zhu, S. T. Nguyen, C. A. Mirkin, J. Am. Chem. Soc. 1999, 121, 462. 75 A. De Sousa Delgado, M. Leonard, E. Dellacherie, Langmuir 2001, 17, 4386. 76 (a) A. N. Semenov, Langmuir, 1995, 11, 3560; (b) U. Raviv, R. Tadmor, J. Klein, J. Phys. Chem. B. 2001, 105, 8125. 77 (a) H. Mori, D. Chan Seng, M. Zhang, A. H. E. Mueller, Langmuir 2002, 18, 3682;

78 79 80 81 82

83 84

85

86 87

(b) K. Ohno, K. Koh, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 8989; (c) S. C. Farmer, T. E. Patten, Chem. Mater. 2001, 13, 3920. J. J. Zhao, S. C. Bae, F. Xie, S. Granick, Macromolecules 2001, 34, 3123. J. I. Amalvy, M. J. Percy, S. P. Armes, H. Wiese, Langmuir 2001, 17, 4770. E. Tadd, A. Zeno, M. Zubris, N. Dan, R. Tannenbaum, Macromolecules 2003, 36, 6497. K.-M. Kim, D.-K. Keum, Y. Chujo, Macromolecules 2003, 36, 867. J. Ren, B. F. Casanueva, C. A. Mitchell, R. Krishnamoorti, Macromolecules 2003, 36, 4188. W. L. Yu, M. Borkovec, J. Phys. Chem. B. 2002, 106, 13106. G. K. Agarwal, J. J. Titman, M. J. Percy, S. P. Armes, J. Phys. Chem. B. 2003, 107, 12497. M. J. Percy, V. Michailidou, S. P. Armes, C. Perruchot, J. F. Watts, S. J. Greaves, Langmuir 2003, 19, 2072. M. R. Gittings, D. A. Saville, Langmuir 2000, 16, 6416. R. C. Dunn, Chem. Rev. 1999, 99, 2891.

213

11

Characterization of Polymer Brushes on Nanoparticle Surfaces Thomas A. P. Seery, Mark Jordi, Rosette Guino, and Dale Huber

Abstract

Polymer brushes have been prepared through metal-mediated, surface-initiated polymerization (SIP) from gold and silica nanoparticles such that all parameters of the resulting nanocomposite material are under synthetic control. Gold and silica nanoparticles provide unique model systems for testing theories of polymer brush synthesis in confined geometries, and are also useful for screening chemical variations for SIP schemes. Specifically, titanium alkoxide-mediated polymerizations of polyhexylisocyanates were performed from gold nanoparticles that were stabilized with mixed layers of thiols containing terminal hydroxyls, and ruthenium alkylidenes were bound to norbornene functionalized Stober silica for the ring-opening metathesis polymerization of norbornene monomers. These two different nanoparticle/ polymer brush systems were analyzed for brush thickness, polymer molecular weight, and particle size. Detailed information on initiation efficiency and on the density of chain attachment points will ultimately lead to future refinements in preparation of polymer brushes using the “grafting from” or SIP approach.

11.1

Introduction

The recent interest in surface-grown polymer brushes can be seen in the large increase in the number of different polymers grown from an ever-increasing variety of surfaces [1–13]. Although most studies have been performed on planar surfaces, there is also significant interest in brushes grown from nanoparticle substrates. Patten and coworkers [14] have prepared polystyrenes using controlled radical methods from silica, CdS, and germanium spheres. While controlled radical methods are currently the predominant approach for surface-initiated polymerizations [15,16], metal-mediated polymerizations provide an alternate approach to preparing surfacebound polymer layers. There are examples of ring-opening metathesis polymerization (ROMP) being initiated from tethered alkylidenes on planar and particle surfaces [17–23]. The first realization of metal-mediated polymerizations on a nanopar-

214

11 Characterization of Polymer Brushes on Nanoparticle Surfaces

A) Nanoparticle substrates are synthesized; B) a stabilizing layer is formed on the surface and seeded with initiators; C) initiators are activated for polymerization; D) a nanocomposite results from polymer growth.

Scheme 11.1

ticle substrate was the preparation of polyhexylisocyanates from thiol-stabilized gold nanoparticles surfaces using tethered titanium alkoxides [11,24,25]. The bulk of this research is directed at attaining new materials with highly controlled compositions and the widest possible range of material properties. The diversity of polymerization mechanisms available for the “grafting from” or “surface-initiated” approach (shown in Scheme 11.1) allows access to a wide range of polymer brushes, and each approach has strengths as regards monomer selection, ease of synthesis, polydispersity, and attainable molecular weights as well as the chemical and physical properties of the target brushes. This wide range of polymerizations also comes with a spectrum of initiating moieties that must be coupled to the substrate for the polymerization to occur. The large surface area of nanoparticles provides an appealing laboratory for the production and study of polymer brushes grown from those surfaces [11,23–28]. Two logical starting points for detailed studies of polymer brush synthesis from nanoparticles are formation of coupling agent layers using thiols on gold and silanes on silica. Due to the wealth of pre-existing literature on the synthesis of nanoscale gold and silica particles, one might expect that the major difficulties in utilizing either system for this purpose had been overcome. Unfortunately, several tools appropriate to the characterization of polymer brushes on gold nanoparticles are ineffective for silica nanoparticles, and vice versa. The synthesis of polymer brushes from initiators tethered on nanoparticle surfaces provides a large number of parameters that can potentially be varied in the production of particle-based nanocomposites: substrate chemistry, particle size, tether length and spacing, polymerization chemistry, polymer composition and chain length. In addition, each of these may be polydisperse in nature so that exerting control over the structure of the final product may be a complex task. The need for a detailed understanding of polymer growth from surfaces is underlined by efforts in our laboratory [20,23,25,28] that have shown that the rate of initiation is sensitive to the structure of the tethered initiator layer when polymerizations are initiated from particle surfaces. The results of this range from broadened molecular

11.2 Experimental

weight distributions to no polymerization at all. The critical variables appear to be the surface density of initiator groups and the steric access at the point of initiation.

11.2

Experimental 11.2.1

Materials

All reagents and solvents were purchased from either Acros Organics or Aldrich Chemical Co., and used without further purification except as noted. Triethylamine (99%) was distilled over CaH2 and stored under N2. Ethyl vinyl ether (99%) and norbornylene (99%) were purified by vacuum transfer and stored under nitrogen at 4 IC. Tetrahydrofuran (THF) (99%) and methanol (absolute) were purchased from J. T. Baker, Inc. THF was freshly distilled over sodium before each use and stored under nitrogen. Silanes such as 5-(Bicycloheptenyl)-triethoxysilane (BCH) and trimethylethoxysilane (TMEOS) were purchased from Gelest Inc., with the exception of 5-norbornen-2-yl(ethyl)chlorodimethylsilane (NCS), which was from Aldrich Chemical Co. Bis(tricyclohexylphosphine) benzylidine ruthenium(IV) dichloride (Grubbs catalyst) was purchased from Strem Chemicals. Ethyl alcohol (dehydrated, 200 proof) was purchased from Pharmco Products Inc., and ammonium hydroxide (28–30 wt%) and hydrofluoric acid (48–50 wt%) were purchased from Fisher Scientific. Ethyl alcohol used in the synthesis of 5-norbornen-2-yl(ethyl)ethoxydimethylsilane (NCSEOS) was dried using Na and stored under nitrogen. Caution: the presence of a large amount of water in ethanol can be dangerous when Na is added. Titanium tetrachloride was distilled under a reduced pressure of nitrogen from copper turnings, and hexyl isocyanate was distilled from calcium hydride under reduced pressure and then degassed. All dry reagents were maintained water and oxygen free by storage under dry nitrogen after their purification. 11.2.2

Instrumentation

Proton NMR spectra were collected on a Bruker DMX 500 FT-NMR spectrometer using standard solution NMR techniques. Deuterated chloroform was used as the solvent, and shifts are referenced to the chloroform residual, defined as 7.27 ppm. Infrared spectrometry (IR) was conducted using a Nicolet-560 Magna-IR spectrometer to characterize the chemical composition of the sample before and after polymerization. Samples were analyzed on KBr disks or as KBr pellets in transmission mode. Nanoparticle samples were prepared for analysis by drying under vacuum after purification. Transmission electron micrographs (TEM) were obtained using a Philips 400 transmission electron microscope. Samples were prepared by deposition onto a car-

215

216

11 Characterization of Polymer Brushes on Nanoparticle Surfaces

bon-coated grid and conducted at several dilutions to insure the reproducibility of the sample morphology. Dynamic light scattering (DLS) measurements were performed with a Brookhaven Instruments BI9000 correlator and a BI200SM goniometer with a 514-nm laser source. Data were acquired at a 90I angle with a 500 mW light intensity. Solution concentrations ranged from ~30 mg mL–1 to 1 mg mL–1, the point at which further dilution did not yield a change in observed radius. Thermogravimetric analyses were obtained on a Perkin-Elmer TGA 7 using Pyris software and a Perkin-Elmer TAC 7/DX thermal analysis controller for data acquisition. Scans were conducted from 50 to 950 IC under nitrogen at a rate of 15 IC per minute. Gel-permeation chromatography (GPC) data were collected using a Waters 150-C ALC/GPC with Millenium software coupled with a PL-ELS 100 and a Waters 2487 Dual k absorbance detector. Samples were analyzed in THF at 35 IC on four 25 cm N 10 mm poly(divinylbenzene) columns (Jordi Assoc. FLP; pore size = 100–105 Q) and calibrations were based on 11 polystyrene standards (Polymer Standard Services = 106–2 000 000; Mw/Mn = 1.0–1.11). 11.2.3

Pyrolysis GC-MS

Samples were analyzed using an HP 6890 series Mass Selective Detector coupled to a HP 6890 series gas chromatograph in conjunction with a pyrolysis sampling device designed at the University of Connecticut. A pyrolysis temperature of 400 IC was used to decompose the samples. The injection port was heated to 300 IC and a temperature program from 35 to 325 IC at 15 IC/min was used to elute sample fragments. HP chem-station software was used for spectral analysis. 11.2.4

Infrared Monitoring of Polymer Formation

A quantity of gold nanoparticles synthesized as described below were freshly freezedried from benzene, placed in a vial, and sealed with a rubber septum. An inert atmosphere was established, and ca. 2 mL of titanium tetrachloride was vacuumtransferred into the vial. The titanium tetrachloride solution was stirred for 15 min, and then dried under vacuum. When the particles were fully dry (<10–2 torr vapor pressure), the vials were taken into a glove box with a dry, oxygen-free nitrogen atmosphere. The septum was removed and 3.0 mL of dry, degassed toluene was added and stirred for 10 min to fully dissolve the particles. An aliquot (1.0 mL) of dry, degassed hexyl isocyanate was added to a stirred solution of nanoparticles, and a timer started. The solution was stirred for 10 min, and then a glass and PTFE syringe was used to withdraw an aliquot of the solution and inject it into a liquid IR sample cell. The cell consisted of two sodium chloride windows, separated by a 15 lm-thick PTFE spacer, with a solution inlet and outlet. After the solution had been injected, the inlet and outlet were sealed with PTFE plugs, and the cell removed from the glove box and taken to the infrared spectrometer.

11.2 Experimental

The IR spectra were collected using a Nicolet Magna-IR 560 FT-IR spectrometer, with a sample shuttle. A macro language allowed automation of the sampling, and was programmed to record spectra at 90-s intervals. During the experiment, the sample shuttle translated the sample out of the beam-path to collect background spectra, and then back into the beam-path for sample spectra. With this arrangement, fresh background spectra were collected for each sample spectrum to account for background drift during the sampling. The sampling was continued uninterrupted for an extended period, typically 72 h. Following collection of the spectra, the Nicolet Omnic software package was used to automatically quantify the area and peak height of the isocyanate peak at 2275 cm–1, with the baseline defined as the line drawn from the values at 2000 to 2500 cm–1. 11.2.5

Synthesis of Alkanethiol-Stabilized Gold Nanoparticles

The gold nanoparticles used above were synthesized utilizing a modification of the two-phase synthesis published by Brust and coworkers [29]. An aqueous solution of hydrogen tetrachloroaurate (30 mL, 0.030 M) was stirred with a toluene solution containing tetraoctylammonium bromide (80 mL, 0.050 M), a phase-transfer agent. After several minutes complete transfer of the Au(III) to the toluene phase was indicated by the deep red color, and a lack of color in the aqueous phase. The alkanethiols, 1-dodecanethiol and 12-hydroxy-1-dodecanethiol (136 mg and 37 mg respectively, a 4:1 molar ratio) were dissolved in a minimum of toluene (ca. 5 mL) and added to the reaction. A freshly prepared aqueous solution of sodium borohydride (25 mL, 0.40 M) was added drop-wise with vigorous stirring. The toluene phase quickly darkened, signaling the reduction of the gold salt to Au(0). The two-phase reaction was stirred for 3 h, and the organic phase then decanted. Approximately 90% of the toluene was evaporated, and the product precipitated in methanol. The precipitate was collected on a nylon filter, redissolved in a minimum of toluene (~1 mL), and reprecipitated in methanol. The product was a dark, dense, waxy solid. The purified product was dissolved in 50 mL of benzene and freeze-dried to remove any remaining solvent. In order to remove the polymer chains from the gold particles, a toluene solution of polymer-coated gold particles (e.g., 100 mg in 5 mL toluene) was stirred with addition of a dilute (~0.1 M) ethanol solution of KCN. The ethanolic cyanide solution was added until the dissolution of the gold cores by the cyanide ions could be observed by the disappearance of the deep purple color that is characteristic of the gold nanoparticles. The toluene solution was extracted five to six times with water to remove all ionic species, after which the polymer was precipitated from the toluene solution in methanol. The cleaved polymer samples had identical IR and NMR spectra to homopolymer prepared using standard methods.

217

218

11 Characterization of Polymer Brushes on Nanoparticle Surfaces

11.2.6

Synthesis of Stober Silica Nanoparticles

Nanoparticles ranging in diameter from 15 to 30 nm were prepared using a method similar to that reported by Stober et al. [30]. As a representative example, absolute ethanol (1.0 L) and ammonium hydroxide (0.481 mol, 32.5 mL) were added to a 1.0L Erlenmeyer flask with vigorous stirring. TEOS (0.134 mol, 32.32 mL) was then added rapidly and the solution was allowed to stir for 12 h. Dynamic light scattering and TEM were used to obtain radii for these particles in the range of 15 to 100 nm, while IR and TGA confirmed the composition of these particles to be primarily silica with some (<3 wt%) residual alkoxide. 11.2.7

Synthesis of NCSEOS

Dry ethanol (20.0 mL), 5-norbornen-2-yl(ethyl)chlorodimethylsilane (5.0 mL), and triethylamine (3.25 mL) were added by syringe to an oven-dried Schlenk reaction flask equipped with a septum. The mixture was stirred for 12 h at 25 IC, during which time a white precipitate formed. The solution was then cooled in an ice-bath, filtered, and excess ethanol removed by distillation to give a brown liquid which was further purified by vacuum transfer at 30 mtorr. The purified product was a clear liquid (~4 mL). FT-IR m = 3059, 2966, 2868, 1448, 1389, 1338, 1250, 1167, 1109, 1082, 947, 831, 779, 717; GC-MS MI = 224. 11.2.8

Synthesis of BCH, NCSEOS, and TMEOS-Coated Nanoparticles

This procedure is an adaptation of that detailed by van Blaaderen et al. [31]. A representative example for fully coated particles is as follows. To a 250-mL, round-bottomed flask equipped with a condenser and a magnetic stirring bar was added 200 mL of the Stober silica nanoparticle solution (0.010 g mL–1) prepared as above, followed by BCH (1.77 N 10–2 mol, 4.36 mL). The solution was stirred for 12 h in a 70 IC constant temperature oil bath. The temperature was then increased to 110 IC and 140 mL of solution was removed using a distilling adapter. Purification of the resulting material was conducted by precipitation in hexane, followed by centrifugation and resuspension in THF. This process was repeated a total of five times, after which the product was re-characterized by DLS for constant size (Rh = 20 nm), IR m = 1091, 2850 demonstrated the presence of Si-O and C-H; PY-GC-MS had Mw = 66 for a cyclopentadiene fragment that is characteristic of norbornene fragmenting from the surface, a TGA shift in weight loss to higher temperature from 0– 150 IC for uncoated to 150–250 IC for coated, and TEM individual nonaggregated particles consistent with the DLS data. A representative example of a controlled initiator density coating is as follows. To a 100-mL, round-bottomed flask equipped with a condenser and a magnetic stirring bar was added 200 mL of the silica nanoparticle solution described above

11.2 Experimental

(c = 0.01 g mL–1) followed by TMEOS (2.67 N 10–3 mol, 42 lL). The solution was then stirred for 12 h in a 70 IC constant temperature oil bath. The temperature was then increased to 110 IC, and 40 mL of solvent was removed using a distilling adapter. The product was allowed to cool to 70 IC, after which NCSEOS was added (1.31 N 10–2 mol, 300 lL) with stirring and the mixture was allowed to stand for 12 h. A second distillation was then conducted at 110 IC, and 20 mL of solution was removed. The product was then cooled to 25 IC. Purification of the resulting material was conducted by precipitation in hexane, followed by centrifugation and resuspension in THF. This process was repeated a total of five times. The product was then characterized by DLS and showed a constant radius (20 nm), IR, m = 1091, 2850, demonstrated the presence of Si-O and C-H; PY-GC-MS had the cyclopentadiene marker, and TEM showed nonaggregated particles. 11.2.9

Synthesis of TMEOS Silica-Polymer Mixture

A 25-mL Schlenk flask was charged with a purified THF nanoparticle dispersion (0.151 g of nanoparticles in 4.0 mL THF). This solution was then evacuated and processed through three freeze-pump-thaw cycles before transfer to a glovebox. The solution was then transferred to a scintillation vial containing Grubbs catalyst (1.21 N 10–5 mol, 0.010 g) and stirred for 3 min. Norbornene was then added (1.2 N 10–3 mol, 0.1144 g), and the solution was allowed to stir for 1 h before the addition of ethyl vinyl ether (0.01 mol, 1 mL). The sample was then analyzed by DLS: diameter = 27.2 nm; TEM, GPC Mw = 20298, PDI = 1.09. 11.2.10

Synthesis of Silica-Poly(norbornene) Nanocomposites

A representative example is as follows. A 25-mL Schlenk flask was charged with a purified THF nanoparticle dispersion (0.302 g of nanoparticles in 8.0 mL THF). This solution was then evacuated and processed through three freeze-pump-thaw cycles before transfer to a glovebox. The solution was then transferred to a scintillation vial containing Grubbs catalyst (3.29 N 10–5 mol, 0.0271 g) and stirred for 3 min. Four precipitation/centrifugation steps were then conducted as previously described, using a sealed centrifugation vial and degassed solvents. The precipitate was then dispersed in degassed THF (22 mL), and this solution was transferred onto norbornylene (3.2 N 10–2 mol, 3.0 g). The solution was then separated into six portions (3.5 mL each) and ethyl vinyl ether (3.0 N 10–3 mol, 0.3 mL) was added to one vial at each of the following times: 2, 15, 30, 60, 90, and 120 min. The samples were then dried under vacuum and the resulting products weighed as 15.2, 49.9, 51.6, 64.2, 60.2, and 72.5 mg, respectively. The resulting material was characterized by TEM to confirm attachment of the polymer layer to the surface and to check particle aggregation state; by DLS to determine composite sizes of 60–350 nm, and by TGA to determine percentage organic as 30–70%. NMR showed the characteristic peaks for polynorbornene, 1H NMR (500 MHz, CDCl3) d5.35 (bs, 2H, olefin trans),

219

220

11 Characterization of Polymer Brushes on Nanoparticle Surfaces

5.19 (bd, 2H, olefin cis), 2.76 (bs, 2H, CH of cyclopentane cis), 2.40 (bs, 2H, CH of cyclopentane trans), 1.85 (bm, 1H, CH2 of cyclopentane), 1.75 (bs, 2H, CH2 of cyclopentane), 1.34 (bs, 2H, CH2 of cyclopentane), 1.04 (bm, 1H, CH2 of cyclopentane) ppm, and IR large increase in alkane region, double bond peak, and silica peak (2948, 2862 alkane), (964 double bond), and (1091 silica). 11.2.11

Isolation of Grafted Polymer Chains

To six 100-mL plastic beakers was added 15.2, 44.9, 46.6, 59.2, 55.5, and 67.5 mg of the dried composite sample for 2, 15, 30, 60, 90, and 120 min, respectively. This was followed by deionized water (20 mL) and hydrofluoric acid (20 mL) to dissolve the silica cores of the nanocomposites. These samples were stirred for 24 h. The white polymer chips were then removed with tweezers, placed on a nylon filter and washed with three portions of deionized water (25 mL) and two portions of methanol (15 mL). Drying under vacuum and weighing resulted in 7.5, 9.2, 7.3, 12.1, 25.7, and 24.0 mg of sample, respectively. The samples were then analyzed by FT-IR to confirm the absence of the silica peak (m = 1091) and presence of alkane peaks (m = 2948, 2861, 964) the same as pure poly(norbornene), TGA to confirm the absence of silica (100% weight loss), and GPC (to determine molecular weight; see Figure 11.7 for values). 11.2.12

Polymer Stability Test

In order to insure that poly(norbornene) does not degrade during the HF treatment used above, an independently synthesized portion of the pure poly(norbornene) (0.260 g) was subjected to the following treatment with HF. The sample was added to a plastic 100-mL beaker and covered with deionized water (20 mL) and hydrofluoric acid (20 mL). The sample was then stirred for 24 h before being placed on a nylon filter. The sample was then washed with three aliquots of deionized water (25 mL) and two aliquots of methanol (15 mL) before being dried under vacuum. The weight of the resulting product was 0.213 g. 1H NMR (500 MHz, CDCl3) d5.35 (bs, 2H, olefin trans), 5.19 (bd, 2H, olefin cis), 2.76 (bs, 2H, CH of cyclopentane cis), 2.40 (bs, 2H, CH of cyclopentane trans), 1.85 (bm, 1H, CH2 of cyclopentane), 1.75 (bs, 2H, CH2 of cyclopentane), 1.34 (bs, 2H, CH2 of cyclopentane), 1.04 (bm, 1H, CH2 of cyclopentane) ppm; FT-IR m = 2943, 2864, 1446, 964, 746 cm–1; GPC Mn = 11 N 103 g mol–1 before treatment and 12 N 103 g mol–1 after treatment, PDI changed from 1.33 to 1.46; TGA inflection point 475 IC, wt% = 98.8.

11.3 Results and Discussion

11.3

Results and Discussion

Our first system of interest is thiol-stabilized gold nanoparticles that have tethered hydroxyl ligands for initiating the polymerization of hexylisocyanate via a titanium alkoxide species. The gold nanoparticles were prepared using the methods of Brust and coworkers [29,32], with thiol stabilizing layers comprising hydroxyl and methyl terminated thiols. After the hydroxyls are transformed into titanium alkoxides, hexylisocyanate can be added and a polyhexylisocyanate-gold nanocomposite results. There are several parameters to be determined in this procedure: nanoparticle radius, hydroxyl concentration, polymer chain length, yield for the formation of titanium alkoxide species, initiation efficiency, and physical characteristics of the ultimate nanocomposite material. The radius of the initiator particles can be observed using TEM, but the dispersity of the particles should be assessed by techniques that sample a larger ensemble. Gold nanoparticle sizes have been studied previously using liquid chromatography techniques [33]. Unfortunately, the chemical stability of the gold thiol bond is not great and, in solution, the particles can aggregate over time in the absence of excess thiol. They are also highly colored, and thus there is the potential to damage separatory columns by increasing background signal intensity. The plasmon absorption of gold nanoparticles also makes them unstable to intense visible radiation, so that light-scattering studies are difficult. With these complications in mind, we turned to thermogravimetric analysis as a means to measure the total surface area in a sample of our thiol-stabilized nanoparticles. We made the assumptions that each gold thiol bond required the same surface area, and that all particles were spherical. Then, by calculating a suitable average molecular weight for mixed thiol layers the fraction of organic material observed using TGA could be uniquely related to a particle radius. The particles were quite stable over time when dried, and could be redissolved in common organic solvents for synthetic manipulations. The nanoparticles in the solution state could be handled much as any other organic reagent, and we were able to characterize the chemical nature of our substrates using standard methods of the organic chemist such as NMR and IR. Proton NMR was particularly valuable as we were able to determine the initiator concentration on the particle surfaces from the ratio of methyl protons to the methylenes on the hydroxyl termini. The interior methylene protons are increasingly conformationally restricted in their relaxation times as they are located closer to the particle surface and thus the integrals for these are not reliable. However, the terminal carbons have similar dynamics and thus have comparable integrals. This can be seen in Figure 11.1, where the NMR spectrum of a mixture of thiols is compared to the spectrum of a nanoparticle with a mixed thiol layer – the free thiols have sharper peaks. Once the particle size and initiator concentration were established, the particles were carried forward using the hydroxyl groups to initiate chain growth polymerization of hexylisocyanate via a titanium-mediated scheme [34]. The polymerization of the hexylisocyanate was followed using FT-IR in situ to continuously monitor the reaction progress. As seen in Figure 11.2, the isocyanate resonance near 2200 cm–1

221

222

11 Characterization of Polymer Brushes on Nanoparticle Surfaces

Upper spectrum: NMR of a 9:1 mixture of decane thiol and 11-hydroxyl undecane thiol. Lower spectrum: NMR of gold nanoparticle synthesized using the same mixture.

Figure 11.1

11.3 Results and Discussion

2.0

B

Absorbance

1.5 C

A

1.0

0.5

0.0 3500

3000

2500

2000

1500

1000

-1

Wavenumbers (cm ) Series of infrared spectra taken during a polymerization run of titanium alkoxide functional gold nanoparticles in a 3:1 solution of toluene:hexylisocyanate.

Figure 11.2

decreases steadily, while the amide peak at 1700 cm–1 increases. This technique is particularly effective for this polymerization because the relevant peaks appear in uncluttered regions of the spectrum. The aromatic C-H stretching of the toluene solvent above 3000 cm–1 provides an internal calibration. The only significant limita3.5

(Ao-Aeq)/(A-Aeq)

3 2.5 2 1.5 1 0.5 0

0.5

1

1.5 Time (h)

2

2.5

Plot of (Ao-Aeq)/(A-Aeq) for the absorbance at 2200 cm–1 during the polymerization of hexylisocyanate for various initiator concentrations [40 (d), 20 (j) and 10(~) mg mL–1] of gold nanoparticles. Figure 11.3

3

223

11 Characterization of Polymer Brushes on Nanoparticle Surfaces

tion is due to the equilibrium nature of the polymerization. This leads to the requirement for a 25% monomer solution to observe substantial polymerization prior to attaining the equilibrium monomer concentration of 10%. As seen in Figure 11.3, the polymerization is second order in monomer, with a 90-min induction period; this was in contrast to the homogeneous polymerization that showed the expected first-order behavior [34,35]. The induction period is likely due to a slight insolubility of the initiating species, and the second-order kinetics to a combination of slow initiation and fast propagation. Neither of these difficulties is fatal to the polymerization as the observed second-order rate constant shows a firstorder dependence on initiator concentration. Preparing four samples in parallel where the polymerization was terminated at 20, 30, 40 and 50% extents of reaction tested this mechanistic hypothesis. These nanocomposites were characterized by TGA and light scattering, after which the gold cores were dissolved using ethanolic solutions of KCN. The polymer in each case was recovered and molecular weight distributions obtained using GPC. The polystyrene equivalent molecular weights have previously been shown to be a factor of 4 larger than the molecular weights obtained using light scattering. This is a result of the rod-like nature of the polyisocyanate molecules. The molecular weights at all extents of reaction were found to be 1.5 € 0.2 N 105 g mol–1. There was a slight decrease in molecular weight with increasing extent of reaction, supporting the contention that the number of active chain ends was increasing as the equilibrium monomer concentration was being approached. The polydispersities were initially ~4–5, but at the highest extents of reaction they approached a value of 2. This is the expected value for a most probable distribution that should arise from a polymerization at equilibrium. We combine the molecular weight data with the TGA results for the nanocomposites and calculate the number of chains per particle to find that it increases linearly with conversion from 11 chains at 20% to 28 chains per particle at 50% monomer conversion, as shown in Figure 11.4. The average number of initiator sites per particle is calculated to be ~60 from the NMR data and particle size measurements. This slow increase in the number of chains per particle at constant 30

Chains per Particle

224

25 20 15 10 5 0 0

0.1

0.2 0.3 0.4 Monomer Conversion

0.5

0.6

Plot of chains per particle against conversion of monomer. Chains per particle derives from TGA mass fraction, particle diameter and molecular weight of polymer arms.

Figure 11.4

11.3 Results and Discussion

weight average molecular weight further supports the hypothesis of a slow initiation and rapid propagation. The homogeneous system contrasts with this, as the trifluoro alkoxy initiator provides high molecular weights and narrow polydispersities at low conversion [34,35]. This is an indicator of fast initiation. The difference between the homogeneous polymerization and the nanoparticle-based polymerization is the steric accessibility of the surface-bound initiator. Efforts are underway to improve the access of monomers to the tethered metal species to test this hypothesis. Gold nanoparticles are a fine substrate for some investigations of surface-initiated polymerizations, but at temperatures above 100 IC the thiol layer is lost and a gold mirror forms on the inner surface of the reaction vessel. Therefore, silica particles were prepared because silane coupling agents are known to be substantially more thermally stable than thiols, so that nanocomposites with better high-temperature capabilities could be accessed. Silica is also a desirable substrate because it has many properties that contrast with those of gold (e.g., insulating, transparent, and cheap). Initial efforts to produce results in parallel with our prior study of gold particles were unsuccessful. The coupling agent coatings were generally hydrolytically unstable when prepared using published methods, and required additional efforts to fully condense them onto the particle surfaces. The addition of a distillation step to the particle preparation allowed us to make stable coatings on particles with radii of 20 nm that could be precipitated and resuspended multiple times in a stepwise procedure for functionalizing the coated surface. At each stage in the synthesis the particle or nanocomposite diameter should be measured to determine how much surface area is available, and also to see whether a particular chemical manipulation has caused the particles to aggregate. For the gold system, our standard method was TGA, though the Stober nanoparticle synthesis uses tetraalkoxysilanes as the silicon source and, under standard conditions, some of the alkoxy groups remain in the interior of the particles. These alkoxy groups oxidize very slowly during TGA experiments, and contribute to a sloping baseline that degrades quantitation of the weight fraction of a surface coverage of coupling agent. Furthermore, the literature on silica particles indicates that the surface silanols do not react completely with coupling agents, even under the most rigorous conditions, so that there was no expected value of the surface area required for each coupling agent. Because of these facts, the mass of coupling agent was not simply related to the surface area, as was the case for thiol-stabilized gold particles. Thus, the inexpensive method for particle size determination that we relied on for the gold nanoparticle system could not be adapted to the silica particles. However, the organic fraction (i.e., the polymer mass) in the nanocomposites could be determined using TGA as the polymer represented a much larger mass fraction in this case. Fortunately, silica particle diameters are easily obtained using dynamic light scattering or TEM, and this provides both the real space image for shape determination as well as the determination of ensemble averages. These particles are generally too large for GPC so that distribution data are not readily accessible. Data processing algorithms for microscopy studies allow the collection of distribution data on sam-

225

11 Characterization of Polymer Brushes on Nanoparticle Surfaces

ples containing large numbers of particles, but this method does not yet have the convenience of the light-scattering experiments. One critical element in these investigations was to establish the initiator concentration, as was established in the gold nanoparticle study. Unfortunately, we were unable to use NMR in the same fashion on silica spheres, as particles with larger radii diffuse at a rate comparable to the Larmor frequency so that solution NMR spectra are significantly broadened. The silica spheres produced by the Stober method are often more than 100 nm in diameter, which places them squarely in the regime where NMR is broadened. Thus, a different tool for establishing initiator densities must be developed in order to study the whole range of nanoparticle sizes. A few reports [15,16,36,37] of polymerizations from smaller spheres were encouraging, but 20 nm is near the lower end of the range. This is significantly larger than the diameter of the gold spheres (~2 nm) however, and has important implications for further analyses. The surface area to volume ratio varies as 1/radius, so that less surface area is available per unit volume. This reduces the sensitivity of most spectroscopic assays while introducing complications due to increasing viscosities at higher volume fractions, even though the difference in density between gold and silica provides a similar surface area per unit mass when comparing 2-nm gold particles to 20-nm silica particles. Infrared spectroscopy provided a measure of the chemical composition, with insufficient resolution for a quantitative determination of initiator density. Our first polymerizations from the silica particles utilized a norbornene coupling agent to tether a ruthenium alkylidene to the surface that would be used for ring-opening

1.2 1.0 0.8

Abs

226

0.6

Cleaved Polymer

0.4 Nanocomposite

0.2 Silica nanoparticle

0.0 4000

3500

3000

2500

2000

1500

-1

Wavenumber, cm

Infrared spectra of: silica nanoparticles coated with norbornene coupling agents before polymerization; silica/polynorbornene nanocomposites; and polynorbornene that has been cleaved from the nanoparticles with HF treatment.

Figure 11.5

1000

11.3 Results and Discussion

metathesis polymerization of norbornene monomer. There is some intensity that may correspond to the olefinic C-H stretch (at m ‡ 3000 cm–1) of the coupling agent and of the polymer in the spectra shown in Figure 11.5, but the main importance of those data was to show the composition of the nanocomposite. These data showed that the samples after polymerization contained silica and poly(norbornene), and that the two large peaks corresponding to silica and hydrocarbon could be used to establish composition in conjunction with TGA of the nanocomposite materials. Because we could not obtain initiator concentrations from IR or NMR, we turned to mass spectroscopy for its potential greater sensitivity. In order to separate the initiator ligands from the silica surface, the particles are initially pyrolyzed and the outgassing is collected on a cold finger that is subsequently warmed and the gasses injected into an otherwise standard GC-MS set-up. Using this method, we have been able to observe unambiguously that our initiator species is only removed from the surface under conditions which indicate that it is covalently bound. This pyrolysis allows us to see that the initiator is present on the surface and to distinguish physisorbed species from chemisorbed species. This distinction is important, as particles can be stabilized with coupling agents that are simply hydrogen-bonded to the surface. Because the role of the surface in the pyrolysis step is unknown, we cannot construct a calibration curve for this system. However, by using GC-MS we can quantitatively assay the coupling agent that does not react with the surface, and set an upper bound on the amount of coupling agent that could have reacted. Measurements in triplicate for several different samples of particles consistently showed that between 79% and 86% of the coupling agent could be recovered from the coating reaction. The remaining 14–21% would constitute a coupling agent coating density of 0.3 to 0.7 coupling agents per nm2, this value being well within the reported range of 0.1 to 1.1. This indirect method is complemented by elemental analysis of particles before and after the coupling agents are applied. The elemental analysis provides values of the surface coverage that correspond to those obtained from the GC-MS study of the coupling agent washes. From these data, we were able to calculate a surface coverage of ~2000 coupling agents per 20 nm diameter particle. The treatment of the coupling agent-coated particles with Grubbs catalyst followed by multiple washings and introduction of norbornene monomer provided silica poly(norbornene) nanocomposites. Published reports of chain transfer via backbiting reactions led us to test whether the polymer was bound to the silica particle surfaces. Mixtures of trimethylsilyl-coated particles and polynorbornene are seen to phase separate in TEM micrographs when cast from ethanol or hexane. Micrographs of nanocomposites cast from the same solvents are shown in Figure 11.6, where particles are clearly dispersed in the polymer matrix. The polymer chains can be quantitatively cleaved from the silica particles using a mild HF treatment, so that the mass fraction of polymer recovered corresponds to the polymer mass fraction seen by TGA. GPC has also been utilized to determine the hydrodynamic size of nanocomposites and their initiator spheres as well as of the polymer arms after they have been cleaved from the surface. The polymerization from the silica particles can be moni-

227

11 Characterization of Polymer Brushes on Nanoparticle Surfaces

TEM micrographs of silica/ polynorbornene nanocomposites cast from ethanol (upper) and hexane (lower) solutions.

Figure 11.6

300

200

250

200

100

150

50

100

3

150

0

0

20

40

60 80 Time (min)

100

Left axis plot: molecular weight of growing polymer chains from silica/polynorbornene nanocomposites against reaction time. Right axis plot: nanocomposite diameter by dynamic light scattering for the same samples.

Figure 11.7

120

50

Diameter (nm)

-1

250

Mw × 10 g mol

228

11.3 Results and Discussion

tored in this fashion by quenching the ring-opening metathesis polymerization using ethyl vinyl ether and cleaving the chains from the surface. A comparison of the increase in chain molecular weight and nanocomposite diameter is seen in Figure 11.7. The composite diameter measured using dynamic light scattering increases steadily throughout the 2-h experiment, while the chain length reaches a maximum after 60 min. In this case, we again attributed the continued increase in composite size to an increase in the number of chains on the particle. The decrease in arm molecular weight was most likely due to the onset of chain transfer reactions. Future efforts are directed at controlling the chain transfer and varying the number of chains per particle in this system.

Summary

Polymer brushes can be prepared through polymerization from nanoparticle surfaces in a well-defined manner such that all parameters of the resulting nanocomposite material are under synthetic control. These platforms provide unique model systems for testing theories of polymer brushes in confined geometries, and are also useful for screening chemical variations for surface-initiated polymerization schemes. This method for preparing nanoparticle composite materials may also lead to new nanocomposite formulations. The method of analyzing polymer nanoparticle composite materials necessarily varies from system to system, depending on the physical and chemical nature of the substrates. The partially inorganic nature of these materials complicates some analyses while simplifying others. The desirable properties of a particular substrate (e.g., conductivity in the case of gold) may provide undesirable obstacles to analysis – for example, light absorption that complicates scattering experiments. Nonetheless, high-quality data can be obtained to characterize these materials at each step in their production. Detailed information on initiation efficiency and on the density of chain attachment points will ultimately lead to future refinements in preparation of polymer brushes using either the “grafting from” or surface-initiated polymerization approach.

Acknowledgments

The authors would like to acknowledge National Science Foundation Grant DMR9876244 for financial support of these studies.

229

230

11 Characterization of Polymer Brushes on Nanoparticle Surfaces

References 1 M. Biesalski, J. Ruehe, Macromolecules 2003,

36, 1222–1227. 2 J. Genzer, T. Wu, K. Efimenko, Mater. Res. Soc. Symp. Proc. 2002, 707, 9–14. 3 S. Minko, S. Patil, J. Pionteck, M. Stamm, Polym. Mater. Sci. Eng. 2001, 84, 877–878. 4 A. H. E. Mueller, G. Cheng, A. Boeker, G. Krausch, Polym. Mater. Sci. Eng. 2001, 84, 91–92. 5 D. Neugebauer, Y. Zhang, T. Pakula, K. Matyjaszewski, Polymer Preprints (Am. Chem. Soc., Div. Polym. Chem.) 2003, 44, 512–513. 6 D. Neugebauer, B. E. Carson, S. S. Sheiko, K. Matyjaszewski, Polymer Preprints (Am. Chem. Soc., Div. Polym. Chem.) 2003, 44, 510–511. 7 S. Qin, K. Matyjaszewski, H. Xu, S. S. Sheiko, Macromolecules 2003, 36, 605–612. 8 R. P. Quirk, R. T. Mathers, T. Cregger, M. D. Foster, Macromolecules 2002, 35, 9964– 9974. 9 W. Senaratne, L. Andruzzi, E. D. Sheets, D. Holowka, B. Ilic, A. Hexemer, B. Baird, E. J. Kramer, C. K. Ober, Polym. Mater. Sci. Eng. 2003, 88, 337–338. 10 D. Usov, M. Stamm, S. Minko, C. Froeck, A. Scholl, M. Mueller, Mater. Res. Soc. Symp. Proc. 2002, 727, 17–22. 11 D. L. Huber, K. Gonsalves, G. Carlson, T. A. P. Seery, in: Interfacial Aspects of Multicomponent Polymer Materials; D. J. Lohse, T. P. Russell, L. Sperling, Eds.; Plenum, 1997, pp. 107–122. 12 X. Fan, C. Xia, T. Fulghum, M. K. Park, J. Locklin, R. C. Advincula, Langmuir 2003, 19, 916–923. 13 Q. Zhou, X. Fan, C. Xia, J. Mays, R. Advincula, Chem. Mater. 2001, 13, 2465–2467. 14 T. E. Patten. Polymer Preprints (Am. Chem. Soc., Div. Polym. Chem.) 2003, 44, 425–426. 15 D. A. Savin, J. Pyun, G. D. Patterson, T. Kowalewski, K. Matyjaszewski, J. Polym. Sci., Part B: Polym. Physics 2002, 40, 2667–2676. 16 J. Pyun, K. Matyjaszewski, Chem. Mater. 2001, 13, 3436–3448. 17 M. Weck, J. J. Jackiw, R. R. Rossi, P. S. Weiss, R. H. Grubbs, J. Am. Chem. Soc. 1999, 121, 4088–4089. 18 M. Bartz, J. Kuther, R. Seshadri, W. Tremel, Angew. Chem., Int. Ed. 1998, 37, 2466–2468.

19 Q. Fu, T. A. P. Seery, Polymer Preprints

(Am. Chem. Soc., Div. Polym. Chem.) 2001, 42, 341–342. 20 Q. Fu, T. A. P. Seery. Mater. Res. Soc. Symp. Proc. 2001, 661, KK8.7/1–KK8.7/6. 21 N. L. Jeon, I. S. Choi, G. M. Whitesides, N. Y. Kim, P. E. Laibinis, Y. Harada, K. R. Finnie, G. S. Girolami, R. G. Nuzzo, Appl. Phys. Lett. 1999, 75, 4201–4203. 22 K. J. Watson, J. Zhu, S. T. Nguyen, C. A. Mirkin, J. Am. Chem. Soc. 1999, 121, 462–463. 23 M. A. Jordi, T. A. P. Seery, Polymer Preprints (Am. Chem. Soc., Div. Polym. Chem.) 2003, 44, 542–543. 24 D. L. Huber, G. Carlson, K. E. Gonsalves, T. A. P. Seery, Polym. Mater. Sci. Eng. 1996, 75, 265–266. 25 D. L. Huber, T. A. P. Seery, Polymer Preprints (Am. Chem. Soc., Div. Polym. Chem.) 1999, 40, 709–710. 26 T. A. P. Seery, P. Dhar, D. L. Huber, F. Vatansever, Polym. Mater. Sci. Eng. 1997, 77, 634–635. 27 Q. Fu, T. A. P. Seery, in: Materials Research Society Symposium Proceedings; A. I. Nakatani, R. P. Hjelm, M. Gerspacher, R. Krishnamoorti, Eds., 2000, 661, KK8.7.1–KK8.7.6. 28 T. A. P. Seery, M. Jordi, Q. F. Adkins, Polymer Preprints (Am. Chem. Soc., Div. Polym. Chem.) 2003, 44, 427–428. 29 M. Brust, M. Walker, D. Bethell, D. J. Shiffrin, R. Whyman, J. Chem. Soc. Chem. Commun. 1994, 801–802. 30 W. Stober, A. Fink, E. Bohn. J. Colloid Interface Sci. 1968, 26, 62. 31 A. Van Blaaderen, A. Vrij, J. Colloid Interface Sci. 1993, 156, 1–18. 32 M. Brust, C. J. Kiely, Colloids Surfaces A: Physicochem. Eng. Aspects 2002, 202, 175–186. 33 J. P. Wilcoxon, J. E. Martin, P. Provencio, Langmuir 2000, 16, 9912–9920. 34 T. E. Patten, B. M. Novak, J. Am. Chem. Soc. 1996, 118, 1906–1916. 35 T. E. Patten, in: Chemistry. University of California: Berkeley, 1994. 36 J. Pyun, K. Matyjasewski, T. Kowalewski, D. Savin, G. Patterson, G. Kickelbick, N. Huesing, J. Am. Chem. Soc. 2001, 123, 9445–9446. 37 T. von Werne, T. E. Patten, J. Am. Chem. Soc. 2001, 123, 7497–7505.

231

12

Spherical Polyelectrolyte Brushes Matthias Ballauff

Abstract

In this chapter, we review recent studies performed with spherical polyelectrolyte brushes that consist of a colloidal core with attached polyelectrolyte chains. Initially, the synthesis is discussed, which is achieved using a photochemical “grafting from” technique (photoemulsion polymerization), after which the theoretical predictions currently available for these systems are compared with experimental data. In particular, the strong confinement of the counterions within the brush layer is noted. The results obtained from this analysis can be extended to explain the flow behavior of dilute suspensions of these particles. Finally, the applications of these particles are discussed, together with their interaction with charged planar surfaces. In addition, the interaction of spherical polyelectrolyte brushes with proteins in solution is reviewed. All results can be traced back to the strong confinement of the counterion within the brush layer on the surface of the particles.

12.1

Introduction

If linear polymer chains are grafted densely to planar surfaces, then the result is a polymer brush [1–3]. In such a layer, the average distances between two grafted chains is much smaller than their dimensions, for example as their unperturbed coil size. In this brush limit the properties of the surface layer is dominated by the mutual interaction of the neighboring chains, and this leads to a marked stretching along the layer normal. A survey of studies conducted with uncharged planar brushes is provided in Chapter 23. If charged chains are grafted to planar surfaces, then polyelectrolyte brushes are generated. These systems have attracted much attention recently (for a review, see Ref. [4]), commencing with the seminal studies of Pincus [5] and continuing with investigations by Borisov, Birshtein and Zhulina [6] on planar polyelectrolyte brushes.

232

12 Spherical Polyelectrolyte Brushes

In planar polyelectrolyte brushes, electrostatic interactions between the chains and the osmotic pressure can become the dominant force. This fact is clear to see for the limiting case of the so-called osmotic brush – that is for a polyelectrolyte brush in the absence of added salt. Pincus [5] showed that the neutralization length f is the decisive length scale for a dense polyelectrolyte brush. Within a distance of f the interaction between the fixed charged and its counterion has decayed strongly, and Pincus argued that f is given by the Debye length in dense brushes. If the thickness L of a brush is much larger than f, the strong electrostatic forces must be balanced within the brush and local electroneutrality results. Therefore, virtually all counterions must be confined within the brush, and the translational entropy can only be increased by a strong stretching of the chains – that is, by a marked increase of L. For planar brushes, these predictions have been confirmed experimentally by the studies of Ahrens et al. [7]. Moreover, Tran et al. [8,9] showed that the distribution of counterions in a planar polyelectrolyte brush follows closely the profile of the monomer units. All findings thus demonstrate that the local electroneutrality is the dominant effect in planar systems. Grafting linear polyelectrolyte chains densely to colloidal core latex particles leads to spherical polyelectrolyte brushes (SPB) in which the curvature of the cores becomes a new decisive length parameter [10]. Figure 12.1 shows the schematic structure of such a spherical polyelectrolyte brush that will form the subject of this chapter. The core is inert and totally impenetrable for the dispersion medium water, and the polyelectrolyte chains may be anionic, as shown in Figure 12.1(a). Chains of a weak polyelectrolyte such as poly(acrylic acid) (PAA) give an annealed brush, while the grafting of strong polyelectrolytes (e.g., poly(styrene sulfonic acid; PSS) leads to quenched brushes [1]. A typical cationic SPB is shown in Figure 12.1(b). All systems are intermediates between the polyelectrolyte star polymers and planar brushes: systems with small cores and long grafted chains closely resemble charged star polymers, whereas small chains grafted to much larger chains can be compared to planar brushes. The SPB are characterized by three parameters: the core radius, R; the contour length, LC of the grafted chains; and the grafting density, r, giving the number of grafted chains per nm2. In order to allow a meaningful comparison with well-characterized planar systems, the size distribution of the core particles should be small. Moreover, all parameters characterizing the SPB should be known with sufficient accuracy. This goal can be achieved by growing the polyelectrolyte chains directly on the surface by generating radicals through a photo-initiator [10,11]. This method, which is termed “photo-emulsion polymerization” allows a brush layer to be generated from virtually any water-soluble monomer. The obvious drawback, however, is the broader polydispersity because of the uncontrolled nature of the radical polymerization thus started. Spherical polyelectrolyte brushes have also been prepared by dissolution of block copolymers in suitable solvents [12]. For suitable systems, this may lead to the formation of spherical micelles with a hydrophobic core and a hydrophilic shell. If the hydrophilic block consists of polyelectrolyte chains, the resulting structure may be

12.1 Introduction

a)

CH CH2 COO CH CH2

s

D SO3-

R a L

b)

Structure of the spherical polyelectrolyte brushes having (a) anionic chains and (b) cationic polyelectrolyte chains on their surface. The core consists of poly(styrene), and has a diameter of ca. 100 nm. The chains are

Figure 12.1

densely grafted to the surface of these cores by a “grafting from” technique (“photoemulsion polymerization”, cf. Refs. [10,11]). The particles are immersed in water.

described by Figure 12.1 in good approximation. It should be noted, however, that the micelles thus formed always have an appreciable polydispersity which can directly be assessed from small-angle neutron scattering (SANS) [13–15]. The finite breadth of the size distribution renders analysis of the various properties (e.g., flow behavior) more difficult. Moreover, the interface between the hydrophobic core and the hydrophilic shell cannot be very sharp, given the fact that the micelles result from self-assembly. In addition to this, the blocks of the copolymer exhibit a finite width of the size distribution, and this difficulty will further aggravate the problem of polydispersity. It should be noted in this context, however, that micelles consisting of hydrophobic blocks and poly(acrylic acid) have served recently as model systems for fundamental studies of strongly curved polyelectrolyte brushes [13–16]. In this chapter, we will review recent investigations carried out on spherical polyelectrolyte brushes, with special focus on systems generated by photo-emulsion po-

233

234

12 Spherical Polyelectrolyte Brushes

lymerization [10,11] (as described in Section 12.2). Special emphasis will also be paid to the comparison with planar systems. Suspensions of SPB in water create a large surface that is of the order of hundreds of m2 per cm3. Hence, suspensions of SPB are wellsuited for studying the interaction of polyelectrolyte brushes with planar surfaces or with dissolved proteins. These topics will be discussed in Section 12.5.

12.2

Synthesis and Characterization 12.2.1

Determination of Core Radius R, Contour Length LC, and Grafting Density r

All systems which are discussed here have been prepared using photo-emulsion polymerization [10,11]. The synthetic route is shown schematically in Figure 12.2. First, near-monodisperse PS cores are made by emulsion polymerization, after which a thin layer of a suitable photoinitiator is polymerized on these cores. As shown in Figure 12.3, the photoinitiator HMEM used until now carries a double bond; it is a vinyl monomer [10]. Thus, the second step proceeds in a seeded emulsion polymerization where the concentration of the new monomer is kept low enough to avoid formation of new particles. In the third step, a water-soluble monomer is added to this core-shell latex and radicals are generated on the surface of the particles by shining light on the suspension [10]. These radicals start a polymerization on the surface, and the chains grow directly from the surface [10,11]. One radical will be generated in the solution, and the polymer growing in solution must be removed after the synthesis of the particles [10].

: PS-core latex

Shell: Photoinitiator Shell composed of linear polyelectrolytes

Schematic representation of the synthesis of defined polyelectrolyte brushes. In the first step, polystyrene latex is generated. In a second step, the latex is covered by a thin layer of the photo-initiator, HMEM. In the third

Figure 12.2

step, photo-initiation in the presence of water soluble monomers (e.g., acrylic acid) leads to polyelectrolyte chains grafted onto the surface of the core particles [10,11].

12.2 Synthesis and Characterization O O

O

C

OH

C O

O O

O

C.

+

. C OH

C O

The photo-initiator HMEM and its decomposition into radicals during photo-initiation [10].

Figure 12.3

The core radius R is fixed by the seed latex used in step 1 of Figure 12.2. The contour length LC of the grafted chains and the grafting density r – that is, the number of grafted chains per nm2 – can be obtained by cleaving off the chains from the surface and analyzing them separately [10,11]. The hydrodynamic radius RH of the SPB in aqueous dispersion is determined by dynamic light scattering. All experiments were carried out at high dilution, so that mutual disturbance of the particles can be ruled out easily [11,17]. As the core radius R is known precisely, the thickness L of the brush layer can be obtained as the function of the concentration of added salt and of the pH. Results obtained on the colloidal systems discussed here can therefore be compared with the well-defined planar polyelectrolyte brushes [7–9,18–20]. 12.2.2

Titration Curve

In the case of weak polyelectrolyte chains (e.g., poly(acrylic acid); PAA), the pH is the main variable that determines the number of charges within a brush. For low pH, most of the carboxyl groups will remain uncharged and the polyelectrolyte will resemble uncharged systems. For pH > pKa of the polyacid, however, a fully charged system will be obtained and the osmotic pressure of the confined counterions must lead to a concomitant increase of L. The strong dependence on pH of the brush layer consisting of PAA chains is directly seen in Figure 12.4 [11,17]. Here, the thickness L = RH – R of the brush layer is plotted against the pH of the solution. It should be borne in mind that the pH was adjusted while keeping the ionic strength constant. There is a strong dependence on the concentration ca of added salt which would obscure the effect of the pH (see below). This also implies that the pH range that can be explored is much smaller at low ionic strength, of course. Figure 12.4 shows that L may increase by almost an order of magnitude when going from low to high pH [11,17]. A study of the respective quenched SPB demonstrates that, as expected, no change takes place upon changing the pH [17]. The pressure which is built up by the counterions in the brush can stretch the chains to almost full length, and for monovalent counterions this is the leading physical effect

235

12 Spherical Polyelectrolyte Brushes 300

200

L [nm]

236

Dependence of brush thickness L on pH for an annealed brush [17]. Parameter of the data is the ionic strength in the solution that had been adjusted by adding KCl. The latex L23 (see Table II of Ref. [17]) has been used. The parameters of this system are: core radius R = 66 nm; contour length of the grafted chains Lc = 228 nm; grafting density r = 0.039 nm–2. Figure 12.4

100

0 0

2

4

6

8

pH

10 12 14

which dominates all other forces in the system. For divalent counterions, the stretching is dramatically reduced: here, the charges of the polyelectrolyte chains are counterbalanced by only half of the number of counterions, and this greatly reduces the osmotic pressure within the brush layer [17]. It is clear that the above reasoning requires that the electrostatic interaction between the polyelectrolyte chains and the counterions is of substantial range – that is, larger than the typical monomer lengths or ion diameters. Raising the ionic strength by adding large amounts of salt, however, will decrease the Debye length so much that electrostatic interaction no longer plays any role. Such a “salted brush” is predicted [5] to resemble the well-studied uncharged systems that are indeed found experimentally for planar systems [7]. The same is found for the strongly curved systems under consideration here [11,17]. Figure 12.4 shows that the stretching is much less pronounced when the titration is carried out at higher concentrations of added salt. At the highest salt concentration (1 M), the effect is virtually nil, and the build-up of charge when the pH is raised is mostly inconsequential. The effect of added salt, which is assessed by measuring L values at constant pH but varying the salt concentration, can be understood quantitatively in terms of the theory of Russel and coworkers [21]. A decisive step of this treatment is the calculation of the total ionic strength within the brush layer in presence of added salt. Assuming local electroneutrality – that is, all counterions are confined within the brush and the particle is electrically neutral – this problem can be treated as the classical Donnan equilibrium. The predictions of the theory of Russel and coworkers allow a quantitative description of L as a function of the added salt. For details of the comparison of theory and experiment, the reader is referred to Refs. [11] and [17].

12.3

Experimental Verification of Theoretical Predictions

There are several features of polyelectrolyte brushes that determine the properties of these systems. The most important one is the confinement of counterions [5], as

12.3 Experimental Verification of Theoretical Predictions

discussed above. The SPB were the first system where this important prediction could be tested in a direct experiment (for discussion, see Section 12.3.1). A more subtle point is the correlation of the counterions to the macroion. Here, recent progress in experimental technique has led to a direct probe of this correlation, the details of which are outlined in Section 12.3.2. 12.3.1

Confinement of the Counterions

The confinement of the counterions can be demonstrated directly by osmotic measurements [22]. In this experiment, salt-free suspensions of quenched SPB bearing PSS chains (see Ref. [17]) are confined in an osmometer cell, and the osmotic pressure is measured against a cell containing pure water. The number of counterions in the system widely exceeds the number of particles, and only the counterions contribute to the measured osmotic pressure [22]. Two limiting cases can be discussed. First, if the counterions were free to move out of the brush layer, then the measured osmotic pressure would be given by the ideal term calculated from van t’Hoff’s law. If, on the other hand, the counterions were fully confined within the particle, the osmotic pressure of the suspension would be exceedingly small. The fraction of osmotically active counterions is expressed through the osmotic coefficient – that is, the ratio of the real to the ideal osmotic pressure. Figure 12.5 illustrates the osmotic coefficient measured for three different SPB. These systems had previously been characterized by dynamic light scattering [17]. All experimental data demonstrate that the osmotic coefficient is of the order of a

0.06

φ

0.04

0.02

0

0

0.1

0.2 c (g L-1)

0.3

The osmotic coefficient of quenched spherical polyelectrolyte brushes as a function of the polyelectrolyte brush concentration [22]. These systems bear chains of poly(styrene sulfonic acid). Three different spherical polyelectrolyte brushes have been

Figure 12.5

0.4 studied (see Ref. [22]): LQ2 (~; core radius R = 68 nm; contour length of the grafted chains Lc = 86 nm; grafting density r = 0.033 nm–2); LQ4 (j; R = 68 nm; Lc = 147 nm; r = 0.037 nm–2); LQ6 (d; R = 68 nm; Lc = 165 nm; r = 0.027 nm–2).

237

238

12 Spherical Polyelectrolyte Brushes

few percent only. This clearly shows that 95–98% of the counterions are confined within the brush layer. It is thus evident that Pincus’ central idea [5] is fully verified by direct experimental measurements. The measured osmotic coefficient diminishes for an increasing contour length of the grafted chains. This can be seen from Figure 12.5, and may be explained by the lower average osmotic pressure for increasing Lc. If Lc is increased linearly, L will also be raised if no salt has been added. The concomitant increase in the volume available for the counterions, however, will be of the order of L2. This will alleviate the osmotic pressure within the brush layer and, as a consequence, the fraction of counterions leaving the layer must decrease. A quantitative treatment of this problem is not yet available, however. 12.3.2

Correlation of the Counterions to the Macroion

Most of the counterions of the highly charged PSS or PAA chains will be strongly correlated to the chains, and therefore remain osmotically inactive. This phenomenon, which has been termed “counterion condensation” by Manning [23] seems to be well understood for linear polyelectrolytes. Recent studies by Likos et al. [24,25] have shown that a high fraction of the counterions within a star-like polyelectrolyte should also be correlated to the chains of the brush. From the results of these studies it can be concluded that the counterions of the brush layer may be subdivided into three classes: 1) Counterions that are strictly correlated to the polyelectrolyte chains; 2) counterions that are free to move within the brush; and 3) a small fraction that can leave the brush and is osmotically active. It is the latter class of counterions that is detected by the osmotic measurements discussed in the previous section. SANS and SAXS are uniquely suited for experimental testing of these predictions. A number of systematic SANS studies have been performed by van der Maarel and coworkers using suitable block copolymers in aqueous solution [13–15]. These block copolymers build up micelles with a hydrophobic core and a corona consisting of polyelectrolyte chains. To enhance the contrast, deuterated counterions based on quaternary ammonium salts had to be used. These investigations show that the counterions are confined within the brush. SANS, however, may be hampered by the necessary choice of rather large hydrophobic counterions. More recently, we were able to demonstrate that SAXS can be used for the same purpose if rubidium counterions are used [26–28]. Here, the K-absorption edge can be reached by use of synchrotron radiation of suitable energy. The contrast of the Rb+ counterions is diminished markedly, while the contrast of the core and the polyelectrolyte chains stays constant. Hence, using the anomalous dispersion of the Rb+ ions the scattering contribution of the macroion and the counterions can be determined separately. This method, termed anomalous small-angle X-ray scattering (ASAXS) [29], was applied successfully to anionic SPB [27,28]. Hence, near the adsorption edge the scattering factor f becomes a function of energy E and can be rendered as [29]:

12.3 Experimental Verification of Theoretical Predictions 0

00

f ¼ f0 þ f ðEÞ þ if ðEÞ

(1)

where i is the imaginary unit, f0 is the scattering factor far below the edge, and f ¢ and f † are the real and the imaginary part of f, respectively. The terms f ¢ and f † differ from zero only in the immediate neighborhood of the edge. The scattering intensity I(q) (q: magnitude of scattering vector; q = (4p/k)sin(h/2); k: wavelength of radiation; h: scattering angle) may then be shown to consist of three additive terms [26–28]: 0

2

00 2

IðqÞ ¼ F0 ðqÞ þ 2f F0 ðqÞvðqÞ þ f v ðqÞ

(2)

Here, F02(q) is the nonresonant part of the scattering intensity – that is, the intensity measured in a conventional SAXS experiment. The second term is the cross term of the scattering amplitude F0(q) of the nonresonant part and the amplitude of the resonant part v(q). The third term is related only to the resonant scattering units. In the present case, the rubidium counterions are the only constituents for which f ¢ and f † differ from zero [27,28]. Hence, v2(q) is the scattering intensity of the counterion cloud surrounding the macroion [27]. We have shown for the first time that all three terms can be obtained experimentally [28]. Details of the theory and the data treatment may be found in references [27,28]. Figure 12.6 provides all three terms as a function of q [28]. It is clear that the terms have the same dependence on q. Moreover, the ratio of term 2 and term 3 is a constant. This demonstrates that most of the counterions are indeed closely correlated to the polyelectrolyte chains, the spatial distribution of the macroion and of the 3

10

2

F0(q)v(q)/v (q)

1

2

F 0(q)

-1

I(q) [cm ]

10

-1

10

F0(q)v(q)

-3

10

2

v (q) -5

10

0

0.2

0.4

0.6

-1

q [nm ] Anomalous small-angle X-ray scattering (ASAXS) applied to spherical polyelectrolyte brushes [28]. Decomposition of the measured intensity according to Eq. (2). The upper curve displays the ratio of the second to the third term of Eq. (2). The term F02(q) represents the intensity that is measured far below the absorption edge. v2(q) presents the term

Figure 12.6

that depends only on the components thatscatter resonantly. In the case of the spherical polyelectrolyte brushes considered here, the measurements have been conducted near the absorption edge of the rubidium counterions. Hence, v2(q) is the scattering intensity solely due to the radial distribution of the counterions.

239

240

12 Spherical Polyelectrolyte Brushes

counterions coincide in good approximation. In this way, the counterions decorate the macroion, and this is in full accord with the predictions of Likos et al. [24,25].

12.4

Flow Behavior

The viscosity of a suspension of spheres is one of the oldest topics in colloid physics. The first fundamental contribution is due to Einstein, who derives his famous relation describing the viscosity of dilute suspensions of spheres in 1905 (cf. Ref. [30]). Binary hydrodynamic interactions have been treated by Batchelor in 1967 [31]. Hence, the zero-shear viscosity g0 of a dilute suspensions can be modeled on an exact theoretical basis. It reads [31] g0 2 ¼ 1 þ 2:5 ueff þ 5:9 ueff þ ::: gs

(3)

where gS is the viscosity of the medium in which the spheres are dispersed and ueff is the effective volume fraction that is directly related to the hydrodynamic radius RH by ueff ¼

4p 3 RH 3

(4)

RH derived from such an analysis can be compared directly to the hydrodynamic radius obtained from dynamic light scattering that has often been applied to colloidal particles (see the discussion in Refs. [32,33]). Both sets of data need not necessarily coincide because the measurements have been made at entirely different concentration regimes. Dynamic light scattering is carried out at high dilution where no interaction of the particles occur. Rheological measurements are carried out at much higher concentrations where binary interaction of the particles can no longer be disregarded. Figure 12.7 displays the relative viscosity g0/gS that was determined in dilute aqueous solution [34]. The effective volume fraction ueff was used as a fitting parameter (cf. Eq. (3)). All data obtained in this way are located on a master curve when plotted against ueff. Figure 12.7 also demonstrates that a wide range of the concentration of added salt has been covered. This shows that ueff is the only decisive variable far beyond the range of concentrations in which Eq. (3) is valid. This surprising result should be explored in more detail. The hydrodynamic radii obtained from ueff using the above procedure – that is, by application of Eqs. (3) and (4) – can again be compared to the data obtained by dynamic light scattering [17], and an excellent agreement is found [34]. As mentioned above, this agreement is not necessarily to be expected because both sets of data were taken at totally different concentrations. Clearly, there is no change in RH when going from the highly dilute regime to concentrations where binary hydrodynamic interactions can no longer be neglected [33,34]. This fact again underscores the finding that ueff is the single decisive parameter over a wide range of concentrations.

12.4 Flow Behavior

Moreover, Figure 12.7 demonstrates that the master curve g0/gS(ueff ) coincides with the master curve derived for a dilute system of hard spheres taken from the work of Meeker et al. (the dashed line in Figure 12.7) [35]). In dilute suspension, the SPB carrying 103 to 104 net charges may hence be treated in terms of a model developed for hard spheres. This finding again points to the fact that most of the charges are localized within the brush layer, as discussed in Section 12.3.1. Only a few percent of the charges can leave this layer, as detected by osmometry (see Figure 12.5). If an appreciable fraction of the counterions were free to move outside the brush, the diffusion coefficient as well as the viscosity would exhibit marked electroviscous effects [30]. Evidently, RH obtained in this case from dynamic light scattering would strongly differ from RH evaluated by means of Eq. (4), but this is not the case. All findings discussed so far – namely the titration curve (Figure 12.4), the osmotic coefficient (Figure 12.5), the spatial distribution of the counterions as revealed by ASAXS (Figure 12.6), as well as the rheology in dilute solution (Figure 12.7) – demonstrate the near- total confinement of the counterions within the brush [5]. If ueff is raised much beyond 0.1, the SPB appear to shrink. This can be argued from the comparison of the master curve g0/gS = g0/gS(ueff ) (Figure 12.7) with the master curve of hard spheres. If the particles exactly retain their hydrodynamic radius RH up to the highest effective volume fractions ueff, all data should be located on this master curve. However, Figure 12.7 shows that g0/gS of the SPB is much lower at higher volume fraction, which may be explained by the onset of shrinking of the brush layer. A small decrease in the hydrodynamic radius of the particle may already explain the result shown in Figure 12.7. The longest polyelectrolyte chains on the surface will retract if they come into close contact with neighboring chains,

2.5 HS-mastercurve 1 mM 500 mM 100 mM 50 mM 10 mM 0.1 mM

η0/ηS

2.0

1.5

1.0

0

0.1

0.2

0.3

φeff Figure 12.7 Relative viscosity g0/gS of a spherical polyelectrolyte brush extrapolated to vanishing shear rates as a function of the effective volume fraction ueff. This parameter has been calculated using Eq. (2), and the hydrodynamic radii RH obtained from dynamic light scattering [34].

241

242

12 Spherical Polyelectrolyte Brushes

and a concomitant shrinking of the surface layer is expected [34]. We reiterate, however, that the effective volume fraction ueff derived from dilute suspensions is still the only decisive parameter in the system. All data superimpose on the master curve g0/gS = g0/gS(ueff ), even when the chains obviously start to retract. Further investigations are needed to clarify this point.

12.5

Applications 12.5.1

Interaction with Charged Surfaces

Practical applications [36], as for example with paints and coatings, often require detailed knowledge of the interaction of latex particles with solid substrates. Atomic force microscopy (AFM) is the method of choice for the investigation of this problem (see the survey of current literature in Ref. [37]). AFM has been applied to the study of the first stage of film formation by polystyrene particles. It was shown that the combination of AFM and SAXS is ideally suited to determine the spatial order of the particles [37,38]. Recently, the interaction of the SPB with negatively charged mica surfaces has been the subject of a comprehensive investigation [39]. Two types of SPB have been used for this study: the anionic systems shown in Figure 12.1(a); and the cationic SPB shown in Figure 12.1(b). Dilute aqueous suspensions of anionic as well as cationic SPB were brought into contact with mica surfaces carrying negative charges. The arrangement of the particles in dry samples was studied by AFM. The anionic system has little interaction with the negatively charged surface, as expected. The resulting arrangements of the particles are therefore akin to earlier observations made on polystyrene particles [37,38]. If, however, the cationic system (Figure 12.1(b)) is brought into contact with a negative surface in the absence of added salt, a strong attraction can be inferred from the observations by AFM [39]. Figure 12.8 shows the main result of this study (see Figure 3 of Ref. [39]). The overall arrangement of the particles is shown in Figure 12.8(a) and (b), with them forming a two-dimensional network on the surface. Moreover, Figure 12.8(b) shows that the tendency of the particles to stick together seems to be small, with only two to three particles aggregating. This contrasts strongly with the usual finding made in studies of negatively charged particles [37,38]. In this case, the strong van der Waals forces between the particles leads to aggregation and the formation of two-dimensional crystals. The interaction between the cationic SPB and the surface seems to be of appreciable magnitude. Once a particle has come into contact with the mica sheet, it exhibits a restricted mobility and can move only locally to come into contact with its neighbors, and no further migration parallel to the surface seems possible. It this way only chains or rather small aggregates can be assembled. The reason for this strong interaction becomes obvious when examining the particles more closely [39]. Figure 12.8(c) shows an enlarged view of a few particles

12.5 Applications

Intermittent contact AFM images of (a,b) topography and (c) phase contrast of the cationic SPB LA2 having polycation chains attached to their cores [39]. The schematic structure of these particles is shown in Figure 12.1(b). The parameters of this system are: core radius R = 45.3 nm; contour length of the grafted chains Lc = 68.6 nm; grafting density Figure 12.8

r = 0.083 nm–2. The solid substrate is mica. (a) Scan size: 6 lm C 6 lm, z-scale: 200 nm; (b) scan size: 1 lm C 1 lm, z-scale: 300 nm; (c) scan size: 1 lm C 1 lm, z-scale: 52D. In (c), the particles show an irregular phase distribution on their surfaces and are surrounded by a corona; as shown in (a) the particles tend to form network-like aggregates on the surface.

obtained by the phase picture of AFM. A corona is seen that extends ca. 10 nm outside the particles and which must be related to the cationic polyelectrolyte chains attached to the surface of the cores of the cationic SPB (see Figure 12.1(b)). These chains interact strongly with the negative mica surface while still being dispersed in the water phase. The solid substrate can replace the counterions confined within the brush, and thus partly release the strong osmotic pressure in the brush layer (cf. Section 12.3.1). Hence, attaching long polyelectrolyte chains to the surface may be used to tune the interaction of latex particles with solid surfaces. However, the implications for possible technical applications remain to be explored in more detail. 12.5.2

Interaction with Proteins in Solution

The adsorption of proteins from solution to solid substrates is an important phenomenon [40]. Often, protein adsorption has to be avoided, for example in the field

243

244

12 Spherical Polyelectrolyte Brushes

of biocompatible materials, and in this case the surface must be modified by suitable groups to resist protein adsorption [41]. Alternatively, proteins may be immobilized on the surface of colloidal particles in order to obtain functional microspheres [42]. Colloidal particles are well-suited to study the adsorption of dissolved proteins to solid surfaces as these particles generate in solution a large surface area with welldefined properties. The process of adsorption onto the surface of the particles can therefore be easily monitored. In this section we discuss the interaction of dissolved proteins with SPB. The previous section has shown that the polyelectrolyte layer attached to the surface of these particles represents a well-characterized brush. Given the high surface created by these colloidal particles, it becomes evident that aqueous suspensions of SPB present a suitable model system for studying the adsorption of proteins from solution. Until now, very few studies of the adsorption of dissolved proteins to polyelectrolyte brushes have been reported. Tran et al. [43] investigated the adsorption of lysozyme and fibrinogen on a planar brush consisting of highly charged poly(styrenesulfonic acid) by neutron reflectivity. The positively charged lysozyme is irreversibly adsorbed to the negatively charged brush, as expected. Tran et al. found, however, that the negatively charged fibrinogen was also strongly adsorbed to the brush [43], and explained this unexpected result by the specific interaction of the fibrinogen molecule with the sulfonate-groups of the brush. The experimental procedure used for studying the protein adsorption onto SPB is shown schematically in Figure 12.9 [44]. Bovine serum albumin (BSA) was used in

Schematic representation of the experiment. Solutions of bovine serum albumin were prepared in buffer solutions (MES) with defined concentrations of added salt.

Figure 12.9

These solutions were added to the SPB dissolved in the same buffer. After equilibration for 24 h, the nonadsorbed protein was removed by careful serum replacement [44].

12.5 Applications

τads [mg/g SPB]

1200 800 400 0 0

4

8

12

csol [mg/ml] Figure 12.10 Adsorption of BSA onto the SPB KpS13 at pH = 6.1, and at a concentration of the MES buffer of 10 mM [44]. The amount of adsorbed BSA sads per unit mass of the particles is plotted against the concentration

csol of the protein left unadsorbed in solution for various concentrations of added salt (NaCl). j: no added salt (7 mM); s: 32 mM; ,: 57 mM; h: 107 mM; d: 157 mM.

these investigations. The SPB solutions, with adjusted pH and of known ionic strength, were mixed with a BSA solution of equal ionic strength and pH. All mixtures were stirred for 24 h, and subsequently the nonadsorbed protein was removed by ultrafiltration against a salt solution of same concentration and pH. The amount of protein removed in this step was determined quantitatively, and in this way the amount of bound protein could be determined from the mass balance. Additional experiments demonstrated that the entire system had reached a well-defined state [44]. Figure 12.10 displays the principal results of this study. The amount of adsorbed BSA sads per unit mass of particles was plotted against the concentration csol of the protein left unadsorbed in solution. The concentration of SPB particles was 1 wt% in all runs. Figure 12.10 shows that BSA is strongly adsorbed if the ionic strength in the system is low, while virtually no adsorption takes place for a high ionic strength. This result was supported by the following observation. The bound protein can be washed from the particles by a salt solution of much higher ionic strength [44]. Thus, flushing the particles carrying BSA by a NaCl solution of 0.5 M concentration removes the protein quantitatively. This is a very important observation inasmuch it shows that the adsorption is not irreversible, but may be reversed upon increasing the ionic strength to values where the electrostatic interaction no longer plays any role. In this way, the SPB presents as a new class of carrier particles, the interaction of which with proteins can be adjusted by the ionic strength. The conformation of BSA desorbed from the SPB in this way can be analyzed in order to detect possible changes, or even denaturation. Circular dichroism measurements of the BSA washed from the particles again showed that only minor changes of the secondary structure had been effected by the process of adsorption [45]. Hence, only be a small conformational change occurs when BSA is adsorbed onto the SPB. All results obtained from the SPB [44,45] were in opposition to the usual finding that a high ionic strength furthers the adsorption of proteins to polymeric surfaces

245

246

12 Spherical Polyelectrolyte Brushes

[40]. In the case of flat substrates, the high ionic strength lowers the electrostatic repulsion between the surface and the dissolved protein. The same repulsion should also operate in the present mixtures, and an increased adsorption is to be expected at high ionic strength. Conversely, a low ionic strength is expected to prevent adsorption of dissolved proteins because both the protein and the particle are both negatively charged. Moreover, the layer of densely grafted polymer chains – that is, the brush layer on the surface of the particles – should exert a strong steric repulsion onto the protein molecules that can be envisioned as small colloidal particles. All currently available data on SPB show that the decisive parameter is the ionic strength, while the pH only modifies the strength of adsorption [44]. The results obtained so far can be explained as follows [44]. The linear polyelectrolyte chains can interact with the positive charges on the surface of the protein molecules and, depending on the number of these positive patches, the interaction may be strong. Thus, the positive patch of the protein becomes a multivalent counterion which neutralizes several negative charges of the linear polyelectrolyte chain. The activity of the counterions within a brush, however, is very low, as demonstrated in Section 12.3.1. For multivalent counterions this effect becomes even stronger (see the discussion of this point in Ref. [17]). Moreover, for each neutralized patch the respective negative counterions together with the positive counterions of the PAA chains are released. This is connected to a concomitant gain of entropy of the entire system. If the proteins were to be released from the brush again, the concomitant number of counterions must be brought back to the brush. The loss of entropy related to such a process renders it much less probable, and the proteins are bound tightly to the brush (“counterion release force”; see the discussion in Ref. [46]). This would also explain the strong adsorption of BSA that persists if the ionic strength in the solution is kept constant: the activity of the protein molecules serving as counterions falls to very low values, and the equilibrium concentration of the proteins outside is, in consequence, negligibly small. If salt is added, however, the effect vanishes and the protein molecules are replaced by the ions of the added salt.

Summary

An overview of experimental results obtained on spherical polyelectrolytes so far has been presented in this chapter. These findings demonstrate that all concepts derived for planar polyelectrolyte brushes [1,47] can also be applied to these strongly curved systems. Most notably, virtually all counterions are confined within the brush layer attached to the surface of the particles. This fact can easily explain all findings related to the titration curve (see Section 12.2.2) or the rheology in dilute solution (see Section 12.4). The high osmotic pressure of the counterions within the brush can be partially released by multivalent counterions [17] or by proteins [44]. The confinement of the counterions can therefore explain the unexpected finding that proteins adsorb to SPB in salt-free solutions. Hence, SPB may serve as novel carrier particles for proteins and enzymes. This also explains the strong interaction of cat-

References

ionic systems with negatively charged solid substrates [39]. If the salt concentration is raised, however, the electrostatic interaction is strongly screened and the SPB behave more or less as uncharged spherical brushes or star polymers [17]. Under these conditions, proteins are repelled by the steric repulsion operating between hairy particles and the surfaces [30].

References 1 G. J. Fleer, M. A. Cohen Stuart, J. M. H. M.

2 3 4 5 6 7 8 9 10 11 12 13

14

15

16

17

Scheutjens, T. Cosgrove, B. Vincent, Polymers at Interfaces. Chapman & Hall, London, 1993. O. Prucker, J. RRhe, Macromolecules 1998, 31, 592–601. O. Prucker, J. RRhe, Macromolecules 1998, 31, 602–613. F. Dubreuil, P. Guenon, Eur. Phys. J. E. 2001, 5, 59–64. P. Pincus, Macromolecules 1991, 24, 2912– 2919. O. V. Borisov, T. M. Birshtein, E. B. Zhulina, J. Phys. II (France) 1991, 1, 521–526. H. Ahrens, S. FSrster, C. A. Helm, Phys. Rev. Lett. 1998, 81, 4172–4175. Y. Tran, P. Auroy, L.-T. Lee, Macromolecules 1999, 32, 8952–8964. Y. Tran, P. Auroy, Eur. Phys. J. 2001, 5, 65–79. X. Guo, A. Weiss, M. Ballauff, Macromolecules 1999, 32, 6043–6046. X. Guo, M. Ballauff, Langmuir 2000, 16, 8719–8726. M. Moffit, K. Khougaz, A. Eisenberg, Acc. Chem. Res. 1996, 29, 95. W. Groenewegen, S. U. Egelhaaf, A. Lapp, J. R. C. van der Maarel, Macromolecules 2000, 33, 3283–3293. W. Groenewegen, A. Lapp, S. U. Egelhaaf, J. R. C. van der Maarel, Macromolecules 2000, 33, 4080–4086. J. R. C. van der Maarel, W. Groenewegen, S. U. Egelhaaf, A. Lapp, Langmuir 2000, 16, 7510–7519. S. FSrster, N. Hermsdorf, Ch. BSttcher, P. Lindner, Macromolecules 2000, 35, 4096– 4105. X. Guo, M. Ballauff, Phys. Rev. E 2001, 64, 051406-1–051406-9.

18 S. W. An, P. N. Thirtle, R. K. Thomas,

F. L. Baines, N. C. Billingsham, S. P. Armes, J. Penfold, Macromolecules 1999, 32, 2731. 19 E. P. K. Currie, A. B. Sieval, M. Avena, H. Zuilhof, E. J. R. SudhSlter, M. A. Cohen Stuart, Langmuir 1999, 15, 7116–7118. 20 M. Biesalski, J. RRhe, D. Johannsmann, J. Chem. Phys. 1999, 111, 7029–7037. 21 R. Hariharan, C. Biver, W. B. Russel, Macromolecules 1998, 31, 7514–7518. 22 B. Das, X. Guo, M. Ballauff, Prog. Colloid Polym. Sci. 2002, 121, 34–38. 23 G. Manning, Annu. Rev. Phys. Chem. 1972, 23, 117. 24 A. Jusufi, C. N. Likos, H. LSwen, Phys. Rev. Lett. 2002, 88, 018301-1–018301-4. 25 A. Jusufi, C. N. Likos, H. LSwen, J. Chem. Phys. 2002, 116, 11011–11027. 26 Q. de Robillard, X. Guo, N. Dingenouts, M. Ballauff, G. Goerigk, Macromol. Symp. 2001, 164, 81–90. 27 N. Dingenouts, R. Merkle, X. Guo, T. Narayanan, G. Goerigk, M. Ballauff, J. Appl. Crystallogr. 2003, 36, 578–582. 28 N. Dingenouts, M. Ballauff, D. Pontoni, T. Narayanan, G. Goerigk, in preparation. 29 H. B. Stuhrmann, G. Goerigk, B. Munk, in: Handbook of Synchrotron Radiation, Vol. 4, Chapter 17, Ed. S. Ebashi, M. Koch, E. Rubenstein. Elsevier, Amsterdam, 1991. 30 W. B. Russel, D. A. Saville, W. R. Schowalter, Colloidal Dispersions. Cambridge University Press, Cambridge, 1989. 31 G. K. Batchelor, J. Fluid Mech. 1977, 83, 97; J. F. Brady, M. Vicic, J. Rheol. 1995, 39, 545. 32 M. Ballauff, Macromol. Chem. Phys. 2003, 204, 220–234. 33 I. Deike, M. Ballauff, N. Willenbacher, A. Weiss, J. Rheol. 2001, 45, 709–720.

247

248

12 Spherical Polyelectrolyte Brushes 34 A. Marra, E. Pleuvrel-Disdier, A. Wittemann,

X. Guo, M. Ballauff, Colloid Polym. Sci. 2003, 281, 491–496. 35 S. P. Meeker, W. C. K. Poon, P. N. Pusey, Phys. Rev. E 1997, 55, 5718. 36 D. Distler (Ed.), Waessrige Polymerdispersionen. Wiley-VCH, New York, 1999. 37 M. Evers, T. Palberg, N. Dingenouts, M. Ballauff, H. Richter, Th. Schimmel, Prog. Colloid Polym. Sci. 2000, 115, 307. 38 M. Evers, H.-J. SchSpe, Th. Palberg, N. Dingenouts, M. Ballauff, J. Non-Cryst. Solids 2002, 307-310, 579–583. 39 Y. Mei, A. Wittemann, G. Sharma, M. Ballauff, Th. Koch, H. Gliemann, J. Horbach, Th. Schimmel, Macromolecules 2003, 36, 3452–3456.

40 Proteins at Interfaces II, Ed. T. A. Horbett,

41

42 43 44 45 46

J. L. Brash, ACS Symposium Series 602, American Chemical Society, Washington DC, 1995. E. Ostuni, R. G. Chapman, R. E. Holmlin, S. Takayama, G. M. Whitesides, Langmuir 2001, 17, 5605–5620; and further references gives therein. H. Kawaguchi, Prog. Polym. Sci. 2000, 25, 1171–1210. Y. Tran, P. Auroy, L.-T. Lee, M. Stamm, Phys. Rev. E 1999, 50, 6984–6990. A. Wittemann, B. Haupt, M. Ballauff, Phys. Chem. Chem. Phys. 2003, 5, 1671–1677. G. Jackler, A. Wittemann, M. Ballauff, C. Czeslik, Spectroscopy, in press. C. Fleck, H. H. von GrRnberg, Phys. Rev. E 2001, 63, 061804-1–061804-5.

249

13

Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up with Oppositely Charged Polyelectrolytes Rupert Konradi, Haining Zhang, Markus Biesalski, and Jrgen Rhe

Abstract

In this chapter, we describe recent investigations on the properties of weak poly(methacrylic acid) brushes and their interaction with low and high molecularweight counterions. The brushes were synthesized via surface-initiated free radical polymerization, using a self-assembled monolayer of azo initiators. Weak polyelectrolyte brushes are systems in which the degree of charges along a chain is variable and can be controlled by external parameters such as the pH value of the surrounding solution. We show that this offers the possibility to readily control the swelling of these brushes in a wide range of thicknesses. Even more dramatic effects were observed when the brush was brought into contact with different salt solutions. The fundamentally different behavior in the presence of monovalent alkaline, bivalent earth alkaline and bivalent transition metal cation solutions is described, and a classification of the different cations with respect to the type of interaction is given. Furthermore, high molecular-weight ions (i.e., polyelectrolytes) were used as counterions. The complex formation behavior of the weak polyacid brush is compared to that of a strong polyelectrolyte brush interacting with an oppositely charged strong polyelectrolyte. Both systems were employed for the build-up of polyelectrolyte multilayer assemblies through a layer-by-layer deposition process. For weak polyelectrolyte brushes, a strong templating effect of the very first brush layer was observed, and the thickness of each subsequently adsorbed layer was closely related to the thickness of the initial brush.

13.1

Introduction

Weak polyelectrolyte (PEL) brushes undergo substantial changes in swelling and contraction in response to a varying environment, such as a change in pH or ionic strength, making them excellent candidates for “smart” surfaces that respond to specific stimuli (see Figure 13.1) [1–4]. Furthermore, these systems can serve as a basic building block for highly sophisticated structures such as templates for protein or cell adhesion or the build-up of

250

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

Schematic depiction of the structure of weak polyelectrolyte brushes in various environments.

Figure 13.1

polyelectrolyte multilayers. In all of these aqueous systems, especially when dealing with biological environments, a mix of monovalent and multivalent ions as well as polyvalent biomolecules is usually present in solution. It is therefore crucial to understand the role that these ions play on the swelling behavior of the polyelectrolyte brush. In the following, the synthesis of the weak polyelectrolyte brushes and the experimental set-up for the determination of the swollen layer volume fraction profiles will be briefly introduced. Detailed aspects of the synthesis and characterization of polyelectrolyte brushes have been recently reviewed [5,6]. Accordingly, only a brief summary is given at this point. Depending on the chemical structure of the monomer, both strong and weak polyelectrolytes can be generated using either one- or two-step approaches. As an example of a densely grafted brush consisting of strong polyelectrolyte molecules, positively charged poly(methyl-4-vinylpyridinium) brushes have been prepared following a two-step approach. Initially, a poly-4-vinylpyridine brush is prepared through surface-initiated polymerization, and charges are introduced as a result of a second, polymer-analogous quaternization step. In a free radical polymerization as described here, the grafting density of the neutral brush can be controlled by varying the conversion of the surface-attached initiator, which can be adjusted by varying the polymerization time. For the same system, the molecular weight of the surfaceattached polymer molecules can be determined by choosing an appropriate monomer concentration, increasing the radical concentration in the growing film or by enhancing chain transfer. By using this technique polyelectrolyte brushes can be prepared with high grafting densities (i.e., distances d between neighboring chains of d » 2–3 nm) and high molecular weights of the attached chains (Mn > 106). However, one-step syntheses of polyelectrolyte brushes can also be performed. The poly(methacrylic acid) (PMAA) brushes described below can be prepared by using the same immobilized azo-initiator monolayer and a surface-initiated poly-

13.2 Synthesis and Data Evaluation

merization from the surface in situ [7–9]. Methacrylic acid can be directly used in the free radical polymerization, and a second step such as the polymer-analogous quaternization described above is not needed. Again, the grafting density and the molar mass of the polyelectrolyte chains can be controlled independently from each other by adjusting the polymerization parameters. In the following, the swelling behavior of PMAA brushes in aqueous media in the presence of different kinds of salts will be discussed. Additionally, the complex formation of polyelectrolyte brushes with oppositely charged polyelectrolytes in solution and the multilayer build-up based on these systems will be described.

13.2

Synthesis and Data Evaluation 13.2.1

Synthesis

PMAA brushes were synthesized via the “grafting from” technique [7–9], as depicted in Figure 13.2. In the first step, a self-assembled monolayer of an azo initiator is covalently linked to the surface of either silicon wafers or glass prisms (LASFN9 glass). The second step comprises of a surface-initiated, free radical chain polymerization where the polymer brush is grown in situ from the surface.

Schematic depiction of the “grafting-from” free radical polymerization, using a self-assembled layer of azo initiators.

Figure 13.2

251

252

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

13.2.2

Multiple-Angle Nulling Ellipsometry

The swelling behavior of the brushes was investigated using a homebuilt multipleangle nulling ellipsometry set-up, as depicted in Figure 13.3. The volume fraction profiles were modeled by using a complementary error function [10,11]. Figure 13.4 compares typical ellipsometric spectra and the corresponding model fits (D only) for

Figure 13.3

Schematic depiction of the multiple-angle nulling ellipsometry set-up.

Representative ellipsometric spectra (D as a function of the incidence angle; W is omitted for clarity) of PMAA brushes on LaSFN9-prisms swollen in aqueous solutions of NaNO3 and Ca(NO3)2. The concentrations

Figure 13.4

are 10–5, 10–4, 10–3, 10–2, 10–1 and 100 mol L–1 top down, and the offset in D is –20, respectively. The solid lines represent model calculations using a complementary error function to describe the segment density profile.

13.3 Swelling Behavior of Weak Polyelectrolyte Brushes in Aqueous Environments

Representative segment density profiles of the collapse of a PMAA brush on a LaSFN9-prism as obtained from model calculations based on the ellipsometric spectra (Figure 13.4). The brush was swollen in aqueous

Figure 13.5

solutions of Ca(NO3)2 having different external salt contents as denoted in the figure. The inset figure shows the profiles at high values of j.

a PMAA brush in aqueous sodium and calcium solutions, respectively. Figure 13.5 shows representative segment density profiles of the PMAA brush obtained from the model fits to the ellipsometric spectra in Figure 13.4 for the calcium case. In the following, twice the first moment of the segment density profiles is taken as a measure for the swollen brush thickness.

13.3

Swelling Behavior of Weak Polyelectrolyte Brushes in Aqueous Environments

In this section, we will introduce the swelling behavior of PMAA brushes in aqueous media. First, the pH-dependence of the swelling of the brushes is discussed, after which the presence of monovalent cations is investigated. Finally, divalent earth alkaline and transition metal cations as well as trivalent cations are addressed. The influence of electrostatic and specific interactions will be elucidated.

253

254

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

13.3.1

The Influence of pH Value

The dependence of the swollen brush height on the pH of the solution without any additional salt present is illustrated in Figure 13.6 [1]. The brush is already in a swollen state at the lowest pH value studied. Lower pH values could not be studied, as under these conditions the lanthanum glass (which is employed as a substrate) begins to dissolve. As expected, the brush thickness increases with increasing pH due to an increase in the number of dissociated carboxylic acid groups on the attached polymer chains – that is, the charge density along the polymer chains increases. A higher charge density leads to a stronger electrostatic repulsion between the charged segments and a higher osmotic pressure of the counterions within the brush along with an increase of the swollen thickness of the brush. The smooth transition can be explained by a variable pKa-value along the segment density profile. It is known for weak polyelectrolytes in solution, that the pKa-value decreases with increasing polymer concentration [12–14]. Since the volume fraction profile of PMAA brushes was not found to be box-like, the segment concentration varies along the z-axis, and accordingly the pKa-value is also variable within the brush.

Swollen thickness of a 20 nm (dry thickness) PMAA brush as a function of the pH of the solution without additional salt. The solid line is a guide to the eye.

Figure 13.6

13.3.2

Interaction with Monovalent Cations

The swelling behavior of a PMAA brush in aqueous solutions of sodium and silver nitrate at neutral pH is compared in Figure 13.7. After swelling in pure water, the brush was first treated with increasingly concentrated silver ion solutions up to a maximum value of 1 M. The silver ion concentration was then gradually decreased, and finally exchanged again for pure water. The same brush was then equilibrated

13.3 Swelling Behavior of Weak Polyelectrolyte Brushes in Aqueous Environments

Swollen thickness of a 42 nm (dry thickness) PMAA brush as a function of the external salt concentration. The brush was swollen in aqueous solutions of sodium (d) and silver (~) nitrate at pH 7. In the case of silver nitrate, the swollen thickness is shown

Figure 13.7

with increasing (~) and decreasing (.) silver nitrate concentration, respectively. The dotted line at low salt concentration (slope m = +1/3) and the solid line at high salt concentration (slope m = –1/3) represent the expected theoretical scaling behavior [15–18].

with sodium ion solutions of increasing concentrations. The difference in the influence of the two monovalent counterions on the swelling behavior of the brush is therefore directly visible. 13.3.2.1 Sodium Ions With respect to sodium ions, the brush thickness passes through a maximum at 3 – 1 mmol L–1. At sufficiently low salt concentrations – the so-called osmotic brush regime – the thickness increases with increasing salt concentration. Here, the protons are confined within the brush in order to keep the surface-attached polymer layer electroneutral. The addition of salt facilitates the dissociation of the acid groups and leads to an increase of the degree of ionization. With increasing charge density along the chains, electrostatic repulsion leads to an increase in the swelling of the brush. However, above a critical ionic strength (in the salted brush regime) the brush thickness decreases with increasing salt concentration due to an increased screening of the charged groups. Theoretical studies based purely on ionic interactions describe qualitatively the behavior more or less correctly [15–19]. The predicted scaling laws are also given in Figure 13.7. In the osmotic brush regime, the brush thickness is expected to increase with the ion concentration with a slope m = +1/3 (dotted line). In the salted brush regime, the brush thickness is expected to decrease with the ion concentration with a slope m = –1/3. The experimental data indicate smaller exponents that may be attributed to the nontrivial segment density profile and the polydispersity of the real brush system. Furthermore, the steric repulsion between the monomeric units was neglected in the scaling approach. However, it has been shown in previous investigations that the influence of such monovalent

255

256

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

salts on the swelling behavior strongly depends on the grafting density of the system [20]. The results for the swelling of such brushes are qualitatively in good agreement with those of other groups on weak polyelectrolyte brushes in different topologies. Currie et al. performed ellipsometric measurements on poly(acrylic acid) brushes that were prepared by an adsorption of poly(styrene)-block-poly(acrylic acid) copolymers on hydrophobically modified Si wafers [21]. These authors also found a maximum in the swollen brush thickness with increasing sodium ion concentration, and the scaling exponent in the osmotic brush regime was determined to be significantly lower than +1/3. Ballauff et al. examined the swelling behavior of spherical poly(acrylic acid) brushes attached to poly(styrene) latex cores [3,4,22]. These brushes were prepared by a grafting from method similar to the one described above [3]. (A detailed description of the results is given by the authors in Chapter 12 of this book.) When the brush is thin compared to the radius of the core particle, the system is expected to behave similarly to a planar brush. Indeed, the authors could show, that the swelling of such a brush reveals a nonmonotonic trend with increasing sodium ion concentration [4]. Cosgrove et al. have synthesized weak polybase brushes through the incorporation of one block of a diblock copolymer inside a latex particle [23]. Similar to the polyacid brushes discussed above, the addition of increasing amounts of sodium ions lead to a maximum in the swollen brush thickness. Silver Ions Silver ions were found to demonstrate a completely different behavior, despite the fact that they are also monovalent. The maximum in the swollen brush thickness is less pronounced and shifted to lower ion concentrations (approx. 0.5 mmol L–1). Further increasing the silver ion concentration up to 1 mol L–1, a strong shrinkage of the brush was observed (Figure 13.7). With decreasing ion concentration, the reswelling occurred only at concentrations below 1 mmol L–1. Still, the silver ions could be completely removed and the initial swollen brush thickness in pure water recovered. The theoretical studies that qualitatively match the experimental findings with regard to the swelling behavior of the PMAA brush in the presence of sodium ions seem to fail when the monovalent alkaline metal sodium is exchanged for silver, a monovalent transition metal. Since the theoretical work is based on ionic interactions alone, no distinction is made between different monovalent salts. Although the experimental results are not yet fully understood, we believe that the difference can be attributed to specific interactions between the carboxylate groups and the silver ions. As the silver ion loading is raised, the methacrylate groups become increasingly hydrophobic due to charge recombination and concurrent dehydration. This leads to an increase of the hydrophobic interactions within the brush, and eventually to a shrinkage of the brush. On decreasing the silver ion concentration in the environment, the hydrophobic interactions can only be overcome at relatively low silver ion concentrations. Such a behavior could explain the observed hysteresis in the swollen brush thickness with respect to the silver ion concentration. Specific interaction of silver ions with poly(acrylic acid) (PAA) and a dehydration effect have also been reported for other systems. Ikegami and Imai investigated the precipitation of 13.3.2.2

13.3 Swelling Behavior of Weak Polyelectrolyte Brushes in Aqueous Environments

bulk PAA on addition of different types of ions [24], while Pohl and Kuhn observed a PAA gel shrinkage upon addition of silver ions with a concurrent change in the UV spectrum [25]. 13.3.3

Interaction with Divalent Cations

We will now go a step further and focus on the interaction of PMAA brushes with bivalent cations. Figure 13.8 shows the swelling behavior of a PMAA brush in aqueous solutions of alkaline earth metal nitrates and copper nitrate at neutral pH. Alkaline Earth Metal Ions PMAA brushes in contact with the different alkaline earth metal ions reveal a similar behavior: the brush exhibits a collapse at intermediate concentrations (around 10–3 mol L–1) and (compared to the monovalent ion case described above) displays no maximum in the brush thickness. Once the brush is collapsed in concentrated alkaline earth metal salt solutions, a reswelling with water is impossible. However, the brush can be reswollen upon multiple exchange of the medium for 0.1 M sodium nitrate solutions. After replacing the sodium solution with water, the initial swollen thickness was obtained and the influence of another counterion could be studied on the same brush. The equilibrium structure of an annealed brush interacting with multivalent cations was theoretically analyzed by Zhulina and Birshtein using a scaling model [26,27]. In these studies, the authors assumed that the salt ions do not bind to the charged groups on the chains – that is, the electrostatic interactions would domi13.3.3.1

Swollen thickness of a 46 nm (dry thickness) PMAA brush as a function of the external salt concentration. The brush was swollen in aqueous solutions of magnesium (j), calcium (d), strontium (r), barium (w) and copper (~) nitrate at pH 7. Figure 13.8

257

258

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

nate. They found that the valency of the salt counterion affects the exponents in the scaling dependences of the degree of neutralization and the swollen brush thickness on the salt concentration. A higher valency is expected to provide a weaker dependency of the brush thickness on the salt concentration in comparison to a monovalent salt; however, a maximum is still expected and should be shifted to lower salt concentrations. This theoretically predicted behavior does not describe our experimental observations where we observe a collapse and no maximum in the swollen brush thickness at intermediate concentrations. The scaling approach, however, does not account for specific interactions between the brushes and the ions and concurrent changes in the solubility of the system. The fact that polycarboxylates show a specific response when exposed to different cations has been known for some time for systems with quite a variety of different topologies [24,28–41]. It is generally accepted that a partial dehydration of both, the polyanion and the cations leads to an increased hydrophobicity of the polymer. Accordingly, hydrophobic interactions within the polymer increase and the solvent power decreases resulting in a shrinkage of the polymer. Ikegami and Imai have carried out refractive index measurements on PAA solutions in the presence of multivalent cations including alkaline earth ions [24,29]. They concluded, that the binding of these ions destroys the first “intrinsic” hydration region of the polyanion leading to partial dehydration. Huber et al. performed light-scattering measurements on a PAA solution in the presence of calcium ions [36]. These authors also explained the reduction of the coil dimensions and the characteristic precipitation behavior with an increasing hydrophobicity of the chains. Finally, Horkay et al. investigated the interaction of fully neutralized PAA gels with a variety of cations [38–40]. For sodium, only a slight contraction due to screening effects was observed. However, with divalent cations at a critical concentration, a sudden volume transition occurred. The critical concentration depends on the type of cation, and is lower for transition metal cations than for alkaline earth metal cations. It was found that the mixing free energy was not altered in the presence of sodium, but was changed in the presence of divalent cations. This strongly suggests specific interactions of the latter ions with the gel, whereas sodium – as expected – does not interact specifically. The authors found a distinct difference in the behavior of alkaline earth metal ions and divalent transition metal cations, such as copper, as well as trivalent cations: only the latter ions were causing a change in the elastic properties of the gel. Therefore, for the binding of alkaline earth metal ions, the authors suggested the promotion of a weak aggregation of the chains in the swollen gel, but excluded the formation of additional more permanent crosslinks. De la Cruz et al. proposed an equilibrium between the formation of monocomplexes (mc) and dicomplexes (dc) in solutions of polyacrylate and divalent cations where the dicomplex is favored at high degrees of neutralization [42]. On the basis of UV measurements, Delsanti et al. found the dicomplex to be predominant in the case of cobalt ions (85% dc), and attributed this as the reason for the precipitation, due to its hydrophobicity, and for an absence of a resolubilization, which they would expect if the more hydrophilic monocomplex was dominant [31]. Furthermore, they supposed that dicomplexation had occurred between neighboring monomers.

13.3 Swelling Behavior of Weak Polyelectrolyte Brushes in Aqueous Environments

Subtle differences in the behavior of the various alkaline earth metals can be seen in Figure 13.8. The degree of dissociation within the weak polyacid brush is not constant, but increases with the distance from the surface [15–17]. Therefore, the equilibrium between mono- and dicomplexes is expected to change normal to the surface. The exact position of the equilibrium as well as the hydrophilicity of the monoand dicomplexes, respectively, probably depends on the exact type of cation present. These effects may account for the differences in the behavior of the various alkaline earth metal salts. Especially interesting in this respect is the behavior of the brush at high salt concentrations. Whereas magnesium shows no influence on the degree of swelling, the addition of high salt concentrations leads to a reswelling for calcium (approx. 20%), and this was even more pronounced for strontium (approx. 60%). Although the experimental technique used here to determine swollen thickness is not very sensitive in this regime, the changes in the ellipsometric spectra are clearly visible. Unfortunately, if barium is used, the concentration region is not experimentally accessible, as the barium salts are insoluble in water at this concentration. A simple mass uptake from salt penetrating into the brush could be considered to be responsible for this effect. This explanation would be in accordance with the finding that the reswelling increases with the ion mass. However, the increase in dry layer thicknesses for all alkaline earth metals was found to be only 14 € 2%, and it therefore seems arguable whether a simple mass uptake can fully account for these observations. On the other hand, a high concentration of cations could shift the equilibrium between mono- and dicomplexes (2 mc <=> dc + free ion) towards the monocomplexes, leading to an increased hydrophilicity of the chains and to reswelling of the brush. The finding that reswelling increases with increasing size of the alkaline earth metal ion reflects a decrease in the complex formation constant for the dimeric complex with increasing ion size. In bulk polycarboxylate systems, a precipitation of the polymer by divalent cations is usually found at concentrations slightly below the isoelectric point [35]. One could question whether the collapse concentration in the PMAA brush system also corresponds to an equimolar concentration of alkaline earth metal cations and carboxylate groups within the brush. Figure 13.9 shows infrared spectra of a PMAA brush on both sides of a silicon wafer that was treated with calcium nitrate solutions of increasing concentration. The carbonyl absorption bands at 1705 cm–1 and 1560 cm–1 correspond to the protonated carboxylic acid group and the antisymmetric stretching mode masym of the deprotonated carboxylate group, respectively. In pure water (at pH 7), only the protonated species is present, but with increasing calcium ion concentration the protons are exchanged for calcium. This exchange appears in the spectra as an increase in intensity of the absorption band due to the deprotonated carboxylate group. Two important observations can be made: first, the increase in complex formation is monotonic and smooth; and second, the brush is not fully deprotonated at concentrations where the collapse occurs. However, this implies that the collapse is not related to an abrupt change in the calcium concentration in the brush, but rather that the solvent quality becomes increasingly worse until it finally becomes a non-solvent and the brush collapses. This again can be explained with an increasing hydrophobicity of the polymer layer.

259

260

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

FT-IR spectra of a PMAA brush on both sides of a silicon wafer. The brush was treated with calcium nitrate solutions of increasing concentration (pure water, 10–4, 10–3, 10–2, 10–1 and 100 mol L–1 from top to Figure 13.9

bottom). The absorption bands at 1705 cm–1 and at 1560 cm–1 correspond to the protonated carboxylic acid group and masym of the deprotonated carboxylate group, respectively.

A remarkably different behavior compared to the same polymers in solution is evident. The collapse concentration in the brush system does not correspond to that of the precipitation phenomenon of the same polymer in solution. In contrast to the bulk polymer system, where precipitation occurs at approximately equimolar amounts of positive charges and carboxylic acid groups, the collapse in the brush system occurs at a much lower ratio. Comparable observations were made by Horkay et al. for gels. These authors found that at a “critical concentration” of divalent cations, a sudden volume change of a PAA-Na gel occurs. Furthermore, the concentrations of both monovalent and divalent cations within the gel vary continuously and smoothly, despite the discontinuous change in the gel volume. 13.3.3.2 Copper Ions The PMAA brush in contact with copper nitrate solutions shows a fundamentally different behavior as compared to the interaction with alkaline earth metal cations (Figure 13.8). The brush already collapses at extremely low concentrations (<10–5 mol L–1). The swollen brush thickness then stays constant with increasing copper ion concentration, and remains elevated in comparison to the alkaline earth case. Horkay et al. also found a strongly different behavior when they studied the interaction of a PAA gel with alkaline earth and transition metal cations. The authors were able to separate contributions to the free energy from the mixing pressure and the elastic modulus. They have shown that alkali metal cations do not affect either

13.3 Swelling Behavior of Weak Polyelectrolyte Brushes in Aqueous Environments

of these terms, and that alkaline earth metals affect the mixing free energy but not the elastic modulus, whereas bivalent transition metal cations impinge on both of these terms. These findings strongly suggest that transition metal cations form bridges between the network chains. Bridges are also likely to be the reason for the early collapse in the PMAA brush system. In order to obtain further information in that respect, we have carried out infrared spectroscopic investigations on PMAA brushes in contact with a number of different metal nitrate salts at 0.1 M concentrations (Figure 13.10). Of special interest is the carbonyl absorption region, which is depicted in the enlargement of Figure 13.10. All brushes show the absorption bands around 1705 cm–1 and 1550 cm–1. The former of these is usually assigned to the carboxylic acid group, and the latter to a chelating bidentate complexation of the carboxylate group (Figure 13.11(a)) [43]. In the case of copper and aluminum complexation, a third absorption band emerges in the carbonyl region around 1620 cm–1; this band has been ascribed to a bridging bidentate complexation (Figure 13.11(b)), and provides further evidence for the formation of bridges between the carboxylate groups along the chains [44–46]. The structure of both, monomeric carboxylate and polycarboxylate copper complexes has been determined [34]. When the carboxylate binds as a chelating bidentate ligand, mononuclear complexes with one or two carboxylates bound to the metal cation are formed. The dicomplex has D2h symmetry with both carboxylate ions and the central ion lying in the symmetry plane (Figure 13.11(a)). In the case of a bridging bidentate binding, a binuclear complex is formed through the association of two copper ions with two pairs of adjacent carboxylate groups. In this dimer, the copper ions are bridged by four carboxylate groups, forming a structure of approximately D4h symmetry with two additional solvent ligands on the C4 axis. The symmetry cen-

Figure 13.10 FT-IR spectra of a 232 nm (dry thickness) PMAA brush on both sides of a silicon wafer. The brush was cut into 1-cm2 pieces and treated with 0.1 M solutions of Mg(NO3)2, Ca(NO3)2, Sr(NO3)2, Ba(NO3)2, Cu(NO3)2, Al(NO3)3 and AgNO3. The enlargement shows the carbonyl absorption region.

261

262

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

Figure 13.11 Structures of metal ion carboxylate complexes. The chelating bidentate and the bridging bidentate coordinations together with examples for the copper ion complexes are shown in (a) and (b), respectively.

ter is halfway between the metal ions, and the carbon atoms of the carboxylate groups lie in the symmetry plane (Figure 13.11(b)). At concentrations above the critical collapse concentration, the PMAA brush shows a higher degree of swelling when the collapse was induced by copper ions than by alkaline earth metal ions. This could be due to the increased steric demands of the binuclear complex causing a loose arrangement of the polymer chains in order to place the carboxylate units in the required positions. Furthermore, this complex could be more hydrophilic in comparison to the chelating bidentate complexes, which becomes plausible if one considers the two additional ligand positions on the C4 axis that are occupied by water molecules. So far, we have elucidated the different mechanisms that lead to the collapse of a PMAA brush in contact with alkaline earth metal and copper ion solutions, respectively. In the former case, a hydrophobization of the polymer was found to be the reason for the collapse, while in the latter case the collapse was ascribed to a bridging of the polymer chains. To prove that the brush swollen thickness is indeed highly sensitive to the bridging phenomenon, an experiment was carried out under steady-flow conditions (Figure 13.12). A PMAA brush was treated with a copper nitrate solution of extremely low concentration (10–6 mol L–1) and the D-value at a fixed angle was recorded as a function of time. This angle was chosen such that the change in D upon a collapse of the brush would be maximal. In addition, angular scans before and after adsorption of the copper ions were performed, and these are shown as insets in Figure 13.12. D immediately starts to fall when the brush is exposed to the copper ion solution, but then decreases over a time period of approximately 8 h, and the angular scans show that the swelling of the brush is drastically reduced during the course of the experiment. At the given ion concentration, the number of ions in the cell is less than the number of surface-bound carboxylic acid groups by a factor of more than 10. This indicates that, under these conditions, cop-

13.3 Swelling Behavior of Weak Polyelectrolyte Brushes in Aqueous Environments

Figure 13.12 Collapse kinetics of a 26 nm (dry thickness) PMAA brush in 10–6 M Cu(NO3)2 solution under steady-flow conditions at pH 7. The evolution of D (s) at 45.54@ (inner angle) is shown as a function of time. The solid line is a guide to the eye. The insets show the D trace (s) of angular scans before the addition of

copper where the brush in swollen in pure water and after 9.3 h of flow measurement, where the brush thickness has reached an equilibrium. The model calculations (solid lines) and the swollen brush thicknesses are also given in the insets.

per ions bind almost irreversibly to the brush, and the brush essentially collects all available ions out of the solution. This extreme sensitivity of brush swelling towards copper ions is unlikely to be due only to a reduction in hydrophilicity. This confirms the suggestion that bridging of the polymer chains by the copper ions is responsible for the strong deswelling of the brush. 13.3.4

Interaction with a Trivalent Cation: Aluminum

As with copper ions, the interaction of a PMAA brush with aluminum ions leads to a collapse at extremely low concentrations (<10–5 mol L–1) (Figure 13.13). Again, the swollen brush thickness remains unaffected by a further increase in the ion concentration, and the degree of swelling is also higher when compared to the alkaline earth metal ion case. As mentioned above, the similarity to the copper ion situation is also found in the infrared spectra of PMAA brushes treated with aluminum and copper nitrate solutions, respectively (see Figure 13.10). Both spectra show an absorption band in the carbonyl region around 1620 cm–1, which is absent for complexes with alkaline earth metal cations and is usually assigned to a bridging bidentate conformation [44,45].

263

264

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

Figure 13.13 Swollen thickness of PMAA brushes as a function of the external salt concentration. The brushes were swollen in aqueous solutions of sodium nitrate (w; dry thickness 42 nm), calcium nitrate (d; dry thickness 46 nm) and aluminum nitrate (j; dry thickness 45 nm) at pH 7.

We therefore believe that the behavior of the PMAA brush in contact with aluminum ion solutions can again be explained by a bridging of the polymer chains through the metal cations. 13.3.5

A Classification

To conclude the study on the interaction of weak polyelectrolyte brushes with low molecular-weight ions, we summarize the results discussed in Sections 13.3.1– 13.3.4 in a classification of the ions with respect to their interaction with PMAA brushes. We can differentiate between three types of interaction, and representatives of each class are shown in Figure 13.13. Mainly Ionic Interactions In this first class, ionic interactions between the ion and the carboxylate groups dominate. Typically, the swollen brush thickness passes through a maximum with increasing salt concentration (Figure 13.13). However, the influence of salt addition is fully reversible – that is, when exposed to low salt concentrations/salt-free conditions, the brush returns to its original height. These experimental observations are qualitatively well understood from a theoretical point of view [15–19]. Sodium and other alkaline metal ions are good representatives of this group. 13.3.5.1

13.3.5.2

Specific Interactions

Dehydration

Specific interactions determine the behavior of the swollen brush in this second class. Typically, the swollen brush thickness undergoes a collapse at intermediate

13.4 Interaction Between Polyelectrolyte Brushes and Oppositely Charged Polyelectrolytes in Solution

concentrations of this type of ion. The collapse at a critical ion concentration is due to a dehydration of the polyelectrolyte. The brush becomes less hydrophilic and, eventually, the solvent power of the aqueous solution with respect to the brush is strongly reduced. The collapse of the brush is not reversible – that is, the brush remains in the collapsed state if re-exposed to low salt concentrations. Typical representatives of this group are the alkaline earth metal ions as well as silver ions. The collapse of the brush in the presence of calcium ions is compared to the sodium case in Figure 13.13. Bridging

The third class is characterized by a bridging of the carboxylate groups through the ions via strong complexation. The brush typically already collapses at trace concentrations of these ions, although the remaining swelling is greater than for the brush collapsed by ions of the second class. Copper and aluminum ions have been discussed as representatives of this class. The evolution of the swollen brush thickness with increasing concentrations of aluminum ions is also shown in Figure 13.13 and compared to the behavior in the presence of sodium and calcium ions as representatives of the first and second classes, respectively.

13.4

Interaction Between Polyelectrolyte Brushes and Oppositely Charged Polyelectrolytes in Solution

In the next section, we will expand the study to high molecular-weight counterions, and describe some recent results on the formation of surface-attached PEL-PEL complexes by using polyelectrolyte brushes as substrates [47–49]. In particular, the differences between strong and weak polyelectrolyte systems will be discussed. In the following processes, we classify our systems as either a “strong system”, in which two strong PEL are used, and a “weak system”, in which at least one weak PEL is used. The surface-attached polyelectrolyte complexes obtained are used for the further layer-by-layer build-up of polyelectrolyte multilayers. 13.4.1

The Formation of Surface-Attached PEL-PEL Complexes

For the formation of PEL-PEL complexes in solution, the process is rather rapid as no diffusion barrier exists, and more or less stoichiometric complexes are obtained [50]. To investigate how fast the formation of PEL-PEL complexes via polymer brushes proceeds, ATR multi-angle null ellipsometry was applied to different systems (MePVP/PSSNa and PMAA/MePVP). The thickness of the film in contact with an oppositely charged polyelectrolyte solution as a function of exposure time for those two systems is shown in Figure 13.14. In the MePVP/PSSNa case, while the thickness of the 13 nm (dry) MePVP brush in pure water is about 660 nm (degree

265

266

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

a) strong system

b) weak system

Figure 13.14 The influence of the exposure time on the layer thickness of (a) a 13-nm MePVP brush in contact with 2 mM PSSNa solution; and (b) a 30-nm PMAA brush in contact with 2 mM MePVP solution. Dashed lines represent the dry thickness of the brush.

of swelling dswell/ddry » 50), immediately after addition of the PSS solution the brush height collapses to about 19 nm, which is roughly the expected value for a dry double layer. In a similar system, when a weak PMAA brush (30 nm, dry state) was immersed into MePVP solution, the layer shrinks slowly from 750 nm to 170 nm in 1 h, but thereafter the layer thickness stays constant (Figure 13.14(b)). The difference between the two systems shows that the MePVP/PSSNa complex does not swell in water, whereas the complex of PMAA/MePVP remains strongly swollen under the same conditions. Another important question is how the thickness of the surface-attached brush influences the layer thickness of the oppositely charged PEL for different systems. To answer this question, a set of experiments has been carried out in which the brush thickness was varied from 5 nm to roughly 200 nm by adjusting the polymerization time, which controls conversion of the initiator and accordingly the grafting

13.4 Interaction Between Polyelectrolyte Brushes and Oppositely Charged Polyelectrolytes in Solution

Figure 13.15 Layer thickness increase due to formation of a PEL-PEL complex during exposure of a MePVP brush to a PSSNa solution (d) or a PMAA solution (j). The lines are guides to the eye. The inset shows an enlargement for the MePVP brush/PSSNa system.

density of the polymer monolayer. The dependence of the adsorbed amount of polyanions on the polycation (MePVP) brush thickness is shown in Figure 13.15. It is evident that adsorption of PMAA onto the MePVP brush is strongly influenced by the initial brush thickness, whereas the adsorption of PSSNa is almost independent of the thickness of the substrate brush. Whilst in the latter case the brush thickness was varied by a factor of about 20, the concurrent change of the thickness of the adsorbed PSSNa layer was only about 1.5-fold. In contrast, if the weak system is viewed, the thickness of the adsorbed PMAA layer is more or less identical to the brush thickness to which it was adsorbed. This difference in the overall layer stoichiometry can be understood if the swelling of the layers in contact with the oppositely charged polyelectrolyte is studied with multiple-angle ellipsometry. Whereas the strong/weak PEL system remains swollen throughout the absorption experiment,

Figure 13.16 Ratio of the adsorbed PEL layer thickness and initial brush thickness as a function of the grafting density of the brush calculated from the initiator decomposition kinetics [10]. The dashed line represents the 1:1 stoichiometry of the complex.

267

268

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

the brush in the strong/strong system collapses immediately upon addition of the oppositely charged PEL. Similar results have been investigated for the formation of weak PEL complexes in different systems, for example if a strong PEL brush/weak free polyelectrolyte system [47] or a weak PEL brush/weak free polyelectrolyte system [48] are studied, as summarized in Figure 13.16. The results obtained indicate that the adsorption of PEL molecules from solution onto PEL brushes to form weak PEL-PEL complexes depends solely on the swelling properties of the surface-attached PEL-PEL complex, and is independent of other details of the system – that is, whether the weak PEL is the brush component or the component coming in from solution. 13.4.2

The Formation of PEL Multilayer Assemblies

Apart from the formation of ultra thin surface-attached PEL-PEL complexes, it is very interesting whether the PEL brushes can also be used for the formation of PEL multilayer assemblies. The so-called layer-by-layer (LBL) technique is a simple and powerful method to form well-defined multilayered structures [51]. For the formation of these multilayer assemblies, the brushes are alternately dipped into polyelectrolyte solutions, one of which consists of a positively charged polyelectrolyte, and the other of a negatively charged polyelectrolyte. It is usually assumed that, in this LBL deposition process, the driving force for each monolayer formation is charge overcompensation [52,53]. The stability of the entire multilayered assembly formed by the LBL process using a charged surface as the substrate in different environments is one of the limitations for this approach. Since the attachment of the first layer depends solely on the interaction of the polymers with surface charges, the whole multilayer assembly can be desorbed by either changing the sign of the surface charge of the substrate or by addition of competing low molecular-weight elec-

a) strong system

Figure 13.17 Film thickness as a function of the layer numbers for: (a) MePVP/PSSNa multilayers using 7.6 nm (j) and 22.9 nm (w) MePVP covalently attached monolayers as the first layer: and (b) PMAA/MePVP multilayers

b) weak system

using 6 nm (j) and 30 nm (d) covalently attached monolayers as the first layer. The solid lines show a linear fit of the dependence of the film thickness on the number of deposited layers.

13.4 Interaction Between Polyelectrolyte Brushes and Oppositely Charged Polyelectrolytes in Solution

trolytes, which can displace the polymer molecules in the first monolayer [53,54]. Here, covalently attached polymer brushes could be advantageous. Another general problem of the LBL method is the low thickness of each single deposited layer, which is on the order of 0.5 nm [51]. Such a small increase in film thickness per deposited layer is rather inconvenient if a thicker PEL multilayer assembly is desired, as in this case many layers must be deposited. The deposition of a MePVP/PSSNa multilayer system and a PMAA/MePVP multilayer system were carried out, using systems as described above. The results of the study are shown in Figure 13.17(a) and (b). From Figure 13.17(a), it can be seen that the increase in thickness due to absorption of the second layer is larger than that of the following layers deposited onto it. However, from the third layer onwards the thickness increases by only about 0.5 nm per deposition cycle (i.e., deposition of two monolayers), and a linear relationship between layer thickness and the number of deposited layers is observed. In this system, after deposition of a total of three or four PEL-layers the brush no longer affects the properties of the outer layers. It is also clearly visible for the PMAA/MePVP system (Figure 13.17(b)) that the film thickness increases linearly with the number of dipping cycles. However, it is also clear that the thickness of each layer in the multilayer assembly depends heavily on the thickness of the initial brush layer. The increase in layer thickness per deposition cycle is more or less identical to that of the thickness of the initial brush monolayer, and even when a very thick brush is used the outermost layer resembles closely the innermost (the brush layer), and a very strong template effect is observed. Even though the overall architecture of the two systems is very similar (surfaceattached PEL brush with electrostatically attached monolayers of alternating charge sign), the film-formation behavior of the two systems is very different, as shown schematically in Figure 13.18. Although this difference is not yet fully understood, it is evident that the basic difference between the systems is the water solubility of the

Figure 13.18 Schematic depiction of the formation of PEL multilayers through PEL brushes. (A) strong/strong system; (B) strong/weak system. No implication is made about phase boundaries between the different layers and interface roughness.

269

270

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up

PEL-PEL complex formed. This difference is directly evident if solutions of the two polyelectrolytes are mixed [55]. While in the first (weak/weak) case the complex remains soluble if solutions containing equimolar amounts of polyanion and polycation are mixed, in the latter (strong/strong) case immediate precipitation occurs due to the hydrophobicity of the neutral complex formed.

Summary

Polyelectrolyte brushes can be used as substrates for the adsorption of polyelectrolytes from solution to form PEL-PEL complexes attached to solid surfaces, and for the subsequent build-up of weak polyelectrolyte multilayers by a LBL deposition process. Because the polyelectrolyte monolayers generated are directly in contact with the surface and are attached to it through chemical bonds, the PEL complexes and the PEL multilayer could, potentially, be very stable. Another advantage of this method is that it allows for fine control of the surface properties of the substrate. The formation of weak PEL-PEL complexes and multilayers at solid surfaces via polymer brushes is quite different from the case where two strong polyelectrolytes are used. Interestingly, in the strong/weak PEL system, each adsorbed layer has the same thickness, which closely resembles that of the innermost brush layer. By contrast, in the strong/strong PEL system the formation of PEL multilayers resembles more the traditional LBL technique in so far as the increase in layer thickness per deposition cycle is well below 1 nm. In the strong/weak case, however, more than 100 nm of polyelectrolyte can easily be deposited per dipping cycle, and this allows the simple generation of very thick PEL multilayer assemblies. However, the exact mechanism and the internal structures of the weak PEL complexes and multilayer assemblies are still not clear. Since many biopolymers (proteins, nucleic acids and DNA) are polyelectrolytes, the systems described here are interesting models for biomaterials. In future investigations, it should be worthwhile to study the internal structure of these layers in more detail using different techniques, including for example neutron reflectivity and fluorescence microscopy.

References

References 1 M. Biesalski, D. Johannsmann, J. RRhe, 2

3 4 5

6 7 8 9 10 11 12

13 14 15 16 17

18 19 20 21

22 23

J. Chem. Phys. 2002, 117, 4988–4994. R. Konradi, J. RRhe, Polymer Preprints (Am. Chem. Soc., Div. Polym. Chem.) 2003, 44, 546–547. X. Guo, A. Weiss, M. Ballauff, Macromolecules 1999, 32, 6043–6046. X. Guo, M. Ballauff, Phys. Rev. E 2001, 64, 051406. M. Biesalski, J. RRhe, R. KRgler, W. Knoll, Handbook of Polyelectrolytes and Their Applications 2002, 1, 39–63. J. RRhe, N. Knoll, J. Macromolec. Sci. – Polym. Rev. 2002, C42, 91–138. O. Prucker, J. RRhe, Macromolecules 1998, 31, 592–601. O. Prucker, J. RRhe, Macromolecules 1998, 31, 602–613. O. Prucker, J. RRhe, Langmuir 1998, 14, 6893–6898. J. Habicht, M. Schmidt, J. RRhe, D. Johannsmann, Langmuir 1999, 15, 2460–2465. M. Biesalski, J. RRhe, D. Johannsmann, J. Chem. Phys. 1999, 111, 7029–7037. I. Borukhov, D. Andelman, R. Borrega, M. Cloitre, L. Leibler, H. Orland, J. Phys. Chem. B 2000, 104, 11027–11034. J. C. Leyte, M. Mandel, J. Polym. Sci. 1964, Pt. A 2, 1879–1891. M. Mandel, Eur. Polym. J. 1970, 6, 807–822. R. Israels, F. A. M. Leermakers, G. J. Fleer, Macromolecules 1994, 27, 3087–3093. G. J. Fleer, Ber. Bunsen-Gesellschaft 1996, 100, 936–942. Y. V. Lyatskaya, F. A. M. Leermakers, G. J. Fleer, E. B. Zhulina, T. M. Birshtein, Macromolecules 1995, 28, 3562–3569. E. B. Zhulina, T. M. Birshtein, O. V. Borisov, Macromolecules 1995, 28, 1491–1499. E. B. Zhulina, O. V. Borisov, Macromolecules 1996, 29, 2618–2626. H. Zhang, J. RRhe, submitted for publication. E. P. K. Currie, A. B. Sieval, G. J. Fleer, M. A. C. Stuart, Langmuir 2000, 16, 8324– 8333. X. Guo, M. Ballauff, Langmuir 2000, 16, 8719–8726. R. D. Wesley, T. Cosgrove, L. Thompson, S. P. Armes, N. C. Billingham, F. L. Baines, Langmuir 2000, 16, 4467–4469.

24 A. Ikegami, N. Imai, J. Polym. Sci. 1962, 56,

133–152. 25 W. G. Pohl, H. J. Kuhn, Kolloid Z Z Polymere

1966, 212, 1–12. 26 E. B. Zhulina, O. V. Borisov, T. M. Birshtein,

Macromolecules 1999, 32, 8189–8196. 27 T. M. Birshtein, E. B. Zhulina, Ber. Bunsen-

Gesellschaft 1996, 100, 929–935. 28 I. Michaeli, J. Polym. Sci. 1960, 48, 291–299. 29 A. Ikegami J. Polym. Sci. 1964, Pt. A 2, 907–

921. 30 I. Sabbagh, M. Delsanti, P. Lesieur, Eur.

Phys. J. B: Condensed Matter Physics 1999, 12, 253–260. 31 I. Sabbagh, M. Delsanti, Eur. Phys. J. E: Soft Matter 2000, 1, 75–86. 32 T. Miyajima, M. Mori, S. Ishiguro, K. H. Chung, C. H. Moon, J. Colloid Interface Sci. 1996, 184, 279–288. 33 T. Miyajima, M. Mori, S. Ishiguro, J. Colloid Interface Sci. 1997, 187, 259–266. 34 J. Francois, C. Heitz, M. M. Mestdagh, Polymer 1997, 38, 5321–5332. 35 Y. Ikeda, M. Beer, M. Schmidt, K. Huber, Macromolecules 1998, 31, 728–733. 36 R. Schweins, K. Huber, Eur. Phys. J. E: Soft Matter 2001, 5, 117–126. 37 B. L. Rivas, E. D. Pereira, I. Moreno-Villoslada, Prog. Polym. Sci. 2002, 28, 173–208. 38 F. Horkay, P. J. Basser, A.-M. Hecht, E. Geissler, Macromolecules 2000, 33, 8329– 8333. 39 F. Horkay, I. Tasaki, P. J. Basser, Biomacromolecules 2000, 1, 84–90. 40 F. Horkay, I. Tasaki, P. J. Basser, Biomacromolecules 2001, 2, 195–199. 41 F. Horkay, A.-M. Hecht, I. Grillo, P. J. Basser, E. Geissler, J. Chem. Physics 2002, 117, 9103–9106. 42 A. V. Ermoshkin, M. Olvera de la Cruz, Phys. Rev. Lett. 2003, 90, 125504/1–125504/4. 43 G. B. Deacon, R. J. Phillips, Coordination Chem. Rev. 1980, 33, 227–250. 44 P. H. McCluskey, R. L. Snyder, R. A. Condrate, J. Solid State Chem. 1989, 83, 332–339. 45 J. W. Nicholson, J. Appl. Polym. Sci. 1998, 70, 2353–2359. 46 J. W. Nicholson, E. A. Wasson, A. D. Wilson, Br. Polym. J. 1988, 20, 97–101.

271

272

13 Weak Polyelectrolyte Brushes: Complex Formation and Multilayer Build-up 47 H. Zhang, J. RRhe, Polymer Preprints (Am.

48 49 50

51

Chem. Soc., Div. Polym. Chem.) 2003, 44, 536–537. H. Zhang, J. RRhe, Macromol. Rapid Commun. 2003, 24, 576–579. H. Zhang, J. RRhe, Macromolecules 2003, 36, 6593–6598. H. Dautzenberg, J. Hartmann, S. Grunewald, F. Brand, Ber. Bunsen-Gesellschaft 1996, 100, 1024–1032. G. Decher, J. D. Hong, Makromol. Chem., Macromol. Symp. 1991, 46, 321–327.

52 S. T. Dubas, J. B. Schlenoff, Macromolecules

1999, 32, 8153–8160. 53 H. G. M. Van de Steeg, M. A. Cohen Stuart,

A. De Keizer, B. H. Bijsterbosch, Langmuir 1992, 8, 2538–2546. 54 M. A. C. Stuart, J. Physique (Paris) 1988, 49, 1001–1008. 55 H. Dautzenberg, W. Jaeger, J. Koetz, B. Phillip, C. Seidel, D. Stscherbina, Polyelectrolytes: Formation, Characterization and Application, 1994.

273

14

Structure and Properties of High-Density Polymer Brushes Yoshinobu Tsujii, Muhammad Ejaz, Shinpei Yamamoto, Kohji Ohno, Kenji Urayama, and Takeshi Fukuda

Abstract

Surface modifications by polymers are becoming increasingly important in a variety of applications ranging from biotechnology to advanced microelectronics. As a group, we were among the first to apply atom transfer radical polymerization (ATRP) – a variant of living radical polymerization – to the surface-initiated graft polymerization of methyl methacrylate, styrene, and some functional monomers on various solid surfaces, to produce a graft layer of low-polydispersity polymer with the highest ever reported graft density. Atomic force microscopic and ellipsometric studies have revealed that, in such a graft layer, the polymer chains are highly extended to almost their full length in good solvent, and that such a high-density polymer brush has characteristic structures and unpredictable properties, in both swollen and dry states, that differ from those of “moderately dense” polymer brushes studied previously. We believe that the current studies on high-density polymer brushes will open up new routes to “precision” surface modification.

14.1

Introduction

Polymers that are densely end-grafted onto a solid surface are obliged to stretch away from the surface, forming a so-called “polymer brush”. Because of their important role in many areas of science and technology – for example, colloid stabilization, adhesion, lubrication, tribology and rheology – polymer brushes have been extensively studied both theoretically and experimentally [1]. Most of the polymer brushes studied so far have been prepared mainly by end-functionalized polymers or block copolymers with a terminal group or one block selectively adsorbed onto the surface. Such systems, however, were found to have a rather low graft density due to the steric hindrance of preadsorbed chains. An alternative method is that of graft polymerization starting with initiating sites fixed on a surface [2–4], but this usually results in a poor control of chain length and its distribution.

274

14 Structure and Properties of High-Density Polymer Brushes

Figure 14.1

Development of a high-density polymer brush.

Recently, living radical polymerization (LRP) techniques – which have attracted much attention as a new route to well-defined, low-polydispersity polymers – were successfully applied to the surface-initiated graft polymerization [5–12]. As a group, we were the first to succeed in applying atom transfer radical polymerization (ATRP) – one of several LRP techniques – to prepare densely grafted low-polydispersity poly(methyl methacrylate) (PMMA) onto various surfaces. This chapter reviews our recent studies with high-density polymer brushes (Figure 14.1), and outlines their controlled synthesis by surface-initiated ATRP. Some experimental results are also presented which indicate that these brushes, in both solvent-swollen and dry states, have characteristic structures and unpredictable properties that differ from those of the “moderately dense” polymer brushes studied previously. Finally, the potential applicability of high-density polymer brushes is discussed, notably with reference to the fabrication of new functional surfaces.

14.2

Controlled Synthesis of High-Density Polymer Brush by ATRP

The key reaction of LRP is the reversible activation-deactivation process [13]. LRP has several branches, each differing in the mechanism of the reversible reaction. In these studies, ATRP was used in which the activation-deactivation process mediated by halogen transfer is catalyzed by Cu/ligand complexes. The technique by which graft polymerization is controlled is shown schematically in Figure 14.2. First, the initiator (e.g., 2-(4-chlorosulfonylphenyl)ethyltrichlorosilane; CTS) is fixed on a silicon substrate by the coupling reaction with a silanol group on the surface. The CuI complex abstracts the chlorine atom of CTS, producing an initiating radical, to which some units of monomer are added until the propagating radical is recapped

14.2 Controlled Synthesis of High-Density Polymer Brush by ATRP

Schematic illustration of surface-initiated atom transfer radical polymerization: the filled circle and X represent a monomer molecule and a halogen atom, respectively.

Figure 14.2

to be a halogen-terminated “dormant” chain. This cycle occurs repeatedly and randomly on the halogenated sites on the surface, thus allowing all graft chains to grow almost simultaneously, and in a controlled fashion. Using this method, we graft-polymerized MMA from a CTS-fixed Silicon wafer as well as from silica particles (Aerosil 200; Nippon Aerosil Co., Japan; average diameter ~12 nm). p-Toluenesulfonyl chloride (TsCl) was added as a free initiator to control the polymerization. The free polymers produced from the free initiator can be a useful measure for the molecular characteristics of the graft polymers. To confirm this, the graft polymers were cleaved from the surface by HF treatment and subjected to gel-permeation chromatography (GPC) measurement. The number-average molecular weight (Mn) and polydispersity index (Mw/Mn) of the graft and free polymers as a function of conversion is illustrated in Figure 14.3. This figure shows that the graft polymerization was controlled as well as the free ATRP, and that the graft polymers have almost the same Mn and Mw/Mn as the free polymers throughout the course of polymerization. A similar result was obtained for graft polymerizations on the silica particle with a larger diameter (~200 nm), which means that surface curvature has little effect on grafting and grafted chains. Hence, the Mn and Mw/Mn values of the free polymers can be reasonably adopted to measure those of the graft polymers. The amount of the polymers on the Silicon wafer and silica particles estimated by ellipsometry and FT-IR spectroscopy, respectively, is plotted against the Mn of the free polymers in Figure 14.4. All of the data are seen to fall into a straight-line, proportional relationship, indicating that the graft density is constant throughout the course of polymerization, and that graft polymerization proceeds in a “living” fash-

275

276

14 Structure and Properties of High-Density Polymer Brushes

Figure 14.3 Plots of Mn and Mw/Mn of the cleaved (d) and free (s) polymers versus monomer conversion.

ion. Based on the slope of the line, the graft density (r) was estimated to be about 0.5 chains nm–2. This is one of the highest ever reported values, being an order of magnitude larger than those of polymer brushes prepared by the adsorption method. The question arises as to why such a high graft density was obtained. In conventional radical graft polymerization, a radical formed on the surface will instantly grow to a high molecular-weight polymer, and the amount of polymer grafted increases with the increasing number of graft chains. In this way, the alreadygrafted chains would sterically screen their neighboring grafting sites. However, in living graft polymerization all graft chains will grow more or less simultaneously, and with a constant graft density. Thus, the problem of steric hindrance will be reduced, and a high graft density achieved. Surface-initiated LRP is useful not only to control chain length and distribution, but also to achieve a high graft density.

Relationship between grafted amount and Mn of free polymers: the open and closed symbols represent the data on the silica particles and silicon wafer, respectively.

Figure 14.4

14.3 Structure and Properties of High-Density PMMA Brushes

14.3

Structure and Properties of High-Density PMMA Brushes 14.3.1

Swollen Brushes

A high-density PMMA brush formed on a Silicon wafer and swollen in toluene was studied using atomic force microscopy (AFM). The interaction force (F) between the graft layer and a silica probe attached on the AFM cantilever was measured as a function of separation (D) between the silicon substrate and silica probe surfaces (Figure 14.5). The measured force F can be reduced to the free energy of interactions (Gf ) between two parallel plates according to the Derjaguin approximation [14], F/R = 2pGf, where R is the radius of the probe sphere (5 lm). Figure 14.5 shows a typical graph with F/R plotted against D. Note that the true distance D between the substrate surface and the silica probe, which is usually difficult to define in AFM experiments, was successfully determined by AFM imaging the sample surface across the boundary of a scratched and an unscratched region of it. The most notable feature of the F/R versus D curve is a rapid increase in the repulsive force with decreasing D. The observed repulsive forces originate from steric interaction between the solvent-swollen brush and the probe sphere. The equilibrium thickness (Le) of the solvent-swollen brushes was determined as the critical distance from the substrate surface beyond which no repulsive force was detectable (cf. Figure 14.5). The scaling and self-consistent mean field approaches predict that Le varies in the manner [15,16]:

Typical F/R versus D curve between the PMMA brush (Ld = 87 nm, Mn = 121 700; Mw/Mn = 1.39) and the silica probe (attached on a AFM cantilever). The arrowheads indicate critical distances: Le is the

Figure 14.5

equilibrium thickness at which a repulsive force is detectable, and D0 is the offset distance beyond which the brush was no more compressible.

277

278

14 Structure and Properties of High-Density Polymer Brushes

Le Lcr1/3

(1)

where Lc is the contour length of the graft chain. This relationship was confirmed by other theoretical calculations, as well as by some experimental data. A series of PMMA brushes with near-equal density and different chain lengths followed this proportional relationship, confirming the brush structure as predicted by Eq. (1) [17a]. More interesting is the change in Le as a function of graft density [17b]. A series of PMMA brushes with the same chain length and different graft densities (0.07 < r (chains nm–2)<0.7) were prepared by the photodecomposition of the surface initiator followed by ATRP grafting. Figure 14.6 shows the plot of Le/Lc,w versus r* in logarithmic scale, where Lc,w is the weight-average full length of the graft chain (in all-trans conformation) and r* is the dimensionless surface density normalized by the monomer cross-sectional area. The weight average, rather than the number average, was adopted by referring to the studies of Milner et al. [16], in which a (moderate-density) polymer brush with a uniform distribution in chain length (Mw/Mn >1) was predicted to be thicker than the equivalent monodisperse brush with the same Mn by a factor of 1+{(3/4)(Mw/Mn – 1)}1/2. Figure 14.6 shows that Le/Lc,w increases with increasing r*, meaning that the graft chains become increasingly extended as the graft density increases. For the highest-density brush with r* = 0.4 (r = 0.7 chains nm–2), the value of Le/Lc,w even approaches 0.9. This surprisingly large value certainly indicates that the graft chains in this brush are extended to an extraordinarily high extent, and compares comparably with fully extended chains. The slope of the curve in Figure 14.6 is much larger than the value of 1/3 expected

Plots of Le/Lc,w versus dimensionless graft density r*; (1) PS-polydimethylsiloxane block copolymers: h (Mw,PS = 60 000) and e (Mw,PS = 169 000) [18]. (2) PEO-PS block copolymers: ~ (Mw,PEO = 30 800) and Figure 14.6

, (Mw,PEO = 19 600) [19]. (3) PMMA brushes: d (Mw = 31 300 ~ 267 400) [17]. The dashed line represents the theoretical prediction for the “semi-dilute” brush.

14.3 Structure and Properties of High-Density PMMA Brushes

theoretically and experimentally for the “semi-dilute” brush (shown by the dashed line in Figure 14.6), suggesting that such a scaling theory derived for the moderatedensity regime is no more applicable to the present “high-density” brushes. The rdependency of Le varies in the manner Le/Lc rn with n increasing from about 1/3 to 1/2 with increasing r. An increase of n in the high-density regime has been predicted by the theory in which higher-order interactions were taken into account [20]. The breakdown of the semi-dilute brush theory was also revealed in the force-distance curve. Using the scaling approach [21], de Gennes derived the equation which relates the interaction force between two parallel plates with a “moderately dense” polymer brush layer, in which graft chains overlap each other but the volume fraction of polymer in the layer is still low. This predicts that the force-distance profiles should be scaled by plotting (F/R) r–3/2 against D/Le for such brushes. In fact, the results for the previously studied block-copolymer brushes were reported to be nearly consistent to this scaling theory. Our system, however, was poorly represented by this scaling theory [17]. With increasing Lc and r, the scaled force curve becomes steeper, meaning that the brush layer is more resistant to compression. For example, the brush layer with the highest r was compressible only to D/Le » 0.7, in contrast to D/Le » 0.2 for the lowest layer. This strong resistance against compression must be characteristic of polymer brushes with an extremely high graft density. In conclusion, the chains in these high-density brushes were highly extended, to almost their full lengths, and were highly resistant against compression. It is believed that these were virtually the first observations of the “real” polymer brush behavior. 14.3.2

Dry Brushes

Ultra-thin polymer films on a solid substrate (supported films) are extremely interesting objects, both scientifically and practically. Detailed knowledge of their structure and properties is essential for the design of advanced materials. Dry polymer brushes are also interesting as a type of supported film. However, the effects on their structure and properties of end-grafting have been studied only for low- to moderate-density brushes. The above-described high-density polymer brushes should be highly anisotropic because their thickness in the dry state (Ld) reaches about 35% of the fully extended chain length (Ld/Lc,w = 0.35). Since the size of the unperturbed chain end-grafted on a repulsive surface is proportional to the squareroot of chain length [22], this large value of Ld means that the chains are already highly extended in the dry state as compared with their unperturbed dimensions. This may result in structure and properties that are quite different from those of previously studied polymer brushes. The glass transition temperature (Tg) of a thin polymer film has been extensively studied [23–26], and Tg has been shown to depend heavily on the thickness of the film and nature of the substrate. We have made the first observations of Tg for highdensity PMMA brushes grafted onto a silicon wafer. The method used was that of

279

280

14 Structure and Properties of High-Density Polymer Brushes

Figure 14.7 Ld-dependency of Tg measured by temperaturevariable ellipsometry. The solid and open circles represent the data for the brushes and the cast films, respectively.

temperature-variable spectroscopic ellipsometry [27]. A plot of Tg-values for the PMMA brushes and corresponding cast films is shown in Figure 14.7. A series of PMMA brushes studied here had a near-constant graft density (~0.7 chains nm–2) and different chain lengths (and hence different Ld). One remarkable finding was the difference in Tg behavior between these two types of ultra-thin films. The molecular characteristics of the polymer forming each cast film was closely similar to those of the polymer forming the brush of (nearly) the same thickness, and therefore the Tg difference could not be ascribed to differences in molecular characteristics such as chain length, chain length distribution, and stereoregularities. Rather, it could be totally ascribed to the effects of grafting – that is, to the chemical binding of one of the chain ends on the substrate surface. Within the range of Ld less than 50 nm, cast films suffer a significant depression in Tg, which can be ascribed not only to the molecular weight effect but also to the interfacial effect. In contrast, the Tg of the brushes steeply increases with decreasing Ld, and clearly, end-grafting restricts the mobility of the chains. One may expect, however, that the effect of end-grafting on chain mobility would become less and less significant as the chain length was increased, and that in the limit of the long chain, the Tg of the graft film would become equal to that of the cast film and hence of the bulk polymer, as all surface effects on the overall Tg of films should not be important in the long chain limit. Figure 14.7 shows, however, that this is not the case. As Ld increased over ~50 nm, the Tg of the brushes reached an almost constant

14.3 Structure and Properties of High-Density PMMA Brushes

value of about 119 RC, which is about 8 RC higher than that of the corresponding cast films. This value strongly suggests that such a difference in Tg between the brushes and cast films would be retained in the long chain limit. We ascribed this marked increase in Tg to an anisotropic structure/chain conformation in high-density brushes, though for low- to moderate-density PMMA brushes with r £ 0.2 chains nm–2, no increment in Tg was observed. Furthermore, by using electromechanical interferometry it could be shown that such an anisotropic structure brought about a large change in elastic properties [28]. Changes in the thickness of a graft film induced by an applied electric field (electrostriction) were measured (using a Nomarski optical interferometer) as a function of temperature [29]. The analysis of the electromechanical and dielectric data yielded the plate compressibility (kp) of the brushes in the glassy and molten states. Comparison of the results for a high-density PMMA brushes and a reference (spin-coated) PMMA layer will reveals differences in both elastic and dielectric properties between the brush composed of highly stretched graft chains and the layer formed by equivalent chains of random-coil conformation. In the glassy state, there was no appreciable difference in kp between the brush and the spin-coated layer, whereas in the molten state, the kp of the brush was markedly (ca. 30~40%) lower. This proved that the molten, high-density PMMA brush was more resistant to compression than the equivalent PMMA melt. An attempt was made to interpret the low compressibility of the molten, high-density brushes in terms of a rubber elasticity theory of a stretched polymer network with entanglements. The analysis suggested that the low compressibility was mainly attributed to a strain-hardening effect of the highly stretched entangled chains, and that there existed a considerable amount of entanglement of different graft chains, which in turn contributed to the elastic modulus. Another example showing the unique properties of high-density brushes is related to the miscibility of the PMMA polymer brush with a chemically identical polymer matrix [30]. Neutron reflectometry (NR) was applied to a series of deuterated PMMA (PMMAd) brushes with a constant chain length (Mn = 46 000; Mw/ Mn = 1.08) and different graft densities (r » 0.7 and 0.06 chains nm–2), on which hydrogenated PMMA (PMMAh) with various molecular weights (2400 < Mn,cast < 780 000) was spin-coated. After annealing at 150 RC for 5 days in vacuum, the NR data were collected at room temperature on the PORE time-of-flight reflectometer at the High Energy Accelerator Research Organization, Tsukuba, Japan. The brush concentration profile was deduced as a function of the distance (z) from the substrate surface by analyzing the q-dependent reflectivity data, where q is the wave vector of incident neutron. A representative result is shown in Figure 14.8, where the brush fraction U is plotted against z/Ld. The figure clearly suggests that the miscibility between the brush and the free polymer depended strongly on graft density. The low-density (semi-dilute) polymer brush (Mn = 46 000; r = 0.06 chains nm–2) was swollen by an oligomeric PMMA (Mn = 4910) to a thickness about four times that of the dry brush. By contrast, the high-density brush (Mn = 46 000; r = 0.7 chains nm–2) was hardly swollen by the same oligomeric PMMA, and maintained its “dry brush” structure. Previously, this phenomenon has been predicted theoretically [31], but never verified experimentally.

281

282

14 Structure and Properties of High-Density Polymer Brushes

Density profiles of polymer brushes swollen with chemically identical polymer matrix. The polymer brushes have a constant chain length (Mn = 46 000; Mw/Mn = 1.08) and different graft densities (r » 0.7 and 0.06 chains nm–2), and the free polymer has Mn = 4910 and Mw/Mn = 1.1. Figure 14.8

14.4

Application of High-Density Polymer Brushes

The simplicity and versatility of LRP enables us to create dense grafts of a variety of well-defined functional polymers [32,33]. As an example, we will briefly describe the polyelectrolytic properties of a high-density poly(methacrylic acid) brush. This was prepared by the surface-initiated ATRP of 1-propoxyethyl methacrylate (PEMA), followed by deprotection (cf. Figure 14.9). The analysis of the simultaneously formed free polymers suggested that the graft polymerization proceeded in a controlled fashion to produce a well-defined, high-density polymer brush with a graft density of ~0.4 chains nm–2. FT-IR analysis revealed that the hemiacetal ester group of poly(PEMA) grafts was quantitatively deprotected by heating at 120 RC in p-xylene containing zinc 2-ethyl hexanoate as a catalyst, producing a high-density polyelectrolyte brush. When glycidyl methacrylate (GMA) was copolymerized with PEMA, the deprotection of PEMA induced cross-linking between the graft chains by reaction of the liberated carboxylic acid with the epoxide group of GMA. The swelling behavior of the polyelectrolyte brushes with and without cross-links was studied in an aqueous solution with differing acidity. Figure 14.9 shows the swelling ratio of these brushes as a function of pH. The ratio increased steeply at around pH 10, to almost the maximum possible, which corresponded to the fully stretched chain length. This means that the apparent pKa was shifted to about 10 – a value much higher than that of the carboxylic acid (pKa = 4–5). This may be ascribed to so-called “charge regulation” [34], an effect which depends heavily on graft density. This successful preparation of a high-density polyelectrolyte brush enabled us to make the

14.4 Application of High-Density Polymer Brushes

Plot of swelling ratio of polyelectrolyte brushes with (s) and without (d) cross-links versus pH of aqueous solution.

Figure 14.9

first observation of such a large pKa shift. The high density also brought about the almost fully stretched conformation of the graft chain. The cross-linking introduced by the copolymerization of 3 mol% GMA enhanced the chemical stability of the graft layer, even in a strong acidic/basic condition. Surface-initiated LRP techniques make it possible to precisely and widely control the structural parameters of polymer brushes, including chain length, chain length distribution, and the monomer sequence distribution along the graft chain (Figure

Figure 14.10

Structural control in high-density polymer brushes.

283

284

14 Structure and Properties of High-Density Polymer Brushes

14.10). The grafting of block or gradient copolymers will produce a layered or gradient structure, even in a good solvent. Such a structure is expected to be stabilized not only by the thermodynamic incompatible interaction of different polymer segments, but also by the highly stretched conformation. The control of surface morphologies has been achieved by patterning the initiator layer and subsequent graft polymerization [8,35]. Grafting block copolymers or homopolymer mixtures have also provided characteristic surface morphologies caused by the phase separation on the nanoscale [7,36,37]. Furthermore, it has been shown that surface-initiated LRP methods are applicable to a variety of materials [38,39]. The results of these investigations will open up a new route to “precision” surface modification, with the ability to produce new types of nanostructured devices, especially when the unique structures and properties of high-density polymer brushes are incorporated into surface design.

Summary

By applying ATRP, we have succeeded in controlling graft polymerization initiated from solid surfaces and grafting a variety of well-defined polymers at exceptionally high graft densities. The high density of grafting was shown to make graft chains stretch to a great extent in the dry state, and to almost their full length in a good solvent. In addition, the process provided brushes, in both dry and swollen states, that had properties quite different and unpredictable from those of previously studied semi-dilute polymer brushes. It is hoped that this will bring about a breakthrough in polymer brush science, and open a new route to the production of welldefined, functional surfaces.

Acknowledgments

These studies were supported in part by a Grant-in-Aid for Scientific Research (Grant-in-Aid 12450385, 14205131, 14350496) from the Ministry of Education, Culture, Sports, Science and Technology, Japan and by Industrial Technology Research Grant Program in 2000 from the New Energy and Industrial Technology Development Organization (NEDO) of Japan.

References

References 1 See, e.g., (a) S. T. Milner, T. Witten, M. Cates,

Science 1991, 251, 905; (b) A. Halperin, M. Tirrell, T. P. Lodge, Adv. Polym. Sci. 1992, 100, 31; (c) M. Kawaguchi, A. Takahashi, Adv. Colloid Interface Sci. 1992, 37, 219; (d) B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677; (e) M. S. Kent, Macromol. Rapid Commun. 2000, 21, 243. 2 G. Boven, M. L. C. M. Oosterling, G. Challa, A. Jan, Polymer 1990, 31, 2377. 3 (a) N. Tsubokawa, K. Maruyama, Y. Sone, M. Shimomura, Polym. J. 1989, 21, 475; (b) N. Tsubokawa, Y. Shirai, K. Hashimoto, Colloid Polym. Sci. 1995, 273, 1049; (c) N. Tsubokawa, M. Koshida, J. Macromol. Sci. -Pure Appl. Chem. 1997, A34(12), 2509. 4 (a) O. Prucker, J. RUhe, Macromolecules 1998, 31, 592; (b) O. Prucker, J. RUhe, Macromolecules 1998, 31, 602. 5 M. Ejaz, S. Yamamoto, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules 1998, 31, 5934. 6 X. Huang, M. J. Wirth, Macromolecules 1999, 32, 1694. 7 (a) B. Zhao, W. J. Brittain, J. Am. Chem. Soc. 1999, 121, 3557; (b) B. Zhao, W. J. Brittain, W. Zhou, S. Z. D. Cheng, J. Am. Chem. Soc. 2000, 122, 2407. 8 (a) M. Hussemann, E. E. MalmstrWm, M. McNamara, M. Mate, D. Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang, T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424; (b) M. Husemann, M. Morrison, D. Benoit, J. Frommer, C. M. Mate, W. D. Hinsberg, J. L. Hedrick, C.J. Hawker, J. Am. Chem. Soc. 2000, 122, 1844. 9 K. Matyjaszewski, P. J. Miller, N. Shukla, B. Immaraporn, A. Gelman, B. B. Luokala, T. M. Siclovan, G. Kickelbick, T. Vallant, H. Hoffmann, T. Pakula, Macromolecules 1999, 32, 8716. 10 (a) T. von Werne, T. E. Patten, J. Am. Chem. Soc. 1999, 121, 7409; (b) T. von Werne, T. E. Patten, J. Am. Chem. Soc. 2001, 123, 7497. 11 B. de Boer, H. K. Simon, M. P. L. Werts, E. W. van der Vegte, G. Hadziioannou, Macromolecules 2000, 33, 349. 12 J.-B. Kim, M. L. Bruening, G. L. Baker, J. Am. Chem. Soc. 2000, 122, 7616. 13 See, e.g., (a) K. Matyjaszewski, T. P. Davis, Eds., Handbook of Radical Polymerization,

John Wiley & Sons, Inc., New York, NY, USA, 2002; (b) K. Matyjaszewski, Ed., Advances in Controlled/Living Radical Polymerization (ACS Symposium Series 854), American Chemical Society, Washington, DC, USA, 2003. 14 B. V. Derjaguin, Kolloid Zeits. 1934, 69, 155. 15 S. Alexander, J. Phys. (Paris) 1977, 38, 983. 16 (a) S. T. Milner, T. Witten, M. Cates, Macromolecules 1988, 21, 2610; (b) S. T. Milner, Europhys. Lett. 1988, 7, 695; (c) S. T. Milner, T. A. Witten, M. E. Cates, Macromolecules 1989, 22, 853. 17 (a) S. Yamamoto, M. Ejaz, Y. Tsujii, M. Matsumoto, T. Fukuda, Macromolecules 2000, 33, 5602; (b) S. Yamamoto, M. Ejaz, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 5608; (c) S. Yamamoto, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 5995. 18 (a) M. S. Kent, L. T. Lee, B. Farnoux, F. Rondelez, Macromolecules 1992, 25, 6240; (b) B. J. Factor, L. T. Lee, M. S. Kent, F. Rondelez, Phys. Rev. E 1993, 48, 2354; (c) M. S. Kent, L. T. Lee, B. J. Factor, F. Rondelez, G. H. Smith, J. Chem. Phys. 1995, 103, 2320. 19 H. D. Bijsterbosch, V. O. de Haan, W. de Graaf, M. Mellema, F. A. M. Leermakers, M. A. Cohen Stuart, A. A. van Well, Langmuir 1995, 11, 4467. 20 (a) P.-Y. Lai, A. Halperin, Macromolecules 1991, 24, 4981; (b) D. F. K. Shim, M. E. Cates, J. Phys. France 1989, 50, 3535. 21 P. G. de Gennes, Adv. Colloid Interface Sci. 1987, 27, 189. 22 (a) T. Tanaka, Macromolecules 1977, 10, 51; (b) P.-G. de Gennes, Scaling Concepts in Polymer Physics, Cornell University Press, NY, Ithaca, 1979. 23 For reviews, see, e.g., (a) R. A. L. Jones, Curr. Opin. Colloid Interface Sci. 1999, 4, 153; (b) J. A. Forrest, K. Dalnoki-Veress, Adv. Colloid Interface Sci. 2001, 94, 167. 24 J. L. Keddie, R. A. L. Jones, Isr. J. Chem. 1995, 35, 21. 25 O. Prucker, S. Christian, H. Bock, J. RUhe, C. W. Frank, W. Knoll, Macromol. Chem. Phys. 1998, 199, 1435. 26 (a) D. S. Fryer, P. F. Nealey, J. J. de Pablo, Macromolecules 2000, 33, 6439; (b) D. S. Fryer, R. D. Peters, E. J. Kim,

285

286

14 Structure and Properties of High-Density Polymer Brushes J. E. Tomaszewski, J. J. de Pablo, P. F. Nealey, C. C. White, W. Wu, Macromolecules 2001, 34, 5627; (c) R. S. Tate, D. S. Fryer, S. Pasqualini, M. Montague, J. J. de Pablo, P. F. Nealey, J. Chem. Phys. 2001, 115, 9982. 27 S. Yamamoto, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 6077. 28 K. Urayama, S. Yamamoto, Y. Tsujii, T. Fukuda, D. Neher, Macromolecules 2002, 35, 9459. 29 (a) H.-J. Winkelhahn, T. Pakula, D. Neher, Macromolecules 1996, 29, 6865; (b) K. Urayama, K. Oliver, R. Bohmer, D. Neher, J. Appl. Phys. 1999, 86, 6367; (c) K. Urayama, M. Tsuji, D. Neher, Macromolecules 2000, 33, 8269. 30 (a) S. Yamamoto, Y. Tsujii, T. Fukuda, N. Torikai, M. Takeda, KENS report 2001–2002, XVI, 204; (b) S. Yamamoto, Y. Tsujii, N. Takeda, T. Fukuda, to be submitted. 31 M. Aubouy, G. H. Fredrickson, P. Pincus, and E. Raphaeel, Macromolecules 1995, 28, 2979. 32 M. Ejaz, K. Ohno, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 2870.

33 C. Perruchot, M. A. Khan, A. Kamitsi,

S. P. Armes, T. von Werne, T. E. Patten, Langmuir 2001, 17, 4479. 34 T. Abe, S. Hayashi, N. Higashi, M. Niwa, K. Kurihara, Coll. Surf. A 2000, 169, 351. 35 (a) M. Ejaz, S. Yamamoto, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 1412; (b) Y. Tsujii, M. Ejaz, S. Yamamoto, T. Fukuda, K. Shigeto, K. Mibu, T. Shinjo, Polymer 2002, 43, 3837. 36 (a) A. Sidorenko, S. Minko, K. Schenk-Meuser, H. Duschner, M. Stamm, Langmuir 1999, 15, 8349; (b) S. Minko, M. Mueller, D. Usov, A. Scholl, C. Froeck, M. Stamm, Phys. Rev. Lett. 2002, 88, 35502. 37 M. Ejaz, K. Ohno, Y. Tsujii, T. Fukuda, Polym. Prepr. (Am. Chem. Soc., Div. Polym. Chem.) 2003, 44, 465. 38 M. Ejaz, Y. Tsujii, T. Fukuda, Polymer 2001, 42, 6811. 39 (a) K. Ohno, K.-M. Koh, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 8989; (b) K. Ohno, K.-M. Koh, Y. Tsujii, T. Fukuda, Angew. Chem. Int. Ed. 2003, 42, 2751.

287

15

Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients on Solid Substrates Tao Wu, Jan Genzer, Peng Gong, Igal Szleifer, Petr Vlcˇek, and Vladimr ubr

Glossary

DS F h H [H+] HDIW H(IS) IS ISmax Ka Mn Mo Mw/Mn N NA NB OB P(c) Q qgraf rg rs SB vgraf vo vw Zg a ao e

degree of swelling Helmholtz free energy dry thickness of surface-anchored polymer wet thickness of surface-anchored polymer proton concentration in bulk solution polymer wet thicknesses in “deionized” water polymer wet thicknesses evaluated at a given IS solution ionic strength solution ionic strength at the transition from OB to SB equilibrium constant of the acid-base equilibrium number average molecular weight of polymer monomer molecular weight polydispersity index degree of polymerization Avogadro’s number neutral brush osmotic brush the probability of a grafted polymer chain being at conformation c total charge density charge of grafted polymer radius of the polymer segment radius of the anions and cations salted brush volume of grafted polymer volume of the monomer unit volume of solvent normalization constant of the probability of the grafted polymer chains degree of dissociation in bulk solution “internal” degree of dissociation dielectric constant of the medium

288

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

l ve s,¥ r j v W

chemical potential electrostatic excluded volume parameter polymer density salt concentration in bulk solution brush grafting density polymer volume fraction Flory-Huggins interaction parameter electrostatic potential

Abstract

We describe experiments on surface-anchored poly(acrylic acid) (PAA) brushes with a gradual variation of the PAA grafting densities on flat surfaces, and provide detailed analysis of their properties. The PAA brush gradients are generated by first covering the substrate with a molecular gradient of the polymerization initiator, followed by the “grafting from” polymerization of tert-butyl acrylate (tBA) from these substrate-bound initiator centers, and finally converting the PtBA into PAA. We use spectroscopic ellipsometry to measure the wet thickness of the grafted PAA chains in aqueous solutions at three different pH values (4, 5.8, and 10) and a series of ionic strengths (IS). Our measurements reveal that at low grafting densities, r, the wet thickness of the PAA brush (H) remains relatively constant, the polymers are in the mushroom regime. Beyond a certain value of r, the macromolecules enter the brush regime. Here, H increases with increasing r. For a given r, H exhibits a nonmonotonous behavior as a function of the IS. At large IS, the H is small because the charges along PAA are completely screened by the excess of the external salt. As IS decreases, the PAA enters the so-called salt brush (SB) regime, where H increases. At a certain value of IS, H reaches a maximum and then decreases again. The latter is a typical brush behavior in so-called osmotic brush (OB) regime. We provide detailed discussion of the behavior of the grafted PAA chains in the SB and OB regimes. Single-chain mean-field theory is applied to simulate selected experimental results and provide molecular-level information about the brush properties, local dissociation inside the brush and the average charge of the polymer as a function of the distance from the surface. The experimental data are found to be reproduced very well by the theory.

15.1

Introduction

The properties of neutral polymer brushes under various conditions have been studied rather extensively over the past two decades, and their behavior is relatively well understood [1]. Both the scaling dependences for the average brush characteristics and the fine details of the intrinsic brush structure predicted theoretically are in rea-

15.1 Introduction

sonable agreement with experimental observations [2–4]. Much fewer investigations have been conducted on charged brushes [5] – that is, polymer assemblies comprising macromolecules containing ionizable groups. There are several motives for seeking to understand the interfacial behavior of charged macromolecules. One reason is to comprehend the properties of polyelectrolyte film surfaces. This is important in applied science, where the use of adsorbed polyelectrolytes is ubiquitous. Practical ramifications of these questions have additional implications for numerous industrial processes that rely on the properties of polymer surfaces. Examples from the chemical biotechnology and food industries include the stabilization and rheology of colloidal dispersions (paints), food emulsions (dairy products), coating of fibers in paper industries, wastewater treatment, mineral processing, and chromatographic separations. There are also evident connections to understanding the physics of charged biomolecules, charged proteins, and nucleic acids, cell attachment onto surfaces (cell growth and separation), and ramifications in applications such as blood clotting and immunoadsorption. Second, it is important for scientific reasons to understand the formation of polyelectrolyte thin films, their stability, and a response to an outside change of their environment. Knowledge gained by studying uncharged polymers is rather difficult to extend to charged systems. Third, understanding and tailoring the behavior of charged molecules at surfaces and interfaces is important in designing and utilizing novel applications (e.g., pH-controlled flow through polymeric micro-membranes), many of which cannot be fabricated using any other set-up. The behavior of polyelectrolyte brushes is quite complex because of a whole battery of parameters, which include both thermodynamic and electrostatic interactions. Specifically, in addition to the parameters governing the performance of neutral polymer brushes – that is, polymer molecular weight (or equivalently the degree of polymerization, N), brush grafting density, r, and the solvent quality (characterized by the Flory-Huggins interaction parameter, v) – the properties of polyelectrolyte brushes depend strongly on the degree of dissociation of the backbone charges (or degree of dissociation), a, counterion volume fraction in the polyelectrolyte solution, counterion valency, q, external salt concentration, s,¥, and in some cases also pH of the solution. Depending on the nature of the electric charges along the polymer backbone, one can distinguish between two types of polyelectrolytes. Strong (“quenched”) polyelectrolytes have a fixed a; their properties thus do not depend on the pH of the solution. On the other hand, in weak (“annealed”) polyelectrolytes, a depends on pH. Various theoretical approaches have been utilized to describe the performance of charged macromolecules at interfaces. In particular, scaling theories pioneered by Pincus [5], Zhulina and coworkers [6,7], and the Wageningen group [1,8–12] laid the foundation of our current understanding of polyelectrolyte brush behavior. These studies revealed that several different regimes of polyelectrolyte brushes could be identified, depending on the concentration of the external salt in solution. More detailed information about the structure of the brush has been obtained through more sophisticated methods based on numerical self-consistent field (SCF) [8–12] and analytical SCF theories [6,10,13]. The following description of the different regimes of

289

290

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

Dependence of the brush thickness reduced by the number of polymer repeat units for monovalent co-ions, H/N, on the concentration of the external salt, s, ¥, for strong (solid line) and weak (dashed line) polyelectroFigure 15.1

lyte brushes in neutral brush (NB), salted brush (SB), and osmotic brush (OB) regimes. a and ao denote the bulk and “internal” (for weak polyelectrolyte brushes only) degree of dissociation, respectively.

weak and strong polyelectrolyte brushes is based on the theoretical scaling approaches. The concentration of the external salt in polyelectrolyte solution has a profound effect on the brush properties. At high s, ¥ the salt concentration inside and outside the brush is about the same, and the electrostatic interactions are largely screened. Under such conditions, the polyelectrolyte brush behaves exactly as a neutral brush (NB) and H/N»(vr)1/3, where H is the height of the polymer brush and v (= 0.5 – v) is the excluded volume parameter. When the external salt concentration decreases, there is an unbalance in the ion concentration inside and outside the brush because the polymer charge density inside the brush (aj, where j is the polymer volume fraction) is no longer negligible with respect to s, ¥. The system enters the so-called salted brush (SB) regime. In the SB regime, H/N scales as H/N » (ver)1/3, where ve (= a2/s,¥) is the electrostatic excluded volume parameter. Due to the electrostatic interactions inside the brush, a salted brush is more extended than a neutral one. As shown schematically in Figure 15.1, the brush expansion increases with decreasing s,¥. If the external salt concentration is further decreased such that s, ¥ << aj, the co-ions are effectively expelled from the brush and H/N » a1/2. In this so-called osmotic brush (OB) regime, a limiting brush thickness is reached, which is independent of s, ¥ and r. Several experimental studies have appeared recently that reported on the interfacial properties of surface-grafted strong polyelectrolytes. Most experiments have focused on sodium polystyrene-sulfonate (PSSNa) [14–16], and recently poly(N-methyl-4-vinylpyridinium iodide) (MePVP) [17–19] brushes. The behavior of weak polyelectrolyte brushes is different from that of strong polyelectrolyte brushes. Here, the number of the backbone charges is not fixed. Specifi-

15.1 Introduction

cally, a depends on the proton concentration in the polymer solution, [H+] = 10–pH, and is given by a/(1 – a) = K/[H+], where K is the dissociation constant. When there is an excess of salt, as in the NB and SB regimes, [H+] inside and outside the brush is approximately equal and the internal degree of dissociation is the same as that in the bulk solution. Hence, the scaling for H/N in the NB and SB regimes is the same as in the case of strong polyelectrolyte brushes: !1=3 2 a 1=3 (1) H~Nr s;¥ When the system enters the OB regime, a significant electric potential difference develops between the brush and the solution in the bulk. In addition, [H+] inside the brush is considerably higher. As a consequence, a portion of the brush charges associate with protons and ao/(1 – ao) = K/(rao1/2), where ao is the “internal” degree of dissociation. This value of ao may be much smaller than the value in the bulk (a); the weak groups respond to the unfavorable electrostatic condition in OB by discharging themselves. The brush height in the OB regime is predicted to scale as [6]: a 1=3 1=3 þ 1=3 H~Nr ð½H þ S;¥ Þ (2) 1a Such a response is impossible for strong brushes, which have a fixed a. Figure 15.1 illustrates the different behavior of weak polyelectrolyte brushes in the OB regime. Because of the discharging process (ao < a), a weak brush in the OB regime is less expanded than the strong brush. As a result, H/N passes through a maximum as a function of s,¥, being small for both high and small s,¥. The unusual feature that at low s, ¥ the brush contracts with decreasing s,¥ is a typical property of weak groups, which can respond to a change in the local environment. As IsraOls suggested [11], by equating the expressions for H given in Eqs. (1) and (2), the value of s,¥ at the transition between the OB and SB regimes scales as: max

b 1=2

s;¥ ~rða Þ

(3)

Several groups investigated the behavior of weak polyelectrolytes anchored at surfaces. For example, Kurihara and Kunitake measured the surface forces between monolayers of anchored poly(acrylic acid) (PAA) [20]. These authors observed that the repulsive forces between two grafted polymeric monolayers increased with increasing salt concentration. Properties of PAA brushes were also studied using surface pressure isotherms and ellipsometry [21,22]. Samples with three different grafting densities were measured at three low pH solutions as a function of the solution ionic strength. Currie and coworkers found that the PAA wet brush thickness is a nonmonotonous function of the ionic strength at a given pH and grafting density. The extent of swelling of the brush increased with increasing pH and grafting density. Although the nonmonotonous behavior agreed qualitatively with theoretical predictions, the mean-field power law for the OB regime at a given pH and r (cf. Eq. (2)), was not observed. Moreover, because of the noncovalent nature of the PAA

291

292

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

grafting to the substrate used in Refs. [21] and [22], which investigated polyelectrolyte brushes comprising polystyrene-poly(acrylic acid) (PS-PAA) diblock copolymers, the copolymer chains partially desorbed from the surface when the copolymer compositional asymmetry increased and the length of the anchoring PS block decreased below a certain critical number of monomers. The goal of this work is to prepare surface-anchored PAA brushes and characterize their interfacial properties as a function of r, pH, and IS. Due to its simplicity, robustness and the ability to synthesize polymer brushes with narrow molecular weight distributions, atom transfer radical polymerization (ATRP) will be used to perform the surface-initiated polymerization [23]. Previous studies have revealed that acrylic-based polymers are difficult to prepare by ATRP because of the interaction of the carboxylic acid functionalities with the ATRP catalyst [24]. Hence, in order to form a surface-anchored PAA with a grafting density gradient, we first used the previously developed method [25] to synthesize a gradient of poly(tert-butyl acrylate) (PtBA) and then converted the PtBA into PAA by acid wash hydrolysis of PtBA [26,27]. We will use spectroscopic ellipsometry to measure the thickness of both dry PAA and PAA exposed to an aqueous solution of various pH and ionic strengths. In addition, we will use single-chain mean-field theory of charged polymer brushes to reproduce selected experimental findings and provide molecular insight into parameters that are either difficult to measure or are not experimentally accessible.

15.2

Experimental Section 15.2.1

Formation of the Gradient of the Polymerization Initiator

Silicon wafers covered with a native silicon oxide (thickness » 1.9 nm) were used as substrates. The silicon wafers were first exposed to an ultraviolet/ozone (UVO) treatment in a commercial UVO chamber (Jelight Company, Inc., Model 42) for 30 min in order to generate surface-bound hydrophilic surfaces comprising predominantly the hydroxy functionalities (-OH) [28]. A molecular gradient of n-octyltrichlorosilane (OTS) (Gelest, Inc.) was formed following the vapor diffusion technique developed by Chaudhury and Whitesides [29]. Specifically, as shown in Figure 15.2, a mixture of OTS and paraffin oil (PO) was placed into an open container that was positioned close to an edge of the silicon wafer. As the OTS evaporated, it diffused in the vapor phase and generated a concentration gradient along the silica substrate. Upon impinging on the substrate, the OTS molecules reacted with the substrate -OH functionalities and formed a selfassembled monolayer (SAM). The breadth and position of the OTS molecular gradient can be tuned by varying the OTS diffusion time and the flux of the OTS molecules. In this work, the OTS:PO ratio was kept 1:2 and the OTS diffusion time was 2 min. Contact angle measurements confirmed that the concentration of the substrate regions unmasked by the OTS molecules increased gradually as one traversed

15.2 Experimental Section

Schematic illustrating the preparation of the ATRP initiator gradient. (a) In the first step, a molecular gradient of n-octyltrichlorosilane (OTS) is prepared on the flat

Figure 15.2

silica-covered substrate. (b) In the second step, the substrate is immersed in a solution of [11-(2-bromo-2-methylpropanoyloxy)undecyl] trichlorosilane (BMPUS).

from the OTS-side of the substrate towards the OTS-unexposed edge. The empty sites on the substrate not occupied by the OTS SAM were filled by [11-(2-bromo-2methylpropanoyloxy)undecyl] trichlorosilane (BMPUS), an ATRP initiator, which was synthesized following a two-step procedure published previously [26]. In a vial, 20 lL of the BMPUS initiator was added into 20 mL anhydrous toluene and mixed thoroughly. The OTS gradient-covered silicon wafers were placed into this solution for 18 h without stirring. After the initiator deposition, the samples were removed from the solution, thoroughly washed with toluene, acetone, and ethanol, and dried with nitrogen. Ellipsometry measurements confirmed that only a single selfassembled monolayer of both OTS and BMPUS formed on the Si substrates. 15.2.2

Preparation of PtBA and Hydrolysis into PAA

The polymerization of tert-butyl acrylate (tBA) was performed by ATRP, as described earlier [27]. Copper(I) bromide (CuBr) (0.1564 g, 1.1 Q 10–3 mol) and copper(II) bromide (CuBr2) (0.012 g, 4.2 Q 10–5 mol) were added to a dry Teflon-capped vial, followed by adding deoxygenated acetone (1.6 g) and tBA (14.0 g, 0.108 mol). After purging the solution with nitrogen for 5 min, 240 lL N,N,N,N¢,N¢-pentamethyldiethylenetriamine (PMDETA) was added. The solution was stirred until the Cu complex formed. After the complex formation, the gradient-covered Si wafers were placed into the solution. The solution was purged with nitrogen for 2 min, after which the vial was sealed and placed in a temperature-controlled oil bath set at 60 RC. After a predetermined reaction time, the samples were removed and washed thoroughly with acetone, methanol and dried with nitrogen.

293

294

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

In order to convert the PtBA into PAA, the silica substrates covered with the PtBA gradient were placed into a flask, which contained a mixture of 20 mL 1,4-dioxane and 3 mL concentrated HCl (37%). The flask was connected to a condenser, and the solution heated to reflux. The samples were removed after 2–5 h (see discussion below) and thoroughly washed with the DI water and methanol. In addition to the surface polymerization, we have also prepared a small amount of PtBA in solution. For the solution polymerization of tBA, the procedure described earlier was duplicated with the exception that after all reagents had been combined and homogeneity reached, methyl 2-bromopropionate (240 lL, 2.2 Q 10–3 mol) was added as a solution initiator instead of BMPUS. The polymerization time of the solution-polymerized polymers was identical to the PtBA formed on the silicon wafer. After polymerization, the solutions were precipitated into a 15-fold excess of a 50:50 water:MeOH (v/v) bath. After decanting off the solvent, the polymer was redissolved in diethyl ether; the solution turned deep blue. Several drops of water were added to extract the copper complexes out of the solution. After separation, the polymer solution was re-precipitated until the polymer remained white. The final polymer was dissolved in acetone and dried under vacuum overnight. The PtBA polymerized in free solution was used to determine the corresponding molecular weight. We note the polymerization rate in the solution is likely faster than that on the flat substrate [30]. Nevertheless, the molecular weight of the solution-based polymer provided a useful estimate for further analysis. In order to convert solution polymerized PtBA into PAA, PtBA (2.5 g) was added to a solution of 20 mL dioxane and 3 mL concentrated HCl (37%). The solution was heated to reflux. After a reaction time of 2–5 h, the solution was cooled and a part of the excess reagents (»10 mL) was removed by evaporation under vacuum. Subsequently, the PAA solution was precipitated into 200 mL of methylethylketone (MEK). After decanting off the solvent, the polymers were dried under high vacuum overnight. 15.2.3

Polymer Characterization

The molecular weights of PtBA and PAA were measured by size-exclusion chromatography (SEC) at the Institute of Macromolecular Chemistry in Prague, Czech Republic. The PtBA SEC experiments were conducted at 20 RC using a Labora set apparatus (Czech Republic) with a two-column separation system (Polymer Standards Service GmbH, Germany; porosities 105 and 103 T). THF was used as the mobile phase. The flow rate was 1 ml min–1, and the concentration of samples for injection was approx. 1% (w/w). The system was calibrated with PMMA standards (PSS GmbH, Germany). Eluograms were analyzed using a software Caliber (Polymer Laboratories) and the Mark-Houwink-Sakurada equation with K = 3.3 Q 10–5 and a = 0.80 [31]. The system is equipped with two detectors of the Czech provenience: differential refractometer RIDK-102 and UV detector LCD-2040 with adjustable wavelength. The polymer weights were calculated from the RI traces. The molecular weight of the PtBA synthesized in solution for 6 h was Mn = 5.94 kDa with Mw/Mn = 1.07. These values are comparable with those reported previously

15.2 Experimental Section

(Mn = 6.0 kDa for PtBA) for polymers prepared under the identical reaction conditions [27]. PtBA polymerized for 10 h in solution had Mn = 8.56 kDa with Mw/ Mn = 1.15. After the conversion of PtBA to PAA, the molecular weight of PAA was also determined by SEC by using Vkta Explorer (Amersham Bioscience) with a SuperoseW 12 column, equipped with a differential refractometer Shodex RI-71 and a multiangle light-scattering detector DAWN DSP-F (Wyatt Technology Corp.). Sodium acetate buffer (0.3 M; pH » 6.5) was used as the mobile phase. The flow rate was 0.5 ml min–1. The molecular weight of PAA for 6 h was Mn = 3.56 kDa with Mw/ Mn = 1.32, and for 10 h was Mw = 4.1 kDa with Mw/Mn = 1.29. In The average molecular weights (Mn) and polydispersity indices (Mw/Mn) of PtBA and PAA obtained using the SEC measurements are listed in Table 15.1. The molecular weights of PAA calculated from the PtBA SEC measurements are in good agreement with the direct PAA measurements (the accuracy of –15% is well within the acceptable limit). In the data analysis that follows, we will be using the degree of polymerization of PAA that was calculated from the PtBA SEC measurements. We justify our choice by the fact that in contrast to PAA, PtBA is a neutral polymer, which is insensitive to the pH and ionic strength of the mobile phase. We thus believe that the PtBA SEC measurements are more accurate. Moreover, as indicated by the relatively narrow molecular weight distribution, the Mn,PtBA values should be more reliable. Comparison of size-exclusion chromatography (SEC) experiment results for PtBA and PAA.

Table 15.1

Polymerization time (h)

Mn,PtBA (kDa)

Mw/Mn

NPtBA a)

Mn,PAA a) (kDa)

Mn,PAA (kDa)

Mw/Mn

NPAA b)

6 10

5.94 8.56

1.07 1.15

» 46 » 67

3.34 4.81

3.56 4.10

1.32 1.29

» 49 » 57

a) b)

Calculated from Mn,PtBA Calculated from Mn,PAA

The thickness of the SAM and the polymer film was measured using a singlewavelength fixed geometry ellipsometer (AutoEL II; Rudolph Technologies) and a variable angle spectroscopic ellipsometry (J. A. Woollam, Inc.). The thickness was evaluated from the experimentally measured ellipsometric angles W and D using the supplied software (DafIMB and WVASE32). The following refractive indices were used for various material: 1.45 for SAMs [25], 1.466 for PtBA [32], and 1.527 for PAA [33]. The wet thickness of the PAA in aqueous solution was measured by placing the samples in a custom-designed solution cell, incubating them for a desired period of time (typically >5 h) and performing the experiments with VASE at u = 70R, where u is the angle between the incoming beam and the sample normal. The ellipsometric angles W and D were collected for a series of wavelengths ranging from 240 to 1000 nm. The wet PAA thickness was evaluated using a graded effective medium approximation model based on linear combination of the optical constants of the DI water and PAA [34].

295

296

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

The pH and the ionic strength of the aqueous solution were adjusted as follows. A desired amount of aqueous solution was removed and replaced with corresponding amounts of concentrated NaCl, HCl, or NaOH solutions. In the following, we illustrate the procedure used to prepare a solution with the pH = 4 and various ionic strengths. First, the ellipsometric cell was filled with 75 mL of aqueous solution, followed by adding 75 lL of 0.1 M HCl in order to obtain a solution of pH 4. We also prepared a stack NaCl solution (concentration 2 M, pH 4). The right-hand column in Table 15.2 lists the corresponding amounts of the aqueous solution that was replaced with the NaCl in order to prepare the ionic strength listed in the left-hand column of Table 15.2. Table 15.2

An example of adjusting the ionic strength of pH = 4 solution.

Ionic strength (IS)

1.0 1.0 5.0 1.0 1.0 5.0 7.0 1.0

Q Q Q Q Q Q Q Q

10–4 10–3 10–3 10–2 10–1 10–1 10–1 100

Amount of aqueous solution (mL) replaced by the NaCl solution (concentration = 2 M, pH = 4) in a 75-mL cell solution – 0.034 0.150 0.188 3.392 15.789 12.500 15.000

Near-edge X-ray absorption fine structure (NEXAFS) was used to study the spatial concentration of PAA on the sample substrates [35]. The NEXAFS experiments were carried out on the U7A NIST/Dow Materials Soft X-ray Materials Characterization Facility at the National Synchrotron Light Source at Brookhaven National Laboratory (NSLS BNL). The NEXAFS spectra were collected in the partial electron yield (PEY) at the so-called “magic” angle (h = 55R) incidence geometries, where h is the angle between the sample normal and the polarization vector of the X-ray beam. Fourier transform infrared (FT-IR) spectroscopy was collected in the transmission mode with Nicolet 750. A total of 1024 scans was made with resolution 8 cm–1 for each measurement, which required a scanning time of approximately 12 min. A bare silicon wafer was used as the blank. The IR spectra for polymer bulk were measured with the KBr pellets. The IR spectra were analyzed using the OMNIC 5.0 software.

15.3

Theory Section

In order to model the behavior of PAA brushes, we use a generalized form of a molecular theory that has been successful in describing the thermodynamic and struc-

15.3 Theory Section

tural properties of uncharged polymer brushes [4]. By comparing the experimental results with the predictions of the theory, we hope to first reinforce the validity of the assumptions made in the development of the theoretical model and, very importantly, through the theory gain access to system parameters that are not easy (or even possible) to measure. Examples of the latter include the local dissociation inside the brush and the average charge of the polymer as a function of the distance from the surface. In the past, the theory has been found to reproduce very faithfully experimental data on pressure-area isotherms of polystyrene-polyethylene oxide diblock copolymers spread at the water-air interface [36], the height and pressure of polystyrene brushes in good solvents [37], and was able to model the ability of short and long polyetheylene glycol tethered layers to reduce the adsorption of lysozyme and fibrinogen [38]. The theory has been recently generalized to treat charged systems [39]. Here, we present a simple derivation of the theory with the emphasis on the PAA brushes that are studied experimentally. We consider a surface of area A with ng polymer molecules grafted at one of their ends to the surface. The surface coverage is r = ng/A and each polymer chain has N segments (acrylic acid groups) that can carry a negative charge when ionized. Each polymer segment is modeled as sphere with a radius rg = 0.3 nm. The surface is in contact with a buffer solution. The bulk solution is composed of salt, at density s,¥, that is dissociated into monovalent anions and cations, each assumed to be sphere with a radius rs = 0.4 nm. The solvent (water) is characterized by a volume vw = 0.03 nm3. The protons are assumed to have the same volume as the solvent, but they carry a positive charge, and the bulk concentration of protons is given by the solution’s pH. Note that in reality what we call protons refers to hydronium ions, i.e. H3O+. The presence of a surface makes the system inhomogeneous. For simplicity, we assume that the inhomogeneity is only in the direction perpendicular to the surface, z. The excess free energy density (per unit area) is given by: R¥ P bF ¼ r PðcÞln PðcÞ þ þ ðzÞ ln þ ðzÞvw 1 dzþ A c 0 R¥ R¥ ðzÞ ln ðzÞvw 1 dz þ w ðzÞ ln w ðzÞvw 1 dzþ 0 ¥

R 0

H

h þ ðzÞ ln

(4)

0

¥ i 1R ðzÞv 1 dz þ QðzÞWðzÞdz w þ H 20

where the first term represents the conformational entropy of the polymer chains, with P(c) being the probability of a grafted polymer chain being at conformation c. The second to fifth terms are the z-dependent translational (mixing) entropies of the cations, anions, solvent, and protons, respectively. The last term represents the electrostatic interactions, with Q(z) being the total charge density at distance z from the surface and w(z) is the electrostatic potential at z. The total charge density is given by: QðzÞ ¼ r

P c

PðcÞqgraf ðz; cÞ þ þ ðz rs Þqþ þ ðz rs Þq þ

H

þ

ðzÞq

H

þ

(5)

297

298

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

where the first term is the average charge density that the grafted polymers contribute to z. qgraf(z;c) is the total charge that a grafted polymer in conformation c contributes to z. The second and third terms represent the contribution from the cations and anions, respectively. The charge of the cation and anion is assumed to be localized at the center of the ions; therefore, the shift in the position of the densities in Eq. (5) by rs. The last term in Eq. (5) is the contribution from the protons. In all of our calculations, we use q+ = –q- = qH+ = 1 in units of the elementary charge, e. Inspection of the free energy, Eq. (4), reveals that we have not included interaction terms beyond the electrostatic one. The reason is that in this version of the theory we assume that the solvent is good for the polymer and the salt. Therefore, the van der Waals attractive interactions among all the species are equal and they do not need to be explicitly included (athermal system). For the repulsive interactions, we assume hard-core repulsions between all molecular species. These are accounted for by packing constraints, expressed by the requirement that the volume available in each layer between z and z + dz is occupied by polymer segments, cations, anions, hydronium, and solvent. This is expressed in the form: 1¼r

P c

PðcÞvgraf ðz; cÞ þ

w ðzÞvw þ

H

þ

R

0

0

0

þ ðz Þvþ ðz; z Þdz þ

R

0

0

0

ðz Þv ðz; z Þdz þ

ðzÞvw

(6)

where the fist term is the average volume that the grafted polymers occupy between z and z + dz. vgraf(z;c) represents the volume that a polymer chain in conformation c occupies between z and z + dz, and explicitly accounts for the size and shape of each monomer of the polymer and thus of the total chain in each conformation c. The cation and anion terms include a sum (integral) over all the molecules that, having their point of closest distance to the surface at z¢, contribute volume to z. This effectively incorporates the exact shape and size of the particles into the theory through the volume distributions v(z;z¢). The probability of chain conformations and the densities of cations, anions, protons and solvent are determined by minimization of the free energy, Eq. (4), subject to the packing constraints, Eq. (6). This is achieved by introducing a set of Lagrange multipliers bp(z) to yield Eq. (7) for the probability of the chains: h R i R 1 PðcÞ ¼ exp bpðzÞvgraf ðz; cÞdz b WðzÞqgraf ðz; cÞdz (7) Zg where Zg is the normalization constant. For the density of cations and anions: h i R 0 0 þ ðzÞvw ¼ exp blþ bpðzÞvþ ðz ; zÞdz bWðz þ rs Þqþ h i R 0 0 ðzÞvw ¼ exp bl bpðzÞv ðz ; zÞdz bWðz þ rs Þq for the density of protons:

(8)

(9)

15.3 Theory Section

h þ ðzÞvw ¼ exp bl H

H

þ

bpðzÞvw bwðzÞq

i H

þ

(10)

and finally for the solvent: w ðzÞvw ¼ exp½bpðzÞvw

(11)

where for the cations, anions, and protons we have accounted for the fact that the system is in equilibrium with a bulk characterized by chemical potentials l+, l-, and lG+, respectively. The solvent chemical potential is not needed because of the constraints equation (see Ref. [4] for details). The minimization is carried out with the explicit consideration of the dependency of the electrostatic potential on the densities. Namely, the fact the electrostatic potentials are the z components of the Coulombic potentials given by: WðrÞ ¼

1 R 4pe

0

Qðr Þ rr 0 dr

(12)

where e is the dielectric constant of the medium and Q(r) is the total charge density at r. The Lagrange multipliers p(z) represent the lateral pressures acting on the molecular species [4], and are related to the local osmotic pressure. Now we need to determine the coupled repulsive (packing) and electrostatic potentials, p(z) and W(z) respectively. To this end, we replace the explicit expression for the density profiles for the cations, anion, protons and solvent (Eqs. (8–11)), and the probability of the polymer chains, Eq. (10), into the packing constraints, Eq. (9). This yields one set of equations, the second set is obtained by realizing that the electrostatic potential has to be the solution of the Poisson equation. Considering only the inhomogeneities in z this is: 2

@ WðzÞ 1 ¼ QðzÞ 2 e @z

(13)

with Q(z) from Eq. (5). The coupled constraint and the Poisson equations are solved by standard numerical methods, as explained in detail in Refs. [39] and [40], with the molecular parameters defined above. We recall that because of the surface-induced inhomogeneity of the system, we expect the distribution of all charged species to also be inhomogeneous, in particular, in the region close to the surface where the polymers are present. The polymer segments are acrylic acid, which have a well defined pKa. Thus, the fraction of polymer segments that are charged will depend on the local concentration of protons. This needs to be explicitly accounted for in the theory for the proper treatment of the system. We consider the acid-base equilibrium in a very similar way that was done with the self-consistent field theory developed by the Wageningen group [8–12]. The acid groups are characterized by an equilibrium of the form: þ

AH > H þ A

(14)

299

300

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

with an equilibrium constant þ ½A H Ka ¼ ½AH

(15)

where the brackets denote molar units. The degree of dissociation per acid group, a, is defined as:

a¼

½A 1 þ ¼ ½AH þ½A 1þ½H

(16)

Ka

The presence of the surface induces an inhomogeneity in the direction perpendicular to it. Therefore the proton concentration, and thus the dissociation, will be a function of z. We denote the local dissociation by a(z) defined by:

aðzÞ ¼

½A ðzÞ 1 ¼ þ ½AHðzÞþ½A ðzÞ 1þ½H ðzÞ

(17)

Ka

Thus, the Rz-dependent charge of a grafted polymer chain in conformation c will be q(z;c) = e n(z; c)a(z)dz, where n(z; c)dz is the number of acid groups that the chain in conformation c has at distance z from the surface. This is explicitly introduced in the probability of the polymer chains to yield: h R i R 1 exp bpðzÞvgraf ðz; cÞdz be WðzÞaðaÞngraf ðz; cÞdz (18) PðcÞ ¼ Zg The acid-base equilibrium induces a coupling between the polymer conformations and the pH. As we will show in the results section, the interplay between the electrostatic interactions, the packing and the acid-base equilibrium results in dramatic changes in the average properties of the polymer layer. In particular, in Section 15.5 we will demonstrate that local pH inside polymer brush varies very dramatically, both as a function of the distance from the surface and the amount of external salt. The numerical methodology that we use to solve the set of coupled non-linear equations as well as the chain model used to generate the polymer conformations needed as input in the constraint and Poisson equations are the same as those described in Refs. [39] and [40]. The number of conformations used in the calculations presented below is 500 000 (for details, see Refs. [39] and [40]).

15.4

Experimental Results

We first verified that the hydrolysis of the PtBA using HCl led to the formation of PAA. The FT-IR KBr spectra of PtBA prepared using the solution polymerization are shown in the bottom and top of Figure 15.3, respectively. Following previous studies [27], we make the following assignments to the FT-IR signals: the peak at 1733 cm–1 can be assigned to the ester carbonyl group (-COO), the peaks at 1254

15.4 Experimental Results

FT-IR spectra from the KBr pellets containing PtBA (bottom) and PAA (top), prepared using the bulk solution polymerization.

Figure 15.3

and 1159 cm–1 are attributed to C-O, and the weak peaks at 2850 and 2925 cm–1 belong to the symmetric and asymmetric vibration modes of the -CH2- groups, respectively. The characteristic peaks of the tert-butyl group, which exist only in PtBA, are at 2979 cm–1 (mas(CH3)) and 1393/1368 cm–1(ds(CH3)). The disappearance of those peaks in the PAA IR spectra confirms the completion of the hydrolysis of PtBA. The broad band at 3200 cm–1 in the PAA IR spectra is attributed to the -OH group formed during the PtBA conversion. In addition to the PtBA-to-PAA conversion measurements on the polymers prepared in the solution, we have also checked the conversion reaction for polymers synthesized on the surface. PtBA brushes were first grown on Si substrates that were covered homogeneously with the surface-bound initiators. The hydrolysis of PtBA on the Si wafer was performed as a function of the reaction time. The spectra for both samples (not shown) exhibit very similar trends reported for the bulk PtBA and PAA specimens (cf. Figure 15.3). As expected, the absolute IR intensities from the latter set of samples are weaker due to the smaller number of polymer chains analyzed. The disappearance of the peaks corresponding to mas(CH3) (at 2979 cm–1) and ds(CH3) (at 1393 and 1368 cm–1) in the PAA/Si spectra, relative to the PtBA/Si samples, provides evidence that the hydrolysis of PtBA took place. The PtBA gradient was prepared by applying the polymerization procedure described previously. The polymerization time was 24 h; the choice of this rather longer polymerization time was motivated by our hope to produce thick PtBA layers that would generate large enough signal for the FT-IR measurements. The FT-IR spectroscopy experiments were performed in order to verify that a gradient in graft-

301

302

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

FT-IR spectra of PtBA grafting density gradient polymerized from the substrate covered with the gradient of the ATRP initiator. The polymerization time was 24 h.

Figure 15.4

ing density of the PtBA was formed on the substrate. In Figure 15.4 are plotted the FT-IR spectra taken at various positions along the PtBA-covered specimen. The intensities of the two characteristic peaks of PtBA corresponding to m(COO) (at 1733 cm–1) and mas(CH3)t-butyl (at 2979 cm–1) are expected to increase while traversing the sample from the OTS-side (small distance along the substrate) towards the PtBA side (larger distance along the substrate). The experimental results confirm the anticipated trend. While FT-IR proved useful in confirming that the conversion of PtBA into PAA took place, it did not provide information about the thickness of the polymers on the substrate. Spectroscopic ellipsometry was used to measure the dry thickness of both the PtBA and PAA gradient samples, which were polymerized for 10 h and hydrolyzed in HCl/dioxane bath for 10 h. In Figure 15.5 are plotted the dry thicknesses of PtBA (solid symbols) and PAA (open symbols) as a function of the position on the substrate. The data in Figure 15.5 reveal that the thickness of both PtBA and PAA increases as one moves from the OTS side (small number on the abscissa) of the sample towards the initiator-covered side (large numbers on the abscissa); in both cases the functional form closely resembles that of a backward diffusion-like profile. Assuming that all chains of both PtBA and PAA have the same degree of polymerization along the substrate (i.e., the polymerization rate was independent of the grafting density of the initiator on the substrate) [25,34], the increase of the polymer dry thickness can be attributed to the increase of the polymer grafting density on the

15.4 Experimental Results

Dry thickness of PtBA (solid symbols) and PAA (open symbols) as a function of the position on the substrate. The solid line represents the PEY NEXAFS intensity measured at E = 531 eV on the PAA sample as a function of the position on the substrate. Figure 15.5

substrate. Interestingly, there is a rapid decrease in the dry polymer thickness after the hydrolysis. Specifically, the thickness of PtBA decreases four- to six-fold upon hydrolysis to PAA. The decrease does not seem to be dependent much on the grafting density of the polymer on the substrate. We will return to this point later in the discussion. Experiments using NEXAFS were conducted in order to evaluate the density of PAA as a function of the position on the substrate. With the X-ray monochromator set to 531 eV, which corresponds to the 1s!p*C=O transition [25,34,35], we collected the partial electron yield (PEY) NEXAFS signal by scanning the X-ray beam along the gradient. The solid line in Figure 15.5 depicts the variation of the PEY NEXAFS intensity corresponding to the C=O bond across the PAA gradient. The NEXAFS results confirm that the C=O intensity, and thus the amount of PAA on the surface, increases as one moves along the gradient. Moreover, the NEXAFS data are in good agreement with the ellipsometric thickness of PAA. In order to study the solution properties of the surface-grafted PAA, we incubated the PAA gradients under aqueous solutions under the different pH values (4, 5.8, and 10) and a series of ionic strengths for each pH. Hence, by measuring the wet thickness of PAA along the gradient at different solution condition, we obtained the wet thickness of the grafted polymer layer as a function of the PAA grafting density, pH, and ionic strength. The ellipsometry experiments described here were carried out on PAA with Mn = 4.8 kDa. The open symbols in Figure 15.6 represent the dry PAA thickness (h) measured as a function of the position on the substrate. The solid line is the PEY NEXAFS intensity scan at E = 531 eV, corresponding to the C=O peak. In the same figure, we also plot the wet thickness of PAA (H) measured at

303

304

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

Dry (h) and wet (H) thickness of PAA measured as a function of the position on the solid substrate. The wet thickness was evaluated at pH = 4, and is plotted as a function of

Figure 15.6

three different ionic strengths. The solid line represents the PEY NEXAFS intensity measured at E = 531 eV on the dry PAA sample.

pH = 4 and three different ionic strengths. The data show that for all ionic strengths, H increases with increasing grafting density of PAA on the substrate. A close inspection of the data in Figure 15.6 reveals that H is a nonmonotonous function of the ionic strength. Specifically, as the ionic strength increases, H also increases, reaches a maximum and then decreases. We will address this behavior in the discussion, which follows.

15.5

Discussion 15.5.1

Surface Hydrolysis of PtBA

We carried out a kinetic study of the PtBA/PAA conversion by measuring the time dependence of the PtBA thickness and contact angle on the reaction time of hydrolysis of PtBA polymerized on substrates with a homogeneous distribution of the surface-bound initiator. The results are summarized in Figure 15.7 for several samples with different initial PtBA dry thicknesses. In Figure 15.7(a) is plotted the dry thickness of PtBA, h, normalized by initial film thickness, h(t = 0), as a function of the time interval of hydrolysis. The data show that for all samples studied, the initial thickness decreases rapidly by about 50% within the first 2 h, followed by a slower decrease at later times. We attribute the sharp drop in the film thickness primarily to the decrease in the volume of the polymer, associated with the removal of the bulky tert-butyl groups.

15.5 Discussion

(An additional contribution to decreasing the occupied volume may also come from the presence of hydrogen bonds in PAA.) We note that similar behavior has been observed by others [27]. We speculate that the slower decrease in PtBA thickness may be associated with possible cleavage of the polymer from the substrate caused by the hydrolysis of the ester group inside the initiator (a primary ester). However, since the reaction rate for tert-alkyl ester acid hydrolysis is much faster than that for primary-ester hydrolysis – this ester cleavage is highly selective – we can complete the conversion of PtBA to PAA before the polymers are completely cleaved from the substrate. We make a simple estimate in order to verify our hypothesis. In Table 15.3 we list the molecular parameters of PtBA and PAA, such as the monomer molecular weight (Mo), the density (), and the volume of the monomer unit (vo). The latter was calculated by using Eq. (19): vo ¼

Mo NA

(19)

where NA is Avogadro’s number. Assuming that the degree of polymerization of PtBA is the same as that of PAA (i.e., the total number of polymer chains upon hydrolysis remains constant), we can use the values in Table 15.3 to estimate the change in thickness of the PtBA film associated with complete conversion of the PtBA into PAA as: v DhPtBA Dvo ¼ ¼ 1 o;PAA » 0:52 hinit;PtBA vo;PtBA vo;PtBA

(20)

where DhPtBA and hinit,PtBA are the PtBA thickness change upon hydrolysis and the initial PtBA thickness, respectively. Table 15.3

Molecular parameters of PtBA and PAA.

Mo (g mol–1) (g cm–3) vo (cm3) * +

PtBA

PAA

128.17 1.05* 2.027 Q 10–22

72.065 1.22+ 9.81 Q 10–23

Value for poly(sec-butyl acrylate) from: J. Brandrup, E. H. Immergut, E. A. Grulke (Eds.): Polymer Handbook, Wiley, New York, 1999. See Ref. [34].

This simple estimate illustrates that complete hydrolysis of PtBA into PAA will result in about 52% decrease of the total PtBA film thickness. This confirms our earlier hypothesis, namely that the rapid drop in polymer thickness within the first 2 h of HCl treatment is associated predominantly with the PtBA to PAA conversion. Based on this estimate, HCl treatment times longer than about 2 h would result in some cleavage of the polymer from the substrate. For example, based on the data in Figure 15.7, after 5 h of hydrolysis, the original thickness of PtBA decreases by 85%.

305

306

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

Dry thickness (a) and contact angle (b) of a PtBA/ PAA sample as a function of the hydrolysis time. The various symbols denote samples with a different initial dry thickness.

Figure 15.7

In Figure 15.7(b) is plotted the negative cosine of the DI water contact angle, hDIW, as a function of the PtBA-to-PAA conversion time. Note that the trend is similar to the PtBA thickness versus time behavior. Specifically, hDIW drops from 88R for a thick PtBA rather rapidly within the first 4 h – a behavior that is associated with the conversion of the tert-butoxy group into the -OH group. The contact angle reaches a minimum at about 5 h, after which time it starts to increase again. The latter can be attributed to the surface exposure of the undecylsilane groups that remained grafted to the substrate after the cleavage of the tert-ester in the initiator molecule. In our gradient samples, we performed the PtBA hydrolysis for 5 h in order to ensure the complete removal of tert-butyl groups. Although some polymer may have been cleaved off the surface during this extended hydrolysis, enough polymer remained on the surface for further analysis. In fact, NEXAFS experiments on PAA samples prepared by hydrolyzing a PtBA sample for about 10 h revealed that a significant amount of PAA still remained on the surface. 15.5.2

Dependence of H on Ionic Strength

In Figure 15.8, is plotted the dependence of the PAA wet thickness (H) on the solution ionic strength (IS) at pH equal to: a) 4, b) 5.8, and c) 10 for three different grafting densities. The squares, circles, and triangles denote grafting density values that are approximately equal for all three samples. Since only NaCl, HCl and NaOH were used to change the solution ionic strength, the salt concentration (s, ¥) in this case is equal to the solution ionic strength (IS). The data in Figure 15.8 reveal that H depends on IS in a non-monotonous fashion. Specifically, as IS increases, H

15.5 Discussion

Wet thickness of PAA (H) as a function of the solution ionic strength (IS) at: (a) pH = 4; (b) pH = 5.8; and (c) pH = 10. The symbols represent different grafting densities of PAA in chains nm–2.

Figure 15.8

increases before reaching a maximum at a certain value, and then starts to decrease. This behavior, which is observed for all pH values at all grafting densities (r), is in accord with the theoretically predicted trends [10] that divide the H versus IS dependence into the osmotic brush (OB) and the salted brush (SB) regimes (cf. Figure 15.1). The ionic strength, at which the transition between the OB and SB regimes occurs (ISmax), is related to r and pH. At pH = 4, ISmax is nearly constant regardless of the r (Figure 15.8(a)). At pH = 5.8, ISmax remains small at low grafting densities, and increases slightly to 0.25 with increasing r (Figure 15.8(b)). At pH = 10, ISmax shifts significantly (Figure 15.8(c)). Specifically, while for low r (» 0.0381 nm–2, not shown) ISmax » 0.25, at high r (» 0.863 nm–2) ISmax»1. Overall, two general trends can be deduced from the data in Figure 15.8. First, with increasing grafting density, the height of the polymer brush increases. Second, at a given value of r, the PAA swelling increases as the solution pH value increases. The latter behavior is associated with the electrostatic charging inside the PAA brush which leads to the increase of the intermolecular repulsions and subsequent brush height increase. The results are in very good agreement with the theoretical prediction by IsraOls et al. [11,12], who proposed that ISmax ~ r(a)1/2. Clearly, ISmax is affected by the solution pH value, which directly influences a. At pH = 4, since a is very small, the grafting density change could not produce any obvious change of r(a)1/2 or ISmax. At pH=10, a is close to 1 (complete ionization), so ISmax ~ r. In Figure 15.9, we plot the ISmax values extracted from Figure 15.8(c) as a function of r for the measurements made at pH = 10. The slope (= 0.925) is close to the expected value of 1. At pH = 5.8, a is about 0.5 [21,22], which explains that the shift in ISmax is relatively small, compared to that at pH = 10.

307

308

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

Values of the ionic strength at the boundary between the OB and SB regimes (ISmax) as a function of grafting density of PAA at pH = 10. Figure 15.9

15.5.3

Dependence of H on the PAA Grafting Density

Earlier, we discussed that weak polyelectrolyte wet thickness has a different dependence on the grafting density for polymers in the SB or OB regimes [6]. In the following section, we will present the results of the experimentally measured H for each regime separately. Since there is no significant variation in the data for different pH values and molecular weights, we will use the results collected at pH = 5.8 and PAA with Mn = 4.8 kDa. Salted Brush (SB) Regime In Figure 15.10(b) is plotted H as a function of the PAA grafting density in the SB regime. The various symbols denote data collected at IS ranging from 0.1 to 0.75. At high polymer grafting densities (r > 0.1 nm–2), H increases with increasing r. This is a typical behavior for the brush conformations. The transition from the brush regime to the mushroom regime occurs at r » 0.08 nm–2. The slope for the brush regime is found to range from 0.29 to 0.31, in good agreement with the theoretically predicted value of 1/3. With increasing IS, H decreases and the slope in the H ~ rn dependence increases. The decrease in polymer swelling is largely due to the screening of the electrostatic interactions by the counter ions inside the polymer brush. The increase in the slope suggests that the solution ions move more easily inside the grafted polymer at lower grafting density. With increasing r, the transport of ions inside the densely packed polymers becomes more difficult. As a consequence, the screening effects weaken. 15.5.3.1

15.5 Discussion

Figure 15.10 Wet thickness at pH = 5.8 for PAA (Mn = 4.8 kDa) as a function of the grafting density and ionic strength of the aqueous solution in the (a) OB regime and (b) SB regime. The symbols represent different IS values.

Osmotic Brush (OB) Regime In Figure 15.10(a) is plotted H as a function of r for IS ranging from 1.56 Q 10–6 to 0.1. We have previously identified that at these IS values the system is in the OB regime. Prior theoretical work predicted that in this regime that wet thickness of polymer brush should decrease with the grafting density as H ~ r–1/3 and should increase with increasing IS [6,11]. Based on theoretical studies, at the transition between the OB to SB regimes (at ISmax), H is independent of the brush grafting density. Similar to earlier experiments by others [21,22], we observe that this scaling relation is somehow flawed. Specifically, by fitting the data in the brush regime to H ~ Nrn, we obtain n that ranges from 0.28 to 0.34, instead of the expected value of –1/3. Close inspection of the data in Figure 15.11 reveals that polymer swelling increases with increasing ionic strength. Interestingly, the value of the exponent n decreases systematically as the solution IS increases. This is in contrast to the performance of PAA in the SB regime, where the value of n increased with increasing IS (cf. Figure 15.10(b)). This behavior reveals that when a small amount of salt is added in the OB regime to polymers with a low r, the grafted polymer swells more relative to PAA at high r. In order to quantify this behavior, we define a degree of swelling ([DS]) of a grafted polymer as: 15.5.3.2

½DS ¼

HðISÞHDIW

100% HDIW

(21)

where H(IS) and HDIW are the PAA thicknesses evaluated at a given IS and in “pure” water (IS!0), respectively. In Figure 15.11 is plotted the degree of swelling

309

310

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

Degree of swelling [H(ISmax) – HDIW]/HDIW for PAA (Mn = 4.8 kDa) as a function of the PAA grafting density in the OB regime at pH equal to (a) 4; (b) 5.8; and (c) 10.

Figure 15.11

at ISmax as a function of the PAA grafting density at different pH values for polymers in the OB regime. By fitting the data to [DS] ~ rn, we find n to be very close to –1/3 in OB regime for pH = 4 and 5.8, and 0 for the pH = 10 data. At low pH, PAA behaves as a weak polyelectrolyte, and the degree of swelling changes with the r. At pH = 10, almost all charges along the polymer backbone are activated and present at the backbone. As a consequence, the polymer behavior closely resembles that of a strong polyelectrolyte, the degree of expansion of which is independent of the polymer grafting density. 15.5.4

Molecular Insight from Calculations

Figure 15.12 presents a comparison of the height of the polymer layer as measured experimentally with the theoretical predictions. The agreement between the two is excellent, considering that there are no fitting parameters used in the calculations. The presence of the maximum has been predicted by the theory of IsraOls and coworkers [8–12], which is the result of the different regimes of the weak polyelectrolyte brush. The importance of our results is that they have been carried out for exactly the same system as the experimental observations. The quantitative agreement between the experimental and calculated brush height validates the prediction of other brush properties (discussed below) obtained from the theory. This is particularly important for the relatively short chain lengths used in the experimental systems here, where scaling-type theories are known to be at best qualitatively correct. We concentrate our attention to show the very large pH gradient that the presence of the polymer layer induces for different bulk salt concentrations. Furthermore, we

15.5 Discussion

Figure 15.12 The height of the polymer as a function of the bulk salt concentration for pH = 4 and r = 0.103 nm–2. The predictions of the theory are depicted by the line, while the symbols are the experimental observations. The calculations use Ka = 10–4 M for N = 50.

will see how the pH gradient induces large changes in the polymer charge, charge distribution and the volume fraction profile. Figure 15.13(a) shows pH as a function of the distance from the surface for four different salt concentrations. As the bulk salt concentration decreases, there are two very important changes in the pH profile. First, the range of inhomogeneity changes more or less according to the change in the Debye length. Namely, as the salt concentration decreases, the charge on the surface due to the charged grafted polymers is felt in a range determined by the screening of the electrostatic interactions. Second, the concentration of protons increases by two orders of magnitude from its bulk value as the salt concentration decreases from 1 M to 0.0005 M. The increase in proton concentration (pH decrease) is due to the decrease of cations from the salt that can adsorb on the grafted layer to reduce the electrostatic interactions. The increase in proton concentration results in a decrease of the number of acrylic acid group charged due to the acid-base equilibrium. We analyze this effect next. Figure 15.13(b) shows the dissociation a(z) calculated using Eq. (17) as a function of the distance from the surface. For the pH of the experiments and the pH of the acrylic acid in the polymer, the bulk dissociation is basically 100%. However, we see that in the region of the pH gradient shown in Figure 15.13(a) results in very low dissociations, particularly at low s, ¥ and in the region where there is polymer. Note that for the lowest salt concentration shown, the dissociation is reduced to 10%. This should have dramatic effects on the structure of the polymer layer, as will be discussed next. Figure 15.14 shows the average charge on the polymer as a function of the distance from the surface. There is a clear change in the overall charge on the polymer

311

312

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

Figure 15.13 (a) The concentration of protons and (b) the local dissociation as a function of the distance from the surface for four different values of s, ¥: 1 M (solid line), 0.1 M (dashed line), 0.0075 M (dot-dashed line), and

0.0005 M (dotted line). The corresponding values of k–1 are: 0.304 nm (for 1 M), 0.961 nm (for 0.1 M), 3.510 nm (for 0.0075 M), and 13.60 nm (for 0.0005 M).

Figure 15.14 The average charge of the polymer as a function of the distance from the surface. The symbols are the same as in Figure 15.13.

15.5 Discussion

layer, as well as in the distribution of charged groups. For the largest salt concentration the polymer is highly charged but, as the salt concentration decreases, the charging decreases by a factor of 5 for the salt range shown.

Summary

We have studied the scaling laws of surface-grafted polyacrylic acid (PAA) as a function of the polymer grafting density (r), solution ionic strength (IS), and pH. In order to facilitate the complete exploration of the r space, we created surface-grafted PAA on flat silica-covered substrate with a spatial variation of the chain grafting density. The surface-bound PAA with a gradual variation of grafting densities was formed by: 1) creating a molecular density gradient of the surface-anchored polymerization initiator; 2) ATRP synthesis of poly(tert-butyl acrylate) (PtBA) by grafting from the surface; and 3) converting the PtBA into PAA by hydrolysis. We used spectroscopic ellipsometry to measure the wet thickness of the PAA as a function of r, IS, and pH. The wet thickness (H) of the surface-grafted PAA brushes was found to have a nonmonotonous dependence on the ionic strength (IS) of the solution. By increasing the concentration of the external salt, the polymer thickness in solution increased and reached a maximum at a certain ionic strength (ISmax), and then further decreased. Guided by the theoretical models of weak polyelectrolyte brushes, we have identified three regimes: the osmotic brush (OB), the salted brush (SB), and the neutral brush (NB) regime. We have discussed how H behaves at different r in the SB and OB regimes. By comparing the swelling of polymer under different pH solution conditions, we concluded that the expansion of the grafted chain at low pH value was much less than that at high pH solution. In the SB regime, H was found to increase with increasing r at high polymer grafting densities, a typical behavior for polymer brushes. The slope for the brush regime ranged from 0.29 to 0.31, in a good agreement with the theoretically predicted value of 1/3. The transition from the brush to the mushroom regime was found to occur at r » 0.08 nm–2. We also established that the slope of H increased with increasing IS, and this behavior was attributed to the less efficient screening effects from solution ions at higher grafting densities. At low IS, the system was in the OB regime. Here, the wet PAA thickness was found to depend strongly on r and pH. Our data revealed that at high r, H followed the scaling law H ~ rn, with n ranging from 0.28 to 0.34. We commented that this observation was in contrast to the theory, which predicts that in the OB regime H ~ r–1/3. We also observed that the degree of polymer swelling increased with increasing IS. The exponent in the H ~ rn dependence decreased with increasing IS. This behavior was exactly opposite to that detected in the SB regime, where n increased with increasing IS. We defined a degree of swelling [DS] parameter as [H(ISmax) – HDIW]/HDIW, where H(IS) and HDIW are the PAA thicknesses evaluated at a given ISmax and in “pure” water (IS!0), respectively. By fitting the data to [DS] ~ rn, we found n to be very close to –1/3 in the OB regime for pH = 4 and 5.8

313

314

15 Behavior of Surface-Anchored Poly(acrylic acid) Brushes with Grafting Density Gradients

and 0 at pH = 10. We rationalized that this behavior was a consequence of the conformational changes in the polymer associated with the concentration of the charges along the backbone. At low pH, not all charges were activated, and PAA behaved as a typical weak polyelectrolyte and [DS] increased with decreasing r. In contrast, at high pH, the whole polymer backbone was decorated with a large number of charges that stayed permanently attached to the backbone. Consequently, PAA behaved like a strong polyelectrolyte, the degree of expansion of which is independent of the grafting density. We also found that the value of the ionic strength at the OB to SB transition (ISmax) depends on the polymer grafting density and the solution pH. At pH = 4, ISmax is independent of r. As the solution pH increases, ISmax also increases; moreover, ISmax also increases with increasing r. These results are in a good agreement with the theoretical predicted scaling law ISmax ~ r(a)1/2. The predictions of the molecular theory can be summarized in the following manner. First, the theory is capable of reproducing the experimental observations in a quantitative manner, without the use of adjustable parameters. The qualitative changes in the structure of the layer are in line with those previously predicted by analytical and numerical self-consistent field theory. What is different here is the ability to predict the behavior of these systems accurately for the exact molecular weight, surface coverage and bulk conditions used in the experimental systems. The agreement between the theory and the experimental observations strongly supports the use of a molecular approach as a predictive tool for both the structural and thermodynamic properties of the weak polyelectrolyte brushes. This is in addition to the previously proven ability of the theory to predict the behavior of uncharged polymer layers. Second, changing the salt concentration of the bulk solutions results in a dramatic change of the local pH. By changing the salt concentration, one can obtain an increase in the local concentration of protons of three orders of magnitude in a length scale of a few nanometers.

Acknowledgments

The studies conducted at NC State University were supported by the National Science Foundation, Grant No. CTS-0209403, The Camille Dreyfus Teacher-Scholar award, and The 3M Non-Tenured Faculty award. The studies at Purdue were supported by the National Science Foundation, Grant No. CTS-0001526, while those at IMC were supported by the Grant Agency of the Czech Republic, Grant # 203/01/ 0513. The NEXAFS experiments were carried out at the National Synchrotron Light Source, Brookhaven National Laboratory, which is supported by the U.S. Department of Energy. The authors thank Dr. Kirill Efimenko (NCSU) and Dr. Daniel Fischer (NIST) for their assistance during the course of the NEXAFS experiments.

References

References 1 A. Halperin, M. Tirrell, T. P. Lodge, Adv.

Polym. Sci. 1991, 100, 31. 2 S. Alexander, J. Phys. 1997, 38, 983; P.-G. De Gennes, J. Phys. (Paris) 1976, 37, 1445; P.-G. De Gennes, Macromolecules 1980, 13, 1069. 3 S.T. Milner, Science 1991, 251, 905. 4 I. Szleifer, M. A. Carignano, Adv. Chem. Phys. 1996, 96, 165. 5 P. Pincus, Macromolecules 1991, 24, 2912. 6 E. B. Zhulina, T. M. Birshtein, O. V. Borisov, Macromolecules 1995, 28, 1491. 7 E. B. Zhulina, O. V. Borisov, J. Chem. Phys. 1997, 107, 5952. 8 R. IsraOls, J. M. H. M. Scheutjens, G. J. Fleer, Macromolecules 1994, 27, 3087. 9 R. IsraOls, F. A. M. Leermakers, G. J. Fleer, E. B. Zhulina, Macromolecules 1994, 27, 3249. 10 G. J. Fleer, Ber. Bunsenges, Phys. Chem. 1996, 100, 936. 11 R. IsraOls, F. A. M. Leermakers, G. J. Fleer, Macromolecules 1994, 27, 3087. 12 R. IsraOls, F. A. M. Leermakers, G. J. Fleer, Macromolecules 1995, 28, 1626. 13 Yu.V. Lyatskaya, F. A. M. Leermarkers, G. J. Fleer, E. B. Zhulina, T. M. Birshtein, Macromolecules 1995, 28, 3562. 14 T. W. Kelly, P. A. Schorr, K. D. Johnson, M. Tirrell, D. C. Frisbie, Macromolecules 1998, 31, 4297. 15 Y. Mir, P. Auroy, L. Auvray, Phys. Rev. Lett. 1995, 75, 2863. 16 P. Guenoun, F. Muller, M. Delsanti, L. Auvray, Y. J. Chen, J. W. Mays, M. Tirrell, Phys. Rev. Lett. 1998, 81, 3872. 17 M. Biesalski, J. R]he, Macromolecules 1999, 32, 2309. 18 M. Biesalski, J. R]he, Langmuir 2000, 16, 1943. 19 M. Biesalski, J. R]he, Macromolecules 2002, 35, 499. 20 K. Kurihara, T. Kunitake, N. Higashi, M. Niwa, Langmuir 1992, 8, 2087. 21 E. P. K. Currie, A. B. Sieval, G. J. Fleer, M. A. Cohen Stuart, Langmuir 1999, 15, 7116. 22 E. P. K. Currie, A. B. Sieval, G. J. Fleer, M. A. Cohen Stuart, Langmuir 2000, 16, 8324.

23 T. E. Patter, J. Xia, T. Abernathy, K. Matyjas-

zewski, Science 1996, 272, 866. 24 V. Coessens, T. Pintauer, K. Matyjaszewski,

Prog. Polym. Sci. 2001, 26, 337. 25 T. Wu, K. Efimenko, J. Genzer, J. Am. Chem.

Soc. 2002, 124, 9394. 26 K. Matyjaszewski, P. J. Miller, N. Shukla,

B. Immaraporn, A. Gelman, B. B. Luokala, T. M. Siclovan, G. Kickelbick, T. Vallant, H. Hoffmann, T. Pakula, Macromolecules 1999, 32, 8716. 27 K. A. Davis, K. Matyjaszewski, Macromolecules 2000, 33, 4039. 28 K. Efimenko, W. E. Wallace, J. Genzer, J. Colloid Interface Sci. 2002, 254, 306. 29 M. K. Chaudhury, G. M. Whitesides, Science 1992, 256, 1539. 30 M. Husseman, E. E. Malmstrom; M. McNamara, M. Mate, D., Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang, T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424. 31 L. Mrkvickov^, J. Danhelka, Appl. Polym. Sci. 1990, 41, 1929. 32 This is the value for poly(n-butyl acrylate) from: J. Brandrup, E. H. Immergut, E. A. Grulke (Eds.): Polymer Handbook, Wiley, New York, 1999. 33 J. Brandrup, E. H. Immergut, E. A. Grulke (Eds.): Polymer Handbook, Wiley, New York, 1999. 34 T. Wu, K. Efimenko, P. Vlcek, V. _ubr, J. Genzer, Macromolecules 2003, 36, 2448. 35 J. St`hr, NEXAFS Spectroscopy, SpringerVerlag, Berlin, 1992. 36 M. C. Faure, P. Bassereau, M. A. Carignano, I. Szleifer, Y. Gallot, D. Andelman, Eur. Phys. J. B 1998, 3, 365. 37 I. Szleifer, Curr. Opin. Colloid Interface Sci. 1996, 1, 416. 38 T. McPherson, A. Kidane, I. Szleifer, K. Park, Langmuir 1998, 14, 176. 39 M. A. Carignano, I. Szleifer, Mol. Phys. 2002, 100, 2993; F. Fang, I. Szleifer, J. Chem. Phys. 2003, 119, 1053. 40 I. Szleifer, Biophysical J. 1997, 72, 595.

315

317

16

Kinetics of Polymer Brush Formation With and Without Segmental Adsorption Lynn S. Penn, Heqing Huang, Roderic P. Quirk, and Tae H. Cheong

Abstract

The kinetics of tethering of chain-end functionalized polystyrene to the surface of an impenetrable solid was investigated. Tethering in the absence of segmental adsorption and in the presence of segmental adsorption was monitored quantitatively in real time. Conditions of no segmental adsorption gave three distinct regimes of kinetics prior to saturation, in contrast to the two-regime kinetics predicted by theory. The largest increases in surface attachment density (chains nm–2) took place in the first and third regimes. The third regime was identified as the one in which the transition from mushroom to brush conformation took place. The kinetics of tethering under conditions that allowed segmental adsorption was entirely different. Rather than distinct regimes of kinetics, the system approached saturation smoothly and reached saturation much faster. The final surface attachment density for a tethered layer formed in the presence of segmental adsorption was about twice that of a tethered layer formed in the absence of segmental adsorption. Either type of kinetics has potential advantages for the experimenter in the preparation of complex, as-designed tethered layers. The first can be exploited for construction of mixed tethered layers, while the second can be exploited for construction of denser polymer brushes.

16.1

Introduction

There are two basic approaches used for introducing polymer chains irreversibly tethered by one end to the surface of a solid. These are broadly termed “graft polymerization” and “polymer grafting”. In graft polymerization, initiator sites are installed on the surface of the solid, and the polymerization of monomers in the surrounding solution is initiated from these sites. In polymer grafting, chemically active sites are installed on the surface of the solid, and the functional ends of already-formed polymer chains in the surrounding solution react chemically with these active sites. The result for both approaches is a solid surface to which many

318

16 Kinetics of Polymer Brush Formation With and Without Segmental Adsorption

polymer chains are all permanently attached by one end. Both approaches can lead to a polymer brush if enough chains are grafted per unit surface area, so that the distance between chains is less than twice the radius of gyration of the polymer. A widely held current opinion is that extremely dense polymer brushes have wider potential applicability and therefore are more desirable than moderately dense brushes. Because, for steric reasons, denser polymer brushes can be formed by means of graft polymerization, more attention is currently focused on graft polymerization than on polymer grafting. However, there are some applications, such as adhesion promotion [1,2], where the brushes need to be able to be easily penetrated by other materials, or bioseparations applications [3–5], where the brushes need to be derivatized with large-sized moieties, that do not require dense brushes. In these cases, moderately dense brushes are the most appropriate, and they are best formed by polymer grafting. Three features of the polymer grafting approach provide versatility in the construction of polymer brushes. First, the surface attachment density can be controlled easily by controlling the grafting time. Second, mixed brushes of controlled proportions can be formed easily by grafting one polymer until a limited surface attachment density is reached, followed by grafting another polymer until the brush stage is reached. Third, a brush of a precise molecular weight can be constructed by use of chain-end functionalized polymer of the selected molecular weight in solution. Obviously, these three features can be combined for construction of multifunctional brushes of complex structure. It is because of the potential for formation of complex brushes that our attention has been drawn to polymer brushes formed by polymer grafting, which we refer to as tethering throughout the remainder of this chapter. Since the attributes of the fully formed polymer brush have been well studied by others, both experimentally [6–28] and theoretically [29–46], we have concentrated on the kinetics of tethering, which has received relatively little attention by comparison. Experimental studies of the kinetics of processes that take place on surfaces are difficult, because the amount of material reacting with the surface is a minute portion of the total mass of the system and therefore approaches the limits of sensitivity of quantitative analysis techniques. However, such studies of kinetics are well worth the effort, because they provide the understanding needed for manipulation of the process to achieve complex, as-designed tethered layers. The kinetics of irreversible tethering of chain-end functionalized polymers from solution has been addressed theoretically [47–49]. Two distinct regimes of kinetics are predicted for the process, as explained briefly in the following discussion. According to theory, the first regime consists of rapid tethering at a rate controlled by center-of-mass diffusion of the chains through the solvent to the bare surface. An abrupt slowdown in rate occurs, ending the first regime, when the surface of the substrate is covered by a layer of nonoverlapping chains, each in the expanded coil, or mushroom, conformation [47–49]. Theory explains that the source of decrease in rate is the mushroom layer itself; its presence introduces an activation barrier to the diffusion of additional free chains to the surface. The predicted second regime is one of slow tethering, at a rate proportional to ln(time). According to theory, the proportionality to ln(time) is associated with a progressive increase in the activation bar-

16.1 Introduction

rier to diffusion as the surface attachment density of the tethered chains increases progressively with time [47]. The second regime is expected to continue until saturation is reached – that is, until further tethering is no longer possible from a thermodynamic standpoint. This point is reached when the energy benefit of chemical bond formation with the surface is offset by the various entropy costs of tethering [47]. The above-described theoretical concepts of the kinetics of tethering provide a backdrop against which experimental studies can be viewed. The two-regime profile of kinetics generated by theory rests on the assumptions that the polymer chains are tethered from solution, by one end, and that no segmental adsorption of polymer to substrate occurs. The present chapter describes experimental studies in which chain-end functionalized polystyrene was tethered from solution in the absence of segmental adsorption and, separately, in the presence of segmental adsorption. What we found was that, surprisingly, tethering in the absence of segmental adsorption did not duplicate the theoretical predictions of two regimes of kinetics, but instead exhibited three regimes. We also found that, as expected, tethering in the presence of segmental adsorption was entirely different than in the absence of segmental adsorption. A hallmark of our studies of the kinetics of tethering is the use of well-characterized and well-controlled polymers and substrates [1,50–52]. Typically, we use monodisperse, chain-end functionalized polymers, and we investigate one molecular weight at a time. As the end-functionalized polymer chains become tethered to the substrate, their depletion from solution is monitored quantitatively in real time. Figure 16.1 shows the chemical reaction between an active site on the surface of the substrate and the functional end of a polymer chain. This reaction occurs readily, even at room temperature, and the resultant chemical bond ensures that the chain is tethered irreversibly. Thus, one end of each tethered chain is fixed to the surface of the substrate, and the other end (not functionalized) is free to explore distances equal or less than the length of the fully stretched chain. O HC

CH2 + H2N

PS

OH HC

CH2 NH

PS

Chemical reaction by which chain-end functionalized-ended polystyrene (PS-NH2) is tethered to the surface of the substrate.

Figure 16.1

319

320

16 Kinetics of Polymer Brush Formation With and Without Segmental Adsorption

16.2

Experimental 16.2.1

Synthesis and Characterization of Amine Chain-End Functionalized Polystyrene

Monodisperse, x-(3-aminopropyl)polystyrene was prepared by means of anionic living polymerization methods under standard high-vacuum conditions in sealed, allglass reactors equipped with breakseals as described elsewhere [53–55]. Polydispersity, Mw/Mn, achieved by this means was <1.04. The primary amine end-group was -(-CH2-)3-NH2. The fraction of chains containing primary amine end-groups was determined by titration in 1/1 (v/v) mixture of chloroform in glacial acetic acid, with perchloric acid (Fisher; 10.1 N in glacial acetic acid) as the titrant and methyl violet 2B (Aldrich; ~75% dye content) as the indicator [56]. Five to seven 0.5-mg samples were analyzed from each batch of polymer, and end-functionalization was found to be >95% for all batches of polymer. 16.2.2

Introduction of Active Sites to Surface of Solid

The solid substrate used was silicate glass in the form of nonporous, spherical beads (Potters Industries, Cleveland, OH, USA) of specific surface area of 0.24 m2 g–1. The beads were cleaned with piranha solution and dried. Epoxide groups were then introduced by exposure to 3-glycidoxypropyltrimethoxysilane (98%; Aldrich, Minneapolis, MN, USA) in toluene under anhydrous conditions. Any unreacted silane that had been adsorbed to the surface was removed by overnight extraction in refluxing toluene in a Soxhlet apparatus. The derivatization procedure resulted in 2.71 – 0.24 epoxide groups per nm2 of glass surface – a surface density value well above that needed for tethering of polymer chains at the highest conceivable surface density [57]. 16.2.3

Tethering Reactions in Good Solvent

Before any tethering reactions were conducted, complete absence of segmental adsorption in good solvent was confirmed by auxiliary experiments [50,52]. These consisted of the exposure of the epoxide-derivatized beads to toluene solutions of polystyrene (with inert, or nonfunctional, chain ends) of two different molecular weights. It is known that adsorption causes the ratio of two different molecular weights of chemically identical dissolved polymer to change over time, because lowmolecular-weight polymer adsorbs at an early time and is displaced by high-molecular-weight polymer after a longer time [58]. Therefore, the absolute and relative values of the two molecular weights in solution were monitored continuously over many hours; all values remained constant over time, confirming the absence of segmental adsorption.

16.2 Experimental

Tethering reactions were run at room temperature, under an argon atmosphere, in glassware that had been previously treated with n-butyltrichlorosilane (an agent that reduces surface energy and prevents segmental adsorption of polystyrene to the glassware). The details of a typical tethering reaction are as follows. A reaction flask was charged with 5.1 mg of amine chain-end functionalized polystyrene of the desired Mn dissolved in 20 mL of reagent-grade toluene. This solution was spiked with a precisely known weight of internal standard. Alkyl-terminated, monodisperse polystyrene (Polymer Standards Service, Silver Spring, MD, USA) was used as an internal standard. (Since the alkyl chain-ends render the internal standard inert to tethering reactions, we term these chains “inert-ended.”) It is important to remember that the internal standard, which is unable to exhibit segmental adsorption in good solvent or to exhibit chemical reaction of its chain ends with the surface of the substrate, remains in solution throughout the tethering process. Before the solid substrate was combined with the polymer solution, two or three aliquots of the solution were subjected to size-exclusion chromatography (SEC) to establish the peak area ratio for chain-end functionalized polymer to internal standard at zero time. (Mass ratio was known from the initial weighing operation.) Then, 18.1 g of surface-derivatized glass beads was combined all at once with the solution to the flask, and the tethering process began immediately. 16.2.4

Tethering Reactions in Poor Solvent

Tethering reactions in poor solvent were set up as described above for good solvent, except that reagent-grade cyclohexane was used as the solvent instead of toluene. Duplicate reactions, each containing a different internal standard, were run. Monodisperse, inert-ended polystyrene served as the internal standard in one reaction, while monodisperse, inert-ended polyisoprene (Polymer Standards Service) served as the internal standard in the duplicate reaction. To be successful as an internal standard, a substance must remain completely in solution throughout the tethering reaction. Auxiliary experiments similar to those used for polystyrene were conducted to verify that polyisoprene remained completely solubilized in cyclohexane, and that segmental adsorption to the substrate or the glassware did not occur. 16.2.5

Monitoring the Tethering Reactions Reactions Run in Good Solvent Representative aliquots (each ~0.3 mL) containing glass beads as well as polymer solution, were removed from the stirring reaction mixture at frequent intervals for quantitative analysis of chains remaining in solution. Immediately after removal from the reaction vessel, each aliquot was treated with a 100-fold excess of trichloroacetylisocyanate to quench the tethering reaction by capping the functional end of each polymer chain. Next, the beads were removed from the aliquot by means of a syringe filter, leaving behind a clear solution that contained the chains not yet teth16.2.5.1

321

322

16 Kinetics of Polymer Brush Formation With and Without Segmental Adsorption

ered and the internal standard. This solution was analyzed by means of SEC using a Waters LC system (Waters Corp., Milford, MA, USA) equipped with two Styragel columns (HR1 and HR3) and ultraviolet and refractive index detectors. From the chromatogram, the mass of chain-end functionalized polystyrene remaining in solution was determined relative to that of the internal standard by comparison of peak areas. Relative peak area at sampling time, t, was normalized by the relative peak area at t = 0 (before addition of beads) to yield mass fraction in solution at time, t. The mass tethered was determined by difference between the mass in solution at t = 0 and sampling time, t. Mass tethered was converted to number of chains tethered by use of the known value of Mn, and division by the total surface area of the substrate yielded surface attachment density (as chains nm–2). Reactions Run in Poor Solvent At frequent intervals, representative aliquots were removed from the duplicate stirring reaction mixtures and immediately quenched with excess trichloroacetylisocyanate, as described above. After quenching, the aliquots removed from reactions containing polystyrene as internal standard and from those containing polyisoprene as internal standard were treated differently. In the former case, the cyclohexane was evaporated and replaced with an equal amount of toluene. The purpose of the toluene was to desorb the polystyrene internal standard and nontethered, chain-end functionalized polystyrene from the beads. (The ability of toluene to completely and quickly desorb segmentally adsorbed polystyrene from epoxide-derivatized surfaces was verified in separate experiments.) Then the beads, containing only those chains that were tethered, were removed from the aliquot by means of a syringe filter, and the clear toluene solution was analyzed by SEC. From the chromatogram, the mass fraction of chain-end functionalized polystyrene chains in solution was determined as described above, and the number of chains tethered (chemically bonded) to the surface of the substrate at the time the aliquot was taken was computed by difference. In the case of aliquots containing polyisoprene as the internal standard, the cyclohexane was not evaporated. Rather, the beads were separated by syringe filter, and the clear cyclohexane solution was analyzed by SEC. This solution contained, in addition to the internal standard (polyisoprene), only those end-functionalized polystyrene chains that were neither tethered nor adsorbed to the substrate. The mass of chain-end functionalized polystyrene remaining in solution at time, t, was computed as described above. Then, the mass of the chain-end functionalized polystyrene both tethered and adsorbed at the time the aliquot was taken was computed by difference. 16.2.5.2

16.3 Results and Discussion

16.3

Results and Discussion 16.3.1

Results in Absence of Segmental Adsorption

2

Surface Attachment Density (chains/nm )

The experimental results for tethering in the absence of segmental adsorption (good solvent) are presented first, and are compared with the kinetics behavior predicted by theory. After this, tethering in the presence of segmental adsorption (poor solvent) is described and is compared with tethering in the absence of segmental adsorption. Figure 16.2 shows typical results for tethering of chain-end functionalized polystyrene (in this case Mn = 4000) from good solvent to the surface of silicate glass substrate. As discussed in the Section 16.2, chain-end functionalized polystyrene exhibited no segmental adsorption whatsoever. In the figure, surface attachment density is plotted versus time, with each data point representing a single measurement. The final surface attachment density shown for Mn = 4000 is consistent with that of a brush according to the often used criterion d < 2Rg [49,59–61], where d is the average distance between tethering points and Rg is the radius of gyration of the polymer in good solvent [62,63]. For the experimental system shown in Figure 16.2, d = 3.9 nm and 2Rg = 4.14 nm, indicating that the final tethered layer is a brush. The results of tethering in the absence of segmental adsorption need to be viewed against the backdrop of the two-regime kinetics provided by theory. The experimen0.08

0.06 rd

3

Saturation

0.04 nd

st

2

1 0.02

0.00

0

10

2

10

3

4

10

Reaction Time (min) Plot of surface attachment density versus time for tethering of monodisperse, chain-end functionalized polystyrene, Mn = 4000, in the absence of segmental adsorption. The x-axis for the first regime is linear, while for the remainder it is logarithmic. Each data point represents one measurement. Figure 16.2

323

324

16 Kinetics of Polymer Brush Formation With and Without Segmental Adsorption

tal system in Figure 16.2 at first appears to follow the behavior predicted by theory; that is, a fast first regime followed by a slow second regime proportional to log(time). However, in disagreement with theoretical predictions, the second regime does not continue steadily until saturation is reached, but is interrupted by an acceleration in rate. The accelerated tethering, like that immediately preceding it, is also proportional to log(time), but has a significantly greater slope. This change in slope distinguishes the experimentally observed accelerated tethering as a third regime, distinct from the second. Figure 16.2 shows that the third regime accounts for a large percentage of the increase in tethered chain density prior to saturation, while the second regime, although lengthy, accounts for only a small percentage of the increase. It is the (unpredicted) third regime then – and not the second – that leads to saturation. The constancy of the last several data points in the plot verifies that saturation has been reached. The three regimes and saturation are labeled in Figure 16.2. The three-regime kinetics was exhibited by other molecular weights [52], indicating that the behavior is general for systems in which segmental adsorption is absent. The previously unpredicted third regime warrants some discussion. The starting point for this discussion is the second regime, where the surface attachment density grows extremely slowly, in proportion to log(time). Each added polymer chain can be assumed to diffuse up to and through the tethered layer at a random location, leading to a spatially uniform increase in surface attachment density in the second regime. The experimentally observed increase in slope on the log(time) axis that marks the third regime suggests a switch to a spatially nonuniform process for the following reasons. First, the mathematical meaning of a higher slope is that the incremental increase in the diffusion barrier presented to incoming chains is less than it was for the lower slope. Second, it is a physical impossibility for the incremental increase in diffusion barrier to lessen if the surface attachment density is increasing uniformly everywhere on the surface of the substrate. A Monte Carlo simulation of tethering as a random sequential adsorption process [64–66] duplicated the three-regime kinetics found experimentally and suggested a cooperative mechanism for the spatially nonuniform increase in surface attachment density in the third regime [52]. The three-regime kinetics offers a valuable advantage for manipulation of the tethering process to create more complex tethered layers. As can be seen from Figure 16.2, about half of the final surface attachment density is attained at the end of the first regime. The second regime, in which relatively little additional tethering takes place, can be regarded as a wide window of opportunity for stopping the tethering at a low value of surface attachment density, or for changing polymer solutions to tether a second polymer of different chemical structure or different molecular weight for construction of mixed layers.

16.3 Results and Discussion

16.3.2

Results in the Presence of Segmental Adsorption

When segmental adsorption occurs simultaneously with tethering, the kinetics is very different from that described above for the experimental system with no segmental adsorption. Results for tethering in the presence of segmental adsorption are presented in Figure 16.3. This figure shows a plot of surface attachment density versus time for Mn = 4000. Chain-end functionalized polymer chains that were merely adsorbed but not tethered were subtracted from each measurement, so that the figure reports only the number of chains actually tethered (chemically bonded by their end-functional groups) to the surface of the substrate. As expected, the appearance of this plot is very different from that in Figure 16.2. There is no evidence of distinct regimes of kinetics; rather, tethering appears to proceed smoothly from beginning to end, and tethering rate declines steadily as saturation is approached. In addition to preventing the appearance of distinct regimes of kinetics, segmental adsorption appears to drive the surface attachment density (of the tethered chains alone) to higher levels. For Mn = 4000, the surface attachment density at saturation went from 0.067 to 0.11 chains nm–2 when the tethering process was conducted in the presence of segmental adsorption. This near two-fold increase was also found for other molecular weights. Also, the distance between tethering points was d = 3.0 nm for tethering from poor solvent, which indicates that an even denser brush was formed in the presence of segmental adsorption than without it. Another apparent consequence of segmental adsorption is that saturation is reached earlier. This can be seen by comparing Figures 16.2 and 16.3. This compar-

2

Surface Tethering Density (chains/nm )

0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 0

100

1000

Reaction Time (min) Plot of surface attachment density versus time for tethering of monodisperse, chain-end functionalized polystyrene, Mn = 4000, in the presence of segmental adsorption.

Figure 16.3

325

326

16 Kinetics of Polymer Brush Formation With and Without Segmental Adsorption

ison reveals that Mn = 4000 reached saturation in about 6000 min in the absence of segmental adsorption (good solvent), whereas it took only 600 min when segmental adsorption was present (poor solvent). This difference is far too large to be explained simply by an increase in the rate of diffusion through solution due to a smaller radius of gyration of a polymer chain in poor solvent than in good solvent. It is more likely that segmental absorption enhances the rate of tethering by increasing the concentration of polymer chains in the vicinity of the surface. Even the adsorption of only a few segments of a free chain would keep the chain in close proximity to the surface. Also, segmental adsorption prevents formation of a mushroom layer and the diffusion barrier it presents to additional incoming chains. Instead, segmental adsorption would lead to more of a pancake conformation [59], and incoming chains would be able to get closer to the surface of the substrate without impairment. Solvent quality might also play a small role, in that polymer-polymer interactions would be preferred over polymer-solvent interactions in poor solvent, possibly leading to a tendency for small clusters of polymer to form on the surface of the substrate, again leading to enhancement of tethering. Although tethering in the presence of segmental adsorption does not offer the distinct three regimes of kinetics, it does have the advantage of yielding nearly twice the final surface attachment density. After tethering is complete, replacement of the original solvent with a solvent in which no segmental adsorption occurs would develop a denser polymer brush than could be constructed directly.

Summary

Investigations of the tethering of monodisperse, chain-end functionalized polystyrene to the activated surface of a silicate glass substrate were conducted. The tethering reactions were monitored quantitatively in real time so that the kinetics of tethering under different conditions could be determined. The kinetics of tethering under conditions of no segmental adsorption gave three distinct regimes of kinetics prior to saturation, in contrast to the two-regime kinetics predicted by theory. The largest increases in surface attachment density (chains nm–2) took place during the first and third regimes. Relatively few chains were tethered in the second regime, even though it was lengthy. This three-regime behavior can be exploited by the experimenter to construct more complex tethered layers. The kinetics of tethering under conditions that allowed segmental adsorption was entirely different. There were no distinct regimes of kinetics, and the process reached saturation much earlier. The final surface attachment density for a tethered layer formed in the presence of segmental adsorption was about twice that of a tethered layer formed in the absence of segmental adsorption, and this can be exploited to achieve much denser polymer brushes.

References

Acknowledgments

These studies were supported in part by grants CTS 9911181 and CTS 0218977 from the National Science Foundation.

References 1 R. Lin, R. P. Quirk, J. Kuang, L. S. Penn,

J. Adh. Sci. Technol. 1996, 10, 341–349. 2 F. Brochard-Wyart, P. G. de Gennes, L. Leger, Y. Marciano, E. Raphael, J. Phys. Chem. 1994, 98, 9405–9410. 3 J. R. Ramirez-Vick, A. A. Garcia, Sep. Purif. Methods 1997, 25, 85–129. 4 F. Denizli, M. Y. Arica, A. Denizli, Reactive and Functional Polym. 2000, 44, 207–217. 5 A. Denizli, K. Kesenci, M. Y. Arica, B. Salih, V. Hasirci, E. Piskin, Talanta 1998, 46, 551– 558. 6 J. Klein, Nature, 1980, 288, 248–249. 7 G. Hadziioannou, S. Patel, S. Granick, M. Tirrell, J. Am. Chem. Soc. 1986, 108, 2869–2876. 8 H. Taunton, C. Toprakcioglu, J. Klein, Macromolecules 1988, 21, 3333–3336. 9 H. Taunton, C. Toprakcioglu, L. J. Fetters, J. Klein, Nature 1988, 332, 712–714. 10 H. Taunton, C. Toprakcioglu, L. J. Fetters, J. Klein, Macromolecules 1990, 23, 571–580. 11 J. Klein, D. Perahia, S. Warburg, Nature, 1991, 352, 143–145. 12 P. Auroy, L. Auvray, L. Leger, Phys. Rev. Lett. 1991, 66, 719–722. 13 A. Budkowski, U. Steiner, J. Klein, L. J. Fetters, Europhys. Lett. 1992, 20, 499–504. 14 E. Kumacheva, J. Klein, P. Pincus, L. J. Fetters, Macromolecules 1993, 26, 6477–6482. 15 B. L. Carvalho, P. Tong, J. S. Huang, T. A. Witten, L. J. Fetters, Macromolecules 1993, 26, 4632–4639. 16 T. Cosgrove, J. S. Phipps, R. M. Richardson, M. L. Hair, D. A. Guzonas, Macromolecules 1993, 26, 4363–4367. 17 T. L. Mansfield, D. R. Iyengar, G. Beaucage, T. J. McCarthy, R. S. Stein, R. J. Composto, Macromolecules 1995, 28, 492–499. 18 G. Fytas, S. P. Anastasiadis, R. Seghrouchni, D. Vlassopoulos, J. Li, B. J. Factor, W. Theo-

19 20

21

22

23

24 25 26

27

28

29 30 31 32 33 34

bald, C. Toprakcioglu, Science 1996, 274, 2041–2044. A. Dhinojwala, S. Granick, Macromolecules 1997, 30, 1079–1085. K. Koutsos, E. W. van der Vegte, E. Pelletier, A. Stamouli, G. Hadziioannou, Macromolecules 1997, 30, 4719–4726. S. K.Satija, P. D. Gallagher, A. Karim, L. J. Fetters, Physica B: Condensed Matter 1998, 248, 204–207. D. L. Anastassopoulos, A. A. Vradis, C. Toprakcioglu,G. S. Smith, L. Dai, Macromolecules 1998, 31, 9369–9371. K. Koutsos, E. W. van der Vegte, G. Hadziioannou, Macromolecules 1999, 32, 1233– 1236. L. L. Cai, S. Granick, Adv. Colloid Interf. Sci. 2001, 94, 135–150. S. M. Kilbey, H. Watanabe, M. Tirrell, Macromolecules 2001, 34, 5249–5259. K. Urayama, S. Yamamoto, Y. Tsujii, T. Fukuda, D. Neher, Macromolecules 2002, 35, 9459–9465. S. M. Sirard, R. R. Gupta, T. P. Russell, J. J. Watkins, P. F. Green, K. P. Johnston, Macromolecules 2003, 36, 3365–3373. P. A. Schorr, T. C. Kwan, S. M. Kilbey, E. S. Shaqfeh, M. Tirrell, Macromolecules 2003, 36, 389–398. P. G. deGennes, Macromolecules 1980, 13, 1069–1075. S. T. Milner, T. A. Witten, M. E. Cates, Macromolecules 1988, 21, 2610–2619. M. Murat, G. S. Grest, Macromolecules 1991, 24, 704–708. M. Murat, G. S. Grest, Phys. Rev. Lett. 1989, 63, 1074–1077. C. M. Wijmans, J. M. Scheutjens, E. B. Zhulina, Macromolecules 1992, 25, 2657–2665. N. Dan, M. Tirrell, Macromolecules 1992, 25, 2890–2895.

327

328

16 Kinetics of Polymer Brush Formation With and Without Segmental Adsorption 35 C. Yeung, A. C. Balazs, D. Jasnow, Macromol-

ecules 1993, 26, 1914–1921. 36 K. Huang, A. C. Balazs, Macromolecules 1993, 26, 4736–4738. 37 G. C. Grest, M. Murat, Macromolecules 1993, 26, 3108–3117. 38 R. Israels, F. A. Leermakers, G. J. Fleer, E. B. Zhulina, Macromolecules 1994, 27, 3249–3261. 39 Y. Y. Lyatskaya, F. A. Leermakers, G. J. Fleer, E. B. Zhulina, T. M. Birshtein, Macromolecules 1995, 28, 3562–3569. 40 D. J. Irvine, A. M. Mayes, L. Griffith-Cima, Macromolecules 1996, 29, 6037–6043. 41 V. A. Pryamitsyn, F. A. Leermakers, E. B. Zhulina, Macromolecules 1997, 30, 584– 589. 42 M. M. Mercurieva, F. A. Leermakers, T. M. Birshtein, G. J. Fleer, E. B. Zhulina, Macromolecules 2000, 33, 1072–1081. 43 B. M. Steels, F. A.Leermakers, C. A. Haynes, J. Chromatogr., B: Biomed. Sci. Appl. 2000, 743, 31–40. 44 C. M. Chen, Y. A. Fwu, Phys. Rev. E: Statist., Nonlinear and Soft Matter Phys. 2001, 63, 011506/1–011506/10. 45 S. W. Sides, G. S. Grest, M. J. Stevens, Macromolecules 2002, 35, 566–573. 46 J. Klos, T. Pakula, J. Chem. Phys. 2003, 118, 7682–7689. 47 C. Ligoure, L. Leibler, J. Phys. France 1990, 51, 1313–1328. 48 R. Hasegawa, M. Doi, Macromolecules 1997, 30, 5490–5493. 49 I. Szleifer, M.A. Carignano, in: Advances in Chemical Physics: Polymer Systems, (Eds.: I. Prigogine, S. A. Rice), John Wiley & Sons, Inc., New York, 1996, Vol. 94, pp. 165–260. 50 L. S. Penn, T. F. Hunter, Y. Lee, R. P. Quirk, Macromolecules 2000, 33, 1105–1107. 51 L. S. Penn, T. F. Hunter, R. P. Quirk, Y. Lee, Macromolecules 2002, 35, 2859–2860.

52 L. S. Penn, H. Huang, M. Sindkhedkar,

53 54

55 56

57 58

59

60 61 62

63

64 65 66

S. E. Rankin, K. Chittenden, R. P. Quirk, R. T. Mathers, Y. Lee, Macromolecules 2002, 35, 7054–7066. M. Morton, L. J. Fetters, Rubber Chem. Technol. 1975, 48, 359–365. R. P. Quirk, in: Comprehensive Polymer Science; (Eds.: S. L. Agarwal, S. Russo), Pergamon Press: Oxford, 1992, First Supplement, pp. 83–106. R. P. Quirk, Y. Lee, Macromol. Symp. 2000, 157, 161–169. J. S. Fritz, G. H. Schenk, Quantitative Analytical Chemistry; 3rd ed., Allyn and Bacon, Boston, 1974. R. Lin, H. Wang, D. S. Kalika, L. S. Penn, J. Adh. Sci. Technol., 1996, 10, 327–339. G. J. Fleer, M. A. Cohen Stuart, in: Fundamentals of Interface and Colloid Science: SolidLiquid Interfaces, (Ed.: J. Lyklema), Academic Press, London, 1995, Vol. 2, Chapter 5. G. J. Fleer, M. A. Cohen Stuart, J. M. Scheutjens, T. Cosgrove, B. Vincent, Polymers at Interfaces, Chapman & Hall, 1993, London, Chapter 8. P. G. de Gennes, J. Physique (Paris) 1976, 37, 1443–1452. S. Alexander, J. Physique (Paris) 1977, 38, 983–987. P. J. Flory, Statistical Mechanics of Chain Molecules, Wiley Interscience, 1969, New York, Chapter 2. P. J. Flory, Principles of Polymer Chemistry, Cornell University Press, 1953, Ithaca, NY, Chapter 14. J. W. Evans, Revs. Modern Physics 1993, 65, 1281–1329. E. L. Hinrichsen, J. Feder, T. Jossang, J. Stat. Phys. 1986, 44, 793–837. J. Talbot, G. Tarjus, P. R. Van Tassel, P. Viot, Coll. Surf. Sci A: Phys. Eng. Aspects 2000, 165, 287–324.

Part III

Applications

331

17

Applications of Polymer Brushes and Other Surface-Attached Polymers Kenneth C. Caster

Abstract

Major recent advances have been made in the synthesis and characterization of many types of surface-attached polymer brushes that have been attached onto different surfaces (e.g., gold, silicon, glass) in multiple configurations (i.e., flat, spherical, tubular) using a variety of polymerization chemistries (e.g., radical, cationic). Polymers with well-defined structure and molecular weight have been prepared using controlled polymerization methodologies such as atom-transfer radical polymerization (ATRP) and ring-opening metathesis polymerization (ROMP). Other techniques (e.g., radiation, plasma) have been used to graft polymers onto surfaces. While these grafting methods do not necessarily provide dense polymer brushes with well-defined structures, they readily provide surface-attached polymers and are thus useful. Considerable effort has gone into the characterization of polymer brushes in order to better understand and predict their properties. This chapter focuses on the use and applications of polymer brushes, and covers both well-defined polymer brushes and surface-grafted polymers in general, as on many occasions it is unclear in published reports whether the grafted polymer is a densely grafted brush or a loosely attached polymer chain. Although the use of these fascinating structures is still in its infancy, they hold great promise for future applications.

17.1

Introduction

From the paintings of early man on cave walls to heat-resistant coatings on engine turbine blades, surface modification has moved from the application of simple coatings to the manufacture of highly engineered surfaces. Along with the development of more complex surface treatments has come the rise in the use of new materials that possess unique properties and, with their introduction and acceptance into technological applications, new, more subtle forms of surface modification have begun to emerge. There is great interest in preparing bulk materials from inexpensive raw materials, and providing the surfaces of those materials with exquisite, highly designed properties.

332

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

The desire to impart new or improved properties to surfaces has been a major undertaking. One method of manufacturing materials with surface properties that are different than those of the bulk involves the application of organic polymer coatings [1]. These coatings may, compositionally, be highly complex and provide excellent properties, depending on the application (e.g., corrosion protection on steel). However, methods for the application of polymeric coatings are not general; rather, there is much that can be done to improve processing, to address environmental release problems, and to improve their performance. Rather than relying on the weak physisorption of a polymer to a surface, much effort has been expended to attach the polymers covalently. Densely packed, end-grafted polymers are known as polymer “brushes”. Hence, as these brushes have emerged in well-characterized form, their use to modify surfaces for real applications has begun to appear both in the open and in patent literature. This chapter addresses the applications and uses of surface-attached polymers, where the polymers are either well-characterized and ordered polymer brushes, or they are less well-characterized end-grafted polymers. As there have been very few actual uses of polymer brushes, but many reports have been related to surfacegrafted polymers, only a few examples will be highlighted here. Thus, for the convenience of the reader, lists of references, sorted by application, are provided in the Appendix at the end of this chapter.

17.2

Surface Modification and Functionalization

Organic coatings or thin films are usually formed by the application of solution or emulsion-based polymers to surfaces by solvent casting, spin casting, painting, rolling, or spraying. The mechanical integrity of the polymer film is determined by cohesive forces within the polymer matrix, whereas adhesion is determined by physisorption, electrostatic interactions, interfacial diffusion, and in some cases covalent bonds with the surface. Surface attachment generally occurs at random points on the polymer backbone. Polymer brushes are densely packed polymeric structures in which one end of the chain is directly attached to the surface, with the bulk of the chain extending into the solution or air interface. 17.2.1

Polymerization Methodologies for Surface-Attached Polymers

Multiple approaches are known for preparing polymer brushes [2,3]. In the method known as “grafting to”, polymer brushes are generated by selective attachment of one end of a polymer chain to a surface using coupling chemistry. Low brush densities result in this entropically disfavored process as the polymer chain must adopt an extended conformational state from a random coil conformation for attachment to the surface to occur (Figure 17.1).

17.2 Surface Modification and Functionalization

Y Y X

X

X

X

X

X

X

Y

Z

Y

Y

Z

Y

Y

Synthesis of polymer brushes via “grafting to” surface attachment of polymers, where “X” and “Y” are reactive functional groups that undergo a coupling reaction to give a new functionality, “Z”.

Figure 17.1

High brush densities are produced in the “grafting from” approach. Here, a small initiator molecule is tethered to the surface using an anchoring group which provides adhesion, a spacer group to separate the surface from the initiator, and an initiator moiety that generates a polymer once initiation has begun (Figure 17.2). “Grafting from” approaches typically reach much higher polymer grafting densities, as a small initiator molecule is first attached to the surface followed by a surfaceinitiated polymerization (SIP). This produces polymer brushes on the surface wherein the polymer chains are extended away from the surface into the bulk solution. Functionalized surfaces using the “grafting from” approach have been prepared using almost all known polymerization techniques (e.g., anion [4], cation [5,6], radiPolymer

Initiator Spacer

Tether Monomer

Anchor

Figure 17.2 Synthesis of polymer brushes via “grafting from” surface attachment of polymers. The tether is comprised of an anchor, a spacer, and an initiating group.

Initiator Polymer Physisorbed Initiator Monomer

Synthesis of surface-attached polymers via “grafting off” polymerization. The initiator is physiosorbed directly onto the surface.

Figure 17.3

333

334

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

cal [7,8], reversible addition-fragmentation chain transfer (RAFT) [9,10], plasma [11], condensation [12], photochemical [13–15], electrochemical [16], and controlled free radical [17–20]). Transition metal-based polymerization methods such as atom-transfer radical polymerization (ATRP) [21–23] and ring-opening metathesis polymerization (ROMP) [24–26] have also been used to prepare surfaces with design properties. Other transition metals have been used to prepare polymers, and these initiators should also provide polymer brushes by SIP [27]. Bulk polymeric surfaces have been “activated” by plasma, UV irradiation, ozone, and flame sources and polymers grown by their exposure to monomers [28]. Another method of introducing an initiator onto a surface involves placement of the initiator molecule directly onto the surface to which a polymer is to be grafted. While relatively new, this technique – which can be referred to as “grafting off” – does not necessarily provide well-defined polymer brushes, as only physisorption holds the initiator at the surface. However, it has been found to provide good adhesion to a variety of substrates using ROMP as the enabling polymerization method (Figure 17.3) [29,30]. This approach has also found use in making free-standing films of conductive polymers [31], in the preparation of coated pigments [32], and in the preparation of fiber-reinforced molded polynorbornene articles [33]. This area is ripe for research investigation to supply understanding into the nature of this type of surface grafting chemistry. Polymer Brushes Many parameters are available to be adjusted when controlled polymerization methodologies such as ATRP and ROMP are used to prepare well-defined polymer brush structures. Molecular weight (Mn) can be predetermined by the monomer to initiator molar level. In addition, controlled polymerization methods usually provide low polydispersity index (PDI) polymers, which is important in making smooth polymer brush films. Brush chain length affects a variety of properties. For example, atomic force microscopy (AMF) studies have revealed that for densely grafted polymer brushes longer chain length leads to increased compression resistance [34]. The length of the brush chain also plays an important role in determining the ultimate mechanical properties of the interface. A longer grafted chain length has been shown to provide large improvements in fracture energy in a polyethylene/glass interface. Interfacial toughness increases as the grafted connecting chains can interdiffuse and form chain entanglements [35]. Depending on the reaction conditions, heavily substituted brushes with a variety of functional groups can be prepared by choosing the correct initiator system. Methods are being developed that allow initiator density to be controlled, which ultimately affects brush grafting density. For example, different concentrations of initiator have been applied to surfaces by dilution of the initiator with unreactive selfassembled monolayers (SAM) forming chains [36,37] or by photodecomposition of surface-attached initiators [38]. Genzer has used vapor phase [39,40], solution phase [41,42], and mechanical [43] methods to vary brush density. Polymer brush density affects other properties including surfactant binding to polyacrylic acid brushes [44] and brush height [45]. 17.2.1.1

17.2 Surface Modification and Functionalization

Tether Design Elements “Grafting from” SIP requires the use of a tether to attach the initiator to the surface. As illustrated in Figure 17.2, three elements must be considered when designing the tether, namely the anchor, spacer, and initiator group. The anchor provides primary adhesion of the polymer brush to the surface, the spacer controls both distance and packing density of the initiator (and hence the polymer brush) from the surface, and the initiator determines the type of polymerization chemistry to be used to prepare the brush. The anchor must provide good surface adhesion for the resulting films to have good mechanical stability. Many polymer brushes have been prepared using different attachment chemistries (e.g., thiols on gold, silanes on silicon and glass). Other functional groups promote extremely strong adhesion to specific surfaces. For example, phosphonates [46] are known to enhance adhesion to galvanized steel [47] and aluminum [48]. Thus, an anchor group must be chosen to promote adhesion of the final surface-attached polymer depending on the substrate to which it is attached and the application for which it is to be used. Besides controlling the distance between the surface and initiator, the tether also affects symmetry, order, and packing-density of the overall assembly. 17.2.1.2

17.2.1.3 Monomers The monomer used to prepare the polymer brush is dictated by the polymerization method, the desired polymer properties, and the ultimate applications. Highly functional polymer brushes have been prepared and investigated [49]. For example, acrylate and styrene derivatives are commonly used with free radical, ATRP, and anionic polymerizations; epoxides with cationic polymerization; and substituted norbornenes with ROMP. With the advent of controlled polymerization methodologies and their application to the synthesis of polymer brushes, the synthesis of block copolymer brushes has been realized. Great promise exists for the application of block copolymers to nanotechnology applications [50]. Although many examples are known, only a few have been shown to illustrate monomer variability. Brittain prepared block copolymer polymer brushes that showed reversible changes in water contact angle simply by changing the solvent. The synthesis used a sequential combination of cationic polymerization followed by ATRP to give a poly[styrene-b-methyl methacrylate] block copolymer brush [51]. Quirk used the “grafting from” approach to prepare poly(isoprene-b-ethylene oxide) block copolymer brushes by anionic polymerization [52]. Polymer chain configuration was examined by neutron reflectivity for poly(2-vinylpyridine-b-deuterated styrene) and poly(methyl methacrylate-b-styrene) diblock copolymer brushes adsorbed at the polystyrene/substrate interface from the melt. Evidence for the existence of three regimes: a “wet brush” regime, a “mushroom” regime, and a broad transition regime in between was revealed [53]. Huck prepared di- and tri-block polyelectrolyte brushes on patterned surfaces using aqueous ATRP. The controlled nature of the polymerization lead to well-defined structures that were characterized by ellipsometry, grazing angle FT-IR, and contact angle measurements. Brush height was found to be influenced by both the presence of salt and nature of its anion during analysis [54].

335

336

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

17.2.1.4 Application Techniques The use of polymer brushes to pattern micro- and nano-structures on surfaces are complementary and related though the scale of their technologies. Besides the bulk coating of surfaces with polymer brushes using “grafting to” and “grafting from” polymerization methodologies [55], many studies have been conducted to investigate the preparation of brushes that have been generated from SAMs [56]. In addition, soft lithography techniques [57] such as micro-contact printing (lCP), stamping, and dip-pen nanolithography (DPN) [58] patterning have all been applied in preparing structured surfaces, and many of these techniques involve polymer brush chemistry. Further details of the use of these methods when preparing polymer brushes may be found in the references cited within this chapter. 17.2.2

Property Control

Whether well-controlled polymer brushes or bulk-grafted polymers are chosen, there are clearly many different synthetic elements available which, through careful design, allow the control of physical properties of resulting surface-attached polymers, and thus their thin films and coatings. Perhaps the most prominent property of polymer brushes is their ability to control surface hydrophophicity/hydrophilicity. Whether on the surface of a flat sheet, a particle, or a fiber, the control of hydrophophicity/hydrophilicity will greatly affect other properties such as wettability, water/ oil adsorption, adhesion, compatibility, and solubility. By adjusting the surface charge on a particle, viscosity and flow properties can be potentially tailored to meet the requirements of a particular formulation. 17.2.3

Impact on Types of Materials

Control of the properties described above directly affects a variety of macroscopic materials. Adhesion of coatings and thin films is important for the performance of the coated substrate. For example, in composite materials poor adhesion between a polymer-coated filler particle or a fiber and the matrix polymer can be responsible for lower than expected mechanical properties, and cause premature failure [59].

17.3

Applications

Perhaps one of the most exciting expectations for the use of polymer brushes is their potential for affecting a variety of different surface properties, ranging from adhesion to tribology on many different substrates, and the ability to tune these properties using an external stimulus. This is expected to impact basic applications such as coatings for corrosion protection to high-tech applications such as controlled-release biocoatings.

17.3 Applications

17.3.1

Adhesion

Whether one considers its promotion or inhibition, adhesion is of fundamental importance whenever surface modification, surface properties, or interfaces are discussed. De Gennes modeled adhesion between two rubber layers in which brushtype polymers extended from one surface across the interface to the other surface. He found adhesion to be a function of the thermodynamic work of adhesion between the two elastomeric surfaces in the absence of brush promoter plus the suction work required to pull out the connecting polymer chains from one layer [60,61]. Considerable effort has gone into understanding the fundamental role that surfaceattached polymer chains play in interfacial adhesion, and this has been reviewed and described in detail by others [62–66]. Functional group placement within a surface-attached polymer has been examined using Scheutjens-Fleer self-consistent mean-field theory. These calculations reveal that an optimum low-energy release surface has functional polymer with adjacent low-energy functional groups located at one chain end, whereas an optimum high-energy adhesive surface is obtained by placing adjacent high-energy functional groups at the center of the polymer chain [67]. Biosurfaces Considerable effort has been made to develop biomaterials that possess good mechanical properties and biocompatibility. While many materials have been developed, they suffer from a variety of problems, including poor surface attachment of cells and tissues. The development of new biomaterials that have all of the desired properties is costly, and current efforts are focused on using presently available biomaterials, but with designed surfaces. Both adhesion and the inhibition of adhesion are important when considering applications involving biosurfaces (e.g., artificial implants, cell culture dishes, biosensors). Many surfaces have been functionalized with proteins and cells by physisorption and “grafting to” methodologies. However, while providing function, they tend to be only weakly ordered systems. Nonetheless, strides have been made to prepare more ordered surface-attached proteins [68,69]. The primary factors that determine how well a block copolymer brush will reduce protein adsorption on a surface were determined via theoretical models, and predicted to be surface coverage of the polymer and the surface-polymer interactions, with polymer chain length being of secondary importance. In the early stages of the process, adsorption depends strongly on polymer molecular weight. Experimental isotherms agreed well with the theoretical models [70]. Poly(vinylidene difluoride) (PVDF) is used as a biomaterial in soft tissue applications and in sutures. Although its material properties are well-suited for this application, improved adhesion of proteins and peptides that promote integrin-mediated cell attachment is desired. Tissue compatibility was engineered by creating poly(acrylic acid) polymer brushes (plasma-induced SIP) on the PVDF surface and converting the acid-functionalized brush to a fibronectin-coated surface by carbodiimide coupling reactions, and studied by comparative exposure of the modified surface to 17.3.1.1

337

338

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

primary human osteoblasts. Prolonged proliferation and survival of the cells was only observed when fibronectin was covalently attached to the surface [71]. Recently, polymer brushes have found use in this arena particularly through the use of surface-attached stimuli-responsive polymers (see also Section 17.3.5) to make “smart” bioconjugates using smart polymers and receptor proteins [72]. The use of external stimuli (e.g., pH, electric field, temperature, solvency) to effect a change in polymer properties has also been found to be very useful for controlling adhesion on biosurfaces. The change usually comes about from a change in conformation which affects hydrophobicity/hydrophilicity and thus the surface energetics of a surface-attached polymer. Many stimuli-responsive polymers are known, and many studies have been made with those based on poly(N-isopropylacrylamide) (poly(NIPAAM)). These polymers exhibit a phase change when taken through the lower critical solution temperature (LCST), typically 32 KC for poly(NIPAAM). The LCST can be manipulated by changing the N-substituted hydrocarbon chain or through the preparation of copolymers [72,73]. At the LCST, reversible dehydration of the hydrocarbon side chain occurs, causing a collapsed conformation and a change from a hydrophilic to a hydrophobic state, which can result in solubility changes for bulk polymers in solution. This stimuli-responsive behavior can be used to promote adhesive Van der Waals or repulsive electrostatic interactions between either hydrophobic or hydrophilic components in a compound or mixture and thus effect a separation (see Section 17.3.6). Huck et. el. investigated the adhesive behavior of patterned poly(NIPAAM) brushes on gold prepared by ATRP using contact-mode AFM imaging. Both expanded (46 – 9 nm) and collapsed (11 – 2 nm) states of the brush were revealed by AFM examination in aqueous medium, both below and above the LCST, respectively. In adhesion force mode, force versus distance curves on these surfaces in aqueous medium showed no adhesion between the Si3N4 AFM tip and the poly(NIPAAM) brushes below the LCST. However, above this value measurable adhesion forces (2.25 nN) were observed, and this was shown to be reproducible on temperature cycling [74]. Zauscher et al. observed a similar differential adhesion on nanopatterned poly(NIPAAM) brushes by using solvent pairs (H2O/MeOH) to traverse through the LCST [75]. Lopez et al. used this property to demonstrate how surface energy (i.e., wettability via contact angle measurement) changed by 19K on progressing through the LCST. The absolute value of the contact angle was changed by using mixed monolayers and thus changing the surface composition of the polymer brush [76]. Temperature-responsive surfaces were created from poly(NIPAAM) polymer brushes (via electron beam-initiated polymerization) on tissue culture polystyrene substrates and used to investigate inflammatory cell adhesion behavior. At elevated temperature, human monocyte and monocyte-derived macrophages were able to adhere, spread, and fuse to form foreign body giant cells (FBGC) on the hydrophobic surface. Cell detachment was accomplished by lowering the temperature of the brush-coated surface below the LCST. Differential macrophage detachment was observed with time (98% after 2 h; 30% after 10 days) in culture. While significant, this approach did not allow the isolation of pure FBGC, as macrophages remained

17.3 Applications

on the surface after 7 days [77]. Biologically interesting temperature-responsive poly(NIPAAM) surfaces have also been prepared using plasma polymerization [78]. While thermally-responsive polymer brushes control the reversible adhesion of proteins and cells, hydrophobic interactions between stimuli-responsive polymers have also found use in bioelectronics by “grafting to” biopolymers to electrodes [79]. Surface-attached biopolymers such as elastin-like polypeptides (ELP) have also been used in this manner [80]. Cell Growth Control Control of cell growth can be accomplished by attaching cells to a surface, allowing them to proliferate and grow, followed by their detachment. Cell attachment and proliferation is a facile process, particularly for hydrophobic surfaces, whereas detachment requires sophistication to recover cells without damage. Thermoresponsive polymer brushes, with their ability to control hydrophobic/hydrophilic properties, were investigated to determine their efficacy in this process. Endothelial cells and hepatocytes were found to attach to poly(NIPAAM) brushes, which had been prepared on polystyrene cell culture dishes by electron beam-initiated polymerization (“grafting from”), and to grow at 37 KC, above the LCST temperature which places the polymer brush in the hydrophobic state. Removal of the cultured cells from the surface was readily accomplished by cooling the matrices below the LCST which hydrates the surface; this was done without the damage usually observed during trypsinization. Different cells were seen to have their own optimal detachment temperatures (hepatocytes, 10 KC; endothelial cells, 20 KC), thus suggesting this to be an important parameter in cell separation processes [81]. These same authors examined grafted carboxyisopropylacrylamide (CIPAAM), which has a similar chain structure to that prepared from a poly(N-isopropylacrylamide)/poly(acrylic acid) copolymer, onto cell culture plastic plates to investigate cell adhesion by controlling the surface hydrophilicity/hydrophobicity via stimuliresponsive polymer brush phase transition. Similar to poly(NIPAAM) grafted polystyrene culture dish surfaces, the poly(CIPAAM) grafted surfaces exhibited relatively weak hydrophobicity and showed good cell adhesion in culture dishes at 37 KC. Again, as with poly(NIPAAM) surfaces, cell detachment was observed below the LCST [82]. Surface-attached polymers (i.e., both “grafting to” and “grafting from”) have been used to control cell growth using protein-repellent micropatterns based on poly(acrylamide)/PEG copolymers [83], comb polymers [84], and polycationic PEG-grafted copolymers [85]. 17.3.1.2

Nonfouling Biosurfaces Recently, polymer brush-coated surfaces have been examined for their ability to provide nonfouling properties. Extracellular proteins adsorb strongly to many surfaces through hydrophobic interactions. While this is useful for making biocoatings, it is problematic when specific surface interactions are to be studied, as the co-adsorbing proteins interfere with the desired surface chemistry. Considerable advances have been made in using SAMs [86–89] to prepare nonfouling biosurfaces through the 17.3.1.3

339

340

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

creation of highly hydrated, hydrophilic layers, usually through the attachment of poly- or oligo-ethylene glycol functionality to the surface. The idea of using polymer brushes composed of short oligoethylene glycol segments to impart nonfouling behavior against proteins was examined [90]. The polymer brushes were prepared by SIP-ATRP of oligoethylene glycol methyl methacrylate (OEGMA) from initiator-terminated alkanethiol SAM-covered gold substrates. Surface plasmon resonance (SPR) revealed that no protein adsorption (i.e., pure solutions of fibronectin, 10% and 100% fetal bovine serum) occurred after exposure of 15-nm poly(OEGMA)-coated surfaces to protein solutions. This nonfouling behavior was employed to make patterns using lCP and DPN to investigate cell growth at the micro- and nano-scale [91]. Microelectronics Adhesion plays a critical role also in microelectronic devices, as poor adhesion can lead to interface failure and shortened device lifetimes. Failure results from a thermal mismatch between the conductive metals and polymers used in printed circuit board manufacture. While little has been done to evaluate the use of polymer brushes in promoting adhesion at such interfaces, it is expected that the use of welldefined surface-attached polymers would be beneficial in solving problems related to decreasing feature size and thermal management. While not necessarily producing well-defined polymer brushes, grafting of polymers onto surfaces has been used to improve adhesion. For example, exposure of a copper foil/argon plasma treated KaptonN HN film to 4-vinylpyridine at elevated temperature resulted in near-doubling of the lap shear strength of the sandwiched laminate [92]. Different amines have produced similar results under photochemically initiated conditions [93]. 17.3.1.4

Miscellaneous Modification of the rubber obtained from recycled tires through surface-grafting reactions has been used to improve adhesion of the tires after they have been ground into particles, which makes them useful as fillers. When added to a polymer formulation, these untreated, recycled elastomers can detrimentally affect mechanical properties because of poor adhesion within the matrix, thus limiting their use in composite materials. By changing the surface energetics of the elastomer through addition of potentially reactive polar functionality, properties such as wetting, adhesion, and phase compatibility improve, which leads to a useful filler. As evidenced by lowered water contact angles, an enhanced wettability of ground rubber films and particles was observed after surface-initiated photopolymerization of methacrylic acid and glycidyl methacrylate. Grafting yields ranging from 3 to 42% were observed for silica-loaded rubber, but these values fell off significantly for carbon black-loaded rubber [94]. These findings suggest that the formation of surfaceattached polymers on recycled tire surfaces will aid in their use as fillers. However, the mechanical properties of many composite interfaces must be evaluated before the generality of this approach to compatibilization can be confirmed. The coating or painting of polyolefin surfaces poses a difficult problem, as wetting the hydrophobic surface is difficult – especially for water-borne coatings. As a result, 17.3.1.5

17.3 Applications

poor adhesion between the freshly applied coating and the surface eventually leads to failure. Surface grafting opens an easy way to modify polyolefin surfaces by making the surface more hydrophilic through addition of polar functional groups, and thus providing favorable surface chemistry for coating formulations. Different processes have been used for this purpose including radiation [95], corona discharge [96], and ROMP [97]. “Grafting off” has been found to give very good adhesion promotion to adhesives either for bonding post-vulcanized elastomers to grit-blasted steel (natural rubber, EPDM, SantopreneN) [29] or for bonding in elastomer/fiber composites (natural rubber; polyester, KevlarN, nylon) [30]. Poly(ethylidenenorbornene) and poly(dicyclopentadiene) polymers were grafted onto a surface coated with Grubbs’ 1st generation catalyst, with adhesion occurring once the catalyst was exposed to the ROMP monomer. This approach has also gave good adhesion when making coatings on metal, plastic, elastomer, and low surface energy substrates [98]. 17.3.2

Tribology

The ability to control surface properties at the nanoscale holds great promise for polymer brushes. The lubrication [99] and friction [100] properties of polymer brushes have been examined quite extensively [101]. Polyelectrolyte polymer brushes have been shown to have superior lubrication properties compared to neutral brushes, and to display effective friction coefficients less than 0.0006–0.001 at low sliding velocities (250–500 nm s–1) and at loading pressures of several atmospheres in aqueous environments. These findings have important implications for artificial implant design and biolubrication [102]. 17.3.3

Stabilization and Compatiblization

The stabilization of colloidal and core-shell particles against agglomeration is important in many industrial processes, and critical in many products. The particles are usually inorganic materials dispersed in an organic matrix, or in an aqueous medium. Their interfacial properties within a matrix, or the nonagglomeration properties within a dispersion, help to determine the ultimate properties of the material or the final properties of a dispersion/formulation, respectively. Stabilization or compatibilization is usually carried out by encapsulating a colloid or particle within a shell of organic polymer; this usually results in a property mismatch between the composite layers (e.g., modulus, polarity). It is the surface properties of the shell that ultimately aid or create the desired stabilization. By compatibilization of colloid or core-shell particles, improved properties can be attained [103,104], an example being the stabilization of latex particles through the use of polymer brushes [105]. Considerable effort has gone into the synthesis and characterization of colloids [106] and nanocomposites [107] that contain polymer brushes. A list of references to publications dealing with the preparation and use of core-shell and colloidal particles is provided in Table 17.1.

341

342

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

Carbon black, silica, clay, metals, and metal oxides are commonly used materials from which colloids or core-shell nanocomposite particles are prepared. Carbon black has many useful properties, and may be found as an ink pigment, a reinforcing filler in tires, and numerous other applications. The formation of stable dispersions in solvents or polymer matrices is important for the resultant final properties of the formulation. Surface-grafted carbon blacks are known [108], but only recently have well-defined polymerization methods been applied to their preparation. Surface-attached poly(n-butyl acrylate) polymer brushes were attached to carbon black using ATRP by both “grafting from” and “grafting to” approaches. Tapping mode AFM on the brush-coated particles revealed a dense corona of polymer chains extending from the core carbon particle for the “grafting to” material, whereas no chains were found to extend from the “grafting to” samples. The “grafting from” particles were readily dispersed in solvents that are good for poly(n-butyl acrylate), and were difficult to isolate by centrifugation unless a poor solvent was added to the mixture [109]. Carbon black colloidal dispersions were functionalized by well-controlled surfaceattached polystyrene brushes, which were prepared by benzoyl peroxide/TEMPOterminated radical chemistry and attached using “grafting to” methodology. The grafted particles formed stable colloidal dispersions in THF. No difference in dispersion stability was noted at different polymer brush molecular weights [110]. An interesting application of polymer brushes involves their use in the attachment of V2O3 onto the surface of semigraphitic carbon black nanoparticles. The composite structure was prepared by nitric acid oxidation of carbon black nanoparticles, followed by “grafting to” esterification of polyethylene glycol (Mn = 4600) and coating with an aqueous dispersion of V2O5 lamellae. The nanocomposite displayed enhanced electrochemical response (i.e., high rates and improved stability) over a wide range of current-densities when compared with standard V2O5 xerogel and V2O5/carbon black electrodes, and is considered a fast transport material based on its kinetic response at high current densities. Such a property makes this nanocomposite suitable for high current/power applications, and it is also being considered for advanced energy storage systems [111]. The stabilization of CdS nanoparticles towards aggregation has been accomplished through the use of sparsely grafted polyoctenamer brushes. The phosphine oxide-coordinated CdS nanoparticle-polymer composites showed excellent interparticle dispersion by transmission electron microscopy when compared to the starting materials. The ROMP-based polymer brush has little effect on the absorption and photoluminescent-emission properties of the nanoparticles [112]. Such particles are important for applications involving quantum dots. Gold nanoparticles, modified using ROMP to yield layered copolymer structures with redox active functionalities, could serve as diagnostic probes for chemical and biochemical detection. Block copolymer polymer brushes were attached to 3-nm gold nanoparticles by self-assembly of a norbornyl containing thiol, followed by conversion to the initiator with Grubbs’ 1st-generation catalyst, and finally ROMP of redox-active ferroceneyl complexes. The newly formed nanoparticles displayed solubility properties expected for the surface-attached polymer complexes. Cyclic voltam-

17.3 Applications

metry showed two distinguishable waves associated with ferroceneyl complexes at E1/2 = 180 mV and –40 mV (versus FcH/FcH+), respectively. This versatile chemistry is expected to be extended to other systems [113]. Polymer brush functionalized particles have also found use as high-loading resins in combinatorial chemistry. The conversion of vinyl polystyrene beads into an active ROMP initiator with Grubbs’ 1st-generation catalyst, followed by polymerization with substituted norbornenes, produces resins with loadings on the order of 3.0 mmol g-1. This polymer support (named ROMP-Spheres) was used as an esterification platform [114]. Polynorbornene has also been grown from gold colloids using surface-attached ROMP catalysts to give a core-shell composite particle. The high catalytic turnovers observed in this process suggest that thick brushes can be prepared in this manner [115]. Surface-initiated living free radical polymerization using a TEMPO initiator produced substituted polystyrene resins that find use as solid supports for solid-phase organic synthesis (named Rasta resins). The microwave-promoted polymerization yielded the resins approximately 150-fold faster than through normal heating. These resins also displayed higher solubility than their crosslinked polystyrene congeners, and are considered to be the largest (550 lm) and highest loading (5.8 mmol g–1) solid supports currently known [116]. 17.3.4

Surface Coatings

The wettability of a surface is an important property for many applications, and is essential for the creation of an adhesive bond when joining two substrates together, during application of a coating to a substrate, and during the creation of almost any interface. Whether the resulting surface is to be hydrophobic or hydrophilic is highly application-dependent. Recently, super hydrophobic surfaces have been created by controlling surface morphology using nanostructures [117,118] and patterned polymers [119]. Polymer brush wetting has also been investigated [120–122], and the use of grafted polymers has been used to control wetting in many applications. The control of fiber surface hydrophobicity, wetting, and adhesion properties is important in composite formation. Polymer brushes were prepared on cellulose fibers by “grafting from” ATRP of methyl acrylate. The coated surface displayed increased hydrophobicity (as shown though advancing water contact angles) which were found to increase with the degree of polymerization (DP), and reached a maximum of ha = 133K (DP = 298). Although the nature of the surface of wood requires that these measurements be made with care, they nicely illustrate that “real” surfaces can be modified from being completely hydrophilic (i.e., no ha for the initial cellulose surface as it adsorbed all the water) to hydrophobic by the grafting of a thin polymer brush [123]. The presence of a polymer brush on a surface has been found to affect the morphology of films adsorbed onto the brush-coated surface. The morphology of ultrathin poly[styrene-b-butadiene-b-styrene] copolymer films deposited on polystyrene

343

344

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

brushes was found to be influenced by grafting density and the DP of the underlying polymer brush layer [124]. Surfaces decorated with poly(4-vinyl-N-methylpyridinum) iodide polyelectrolyte brushes served as substrates for the preparation of welldefined polyelectrolyte multilayers via layer-by-layer deposition. Strong electrostatic forces and low solubility of the surface-bound polycation/solution-phase polyanion complex resulted in nonstoichiometric film formation and collapse of this newly formed film to thicknesses near the dry film thickness. Film thickness could be modulated slightly by the inclusion of low-molecular-weight salts during film formation. The build-up of film thickness using the normal layer-by-layer process followed normal trends once the initial film was formed. This process is robust, and little affected by changes in deposition variables [125]. Silica nanoparticle (8 nm) adsorption has been examined on poly(ethylene oxide) brushes. Adsorption was found to occur rapidly, and the amount of adsorbed particles was strongly influenced by pH and brush grafting density, with a maximum amount adsorbed at low pH and intermediate grafting densities (pH = 2; r = 0.15 nm–2). Reversible adsorption was observed by switching pH regions [126]. In the presence of the nematic liquid crystal pentacyanobiphenyl, a polystyrene brush exhibited various structures and ordering profiles, depending on its molecular weight and grafting density. The isotropic phase is stabilized by the surface-attached polystyrene, which is an important feature for the fabrication of bistable liquid crystal (LC) displays [127]. Well-defined poly(e-caprolactone) brushes have been used to control nanoscale topography and thus orient 4-cyano-4¢-pentylbiphenyl liquid crystals on gold surfaces. LC orientation changed with increasing ellipsometric thickness of the brush from planar and azimuthally uniform to homeotropic or nonuniform. Interestingly, this change in anchoring with brush thickness was used to image patterned poly(e-caprolactone) brushes on the surface [128]. The stability of thin polystyrene films formed on top of chemically identical surface-attached polystyrene brushes was investigated as a function of grafting density and thickness. Brush thicknesses of the order of 20–35 nm stabilized the polystyrene film towards dewetting; however, dewetting was observed for brush thicknesses either above or below this value [129]. Autophobic dewetting of such systems can be suppressed by using bimodal brushes that contain a small number of long chains [130]. Coatings have been prepared on electrically conductive substrates using electrochemical polymerization. The coatings prepared by this process tend to have highly desirable properties such as good adhesion. Moreover, they can be formed on virtually any shaped substrate, and processing can be simplified by the elimination of primers [131]. However, this process is somewhat limited by the final coating film thickness, as the electropolymerization is self-limiting and yields only thin films. This is especially problematic in real use, where the surface may be easily scratched or gouged. Thicker coatings have been produced by sequentially coupling cathodic electropolymerization with another polymerization method. In this way, polymer brushes have been produced on electrically conductive surfaces (e.g., steel, copper wire, carbon plates, carbon fibers) using acrylate-substituted monomers that are functionalized to undergo ATRP [132] and ROMP [133] “grafting from” methodolo-

17.3 Applications

gies. Strongly adhering polystyrene (200 nm to 1 lm) and polynorbornene (6 lm) films were formed on steel surfaces. In addition to producing the thicker film, the polymer brush layer now serves as a primer for further chemistry, should that be desired. As a variation, polystyrene brushes were grown by radical polymerization after the conversion of carboxylate-substituted poly(pyrrole) coatings, which had been prepared by electropolymerization of carboxylated pyrroles, to azo initiators. This procedure also gave uniform coatings that showed good adhesion to the steel electrodes [134]. 17.3.5

Stimuli-Responsive and Switchable Surfaces

As discussed above in Sections 17.3.1.1 and 17.3.1.2, the use of stimuli-responsive polymer brushes is very useful in the control of adhesion, particularly in biological applications. Applications involving stimuli-responsive, surface-attached polymers will be further elaborated on in this section, and in Section 17.3.6. Perhaps one of the most exciting and promising areas for the development of polymer brush applications is that involving switchable, stimuli-responsive, or “smart” surfaces. This is especially true for block copolymer brushes. Brittain et al. observed reversible changes in surface morphology and water contact angle simply by changing the solvent to which the block copolymer brush was exposed. Polystyrene-b-poly(methyl methacrylate) (PS-b-PMMA) brushes were smooth (RMS roughness = 0.77 nm; contact angle = 74K) when exposed to CH2Cl2, but became rougher (RMS = 1.79 nm; contact angle = 99K) after exposure to cyclohexane. These changes, which were induced by selective solvation of one phase, were believed to result from large conformational changes in the outermost layer of the copolymer, and were found to be reversible. X-ray photoelectron spectroscopy (XPS) corroborated this hypothesis. Very rough, hydrophobic surfaces (RMS = 13.08 nm; contact angle = 120K) populated with ellipsoid structures were prepared through manipulation of solvent polarity [51,135,136]. Similar behavior has been reported for other block copolymer [9,137], Y-shaped [138], terpolymer [139,140], and binary polymer brushes [141]. An interesting application of stimuli-responsive polymer brush surfaces uses a mixed brush composed of poly(2-vinylpyridine) and polyisoprene to write permanent patterns onto a surface that has been patterned via photolithography – a process termed “environment-responsive lithography”. Solvent switching provides both the stimulus for creating and erasing the pattern. UV radiation during the photolithography step crosslinks the polyisoprene in the mixed brush, and this causes a loss of switching properties for the surface in that region. Imaging relies on the contrast that develops between parts of the surface that have been irradiated and masked when exposed to solvent. Thus, the exposure of a patterned surface to toluene creates hydrophobicity in regions masked in the photolithography, while ethanol yields a uniform surface composition, and water (pH = 2) creates a hydrophilic surface in the masked region. This process can be repeated, and the patterns that are formed are easily erasable by switching solvents. In addition, they are stable for

345

346

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

at least one month when stored in the acid form. Manipulation and tuning of regional polymer brush properties (i.e., wetting, adhesion, adsorption) has allowed the surface to be used as a pH sensor, to visualize adsorbed silica nanoparticles, and to prepare a gated microfluidic channel [142]. Environmentally smart (co)polymer brush surfaces have been prepared that respond to changes in humidity with changes in their wetting behavior [143]. Two-level self-adaptive surfaces that show exaggerated wetting behavior have been prepared from polymer brushes composed of carboxyl-terminated poly(styrene-co2,3,4,5,6-pentafluorostyrene) and carboxyl-terminated poly(2-vinylpyridine). These copolymer brushes were “grafted to” poly(tetrafluoroethylene) (PTFE) that had been roughened by plasma etching so that two levels of structure had formed (i.e., microsize PTFE needle-like features coated with nanoscale demixed copolymer polymer brushes). The larger features amplify the stimuli-responsive behavior of the brushes when the solvent is changed. Water contact angle measurements show expected switchable responses on control one-level surfaces (toluene, H = 118K; water, pH = 3, H = 25K), whereas the two-level surfaces show ultrahydrophobic responses after toluene exposure (H = 160K) and complete wetting by water (pH = 3). The microstructure amplifies the nanoresponse to the external stimulus. Controllable adhesion between the two surface types was also observed in a simple peel test [144]. 17.3.6

Separations

The separation of mixtures into their components is an extremely important process that impacts on all branches of chemistry, and especially on biological areas where the isolation of pure substances is critical to their use in humans. It has been discussed in Sections 17.3.1.1 and 17.3.12 how the use of an external stimulus to effect a change in polymer brush properties is valuable in the separation of proteins, cells, and biomolecules as applied to specific adhesion and nonfouling behavior. Examples of stimuli-responsive behavior and other topics where polymer brushes are critical to separation processes will be discussed throughout the present section. Specific reviews regarding the use of polymer brushes in the aqueous separation of biological compounds [145], as chiral stationary phases for high-performance liquid chromatography [146], and in the analysis of particle/brush interactions for separations [147] are available. 17.3.6.1 Chromatographic Separation The use of polymer brushes in chromatographic protein separation and purification has been recently reviewed [148]. The focus of the discussion was on adsorptive membrane chromatography using polymer brushes decorated with ion-exchange, hydrophobic and affinity groups. These authors found that increasing the permeation rate of protein solution through the membrane resulted in an accelerated overall adsorption rate of protein. Microporous hollow-fiber membranes decorated with ion-exchange groups showed repeatable protein adsorption and elution processes,

17.3 Applications

whereas the hydrophobic- and affinity ligand-immobilized membranes showed deteriorated protein binding capacity after repeated cycles. The hydrophobic-ligand containing membranes could be regenerated with alkaline solution. Van Zanten used a Flory-type mean-field analysis for mixing of a multicomponent solvent with surface-attached polymers to determine their chromatographic properties. For dilute solution, analytical expressions were developed that described the partitioning and retention of solute molecules at the solvent/brush interface, which was found to be dependent on chain configuration, entropy of mixing, and contact interactions. Depending on specific solute, solvent, and polymer brush interactions, separation depends on surface density and chain length of the surface-attached polymer chains, solvent size, and polarity. The theory allows for the calculation of average or global properties such as the polymer, solvent, and solute volume fractions in the interphase, the interphase thickness, and solute partition coefficients and retention factors. Size exclusion and enhancement, affinity, and gradient chromatography are considered [149]. Polymer brushes have been used in the preparation of molecularly imprinted polymers (MIPs) for chromatographic supports. The brushes were prepared using different grafting techniques (i.e., “grafting to” and “grafting from” with both covalent and physisorbed azo initiators) by photoinitiation of mixtures of methacrylic acid, ethylene glycol dimethacrylate, l-phenylalanine, and silica particles. The chromatographic properties of these MIP supports were examined as a function of polymer brush thickness, solvent, support diameter, crosslink density, and composition of the mobile phase. The MIPs prepared by “grafting from” were easily synthesized and showed superior, reproducible resolution. For porous particles, column efficiency was strongly dependent upon the amount of grafted polymer brush. Particles of 10 nm average pore diameter coated with 0.8 nm polymer brush had the highest column efficiency (plate number N = 700 m–1, antipode ca. 24 000 m–1). Properties such as sample load capacity tended to fall off as brush film thickness increased. The tunable film thickness allows the supports to be optimized for either high analytical efficiency or high preparative scale separations [150]. Porous silica decorated with poly(acrylamide) brushes can act as a stationary support for size-exclusion separation of lysozyme, a strongly basic protein. SIP-ATRP of acrylamide provided well-controlled, uniform films on nanoporous silica [151]. Polyethylene films which had been radiation grafted acrylic acid showed excellent adsorption of Eu, Gd, Tb, and Dy ions. Besides potential application as water filtration membranes, these films may also find use as fluorescence emission displays [152]. 17.3.6.2 Membranes The attachment of polymer brushes to membranes can impact a variety of fluid flow properties. One might envision that appropriately functionalized membrane surfaces can improve or enhance separation and resolution through selective adsorption of one component in a mixture. Chiral surfaces could be used for resolving enantiomeric mixtures of medicinal products. Alternatively, such coatings could pro-

347

348

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

vide nonfouling biosurfaces against intracellular proteins, and thus extend the life of blood dialysis membranes and equipment. Another application of polymer brushes involves their use as microvalves to control flow. This idea of using two closely spaced polymer brushes as a gate to control fluid flow has been explored both theoretically and experimentally. Theoretical investigation of the flow of a good solvent through two closely spaced polymer brush-coated surfaces revealed that the brushes respond to the flow by expanding in response to a shear flow, which causes a decrease in the cross-sectional flow area and a reduction in the flux rate. This pressure-sensitive behavior allows the polymer brush to act as both a sensor and a self-regulating valve. Modeling this behavior revealed that several flow regimes exist, and these were found to be dependent on brush height and gap distance between closely spaced parallel plates: 1) The brush pair acted as a constant-discharge microvalve for brushes that fill the gap moderately; 2) the brush pair acted as a cut-off microvalve and limited the maximum discharge for brushes that fill the gap (i.e., brush pairs that fill the gap >80%). It was also found that more sensitive microvalves would result with loosely grafted brushes than with densely grafted brushes [153]. Theory also predicts that polymer brushes can be used to create channels that can be opened and closed by controlling solvent properties and pH [154]. Porous polyethylene membranes were coated with the stimuli-responsive polymer poly(NIPAAM) using plasma-initiated grafting. High grafting rates were observed for up to 95 h after plasma treatment of the membrane surface. Surface analyses (SEM, XPS, FT-IR) showed that grafted polymer was found on both outer surfaces and inner pores of the membrane. Significant changes in aqueous flow during temperature cycling near the LCST of the poly(NIPPAM) were revealed by permeation studies. Flow increased significantly above the LCST with permeability ratios showing grafting density dependence. In contrast with the grafted membrane, the unfunctionalized membrane showed a linear increase in flow with increasing temperature which has been ascribed to reduced viscosity of the medium. The effective pore radii of the grafted membranes could be modeled both above and below the LCST using Hagen-Poiseuille’s law [155]. On a similar system, others reported decreased aqueous permeability below the LCST [156]. Poly(methacrylic acid) polymer brushes prepared by plasma polymerization of methacrylic acid on track-etch porous polycarbonate membranes displayed pH-dependent flow characteristics. This behavior results from variable conformational properties of the polyelectrolyte brush, which change depending on the pH of the medium. At low pH, the polymer is protonated and contracted, whereas at neutral or high pH the polymer chain is deprotonated and extended. This was used to effect a channel-gating process within the membrane and to modulate aqueous flow. Changes in water permeation rate of 8 mL min–1 at low pH to less than 1 mL min–1 at neutral pH were reported. This behavior contrasted strongly with the ungrafted membrane, which showed a water permeation rate near 13 mL min–1 over the entire pH range. In-situ AFM of the membranes revealed large changes in the pores under low and neutral pH conditions, thus reflecting the gating process of the polymer brush [157].

17.3 Applications

Redox control has been used to modulate aqueous flow through poly[3-carbamoyl-1-(p-vinylbenzyl)pyridinium chloride] which had been radical-grafted onto a fluoropolymer membrane from peroxide sites generated by glow-discharge. Decreased water permeation rates were observed when the extended polymer brush (which was in its oxidized, ionized state) was believed to cover the membrane pores. However, increased permeation rates were observed when the polymer brush was in its reduced, deionized state. The heights of extended and contracted states were calculated to be 44 and 7 nm, respectively, using Hagen-Poiseuille’s law. Ionic strength was found to affect permeation rate, particularly when the brush was in its reduced state. The redox process was reversible when performed under high ionic strength conditions which precluded formation of an insoluble reduced polymer complex [158]. Poly(glutamic acid) grafted onto poly(tetrafluoroethylene) membranes also showed aqueous flow rate dependence on pH. The density and length of the grafted chains played an important role in determining the permeation rate [159]. Polymer brushes and surface-attached polymers have also been prepared for use as pH-sensitive microfiltration membranes [160], to improve polyethylene membrane thermal stability [161], in the construction of fuel cells [162], in cation-exchange membranes for metal recovery from aqueous solution [163], and for binding of ionic surfactants to charged polymer brushes grafted onto porous hollow-fiber membranes [164]. 17.3.6.3 Microfluidics The development of microfluidic devices is a rapidly growing field which has important implications for bioanalytical analysis, studying reactions in microreactors, and understanding fluid mixing under flow [165]. Interest exists in the possibility that, through the use of patterned polymer brushes in a microfluidics channel, mixing and fluid flow in the device can be controlled. One study examined different parameters that would need to be controlled to prepare microfluidic channels functionalized with polymer brushes by photografting chemistry. The process was reported to be a simple and versatile approach that worked on a variety of commercially available polymer substrates with many different types of monomers [166]. Others have used surface-attached poly(dimethylacrylamide) to suppress electroosmotic flow in a sidearm channel of a microfluidic device, thus producing hydraulic pumping in that channel with resultant differential ion transport between channels [167]. 17.3.7

Nanofabrication

IBM researchers used a combination of “top down” and ”bottom up” approaches to pattern and prepare polymer brushes with controlled composition and size of nanoscale features. Contact-molding was used to transfer the pattern from an electron beam-fabricated silicon wafer master to a photopolymer matrix, which contained inimers (inimers = initiators/monomers) as one component of the formulation. Once cast, the surface-exposed inimers were used as sites for “grafting from” ATRP and nitroxide living free radical polymerization chemistries to produce well-con-

349

350

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

trolled polymer brushes with styrene, methyl methacrylate, and hydroxyethyl methacrylate. Through this process, features ranging in size from less than 60 nm up to lm dimensions could be replicated, with thicknesses ranging from 10 to 143 nm on both flat and nanopatterned surfaces. Significantly, this approach controlled nanoscopic features down to 20 nm. Current studies are aimed at the application of this approach to patterned magnetic media and molecular electronics [168]. Photolithography has been used to prepare hydrophobic and hydrophilic patterns of well-defined poly(t-butyl acrylate) and poly(acrylic acid) polymer brushes. The ester-containing brush was prepared by SIP on silicon wafers, and then converted to the acid after a sacrificial polystyrene/bis(t-butylphenyl)iodonium triflate photoresist layer was spin-coated on the brush layer, masked, and converted by photolithography. AFM revealed topographical differences between the ester (130 nm) and acid (80 nm) portions of the thin film. The hydrophobic and hydrophilic regions had water contact angles of 92K and 15K, respectively [169]. Other modes of nanofabrication that use polymer brushes have been investigated including patterning by chemical lithography [170], contact printing using ROMP [171], and fabrication of protein nanostructures by DPN [172]. Polymer brushes prepared by layer-by-layer grafting [173] show promise for making functionally gradient surfaces with tuned properties. Such interfaces should dramatically improve adhesion within composite materials through designed gradients in the modulus. Nanoporous surfaces have been prepared from crosslinked polymer brushes [174]. 17.3.8

Surfaces for Electronics

Polymer brushes have been used to make both insulating and conducting surfaces. While investigations of conductive polymer brushes for these types of applications are just beginning to appear in the literature, surface-attached thin films should be very useful as they will remain adhered to the surface throughout processing steps. In one application, they have served as templates in the fabrication of conducting polymer and complementary gold microstructures by being the insulating layer during electrodeposition [175]. Polymer brushes attached directly to silicon surfaces can eliminate the electrically defective silicon oxide layer at the molecular level, thereby improving a number of semiconductor processing issues [176]. Conductive polymers have been grafted onto polyethylene and poly(styrenesulfonic) acid films to give conductive poly(thiophene) [177,178] and poly(ethylenedioxythiophene) [179] surfaces, respectively. For use as chemical sensors, semiconductive poly(p-phenylene ethynylene) brushes have been “grafted from” oxidized silicon surfaces using ROMP. In addition to showing improved stability, the brushes also displayed higher emission quantum yields relative to spin-cast films because of the lack of aggregation [180].

17.4 Future Prospects

17.3.9

Other Uses

Advances in using controlled polymerization methodologies to prepare well-defined polymer brushes (i.e., high functionality, defined brush thickness, patterned surfaces, morphology) open new approaches to making materials with surfaces that have been designed on the nanoscale. This is especially true now that micro- and nano-scale surface features have been found to have considerable importance in cell behavior and function [181–183]. Well-controlled polymer brushes have been recently “grafted from” natural surfaces, including chitosan [184,185] and b-cyclodextrin [186] by ATRP and “grafted to” cellulose [187] by cationic polymerization [188]. Grafted polymers have also been attached to cellulose, hydroxypropyl cellulose, starch, potato starch, mercaptochitin, chitosan, casein proteins, and coconut husk by other methods, and have been found in a variety of applications including biodegradable polymers [189], improvement of resistance to bio- [190] and enzymatic [191] degradation, improvement of fiber dyeability [192], improvement of fiber water absorption [193], improved thermal stability and temperature-dependent swelling [194], improvement of blend compatibility [195], modification of biopolymer thermal and solubility properties [196], heavy metal removal from industrial waste water [197], altering membrane properties [198], and drug delivery [199,200]. Surfaceattached polymers grafted from non-natural materials have been used to make stable composite materials on silicon for multilayer dielectric mirrors [201] and to enhance dyeability of polyethylene after radiation grafting of N-phenyl- and p-hydroxy-N-phenylmaleimide [202].

17.4

Future Prospects

While the use of surface-attached polymers and polymer brushes in commercial applications is only just in its infancy, the impact of these materials on future applications holds great promise. The field is ripe for exploitation, as many applications and uses have yet even to be examined, or have only been demonstrated as proof-ofprinciple. This is clearly seen by the lack of patents or patent applications in this field. When the literature for this chapter was first assembled, a search was carried out via the US Patent and Trademark website for patents and patent applications in which “polymer brush” and “applications” were used as keywords. Of the 31 patents or patent applications uncovered in this search, only five dealt with polymer brushes as the key technology, and these involved bioseparations, bioadhesion, patterning, and antifogging coatings. Since then, several more applications have issued, but these have dealt primarily with process and composition of matter applications. Before the potential of polymer brushes can be truly realized, considerable advances in their synthesis must occur. Currently, polymer brush synthesis and characterization uses large excesses of reagents for performing surface-attachment reactions, whether they be self-assembly processes, coupling reactions on tethers, or

351

352

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

“grafting to”/“grafting from” chemistry. Cleaning these thin films requires incredible volumes of solvents – especially when one considers how little material actually is present on the surface. Improved reactions that use a fraction of the reagents are needed, and cleaning procedures that are fast, efficient, and minimize solvent recycle and waste must be developed before polymer brushes are considered seriously by technology industries. As the technology develops, cheap bulk polymers or polymers from renewable resources will be fabricated with highly designed, value-added surface properties such as stain and abrasion resistance, flame retardency, odor reduction, dyeability and color fastness. The marketplace will see materials with improved properties resulting in better interface control in adhesives, coatings, and composites. The emergence of new medical devices (e.g., catheters, eye lenses, biosensors, drug delivery vehicles) with improved properties and response behaviors will come about though application of polymer brush technologies.

Acknowledgments

These studies were supported by the Center for Biologically Inspired Materials and Materials Systems (CBIMMS) at Duke University and through funding from the National Science Foundation grant EEC-02-10590.

Substrate Silica particles

Silica particles (490 nm)

Silica nanoparticles (20 nm)

Silica nanoparticles (12 nm)

Silica nanoparticles (12 nm)

Crosslinked core-shell nanocomposites and hollow nanoparticles

Biodegradable nanocomposites particles

Homo- and block copolymer core-shell nanocomposites

Core-shell nanoparticles, filmforming properties

Core-shell hyperbranched nanocomposites with controlled hydrophobicity

“Grafting from” radical surfaceinitiated polymerization using VA, GlyMA, HEMA, MMA, NVP, Sty

Cu-based ATRP using nBA

Cu-based “grafting from” ATRP using Sty, nBA, MMA

“Grafting from” nitroxide-based polymerization using Sty, vinyl benzocyclobutene, or maleic anhydride Surface-initiated ring-opening polymerization of p-dioxanone

Methoda)

Applications of nanocomposites prepared using polymer brushes or surface-attached polymers.

Application

Table 17.1

Appendix

Reference

Formation of Silica/Poly(p-dioxanone) Microspheres by SurfaceInitiated Polymerization. Synthesis and Characterization of Organic/Inorganic Hybrid Nanoparticles: Kinetics of Surface-Initiated Atom Transfer Radical Polymerization and Morphology of Hybrid Nanoparticle Ultrathin Films. Atom Transfer Radical Polymerization of n-Butyl Acrylate from Silica Nanoparticles. Grafting of Hyperbranched Polymers onto Ultrafine Silica: Postgraft Polymerization of Vinyl Monomers Initiated by Pendant Initiating Groups of Polymer Chains Grafted onto the Surface.

207

206

205

204

Production of Crosslinked, Hollow 203 Nanoparticles by Surface-Initiated Living Free-Radical Polymerization.

Title

Appendix 353

Substrate

Alumina nanoparticles (10.4 nm)

Enhancement of interfacial interactions and the preparation of nanocomposites Amphiphilic metal-polymer nanocomposites with core-shell morphology

Gold nanoparticles (12 nm) on carbon and mica surfaces

Films of ordered arrays of gold nanoparticle hybrids

Water-soluble, dispersed gold Gold nanoparticles (12 nm) nanoparticles; nonspecific adhesion of biomolecules

Gold colloids

Core-shell nanocomposites

Gold nanoparticles

Silica nanoparticles (7 nm)

Core-shell nanocomposites

Stabilization of core-shell nanoSilica nanoparticles (12 nm) composites dispersions and controlling surface hydrophobicity

Application

Title

Reference

Effective Grafting of Polymers onto 208 Ultrafine Silica Surface: Photopolymerization of Vinyl Monomers Initiated by the System Consisting of Trichloroacetyl Groups on the Surface and Mn2(CO)10. Cationic surface-initiated polymer- Surface Functionalization of Silica 209 ization using 2-vinylfuran with 2-Vinylfuran by Cationic Polymerization. Radical-based “grafting onto” using Graft polymerization of Vinyl 210 AIBN and ammonium persulfate Monomers onto Nanosized Alumina with Sty and acrylamide Particles. “Grafting from” living cationic Nanocomposites by Surface-Initia- 211 polymerization ted Living Cationic Polymerization of 2-Oxazolines on Functionalized Gold Nanoparticles. “Grafting from” Cu-based ATRP Gold Nanoparticles with Covalently 212 with nBA Attached Polymer Chains. “Grafting from” Cu-based ATRP Fabrication of Ordered Arrays of 213 with MMA Gold Nanoparticles Coated with High-Density Polymer Brushes. “Grafting to” attachment of polySynthesis and Properties of Water- 214 (N-tris-(hydroxymethyl)methylSoluble Gold Colloids Covalently acrylamide and poly(N-isopropyl)- Derivatized with Neutral Polymer acrylamide Monolayers.

“Grafting from” photopolymerization using MMA and Sty

Methoda)

354

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

Substrate

Magnetite particles (0.23 lm)

Magnetite nanoparticles (10 nm)

Montmorillonite layered silicates

Montmorillonite clay

Montmorillonite clay adsorbed onto silicon and gold surfaces Montmorillonite

Magnetic separation of adenosinesubstituted sugars

Core-shell dispersed nanocomposites

Exfoliated montmorillonite nanocomposites

Exfoliated polymer-layered silicate nanocomposites

Polymer-layered silicate nanocomposites

Self-assembled polymer nanocomposites

Stabilized germanium nanoclusters Germanium nanoclusters (4 nm) for photoluminescence

Application

Title

Reference

Synthesis of Germanium Nanoclus- 215 ters with Irreversibly Attached Functional Groups: Acetals, Alcohols, Esters, and Polymers. Ce-redox-promoted “grafting from” Sugar-binding Property of Magnetite 216 using AA followed by coupling with Particles Modified with Dihydroxy3-aminophenylboronic acid borylphenyl Groups via Graft Polymerization of Acrylic Acid. “Grafting from” nitroxyl-mediated Polystyrene-Grafted Magnetite 217 living free radical polymerization of Nanoparticles Prepared through Sty using phosphoric ester anchor Surface-Initiated Nitroxyl-Mediated chemistry Radical Polymerization. Radical “grafting from” with nBA Nanocomposites from Layered 218 Silicates: Graft Polymerization with Intercalated Ammonium Peroxides. “Grafting from” radical polymeriGrafting of Polymers from Clay 219 zation of Sty Nanoparticles via in Situ Free Radical Surface-Initiated Polymerization: Monocationic versus Bicationic Initiators. “Grafting from” radical polymeriPolymer Brushes Grafted from Clay 220 zation of Sty Nanoparticles Adsorbed on a Planar Substrate by Free Radical SurfaceInitiated Polymerization. “Grafting to” thermal or photoSelf-Assembly of an Environmen221 chemical polymerization of dodecyl tally Responsive Polymer/Silica methacrylate and NIPPAM Nanocomposite.

ATRP of Sty (Cu) and MMA (Ni)

Methoda)

Appendix 355

Polystyrene core shell polymer on mica

Amino functionalized latex particles (70–245 nm)

Poly(divinylbenzene-80) microspheres; swellable, lightly crosslinked poly(divinylbenzene-cohydroxyethylmethacrylate) microspheres

Hydrophilic latex particles

Homopolymer and polyelectrolyte microspheres

“Grafting from” ATRP using HEMA, (dimethylamino)ethyl methacrylate, MMA, GlyMA, trimethylammonium ethylmethacrylate

Photoemulsion polymerization – “Grafting from” UV/vis radiation of 2-(acryloyloxy)ethyltrimethyl ammonium chloride onto polystyrene particles “Grafting to” attachment of poly(N-acryloylmorpholine)

Latex Particles Bearing Hydrophilic 226 Grafted Hairs with Controlled Chain Length and Functionality Synthesized by Reversible Addition–Fragmentation Chain Transfer. Formation and Morphology of 227 Methacrylic Polymers and Block Copolymers Tethered on Polymer Microspheres.

Cu-based “grafting from” ATRP using MMA and EA

Polyelectrolyte core-shell particles for controlling the interaction of latex particles with solid substrates

Grafting on Crosslinked Polymer 224 Beads by ATRP from Polymer Supported N-Chlorosulfonamides. Engineering the Interaction of Latex 225 Spheres with Charged Surfaces: AFM Investigation of Spherical Polyelectrolyte Brushes on Mica.

Cu-based ATRP using N,Ndimethyl-acrylamide, NIPAAM, PEGMA, HEMA

Reference

Core-shell nanocomposite particles Crosslinked polystyrene beads (100 lm) and poly(ethylene glycol) beads (100–200 lm, 10 lm, 150–200 lm Core-shell nanocomposites beads Spherical polystyrene beads with retention of bead shape (420–590 lm)

Title

Synthesis and Characterization of 222 Polymer Brushes of Poly(N,N-dimethylacrylamide) from Polystyrene Latex by Aqueous Atom Transfer Radical Polymerization. 223 Thick Coating and Functionalization of Organic Surfaces via ATRP in Water.

Methoda) “Grafting from” Cu- based ATRP with N,N-dimethylacrylamide

Substrate

Core-shell nanocomposite particles Polystyrene latex particles

Application

356

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

a)

MMA = methyl methacrylate, nBA = n-butyl acrylate, Sty = styrene, MA = methyl acrylate, AA = acrylic acid, HEMA = hydroxyethyl methacrylate, NIPAAM = N-isopropylacryl-

Improved dispersibility of organic pigment particles

Polymer-coated carbon nanotubes

Single-walled carbon nanotubes

Water-soluble carbon nanotubes composites

Title

amide, VA = vinyl acetate, NVP = N-vinyl pyrrolidinone, GlyMA = glycidol methacrylate, PEGMA = polyethylene glycol monomethacrylate, EA = ethyl acrylate.

Cu-based ATRP using MMA and t-butyl acrylate

Methoda) Reference

Polymerization from the Surface 228 of Single-Walled Carbon Nanotubes – Preparation and Characterization of Nanocomposites. Single-walled carbon nanotubes “Grafting from” ROMP using Ring-Opening Metathesis Poly229 pyrene-based anchoring groups merization on Non-Covalently Functionalized Single-Walled Carbon Nanotubes. Quinacridone, diketopyrrolopyrrole, Radical grafting using Sty, MMA, Graft Polymerization of Vinyl 230 and anthraquinone pigments and 2-isocyanatoethyl methacrylate Monomers Initiated by Azo Groups Introduced on Organic Pigment Surface.

Substrate

Application

Appendix 357

Substrate Loofah fibers

Nylon fibers

Polyethylene terephthalate (PET)

Polypropylene film

Polyaniline films

Polymethyl methacrylate lenses

Antibacterial activity

Light-activated antimicrobial materials

Blood compatibility, platelet adhesion

Improved blood compatibility of functionalized polyolefin films

Nonfouling biosurfaces; protein adsorption, platelet adhesion

Nonfouling biosurfaces for intraocular lenses

231

Graft Copolymerization of Vinyl Monomers Bearing Positive Charges or Episulfide Groups onto Loofah Fibers and Their Antibacterial Activity Porphyrin-Based, Light-Activated Antimicrobial Materials.

236

235

234

233

232

Reference

Title

Platelet Adhesion on Laser-Induced Acrylic Acid–Grafted Polyethylene Terephthalate. Radiation-induced “grafting from” Surface Modification of Polypropyusing 2,3-epoxypropyl methacrylate lene Film by Radiation(EPMA) Induced Grafting and Its Blood Compatibility. UV-induced photografting Surface modification of polyaniline film by grafting of polyethylene glycol for reduction in protein adsorption and platelet adhesion. UV-induced photografting Surface Modification of Polymethyl Methacrylate Intraocular Lenses with the Mixture of Acrylic Acid and Acrylamide via Plasma-Induced Graft Copolymerization.

Radiation grafting acrylic acid followed by reaction with protoporphyrin IX and zinc protoporphyrin IX Irradiation with a CO2 pulsed laser

Initiated with cerium ammonium nitrate or H2O2

Methoda)

Biological applications involving “grafting from” polymerization.

Application

Table 17.2

358

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

Substrate

Silica gel

Solid supports for nucleic acid chemistry

Sty = styrene, PET = polyethylene terephthalate.

Starch

Stabilization

a)

Polypropylene fibers

Wound dressings

Monitoring protein adsorption on Gold-coated silicon wafers polymers by SPR, patterning cells on micropatterned polymers fabricated using lCP and SIP Biomaterials for hepatocyte culture PET – argon-plasma-treated PET films

Lysozyme-resistant surface, stabili- Poly(dimethylsiloxane) (PDMS) zation of hydrophilic PDMS surface plates

Application

Title

Chemical Modification of the Surface of Poly(dimethylsiloxane) by Atom-Transfer Radical Polymerization of Acrylamide. Radical polymerization of Sty Surface-Initiated Free Radical Polymerization of Polystyrene Micropatterns on a Self-Assembled Monolayer on Gold. UV-induced photografting Immobilization of Galactose Ligands on Acrylic Acid Graft-Copolymerized Poly(ethylene terephthalate) Film and Its Application to Hepatocyte Culture. Radiation Development of a Poly(N-vinyl-2pyrrolidone)/Poly (ethylene glycol) Hydrogel Membrane Reinforced with Methyl Methacrylate-Grafted Polypropylene Fibers for Possible Use as Wound Dressing. Electron-beam irradiation Physical Stabilization of StarchAllylurea Blends by EB-Grafting: a Compositional and Structural Study. Cu-based ATRP using 5¢-methacry- Copper(I) Mediated Radical Polyloyluridine and 5¢-methacryloylade- merization of Uridine and Adenosine Monomers on a Silica Support. nosine

Cu-based ATRP using acrylamide

Methoda)

242

241

240

239

238

237

Reference

Appendix 359

Polystyrene

Blood compatibility, biocompatible materials, biomedical polymers Reduced fibrinogen adsorption

Polyethylene, polystyrene, poly(methyl methacrylate), and poly(ethylene terephthalate) films Polystyrene, polyethylene terephthalate, poly(methyl methacrylate), gold Polyethylene terephthalate film

Patterning of biomolecules on surfaces with microscale resolution

Fabricated micropatterned surfaces for cell adhesion Patterning of cell-adhesive peptide using elastomeric microwell reservoirs (Wellpat)

Highly antithrombogenic biomaterials with improved hemocompatibility

Nylon

Antimicrobial

Fluorinated ethylene propylene copolymer surface Polysulfone membranes

Substrate

Biological applications involving “grafting to” polymerization.

Application

Table 17.3

Grafting of Light-Activated Antimicrobial Materials to Nylon Films. Grafting of Oligopeptide on Poly(aminostyrene)s and Characterization of the Polymers. Surface Immobilization of Poly(Ethylene Oxide): Structure and Properties. Surface Modification of Polysulfone Membranes by Low-Temperature Plasma–Graft Poly(ethylene glycol) onto Polysulfone Membranes. Microstamping on an Activated Polymer Surface: Patterning Biotin and Streptavidin onto Common Polymeric Biomaterials. Universal Route to Cell Micropatterning Using an Amphiphilic Comb Polymer. Micropatterning Biological Molecules on a Polymer Surface Using Elastomeric Microwells.

Title

249

248

247

246

245

244

243

Reference

360

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

Nomex, Kermel, PBI/Kevlar blend fibers

Nylon-66, polyester (PET), polypropylene, acrylic (Orlon), polyester/cotton blend (PET/Cotton), cotton print cloth Polyester fibers

Antibacterial activity

Antimicrobial fabrics

High-pressure mercury lamp with acrylic acid K2S2O8/CuSO4 Initiated radical polymerization

Polyethylene terephthalate (PET) membranes Polyamide fabrics

Increased hydrophilicity, improved aqueous flow, higher protein resistance Improved moisture regain to enhance antistatic effect

Titanium(III) chloride–potassium persulfate redox initiator

Silk fibers

Radical polymerization with benzoyl peroxide

Radical initiation

Radical initiation

Method

Improved mechanical properties, and water readsorption

Improved water and oil repellancy and soil resistance

Substrate

Miscellaneous applications involving surface-attached polymers.

Application

Table 17.4

254

253

252

251

250

Reference

Grafting and Quaternization of 2255 (Dimethylamino)Ethyl Methacrylate onto Polyamide-6 Fabric Pretreated with Acetone.

Novel Refreshable N-Halamine Polymeric Biocides: Grafting Hydantoin-Containing Monomers onto High Performance Fibers by a Continuous Process. Durable and Regenerable Antimicrobial Textile Materials Prepared by a Continuous Grafting Process. Graft Copolymerization of a Mixture of Perfluorooctyl-2 Ethanol Acrylic and Stearyl Methacrylate onto Polyester Fibers using Benzoyl Peroxide as Initiator. Studies of Mechanical and Moisture Regain Properties of Methyl Methacrylate Grafted Silk Fibers. Photografting Modification of PET Nucleopore Membranes.

Title

Appendix 361

Substrate Coir fibers

Poly(ethylene terephthalate) fibers

Nylon fabrics

Polypropylene fabric

Glass fibers

Glass wool

Polypropylene nonwoven fabrics

Application

Improved hydrophobicity and mechanical properties

Improved water repellancy

Improved moisture regain and increased dyeability

Improve functionalization and dyeability

Improved interfacial properties

Property improvement of polypropylene glass wool composites

Odor-absorbing fabrics

K2S2O8/CuSO4 Initiated radical polymerization

Chemical Grafting of 2-Ethyl 258 Methacrylate Phosphoric Acid onto Nylon 6 Fabric. c-irradiation Synthesis and Characterization 259 of Novel Grafted Amphoteric Poly(propylene) Fabrics. Radical polymerization with methyl Radical Grafting from Glass Fiber 260 methacrylate, styrene, N-vinylcarba- Surface: Graft Polymerization of zole, acrylic acid Vinyl Monomers Initiated by Azo Groups Introduced onto the Surface. Potassium persulfate (PPS) and Preparation, Characterization, and 261 PPS/acetone sodium bisulfite Some Physical Properties of Polypro(ASBS) redox-pair initiation pylene/Poly-(methyl acrylate)-Grafted Glass Wool Composites. Photoinitiated radical polymeriAmmonia Adsorption Behavior of 262 zation Polypropylene Nonwoven Fabric Grafted with Acrylic Acid.

Benzoyl peroxide radical initiation

Surface Modification of Coir Fibers. 256 II. Cu(II)- IO4-Initiated Graft Copolymerization of Acrylonitrile onto Chemically Modified Coir Fibers. Wettability of Grafted Poly(ethylene 257 terephthalate) Fibers.

CuSO4/NaIO4

Reference

Title

Method

362

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

Substrate Poly(vinylidene fluoride)graft-poly(vinylbenzyl chloride) (PVDF-g-PVBC) membranes

Gold-coated surfaces

Glass slides

Application

Novel membranes for fuel-cell applications

Polylactide brushes and chiral surfaces

Liquid crystalline polymer brushes, displays

Title

Synthesis of Proton-Conducting Membranes by the Utilization of Preirradiation Grafting and Atom Transfer Radical Polymerization Techniques. Sn(OTf)2 catalyzed ring-opening Stannous(II) Trifluoromethane polymerization of lactide isomers, Sulfonate: A Versatile Catalyst for SAM initiator the Controlled Ring-Opening Polymerization of Lactides: Formation of Stereoregular Surfaces from Polylactide “Brushes”. Radical SIP using liquid crystalline Polymer Brushes with Liquid substituted methacrylate ester Crystalline Side Chains.

ATRP

Method

265

264

263

Reference

Appendix 363

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

364

References 1 As this review is concerned with organic

2 3

4

5

6 7 8

9 10 11

12

13

14

15 16 17 18 19

polymer brushes, the focus of this discussion is limited to organic polymers and neglects inorganic polymers or coatings. B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677–710. A. M. Granville, W. J. Brittain, in: Polymer Brushes: From Synthesis to Functional Microstructures (Ed.: R. C. Advincula, W. J. Brittain, K. C. Caster, J. RUhe), Wiley-VCH, Weinheim, Germany, 2004, Chapter 1. R. Jordan, A. Ulman, J. F. Kang, M. H. Rafailovich, J. Sokolov, J. Am. Chem. Soc. 1999, 121, 1016–1022. M. D. Ingall, C. H. Honeyman, J. V. Mercure, P. A. Bianconi, R. R. Kunz, J. Am. Chem. Soc. 1999, 121, 3607–3613. Z. Pi, J. P. Kennedy, J. Polym. Sci.: Part A: Polym. Chem. 2001, 39, 1675–1680. O. Prucker, J. RUhe, Langmuir 1998, 14, 6893–6898. K. Fujiki, M. Sakamoto, A. Yoshida, H. Maruyama, J. Polym. Sci., Part A: Polym. Chem. 1999, 37, 2121–2128. M. Baum, W. J. Brittain, Macromolecules 2002, 35, 610–615. Y. Tsujii, M. Ejaz, K. Sato, A. Goto, T. Fukuda, Macromolecules 2001, 34, 8872–8878. W. Chen, A. Y. Fadeev, M. C. Hsieh, D. Vner, J. Youngblood, T. J. McCarthy, Langmuir 1999, 15, 3395–3399. M. Husemann, D. Mecerreyes, C. J. Hawker, J. L. Hedrick, R. Shah, N. L. Abbott, Angew. Chem. Int. Ed. 1999, 38, 647–649. O. Prucker, C. A. Naumann, J. RUhe, W. Knoll, C. W. Frank, J. Am. Chem. Soc. 1999, 121, 8766–8770. N. Luo, J. B. Hutchison, K. S. Anseth, C. N. Bowman, J. Polym. Sci.: Part A: Polym. Chem. 2002, 40, 1885–1891. H. J. Lee, Y. Nakayama, T. Matsuda, Macromolecules 1999, 32, 6989–6995. C. Gurtner, A. W. Wun, M. J. Sailor, Angew. Chem. Int. Ed. 1999, 38, 1966–1968. J.-B. Kim, M. L. Bruening, G. L. Baker, J. Am. Chem. Soc. 2000, 122, 7616–7617. K. Ohno, K.-M. Koh, Y. Tsujii, T. Fukuda, Macromolecules 2002, 35, 8989–8993. S. Pascual, B. Coutin, M. Tardi, A. Polton, J.-P. Vairon, Macromolecules 1999, 32, 1432– 1437.

20 R. R. Shah, D. Merreceyes, M. Husemann,

I. Rees, N. L. Abbott, C. J. Hawker, J. L. Hedrick, Macromolecules 2000, 33, 597–605. 21 K. Matyjaszewski, J. Xia, Chem. Rev. 2001, 101, 2921–2990. 22 T. E. Patten, K. Matyjaszewski, Acc. Chem. Res. 1999, 32, 895–903. 23 D. Bontempo, N. Tirelli, K. Feldman, G. Masci, V. Cresenzi, J. A. Hubbell, Adv. Mater. 2002, 14, 1239–1241. 24 K. J. Ivin, J. C. Mol, Olefin Metathesis and Metathesis Polymerization, Academic Press, New York, 1997, Chapter 11. 25 M. Weck, J. J. Jackiw, R. R. Rossi, P. S. Weiss, R. H. Grubbs, J. Am. Chem. Soc. 1999, 121, 4088–4089. 26 N. Y. Kim, N. L. Jeon, I. S. Choi, S. Takami, Y. Harada, K. R. Finnie, G. S. Girolami, R. G. Nuzzo, G. M. Whitesides, P. E. Laibinis, Macromolecules 2000, 33, 2793–2795. 27 M. Kamigaito, T. Ando, M. Sawamoto, Chem. Rev. 2001, 101, 3689–3745. 28 K. Kato, E. Uchida, E.-T. Kang, Y. Uyama, Y. Ikada, Prog. Polym. Sci. 2003, 28, 209–259. 29 K. C. Caster, E. F. Tokas, C. G. Keck, M. E. Hontz, J. Mol. Catal. A: Chem. 2002, 190, 65–77. 30 K. C. Caster, R. D. Walls, Adv. Synth. Catal. 2002, 344, 764–770. 31 F. L. Klavetter, R. H. Grubbs, J. Am. Chem. Soc. 1988, 110, 7807. 32 J. H. Chen, F. E. Schubert, (Copytele, Inc.), US Patent, 6,194,488 B1, 2001. 33 B. L. Goodall, J. V. Standish, (B. F. Goodrich), European Patent, 0,424,833 B1, 1990. 34 S. Yamamoto, M. Ejaz, Y. Tsujii, M. Matsumoto, T. Fukuda, Macromolecules 2000, 33, 5602–5607. 35 J. Duchet, J.-P. Chapel, B. Chabert, J.-F. Gerard, Macromolecules 1998, 31, 8264–8272. 36 D. M. Jones, A. A. Brown, W. T. S. Huck, Langmuir 2002, 18, 1265–1269. 37 S. Mendez, L. K. Ista, G. P. LWpez, Langmuir 2003, 19, 8115–8116. 38 S. Yamamoto, M. Ejaz, Y. Tsujii, T. Fukuda, Macromolecules 2000, 33, 5608–5612. 39 T. Wu, K. Efimenko, J. Genzer, J. Am. Chem. Soc. 2002, 124, 9394–9395. 40 T. Wu, K. Efimenko, P. Vlcek, V. Subr, J. Genzer, Macromol. 2003, 36, 2448–2453.

References 41 M. R. Tomlinson, J. Genzer, Chem. Commun.

2003, 1350–1351. 42 M. R. Tomlinson, J. Genzer, Macromol. 2003, 36, 3449–3451. 43 T. Wu, K. Efimenko, J. Genzer, Macromolecules 2001, 34, 684–686. 44 O. Pyshkina, V. Sergeyev, A. Zezin, V. Kabanov, D. Gage, M. Cohen Stuart, Langmuir 2003, 19, 2000–2006. 45 D. M. Jones, A. A. Brown, W. T. S. Huck, Langmuir 2002, 18, 1265–1269. 46 G. Busch, E. Jaehne, X. Cai, S. Oberoi, H.-J. P. Adler, Synthetic Metals 2003, 137, 871–872. 47 C. Bressy-Brondino, B. Boutevin, Y. Hervaud, M. Gaboyard, J. Appl. Polym. Sci. 2002, 83, 2277–2287. 48 I. Maege, E. Jaehne, A. Henke, H.-J. P. Adler, C. Bram, C. Jung, M. Stratmann, Prog. Org. Coatings 1998, 34, 1–12. 49 J. RUhe, W. Knoll, J. Macromol. Sci.-Polym. Rev. 2002, C42, 91–138. 50 C. Park, J. Yoon, E. L. Thomas, Polymer 2003, 44, 6725–6760. 51 B. Zhao, W. J. Brittain, J. Am. Chem. Soc. 1999, 121, 3557–3558. 52 R. P. Quirk, R. T. Mathers, T. Cregger, M. D. Foster, Macromolecules 2002, 35, 9964–9974. 53 H. Retsos, A. F. Terzis, S. H. Anastasiadis, D. L. Anastassopoulos, C. Toprakcioglu, D. N. Theodorou, G. S. Smith, A. Menelle, R. E. Gill, G. Hadziioannou, Y. Gallot, Macromolecules 2002, 35, 1116–1132. 54 V. L. Osborne, D. M. Jones, W. T. S. Huck, J. Chem. Soc., Chem. Commun. 2002, 1838– 1839. 55 D. J. Dyer, Adv. Funct. Mater. 2003, 13, 667– 670. 56 C. D. Bain, E. B. Troughton, Y.-T. Tao, J. Evall, G. M. Whitesides, R. G. Nuzzo, J. Am. Chem. Soc. 1989, 111, 321–335. 57 Y. Xia, G. M. Whitesides, Angew. Chem., Int. Ed. Engl. 1998, 37, 550–575. 58 R. D. Piner, J. Zhu, F. Xu, S. Hong, C. A. Mirkin, Science 1999, 283, 661–663. 59 W.-Y. Chiang, W.-D. Yang, J. Appl. Polym. Sci. 1988, 35, 807–823. 60 E. RaphaXl, P.-G. de Gennes, J. Phys. Chem. 1992, 96, 4002–4007. 61 H. Ji, P.-G. de Gennes, Macromolecules 1993, 26, 520–525.

62 L. LYger, E. RaphaXl, H. Hervet, Adv. Polym.

Sci. 1999, 138, 186–225. 63 M. Deruelle, L. LYger, M. Tirrell, Macromole-

cules 1995, 28, 7419–7428. 64 L. LYger, Macromol. Symp. 2000, 149, 197–

205. 65 W. S. Gutowski, J. Adhesion 2003, 79, 445–

482. 66 W. S. Gutowski, L. Sheng, C. Filippou,

P. Hoobin, S. Petinakis, J. Adhesion 2003, 79, 483–519. 67 P. A. V. O’Rourke-Muisener, J. T. Koberstein, S. Kumar, Macromolecules 2003, 36, 771–781. 68 J. R. Long, N. Oyler, G. P. Drobny, P. S. Stayton, J. Am. Chem. Soc. 2002, 124, 6297–6303. 69 W. Frey, D. E. Meyer, A. Chilkoti, Langmuir 2003, 19, 1641–1653. 70 I. Szleifer, M. A. Carignano, Macromol. Rapid Commun. 2000, 21, 423–448. 71 D. Klee, Z. Ademovic, A. Bosserhoff, H. Hoecker, G. Maziolis, H.-J. Erli, Biomaterials 2003, 24, 3663–3670. 72 A. S. Hoffman, P. S. Stayton, V. Bulmus, G. Chen, J. Chen, C. Cheung, A. Chilkoti, Z. Ding, L. Dong, R. Fong, C. A. Lackey, C. J. Long, M. Miura, J. E. Morris, N. Murthy, Y. Nabeshima, T. G. Park, O. W. Press, T. Shimoboji, S. Shoemaker, H. J. Yang, N. Monji, R. C. Nowinski, C. A. Cole, J. H. Priest, J. M. Harris, K. Nakamae, T. Nishino, T. Miyata, J. Biomed. Mater. Res. 2000, 52, 577–586. 73 Y. G. Takei, T. Aoki, K. Sanui, N. Ogata, Y. Sakurai, T. Okano, Macromolecules 1994, 27, 6163–6166. 74 D. M. Jones, J. R. Smith, W. T. S. Huck, C. Alexander, Adv. Mater. 2002, 14, 1130– 1134. 75 M. Kaholek, W.-K. Lee, B. LaMattina, K. C. Caster, S. Zauscher, in: Polymer Brushes: From Synthesis to Functional Microstructures (Ed.: R. C. Advincula, W. J. Brittain, K. C. Caster, J. RUhe), Wiley-VCH, Weinheim, Germany, 2004, Chapter 19. 76 S. Mendez, L. K. Ista, G. P. LWpez, Langmuir 2003, 19, 8115–8116. 77 T. O. Collier, J. M. Anderson, A. Kikuchi, T. Okano, J. Biomed. Mater. Res. 2001, 59, 136–143. 78 Y. V. Pan, R. A. Wesley, R. Luginbuhl, D. D. Denton, B. D. Ratner, Biomacromolecules 2001, 2, 32–36.

365

366

17 Applications of Polymer Brushes and Other Surface-Attached Polymers 79 I. Willner, E. Katz, Angew. Chem. Int. Ed.

2000, 39, 1180–1218. 80 N. Nath, A. Chilkoti, Anal. Chem. 2003, 75, 709–715. 81 T. Okano, N. Yamada, M. Okuhara, H. Sakai, Y. Sakurai, Biomaterials 1995, 16, 297–303. 82 M. Ebara, M. Yamato, M. Hirose, T. Aoyagi, A. Kikuchi, K. Sakai, T. Okano, Biomacromolecules 2003; 4, 344–349. 83 A. Tourovskaia, T. Barber, B. T. Wickes, D. Hirdes, B. Grin, D. G. Castner, K. E. Healy, A. Folch, Langmuir 2003, 19, 4754–4764. 84 J. Hyun, H. Ma, P. Banerjee, J. Cole, K. Gonsalves, A. Chilkoti, Langmuir 2002, 18, 2975–2979. 85 G. Csucsa, R. Michelb, J. W. Lussic, M. Textor, G. Danuser, Biomaterials 2003, 24, 1713–1720. 86 C. Pale-Grosdemange, E. S. Simon, K. L. Prime, G. M. Whitesides, J. Am. Chem. Soc. 1991, 113, 12–20. 87 S. Herrwerth, W. Eck, S. Reinhardt, M. Grunze, J. Am. Chem. Soc. 2003, 125, 9359–9366. 88 P. Kingshott, J. Wei, D. Bagge-Ravn, N. Gadegaard, L. Gram, Langmuir 2003, 19, 6912–6921. 89 M. Zheng, F. Davidson, X. Huang, J. Am. Chem. Soc. 2003, 125, 7790–7791. 90 A. Chilkoti, H. Ma, J. Hyun, N. Nath, Polymer Preprints (American Chemical Society, Division of Polymer Chemistry) 2003, 44(1), 455–456. 91 H. Ma, J. Hyun, P. Stiller, A. Chilkoti, Adv. Mater. 2004, 16, 338–341. 92 M. B. Chan-Parka, S. S. Tan, Int. J. Adhesion Adhesives 2002, 22, 471–475. 93 Z. J. Yu, E. T. Kang, K. G. Neoh, Polymer 2002, 43, 4137–4146. 94 I. Fuhrmann, J. Karger-Kocsis, J. Appl. Polym. Sci. 2003, 89, 1622–1630. 95 G. S. Srinivasa Rao, R. C. Jain, J. Appl. Polym. Sci. 2003, 88, 2173–2180. 96 J. Lei, X. Liao, J. Appl. Polym. Sci. 2001, 81, 2881–2887. 97 P. Preishuber-PflUgl, R. Podolan, F. Stelzer, J. Mol. Catal. A: Chemical 2000, 160, 53–61. 98 E. F. Tokas, K. C. Caster, US Patent Application, 20020053379 A1, May 9, 2002. 99 J.-F. Joanny, Langmuir 1992, 8, 989–995. 100 A. Casoli, M. BrendlY, J. Schultz, P. Auroy, G. Reiter, Langmuir 2001, 17, 388–398.

101 G. S. Grest, Adv. Polym. Sci. 1999, 138, 149–

183. 102 U. Raviv, S. Giasson, N. Kampf, J.-F. Gohy,

R. JYrZme, J. Klein, Nature 2003, 425, 163– 165. 103 M. Garvey, Chemistry in Britain 2003, 39(2), 28–32. 104 P.-G. de Gennes, Nature 2001, 412, 385. 105 M. Ballauff, Macromol. Chem. Phys. 2003, 204, 220–234. 106 J. Pyun, K. Matyjaszewski, Chem. Mater. 2001, 13, 3436–3448. 107 E. P. K. Currie, W. Norde, M. A. Cohen Stuart, Adv. Colloid Interface Sci. 2003, 100– 102, 205–265. 108 N. Tsubokawa, Prog. Polym. Sci. 1992, 17, 417–470. 109 T. Liu, S. Jia, T. Kowalewski, K. Matyjaszewski, R. Casado-Portilla, J. Belmont, Langmuir 2003, 19, 6342–6345. 110 S. Yoshikawa, S. Machida, N. Tsubokawa, J. Polym. Sci. A: Polym. Chem. 1998, 36, 3165–3172. 111 H. Huang, L. F. Nazar, Angew. Chem. Int. Ed. 2001, 40, 3880–3884. 112 H. Skaff, M. F. Ilker, E. B. Coughlin, T. Emrick, J. Am. Chem. Soc. 2002, 124, 5729–5733. 113 K. J. Watson, J. Zhu, S. T. Nguyen, C. A. Mirkin, J. Am. Chem. Soc. 1999, 121, 462–463. 114 A. G. M. Barrett, S. M. Cramp, R. S. Roberts, Organic Lett. 1999, 1, 1083–1086. 115 M. Bartz, J. KUther, R. Seshadri, W. Tremel, Angew. Chem. Int. Ed. 1998, 37, 2466–2468. 116 D. D. Wisnoski, W. H. Leister, K. A. Strauss, Z. Zhao, C. W. Lindsley, Tetrahedron Lett. 2003, 44, 4321–4325. 117 L. Feng, Y. Song, J. Zhai, B. Liu, J. Xu, L. Jiang, D. Zhu, Angew. Chem. Int. Ed. 2003, 42, 800–802. 118 L. Feng, S. Li, Y. Li, H. Li, L. Zhang, J. Zhai, Y. Song, B. Liu, L. Jiang, D. Zhu, Adv. Mater. 2002, 14, 1857–1860. 119 Z.-Z. Gu, A. Fujishima, O. Sato, Angew. Chem. Int. Ed. 2002, 41, 2068–2070. 120 C. Gay, Macromolecules 1997, 30, 5939–5943. 121 J. I. Martin, Z.-G. Wang, M. Schick, Langmuir 1996, 12, 4950–4959. 122 J. H. Maas, F. A. M. Leermakers, G. J. Fleer, M. A. Cohen Stuart, Macromol. Symp. 2003, 191, 69–79. 123 A. Carlmark, E. Malmstr[m, J. Am. Chem. Soc. 2002, 124, 900–901.

References 124 I. Luzinov, V. V. Tsukruk, Macromolecules 125 126 127 128

129 130 131

132

133

134 135

136

137 138

139

140

141

142

2002, 35, 5963–5973. H. Zhang, J. RUhe, Macromolecules 2003, 36, 6593–6598. R. A. Gage, E. P. K. Currie, M. A. Cohen Stuart, Macromolecules 2001, 34, 5078–5080. F. Brochard-Wyart, O. O. Ramdane, Y. Tran, P. Auroy, Macromol. Symp. 2003, 191, 59–68. Y. Gu, F. Nederberg, R. K\nge, R. R. Shah, C. J. Hawker, M. M[ller, J. L. Hedrick, N. L. Abbott, ChemPhysChem 2002, 448–451. A. Voronov, O. Shafranska, Polymer 2003, 44, 277–281. S. R. Edgecombe, J. M. Gardiner, M. W. Matsen, Macromolecules 2002, 35, 6475–6477. S. L. Cram, G. M. Spinks, G. G. Wallace, H. R. Brown, J. Appl. Polym. Sci. 2003, 87, 765–773. M. Claes, S. Voccia, C. Detrembleur, C. JYrZme, B. Gilbert, Ph. Lecl]re, V. M. Geskin, R. Gouttebaron, M. Hecq, R. Lazzaroni, R. JYrZme, Macromolecules 2003, 36, 5926– 5933. C. Detrembleur, C. JYrZme, M. Claes, P. Louette, R. JYrZme, Angew. Chem. Int. Ed. 2001, 40, 1268–1271. S. Roux, A.-S. Duwez, S. Demoustier-Champagne, Langmuir 2003, 19, 306–313. B. Zhao, W. J. Brittain, W. Zhou, S. Z. D. Cheng, J. Am. Chem. Soc. 2000, 122, 2407–2408. B. Zhao, W. J. Brittain, W. Zhou, S. Z. D. Cheng, Macromolecules 2000, 33, 8821–8827. B. Zhao, W. J. Brittain, Macromolecules 2000, 33, 8813–8820. D. Julthongpiput, Y.-H. Lin, J. Teng, E. R. Zubarev, V. V. Tsukruk, J. Am. Chem. Soc. 2003, 125, 15912–15921. S. G. Boyes, W. J. Brittain, X. Weng, S. Z. D. Cheng, Macromolecules 2002, 35, 4960–4967. J. Wang, S. Kara, T. E. Long, T. C. Ward, J. Polym. Sci.: Part A: Polym. Chem. 2000, 38, 3742–3750. S. Minko, S. Patil, V. Datsyuk, F. Simon, K.-J. Eichhorn, M. Motornov, D. Usov, I. Tokarev, M. Stamm, Langmuir 2002, 18, 289–296. L. Ionov, S. Minko, M. Stamm, J.-F. Gohy, R. JYrZme, A. Scholl, J. Am. Chem. Soc. 2003, 125, 8302–8306.

143 S. H. Anastasiadis, H. Retsos, S. Pispas,

N. Hadjichristidis, S. Neophytides, Macromolecules 2003, 36, 1994–1999. 144 S. Minko, M. MUller, M. Motornov, M. Nitschke, K. Grundke, M. Stamm, J. Am. Chem. Soc. 2003, 125, 3896–3900. 145 A. Kikuchi, T. Okano, Prog. Polym. Sci. 2002, 27, 1165–1193. 146 R. Dappen, H. Arm, V. R Meyer, J. Chromatogr. 1986, 373, 1–20. 147 B. M. Steels, J. Koska, C. A. Haynes, J. Chromatogr. B, 2000, 743, 41–56. 148 T. Kawai, K. Saito, W. Lee, J. Chromatogr. B 2003, 790, 131–142. 149 J. H. van Zanten, Macromolecules 1994, 27, 6797–6807. 150 C. Sulitzky, B. RUckert, A. J. Hall, F. Lanza, K. Unger, B. Sellergren, Macromolecules 2002, 35, 79–91. 151 X. Huang, M. J. Wirth, Anal. Chem. 1997, 69, 4577–4580. 152 S.-H. Choi, K.-P. Lee, S.-H. Sohn, J. Appl. Polym. Sci. 2003, 87, 328–336. 153 E. M. Sevick, D. R. M. Williams, Macromolecules 1994, 27, 5285–5290. 154 R. Israels, D. Gersappe, M. Fasolka, V. A. Roberts, A. C. Balazs, Macromolecules 1994, 27, 6679–6682. 155 J. Huang, X. Wang, X. Chen, X. Yu, J. Appl. Polym. Sci. 2003, 89, 3180–3187. 156 S.-Y. Kim, T. Kanamori, T. Shinbo, J. Appl. Polym. Sci. 2002, 84, 1168–1177. 157 Y. Ito, Y. S. Park, Y. Imanishi, J. Am. Chem. Soc. 1997, 119, 2739–2740. 158 Y. Ito, S. Nishi, Y. S. Park, Y. Imanishi, Macromolecules 1997, 30, 5856–5859. 159 Y. Ito, Y. Ochiai, Y. S. Park, Y. Imanishi, J. Am. Chem. Soc. 1997, 119, 1619–1623. 160 R. F. Childs, J. Weng, M. Kim, J. M. Dickson, J. Polym. Sci.: Part A: Polym. Chem. 2002, 40, 242–250. 161 B. Gupta, N. Anjum, J. Appl. Polym. Sci. 2001, 82, 2629–2635. 162 T. N. Danks, R. C. T. Slade, J. R. Varcoe, J. Mater. Chem. 2003, 13, 712–721. 163 N. M. El-Sawy, F. A. Al Sagheer, J. Appl. Polym. Sci. 2002, 85, 2692–2698. 164 H. Ogawa, K. Sugita, K. Saito, M. Kim, M. Tamada, A. Katakai, T. Sugo, J. Chromatogr. A 2002, 954, 89–97. 165 R. F. Ismagilov, Angew. Chem. Int. Ed. 2003, 42, 4130–4132.

367

368

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

166 T. Rohr, D. F. Ogletree, F. Svec, J. M. J. FrY-

chet, Adv. Funct. Mater. 2003, 13, 264–270. 167 C. T. Culbertson, R. S. Ramsey, J. M. Ramsey, Anal. Chem. 2000, 72, 2285–2291. 168 T. A. von Werne, D. S. Germack, E. C. Hagberg, V. V. Sheares, C. J. Hawker, K. R. Carter, J. Am. Chem. Soc. 2003, 125, 3831–3838. 169 M. Husemann, M. Morrison, D. Benoit, J. Frommer, C. M. Mate, W. D. Hinsberg, J. L. Hedrick, C. J. Hawker, J. Am. Chem. Soc. 2000, 122, 1844–1845. 170 U. Schmelmer, R. Jordan, W. Geyer, W. Eck, A. G[lzh^user, M. Grunze, A. Ulman, Angew. Chem. Int. Ed. 2003, 42, 559–563. 171 Y. Harada, G. S. Girolami, R. G. Nuzzo, Langmuir 2003, 19, 5104–5114. 172 J. Hyun, S. J. Ahn, W. K. Lee, A. Chilkoti, S. Zauscher, Nano Lett. 2002, 2, 1203–1207. 173 B. Sellergren, B. RUckert, A. J. Hall, Adv. Mater. 2002, 14, 1204–1208. 174 U. Jeong, D. Y. Ryu, J. K. Kim, D. H. Kim, T. P. Russell, C. J. Hawker, Adv. Mater. 2003, 15, 1247–1250. 175 F. Zhou, W. Liu, J. Hao, T. Xu, M. Chen, Q. Xue, Adv. Funct. Mater. 2003, 13, 938–942. 176 A. Juang, O. A. Scherman, R. H. Grubbs, N. S. Lewis, Langmuir 2001, 17, 1321–1323. 177 D. E. Bergbreiter, M. L. Liu, J. Polym. Sci.: Part A: Polym. Chem. 2001, 39, 4119–4128. 178 N. Chanunpanich, A. Ulman, Y. M. Strzhemechny, S. A. Schwarz, J. Dormicik, A. Janke, H. G. Braun, T. KratzmUller, Polym. Int. 2003, 52, 172–178. 179 K. L. Mulfort, J. Ryu, Q. Zhou, Polymer 2003, 44, 3185–3192. 180 J. H. Moon, T. M. Swager, Macromolecules 2002, 35, 6086–6089. 181 R. G. Flemming, C. J. Murphy, G. A. Abrams, S. L. Goodman, P. F. Nealey, Biomaterials 1999, 20, 573–588. 182 A. Thapa, D. C. Miller, T. J. Webster, K. M. Haberstroh, Biomaterials 2003, 24, 2915–2926. 183 A. Brock, E. Chang, C.-C. Ho, P. LeDuc, X. Jiang, G. M. Whitesides, D. E. Ingber, Langmuir 2003, 19, 1611–1617. 184 D. W. Jenkins, S. M. Hudson, Chem. Rev. 2001, 101, 3245–3273. 185 K. El Tahlawy, S. M. Hudson, J. Appl. Polym. Sci. 2003, 89, 901–912. 186 D. M. Haddleton, K. Ohno, Biomacromolecules 2000, 1, 152–156.

187 N. Tsubokawa, T. Iida, T. Takayama, J. Appl.

Polym. Sci. 2000, 75, 515–522. 188 S. Spange, Prog. Polym. Sci. 2000, 25, 781–

849. 189 C. Wang, Y. Dong, H. Tan, J. Polym. Sci.: Part

A: Polym. Chem. 2003, 41, 273–280. 190 F. E. Okieimen, J. Appl. Polym. Sci. 2003, 89,

913–923. 191 K. Kurita, M. Inoue, M. Harata, Biomacro-

molecules 2002, 3, 147–152. 192 M. W. Sabaa, S. M. Mokhtar, Polymer Testing

2002, 21, 337–343. 193 M. I. Lor_a-Bastarrachea, H. J. Carrillo-Esca-

194 195 196 197 198 199

200

201

202 203

204 205

206

207

lante, M. J. Aguilar-Vega, J. Appl. Polym. Sci. 2002, 83, 386–393. K. C. Gupta, K. Khandekar, Biomacromolecules 2003, 4, 758–765. A. Olivier, F. Cazaux, C. Gors, X. Coqueret, Biomacromolecules 2000, 1, 282–289. Q. Dong, Y.-L. Hsieh, J. Appl. Polym. Sci. 2000, 77, 2543–2551. M. K. Sreedhar, T. S. Anirudhan, J. Appl. Polym. Sci. 2000, 75, 1261–1269. H. Liu, Y.-L. Hsieh, J. Polym. Sci.: Part B: Polym. Physics 2003, 41, 953–964. S. Geresh, Y. Gilboa, J. Peisahov-Korol, G. Gdalevsky, J. Voorspoels, J. P. Remon, J. Kost, J. Appl. Polym. Sci. 2002, 86, 1157– 1162. C. Radhakumary, G. Divya, P. D. Nair, S. Mathew, C. P. R. Nair, J. Macromol. Sci., Part A: Pure Appl. Chem. 2003, A40, 715–730. M. S. Yoon, K. H. Ahn, R. W. Cheung, H. Sohn, J. R. Link, F. Cunin, M. J. Sailor, Chem. Commun. 2003, 680–681. R. O. Aly, T. B. Mostafa, S. M. Mokhtar, Polymer Testing 2002, 21, 857–865. S. Blomberg, S. Ostberg, E. Harth, A. W. Bosman, B. Van Horn, C. J. Hawker, J. Polym. Sci.: Part A: Polym. Chem. 2002, 40, 1309–1320. K. R. Yoon, Y.-J. Koh, I. S. Choi, Macromol. Rapid Commun. 2003, 24, 207–210. J. Pyun, S. Jia, T. Kowalewski, G. D. Patterson, K. Matyjaszewski, Macromolecules 2003, 36, 5094–5104. G. Carrot, S. Diamanti, M. Manuszak, B. Charleux, J.-P. Vairon, J. Polym. Sci.: Part A: Polym. Chem. 2001, 39, 4294–4301. S. Hayashi, K. Fujiki, N. Tsubokawa, Reactive & Functional Polymers 2000, 46, 193–201.

References 208 Y. Shirai, K. Shirai, N. Tsubokawa, J. Polym.

226 F. D’agosto, M.-T. Charreyre, C. Pichot,

Sci.: Part A: Polym. Chem. 2001, 39, 2157– 2163. 209 S. H[hne, S. Spange, Macromol. Chem. Phys. 2003, 204, 841–849. 210 M. Z. Rong, Q. L. Ji, M. Q. Zhang, K. Friedrich, Eur. Polym. J. 2002, 38, 1573–1582. 211 R. Jordan, N. West, A. Ulman, Y.-M. Chou, O. Nuyken, Macromolecules 2001, 34, 1606– 1611. 212 S. Nuß, H. B[ttcher, H. Wurm, M. L. Hallensleben, Angew. Chem. Int. Ed. 2001, 40, 4016–4018. 213 K. Ohno, K. Koh, Y. Tsujii, T. Fukuda, Angew. Chem. Int. Ed. 2003, 42, 2751–2754. 214 C. Mangeney, F. Ferrage, I. Aujard, V. Marchi-Artzner, L. Jullien, O. Ouari, E. D. RYkai, A. Laschewsky, I. Vikholm, J. W. Sadowski, J. Am. Chem. Soc. 2002, 124, 5811–5821. 215 R. S. Tanke, S. M. Kauzlarich, T. E. Patten, K. A. Pettigrew, D. L. Murphy, M. E. Thompson, H. W. H. Lee, Chem. Mater. 2003, 15, 1682–1689. 216 M. Shimomura, T. Abe, Y. Sato, K. Oshima, T. Yamauchi, S. Miyauchi, Polymer 2003, 44, 3877–3882. 217 R. Matsuno, K. Yamamoto, H. Otsuka, A. Takahara, Chem. Mater. 2003, 15, 3–5. 218 M. Albrecht, S. Ehrler, A. MUhlebach, Macromol. Rapid Commun. 2003, 24, 382–387. 219 X. Fan, C. Xia, R. C. Advincula, Langmuir 2003, 19, 4381–4389. 220 X. Fan, C. Xia, T. Fulghum, M.-K. Park, J. Locklin, R. C. Advincula, Langmuir 2003, 19, 916–923. 221 G. Garnweitner, B. Smarsly, R. Assink, W. Ruland, E. Bond, C. J. Brinker, J. Am. Chem. Soc. 2003, 125, 5626–5627. 222 K. N. Jayachandran, A. Takacs-Cox, D. E. Brooks, Macromolecules 2002, 35, 4247–4257. 223 D. Bontempo, N. Tirelli, G. Masci, V. Crescenzi, J. A. Hubbell, Macromol. Rapid Commun. 2002, 23, 417–422. 224 B. F. Senkal, N. Bicak, Eur. Polym. J. 2003, 39, 327–331. 225 Y. Mei, A. Wittemann, G. Sharma, M. Ballauff, Th. Koch, H. Gliemann, J. Horbach, Th. Schimmel, Macromolecules 2003, 36, 3452–3456.

R. G. Gilbert, J. Polym. Sci.: Part A: Polym. Chem. 2003, 41, 1188–1195. 227 G. Zheng, H. D. H. St[ver, Macromolecules 2003, 36, 1808–1814. 228 Z. Yao, N. Braidy, G. A. Botton, A. Adronov, J. Am. Chem. Soc. 2003, 125, 16015–16024. 229 F. J. GWmez, R. J. Chen, D. Wang, R. M. Waymouth, H. Dai, Chem. Commun. 2003, 190–191. 230 N. Tsubokawa, M. Kobayashi, T. Ogasawara, Prog. Org. Coatings 1999, 36, 39–44. 231 T. Nonaka, E. Noda, S. Kurihara, J. Appl. Polym. Sci. 2000, 77, 1077–1086. 232 J. Bozja, J. Sherrill, S. Michielsen, I. Stojiljkovic, J. Polym. Sci.: Part A: Polym. Chem. 2003, 41, 2297–2303. 233 H. Mirzadeh, M. Dadsetan, N. Sharifi-Sanjani, J. Appl. Polym. Sci. 2002, 86, 3191–3196. 234 O. H. Kwon, Y. C. Nho, J. Chen, J. Appl. Polym. Sci. 2003, 88, 1726–1736. 235 Y. Chen, E. T. Kang, K. G. Neoh, P. Wang, K. L. Tan, Synthetic Metals 2000, 110, 47–55. 236 T. S. Suh, C.-K. Joo, Y. C. Kim, M. S. Lee, H. K. Lee, B. Y. Choe, H. J. Chun, J. Appl. Polym. Sci. 2002, 85, 2361–2366. 237 D. Xiao, H. Zhang, M. Wirth, Langmuir 2002, 18, 9971–9976. 238 J. Hyun, A. Chilkoti, Macromolecules 2001, 34, 5644–5652. 239 L. Ying, C. Yin, R. X. Zhuo, K. W. Leong, H. Q. Mao, E. T. Kang, K. G. Neoh, Biomacromolecules 2003, 4, 157–165. 240 L. C. LopYrgolo, A. B. Lug¼o, L. H. Catalaini, J. Appl. Polym. Sci. 2002, 86, 662–666. 241 A. Olivier, F. Cazaux, C. Gors, X. Coqueret, Biomacromolecules 2000, 1, 282–289. 242 A. Marsh, A. Khan, M. Garcia, D. M. Haddleton, Chem. Commun., 2000, 2083–2084. 243 J. Sherrill, S. Michielsen, I. Stojiljkovic, J. Polym. Sci.: Part A: Polym. Chem. 2003, 41, 41–47. 244 K. Se, R. Ohtomo, J. Appl. Polym. Sci. 2000, 77, 1558–1567. 245 X. Gong, L. Dai, H. J. Griesser, A. W. H. Mau, J. Polym. Sci.: Part B: Polym. Physics 2000, 38, 2323–2332. 246 Y.-Q. Song, J. Sheng, M. Wei, X.-B. Yuan, J. Appl. Polym. Sci. 2000, 78, 979–985. 247 J. Hyun, Y. Zhu, A. Liebmann-Vinson, T. P. Beebe, A. Chilkoti, Langmuir 2001, 17, 6358–6367.

369

370

17 Applications of Polymer Brushes and Other Surface-Attached Polymers

248 J. Hyun, H. Ma, T. P. Beebe, A. Chilkoti,

Adv. Mater. 2003, 15, 576–579. 249 J. Hyun, A. Chilkoti, J. Am. Chem. Soc. 2001, 123, 6943–6944. 250 Y. Sun, G. Sun, J. Appl. Polym. Sci. 2003, 88, 1032–1039. 251 Y. Sun, G. Sun, J. Appl. Polym. Sci. 2002, 84, 1592–1599. 252 D. Sabhi, A. El-Achari, A. Ghenaim, C. CazY, Polymer Testing 2002, 21, 607–612. 253 T. K. Maji, D. Basu, C. Datta, A. Banerjee, J. Appl. Polym. Sci. 2002, 84, 969–974. 254 B. Yang, W. Yang, J. Macromol. Sci, Part A: Pure Appl. Chem. 2003, A40, 309–320. 255 A. M. Ramadan, S. Mosleh, S. M. Gawish, J. Appl. Polym. Sci. 2001, 81, 2318–2323. 256 J. Rout, M. Misra, S. S. Tripathy, S. K. Nayak, A. K. Mohanty, J. Appl. Polym. Sci. 2002, 84, 75–82. 257 D. Sabhi, A. El-Achari, A. Ghenaim, C. CazY, Polymer Testing 2002, 21, 615–618.

258 S. Mosleh, J. Appl. Polym. Sci. 2000, 77,

1357–1361. 259 S. Mosleh, S. M. Gawish, J. Appl. Polym. Sci.

2003, 88, 2504–2510. 260 K. Fujiki, M. Sakamoto, A. Yoshida, H. Mar-

uyama, J. Polym. Sci.: Part A: Polym. Chem. 1999, 37, 2121–2128. 261 G. M. Turky, M. M. Madani, M. H. AbdelRehim, A. S. Badran, J. Appl. Polym. Sci. 2003, 87, 723–732. 262 S. Y. Kim, C. N. Choi, J. Appl. Polym. Sci. 2002, 84, 295–301. 263 S. Holmberg, P. Holmlund, C.-E. WilYn, T. Kallio, G. Sundholm, F. Sundholm, J. Polym. Sci.: Part A: Polym. Chem. 2002, 40, 591–600. 264 M. M[ller, F. Nederberg, L. S. Lim, R. K\nge, C. J. Hawker, J. L. Hedrick, Y. Gu, R. Shah, N. L. Abbott, J. Polym. Sci.: Part A: Polym. Chem. 2001, 39, 3529–3538. 265 B. Peng, D. Johannsmann, J. RUhe, Macromolecules 1999, 32, 6759–6766.

371

18

Polymer Brushes: Towards Applications Gregory L. Whiting, Tamer Farhan, and Wilhelm T. S. Huck

Abstract

Recent research in the area of polymer brushes has advanced the field to such a point that possible applications for this material can begin to be investigated. Here, we examine potential uses of polymer brushes for organic electronic devices. The synthesis of polymer brushes from flexible plastic substrates is explored, as this development may assist research into all-polymeric devices. In addition, the synthesis of a hole-transporting polymer brush from an indium tin oxide surface, which will form active components in future polymeric devices, is also demonstrated.

18.1

Introduction

The formation of polymer brushes using surface-initiated polymerizations has seen an enormous rise in popularity over the past five years. Polymer brushes have important advantages over self-assembled monolayers in surface modification applications. Polymers are chemically and mechanically more robust, and they can incorporate a greater diversity of functional groups and introduce specific properties that cannot be obtained with small molecules. There are numerous polymerization techniques that can be applied to the synthesis of polymer brushes. These include uncontrolled techniques, such as free radical polymerization [1] as well as controlled techniques, such as reversible addition-fragmentation polymerization (RAFT) [2], nitroxide-mediated radical polymerization (NMP) [3], ionic polymerization [4], and ring-opening metathesis polymerization (ROMP) [5]. For this chapter, the most important technique used for the synthesis of polymer brushes is atom transfer radical polymerization (ATRP), particularly in aqueous or very polar solvents. Early examples of polymer brushes synthesized via ATRP are shown by Huang et al. [6], though this polymerization suffers from a lack of control over the polymer brush thickness. A method to control the growth of polymer brushes, by using a sacrificial initiator, was demonstrated by Husseman et al. [3]. A problem with this method of controlling polymer brush growth is that initiation occurs in solution,

372

18 Polymer Brushes: Towards Applications

giving rise to free polymer that must be removed. Another method to control the growth of polymer brushes via ATRP is to add a Cu(II) salt to the polymerization [7]. This method of control is considered surface-confined, such that no polymer is generated in solution. A combinatorial study of the effect of initiator density was carried out by Wu et al. [8]. The results showed that as initiator density is decreased, the brush thickness also decreases, until a certain point where there is a crossover to a mushroom morphology. At the crossover point, the density of initiator is not dense enough to force the polymer chains into a stretched conformation. In this regime the thickness of the polymer layer does not change with a change in initiator density. In related investigations, Huck and coworkers showed that the initiator efficiency on the surface is relatively low, as even on 100% initiator surfaces, only a maximum of 10% of initiating sites carries a polymer brush [9]. An example of a responsive surface using polymer brushes was shown by Jones et al. [10]. In this study, poly(N-isopropyl acrylamide) (PNIPAM) brushes were synthesized, and the polymer showed interesting phase behavior in aqueous solutions. Poly(N-isopropyl acrylamide) can be considered to be hydrophilic below its lower critical solution temperature, and hydrophobic above it. The brushes grown in this study are investigated using atomic force microscopy (AFM) and show a significant change in surface properties when the surface is brought above the lower critical solution temperature. Although the synthesis of polymer brushes is now well established, applications of polymer brushes are only just beginning to be explored. Potentially, these materials could have applications in many areas where polymer surface coatings are employed. In this chapter, we will demonstrate that highly controlled surface initiated and surface-controlled polymerizations are powerful synthetic tools that could lead to the use of polymer brushes in optoelectronic devices, or be used to modify polymer substrates that could be utilized for all-polymeric devices.

18.2

Experimental 18.2.1

Materials

Methyl methacrylate (C5H8O2, 99%, MEHQ stabilized; Aldrich) was passed through an alumina column prior to polymerization. Triethylamine (Lancaster) was distilled over KOH prior to use. All solvents were distilled before use, except DMF (Fluka; anhydrous) and isopropanol (Aldrich; HPLC grade), which were used as received. All water used was either demineralized or purified using a Millipore system (18.2 MX). All other reagents were purchased from Fluka, Aldrich, Acros and Lancaster, and used as received. Poly(ethylene naphthalate) (PEN) and poly(ethylene terephthalate) (PET) films were provided by Dupont Teijin Films. Silicon wafers were obtained from Compart Technology Ltd. (100 mm diameter, boron-doped,

18.2 Experimental

<100> orientated, polished one side). Glass slides pre-coated with indium tin oxide (ITO) were purchased from Donnely, Inc. 18.2.2

Characterization

Ellipsometric measurements were performed on a DRE ELX-02C ellipsometer at a wavelength of 632.8 nm at an incident angle of 70K. Surface topographical images were obtained using a Digital Instruments Nanoscope IV Dimension 3100 Atomic Force Microscope in tapping mode. Static contact angle data were determined using a system composed of a homemade stage, a kd Scientific computer-controlled microsyringe, and a COHU high-performance CCD camera. 18.2.3

Synthesis of Triphenylamine Acrylate (TPAA) Monomer

TPAA synthesis was carried out using a method similar to that of Tamada et al. [11]. 4-(Diphenylamino)benzaldehyde was reduced to the corresponding alcohol using LiAlH4. This alcohol was then coupled to acryloyl chloride in the presence of triethylamine to give the solid yellow TPAA monomer. 18.2.4

Synthesis and Deposition of Trichlorosilane ATRP Initiator

The ATRP initiator (3-trichlorosilylpropyl 2-bromo-2-methylpropanoate), which was used for all surface-initiated polymerizations, was synthesized following a procedure described by Husseman et al. [3]. Prior to immobilization of the initiator, plasma oxidation of polyester substrates was carried out using an Emitech K1050X plasma asher. Silicon slides were treated either in the plasma asher, or by using a standard RCA1 clean before initiator immobilization. ITO slides were first sonicated in acetone, followed by sonication in isopropanol, and then treated with a standard RCA-1 clean. Deposition of the initiator SAM was carried out by immersing the substrates in a solution of 3-trichlorosilylpropyl 2-bromo-2-methylpropanoate (1 mmol) and an excess of triethylamine in freshly distilled toluene at room temperature. Subsequently, the wafer was removed from solution and cleaned with a series of solvents: washed and sonicated (1 min) with toluene, and then washed with acetone followed by ethanol. The substrates were dried under a stream of nitrogen and stored under nitrogen until further use. 18.2.5

Surface-Initiated Polymerizations

Polymer brush growth of N-isopropyl acrylamide was carried out following published procedures [10]. Synthesis of poly(triphenylamine acrylate) (PTPAA) brushes was completed in a similar manner, though the polymerization was carried out at 90 KC, with DMF as the solvent.

373

374

18 Polymer Brushes: Towards Applications

18.3

Results and Discussion 18.3.1

Kinetics of Surface-Initiated ATRP of MMA from Silicon

While ATRP is typically carried out at elevated temperature, room temperature polymerization is desirable, since at room temperature thermal polymerizations are

a)

b)

(a) Surface-initiated atom transfer radical polymerization (ATRP) from Si/SiO2. No deactivator was added to the polymerization bath. (b) Surface-initiated ATRP from silicon 5 mol% deactivator (Cu(II)Br2) added to a polymerization bath consisting of unpurified Cu(I)Br. Figure 18.1

18.3 Results and Discussion

minimized. One method for increasing polymerization rates, and therefore decreasing the temperature at which the reaction is carried out, is to use aqueous solvent systems. It is thought that the high dielectric constant of the polar solvent increases the activity of the catalyst. The following study focuses on surface-initiated aqueousbased ATRP of homogeneous brushes on Si/SiO2. Specifically, it gives insight into the controlled growth of methyl methacrylate in aqueous media. The aim of this study was to grow as thick brushes as possible by extending the lifetime of the “living” polymerization. This was achieved by varying the conditions of the polymerization, specifically the concentration of the deactivator (i.e., Cu(II)) species in solution. The plot in Figure 18.1(a) represents the thicknesses of homogeneous poly(methyl methacrylate) (PMMA) brushes synthesized as a function of polymerization time. In this experiment, the Cu(I) catalyst was used as received and therefore unpurified (i.e., Cu(I) with trace amounts of Cu(II)), but no deactivator was added to the polymerization bath and termination can be seen to occur after 2 h, with thicknesses of ~50 nm. Studies by Bruening et al. [12] suggest that 30 mol% of Cu(II) complex, with respect to Cu(I), provides a more controlled/“living” polymerization process. An experiment involving 5 mol% Cu(II)Br2 with respect to the same unpurified Cu(I)Br is shown in Figure 18.1(b). This plot clearly shows that the addition of 5 mol% deactivator leads to a much longer controlled/“living” polymerization. Termination starts to play a major role after 48 h, as opposed to 2 h when no deactivator was used (see Figure 18.1(a)). As the data show, adopting this method allows the growth of PMMA brushes of up to 230 nm thickness. It should also be noted that the addition of 30 mol% Cu(II) did not improve the controlled character of this polymerization, but did slow the overall growth of polymer brushes. These experiments show that aqueous ATRP of MMA is well controlled, and generates brushes of sufficient thickness to introduce (if patterned) topographical features on surfaces. In these investigations, we are interested in modifying polyester films of PET and PEN. Micro-contact printing (lCP) [13,14] is used to immobilize an initiator SAM to a substrate to create patterned polymer brushes via ATRP. In order for surface-initiated polymerization from polymeric substrates to occur, their surfaces must first be modified by plasma oxidation in order to expose surface -OH groups which serve as ideal attachment points for the initiator. The adoption of polymer-based films as substrates presents a range of new challenges which include, most importantly, their characterization. 18.3.2

Surface-Initiated ATRP from Polymeric Substrates

Until now, very little research has been conducted into the use of polymeric substrates for surface-initiated ATRP. To date, studies have focused on brush growth from polymer microspheres [15–18] rather than macroscale polymer surfaces. The most notable exception is the investigation by Genzer and co-workers involving surface-initiated polymerization from polydimethylsiloxane (PDMS) substrates [19], and another example from PET films by S. Roux et al. [20]. This study on PET films

375

376

18 Polymer Brushes: Towards Applications

explores different methods of preparing PET surfaces, with the aim of carrying out surface-initiated free-radical polymerization of styrene. Here, we report the growth of patterned polymer brushes of the thermo-responsive poly(N-isopropyl acrylamide) (PNIPAM) from PET and PEN substrates via aqueous-based ATRP. AFM Measurements AFM images of PNIPAM brushes from initiator-modified PET and PEN are shown in Figure 18.2. These data clearly show that thick, continuous films can be grown from these substrates. The polymer bush thicknesses are much less than those grown from silicon substrates with respect to the same polymerization times. For instance, the average brush thicknesses of patterned PNIPAM grown from PET is ~100 nm, as opposed 18.3.2.1

(a) 2 - 2 lm patterned poly(N-isopropyl acrylamide) (PNIPAM) on plasma-treated PET; (b) 2 - 4 lm patterned PNIPAM on plasma-treated PET; (c) 2 - 4 lm patterned PNIPAM plasma- treated PEN.

Figure 18.2

18.3 Results and Discussion

to homogeneous brushes of silicon, which are over 200 nm thick. A possible reason for this is the roughness of the polyester films, which is significantly higher than that of the silicon substrates. Such a property would hinder conformal contact between the PDMS stamp and render the film nonuniform when carrying out lCP for depositing SAMs of the initiator. This would mean a much lower grafting density of the SAM initiator, and could also explain the rather ill-defined nature of the patterns. 18.3.3

Synthesis of Conjugated Polymer Brushes from ITO

In recent years, polymeric organic semiconductors have been used to fabricate a wide range of electronic devices such as photovoltaic cells, electroluminescent diodes, and transistors. There are substantial benefits in using organic materials, including low costs of manufacture, easy processability, and good mechanical properties, as compared with their inorganic counterparts. Continued research into these materials has led to significant improvements in the performance characteristics of these devices. Surface-initiated ATRP provides a promising route to producing thin films of aligned polymer chains tethered to a surface. As the polymerization from polymer surfaces is well-controlled (see above), the aim is to demonstrate in the next step that the growth of electro-active brushes is also feasible. Ultimately, these two processes can be coupled in the fabrication of polymer devices via a bottom-up, surfaceinitiated polymerization method. In order to fabricate an optoelectronic device containing an active polymer brush component, a suitable monomer must be selected that can be polymerized via an ATRP reaction, and which can generate a conjugated polymer with good conducting properties. Also, the polymer brush must be synthesized from a surface that has a suitable work function to act as an electrode in the completed device. In this case, it was decided to grow the polymer brush from an ITO surface, which is used as an electrode in many organic electronic devices. A triphenylamine acrylate was chosen as the monomer to form the hole-transporting PTPAA brush. This material was chosen because similar molecules have shown good hole-conducting properties [11,21], and because of its ease of synthesis. The subsequent synthesis of PTPAA brushes using the TPAA monomer is shown in Scheme 18.1. These brushes were initially synthesized from a silica surface, so that simple characterization techniques could be used in order to optimize the reaction conditions. An interesting observation was made when the concentration of CuBr in the solution was varied. As the CuBr concentration was increased, a decrease in PTPAA brush thickness was noted. An increase in thickness with decreasing copper concentration was not expected, as it has been previously shown, with a styrene monomer, that as copper concentration is increased, the brush thickness also increases [22]. In order to examine this unexpected phenomenon further, a kinetic study was carried out to show how thickness changes with time. The results of this study are shown graphically in Figure 18.3. The shapes of the kinetic plots are characteristic of termination for a polymer brush synthesized via ATRP [7]. The brushes initially grow

377

378

18 Polymer Brushes: Towards Applications

Scheme 18.1

Synthesis of poly(triphenylamine acrylate) (PTPAA) brushes.

until a value where the plots plateau, and at which point the growing polymer chains have terminated. In these plots, complete termination occurs more quickly for polymerization systems with higher Cu(I) concentrations. This early termination may explain why an inverse relationship between copper concentration and thickness is observed. A higher Cu(I) concentration will lead to a higher concentration of radicals, and an increased probability that two radicals will recombine, and terminate the growing polymer chain. While the addition of more catalyst will speed up the reaction, and lead to thicker brushes in a shorter period of time in some systems, for this particular system an increase of radicals due to higher catalyst concentrations only causes more termination of the growing chains and therefore thinner brushes. In a recent study, Bruening et al. [23] also observed an inverse relationship between catalyst concentration and thickness for the synthesis of PMA brushes via ATRP, which they also attributed to early termination of the growing polymer chains. In their study, they followed up their experimental data on the effect of cata-

Kinetic study of poly(triphenylamine acrylate) (PTPAA) brush growth, at varying CuBr concentrations.

Figure 18.3

18.3 Results and Discussion

lyst concentration on the growth of PMA brushes, with a mathematical simulation of polymer brush growth using the basic kinetic expressions of ATRP. Using this simulation, they were able to predict the inverse relationship between thickness and catalyst concentration that we have observed in the synthesis of PTPAA brushes. While the growth of polymer brushes via ATRP is well known on many surfaces, it has not yet been reported from an ITO surface. In order to synthesize PTPAA brushes from this surface, a SAM must first be generated on the surface. While the formation of a trichlorosilane SAM on silica is well known [24], it is complicated on an ITO surface due to its high surface roughness and low coverage of hydroxyl groups [25]. Nevertheless, some methods have been produced for generating a SAM [25–28]. Markovich and Mandler [25] showed that after the ITO had been refluxed in a trimethoxysilane solution for 7 days, about 90% coverage of the ITO surface was achieved. Considering the higher reactivity of the trichlorosilane used (compared with a trimethoxysilane), the reaction was carried out room temperature, to prevent polymerization, for 7 days. Triethylamine was used, as in the method of Husseman et al. [3], to drive the reaction. While it is likely that this method of SAM deposition on ITO does not provide complete surface coverage, and considering that only 10% (estimated) of surface-bound initiating molecules initiate a polymer chain [9], this level of incomplete surface coverage should not be a significant factor in polymer brush growth. Static contact angles were used to confirm the presence of the SAM on ITO. An increase in static contact angle from 26K to 64K was observed for an ITO surface after an RCA-1 clean, and after SAM deposition respectively. For comparison, an increase from 22K to 67K was observed for a silica surface after the same treatments. Following SAM deposition, polymer brushes were synthesized on ITO in the same manner as with silica. A silica substrate functionalized with an ATRP initiator trichlorosilane SAM, and an ITO substrate functionalized with the same SAM, were polymerized under identical conditions. Brush thicknesses on these two different surfaces were measured using AFM, and were observed to be identical. This result shows that PTPAA brushes grow on ITO in a similar manner as on silica. A static contact angle of 81K was measured for the PTPAA brush layer on an ATRP-initiating, SAM-modified ITO surface. This contact angle, which is consistent with that of the brush layer on a silica surface, and higher than that of the SAM-modified surfaces before brush growth, provides further evidence that the PTPAA brushes have formed on the ITO substrate. We are currently investigating devices incorporating these polymer brushes as the hole-conducting layer, and one of the first targets will be photovoltaic devices. In order to produce working solar cells, an electron-conducting and light-absorbing material (such as perylene or a semiconducting polyfluorene polymer) must be added to the PTPAA brushes, and a cathode evaporated on top of the organic films. At present, we are studying transport properties in such devices, and are optimizing the external quantum efficiency. Summary

Controlled brush growth of poly(meth)acrylates is possible on a variety of substrates, using a number of functional groups as side chains. Here, we have highlighted

379

380

18 Polymer Brushes: Towards Applications

some recent efforts to include such well-defined polymer brush films as passive or active components in devices. Our initial focus has been on (opto)electronic devices, where our aim is to grow dielectric or hole-conducting films. We have demonstrated the synthesis of polymer brushes from polymeric substrates, which could provide a route to flexible, all-polymer devices incorporating polymer brushes. We have also examined the synthesis of a conjugated polymer from an ITO surface. The high level of order of this polymer brush material may lead to future applications in optoelectronic devices. Although studies into the applications of polymer brushes in such devices are in their infancy, we are confident that because of their special nature (all polymers aligned, narrow polydispersity, blocks in layers), polymer brushes will play important roles in future polymeric devices.

References 1 O. Prucker, J. RRhe, Langmuir 1998, 14,

6893–6898. 2 M. Baum, W. J. Brittain, Macromolecules 2002, 35, 610–615. 3 M. Husseman, E. E. MalmstrSm, M. McNamara, M. Mate, D. Mecerreyes, D. G. Benoit, J. L. Hedrick, P. Mansky, E. Huang, T. P. Russell, C. J. Hawker, Macromolecules 1999, 32, 1424–1431. 4 R. Jordan, A. Ulman, M. H. Rafailovick, J. Sokolov, J. Am. Chem. Soc. 1999, 121, 1016–1022. 5 A. Juang, O. A. Scherman, R. H. Grubbs, N. S. Lewis, Langmuir 2001, 17, 1321–1323. 6 X. Huang, M. J. Wirth, Anal. Chem. 1997, 69, 4577–4580. 7 W. Huang, J.-B. Kim, M. L. Bruening, G. L. Baker, Macromolecules 2002, 35, 1175– 1179. 8 T. Wu, K. Efimenko, J. Genzer, J. Am. Chem. Soc. 2002, 124, 9394–9395. 9 D. M. Jones, A. A. Brown, W. T. S. Huck, Langmuir 2002, 18, 1265–1269. 10 D. M. Jones, J. R. Smith, W. T. S. Huck, C. Alexander, Adv. Mater. 2002, 14, 1130– 1134. 11 M. Tamada, H. Koshikawa, F. Hosoi, T. Suwa, H. Usui, A. Kosaka, H. Sato, Polymer 1999, 40, 3061–3067. 12 W. Huang, G. L. Baker, M. L. Bruening, Angew. Chem. Int. Ed. 2001, 40, 1510–1512. 13 A. Kumar, G. M. Whitesides, Appl. Phys. Lett. 1993, 63, 2002–2004. 14 M. Geissler, A. Bernard, A. Bietsch, H. Schmid, B. Michel, E. Delamarche, J. Am. Chem. Soc. 2000, 122, 6303–6304.

15 D. Bontempo, N. Tirelli, K. Feldman,

G. Masci, V. Crescenzi, J. A. Hubbell, Adv. Mater. 2002, 14, 1239–1241. 16 G. Zheng, H. D. H. StSver, Macromolecules 2002, 35, 7612–7619. 17 M. M. Guerrini, B. Charleux, J.-P. Vairon, Macromol. Rapid Comm. 2000, 21, 669–674. 18 K. N. Jayachandran, A. Takacs-Cox, D. E. Brooks, Macromolecules 2002, 35, 4247– 4257. 19 T. Wu, K. Efimenko, P. Vlcek, V. Subr, J. Genzer, Macromolecules 2003, 36, 2448– 2453. 20 S. Roux, S. Demoustier-Champagne, J. Polym. Sci. Polym. Chem. 2003, 41, 1347–1359. 21 M. Stolka, D. M. Pai, D. S. Renfer, J. F. Yanus, J. Polym. Sci. Polym. Chem. 1983, 21, 969– 983. 22 J. D. Jeyaprakash, S. Samuel, R. Dhamodharan, J. Ruhe, Macromol. Rapid Comm. 2002, 23, 277–281. 23 J. B. Kim, W. Huang, M. D. Miller, G. L. Baker, M. L. Bruening, J. Polym. Sci. Polym. Chem. 2003, 41, 386–394. 24 A. Ulman, Chem. Rev. 1996, 96, 1533–1554. 25 I. Markovich, D. Mandler, J. Electroanal. Chem. 2001, 500, 453–460. 26 Y. Koide, M. W. Such, R. Basu, G. Evmenenko, J. Cui, P. Dutta, M. C. Hersam, T. J. Marks, Langmuir 2003, 19, 86–93. 27 J. Lee, B. J. Jung, J. I. Lee, H. Y. Chu, L. M. Do, H. K. Shim, J. Mater. Chem. 2002, 12, 3494–3498. 28 C. K. Luscombe, H. W. Li, W. T. S. Huck, A. B. Holmes, Langmuir 2003, 19, 5273– 5278.

381

19

Polymerization, Nanopatterning and Characterization of SurfaceConfined, Stimulus-Responsive Polymer Brushes Marian Kaholek, Woo-Kyung Lee, Bruce LaMattina, Kenneth C. Caster, and Stefan Zauscher

Abstract

In this chapter we report the surface-initiated atom transfer radical polymerization (ATRP) of poly(N-isopropylacrylamide) (pNIPAAM), a stimulus-responsive polymer, from monolayers of x-mercapto-undecyl bromoisobutyrate on gold-coated surfaces. NIPAAM was polymerized in aqueous solution at low methanol concentrations ([MeOH] <3% by volume) at room temperature to maintain the pNIPAAM chains in a hydrophilic and extended conformational state. Under these conditions, thick polymer brush layers (up to 500 nm) are produced after 1 h of polymerization. Using dip-pen nanolithography (DPN), we fabricated for the first time patterned, stimulusresponsive pNIPAAM brushes with feature sizes on the order of 500 nm. The stimulus-responsive conformational change of bulk and patterned brushes was demonstrated by inverse transition cycling in water, and water-methanol mixtures (1:1, v:v). Diffusional processes are shown to be important for DPN fabrication of both dot and line patterns. Surface force measurements between pNIPAAM brush surfaces and a pNIPAAM decorated cantilever tip revealed large adhesion forces for the collapsed polymer brushes, suggesting strong hydrophobic interactions. The present studies are important because the triggered control of interfacial properties on the nanometer scale holds significant promise for actuation in bionanotechnology applications where polymeric actuators may manipulate the transport, separation, and detection of biomolecules.

19.1

Introduction

One central goal of materials engineering on the nanometer and micrometer length scale is to produce materials that are ordered over a range of length scales, and in which larger scale structural and physico-chemical properties are controlled by molecular characteristics [1]. For example, growing polymer brushes with thicknesses on the molecular scale from solid surfaces, allows one to tailor the surface properties of materials by imparting desirable energetic, mechanical and electrical functional-

382

19 Polymerization, Nanopatterning and Characterization

ities [2]. The in-situ formation of polymer brushes is possible through a “graftingfrom” approach [3], using surface-initiated polymerization, which achieves higher packing densities than those possible in a “grafting-to” approach [3–6]. Although the templated fabrication of polymer brushes has been prototypically demonstrated, and many methods to initiate polymerization reactions have been used (e.g., anion [7], cation [8], radical [9], plasma [10], condensation [11], photochemical [12,13], electrochemical [14], and ring-opening metathesis polymerization [15-20]), the preparation of precisely patterned, surface-attached polymeric nanostructures with controlled chain lengths, conformational geometries, functionality, and properties is still in its infancy. Here, we present a new and simple strategy to fabricate patterned surface confined, “smart” polymer brushes from N-isopropylacrylamide for applications in biosensors, proteomic chips, and nanofluidic devices. Poly(N-isopropylacrylamide) (pNIPAAM) is a stimulus-responsive polymer (SRP) with a lower critical solution temperature (LCST) of about 32 IC [21] that has been studied widely as a substrate for tissue and cell-growth surfaces [22–24], separations [25,26] and, because of the potential of its hydrogels, for drug delivery [27]. The functionality of pNIPAAM on surfaces can be separated into two categories: 1) triggered changes in polymer conformation; and 2) triggered changes in polymer surface energetics. The change in pNIPAAM conformation associated with a phase transition is central to imparting stimulus-responsive behavior to surfaces, and potentially can be exploited for force generation and control of surface energy in actuation devices on the nano- and micro-scales. Examples of these include devices for protein affinity separations [28] and the manipulation of fluid flow in nanofluidic devices [29]. In order to fabricate patterned, surface-confined, SRP brushes, we have developed novel methods that combine dip-pen nanolithography [30–32] with surface-initiated polymerizations using atom transfer radical polymerization (ATRP) [33–36]. So far, ATRP has been the workhorse polymerization methodology used by researchers attempting to prepare surface-attached polymer brushes of controlled structure, and has also been applied to NIPAAM [37,38]. This transition metal-based, controlled radical polymerization chemistry produces functional polymers with defined molecular weight and polydispersity, and as a result of the “living” nature of the initiator, allows the ready synthesis of block copolymers.

19.2

Experimental 19.2.1

Materials

N-Isopropylacrylamide (NIPAAM) monomer, copper(I) bromide (Cu(I)Br, 99.9%), methanol (MeOH, 99.9%) were obtained from Sigma-Aldrich (Milwaukee, WI, USA). NIPAAM was purified by recrystallization from toluene:n-hexane before use. Milli-QL (Millipore, Billerica, MA, USA) water (18 MX cm–1) and methanol were

19.2 Experimental

used as polymerization solvents. N,N,N¢,N¢,N†-Pentamethyldiethylenetriamine (PMDETA) was used as received from Acros Organics (Hampton, NH, USA). 19.2.2

Substrates

To immobilize the initiators for surface-initiated polymerization, gold substrates with an average grain diameter of 30 nm were prepared by thermal evaporation under a vacuum of 5.3 M 10–2 mPa. For this purpose, an adhesion layer of chromium (5.0 nm) followed by a layer of gold (50 nm) was evaporated onto silicon wafers. For interaction force measurements, the underside of atomic force microscopy (AFM) silicon-nitride (Si3N4) cantilevers (NanoprobeL; Veeco, Santa Barbara, CA, USA) were coated by the same process. Before deposition, silicon wafers and AFM cantilevers were cleaned in a mixture of H2O2:H2SO4, 1:3 (v:v) at 80 IC (“piranha solution”) for 10 min and washed thoroughly with Milli-QL grade water. (Caution: Piranha solution reacts violently with organic matter!) 19.2.3

Preparation of Initiator Monolayers

The ATRP thiol initiator, x-mercapto-undecyl bromoisobutyrate (BrC(CH3)2COO (CH2)11SH), was synthesized as reported previously [39]. A self-assembled monolayer (SAM) of the bromo-initiator was obtained by immersing clean, gold-coated Si substrates and gold-coated AFM cantilevers in a 1 mM ethanolic solution of the thiol initiator for 1 day. After incubation, the substrates and cantilevers were washed with copious amounts of ethanol, dried with nitrogen, and immediately transferred into the polymerization solution. 19.2.4

Nanopatterning of Initiator

DPN [30,32] was used to fabricate spatially confined, self-assembled monolayers (SAMs) of functionalized alkanethiols that serve as initiator platforms for ATRP. The bromo-thiol initiator was patterned directly on gold surfaces with DPN after incubating the tip (NanoProbeL; k ~ 0.12 N m–1) in a 100 lM initiator solution in hexane for 1 min. Line and dot arrays with periodic features were drawn by programming the XY-motion of the AFM tube scanner using a customized nanolithography program (NanoScriptL; Digital Instruments, Santa Barbara, CA, USA). Initiator line patterns were generated by “dynamic” DPN, repeatedly scanning back and forth on one scanline with a scan speed of 10 lm s–1 for 1 min. Dot patterns were generated by “static” DPN, holding the tip in contact with the sample substrate for a predetermined amount of time. The relative humidity during lithography was 25–35%.

383

384

19 Polymerization, Nanopatterning and Characterization

19.2.5

NIPAAM Polymerization

The synthetic pathway for the preparation of pNIPAAM brushes on gold surfaces by surface-initiated ATRP is outlined in Scheme 19.1.

Au

S

CH2

O 11

O

CH3

C

C

O Br

+ CH2

C

CH

NHCH(CH3)2

CH3 H2O-MeOH (38:1, v/v) CuBr/PMDETA RT

Au

S

CH2

O 11

O

CH3

C

C CH3

CH2

CH C

(CH3)2CHNH

Br O

n

Surface-initiated polymerization of N-isopropylacrylamide (NIPAAM) on gold surfaces using atom transfer radical polymerization (ATRP) to yield poly(N-isopropylacrylamide) (pNIPAAM) polymer brushes.

Scheme 19.1

In this process, gold substrates decorated with SAMs of initiator-functionalized thiol are placed in contact with a polymerization solution for a specified time period in a nitrogen atmosphere. Prior to use, all solutions and flasks were thoroughly flushed with nitrogen to remove oxygen. A monolayer film of the thiol initiator, BrC(CH3)2COO(CH2)11SH, was prepared by immersion of gold-coated wafers for 24 h in a 1 mM ethanolic solution of the initiator thiol. The polymerization solution is produced by combining an organometallic catalyst with a solution of NIPAAM monomer. The organometallic catalyst was formed in a nitrogen atmosphere by adding Cu(I)Br and PMDETA in a 1:5 molar ratio to 1 mL of MeOH as solvent. The mixture was then sonicated for 1–2 min to facilitate the formation of the Cu(I)Br/ PMDETA complex. Next, a 18% (wt) aqueous solution of NIPAAM was filtered into the catalyst-complex solution through a 0.45 lm Millipore Millex filter. The molar ratio of NIPAAM to Cu(I) was fixed at 4300:1 at a volume ratio of MeOH to water of 1:38 for all polymerizations. The polymerization solution was then transferred into flasks containing the gold sample substrates. The flasks were sealed with rubber septa and kept at room temperature under nitrogen. To obtain different brush heights, polymerization times were varied from 5 min to 1 h, without stirring. After the desired polymerization time, substrates were removed from the polymerization solution, rinsed extensively with Milli-Q water to remove all traces of the polymer-

19.2 Experimental

ization solution, and then dried under a flow of nitrogen. No precipitation of polymer was observed after pouring an aqueous polymerizing solution into an equal volume of MeOH at room temperature, indicating minimal polymerization in solution. In order to prepare the nanopatterned pNIPAAM brushes, the initiator-patterned surfaces were immersed for 60 min in a polymerization solution of the same composition as that used for the preparation of bulk polymer brushes. The pNIPAAM brushes on gold-coated AFM cantilevers were prepared by a similar procedure. 19.2.6

Polymer Characterization Reflectance FT-IR Spectroscopy Reflectance FT-IR spectroscopy was performed using a Thermo Nicolet Nexus 670 spectrometer with an attenuated total reflectance (ATR) accessory, fitted with a nitrogen-cooled MCT detector. For each spectrum, 128 scans with a nominal resolution of 4 cm–1 were collected. 19.2.6.1

Ellipsometry Ellipsometric measurements of the polymer brush height in water and in water/ methanol mixtures were made on a customized Rudolph Research null ellipsometer (Model 43603, 200E) at a wavelength of 401.5 nm [40]. The optical properties of the gold substrate and the thiol initiator layer were determined by measuring the ellipsometric angles in two different media (air and MQ-grade water) using four-zone null averaging [40]. The mean refractive index and the average polymer brush thickness were calculated numerically from the ellipsometric angles W and D using an optical four-layer model [41]. 19.2.6.2

Atomic Force Microscopy pNIPAAM brush substrates were rinsed with Milli-Q water, dried under a stream of nitrogen, and mounted on steel sample discs prior to AFM measurements. AFM height images were obtained in contact and TappingModeL using V-shaped silicon nitride cantilevers (NanoProbe; Digital Instruments; k ~0.12 N m–1, tip radius: 20– 60 nm) using a MultiModeL (Digital Instruments) scanning probe microscope (SPM). Topographic imaging was performed in air, in water, and in water-methanol (1:1, v:v) mixtures using a fluid cell. Imaging forces were kept at or below 1 nN to minimize compression and damage to polymer brushes. Patterned areas were located accurately and repeatedly by pixel correlation using still-video micrographs captured during lithography. To measure brush thickness, samples were carefully scored with a razor-blade tip, thus removing only the brush and Au/Cr layer. Several control experiments on bare, clean silica wafers and on gold-coated silica wafers showed that it was not possible to scratch the pure silica substrate by gentle scratching, whereas the Au/Cr layer was easily scratched to the bare silica surface. Brush height, Bh, was determined from cross-sectional analysis of AFM height images taken at the boundary between the scratched and nonscratched regions using Eq. (1), 19.2.6.3

385

386

19 Polymerization, Nanopatterning and Characterization

Bh = Sh – Gh

(1)

where Sh is the average height of the scratch (i.e., the combined thickness of the polymer brush and the Au/Cr layer), and Gh is the average height of the Au/Cr layer, measured before polymerization [42]. Surface Force Measurement Interaction forces between pNIPAAM brushes and pNIPAAM-decorated cantilever tips were measured by AFM in force spectrometry mode (MultiModeL AFM with a Nanoscope IIIa controller; Digital Instruments). Cantilever stiffness was estimated from the power spectral density of the thermal noise fluctuations [43,44]. The sensitivity of the photodiode detector was determined from the constant compliance regime upon approach at large applied normal force. The zero of separation was customarily chosen to coincide with the constant compliance regime. Approach and retraction rates were kept below 1 lm s–1 to minimize hydrodynamic drag forces, and measurements were performed in a background electrolyte concentration of 0.01 M NaCl to decrease electric double layer interactions. To obtain good statistics, at least 20 force-distance curves were recorded at each position and measurements were repeated at several locations on the sample. To investigate the reversibility of the LCST behavior, force measurement were repeated several times, switching from water to 1:1 (v:v) water-methanol mixtures to induce the phase transition. 19.2.6.4

19.3

Results and Discussion 19.3.1

Surface-Initiated Bulk Polymerization

Previous approaches to surface-initiated polymerization of pNIPAAM, using standard redox initiators, were only partially successful, producing brush heights of less than 10 nm [45]. To improve brush growth, Lopez et al. [38] used ATRP in DMF, and obtained a pNIPAAM brush with a dry thickness of about 50 nm. Although an improvement over free-radical polymerization, the full possibilities of achieving large brush heights have not been exploited. It is well known that ATRP of hydrophilic monomers can be accelerated greatly in aqueous media [46,47], and recently Huck et al. [37] demonstrated the rapid, controlled polymerization of pNIPAAM brushes with brush heights up to 100 nm, from SAM-bound initiators using ATRP in aqueous media. Under the reaction conditions chosen by Huck et al. [37], the pNIPAAM brush is in a collapsed conformational state and brush growth is likely limited. Our polymerization conditions differ from previous approaches in that they employ a low methanol concentration and a high monomer-to-catalyst ratio. By tuning the reaction conditions (vide infra) and performing the polymerization while the

19.3 Results and Discussion

growing polymer brush is in an extended conformational state, we were able to grow pNIPAAM brushes with heights up to 500 nm (wet state). A detailed analysis of the solvent conditions on pNIPAAM brush growth is currently underway. As outlined in the synthetic pathway for the preparation of pNIPAAM brushes (Scheme 19.1), initiator monolayers were prepared by immersion of gold-coated wafers in a 1 mM ethanolic solution of the bromo-initiator thiol for 24 h. Reflectance FT-IR spectra show the appearance of a carbonyl peak at 1730 cm–1 (Figure 19.1; spectrum (a)), confirming initiator immobilization on the substrate. Immersion of the substrate in an aqueous solution of NIPAAM with low MeOH content (less than 3% by volume) containing the CuBr/PMDETA catalyst (vide infra) initiates the polymerization. The major IR absorption peaks of the resulting pNIPAAM brushes (Figure 19.1; spectrum (b)) are identical to those for linear pNIPAAM polymerized in aqueous solution [48], thus verifying the presence of pNIPAAM on the surface. The absorption peak at 3300 cm–1 in Figure 19.1(b) can be attributed to the stretch of the hydrogen-bonded NH group. The anti-symmetric stretching vibration of the CH3 group occurs at 2970 cm–1, the secondary amide C=O stretching gives rise to a strong band at 1640 cm–1, and the anti-symmetric bending deformation of CH3 occurs at 1460 cm–1. The two bands at 1370 cm–1 and 1390 cm–1 of

Absorbance

1730

1640

1460 3300

3500

1370

2970

3000

2500

2000

1500

Wavenumber [cm-1] Figure 19. 1 Reflectance FT-IR spectra of: (a) bromothiol initiator monolayer on gold; and (b) poly(N-isopropylacrylamide) (pNIPAAM) brush grown by surface-initiated ATRP on gold.

387

388

19 Polymerization, Nanopatterning and Characterization

almost equal intensity are assigned to the two methyl groups in the isopropyl functionality. The polymerization of acrylamide monomers is more complicated than that of other vinyl monomers because of the possible complexation of the copper catalyst with the amide functionality of the growing polymer. This potential side reaction leads to uncontrolled polymerization [49,50], lowers conversion, and limits brush growth. Recently, Matyjaszewski et al. [51] reported an increase in polymerization rate for 4-vinylpyridine when the ligand to catalyst molar ratio was increased from 1:1 to 6:1. In order to suppress competitive coordination of pNIPAAM to copper, we made use of this acceleration strategy, using a PMDETA to CuBr molar ratio of 5:1, to maintain a relatively fast polymerization reaction and to obtain high conversion of monomer. We chose PMDETA, which is a strong complexing ligand, because the coordination complex that forms between copper and simple amines has a relatively small redox potential, which results in generation of more radicals in the system, thus accelerating polymer growth. The effective amount of initiator immobilized on the surface in surface-initiated ATRP is small, so that only a low concentration of catalyst Cu(I) is needed to maintain rapid brush growth. To determine a viable catalyst concentration, we assumed that the active thiol initiators were distributed on the surface with an area density of one initiator molecule per 0.4 nm2 [52] to yield an initiator concentration of 0.5 nmol cm–2 substrate. We chose to work with a catalyst concentration of 3.5 lmol per substrate – an amount that was enough for ATRP but which decreased the steady-state radical concentration sufficiently to minimize bimolecular termination reactions. Baker et al. [52] analyzed a series of surface-initiated ATRPs of methyl acrylate at different catalyst concentrations, and found that the thickest PMA-polymer film was formed at a molar ratio of methyl acrylate to Cu(I) of 20000:1. We chose to work with a molar ratio of NIPAAM monomer to Cu(I) of 4300:1. At this ratio, a pNIPAAM brush can be prepared with a height of 500 nm (wet state) after about 1 h reaction time (Figure 19.2). Our catalyst amount, however, was significantly smaller (approximately 50-fold less) than that reported by others [37,38]. Huck et al. [37] carried out surface-initiated ATRP of pNIPAAM in a 1:1 (v:v) mixture of water-MeOH, and the pNIPAAM brushes formed on the surface varied in height from 13 nm to over 100 nm in the dry state, corresponding to 5 min and 100 min reaction times, respectively. Under these conditions, polymer was likely formed in a more hydrophobic (collapsed) state where termination reactions between growing polymer chains limited brush growth. It has been shown that the physical properties of pNIPAAM polymerized above the LCST differ significantly from those of polymer polymerized below the LCST [53]. To optimize brush growth by performing the polymerization in the conformationally extended state, we polymerized pNIPAAM at low MeOH concentrations (<3% by volume) that do not affect the LCST behavior significantly [54]. A small amount of MeOH, however, is needed to achieve good solubility of the initial Cu(I)/PMDETA complex. We measured the thickness of the pNIPAAM brushes by cross-sectional analysis from AFM-height images (Figure 19.2). These measurements revealed that the

19.3 Results and Discussion

water-swollen polymer brushes are smooth and homogeneous; for example, a 500 nm-high brush had a RMS roughness of <5 nm over a 400 lm2 area. Although the measured brush thickness in aqueous media depends on the applied scanning force below the LCST, we found that it is largely unaffected by the applied scanning force above the LCST, suggesting a considerable stiffening of the brush in its collapsed state [55]. To insure consistent height measurements, brush height was measured as a function of polymerization time in the dry (fully collapsed) state.

Scratched region

pNIPAAM brush

Sh

Poly(N-isopropylacrylamide) (pNIPAAM) brush grown by surface-initiated atom transfer radical polymerization (ATRP) on gold after 60 min reaction time. (a) Atomic force microscopy (AFM) height image (contact mode in MQ-grade water at 25 .C); (b) corresponding average cross-section. The average

Figure 19.2

brush height is 505 nm and the RMS roughness over a 400 lm2 area is 4.3 nm. The brush height was obtained by averaged cross-sectional analysis of the AFM height image taken on a region of the sample where part of the brush and the underlying gold and chromium layer (of known thickness) have been removed.

389

19 Polymerization, Nanopatterning and Characterization

AFM polymer brush thickness [nm]

390

250

200

Regime II

Regime I

150

100

50

0 0

10

20

30

40

50

60

Reaction time [min]

The average pNIPAAM dry brush height plotted as a function of polymerization time, using a molar ratio of NIPAAM to Cu(I) of 4300:1 and a volume ratio of MeOH to

Figure 19.3

water of 1:38. The error bars represent the RMS roughness of pNIPAAM brushes over a 3600 lm2 area. Each data point represents a measurement on a different substrate.

The data in Figure 19.3 show that brush growth occurs in two regimes. In regime I (0 < t < 30 min), after a short induction time, growth is rapid and almost linear, while in regime II (t > 30 min) the growth rate declines sharply. The growth behavior in regime I is consistent with controlled chain growth from the surface, and suggests that we can obtain a brush with predictable thickness up to 250 nm in the dry state. The loss of growth rate in regime II possibly results from chains that become buried within the film and thus become inaccessible to monomers, which ultimately limits surface-initiated polymerization. Another possibility to explain the stagnant growth is the inactivation of the catalyst by forming a competitive complex with growing polyacrylamides, similar to that previously reported for the polymerization of (meth)acrylamides using linear amines [49,50]. We noticed that the achievable brush height for a given polymerization time depends strongly on the roughness and quality of the gold substrate, likely because the effective initiator density is affected by surface roughness. 19.3.2

Phase Behavior and Mechanical Characterization

The interesting phase behavior of pNIPAAM in solution reflects the balance of like and unlike interactions among its own segments and the surrounding solvent molecules. The inverse solubility upon heating of pNIPAAM in solution likely arises from changes in the number or strength of hydrogen bonds that develop between the solvent and polar groups on the polymer, as water molecules must reorient around nonpolar regions on the polymer backbone, having no opportunity to hydrogen bond [56]. This hydrophobic effect causes a negative entropy of mixing. The enthalpy of mixing, however, is positive and relatively large as it arises from the for-

19.3 Results and Discussion

mation of hydrogen bonds between solvent and polymer. The requirement for phase separation is a positive Gibbs free energy of mixing that results when a negative excess entropy of mixing is combined with the positive enthalpy of mixing. The required changes in the excess of the entropy are likely coupled to changes in the local hydrogen bonding state of pNIPAAM in water where solubility is maintained as long as the state of hydrogen bonding (below the LCST) is not affected significantly. In addition to temperature, cosolvents can cause an inverse phase transition in pNIPAAM. For example, addition of methanol to aqueous pNIPAAM solutions in the range from 10% to 65% (by volume) leads to co-nonsolvency [21,54,57], effectively shifting the LCST of pNIPAAM in solution to lower temperatures. Increasing the methanol concentration above 65% causes a dramatic increase in the LCST to values greater than that in pure water. This reentrant phase behavior can be explained as follows: addition of methanol leads to a decrease in hydrogen bond enthalpy, indicating that the cosolvent lowers the number or strength of polymer water contacts, and solvent-solvent interactions dominate at the middle of the phase diagram, promoting larger numbers of polymer-polymer contacts. Associated with the solubility change at the LCST is also a conformational change. Below the LCST, pNIPAAM is hydrated and the chains are in an expanded conformation. Above the LCST, pNIPAAM is in a hydrophobically collapsed conformational state.

∆h = 58%

300

Cycle 2

∆h = 56%

400

∆h = 56%

Cycle 1

500

∆h = 61%

Polymer Brush Height [ nm ] (AFM Measurements)

600

200 100 0 Air Water (Dry State)

Water

The effect of solvent conditions on pNIPAAM brush height measured by AFM. (a) Brush in a dry, fully collapsed state (Bh = 182 nm) and after immersion into MQgrade water in a swollen, fully extended conformational state (Bh = 470 nm). (b) Reversible, inverse transition cycling of the brush by repeated exposure to a 1:1 (v:v) MeOH/water Figure 19.4

Water

MeOH/Water (1:1, v/v)

MeOH/Water (1:1, v/v)

mixture. Cycle 1: Brush in a fully extended conformational state in MQ-grade water (Bh = 470 nm) and after immersion into a 1:1 (v:v) MeOH/water mixture in a collapsed conformational state (Bh = 205 nm). Cycle 2: Brush in MQ-grade water (Bh = 467 nm) and after immersion into a 1:1 (v:v) MeOH/water mixture state (Bh = 194 nm).

391

19 Polymerization, Nanopatterning and Characterization

The reversible conformational mechanics of two different pNIPAAM brushes were studied using AFM (Figure 19.4) and ellipsometry (Figure 19.5), in solventswelling experiments. The data in Figures 19.4(a) and 19.5(a) show that a pNIPAAM brush in the dry, fully collapsed state swells significantly when exposed to water at room temperature. The average brush height change is 61% and 73% for the samples in Figures 19.4(a) and 19.5(a), respectively. As discussed above, the brushes are also responsive to the solvent composition. In pure water, the pNIPAAM brushes are in a good solvent at temperatures below the LCST, and the brush is likely in a fully extended conformational state (Figures 19.4(b) and 19.5(b)). After exposure to a water-MeOH (1:1, v:v) mixture (a poor solvent), the brush adopts a collapsed conformation. The corresponding brush height change is 56% and 30% for the samples in Figures 19.4(b) and 19.5(b), respectively. We demonstrate the reversibility of these conformational changes by inverse transition cycling (Figures 19.4(b) and 19.5(b)). The ellipsometric brush height measurements (Figure 19.5(b)) revealed that the conformational collapse increases from 30% to 43% in the second cycle, suggesting that the fully collapsed state is not reached immediately. Interestingly, the brush height in the expanded conformational state of the brush was unaffected by repeated cycles. The AFM brush height measurements did not reveal such a dependence on cycle number, most likely because, in contrast to ellipsometry, the AFM measurements involve 180

80

∆h = 43%

100

∆h = 73%

120

Cycle 2

∆h = 29%

140

∆h = 30%

Cycle 1

∆h = 43%

160 Polymer Brush Height [ nm ] (Ellipsometric Measurements)

392

60 40 20 0 Air Water Water Water Water (Dry State) MeOH/Water MeOH/Water (1:1, v/v) (1:1, v/v)

The effect of solvent conditions on pNIPAAM brush height measured by ellipsometry. (a) Brush in a dry, fully collapsed state (Bh = 39 nm) and after immersion into MQgrade water in a swollen, fully extended conformational state (Bh = 145 nm). (b) Reversible, inverse transition cycling of the brush by repeated exposure to a 1:1 (v:v) MeOH/water mixture. Cycle 1: Brush in a fully extended Figure 19.5

conformational state in MQ-grade water (Bh = 145 nm) and after immersion into a 1:1 (v:v) MeOH/water mixture in a collapsed conformational state (Bh = 101 nm). Cycle 2: Brush in MQ-grade water (Bh = 143 nm), after immersion into a 1:1 (v:v) MeOH/water mixture state (Bh = 81 nm), and after exposure to MQ-grade water (Bh = 143 nm).

19.3 Results and Discussion

a direct, mechanical compression of the brush during imaging, rendering the measurements relatively insensitive to subtle conformational changes in the brush. While SRP networks can shrink in all three dimensions [58], shrinkage in surfaceconfined polymer brushes is much less because the mobility of polymer chains is restricted largely to one dimension perpendicular to the substrate [59,60]. 19.3.3

Surface Force Measurements

In order to elucidate further the conformational and surface energetic changes of pNIPAAM brushes as a function of solvent conditions, we performed AFM surface force measurements between pNIPAAM brush surfaces and pNIPAAM-decorated cantilever tips in water (0.01 M NaCl) and a MeOH/water (1:1 v:v) mixture. This is shown schematically in the insets in Figure 19.6. 20 Approach Retraction

Force [ nN ]

15 10 5 0 -5 0

200

400

600

800

1000

800

1000

Separation [ nm ] 15 Approach Retraction

Force [ nN ]

10 5 0 -5 -10 -15 0

200

400

600

Separation [ nm ] Typical force-separation profiles for a pNIPAAM brush and a pNIPAAM decorated cantilever tip interacting in (a) 0.01 M NaCl and in (b) a 1:1 (v:v) MeOH/0.01 M NaCl mixture. The arrows indicate the onset

Figure 19.6

of repulsive force interactions for each case. The insets illustrate schematically the interaction between a pNIPAAM brush and a pNIPAAM decorated AFM cantilever tip.

393

394

19 Polymerization, Nanopatterning and Characterization

Interactions between two polymer-decorated surfaces can be either attractive or repulsive, and depends on the dynamics of the measurement [61–63]. Although AFM offers a convenient way to study these surface interactions, it must be recognized that force measurements are performed on time scales that are likely too short for thermodynamic equilibrium of the pNIPAAM chains to be reached. In the present case, the grafting density of pNIPAAM was high, likely leading to an extended polymer conformation. When two such surfaces are brought into increasingly compressive contact, repulsive, steric forces arise from the restriction of conformational degrees of freedom of the thermally mobile polymer chains. In order to compare force-distance data, the constant compliance regime was chosen to coincide with the zero of separation. In the present case, where tip and surface are covered with pNIPAAM, this results in an offset error in surface separation on the order of the thickness of the compressed polymer layers. The repulsive force contribution due to electric double layer overlap is expected to be small and of short range (k–1 ~ 3 nm), as experiments were performed in a background of 0.01 M NaCl. Figure 19.6(a) shows that the interaction upon approach and retraction was monotonically repulsive below the LCST (0.01 M NaCl), and that the onset of steric interactions occurred at about 200 nm, in agreement with the estimated combined brush thickness of the pNIPAAM layer on substrate and cantilever (vide infra). Assuming a tip radius of 100 nm, we estimate the interaction energy per unit area to be on the order of 0.1 N m–2. This corresponds to a pressure on the brush that is significantly greater than that reported for brush compressibility studies using AFM [42] or the surface forces apparatus [62,63], and we assume that at these pressures the brush is strongly compressed. Figure 19.6(b) shows that the force upon approach in the water/MeOH mixture goes through an attractive minimum, followed by a significantly reduced steric repulsion regime when compared to the interaction in water alone and that, upon retraction, a large unspecific adhesion force occurs, indicating a hydrophobically collapsed surface state of the pNIPAAM brush. The minimum on approach likely arises from attractive polymer segment interactions, the number of which increases with increasing compression of the brush surfaces until, with further compression, the restriction in conformational degrees of freedom finally dominate and give rise to strong, steric-repulsive forces. The significant change in polymer conformation associated with the collapse of the polymer brush during a phase transition can be inferred from the decrease in decay length, k–1, from about 38 nm to about 3 nm, obtained by fitting an inverse exponential function (Eq. (2)) F(D) e–kD

(2)

to the data, where F is the force and D is the separation distance (Figure 19.7). The repulsive pressure, P(D), that develops in a good solvent between two brushbearing, parallel surfaces can be further characterized by the scaling approach developed by Alexander and de Gennes [64,65] (Eq. (3)),

19.3 Results and Discussion

100 0.01M NaCl Decay Length: κ ~ 38 nm -1

Force [ nN ]

10

Alexander-DeGennes Scaling Approach Average Brush Thickness: L ~ 120 nm

1 Decay Length: -1 κ ~ 3 nm MeOH/0.01M NaCl (1:1, v/v)

0.1

0.01 0

50

100

150

200

250

300

Separation [ nm ] Repulsive force in 0.01 M NaCl and in a 1:1 (v:v) MeOH/0.01 M NaCl mixture plotted as a function of separation and fitted to a decaying exponential function. The significant change in polymer conformation associated with the collapse of the polymer brush

Figure 19.7

during a phase transition can be inferred from the decrease in decay length, k–1, from about 38 nm to about 3 nm. An average brush thickness of 120 nm in 0.01 M NaCl is predicted from the scaling approach developed by Alexander and de Gennes.

P(D) [(2L/D)9/4 – (D/2L)3/4] for D < 2L

(3)

where L is the average brush height. Integration of Eq. (3) from L to D yields the interaction free energy per unit area, W(D), which is proportional to the measurable force between a flat plate and a sphere with radius R, according to the Derjaguin approximation (Eq. (4)) [64], RD F(D) = 2pRW (D) 2pR P(D)dD

(4)

L

For D/2L in the range from 0.2 to 0.9, Eq. (3) can be approximated by a decaying exponential function [64] (Eq. (5)), P(D) e–pD/L

(5)

that yields a relationship between F and D of the form (Eq. (6)), F(D) L/p e–pD/L

(6)

An estimate of the average brush thickness L can be obtained by fitting the measured force-separation data to Eq. (6) using an arbitrary pre-exponential factor. An average brush thickness of 120 nm was predicted from such a fit (Figure 19.7). Using this value and the known brush thickness on the substrate of 60 nm (hydrated brush thickness), we estimated the unknown brush thickness on the cantilever tip to be about 180 nm, which was in reasonable agreement with the height expected for a polymerization time of 60 min.

395

19 Polymerization, Nanopatterning and Characterization

We also measured the reversible change in surface energy associated with a phase transition by measuring pull-off forces (i.e., the maximum force required to liberate the cantilever from surface contact). This is a good measure of adhesion, and thus surface energy, as it does not contain contributions from elastic surface deformation. There was no adhesion between the polymer brushes in water (0.01 M NaCl) alone (Figure 19.6(a)). In the MeOH/water mixture, the average adhesion force between pNIPAAM brushes was on the order of 12 nN (Figure 19.6(b)). The adhesion force distributions obtained from many (n = 150) approach-retraction cycles in water and in MeOH/water are shown in Figure 19.8. This adhesion likely arises from van der Waals forces that can effectively act between the hydrophobically collapsed polymer segments on substrate and tip [66]. The reversibility of these adhesion forces was demonstrated for three solvent exchange cycles, where essentially the same adhesion force distributions were obtained in each of the three cycles, suggesting that the effect of a phase transition on adhesion is entirely reversible (data not shown).

100 Occurrence [ % ]

396

MeOH/0.01M NaCl (1:1, v/v) 0.01M NaCl

80 60 40 20 0 0

10

12

14

16

Pull-Off Force [ nN ] Adhesion force distributions for the interaction of a pNIPAAM brush and a pNIPAAM-decorated cantilever tip in 0.01 M NaCl and in a 1:1 (v:v) MeOH/0.01 M NaCl mixture. The data in each case reflect 150 approach-retraction cycles on different locations of the sample. Figure 19.8

19.3.4

Nano-Patterning

Figure 19.9 shows an AFM height image (contact mode) and the corresponding, averaged height profile for a dot-array nanopattern of pNIPAAm brushes after 1 h of polymerization. The corresponding initiator thiol dot pattern was obtained with 5 s contact time per dot. DPN patterned, chemically functionalized thiols can often be imaged by AFM in lateral force mode, if the chemical functionality of the thiol is sufficiently different from that of the background [67,68]. We were unable to obtain reasonably

19.3 Results and Discussion

µm

12.5

0 µm

0

12.5

610 nm

nm

18 nm

µm

Contact mode AFM height image in air and the corresponding, averaged height profile of a pNIPAAM brush dot array prepared on gold-coated silicon substrates (polymerization time 60 min). The tip-sample contact time during “static” DPN was 5 s per dot.

Figure 19.9

well-defined images of the likely somewhat hydrophobic initiator thiol patterns. The apparent brush heights of the patterned brushes were significantly smaller, at equal reaction times, when compared with bulk pNIPAAM brushes. Our height measurements revealed that a bulk pNIPAAM brush has formed in the vicinity of the pattern, likely because of fast diffusion of the initiator thiol (vide infra). Figure 19.10 shows AFM height images (TappingModeL AFM) and the corresponding, averaged height profiles in (a) air, (b) in water, and (c) in a mixture of water-MeOH (1:1, v:v) for dynamically nanopatterned pNIPAAM brushes after 1 h of polymerization at room temperature. The rectangles depicted in the AFM height images show the area selected for averaged cross-sectional analysis. The averaged heights of selected lines are 44 nm, 54 nm, and 86 nm in air, water, and water-MeOH, respectively. The averaged widths of the pNIPAAM line patterns are 470 nm, 490 nm, and 390 nm in air, water, and

397

19 Polymerization, Nanopatterning and Characterization

44 nm 470 nm

µm

nm

30

µm

0 0

µm

30

54 nm 490 nm

µm

nm

25

µm

0 0

µm

25

30 nm

86 nm 390 nm

µm

398

µm

0 0

µm

30

TappingModeA AFM height images and corresponding averaged height profiles of a pNIPAAM brush line-pattern prepared on a gold-coated silicon substrate (polymerization time 60 min) in (a) air, (b) MQFigure 19.10

grade water, and (c) a mixture of MeOH/water (1:1, v:v), all at room temperature. The writing speed for each line was 10 lm s–1. The rectangle in each image indicates the area over which the cross-sectional data were averaged.

Summary

water-MeOH, respectively, indicating that the width change is significantly smaller compared with the height change of these features. The lengths (8.0–8.5 lm) of the pNIPAAM lines remained nearly constant in air, water, and the water/MeOH mixture. Our findings are consistent with the behavior of laterally confined and terminally attached polymer chains, where the average chain height can be affected significantly more by application of external stimuli than the lateral dimensions of the patterned features because the mobility of polymer chains is restricted largely to one dimension perpendicular to the substrate [59,60]. The height of the dry nanopatterned brushes was increased by about 23% after exposure to water, and surprisingly also by about 59% (from 54 nm to 86 nm) after addition of 50% (by volume) MeOH as a cosolvent to pure water. The latter observation can be explained, again, by the formation of bulk pNIPAAM brushes in the direct vicinity of the patterned areas. AFM height measurements (in air) revealed that this bulk polymer brush layer has a thickness of around 65 nm. It is likely that the bulk polymer brush responds to the external stimulus more than the nanopatterned features, so that after addition of 50% (by volume) of MeOH to pure water the bulk pNIPAAM brush collapsed more than the nanopatterned lines, resulting in a relative height increase of the patterned features. The formation of a bulk polymer brush in the direct vicinity of pNIPAAM patterns can be explained as follows. The ATRP initiator was applied to the gold surface by DPN, and with this technique the initiator is likely not only chemisorbed on the gold surface but also physisorbed, leaving diffusionally mobile initiator molecules on the surface. Although the gold substrates were washed with copious amounts of ethanol immediately after patterning and then stored in anhydrous ethanol for 1 h before transfer into the polymerization solution, the physisorbed initiator may not have been removed quickly enough, facilitating the diffusion of the physisorbed initiator away from the patterned area onto the surrounding gold which led to a sufficiently dense layer of ATRP initiator to initiate polymerization.

Summary

We have demonstrated the surface-initiated ATRP of stimulus-responsive poly(N-isopropylacrylamide) brushes from monolayers of x-mercaptoundecyl bromoisobutyrate on gold-coated surfaces. Our polymerization conditions differed from previous approaches in that they employed a small methanol concentration and a high monomer-to-catalyst ratio. By adopting low methanol concentrations ([MeOH] <3% by volume) during polymerization, the pNIPAAM chains were maintained in a hydrophilic and extended conformational state, yielding thick polymer brush layers. We fabricated, for the first time, patterned, stimulus-responsive pNIPAAM brushes with feature sizes on the order of 500 nm using DPN. While DPN is a simple method that enables the patterning of functionalized alkane thiols, this method, when used with the initiator thiols of this research, is troubled by problems arising from the lack of diffusional control. This causes unwanted polymer growth in the

399

400

19 Polymerization, Nanopatterning and Characterization

vicinity of patterned features. Recent studies in our laboratory, using alternate nanolithographic methods, have yielded well-defined, stimulus-responsive pNIPAAM brushes with high aspect ratios [69]. We were able to demonstrate the stimulus-responsive conformational change of bulk and patterned brushes by inverse transition cycling in water, and water-methanol mixtures (1:1, v:v). Surface force measurements between pNIPAAM brush surfaces and a pNIPAAM-decorated cantilever tip revealed large adhesion forces for the collapsed polymer brushes, suggesting strong hydrophobic interactions between polymer segments. The fabrication and characterization of SRP brushes and nanostructures is important, because the triggered control of interfacial properties on the nanometer scale holds significant promise for actuation in bionanotechnology applications where polymeric actuators may manipulate the transport, separation, and detection of biomolecules.

Acknowledgments

The authors thank the National Science Foundation for support through grants NSF EEC-021059, NSF DMR-0239769 CAREER AWARD, and ARO DAADG55-98-D0002. They also thank Hongwei Ma (Department of Biomedical Engineering, Duke University) for synthesis of the bromo-initiator, and Dr. Tommy Nylander and Yulia Samoshina (Department of Physical Chemistry I, Lund University, Sweden) for their generous help with ellipsometric measurements.

References

References 1 N. Dan, Trends Biotechnol. 2000, 18, 370–374. 2 B. Zhao, W. J. Brittain, Prog. Polym. Sci. 3 4

5 6 7

8

9 10

11

12

13 14 15 16

17

18 19 20 21

2000, 25, 677–710. O. Prucker, J. RWhe, Macromolecules 1998, 31, 592–601. K. Fujiki, M. Sakamoto, A. Yoshida, H. Maruyama, J. Polym. Sci. Polym. Chem. 1999, 37, 2121–2128. O. Prucker, J. RWhe, Macromolecules 1998, 31, 602–613. O. Prucker, J. RWhe, Langmuir 1998, 14, 6893–6898. R. Jordan, A. Ulman, J. F. Kang, M. H. Rafailovich, J. Sokolov, J. Am. Chem. Soc. 1999, 121, 1016–1022. M. D. K. Ingall, C. H. Honeyman, J. V. Mercure, P. A. Bianconi, R. R. Kunz, J. Am. Chem. Soc. 1999, 121, 3607–3613. O. Prucker, J. Ruhe, Macromolecules 1998, 31, 592–601. W. Chen, A. Y. Fadeev, M. C. Hsieh, D. Oner, J. Youngblood, T. J. McCarthy, Langmuir 1999, 15, 3395–3399. M. Husemann, D. Mecerreyes, C. J. Hawker, J. L. Hedrick, R. Shah, N. L. Abbott, Angew. Chem. Int. Ed. 1999, 38, 647–649. O. Prucker, C. A. Naumann, J. RWhe, W. Knoll, C. W. Frank, J. Am. Chem. Soc. 1999, 121, 8766–8770. H. J. Lee, Y. Nakayama, T. Matsuda, Macromolecules 1999, 32, 6989–6995. C. Gurtner, A. W. Wun, M. Sailor, Angew. Chem. Int. Ed. 1999, 38, 1966–1968. M. Bartz, J. Kuther, R. Seshadri, W. Tremel, Angew. Chem. Int. Ed. 1998, 37, 2466–2468. M. Weck, J. J. Jackiw, R. R. Rossi, P. S. Weiss, R. H. Grubbs, J. Am. Chem. Soc. 1999, 121, 4088–4089. N. Y. Kim, N. L. Jeon, I. S. Choi, S. Takami, Y. Harada, K. R. Finnie, G. S. Girolami, R. G. Nuzzo, G. M. Whitesides, P. E. Laibinis, Macromolecules 2000, 33, 2793–2795. M. R. Buchmeiser, N. Atzl, G. K. Bonn, J. Am. Chem. Soc. 1997, 119, 9166–9174. F. Sinner, M. R. Buchmeiser, Macromolecules 2000, 33, 5777–5786. F. M. Sinner, M. R. Buchmeiser, Angew. Chem. Int. Ed. 2000, 39, 1433–1436. H. G. Schild, M. Muthukumar, D. A. Tirrell, Macromolecules 1991, 24, 948–952.

22 C. S. Chen, M. Mrksich, S. Huang,

G. M. Whitesides, D. E. Ingber, Science 1997, 276, 1425. 23 O. H. Kwon, A. Kikuchi, M. Yamato, T. Okano, Biomaterials 2003, 24, 1223–1232. 24 D. Cunliffe, C. D. Alarcon, V. Peters, J. R. Smith, C. Alexander, Langmuir 2003, 19, 2888–2899. 25 X. L. Wang, J. Huang, X. Z. Chen, X. H. Yu, Desalination 2002, 146, 337–343. 26 H. Kanazawa, T. Sunamoto, E. Ayano, Y. Matsushima, A. Kikuchi, T. Okano, Anal. Sci. 2002, 18, 45–48. 27 R. Duncan, J. Kopecek, Adv. Polym. Sci. 1984, 57, 53. 28 J. Hyun, W.-K. Lee, N. Nath, A. Chilkoti, S. Zauscher, JACS 2004, in press. 29 D. J. Beebe, J. S. Moore, Q. Yu, R. H. Liu, M. L. Kraft, B. H. Jo, C. Devadoss, Proc. Natl. Acad. Sci. USA 2000, 97, 13488–13493. 30 R. D. Piner, J. Zhu, F. Xu, S. H. Hong, C. A. Mirkin, Science 1999, 283, 661–663. 31 S. H. Hang, C. A. Mirkin, Science 2000, 288, 1808–1811. 32 S. H. Hong, J. Zhu, C. A. Mirkin, Langmuir 1999, 15, 7897–7900. 33 K. Matyjaszewski, J. H. Xia, Chem. Rev. 2001, 101, 2921–2990. 34 T. E. Patten, K. Matyjaszewski, Acc. Chem. Res. 1999, 32, 895–903. 35 D. J. Kim, J. Y. Heo, K. S. Kim, I. S. Choi, Macromol. Rapid Commun. 2003, 24, 517– 521. 36 J. N. Kizhakkedathu, D. E. Brooks, Macromolecules 2003, 36, 591–598. 37 D. M. Jones, J. R. Smith, W. T. S. Huck, C. Alexander, Adv. Mater. 2002, 14, 1130– 1134. 38 S. Balamurugan, S. Mendez, S. S. Balamurugan, M. J. O’Brien, G. P. Lopez, Langmuir 2003, 19, 2545–2549. 39 D. M. Jones, A. A. Brown, W. T. S. Huck, Langmuir 2002, 18, 1265–1269. 40 M. Landgren, B. JYnsson, J. Phys. Chem. 1993, 97, 1656–1660. 41 F. Tiberg, M. Landgren, Langmuir 1993, 9, 927–932. 42 S. Yamamoto, M. Ejaz, Y. Tsujii, M. Matsumoto, T. Fukuda, Macromolecules 2000, 33, 5602–5607.

401

402

19 Polymerization, Nanopatterning and Characterization 43 J. P. Cleveland, S. Manne, D. Bocek,

57 F. M. Winnik, M. F. Ottaviani, S. H. Boss-

P. K. Hansma, Rev. Sci. Instrum. 1993, 64, 403–405. 44 J. L. Hutter, J. Bechhoefer, Rev. Sci. Instrum. 1993, 64, 1868–1873. 45 L. K. Ista, S. Mendez, V. H. Perez-Luna, G. P. Lopez, Langmuir 2001, 17, 2552–2555. 46 X. S. Wang, S. F. Lascelles, R. A. Jackson, S. P. Armes, Chem Commun 1999, 1817– 1818. 47 X. S. Wang, S. P. Armes, Macromolecules 2000, 33, 6640–6647. 48 Y. V. Pan, R. A. Wesley, R. Luginbuhl, D. D. Denton, B. D. Ratner, Biomacromolecules 2001, 2, 32–36. 49 M. Teodorescu, K. Matyjaszewski, Macromolecules 1999, 32, 4826–4831. 50 J. T. Rademacher, R. Baum, M. E. Pallack, W. J. Brittain, W. J. Simonsick, Macromolecules 2000, 33, 284–288. 51 J. H. Xia, X. Zhang, K. Matyjaszewski, Macromolecules 1999, 32, 3531–3533. 52 J. B. Kim, W. X. Huang, M. D. Miller, G. L. Baker, M. L. Bruening, J. Polym. Sci. Polym. Chem. 2003, 41, 386–394. 53 E. S. Matsuo, M. Orkisz, S. T. Sun, Y. Li, T. Tanaka, Macromolecules 1994, 27, 6791– 6796. 54 F. M. Winnik, H. Ringsdorf, J. Venzmer, Macromolecules 1990, 23, 2415–2416. 55 S. Kidoaki, S. Ohya, Y. Nakayama, T. Matsuda, Langmuir 2001, 17, 2402–2407. 56 H. G. Schild, Prog. Polym. Sci. 1992, 17, 163– 249.

mann, M. Garciagaribay, N. J. Turro, Macromolecules 1992, 25, 6007–6017. 58 M. Shibayama, T. Tanaka, Adv. Polym. Sci. 1993, 109, 1–62. 59 N. Dingenouts, C. Norhausen, M. Ballauff, Macromolecules 1998, 31, 8912–8917. 60 S. Seelenmeyer, I. Deike, S. Rosenfeldt, C. Norhausen, N. Dingenouts, M. Ballauff, T. Narayanan, P. Lindner, J. Chem. Phys. 2001, 114, 10471–10478. 61 G. J. Fleer, M. A. Cohen Stuart, J. M. H. M. Scheutjens, T. Cosgrove, B. Vincent, Polymers at Interfaces, Chapman & Hall: London, 1993. 62 J. Klein, P. F. Luckham, Macromolecules 1984, 17, 1041–1048. 63 P. F. Luckham, J. Klein, Macromolecules 1985, 18, 721–728. 64 J. N. Israelachvili, Intermolecular and surface forces, 2nd ed., Academic Press: London, San Diego, 1991. 65 P. G. DeGennes, Adv. Colloid Interface Sci. 1987, 27, 189–209. 66 F.-J. Schmitt, C. Park, J. Simon, H. Ringsdorf, J. Israelachvili, Langmuir 1998, 14, 2838–2845. 67 J. Y. Jang, S. H. Hong, G. C. Schatz, M. A. Ratner, J. Chem. Phys. 2001, 115, 2721–2729. 68 P. E. Sheehan, L. J. Whitman, Phys. Rev. Lett. 2002, 88, 156104-1–156104-4. 69 M. A. Kaholek, S.-J., Lee, W.-K., Caster, K. C., Zauscher, S,. Nano Lett. 2004, 4, 373–376.

403

20

Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale Sergiy Minko, Marcus Mller, Valeriy Luchnikov, Mikhail Motornov, Denys Usov, Leonid Ionov, and Manfred Stamm

Abstract

In this chapter, we review our recent results on the synthesis and study of a new kind of polymer brush which comprises two immiscible polymers covalently grafted to a solid substrate. The mixed brushes represent a fascinating world of responsive materials that demonstrate switching behavior upon change of their environmental medium. They show rich phase behavior when segregated nanoscopic domains allow for reversible chemical patterning of surfaces. The behavior of mixed brushes open new possibilities for the design of materials with reversibly switching wettability, adhesion, adsorption and other properties related to the surface composition and film morphology.

20.1

Introduction

Tailoring materials with smart responses to external fields is a major goal of modern material science. Devices such as sensors, switches or microactuators rely essentially on the response of their physical properties (surface composition and energy, optical and electrical properties, etc.) to external stimuli. This response may be employed to tune stability, adhesion, wettability, and to regulate interactions with cells and proteins in biomaterials, or membrane permeability. The high sensitivity of polymeric and biopolymeric systems to external fields makes them perfect candidates for responsive materials. For example, polymers respond by large conformational changes even to very small external fields in order to minimize free energy. Typical examples are the collapse of a chain in a poor solvent or the coagulation of proteins. Mixed polymer brushes represent a polymeric system with remarkable responsive properties. In contrast to brushes, which consist of one type of homopolymer, mixed polymer brushes can amplify the response due to a combination of conformational changes and microphase separation. If a mixed brush of hydrophilic and hydrophobic homopolymers is exposed to a hydrophilic solvent, the hydrophilic component preferentially segregates to the top of the film, and the surface becomes hydrophilic.

404

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

Schematic illustration of two possible morphologies of a mixed brush irreversibly grafted to solid substrates (crosssection of the layer). (a) “Ripple” morphology in a nonselective solvent; (b) “dimple” morphology in a solvent that is a poor one for the black chains. (Reprinted with permission from Ref. [3b].)

Figure 20.1

However, exposing the same brush to a hydrophobic solvent reversibly switches the surface from a hydrophilic to a hydrophobic state. This adaptive behavior is very promising for the engineering of smart surfaces for biomedical applications and micro- and/or nanodevices. Another particular property of mixed brushes which distinguishes them from one-component brushes is their lateral morphology. The phase behavior of a binary brush is determined by a competition of the mixing entropy, which favors a homogeneously mixed state, and the interaction energy, which is reduced by spatial separation of the incompatible polymers. By anchoring the polymer chains, we prevent macroscopic segregation of the incompatible species. In order to reduce the number of energetically unfavorable interactions between distinct species, the molecules self-assemble into complex morphologies [1–3]; the different morphologies are shown schematically in Figure 20.1. The characteristic length of the lateral morphology is of the order of the radius of gyration of the chains in h-solvent – that is, several nanometers. This opens the possibility of using mixed brushes for chemical patterning at the nanoscale. During the past decade, mixed polymer brushes have become the subject of intensive theoretical and experimental research. In this chapter, we report our recent results on the theory and experimental investigations of mixed brushes.

20.2

Theory of Mixed Polymer Brushes

The first study of mixed polymer brushes by Marko and Witten [1] employed the strong stretching approximation (SST). The basic assumption of this is that fluctuations of the chain conformations around the conformation that minimizes the free energy can be neglected. This approximation has been introduced by Semenov [4], who studied the structure of micelles in the melts of the block-copolymers. The approximation becomes better if the chain extension perpendicular to the grafting surface – that is, the brush height, h – greatly exceeds the unperturbed chain extension, Re. This condition can be fulfilled at very high grafting density and/or high molecular weight. By invoking the SST approximation, Milner, Witten and Cates [5] and Zhulina and coworkers [5] obtained an explicit expression for the pressure and density profile of a one-component brush:

20.2 Theory of Mixed Polymer Brushes 2 2

ðzÞ ¼

p h 2 2 ½1 ðz=hÞ 8mN

(1)

where v denotes the segmental excluded volume. The brush height takes the form h = (4mrb2/p2)1/3, where r is the number of chains per unit surface (i.e., the grafting density) and b denotes the statistical segment length of the unperturbed chain. By employing the SST, Marko and Witten [1] demonstrated that a mixture of immobile, anchored polymer chains of two chemically distinct types undergoes microphase separation, when the incompatibility between different species is sufficiently high. They predicted that the species would segregate into lamellar domains – parallel cylinders which run parallel to the grafting surface – and calculated the onset of lateral ordering (spinodals) in a melt brush. Within SST, density and composition fluctuations decouple – that is, microphase separation does not affect the parabolic (total) density profile. Microphase separation occurs due to a balance between the reduction of energetically unfavorable contacts between unlike species and the loss of entropy of free chain ends as they are confined into regions in which the appropriate species is enriched. The wavelength of the “ripples” is approximately twice as large as the unperturbed end-to-end distance Re = bN1/2, and similar to the length scale observed in the morphologies of diblock copolymers. Marko and Witten denoted this morphology as the “ripple” phase. Intriguingly, this laterally structured morphology is thermodynamically more stable than the “sandwich” morphology, in which one species segregates to the grafting surface while the other is enriched at the top of the brush, but the system remains laterally homogeneous. The structure in solution has been explored by Balazs and co-workers [6] using scaling considerations and two-dimensional self-consistent field (SCF) calculations. These calculations are not restricted to very high grafting densities. Balazs and colleagues found that in poor solvent and for low grafting density, the brush self-assembles into an ordered array of clusters, which have “onion” or “garlic-like” structures. The characteristics of the morphology can be controlled via chain length, composition and solvent selectivity. These authors suggested that mixed brushes could possibly be used as coatings for colloids as they enabled some fine-tuning of the colloid-colloid interaction [2]. Several numerical studies of binary brushes have been undertaken in addition to the SCF approach. Soga et al. [7] used a coarse-grained simulation method involving direct calculation of the Edward’s Hamiltonian to study the behavior of the brush in a wide range of solvent conditions. They presented evidence for more complex morphologies in bad solvents. Effects of varying the composition have been explored in Monte Carlo simulations of the bond fluctuation model by Lai [8]. Recently, we have calculated the phase diagram of mixed polymer brushes using three-dimensional SCF calculations [3a]. The calculation technique was designed initially by Matsen and Schick [9] for the calculation of the phase diagram of blockcopolymers. The method employs a Fourier representation of the monomer densities and effective fields, and allows for an analysis of the stability of different lateral morphologies, such as lamellar “ripple”, checkerboard and hexagonal “dimple” phases (Figure 20.2(a–c)). We have obtained the location of the phase transitions

405

406

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

(Figure 20.2(d)) between different morphologies as a function of experimentally accessible parameters (e.g., incompatibility (v), grafting density (r), or solvent selectivity (n)). All SCF-like descriptions of polymer brushes [1–5,6] start from a hypothetical unperturbed chain of Gaussian statistics, and describe the segmental interactions via a second- or third-order virial expansion. This is, of course, a great simplification. On one hand, we expect the virial expansion only to capture the gross features of an equation of state of two polymers in a common solvent. On the other hand, the size of the hypothetical Gaussian chain is difficult to identify for experimental systems [10]. Nevertheless, SCF calculations provide valuable insights into the qualitative behavior and correlations between experimentally accessible parameters (e.g., solvent quality) and morphologies. The different morphologies that we have investigated in our three-dimensional SCF calculations are presented in Figure 20.2(a–c) for a brush of symmetric composition in a non-selective solvent [3a]. The morphologies are truly three-dimensional, periodic in the two lateral directions, and they possess a brush-like structure perpendicular to the grafting surface. If the brush is not strongly stretched, then the density and composition fluctuations will be coupled – the density is higher in segregated regions and lower at the internal interfaces of the morphology. Figure 20.2(a) depicts contours of equal density in the “ripple” phase, where the cylinders are enriched alternatively with A-polymers and B-polymers. Figure 20.2(b) shows the checkerboard morphology, where both components collapse into dense clusters (“dimples”) which arrange on a quadratic lattice. In both morphologies, the symmetry between the two components A and B is retained. The interaction with macroscopic bodies (e.g., large water drops) depends only on laterally averaged properties, which do not exhibit any perpendicular variation in those two morphologies in a nonselective solvent. Although the brush is symmetric, the symmetry between the A and B components can be broken spontaneously. One possibility would be the perpendicularly segregated but laterally homogeneous “sandwich” morphology. In accord with Marko and Witten’s calculations [1], we find this morphology to be pre-empted by laterally structured phases. Another morphology, which spontaneously breaks the AB symmetry and is thermodynamically stable for certain parameters, is the hexagonal “dimple” morphology, which is depicted in Figure 20.2(c). One component forms clusters which arrange on a triangular lattice, while the other component fills the interstitials. In this morphology, lateral segregation goes along with a laterally averaged perpendicular segregation: the cluster-forming component is located closer to the grafting surface, while the other component is enriched at the brush’s top. The complete phase diagram of a symmetric brush as a function of incompatibility v between the species and inverse stretching d~(Re/h)2 is shown in Figure 20.2(d). At low incompatibility or small stretching (e.g., low grafting density), the brush remains disordered, but upon increasing incompatibility we encounter different laterally segregated morphologies. Transitions between the disordered phase and the “ripple” phase are second-order, while all other transitions are weakly first-order. For a particular value of stretching and incompatibility, all three laterally ordered phases coexist and the density contour plots in Figure 20.2(a–c) correspond to this triple point.

20.2 Theory of Mixed Polymer Brushes

The morphology of the brush can be controlled by various parameters. We emphasize that there is a coupling between the laterally averaged perpendicular composition profiles, which are important for the adhesion or wetting properties, and the lateral morphologies: small changes in an external control parameter might

a

b

c

d

Contour plots of the density in the different morphologies. (a) “Ripple” morphology; (b) checkerboard morphology; (c) hexagonal “dimple” morphology. Panel (d) shows the phase diagram of a symmetric mixed brush (i.e., symmetric in composition and chain architecture in a nonselective solvent) as a

Figure 20.2

function of the incompatibility and the inverse stretching. The line DIS-1D represents the second-order transition between the disordered phase and the sandwich morphology; it is, however, pre-empted by laterally structured phases. (Reprinted with permission from Ref. [3a].)

407

408

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

give rise to a transition between distinct lateral morphologies, which in turn, result in a significant change of the laterally averaged surface properties. Two experimentally accessible examples have been explored within SCF calculations: a solvent which is selectively favorable to one component can alter the morphology [11]. Nonselective solvents stabilize the “ripple” phase, which does not exhibit any spontaneous perpendicular segregation, while selective solvents might induce a transition to a hexagonal “dimple” structure, where the component for which the solvent is worse forms the dimples. Recently, we have extended the theory of binary mixed polymer brushes to incorporate selective interactions with external confining walls [12]. The phase diagram of a symmetric brush as a function of polymer incompatibility and interaction with the top surface, is presented in Figure 20.3. The phase diagram comprises a region of laterally disordered phase in which the densities do not depend on the lateral coordinates, but the preferred component segregates to the top surface (sandwichlike structure). There are regions of stability for the “ripple”, checkerboard and hexagonal “dimple” morphologies. In the dimple-A phase, the A component segregates into clusters, which arrange laterally on a hexagonal lattice, while the B component (matrix) is less dense and fills the space between the clusters. Increasing the magnitude of the surface field Uw enhances the perpendicular segregation and leads to enrichment of the preferred component at the top layer of the brush. At high incompatibilities, the first-order phase transition between the dimple-A and dimple-B phases at Uw = 0 is accompanied by a jump of the surface chemical composition. Eventually, strong surface fields destroy the lateral order.

Phase diagram of a binary mixed brush interacting with a confining wall. The dashed line, ending in a dot, marks the pre-empted coexistence of “sandwich” structures.

Figure 20.3

20.3 Synthesis of Mixed Brushes

20.3

Synthesis of Mixed Brushes

The two main approaches employed for synthesizing one-component polymer brushes are also used to fabricate mixed polymer brushes, namely “grafting to” [13] and “grafting from” [14]. The main difference between the fabrication of one-component and binary brushes consists of twice repeating the grafting procedure; the grafting of the first polymer is followed by grafting of the second polymer. One of the few exceptions to this is represented by methods where two polymers are grafted in one step from a blend [15], or the grafting of block-copolymers where the anchoring group is located near the bond connecting two different blocks in the block-copolymer [16]. All methods potentially face the same problems of inhomogeneous grafting. Laterally inhomogeneous or spatially correlated grafting introduced by synthesis might seriously affect the morphology; thus, all synthetic routes aim to overcome this problem. 20.3.1

The “Grafting To” Method

This method of synthesizing binary polymer brushes is based on subsequent grafting of end-functionalized polymers to a solid substrate. Before grafting, a surface treatment is usually applied to introduce appropriate functional groups onto the substrate surface. We use a plasma treatment to introduce amino or hydroxyl functional groups on polymeric substrates [17,18] or x-functional silanes to introduce epoxy, amino, and hydroxyl functional groups onto Si-wafers [19,20]. The grafting of a mixed brush is shown schematically in Figure 20.4. The first step consists of chemisorption of 3-glycidoxypropyltrimethoxysilane (GPS) onto cleaned silica,

Figure 20.4

Sketch of the “grafting to” method. See text for details.

409

410

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

which forms a layer with very high concentration of epoxy groups on the surface. Its ellipsometric thickness is 0.7–1.1 nm, corresponding to 1–1.5 theoretical monolayers. The second and third steps consist of grafting the carboxyl terminated poly(2-vinylpyridine) (P2VP-COOH) and polystyrene (PS-COOH), respectively. The polymers were grafted from melt. A thin 10–50 nm PS-COOH film is spin-coated onto the substrate, and the polymer is grafted by annealing the sample at a temperature higher than glass transition temperature (Tg). The nongrafted polymer is removed by Soxhlet extraction with THF. Subsequently, a 50 nm- thick film of the second polymer P2VP-COOH is spin-coated on top of the PS-COOH brush and grafted upon heating above Tg. The kinetics of grafting of PS-COOH and P2VP-COOH in terms of the ellipsometric thickness of the layer is presented in Figure 20.5(a) and (b), respectively. In this figure, plot Lc’ shows the grafting kinetics of P2VP-COOH after grafting a 3.5-nm layer of PS-COOH. It is worth noting that the second polymer can be successfully grafted only if the first polymer has a much smaller affinity to the solid substrate than the second. In this case, the strong affinity of the second polymer to the solid substrate acts as a driving force for chains of the second species to penetrate the brush layer formed by the first polymer. Binary brushes that are synthesized by the “grafting to” method, are macroscopically homogeneous, as demonstrated by ellipsometric mapping (Figure 20.6, level D) and atomic force microscopy (AFM), but they exhibit phase segregation on a nanoscopic length scale. The composition of the brush (the amount of each polymer grafted) can be controlled by conditions of the grafting procedure, including time, temperature, and thickness of the spin-coated films. The drawback of the “grafting to” procedure is that only a relatively small amount of polymer can be grafted onto the surface, although the number of grafted chains of polymers is much smaller than the number of functional reactive groups present on the surface. Diffusion of polymer chains to the surface is strongly limited by the

Growth of the film thickness as measured by ellipsometry for a mixed brush of carboxyl-terminated polystyrene (PS) and poly(2-vinylpyridine) (P2VP). Grafting kinetics for: (a) PS homopolymer brush; (b) P2VP

Figure 20.5

homopolymer brush; (c) for grafting of the mixed brush, when P2VP is being grafted after initial grafting of 3.5 nm of PS. (Reprinted with permission from Ref. [13].)

20.3 Synthesis of Mixed Brushes

Ellipsometric mapping of a Si wafer following each grafting step. The Z-axis represents the increasing layer thickness starting from the Si substrate (zero point). (A) SiO2 native layer after cleaning; (B) grafted GPS; Figure 20.6

(C) grafted PS-COOH (Mn = 16 000); (D) grafted binary brush PS-COOH + P2VPCOOH (Mn = 39 200). (Reprinted with permission from Ref. [13].)

already grafted polymer chains. The maximal thickness of binary brushes that can be achieved with the “grafting to” method usually does not exceed 10 nm, depending on the molecular mass of the grafted polymers. Nevertheless, this method achieves a grafting density of 0.1–0.2 nm–2, which is quite high for polymers with molecular mass ranging from 10 000 to 100 000 g mol–1. The concomitant end-to-end distance of the nongrafted chains in h solvent ranges from 5 to 10 nm and, hence, the chains strongly interdigitate on the surface. 20.3.2

The “Grafting From” Method

The “grafting from” approach is based on surface-initiated polymerization. From the outset, the substrate surface is functionalized to anchor the initiator for radical polymerization [21–23]. Recently, we identified a rather effective and reproducible method of anchoring the initiator for radical polymerization on silica substrate. Initially, the cleaned Si-wafer is modified with GPS (see Section 20.2), which results in a surface functionalized with epoxy and hydroxyl groups. The surface is then treated with ethylene diamine to open the epoxy rings and substitute them with more reactive amine groups. We then graft an acid chloride derivative of 4,4¢-azobis(4-cyanopentanoic acid) from dichloromethane solution containing catalytic amounts of triethylamine. This procedure results in a dense monolayer of the initiator attached to the solid substrate. Subsequently, the substrate is embedded in a monomer solution to graft the first polymer (see synthesis scheme in Figure 20.7). The grafted amount of the first polymer is controlled through the known decomposition rate of the initiator immobi-

411

412

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

Scheme of synthesis of binary polymer brushes via the “grafting from” approach by the two-step polymerization with controlled rate of surface-initiation. The same initiator is used for the both steps.

Figure 20.7

lized on the surface. Once the polymerization is stopped and the nongrafted polymer is removed by cold Soxhlet extraction, the residual amount of initiator can be used for polymerizing the second monomer. By using this “grafting from” approach, we have synthesized mixed brushes with grafting densities of 0.1–0.2 nm–2 for polymers of high molecular mass (up to 800 000 g mol–1). The brushes prepared in this manner were much thicker (100– 200 nm for dry film) than those prepared using the “grafting to” method. The characteristics of the mixed brushes synthesized using the “grafting from” method are listed in Table 20.1. In the first step, we have grafted a random copolymer poly(styrene-co-2,3,4,5,6-pentafluorostyrene) (PSF) (25% of Fluor-containing monomer units), whilst in the second step we have grafted poly(methylmethacrylate) (PMMA). A comparison of the distance between grafting points and end-to-end distance of the grafted chains provides clear evidence for dense brushes. Table 20.1

Characteristics of the PSF/PMMA brushes prepared by “grafting from” approach.

Sample Molecular no. weight Mw 1 10–3 g mol–1 PSF/PMMA

Grafted thickness (nm) PSF/PMMA/ (PSF+PMMA)

Grafted amount (mg m–2) PSF/PMMA

Grafting density (1102 nm–2) PSF/PMMA

Distance between grafted points (nm) PSF/PMMA/ (PSF+PMMA)

End-to-end distance for the non grafted chains RePSF/RePMMA (nm)

1 2 3 4 5

23.4/42.5/65.9 23.9/30.0/53.9 23.4/30.6/54.0 23.9/29.2/53.1 54/68/122

28/51 28/36 28/37 28/35 65/82

3.6/3.8 3.8/2.7 3.6/2.7 3.8/2.6 10.3/5.9

6.0/5.8/4.2 5.8/6.9/4.4 6.0/6.8/4.5 5.8/7.0/4.5 3.5/4.7/2.8

48/58 47/58 48/58 47/58 43/59

475/811 452/811 475/811 452/811 380/840

20.4

Experimental Study of Phase Segregation in Mixed Brushes

Two incompatible polymers in a mixed brush tend to segregate. Macroscopic segregation is prevented by covalent grafting of the chains, which arrange to avoid unfa-

20.4 Experimental Study of Phase Segregation in Mixed Brushes

vorable interactions by microphase separation. This behavior is a key to the tailoring of surface properties, and is an important ingredient for the switching mechanism of mixed brushes. The phase segregation depends strongly on environment. On the one hand, this property is important for the application of responsive surfaces, but on the other hand it creates problem for investigating the phenomenon. In our experiments, the assumption is made that the morphologies observed in dry films are also characteristic for the morphology of swollen films (i.e., under solvent). There is much evidence for this: 1.

After exposure to a particular solvent, the film is rapidly dried by a flow of nitrogen within a few seconds. The time of solvent evaporation is very much shorter than the characteristic time for transforming one morphology into a different one upon changing the solvent. The latter time scale ranges from several minutes to hours. Hence, rapid solvent evaporation freezes the morphology as it was in the solvent. Under these conditions no effect of the drying rate has been observed.

Morphologies of the mixed PSF/ PMMA brush step by step exposed to solvents of different selectivity: (a) toluene; (b) chloroform; and (c) acetone. Atomic force microscopy (AFM) repulsive tapping mode, set-point

Figure 20.8

ratio A/A0 50%, A0 45 nm. Left: topography (black horizontal lines mark origins of topography profiles); middle: phase contrast. Scale 2 B 2 lm. Right: topography profiles. (Reprinted with permission from Ref. [11].)

413

414

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

2.

3. 4.

The morphologies after treatment by a particular solvent are reproducible, and the morphology reversibly switches upon exposure to another solvent. Therefore, morphologies in our experiment are (close to) equilibrium structures. The morphology after drying does not change in a poor solvent, and is stable for a long period of time. Experimental evidence for the similarity of the morphologies of swollen and dry mixed brushes was obtained only recently [24].

In the following, we exemplify transitions between morphology using a mixed PSF/PMMA brush. It is found that this brush segregates into very distinct morphologies on exposure to solvents of different selectivity (Figure 20.8). For example, treatment with toluene (nonselective solvent) and chloroform (increased selectivity for PMMA) results in a “ripple”-like brush morphology (Figure 20.8(a) and (b)), whereas treatment with acetone (selective solvent) produces clusters (“dimple”-like morphology) of PSF embedded in the PMMA matrix (Figure 20.8(c)). Unlike the theoretical calculations, there is no long-ranged order. This conclusion has been corroborated by X-ray photoemission electron microscopy (XPEEM), which utilizes near-edge X-ray absorption fine structure (NEXAFS) contrast to probe directly the chemical composition of the top layer (~5–15 nm) of thin polymer films. In the XPEEM microscope (beamline 7.3.1.1 of the Advanced Light Source at Ernest Orlando Berkeley National Laboratory, CA, USA), a polymer film, which has been immobilized on a conductive substrate (Si), is irradiated with monochromatic X-ray at an angle of incidence of 30Q. Photoelectrons emitted by the sample are focused by a system of electrostatic lenses onto a phosphorus screen, and the image produced is recorded by a CCD camera. The C1sﬁ p* absorption

Near-edge X-ray absorption fine structure (NEXAFS) spectra recorded with the XPEEM microscope from one-component poly(styrene-co-2,3,4,5,6-pentafluorostyrene) (PSF) and poly(methylmethacrylate) (PMMA) brushes at the 1s C edge.

Figure 20.9

20.4 Experimental Study of Phase Segregation in Mixed Brushes

Figure 20.10 X-ray photoemission electron microscopy (XPEEM) images of a mixed PSF/ PMMA brush after exposure to: (a,b) toluene and (c,d) acetone. Scale 3 B 3 lm. The contrast and brightness were optimized. The brush exposed to toluene revealed two kinds of elongated features which have inverse contrast

at photon energies (a) 286.1 and (b) 289.3 eV. No contrast inversion was found at these photon energies for the same brush exposed to acetone, (c) and (d), respectively. The lighter tone of features correspond to higher photoelectron flux. (Reprinted with permission from Ref. [11].)

peaks of PSF and PMMA are shifted by ~3 eV (Figure 20.9), thus providing a sufficient contrast between these two polymers. Two XPEEM micrographs were recorded at photon energies of 286.1 and 289.3 eV, respectively, from the same region on a PSF/PMMA brush, which had been dried after exposure to toluene. The same “ripple”-like morphology was found on both micrographs. The features enriched with PSF appeared light-gray at 286.1 eV (Figure 20.10(a)) and dark at 289.3 eV (Figure 20.10(b)), while the features enriched with PMMA appeared dark at 286.1 eV and light-gray at 289.3 eV. Thus, the observed morphology consisted of alternating domains composed of different polymers, and was in good agreement with the theoretical prediction. No chemical contrast between features was found on the micrographs recorded at 286.1 and 289.3 eV (Figure 20.10(c,d)) after exposure of the brush to acetone. The AFM image (see Figure 20.8(c)) is affected by the topographical relief of the film formed during solvent evaporation. Acetone is a selective solvent for PMMA, and this polymer is highly swollen; PSF forms clusters embedded in the swollen

415

416

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

PMMA. During the evaporation of acetone, the PMMA layer undergoes a large contraction which gives rise to the topographical relief, and it was due to this effect that the cluster morphology of the film was observed. XPEEM is sensitive to the chemical contrast of the surface morphology. The absence of any contrast on XPEEM images suggests that the top of the brush is preferentially occupied by PMMA, and thus a combination of AFM and XPEEM supports the theoretical prediction of “dimple” morphology in a selective solvent. Lateral and perpendicular segregation occurs simultaneously: clusters form of the component which is disfavored by the solvent, while the other component, for which the solvent is good, preferentially occupies the top layer. The size of the microsegregated domains is an important probe for homogeneous grafting of two polymers in the mixed brush. Theory predicts the size of laterally segregated domains to be about 1.85 Re for ripple and 2.15 Re for dimple phases. The results of the experimental study with AFM are presented in Table 20.2. By applying Fourier transformation of the topography AFM images, obtained from PSF/PMMA brushes exposed to toluene and acetone, we were able to identify the mean periods between the ripple phases (dRFT) and the dimples (dDFT), respectively, and compare them with the values predicted by the SCF theory. The latter values were approximated by: dSCF = (Re PSF + Re PMMA) T 0.5k, where RePSF and RePMMA are the root-mean-square end-to-end distances in h-conditions for PSF and PMMA, respectively, and k is the coefficient obtained from the SCF theory (which equals 1.85 and 2.15 for the ripple and dimple phases, respectively). The values of RePSF and RePMMA were calculated from the weight-averaged molecular masses (Mw) of PSF and PMMA. The experimentally obtained values were, on average, not more than 50% higher than the SCF prediction. In view of the rather large uncertainties in identifying the parameters of the SCF calculations, and the possible effects due to polydispersity or slight spatial correlations in the grafting points, we consider these qualitative agreements as gratifying. Period for the dimple and ripple phases predicted by the SCF theory (dDSCF and dRSCF, respectively) with the corresponding experimental values obtained from Fourier transformation of AFM images of the PSF/PMMA brushes exposed to toluene and acetone (dDFT(acetone) and dRFT(toluene), respectively).

Table 20.2

Sample no.

hhPSF (nm)

hhPMMA (nm) dDSCF (nm)

dRSCF (nm)

dDFT (nm)

dRFT (nm)

1 2 3 4 5

48 47 48 47 43

58 58 58 58 59

105 104 105 97 94

155 151 168 110 154

151 168 111 115 154

117 116 117 116 112

20.5 Adaptive Responsive Behavior: Regulation of Wetting and Adhesion

20.5

Adaptive Responsive Behavior: Regulation of Wetting and Adhesion

Phase segregation in mixed brushes upon exposure to solvents of different qualities causes a strong alternation of the surface chemical composition of the brushes. In this way, the mixed brush adapts to its environment and, at the same time, the surface behavior of the material is dramatically changed (switched). This switching is a reversible process such that the surface properties of the mixed brush (wetting behavior, adhesion, adsorption, etc.) can be changed identically, an on many occasions. This can be demonstrated with a very simple wetting experiment. After exposure of the mixed brush to different solvents, we freeze the morphology due to the rapid solvent evaporation. Then, we probe the wetting behavior by using a fast probe (within 1 min), namely the measurement of the advancing contact angle of a water drop set on the brush. The switching properties with PSF/PMMA and PS/P2VP mixed brushes are exemplified in Figures 20.11 and 20.12. The experiments demonstrate the reversibility and the reproducibility of the switching behavior as well as the high sensitivity of the mixed brushes to different solvents. For example, in the case of PS/P2VP mixed brush (Figure 20.12), the advancing contact angle changes from 90Q (upon exposure to toluene) to 20Q (upon exposure to acidic water). The changes of the solvent are accompanied by a change in the morphology and the laterally averaged composition of the top layer. For example, in toluene PS is enriched in the top layer, while P2VP predominantly occupies the top layer upon exposure to acidic water. Switching kinetics depends on the composition of the mixed brush and the molecular mass of its polymers. The time of switching ranges from seconds to hours (cf. Figure 20.13).

Figure 20.11 Switching of the surface state of a mixed PSF/ PMMA brush upon exposure to various solvents. The water contact angles were measured with the sessile drop technique on the brush after exposure to the solvent and rapid drying in nitrogen flux.

417

418

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

Figure 20.12 Switching of the surface state of the mixed PS/ P2VP brush upon exposure to various solvents.

Chloroform is a more selective solvent for PMMA, and switches the brush to the state when the top layer is enriched with PMMA. At the same time, the ripple morphology of the film (see Figure 20.8) suggests that both of polymers occupy the top layer. This is an example of a boundary situation, when solvent still stabilizes the ripple morphology, but the top layer tends to be preferentially occupied by one of the polymers. This mixed brush is switched with the solvents very rapidly, and we may assume that the switching rate is of the same order as the solvent evaporation rate. However, the monotonous kinetics of the switching provided evidence that no specific changes in the top layer composition were introduced while the solvent evaporated.

Figure 20.13 Kinetics of switching of the mixed PSF/PMMA brush from the hydrophobic to hydrophilic state by chloroform and in the opposite direction by toluene. The switching time in both cases was less than 6 s.

20.5 Adaptive Responsive Behavior: Regulation of Wetting and Adhesion

Figure 20.14 Schematic representation of the two-level morphology of the polymer film with grafted binary polymer brush. (a) The arrow indicates the location of the mixed brush on

the surface of the needle-like PTFE substrate; (b) illustrated with a scanning electron microscopy image of size 20 B 20 lm2. (Reprinted with permission from Ref. [25].)

Hydrophobic and hydrophilic properties can be strongly amplified by the roughness of the substrate [25]. We exploit this amplification of the switching effect of the wetting behavior by synthesizing a mixed PSF/P2VP brush via the “grafting to” technique onto a plasma-etched PTFE foil (Figure 20.14) with a roughness on the micrometer scale. The mixed polymer brush forms domains of nanometer size. The substantial amplification of the switching range by the needle-like micrometer roughness of the etched PTFE substrate is demonstrated in Figure 20.15: an advancing contact angle of 160Q is measured after exposure to toluene. A drop of water is then able to roll easily on the surface – a fact that indicates also a very small hysteresis of the contact angle (Figure 20.15(a)). However, following immersion of the same sample in acidic water (pH 3) for several minutes to switch the brush and dry the sam-

Figure 20.15 Rolling of a water drop on PTFE with grafted PSF/P2VP binary brush. (a) After exposure to toluene; (b) wicking after exposure of the brush to acidic water. (Reprinted with permission from Ref. [25].)

419

420

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

Figure 20.16 Switching of adhesion. The plot presents the change of the force applied to the TesaH band defoliated from PTFE substrate with grafted PS/P2VP binary brush versus distance (x) from the starting point. The dashed

line marks the border between ultra-hydrophobic (dark) and hydrophilic (gray) area on the brush switched with toluene and acidic water, respectively. (Reprinted with permission from Ref. [25].)

ple, a drop of water is able to spread on the surface because of the wicking effect (Figure 20.15(b)). The mixed brushes can be utilized for constructing sophisticated functional materials. Switching of the brush results in tuning of various kinds of interaction mechanisms with its surroundings. An example of practical importance is van der Waals interactions, which can be used to regulate adhesion. We have performed a simple adhesion test in which we glue TesaU tape to a plate covered with a PS/P2VP mixed brush. Half of the plate is in the hydrophilic state, and the other half is in the hydrophobic state. A sharp decrease in adhesion is observed when crossing the border between the ultra-hydrophobic and hydrophilic areas on the sample (Figure 20.16).

20.6

Patterning of Mixed Brushes

Patterning of the mixed brush allows fabrication of devices that exploits their switching behavior. We provide an example where a mixed brush is patterned due to the local “freezing” of morphology by crosslinking of polymers [26]. A thin film of the mixed polymer brush, prepared from polyisoprene (PI) and P2VP, is exposed to a selective solvent and illuminated through a photomask. In the illuminated areas,

20.6 Patterning of Mixed Brushes

Figure 20.17 Scheme of photolithography of mixed polymer brushes (see text for details) (Reprinted with permission from Ref. [26].)

the brush is crosslinked, but in the dark areas the chains retain their capability of switching conformation and properties. Consequentially, the designed pattern can be developed by exposing the brush to a selective solvent for one constitutive polymer. The treatment with solvent changes the chemical composition of the top layer in the dark areas as a result of phase segregation, but in the illuminated areas the brush remains unchanged. Whenever the patterned mixed brush is exposed to a nonselective solvent, any contrast in the chemical composition of the top layer disappears and the image is erased. This process is reversible, and so pattern development and erasure can be repeated many times (Figure 20.17). Figure 20.18 shows the characteristic features of the patterned brush after interaction with different solvents. After illumination through the photomask, the pat-

421

422

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

Figure 20.18 Optical microscopy of water droplets on a mixed polymer brush with (a) developed and (b) erased pattern. The inset presents a detailed view. (Reprinted with permission from Ref. [26].)

terned brush is washed with ethanol for several minutes, dried, and then exposed to water vapor. No pattern develops on the surface (Figure 20.18(b)). The brush is then treated with acidic water (pH = 2) and dried. After exposure to water vapor, the local condensation of water droplets reveals quite clearly the image imprinted onto the brush (Figure 20.18(a)). The inset of the figure demonstrates the difference in wetting properties of the dark and illuminated areas. Water barely wets the surface of the illuminated areas (semi-spherical droplets), but spreads more extensively over the surface of the dark areas. An image with strong contrast is formed because the light reflection changes with the size and shape of the water droplets. The image can be erased merely by washing with ethanol or neutral water (pH = 6.5). This process provides evidence that the film is sensitive to acidic water, and this can be repeated many times. It has been noted that illuminated mixed PI/P2VP brushes stored for more than 1 month after exposure to one selective solvent (e.g., acidic water) preserve their morphology and the concomitant information. These samples, when exposed to water vapor, reveal the photoprinted images. However, the image can be erased upon heating above 80 QC (the Tg of PI is below room temperature, and PI occupies the top of the brush upon heating to a temperature slightly below Tg = 85 QC of P2VP). Further exposure to acidic water, followed by drying and contact with water vapor, restores the image. Therefore, heating the sample erases the image without destroying the brush. Two different examples for practical application of this environment-responsive lithography can be suggested. The first example shows that this approach can be used for fabricating tunable microchannels. The tunable channel is prepared as shown in Figure 20.19, with the mixed brush (1) being grafted between two hydrophilic channels (2) (Figure 20.19(a)) on a solid substrate. The channel is fabricated (Figure 20.19(b)) by photo-crosslinking of the brush through the photomask (3), such that the irradiated areas (4) lose their switching ability and serve as walls of the

20.6 Patterning of Mixed Brushes

Figure 20.19 Fabrication of a tunable channel employing a mixed brush (see text for details). The images (e) and (f) show open and closed states of the tunable channel, respectively, as they appear in optical microscopy. (Reprinted with permission from Ref. [26].)

channel. The bottom of the channel (6) becomes hydrophilic upon exposure to acidic solution, and water can flow through the channel (Figure 20.19(c,e)). Upon heating above 80 QC, or by changing the pH (Figure 20.19(d,f)), the bottom surface of the channel switches to a hydrophobic state, closing the channel (5). Although this example demonstrates the general principle, much more complicated systems can be fabricated using photolithography on mixed brushes. The second experiment involves the fabrication of smart sensors to directly visualize the test result. For example, if the pattern is visualized at a particular pH, the smart surface can display the value of pH directly as an image (cf. Figure 20.20).

Figure 20.20 Example of a smart sensor from a mixed brush grafted to Si wafer which displays the result of the analysis of acidic aqueous solution: the wafer was exposed to neutral water (top) and to water with pH 2.3

(bottom). The image appeared upon exposure to water vapor only if the sample had been treated with acidic water solution of pH <2.5. (Reprinted with permission from Ref. [26].)

423

424

20 Mixed Polymer Brushes: Switching of Surface Behavior and Chemical Patterning at the Nanoscale

Figure 20.21 AFM image (a) of the oxidized surface of PDMS stamp and (b) XPEEM-image of the binary PS/P2VP brush after the contact with the stamp. Light patterns correspond to higher fraction of PS; dark patterns indicate the higher fraction of P2VP.

An alternative way of reversibly patterning mixed brushes is based on the interaction with a confining wall [12]. The binary PS/P2VP brush is brought into contact with a rubber polydimethylsiloxane (PDMS) stamp, the surface of which is hydrophilized with oxygen plasma. The stamp surface is profiled with shallow square holes (30 nm deep and 5 lm wide), arranged in a square order 12 lm apart from each other. The stamp is removed from the brush surface after evaporation of the solvent (chloroform). The XPEEM image of the patterned brush surface in comparison with the AFM image of the oxidized surface of PDMS stamp is shown in Figure 20.21. The equilibrium state of the PS/P2VP brush, swollen by chloroform, is changed upon contact with the stamp. The hydrophilized stamp attracts P2VP and repels PS, this effect being greater in the areas of stronger contact between the stamp and the brush, and smaller where the contact is weaker (under the holes). The network of “cracks” seen on the XPEEM image reproduces the topography of a similar network on the stamp surface (the cracks on the stamp appear upon oxidation in plasma due to losses in elasticity of oxidized PDMS). As the crack width is about 100 nm, this provides an estimate for the resolution of the pattern transfer.

Summary

We have demonstrated that traditional “grafting to” and “grafting from” approaches can be employed to fabricate mixed polymer brushes from polymers of different composition and molecular mass. Due to lateral microphase segregation, the mixed brushes exhibit a very rich behavior which constitutes the basis for adaptive switching. Gratifyingly, we have found qualitative agreement between theory and experiment – a fact that demonstrates the salient features of this complex experimental

References

system may be captured by the highly simplified model of SCF theory. We provide several examples to illustrate the usefulness of responsive behavior of the mixed brushes for chemical patterning, microfluidic technologies, sensors, or regulation of wetting behavior, adsorption and adhesion.

References 1 J. F. Marko, T. A. Witten, Phys. Rev. Lett.

1991, 66, 1541–1544. 2 C. Singh, G. T. Pickett, E. B. Zhulina, A. C. Balazs, J. Phys. Chem B 1997, 101, 10614–10624. 3 (a) M. MWller, Phys. Rev. 2002, E65, 030802(R); (b) S. Minko, I. Luzinov, V. Luchnikov, M. MWller, S. Patil, M. Stamm, Macromolecules 2003, 36, 7268–7279. 4 N. Semenov, JETP Lett. 1985, 61, 733. 5 (a) S. T. Milner, T. A. Witten, M. E. Cates, Macromolecules 1988, 21, 2610–2619; (b) S. T. Milner, T. A. Witten, M. E. Cates, Europhys. Lett. 1988, 5, 413–418; (c) A. M. Skvortsov, A. A. Gorbunov, V. A. Pavlushkov, E. B. Zhulina, O. V. Borisov, V. A. Pryamitsyn, Polym. Sci. USSR 1988, 30, 1706; (d) E. B. Zhulina, O. V. Borisov, V. A. Pryamitsyn, J. Colloid Interf. Sci. 1990, 137, 495. 6 (a) E. B. Zhulina, C. Singh, A. C. Balazs, Macromolecules 1996, 29, 6338–6348; (b) E. B. Zhulina, C. Singh, A. C. Balazs, Macromolecules 1996, 29, 8254–8259; (c) E. B. Zhulina, A. C. Balazs, Macromolecules 1996, 29, 2667–2673. 7 K. G. Soga, M. J. Zuckermann, H. Guo, Macromolecules 1996, 29, 1998–2005. 8 P. Y. Lai, J. Chem. Phys. 1994, 100, 3351– 3357. 9 M. W. Matsen, M. Schick, Phys. Rev. Lett. 1994, 72, 2660–2663. 10 T. Kreer, S. Metzger, M. MWller, K. Binder, J. Baschnagel, J. Chem. Phys. 2004, 120, 4012. 11 S. Minko, M. MWller, D. Usov, A. Scholl, C. Froeck, M. Stamm, Phys. Rev. Lett. 2002, 88, 035502. 12 S. Minko, D. Usov, V. Luchnikov, M. MWller, L. Ionov, A. Scholl, G. PfWtze, M. Stamm, Polymer Preprints 2003, 44(1), 478–479.

13 S. Minko, S. Patil, V. Datsyuk, F. Simon,

14

15

16

17

18

19 20

21

22

23

24

25

26

K.-J. Eichhorn, M. Motornov, D. Usov, I. Tokarev, M. Stamm, Langmuir 2002, 18, 289–296. A. Sidorenko, S. Minko, K. Schenk-Meuser, H. Duschner, M. Stamm, Langmuir 1999, 15, 8349–8355. J. Draper, I. Luzinov, L. Ionov, S. Minko, S. K. Varshney, Polymer Preprints 2003, 44(1), 570–571. J. Wang, S. Kara, T. E. Long, T. C. Ward, J. Polym. Sci. Polym. Chem. 2000, 38, 3742– 3750. M. Motornov, S. Minko, M. Nitscke, K. Grundke, M. Stamm, Polymer Preprints 2002, 43, 379–380. M. Motornov, S. Minko, M. Nitschke, K. Grundke, M. Stamm, Polym. Mater. Sci. Eng. 2003, 88, 264–265. S. Minko, S. Patil, J. Pionteck, M. Stamm, Polym. Mater. Sci. Eng. 2001, 84, 877–878. S. Minko, I. Luzinov, S. Patil, V. Datsyuk, M. Stamm, Polym. Mater. Sci. Eng. 2001, 85, 314–315. S. Minko, A. Sidorenko, E. Goreshnik, D. Usov, M. Stamm, Polym. Mater. Sci. Eng. 2000, 83, 448–449. S. Minko, M. Stamm, E. Goreshnik, D. Usov, A. Sidorenko, Polym. Mater. Sci. Eng. 2000, 83, 533–534. S. Minko, D. Usov, E. Goreshnik, M. Stamm, Macromol. Rapid. Commun. 2001, 22, 206–211. M. C. LeMieux, S. Minko, D. Usov, H. Shulka, M. Stamm, V. V. Tsukruk, Polym. Mater. Sci. Eng. 2004, 90, 272–373. S. Minko, M. MWller, M. Motornov, M. Nitschke, K. Grundke, M. Stamm, J. Am. Chem. Soc. 2003, 125(13), 3896–3900. L. Ionov, S. Minko, M. Stamm, J. F. Gohy, R. JXrYme, A. Scholl, J. Am. Chem. Soc. 2003, 125, 8302–8306.

425

427

21

Local Chain Organization of Switchable Binary Polymer Brushes in Selective Solvents Melbs C. LeMieux, Denys Usov, Sergiy Minko, Manfred Stamm, and Vladimir V. Tsukruk

Abstract

Binary polymer brushes were exposed to selective solvents for each of the two components, and the morphological state, local structure reordering, and modes of nanophase segregation investigated with atomic force microscopy (AFM) imaging and force mapping. Two incompatible polymers, polymethylacrylate (PMA) and poly(styrene-co-2,3,4,5,6-pentafluorostyrene) (PSF), were randomly grafted, using a two-step procedure, onto a silicon wafer using the “grafting from” method, to produce thick (20–150 nm), dense, mixed brush layers. The resulting layers possessed a nanostructured surface exhibiting either lateral, or vertical microphase segregation of the two components. The lateral and vertical reordering of the mixed brush layer was rapid (a few minutes) and reversible for at least 100 “switches” between good and bad solvent states for each component. Direct measurements in selective solvent conditions were compared with results in the dry state immediately after selective solvent exposure. Thus, it is shown directly that the morphology and chain ordering in the dry state after exposure to specific conditions is closely associated with the corresponding solvated case.

21.1

Introduction 21.1.1

Polymer Surface Modification

Polymer surface and interface science has emerged over the past two decades as a critical issue in fields ranging from materials science and surface chemistry, to nanotechnology and bio-inspired research. Traditionally, polymer surfaces have been instrumental in physical and chemical processes to stabilize colloidal suspensions, promote lubrication and control adhesion, and to act as water and oil repellants [1–3]. More specifically, in recent years, a vast number of theoretical and experimental studies has been devoted to applications of thin (1 to 500 nm) polymer films

428

21 Local Chain Organization of Switchable Binary Polymer Brushes in Selective Solvents

constructed from well-organized or patterned polymer chains that are chemically attached (grafted) at one end to a solid surface [4–7]. Today, one of the primary aims of polymer surface science is to create dynamic versatile polymer layers that serve so much more purpose than mere lubrication, repellant, or surface protection. The next generation of sophisticated nanomachines is being designed with integrated mechanical, optical, microfluidic flow, and biological devices to perform complex sensing functions [8]. Such systems require “adaptive surfaces”, or surfaces that have the ability to respond rapidly to external stimulus so that on-demand physical properties are possible. Polymer brushes are prospective materials for such applications because as a result of the high grafting density and uniformity in composition and chain height throughout the brush, the layer responds communally to very subtle changes in the surrounding environment such as fluctuations in pH [9], temperature [10], and solvent quality [11–13]. 21.1.2

Polymer Brushes

Polymer brush refers to a system where polymer chains are strongly adsorbed, or tethered at one end to a surface, or interface, with a sufficiently high enough grafting density. Moreover, the chains act to alleviate overlapping by stretching away from the surface and forming a brush-like structure [14–16]. Irreversible (chemical) grafting of polymer chains to a solid interface generally can result in one of three possible conformations depending upon the grafting density of the chains [17,18]. At low surface concentrations, the chains lie on the surface and form the “pancake” structure. In the case when there is a relatively medium grafting density (or distance between grafting sites, d), then d/Rm » 1, where the free end of the chain tends to

Model comparing the conformation of a free polymer chain (top left) and grafted chains as a function of the grafting density (d) on a solid surface. The chain conformation goes from the pancake, to the mushroom,

Figure 21.1

and finally to the brush conformation (far right) at the highest grafting density. The model for the mushroom structure is oversimplified, in that there is no free volume within the layer underneath the chains.

21.1 Introduction

form a mushroom-like structure [19] with radius Rm (Figure 21.1). However, when the tethering density becomes high and crosses a certain threshold in which d/Rm becomes very small, neighboring chains crowd one another. As a result, densely grafted chains will be more apt to stretch away from the grafting site, and strong deformations of the average dimensions will occur [20]. The resulting layer architecture is known as a “polymer brush”. This situation, in which the polymer chains stretch along the surface normal, is quite different from typical flexible polymer chain behavior in solution, where the well known, random-walk (Gaussian coil) configuration is found (Figure 21.1). In other words, the equilibrium conformation is a highly stretched (or shrunken) conformation. Sometimes, this stretching (or collapse) is very much farther than the typical unstretched size of a chain (often more than five times), especially in the presence of a good (or bad) solvent [21]. Thus, the brush structure is responsible for the physical properties that are important in applications of colloid stabilization [22], drug delivery and biomimetic materials [23,24], chemical gates [25], and tuning lubrication, friction, adhesion, and wettability for tailored surfaces [26–28]. 21.1.3

Binary Polymer Brushes

There are effectively two pathways to develop surfaces with tunable or tailored surface properties: 1) controlling the morphology of the layer; or 2) controlling the chemical composition of the surface in the brush layer. For homopolymer brushes, only the thickness and morphology of the layer can change as a function of the surrounding environment. In the case of two-component or binary polymer brush layers, both the structure and surface chemistry can be altered, and the variety of surface morphologies possible is greatly increased, depending upon the chemistry used. If the two grafted polymers are incompatible, each chain will interact very differently with the surroundings, and switching of phase segregations, morphologies, and the identity of the polymer at the top of the mixed brush can result (Figure 21.2) [29,30]. Surface composition – and hence properties such as surface energy, adhesion, elasticity, and wettability – have the possibility of being precisely tailored to the necessary state. Indeed, the recent development of binary brushes with proven reversibility between morphological states has allowed for much more diverse applications [30,31]. A limited number of theoretical studies have been contrived regarding the phase diagrams and predictions of chain organization in a binary polymer brush film. Based on self-consistent field (SCF) theory, it was proposed that the grafting of two incompatible polymers at high densities produced either a “layered profile”, or a “ripple profile” under melt conditions [29]. The layered profile describes a vertically segregated system where only one of the components is found at the top, and the other at the polymer/inorganic interface. While it is known that sufficiently random, irreversible grafting of incompatible polymers prevents lateral macrophase separation, the ripple profile describes a system in which the two components are laterally segregated, with the dimensions of the lateral structures on the order of the free

429

430

21 Local Chain Organization of Switchable Binary Polymer Brushes in Selective Solvents

Schematic demonstrating how a wide range of surface composition is possible in a binary polymer brush system. The top two schemes show selective environments for each polymer, bringing the favored one to the

Figure 21.2

surface while the other collapses to avoid interaction. At bottom, the both chains favor the surrounding environment, and attempt to increase interaction with it by swelling.

radius of the chains [29]. In these calculations, it was found that the more likely transition would be to the “rippled” state, because it was found that AB interactions are energetically favored over the case where some B chains will be submerged under A tails (or vice versa). Additionally, Soga et al. have also shown that under various solvent conditions, microphase separation (ripple profile) dominates in a brush of two incompatible species, although the interaction of solvent is considered as equal for both species – that is, nonselective solvents [32]. The layering profile has been predicted in some cases, as described by Brown et al. [33,34]. More recently, a new phase diagram has been proposed for a binary polymer brush [35]. In these investigations, the calculations were limited to solvents of good quality, and it was found that in systems with very high polymer incompatibilities, a never-before seen “dimple phase” is thermodynamically stable. When the solvent quality is decreased for species “A”, this species collapses and forms round “A” clusters that supposedly arrange in a hexagonal lattice [35]. This means that, in addition to the ripple phase, the calculations for a binary brush of symmetrical composition surprisingly find a new dimple phase. Additionally, contour plots of the composition at small and large solvent selectivity have been produced and, for the first time, a morphology was predicted that combines both lateral segregation and a pronounced perpendicular (layering) segregation [36]. Up until these most recent studies, which were published in 2003, very little theoretical or experimental work has been devoted to characterizing the structure and morphology of binary polymer brushes in selective solvents for each component, which is quite surprising considering that this type of environment would be found in the majority of applications for such materials.

21.2 Experimental

Here, we report for the first time results of AFM imaging and simultaneous AFM force mapping measurements on binary polymer brushes comprised of glassy poly(styrene-co-2,3,4,5,6-pentafluorostyrene) (PSF) and rubbery polymethylacrylate (PMA) components in selective solvent conditions. We show that the morphology of the brush can be “switched” between different states in which either PMA or PSF are predominantly found at the top of the layer after exposure to their respective good selective solvents. Furthermore, we illustrate that the switching observed in the dry state is quite similar to switching in solvents. While we do not claim that the morphology is exactly identical for a specific state of the mixed brush in dry and solvated conditions, in terms of the resulting chain organization and segregation in the mixed brush, we found identical behavior for each environment (solvated, dry). Although it has been argued earlier that the brush morphology resulting from exposure to a particular solvent should be kinetically “frozen” into place if dried quickly, such that the dry morphology can be closely associated with this particular equilibrium state in solvent [37], there have been no experimental studies conducted to prove that the dry binary brush morphology is similar to that of the same solvated brush. In a series of studies, Koutos et al. observed the morphology of PS brushes only in bad solvent conditions, and the corresponding dry brush was not imaged [38,39]. Kelley et al. studied PS brushes in a good solvent and observed that, due to the excessive chain stretching, overall homogeneous images were obtained [40]. In this chapter, results from AFM imaging of the binary brush in acetone and toluene are presented, and the resulting morphology is compared with that found in the dry state.

21.2

Experimental 21.2.1

Materials and Synthesis

The gel-permeation chromatography (GPC) determined molecular weights of the grafted polymers were completed on a Breeze 1500 instrument (Waters), using polystyrenes as calibration standards. It was assumed that the polymers in bulk have the same molecular weight as polymers grafted onto the substrate. Although contradicting reports have been published regarding this assumption, the kinetics scheme suggests nearly the same molecular weight for grafted chains and chains in the bulk [41], though in some experiments an increased molecular weight and larger polydispersity index for the grafted polymer as compared to the bulk polymer was documented due to the Trommsdorff effect. A detailed explanation of the “grafting from” procedure and molecular weight characterization is given elsewhere [42]. Briefly, we grafted PMA (Mw 505 000 g mol–1) at the first polymerization step, and then PSF (Mw 372 000 g mol–1) at the second step using the residual amount of the azo-initiator on the silicon substrates. Oxygen was removed from the methyl acrylate (MA) monomer solution in toluene (5 mol L–1), or a mixture of styrene (PS)

431

432

21 Local Chain Organization of Switchable Binary Polymer Brushes in Selective Solvents

and 2,3,4,5,6-pentafluorostyrene (FS) in ratio 4:1 wt% in THF (5 mol L–1) and 4,4¢azobis(isobutyronitrile) (AIBN) (4.4 K 10–4 mol L–1) using five freeze-pump-thawcycles. The silicon wafers with the chemically attached azo-initiator were placed into a reactor with the monomer solution under an argon atmosphere. The reactor was immersed in a water bath (60 – 0.1 LC) for 12 h. The silicon wafers were rinsed several times with toluene. The non-grafted polymer was removed by cold Soxhlet extraction in THF for 1 h. The same procedure was used to graft the second polymer. The non-grafted amount of the second polymer was removed by a hot Soxhlet extraction in THF for 12 h. Since complete extraction of any ungrafted chains is a critical issue in this type of system, we have extensively studied the extraction, and have found that layer thickness does not change significantly after 4 h of extraction time. The grafted amount of the polymers was monitored after each polymerization step, using ellipsometry [42]. Typical grafting parameters for PSF and PMA, as well as of the overall binary brush are listed in Table 21.1. Table 21.1

Characteristics of the grafted polymers in the mixed brush.

Polymer

Mw (g mol–1)

Grafted amount (mg m–2)

Grafting density (chains nm–2)

Grafting distance (nm)

PMA PSF PMA+PSF binary brush

505 000 372 000 NA

25.2 43.2 68.4

0.13 0.12 0.15

3.1 3.3 2.9

NA = not applicable.

21.2.2

Methods

To switch the binary brushes with solvents, a sample was immersed in a selective solvent (toluene for PSF, acetone for PMA) for 5 min, and then rapidly dried under a dry nitrogen flux. Samples were immediately scanned with AFM in air using Dimension 3100 and Multimode microscopes (Digital Instruments, Inc.). An indepth description of scanning parameters, AFM tip characterization, and force mapping procedures are detailed elsewhere [42–44]. For measurements in solvents, the AFM tip was brought very close (a few mm) to the sample and, using a micropipette, approximately 40 lL of solvent was placed on to the sample and held in place due to capillary forces between the tip and sample. The binary brush morphology was imaged under solvent conditions in contact mode using a soft tip (0.05–0.1 N m–1) to minimize applied forces. Once the solvent was added, the binary brush was allowed to equilibrate with the solvent; this took less than 5 min based on contact angle and AFM analysis, but longer times were allowed to ensure minimal repulsion/attraction with the AFM tip. AFM scanning was very stable after an equilibration time of 20 min. Since toluene and acetone evaporate rapidly in air, it was neces-

21.3 Results and Discussion

sary during this period to keep adding solvent onto the sample with a micropipette, until the sample engaged with the AFM tip.

21.3

Results and Discussion 21.3.1

Dry State Analysis

The results of our previous studies showed that dry state PSF and PMA homopolymer brushes possess elastic moduli of approximately 1 GPa and 50 MPa respectively, based on direct force measurements with AFM [45]. In addition, the sticky PMA was found to display five times more adhesive strength than PSF as measured by AFM force-distance curves [45]. Thus, these two very different polymers were ran-

Atomic force microscopy (AFM) images in the dry state (left-topography, right-phase) of the glassy (top) and rubbery (bottom) state of the binary brush at 3 & 3 lm. Top picture: Z scale for height is 10 nm, Z scale for phase is 10*. Bottom picture: Z scale for height is 150 nm, Z scale for phase is 40*.

Figure 21.3

433

434

21 Local Chain Organization of Switchable Binary Polymer Brushes in Selective Solvents

domly grafted onto a silicon surface with the goal of creating a single polymer layer that can switch on demand between a soft, rubbery, high-energy surface and a glassy, stiff, slippery fluorinated surface [42]. The binary brush possessed a nanostructured surface in the dry state, and had very different morphologies after exposure to acetone and toluene (Figure 21.3). Micromechanical analysis (MMA) and high-resolution tapping mode/phase imaging verified the complete reversible switching between a “rubbery state” (PMA completely segregated to the surface and occupying the top layer) and a “glassy state” (PSF predominantly found at the top layer) [42]. Cross-sectional analysis and AFM scratch tests revealed that, under dry conditions, the thickness in the rubbery state was close to ten times that of the glassy state [42]. 21.3.2

Morphology in Solvent

AFM images of solvated brushes in toluene (glassy state) and acetone (rubbery state) are presented in Figures 21.4 and 21.5, respectively. These figures can be compared with the corresponding dry state images in Figure 21.3. For the glassy state, the solvent image shows a rough surface with only a very limited number of deep holes. While the nanodomain structure is different from the dry state, the same “type” of morphology is found. As in the dry state, the AFM image under toluene reveals lateral phase separation, although not to the extent as in the dry state, along with incomplete vertical segregation. Furthermore, from section analysis, the holes in the glass state in toluene are on the order of 50 nm, much larger than for the 10-nm scale in the dry state. The surface roughness under toluene is about 3 nm RMS for a 1 K 1 lm area, as compared with 1 nm RMS in the dry state. For the rubbery state, a network, or cellular-like structure of the dry brush is absent in the AFM image under acetone (Figure 21.5). Apparently, in acetone, the chain swelling is so strong that a more enhanced homogeneous PMA layer forms over collapsed PSF, while in the dry state some PMA collapses in air, resulting in the more porous structure. The roughness value of 0.6 nm RMS in acetone is also indicative of a very homogeneous top layer. Although the web-like structure found in the dry rubbery state possesses high roughness, MMA has shown that there is indeed a complete PMA layer over the collapsed PSF – that is, complete vertical segregation was reached. Thus, the conclusion can be reached that the dry state morphology is associated with the solvated morphology. For the dry glassy state, we observed lateral and partial vertical segregation, which was also observed in the solvated state, albeit the vertical segregation was more pronounced in solvent with the deeper holes. For the rubbery state, complete vertical segregation was achieved in both the solvated and dry state. For both the rubbery and glassy state, the degree of vertical layering (segregation) is greatly enhanced in the presence of selective solvents, which is in agreement with theory [36]. It is worth noting that when comparing the solvated images (Figures 21.4 and 21.5) with dry state images of PSF and PMA homobrush layers (see Ref. [45]), a strong resemblance is observed, indicating near-complete vertical reordering

21.3 Results and Discussion

Figure 21.4 Atomic force microscopy (AFM) image of the binary brush in toluene at 2 & 2 lm (top, left) and 10 & 10 lm (top, right) with corresponding three-dimensional and crosssectional analysis plots from the 10 lm image.

435

436

21 Local Chain Organization of Switchable Binary Polymer Brushes in Selective Solvents

Figure 21.5 Atomic force microscopy (AFM) image of the binary brush in acetone at 2 & 2 lm (top, left) and 10 & 10 lm (top, right) with corresponding three-dimensional and cross-sectional

analysis plots from the 10 lm image. The scales in Figures 21.4 and 21.5 are the same, so direct comparison can be made between the solvated glassy and rubbery states.

21.3 Results and Discussion

of PSF and absolutely complete reordering of PMA in the binary brush upon exposure to toluene and acetone, respectively. There are several reasons why complete vertical segregation and a homogeneous top layer are not achieved in the glassy state, but only in the rubbery state. First, the most critical factor concerning swelling of brushes is the Flory-Huggins interaction parameter (r1,2), which takes into account the solubility parameters of a specific solvent and a polymer, and the enthalpy of mixing. The solubility parameters are as follows: PSF = 8.0, toluene = 8.9 (a difference of 0.9); and PMA = 9.7, acetone = 9.8 (difference of 0.1). From this, the r1,2 value is smaller, and more favorable for the PMA/acetone combination [19,47,48]. Second, PSF is more sterically constrained, with its bulky phenyl groups and larger van der Waals radius. Third, PMA was found to have a higher graft density than that of PSF, and it is well known that the higher the graft density, the more homogeneous the brush layer will be [39,40,48,49]. And finally, the glass transition temperature (Tg) of PSF is 108 LC, whereas Tg for PMA is well below room temperature (5 LC), allowing PMA chains to be much more mobile than PSF. Therefore, the experimentally determined morphology of the solvated states is valid based on these arguments, and the theory that the morphology in selective solvent conditions can be “frozen” into place is verified. 21.3.3

Mechanical Response in Solvent

MMA measurements in fluid can further verify the vertical reordering in the binary brush upon switching from the glassy to the rubbery state. Since the solvent evaporates in usually 30 min or less, MMA measurements had to be limited to 16 K 16 pixel resolution, compared with 64 K 64 resolution in the dry state. However, this should not be critical since, in the solvent case it is expected that complete vertical segregation of the favored component takes place, and that there is no lateral segregation of components at the very top of the brush. The MMA data are presented in Figures 21.6 and 21.7 for the glassy state and rubbery state, respectively. When the binary brush is exposed to toluene, PSF swells and becomes very soft in its solvated state, and PMA is already quite soft in the dry state. Thus, even to the highly sensitive AFM tip used in this experiment, PSF and PMA will effectively have the same mechanical response in the glassy state. In the rubbery state however, PSF will collapse in the bad solvent and become moderately stiff (approaching its mechanical behavior in air, a bad solvent). Moreover, if true reordering of the binary brush occurs in which PMA forms a homogeneous layer over PSF, then the depth profiling of the Young’s modulus will show this. Results from MMA indicate that in the glassy state, at higher indentation depths, the effect of silicon is observed around 140–160 nm. This means that, under toluene, the total thickness of the binary brush is around 160 nm (Figure 21.6), compared to 50–60 nm, which is thickness of the dry glassy state [42]. For the rubbery state, the force-distance curve initially shows a soft surface with extremely low resistance for about the first 200 nm of indentation (Figure 21.7). However, at higher

437

8

10

6

8

Elastic Modulus, MPa

Deflection, nm

21 Local Chain Organization of Switchable Binary Polymer Brushes in Selective Solvents

4 2 0 -2 -4

0

50

100 150 Distance, nm

6 4 2 0

200

0

20

40

60

80

100 120 140 160

Depth, nm

Figure 21.6 Micromechanical analysis (MMA) results on the binary brush in toluene (solvated glassy state). Left: typical force-distance curve (gray is approach; black is retracting curve). Right: depth profiling of the elastic modulus at a relatively high indentation force.

indentations, the effect of the collapsed PSF clusters is felt as the modulus increases over the last 50 nm of compression. Although the elastic modulus increases, the values should be much higher than 8 MPa for the collapsed PSF cluster. This low value is entirely a result of using too soft an AFM tip to probe the relatively stiff PSF. The spring constant of the tip is too low, and it is well known in AFM measurements that the combination of soft tip/stiff surface in force mapping measurements will produce unstable probing or unreasonable values [50,51]. The soft tip was necessary in this case because the solvated PMA is highly compliant; thus, pure quantitative values cannot be achieved for the PSF clusters in terms of the elastic modulus, but results do indicate the expected chain reordering as PMA swells to nearly 8

-4

7

Elastic Modulus, MPa

-6 Deflection, nm

438

-8 -10 -12 -14 -16 -18

5 4 3 2 1

-20 -22

6

0

50

100 150 200 250 300 350 Distance, nm

0

0

50

Figure 21.7 Micromechanical analysis (MMA) results on the binary brush in acetone (solvated rubbery state). Force-distance curve and corresponding depth profile reveal a highly non-linear elastic response, indicating vertical segregation of layers with very different elastic response.

100

150

200

Depth, nm

250

300

References

250–275 nm over collapsed PSF. Thus, the total thickness of the binary brush in the solvated rubbery state exceeds 300 nm, compared with 110–150 nm found in the dry state [42]. Therefore, in-situ MMA measurements under solvent conditions have allowed the determination of brush thickness in the solvated state and depth profiling of the elastic modulus, which confirms the vertical layering and depth of the PMA/PSF interface in the mixed brush.

Acknowledgments

This research is supported by the National Science Foundation, CMS-0099868 Grant, Grant M01-C03 from Department of Commerce through National Textile Center, and DFG, Grant SFB 287, B10. The authors thank D. Julthongpiput and S. Peleshanko for stimulating technical discussions and assistance.

References 1 Y. Ikada, Y. Uyama, Lubricating Polymer Sur-

faces, Technomic, Basel, 1993. 2 R. P. Wool, Polymer Interfaces: Structure and Strength, Hanser, Munich, 1995. 3 J. C. Berg, Wettability, Marcel Dekker, New York, 1993. 4 G. J. Fleer, S. M. Cohen, J. Scheutjens, T. Cosgrove, B. Vincent, Polymers at Interfaces, Chapman & Hall, New York, 1993. 5 P. G. de Gennes, Soft Interfaces, Cambridge University Press, Cambridge, 1997. 6 K. Kato, E. Uchida, E. Kang, Y. Uyama, Y. Ikada, Prog. Polym. Sci. 2003, 28, 209. 7 F. Garbassi, M. Morra, E. Occhiello, Polymer Surfaces, Wiley, New York, 1994. 8 (a) R. S. Muller, in: Micro/Nanotribology and Its Applications (Ed.: B. Bhushan), Kluwer Academic Press, Dordrecht, 1997, p. 579; (b) R. S. Muller, in: Tribology Issues and Opportunities in MEMS (Ed.: B. Bhushan), Kluwer Academic Publishers, Dordrecht, 1997; (c) V. V. Tsukruk, in: Nanotribology (Eds.: S. M. Hsu, C. Z. Ying), Kluwer Academic Press, Boston, 2002, p. 347. 9 R. IsraRls, F. A. M. Leermakers, G. J. Fleer, E. B. Zhulina, Macromolecules 1994, 27, 3249.

10 Y. G. Takei, T. Aoki, K. Sanui, N. Ogata,

11 12 13 14 15 16 17 18

19 20

21 22 23

Y. Sakurai, T. Okanao, Macromolecules 1994, 27, 6163. P. Auroy, L. Auvray, L. Leger, Phys. Rev. Lett. 1991, 66, 719. U. Raviv, R. Tadmor, J. Klein, J. Phys. Chem. B 2001, 105, 8125. G. S. Grest, M. Murat, Macromolecules 1993, 26, 3108. S. J. Alexander, J. Phys. 1977, 38, 977. P. G. de Gennes, Macromolecules 1980, 13, 1069. S. T. Milner, Science 1991, 251, 905. J. Klein, Annu. Rev. Mater. Sci. 1996, 26, 581. S. Granick, in: Physics of Polymer Surfaces and Interfaces (Ed.: I. Sanchez), Manning, New York, 1993. L. H. Sperling, Introduction to Physical Polymer Science, Wiley, New York, 2001. A. Halperin, M. Tirrel, T. P. Lodge, Advances in Polymer Science, Springer, 1992, Vol. 100, p. 31. J. Wittmer, A. Johner, J. F. Joanny, Coll. Surf. A: Physicochem. Eng. Aspects 1994, 86, 85. P. Pincus, Macromolecules 1991, 24, 2912. I. Galaev, B. Mattiasson, Trends Biotechnol. 1999, 17, 335.

439

440

21 Local Chain Organization of Switchable Binary Polymer Brushes in Selective Solvents 24 A. Aksay, M. Trau, S. Manne, I. Honma,

N. Yao, L. Zhou, F. Fenter, P. M. Eisenberger, S. M. Gruner, Science 1996, 273, 892 25 Y. Ito, Y. Ochiai, Y. S. Park, Y. Imanishi, J. Am. Chem. Soc. 1997, 119, 1619. 26 L. Leger, E. RaphaRl, H. Hervert, Adv. Polym. Sci. 1999, 138, 185. 27 M. Ruths, D. Johannsmann, J. RShe, W. Knoll, Macromolecules 2000, 33, 3860. 28 P. Mansky, Y. Liu, E. Huang, T. P. Russell, C. J. Hawker, Science 1997, 275, 1458. 29 J. F. Marko, T. A. Witten, Phys. Rev. Lett. 1991, 66, 1541. 30 A. Sidorenko, S. Minko, K. Schenk-Meuser, H. Duschner, M. Stamm, Langmuir 1999, 15, 8349. 31 B. Zhao, W. J. Brittain, Prog. Polym. Sci. 2000, 25, 677. 32 K. G. Soga, M. J. Zuckermann, H. Guo, Macromolecules 1996, 29, 1998. 33 G. Brown, A. Chakrabarti, J. F. Marko, Europhys. Lett. 1994, 25, 239. 34 P. K. Lai, J. Chem. Phys. 1994, 100, 3351. 35 M. MSller, Phys. Rev. E 2002, 65.030802. 36 S. Minko, M. MSller, D. Usov, A. Scholl, C. Froeck, M. Stamm, Phys. Rev. Lett. 2002, 88.035502. 37 S. Minko, D. Usov, E. Goreshnik, M. Stamm, Macromol. Rapid Commun. 2001, 22, 206. 38 V. Koutsos, E. W. van der Vegte, G. Hardziioannou, Macromolecules 1997, 30, 4719. 39 V. Koutsos, E. W. van der Vegte, G. Hardziioannou, Macromolecules 1999, 32, 1233. 40 T. W. Kelley, P. A. Schorr, K. D. Johnson, M. Tirrell, C. D. Frisbie, Macromolecules, 1998, 31, 4297.

41 S. Minko, G. Gafijchuk, A. Sidorenko,

S. Voronov, Macromolecules 1999, 32, 4525. 42 M. C. LeMieux, S. Minko, D. Usov, M.

43

44

45

46 47

48 49

50 51

Stamm, V. V. Tsukruk, Macromolecules, 2003, 19, 7244. (a) V. V. Tsukruk, Rubber Chem. Technol. 1997, 70, 430; (b) V. V. Tsukruk, D. H. Reneker, Polymer 1995, 36, 1791. (a) V. V. Tsukruk, V. V. Gorbunov, Probe Microscopy 2002, 3-4, 241; (b) Z. Huang, S. A. Chizhik, V. V. Gorbunov, J. Mater. Sci. 1998, 33, 4905; (c) V. V. Tsukruk, Z. Huang, Polymer 2000, 41, 5541. M. C. LeMieux, S. Minko, D. Usov, M. Stamm, V. V. Tsukruk, Langmuir, 2003, 15, 6126. D. W. Van Krevelen, Properties of Polymers, Elsevier, Toronto, 1997. J. Brandrup, E. H. Immegurt, E. A. Grulke, Polymer Handbook, 4th ed., John Wiley & Sons, Inc., New York, 1999. D. F. Siqueira, K. KThler, M. Stamm, Langmuir 1995, 11, 3092. A. Karim, V. V. Tsukruk, J. F. Douglas, S. K. Satija, L. J. Fetters, D. H. Reneker, M. D. Foster, J. Phys., II 1995, 5, 1441. V. V. Tsukruk, A. Sidorenko, V. V. Gorbunov, S. A. Chizhik, Langmuir 2001, 17, 6715. V. V. Tsukruk, V. V. Gorbunov, I. Luzinov, Z. Huang, N. Fuchigami, in: Interfacial Properties on the Submicrometer Scale (Eds.: J. Frommer, R. M. Overney), American Chemical Society Symposium Series, Washington, DC, USA, 2001.

441

22

Motion of Nano-Objects Induced by a Switchable Polymer Carpet Svetlana Prokhorova, Alexey Kopyshev, Ayothi Ramakrishnan, and Jrgen Rhe

Abstract

We present a method to transport nano-objects across a surface, utilizing the unique properties of a supporting polymer substrate. Instead of moving individual nanoobjects, we propose to induce motion of many objects. The changes in surface topography are induced by an external stimulus (solvent exposure), and the objects move on the surface in the same way that a “traveling fold” in a carpet carries objects lying on top of it. The particular polymer systems we focus on are polymer brushes; these are built from diblock-copolymer chains covalently attached at high grafting density to a solid substrate. The polymer carpet is supposed to guide the motion of nano-objects, utilizing switchable topographical patterns. As a first example, we demonstrate how an ensemble of silica particles can be raked together on a poly(methyl methacrylate-b-glycidyl methacrylate) diblock-copolymer brush.

22.1

Introduction

The need to build nanomotors or engines, that could actively and independently manipulate small structures has been prompted by the dramatic growth of research on nanotechnology over the past 20 years [1]. One of the challenging steps in building nano-world elements is to establish a nano infrastructure, that could deliver or remove parts of nanomotors into place by assembly lines. For moving nano-objects or even single atoms, there are to date two different approaches. The first one uses a macroscopic device, such as an AFM tip. The problem here is, that for directed transport not only does one have to locate the object first, but also to fix it, and then carefully push it across a surface using either a manual or computer-controlled piezo system. Recently, several groups have shown that it is possible to arrange a small number of clusters or atoms on a flat surface in simple geometrical shapes, such as lines or squares [2]. However, even for these seemingly simple tasks, the procedure is difficult, and carrying it further would require significant advances in automation of the nanomanipulation procedure.

442

22 Motion of Nano-Objects Induced by a Switchable Polymer Carpet

Another approach is to resort to biological pathways to induce movement, and to use highly specific protein complexes such as kinesin/microtubules or myosin/ actin, thus building the basis for contraction-expansion processes in muscle cells, transport of organelles, and segregation of chromosomes during mitosis. These proteins convert chemical energy into mechanical, translational work by the hydrolysis of adenosine 5-triphosphate (ATP). Applications of these biological motors in nano engineering show much promise, as they possess a high efficiency, are small in size, and available in large quantities [3]. However, the specifics of the chemical process – how mechanical work is retrieved from chemically stored energy – are not well understood at this time [4]. Furthermore, one must take into account the fact that protein complexes are fully functional only in aqueous solution, and within a narrow temperature range. This significantly restricts their usefulness for a variety of practical applications, for example, when nanomachines are to be integrated into electronic systems, where no aqueous environment is permitted. In this chapter, we report on a method to transport nano-objects across a surface, utilizing the unique properties of a supporting polymer substrate. We propose to use changes in surface topography to move nanoparticles, in a similar fashion to the way a traveling fold in a carpet carries objects that are lying on top of it. We propose to build the “carpets” from homopolymer-, diblock-copolymer- and mixed brushes, which all consist of polymer chains covalently attached to the surface at a high grafting density. The surface of the carpet systems can be reversibly switched into different morphologies with distinct topographical and chemical characteristics by exposing them to different external conditions, which in turn cause conformational changes of the surface-attached polymer molecules. The essential idea here is that different topological and chemical configurations, together with drastic changes in surface energy and potential landscape, can cause the polymer brushes to “grasp or release” the nano-object periodically, moving it across a surface (Figure 22.1).

(a) Scheme of a diblock copolymer brush carrying a cargo. (b) During shape switching and change in interfacial energy, the “arms” of the brush grasp the nano-cargo and propel it along a surface. This can either lead

Figure 22.1

to random sliding (c1) or directed motion (c2). The latter can either be induced by pre-structuring the surface (c2.1) or by inducing local perturbations caused by a probe, e.g. local heating, light, etc. (c2.2).

22.2 Materials

Polymer brushes form when polymer chains are grafted with one end to a surface at high grafting density, which leads to significant overlap of the neighboring chains. The stretching of the chain in a normal direction away from the surface introduces diverse, interesting properties, in contrast to simply adsorbed polymer chains [5,6]. One particularly interesting system is that of diblock-copolymer brushes, which show peculiar topographical structures that result from the interplay of the tendency of block copolymers to microphase separate, and the restricted motion induced by covalent bonding to the substrate [7,8]. The brushes exhibit complex behaviors that depend on many factors such as overall molecular weight, volume fraction of a single block, flexibility of the chains, grafting density, the FloryHuggins interaction parameter, surface free energy of each block, and environmental conditions. By controlling all of these parameters, it is possible to obtain a variety of topological patterns termed “onion”, “garlic”, “dumbbell”, “flowerlike”, or “checkerboard” structures [9]. Moreover, the patterns introduce not only topological changes but also variations in surfaces energies and chemical potentials. From a theoretical standpoint, the system has been comprehensively analyzed, although experimental investigations of block-copolymer brushes have only been initiated with the introduction of atom transfer radical polymerization (ATRP) [10]. This living polymerization allows easily adjustable molecular parameters and composition of the brush, and possesses the ability to prepare block copolymers by the sequential activation of the dormant chain end in the presence of different monomers. To date, several groups have succeeded in preparing di- and tri-block copolymer brushes [11]. In our laboratory, we are using a similar synthesis approach, the aim being to design a series of diblock copolymer brushes that exhibit different patterned structures.

22.2

Materials

In seeking a means to develop pattern switching as a transport mechanism, we first studied the microphase behavior of diblock copolymer brushes consisting of poly(methyl methacrylate-b-glycidyl methacrylate) (PMMA-b-PGMA) with PMMA as the attached block (Figure 22.2). The synthesis of tethered PMMA-b-PGMA on a silicon substrate was carried out by a sequential process involving ATRP of methyl methacrylate, followed by ATRP of glycidyl methacrylate. In this chapter, we illustrate the concept of transporting objects with just one example among a series of polymer brushes that we have investigated, by varying the molecular weight of the first PMMA block. A comprehensive analysis of all types of brushes has been published elsewhere [12].

443

444

22 Motion of Nano-Objects Induced by a Switchable Polymer Carpet

Figure 22.2

Chemical structure of the diblock copolymer brush PMMA-b-PGMA.

22.3

Results and Discussion

During our studies it was found that, depending on molecular weight, the brushes show different morphologies, ranging from spherical to worm-like, with several intermediate patterns [12]. In the specimen used here (molecular weight of PMMA and PGMA blocks were 8 I 104 g mol–1 and 4 I 104 g mol–1, respectively) the morphology is spherical if the sample is exposed to a “bad” solvent for the second PGMA block. This structure (as any morphology) can be turned into a featureless flat surface by exposing the brush in a chloroform solution that is a “good” solvent for both blocks. This transition is reversible over many cycles (Figure 22.3). Having understood how the formation of nanostructured polymer films can be controlled in a reversible manner, we have in turn investigated the organization of nanoparticles on the polymer surface (Figure 22.4). This study was started with silica nanoparticles with a diameter of 50 nm, which is larger than the average thickness of the brush (30 nm), but corresponds to the characteristic length scale of a pattern, which is the distance between two neighboring constituents (spheres). It

22.3 Results and Discussion

Tapping mode atomic force microscope (AFM) micrographs of p(MMA-b-GMA) diblock copolymer brush. a) The brush forms a spherical pattern, that (b) can be turned into a featureless flat surface upon exposure to chloroform.

Figure 22.3

445

446

22 Motion of Nano-Objects Induced by a Switchable Polymer Carpet

AFM micrographs of the organization process of silica nanoparticles on the diblock copolymer brush p(MMA-b-GMA). (a) Nanoparticles adsorbed from ethanol solution formed no agglomerates on the microphase separated brush; (b) after exposure of the substrate to a chloroform solution (the good solvent), the carpet pattern disappears, whereas some of the particles have started to

Figure 22.4

form islands consisting of two, three or sometimes five objects; (c) restoration of the pattern structure of the brush and further agglomeration into island of the nanoparticles were observed after treatment of the substrate with the toluene solution (bad solvent for the second block). The micrograph in (c) shows the situation after 26 cycles.

was found that, just after adsorption from the ethanol solution, silica nanoparticles showed no aggregation, and were located as single specimens on top of the ordered polymer surface (Figure 22.4a). It was verified that treatment with ethanol did not have any influence on the polymer carpet. Thereafter, we performed a first cycle of ex-situ switching of the structure from spherical to featureless and back, when the substrate was first exposed to a good solvent (chloroform) for the first block and then to a poor solvent (toluene). The formation of small clusters of nanoparticles was observed after only the first cycle of switching [12], and reiteration over more cycles led to the formation of large, elongated aggregates (Figures 22.4(b,c)). Figures 22.4(a,c) show snapshots of the carpet in the structured state, and for comparison the carpet is shown after chloroform treatment (Figure 22.4(b)) in the flat state. To check whether the aggregation appeared because of the pattern switching, or for other reasons, we conducted a series of experiments with homopolymer brushes of PMMA and PGMA, with molecular weight and grafting density similar to the diblock copolymer brush. In the case of homopolymer brushes, no aggregation of the nanoparticles was found [12], and additional annealing in the good solvent (chloroform) at room temperature during 24 h did not lead to any alteration of the structure. These results clearly indicated that periodic switching of the structure of the polymer surface induces organization of the nanoparticles, whereas the homopolymer brush does not influence the distribution of the nanoparticles – at least those with a diameter greater than the thickness of the brush [12]. As yet, we do not have any deeper understanding of the physical processes of the motion of the nanoparticles during pattern switching, but we are able to speculate on the basic mechanisms – that is, if the motion is proceeding diffusively, or

22.3 Results and Discussion

whether it is a cooperative process between the object and the carpet, or even an interplay between the two. Our current picture of this phenomenon is closely related to one of the essential transport mechanisms suggested here. In the structured state of a carpet built from diblock-copolymer brushes, the surface energy fluctuates on the scale of the pattern structure, allowing for a “hopping” activity of the grains. Whilst this appears to be a random process on a small scale, on a larger scale the grains tend to be captured by extended regions of surface energy minima, caused by an initial inhomogeneous distribution of nano-objects.

Summary

During our studies, we were able to establish an alternative approach to move nanoparticles using a switchable polymer carpet. We have shown that motion and organization of an ensemble of silica nanoparticles on a diblock-copolymer carpet is possible over large distances. However, in attempting to apply these findings in the context of nano-transport devices, a deeper understanding of several aspects is necessary: .

.

.

A suitable topography and energy landscape of the pattern must be established that causes a nano-object to move. The relationship between the cargo mass/size and surface properties, as well as the molecular-compositional parameters of the brush, must be analyzed. It needs to be elucidated how the direction of motion can be controlled.

Among these questions, the final one is the most challenging. In order to impose a direction, one must introduce asymmetry into the potential energy landscape, and this can be achieved by different means. Examples include: (i) pre-shaping the polymer surface by splitting the surface to which the carpet is attached into channels or “wells” that can be contacted by electrodes; (ii) introducing a pattern of bias potentials by an adjustable chemical protocol, resulting in a gradient (i.e., grafting density) and causing the length of the induced morphology to vary; and (iii) applying local perturbations, for example by probing the surface with heated or charged AFM tips.

447

448

22 Motion of Nano-Objects Induced by a Switchable Polymer Carpet

References 1 K. E. Drexler, Engines of Creation 1986,

http://www.foresight.org/EOC/ 2 (a) J. A. Stroscio, D. M. Eigler, Science 1991, 254, 1319; (b) I.-W. Lyo, P. Avouris, Science 1991, 253, 173; (c) H. Uchida, D. H. Huang, J. Yoshinobu, M. Aono, Surface Science 1993, 287/288, 1056; (d) T. Junno, K. Deppert, L. Montelius, L. Samuelson, Appl. Phys. Lett. 1995, 66, 3627; (e) S. L. Brandow, W. J. Dressick, C. S. Dulcey, T. S. Koloski, L. M. Shirey, J. Schmidt, J. M. Calvert, J. Vac. Sci. Technol. B 1997, 15, 1818; (f) C. Baur, B. C. Gazen, B. E. Koel, T. R. Ramachandran, A. A. G. Requicha, L. Zini, J. Vac. Sci. Technol. B 1997, 15, 1577; (g) D. M. Schaefer, R. Reifenberger, A. Patil, R. P. Andres, Appl. Phys. Lett. 1995, 66, 1012; (h) P. Sheehan, C. M. Lieber, Science 1996, 272, 1158; (i) S. C. Hsieh, S. Meltzer, C. R. C. Wang, A. A. G. Requicha, M. E. Thompson, B. E. Koel, J. Phys. Chem. B 2002, 106, 231; (j) P. Avouris, T. Hertel, R. Martel, T. Schmidt, H. R. Shea, R. E. Walkup, Appl. Surf. Sci. 1999, 141, 201; (k) D. M. Kolb, R. Ullmann, T. Will, Science 1997, 275, 1097; (l) M. B. Mohamed, V. Volkov, S. Link, M. A. El-Sayed, Chem. Phys. Lett. 2000, 317, 517. 3 (a) H. Hess, J. Clemmens, D. Qin, J. Howaed, V. Vogel, Nano Lett. 2001, 1, 235; (b) J. T. Yang, W. M. Saxton, R. J. Stewart, E. C. Raff, L. S. B. Goldstein, Science 1990, 249, 42; (c) D. L. Coy, M. Wagenbach, J. Howard, J. Biol. Chem. 1999, 274, 3667; (d) R. J. Stewart, J. P. Thaler, L. S. B. Goldstein, Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5209. 4 (a) M. O. Magnasco, Phys. Rev. Lett. 1994, 72, 2656; (b) T. Duke, T. E. Holy, S. Leibler, Phys. Rev. Lett. 1995, 74, 330; (c) G. N. Straropoulos, T. E. Dialynas, G. P. Tsironis, Phys. Lett. A 1999, 252, 151.

5 (a) P. G. de Gennes, Adv. Colloid Interface Sci.

1987, 27, 189; (b) S. J. Alexander, Phys. (Paris) 1977, 38, 983; (c) S. T. Milner, Europhys. Lett. 1988, 7, 695; (d) E. B. Zhulina, O. V. Borisov, V. A. Pryamitsyn, T. M. Birshtein, Macromolecules 1991, 24, 140; (e) S. T. Milner, T. A. Witten, M. E. Cates, Macromolecules 1989, 22, 853; (f) J. RNhe, W. Knoll, Supramolecular Polymers, (Ed.) A. Ciferri. Marcel Dekker, New York, Basel, 2000, p. 565– 613. 6 (a) D. R. M. Williams, J. Phys. II 1993, 3, 1313; (b) J. F. Marko, T. A. Witten, Macromolecules 1992, 25, 296. 7 H. Dong, J. F. Marko, T. A. Witten, Macromolecules 1994, 27, 6428. 8 (a) G. Brown, A. Chakrabarti, J. F. Marko, Europhys. Lett. 1994, 25, 239; (b) J. F. Marko, T. A. Witten, Phys. Rev. Lett. 1991, 66, 1541; (c) S. Minko, M. MNller, D. Usov, A. Scholl, C. Froeck, M. Stamm, Phys. Rev. Lett. 2002, 88, 035502-1. 9 E. B. Zhulina, C. Singh, A. C. Balazs, Macromolecules 1996, 29, 6338. 10 (a) B. Zhao, W. J. Brittain, J. Am. Chem. Soc. 1999, 121, 3557; (b) K. Matyjaszewski, P. J. Miller, N. Shukla, B. Immaraporn, A. Gelman, B. B. Luokala, T. M. Siclovan, G. Kickelbick, T. Vallant, H. Hoffmann, T. Pakula, Macromolecules 1999, 32, 8716. 11 (a) S. G. Boyes, W. J. Brittain, X. Weng, S. Z. D. Cheng, Macromolecules 2002, 35, 4960; (b) B. Zhao, W. J. Brittain, W. Zhou, S. Z. D. Cheng, J. Am. Chem. Soc. 2000, 122, 2407. 12 (a) S. A. Prokhorova, A. Kopyshev, A. Ramakrishnan, H. Zhang, J. RNhe, Polymer Preprints 2003, 44(1), 453; (b) S. A. Prokhorova, A. Kopyshev, A. Ramakrishnan, H. Zhang, J. RNhe, Proceeding of SPIE, Nanotechnology 2003, 30, 5118; (c) S. A. Prokhorova, A. Kopyshev, A. Ramakrishnan, H. Zhang, J. RNhe, Nanotechnology 2003, 14, 1098.

449

23

Photochemical Strategies for the Preparation and Microstructuring of Densely Grafted Polymer Brushes on Planar Surfaces Oswald Prucker, Rupert Konradi, Martin Schimmel, Jrg Habicht, In-Jun Park, and J rgen R he

Abstract

In this chapter, we present strategies for the generation of microstructured polymer brushes that are based on different combinations of photolithographic techniques with the concept of surface-initiated free radical polymerization. Positive and negative images of the mask can be generated by choosing appropriate irradiation conditions during the photolithographic process. Chemically patterned layers (i.e., multifunctional structures) can be generated through step and repeat processes. Furthermore, we demonstrate that microstructured polymer brushes are surface architectures that are potentially useful for a variety of applications such as guided cell outgrowth or as a resist material for X-ray lithography or reactive-ion etching.

23.1

Introduction 23.1.1

Topological and Chemical Patterning of Surfaces

The generation of topological microstructures (Figure 23.1(A)) written into thin polymer layers through the use of light is a well-established process for the fabrication of silicon chips and microdevices [1]. Here, polymeric photoresist solutions are spun onto the surfaces of the chip material and illuminated through masks. Photochemical reactions caused by the illumination alter the solubility of polymer in the illuminated areas in such a way that the polymer becomes either more or less soluble in the solvent, which is used in the following development step. An enhanced solubility is achieved either through light-induced cleavage of certain groups in the polymer, which changes the overall polarity of the system, or via light-induced scission of the polymer chains, which leads to a reduction in the molecular weight and thus altered solubility. The opposite case – a decrease in the solubility of the system – is usually achieved through a crosslinking process within the polymer. Through this developing step a relief is generated, which represents either a positive

450

23 Photochemical Strategies for the Preparation and Microstructuring

(A)

Figure 23.1

(B)

Schematic representation of a topological (A) and chemical (B) micropattern.

or a negative image of the mask written into the polymer layer. In a following step, the micropatterned surface is exposed to etching conditions. The polymer located on top of the substrate withstands the etching process (hence the name “photoresist”), and the noncoated areas are etched at a much faster rate than the coated ones. As a result, the relief pattern is transferred into the underlying chip surface. In the last step – the so-called “stripping process” – the polymer remaining on top of the chip is removed or burned away through reactive-ion etching. The polymer pattern is therefore only used in a sacrificial way, and issues such as the durability of the pattern or the layer are unimportant. However, in some cases it is not so much the generation of topological height differences but rather the writing of chemical structures into the surface of a substrate that is of interest. In those cases, it is desirable that the surface possesses chemical (and thus physical) properties which, in certain selected areas of a substrate, are different from those in other areas. It is also desirable that these areas can be easily addressed during the chip production process. Examples of this include DNA- or protein microarrays [2,3], where there is a need to attach different probe molecules to different areas of the substrate. The writing of chemical information into the chip surface is also of utmost interest for the generation of many other chemical or biochemical sensors. The generation of functional pattern on surfaces is a topic that has been under intense investigation for some decades. Several strategies have been developed, some of which are based on so-called self-assembled monolayers [4]. This method of film formation makes use of difunctional molecules in which one function is used for surface attachment (i.e., for the formation of a covalent bond to the surface), while the other function provides the desired functionality. Typical anchor groups are chloro- or alkoxysilane moieties for the attachment to silicon oxide surfaces, or sulfur-containing groups for the modification of gold surfaces. For example, such methods can be used to control the wettability of the underlying surfaces by the generation of a hydrophobic alkylsilane monolayer on top of hydrophilic silicon surfaces. The microstructuring of such layers can be achieved by deep UV (DUV) ablation of the layer in predefined areas through the use of masks. In another general approach, the microstructure can also be generated via the so-called l-contact printing process established by Whitesides and coworkers [5]. The latter approach is typically used for the spatially resolved deposition of thiols onto gold substrates. Due to the fact that the layers generated by any of these methods are very thin (mono-

23.1 Introduction

layers/molecular dimensions), the properties of these patterns are not so much dominated by the topological features on the surfaces but rather can be considered as a chemical pattern (Figure 23.1(B)) written into the surfaces. In general, a more sophisticated pattern can be produced on surfaces if molecules are deposited that carry chemically reactive groups which are then used for further reaction steps. However, as the anchor groups are often highly reactive (e.g., chlorosilane groups for anchoring to silicon surfaces), the number of functionalities that can be tolerated within one molecule is often quite limited. In some cases it was also reported that certain functional groups tend to bind strongly to surface groups, leading to an undesired “upside-down” type of deposition [6–8]. An example of systems that show such a complication are the widely used aminopropyl silanes that can form surface salts between the basic amino groups of the silane and the rather acidic silanol groups at the surface of the substrates. An alternative to the self-assembled monolayer (SAM) approach is to attach polymer molecules to a surface with high spatial resolution. Such architectures are especially interesting as they are potentially useful for (bio)sensor applications in which probe groups are presented at a surface within a polymer layer. In this way, the strictly two-dimensional (2D) presentation of these groups is extended over the entire film thickness of the polymer coating. This would allow the probe molecules to be no longer in a strict 2D arrangement, but a large number of probe molecules per surface area could be attached, following such a “skyscraper” approach. In order to take advantage of all the attached probe molecules, however, the polymer layer must be significantly swollen during exposure to the analyte solutions. It is clear that such a format asks for covalent attachment of the polymer to the surfaces, as the polymer would otherwise be more or less rapidly removed from the surfaces. Based upon these guidelines, we sought novel synthetic strategies that would allow the generation of micropatterns of thin layers (nm to lm thickness) of polymers that are covalently attached to the surfaces of the substrates. Since a wider range of thicknesses can only be realized by using “grafting from” techniques (see Chapter 1) – that is, techniques in which the polymers are grown on the surfaces by means of surface-initiated polymerization – we are especially interested in synthetic pathways whereby this means of layer formation is combined with photolithographic techniques, for example in the sense of using a photochemical initiation of the layer growth. 23.1.2

Photochemical Pathways for Grafting Polymers onto Surfaces: A Literature Survey

The modification of surfaces via photochemical grafting of polymers is by no means a new area of research, and in fact pioneering studies were carried out by Oster and Oster during the late 1950s [9]. These authors established two methods for modifying the surfaces of polyethylene and polyethylene terephthalate via a surface-initiated or surface-bound polymerization of various monomers such as styrene and acrylamide. For both strategies, benzophenone was physically incorporated into the surfaces of the polymeric substrates. In one approach, these substrates were then

451

452

23 Photochemical Strategies for the Preparation and Microstructuring

simply illuminated with UV light in air. During this process, the benzophenone molecules near the surfaces of the substrates were excited to form triplets. These biradicals then abstracted hydrogen atoms from the polymers, thereby creating radical centers along the chains. Some of these radicals recombined to form crosslinks, but others reacted with the oxygen of the surrounding air to form surface-bound peroxide groups. The latter could then be used as initiating sites for a subsequent surface-initiated free radical polymerization. In a second system, Oster and Oster used the radicals created via the reaction of the benzophenone triplets with the polymer chains of the substrate directly for the graft polymerization, simply by illuminating the benzophenone-doped substrates through a thin film of monomer. When using these techniques, there was a gain in the total mass of the film substrate of up to 20%, indicating that indeed substantial grafting had taken place. The latter system was then improved by Tazuke and Kimura [10,11], whose aim was to provide hydrophobic polymers with a hydrophilic coating. They used polyolefin, cellulose and PVC substrates, covered them with a solution of a monomer (e.g., acrylamide), which also contained photosensitizers such as benzophenone, and illuminated the samples from the rear (i.e., through the transparent substrates). The advantage of this system was that modification proceeded faster (or could be triggered by a weaker light source) as less light energy was lost in the monomer solution. The results of these investigations showed that the surfaces of the substrates could indeed be made hydrophilic, in that quite large amounts of the respective polymers could be grafted onto the substrates. The group of RJnby [12] continued research into these systems, and developed a method which deposited monomer and sensitizer from gas phase by using a solvent vapor (acetone or ethanol) as a carrier. By following this route, a further increase in the rate of the modification reaction could be achieved, and the amount of free (i.e., unbound) polymer could also be reduced. Furthermore, the group showed that functional polymer coatings could be generated by using functional monomers that carry epoxy groups. When examining all of these systems, it is clear that they are all motivated from an engineering point of view, and are designed mostly to create hydrophilic surfaces that can be used to improve the appearance of polymeric objects through the application of a paint or other coatings. Knowing this background, the choice of substrates becomes clear, and it is also very clear as to why simple and relatively cheap initiator systems have been used. As another consequence, the main emphasis of the research was aimed at the paintability and wettability of the samples – along with the long-term stability of the layers – and in many cases the results were satisfactory. From an academic point of view, however, the systems were not suited to the study of any influence of surface confinement on the polymerization process, for a number of reasons. For example, the initiation process was certainly not welldefined, and there was no chance to determine important layer features such as the graft density of the resulting layers and the molecular weight of the surface-attached polymers. In our studies [13–15], we concentrate on a somewhat different approach, with the goal of using a photo-initiated “grafting from” process in conjunction with

23.2 General Features of Surface-Initiated Polymerization from Monolayers of Azo Initiators

Schematic description of the “grafting from” system used for thermal and photochemical generation of polymer brushes. The degrafting step can be used to make the polymers available for general analytical techniques of polymer characterization, for example, for the determination of the molecular weight.

Figure 23.2

photolithographical techniques to generate microstructured polymer brushes on planar surfaces. The system we use is directly adapted from earlier investigations made by our group, where monolayers of aliphatic azo compounds were used to grow polymer brushes via a thermally initiated free radical polymerization process [16–18]. The system is depicted schematically in Figure 23.2. Details on the general features of this system will be described in the next section of this chapter, followed by a description of several pathways used to generate patterned polymer brushes and, finally, multifunctional microstructures.

23.2

General Features of Surface-Initiated Polymerization from Monolayers of Azo Initiators

The molecules used to generate the initiator monolayers consist of three basic structural features: an initiator moiety for starting the polymerization reaction; an anchor group for surface attachment; and a cleavable group, which allows for the detachment of the polymer chains formed after polymerization [17,18]. The cleaved-off polymer is then available for standard techniques of polymer characterization, for example, gel-permeation chromatography (GPC).

453

454

23 Photochemical Strategies for the Preparation and Microstructuring

Figure 23.3

Initiator systems used in this study.

The concrete chemistries of some initiators used in our laboratories are displayed in Figure 23.3 [13–15]. In all three cases discussed here, an azo group is used as the initiating unit. Systems 1 and 2 are initiators that closely resemble the structure of azobis-isobutyronitrile (AIBN), one of the most prominent initiators for free radical polymerization reactions. System 3 is a semi-aromatic azo compound which has a higher extinction coefficient. The anchor group of systems 1 and 3 are monochlorosilane moieties which are suitable for the modification of many oxide surfaces such as the oxide layers of silicon wafers, glass, evaporated or sputtered SiO2. System 2 contains a disulfide group for the attachment to gold surfaces. For the AIBN type systems it was found [17,18] that the decomposition kinetics are neither influenced by the additional structural features next to the azo group, nor by the attachment of these layers to a solid surface, and the surface-attached monolayers essentially behave like AIBN. It is a general feature of all the systems that the initiating group is only attached to the surface by one end. This means that, after decomposition, two different kinds of radicals are generated: one that is attached to the surface, and a second one that diffuses out into the monomer solution. Consequently, a fair amount of free (i.e.,

23.3 Photolithographic Procedures for the Generation of Microstructured Polymer Brushes

non-bound) polymer is generated besides the brush. It has been shown, however, that this free polymer can be washed off easily [19]. The polymers deposited on planar substrates are in general of high molecular weight. The amount of surface-attached polymer can be well predicted by standard models for radical polymerization. Depending on the exact details of the polymerization mechanism, the polydispersities are around 1.5 to 2, as long as the monomer conversion remains low [17,18]. It should be noted that the free radical approach offers the possibility to vary independently from each other the graft density and the molecular weight of the polymer brushes [17–19]. The graft density of the chains is a function of the initiator conversion, as more chains are formed at the surface if more initiator molecules are decomposed. The initiator conversion is easily controlled by variation of the polymerization time. Both, the rate of decomposition of the initiator and the radical efficiency (i.e., the number of initiator radicals that actually start a polymerization) are known, and therefore the graft density of the resulting brush can be controlled within values that correspond to an average distance between two chain anchors of about 6 to 2.5 nm. Lower graft densities can be achieved if the initiator is diluted on the surface by means of coimmobilization of inactive silanes or sulfides, or through controlled, partial deactivation of initiator prior to polymerization. The molecular weight of the chains can be influenced by controlling a set of parameters, which are identical to that of free radical polymerizations in solution. In general, the molecular weight can be decreased if the polymerization is carried out at higher temperatures, or if transfer agents are added. The most convenient way, however, to modify the molecular weight of the brushes is simply to vary the monomer concentration used during polymerization [17–19].

23.3

Photolithographic Procedures for the Generation of Microstructured Polymer Brushes on Planar Surfaces 23.3.1

Photoablation of Polymer Brushes 23.3.1.1 Description of the Procedure The first, and most simple, approach [13,14] for the generation of microstructured layers of polymer brushes on surfaces is to deposit a homogeneous layer of the brush (using any procedure reported for the preparation of such architectures), and then to use high-energy irradiation either in combination with masks (deep UV, Xrays, etc.) or by direct writing (e.g., ion beams) to essentially burn off the polymer in the desired areas. A illustration of such an experiment for the case of a maskassisted photoablation is shown in Figure 23.4. In the following, we will first describe experiments that are aimed at understanding and optimizing the parameters involved in the photoevaporation of the polymer. This involves the investigation of the photoablation process itself in reactive and

455

456

23 Photochemical Strategies for the Preparation and Microstructuring

Schematic illustration of the deep UV (DUV) ablation process for the generation of l-structured polymer brushes.

Figure 23.4

inert atmospheres (air and nitrogen, respectively). We will then describe the investigations of X-ray photolithographic techniques and direct electron-beam writing procedures for this purpose. The Ablation Process in Air and Argon Prior to the patterning experiments, the photoablation process (deep UV) of polystyrene brushes was investigated in air and in nitrogen [13]. As a lamp with only a low flux of photons (pen ray, LOT, 17 mA) was used, the ablation process was slow enough to interrupt the photoevaporation after certain periods of illumination and to measure the residual film thickness, for example, by using surface plasmon resonance spectroscopy (SPR; reflectivity curves shown in Figure 23.5(A)). The results from two sets of ablation experiments for polystyrene brushes of comparable film thickness are shown in Figure 23.5(B); these had been prepared via surface-initiated polymerization from monolayers of the azoinitiator 1. From the X-ray photoelectron spectroscopy (XPS) spectra (Figure 23.6) of the films irradiated in air, it can be concluded that strong oxidation processes occur, but these are restricted to more or less the topmost part of the monolayer. However, the similar rates of the film thickness decrease upon irradiation in air and under argon seem to indicate that the rate-limiting step of the ablation reaction is not so much the oxidative degradation of the polymer, but rather the chopping up of the polymer chains into fragments short enough that they exhibit a significant vapor pressure at room temperature. 23.3.1.2

23.3 Photolithographic Procedures for the Generation of Microstructured Polymer Brushes

(A)

(B)

(A) Surface plasmon resonance curves taken after various times of DUV illumination of a polystyrene (PS) brush. (B) Thickness as a function of irradiation time for the photoablation of PS brushes in air and argon. The straight lines indicate the initial ablation rates.

Figure 23.5

X-ray photoelectron spectroscopy (XPS) detail scan of the C(1s) region of a polystyrene (PS) brush during DUV ablation experiments.

Figure 23.6

457

458

23 Photochemical Strategies for the Preparation and Microstructuring

Patterned Polymer Brushes via DUV Ablation through Masks Microstructuring experiments via photoablation of preformed polymer brushes were performed [13] by using simple electron microscopy grids as masks. The illumination times were adjusted based on the findings of the experiments described before, where the illumination was carried out without the use of masks. An optical micrograph obtained from a 83 nm-thick poly(dimethyl acrylamide) (PDMAA) brush deposited onto a silicon wafer using a monolayer of the azo system 1 is shown in Figure 23.7. A copper grid (mesh 400, width of quadratic holes: 63.5 nm) was used as a mask, and irradiation was carried out using a low-pressure Hg lamp (LOT pen ray) operated at 17 mA and placed about 2.5 cm above the sample. The pattern is clearly identified on the micrograph. For a more quantitative analysis, we used imaging SPR to study the residual film thicknesses in the illuminated and shaded areas of the sample after the patterning experiment. An example of two such micrographs obtained by taking images at different angles of incidence is shown in Figure 23.8. In this experiment, a polystyrene (PS) brush grown as described above for the PDMAA sample with an initial film thickness of 34 nm and again a copper grid, were used. The image in Figure 23.8(A) was taken at an angle of incidence of the laser beam of 48.5M. In this micrograph, 23.3.1.3

Optical micrograph of a microstructure generated within a poly(dimethyl acrylamide) (PDMAA) brush via DUV ablation.

Figure 23.7

Images obtained by surface plasmon microscopy from pattern written into a polystyrene (PS) brush via photoablation. Details are described in the text.

Figure 23.8

23.3 Photolithographic Procedures for the Generation of Microstructured Polymer Brushes

the DUV-illuminated areas are dark, indicating the excitation of a surface plasmon. This angle is identical with the angle of the plasmon resonance signal of the bare substrate as determined prior to any surface modification. This result, therefore, indicates that the PS film was completely removed in these areas during the microstructuring experiment. The image in Figure 23.8(B) was taken at an incidence angle of 60.1M, which was also the angle of plasmon resonance of the as-obtained PS film measured after polymer brush deposition and prior to the ablation. This indicates that the brush stayed intact in areas that were covered by the mask during illumination. Note, that all images shown were taken after the initially formed microstructures were again subjected to solvent extraction (methanol for the PDMAA sample, toluene in the case of PS). The fact that the patterns are still intact after such a treatment demonstrates the superior adhesion of the films due to the covalent bond between the polymer and the surface. Any physisorbed film would have been washed away, even if subjected to simple rinsing procedures. We have demonstrated this in a reference experiment in which a spin-cast film of PS of similar thickness was microstructured in the same way as described for the brushes. Patterned Polymer Brushes via X-Ray Exposure In addition to the use of DUV irradiation, we have also studied the behavior of polymer brushes against irradiation with X-ray beams. For these studies, polymethylmethacrylate (PMMA) brushes were used, as this polymer is commonly used for Xray lithography due to its fairly rapid decomposition in the X-ray beam. In one experiment, we used a PMMA brush with an initial film thickness of about 800 nm. A copper grid was again placed on the film as a mask, and the sample was irradiated with an X-ray beam (Cu Ka) for 72 h at a count rate of 1.6 N 106 cps. The resulting dose was thus 3.3 N 1015 eV. It is interesting to note that no pattern was visible on the substrate after irradiation, but one developed quickly after rinsing the sample with toluene. An optical micrograph of the resulting pattern is shown in Figure 23.9. The thickness in the irradiated area was reduced from 800 nm to about 200 nm during the X-ray and subsequent solvent exposures. The experiments show that even a very low dose of radiation leads to a very significant decrease in film thickness. In strong contrast to conventional X-ray resists, the polymer layer can be 23.3.1.4

Optical micrograph obtained from micropattern generated by X-ray irradiation of a polymethylmethacrylate (PMMA) brush through an electron microscopy grid.

Figure 23.9

459

460

23 Photochemical Strategies for the Preparation and Microstructuring

exposed to a good solvent after irradiation, and all parts of the polymer which lost contact with the surface through a chain scission process are washed out. Therefore, if only one chain scission occurs per chain, the film thickness is already reduced to half of the initial value and a strong contrast is obtained. If a film of nonattached polymers would be exposed to the same dose, such a decrease in molecular weight would not lead to any significant differences in the solubility of the polymer between the irradiated and nonirradiated areas. 23.3.1.5 Patterned Polymer Brushes via Direct Writing with an Ion Beam The patterning of polymer brushes without the use of masks is possible via electronbeam or ion-beam direct writing. In a first experiment, a 20 nm-thick PDMAA brush was subjected to a 10 lm-wide ion beam generated by a van der Graaf accelerator. The pattern written into the brush is shown in Figure 23.10. To enhance the contrast of this optical micrograph, the sample was placed in a high-humidity atmosphere, which led to a slight swelling of the PDMAA brush. Again, it is worthwhile noting that the image of this surface microstructure was taken after extraction of the structured brush with a good solvent (methanol).

Figure 23.10 Optical micrograph of a pattern generated within a PDMAA brush via direct ionbeam writing. The width of the vertical line is approximately 10 lm.

23.3.2

Photoablation/Photodecomposition of the Initiator Layer followed by Thermal Polymerization

Even though a direct patterning of polymer brushes via photoablation of a preformed brush is simple to perform, it is also clear that either long irradiation times or high intensities of the high-energy light are needed to obtain the microstructures. This disadvantage can be circumvented if the initiator monolayer is already lithographically structured [13,14]. The initiator layer has a thickness of only 1–2 nm, and ablation will be complete after only very short irradiation times, even for rather weak radiation sources. Here, the microstructuring process can be described by first writing a latent image into the surface; this is subsequently developed to the final pattern by means of a thermally triggered, surface-initiated polymerization from the remaining initiator patches on the surface. A schematic depiction of the entire procedure is shown in Figure 23.11.

23.3 Photolithographic Procedures for the Generation of Microstructured Polymer Brushes

Figure 23.11 Schematic representation of the procedure used for the microstructuring of polymer brushes via photovolatilization of the initiator monolayer through a mask and subsequent deposition of the polymer via thermally initiated polymerization.

An example of microstructures generated in our laboratories via this pathway is shown in Figure 23.12. The optical micrographs shown were obtained by subjecting a monolayer of the azo initiator 1 to DUV light for about 5 min in the presence of a mask [conditions as described for the ablation experiments on preformed brushes; feature sizes of the masks: mesh 200 with 127 lm line spacing for (A) and mesh 400 with 63.5 lm line spacing for (B) and (C) followed by a surface-initiated polymerization of DMAA on these substrates; conditions: DMAA/water (1/2, v/v), 2 h (A) and 17 h (B and C), 60 MC; extraction with MeOH]. In order to quantify the layer thicknesses and step heights achieved via this twostep approach of microstructuring, we investigated the samples using an imaging ellipsometer (iElli2000; nanofilm Technologies, GOttingen, Germany). Figure 23.13 shows three images obtained from the samples from which the optical micrographs (B) and (C) from Figure 23.12 were taken. In these ellipsometry experiments, the images were recorded with analyzer and polarizer settings chosen such that the

(A)

Figure 23.12 Optical micrographs obtained from two samples (PDMAA brushes) that were microstructured via the technique depicted in Figure 23.11. Images were taken from two different samples prepared under the same

(B)

(C)

conditions, but using different masks; images (B) and (C) are from the same sample, but were recorded at different magnifications. Details are given in the text.

461

462

23 Photochemical Strategies for the Preparation and Microstructuring

(A)

(B)

(C)

Figure 23.13 Ellipsometrically recorded images obtained from a microstructured PDMAA brush. Details are given in the text.

intensity of the reflected light was minimized either in the shaded (A) or illuminated (B) areas. By using the polarizer/analyzer settings, the ellipsometric parameters for each could be determined, and the thickness of the polymer layer was calculated based on the assumption of a simple box profile. For the sample shown in Figure 23.13, a layer thickness of 69 nm was obtained, and this was in good agreement with values obtained from samples where a homogeneous layer was grown under identical conditions. Figure 23.13(C) illustrates a three-dimensional impression of this sample. 23.3.3

Patterning by Photopolymerization The Photopolymerization Process Although in the previous sections a variety of techniques has been described for llithographic patterning, the polymer was in all cases grown in a conventional manner. That is, the surface-initiated polymerization reaction was initiated through thermal decomposition, and growth of the polymer through a free radical process. However, it is also possible to trigger the surface-initiated polymerization reaction directly through light exposure if appropriate initiator systems and wavelengths are used [13–15]. Azo compounds, such as those used in this study, exhibit an n,p*-transition when irradiated with UV light with a wavelength of around 350 nm [20–22]. This transition then leads to a decomposition of the compounds via loss of nitrogen, leaving behind two radicals identical to those generated thermally. Hence, these radicals will initiate a polymerization reaction if appropriate monomers are present. The use of AIBN and related aliphatic azo compounds as photoinitiators has been studied in some detail, and the low extinction coefficient of these compounds was found to limit their use for this purpose [22]. Nevertheless, polymers can be generated if illumination times are chosen appropriately. Systems such as the semi-aromatic initiator 3 have a much higher extinction coefficient, and thus are much more sensitive photoinitiators. However, it should be noted, that for growth of the polymer molecules it is not only the extinction coefficient that is an important parameter. Indeed, 23.3.3.1

23.3 Photolithographic Procedures for the Generation of Microstructured Polymer Brushes

(A)

(B)

Figure 23.14 Results obtained from the investigation of the photoinitiated “grafting from” polymerization of styrene on gold using initiator system 2. (A) Thickness as a function of polymerization time; (B) thickness as a function of integral light intensity.

even more important is the so-called quantum efficiency, which describes how many polymer chains are grown per absorbed photon. This parameter is indeed quite high for the azo compounds described here, and values of ~0.5 are observed, though the exact depends on the individual reaction conditions [22]. It should be further noted that a rapid polymerization reaction caused by fast photodecomposition of the initiator is not at all favorable if thicker polymer brushes (d > 20 nm) are desired, as a high concentration of simultaneously growing sites leads to a rapid bimolecular termination, and the molecular weights of the polymers generated at the surface stays low. This would also keep the obtainable film thickness at an inherently low level, even if a high graft density of the chains can be achieved. In order to investigate the photochemically triggered polymerization using monolayers of azo initiators, we performed two sets of experiments [13–15]. In both cases, monolayers based on initiator system 2 were used for the bulk polymerization of styrene on gold. In the first set of experiments the intensity of the UV light was kept constant, and the polymerization time was varied from 0.5 to 12 h. In the second set, the situation was reversed and samples were polymerized for the same period of time (3 h) but at varying light intensities. The variation of the light intensity was achieved by using gray filters that control transmission. The resulting film thicknesses of the samples as determined by SPR and ellipsometry are plotted as a function of either polymerization time or integral light intensity in Figure 23.14. From these studies, it was found that PS brushes with thicknesses of up to about 350 nm can be prepared via photoinitiated polymerization from monolayers of azo initiators.

463

464

23 Photochemical Strategies for the Preparation and Microstructuring

Patterned Polymer Brushes via Photopolymerization through Masks In order to utilize this photochemical “grafting from” process for the generation of a micro-pattern at the surface, we again used contact masks to shade certain areas on the sample during the photopolymerization process [13–15]. A schematic representation of the process is given in Figure 23.15. Again, it was our goal to demonstrate the feasibility of this approach, and accordingly simple copper grids were used as masks and no efforts were made to explore the limits of the system in terms of line width or edge characteristics. Among the many different samples prepared, Figure 23.16 shows micrographs obtained by imaging ellipsometry from a typical sample. This sample was prepared by using initiator system 1 for the surface-initiated polymerization of DMAA on top of a silicon wafer (conditions: DMAA/water (1/4, v/v), 3 h, extraction with methanol). The images displayed in this figure were obtained by intensity nulling for the areas that were either shaded (Figure 23.16(A)) or illuminated (Figure 23.16(B)) during polymerization. Based on the ellipsometric angles determined from the analyzer/polarizer settings found for these images, the thickness of the deposited PDMAA brush was 131 nm. A three-dimensional impression of the pattern is illustrated in Figure 23.16(C). 23.3.3.2

Figure 23.15 Schematic illustration of the process used for the generation of microstructured polymer brushes via photoinitiated “grafting from” polymerization.

(A)

(B)

Figure 23.16 Ellipsometrically recorded images obtained from a microstructured PDMAA brush. The microstructuring was realized via photoinitiated “grafting from” polymerization through a mask (cf. Figure 23.15). Details are given in the text.

(C)

23.4 Multifunctional Patterns

23.4

Multifunctional Patterns

The fact that the polymers forming the pattern are covalently attached to the surface of the substrate offers the unique opportunity to use the unmodified areas of the substrate for further modification, for example, with a second polymer. In principle, any of the processes described above (and combinations thereof) can be utilized for this purpose. In order to demonstrate the principle [13–15] of such “step-and-repeat” combinations for the preparation of multifunctional polymer pattern, we examined the following combination: .

. . .

the preparation of a homogeneous brush via thermally initiated “grafting from” polymerization of styrene; pattern generation via DUV polymer ablation through a copper mask; redeposition of initiator in the ablated areas; and a second thermal “grafting from” polymerization of a mixture of styrene and a fluorescence labeled methacrylate to fill the ablated areas.

The overall scheme of this process is shown in Figure 23.17. The monomer combination for the second polymerization was chosen because a selective deposition of

Figure 23.17 Schematic representation of the step-and-repeat process used for the multifunctional modification of surfaces via the spatially resolved deposition of different polymers. Details are given in the text.

465

466

23 Photochemical Strategies for the Preparation and Microstructuring

Figure 23.18 Fluorescence microscopy image obtained from a sample in which two different polymers were deposited in different areas of a substrate. The structures of the polymers are also shown. Only those areas that carry the fluorescence dye appear bright in the image.

this monomer in the ablated areas can be readily visualized using fluorescence microscopy. The image obtained by this method is shown in Figure 23.18, together with the structures of the polymer brushes deposited in the respective areas on the sample. The areas carrying the fluorescence dye appear bright due to the emission from the dye, whereas the areas carrying the original PS brush appear dark, indicating that little or no polymer was deposited there during the second surface-initiated polymerization.

23.5

Applications of Photostructured Polymer Brushes

Of course, the l-patterned polymer monolayers obtained by surface-initiated growth can be used in the very same way as conventional photoresists are used. In the experiments described in Figure 23.19, we took advantage of the fact that PS monolayers exhibit a rather strong resistance against etching with fluorine plasmas due to the relative large amount of aromatic moieties contained in such films [23]. To generate the pattern, a PS monolayer was initially grown at the surface of a silicon wafer through surface-initiated growth using the initiator system 3 described above. The substrate was shaded in selected areas by using a mask. The l-structured polymer monolayer obtained was exposed to a microwave plasma containing fluorine. In all noncoated areas the silicon (oxide) is etched away quite rapidly by the reactive ion process. The PS monolayer, however, resists the etching quite strongly, so that even during rather short exposure to the plasma strong topological height differ-

23.5 Applications of Photostructured Polymer Brushes

Figure 23.19 Atomic force microscopy (AFM) images obtained from a microstructured PS brush before (A) and after (B) fluorine etching. Details are given in the text.

ences can be obtained. In the example shown in Figure 23.19, a 2.6-lm height difference is etched into the silicon using this process. Chemical microstructures written into the surface of a substrate can also be used to control the adhesion of biological cells [19,24–26]. In the example shown in Figure 23.20, in the first step a PS monolayer is deposited through surface-initiated photopolymerization. Thus, the polymer is grown only in the irradiated areas. The l-structured surface is, in a second step, exposed to a short KOH etch; this removes the initiator in the nonirradiated areas, which have not been coated by polymer. However, under the conditions employed, the polymer monolayer is in the glassy state and does not permit the transport of any base to the substrate surface. After removing the etching solution and washing with pure water, the substrate surface now consists of a clear hydrophilic/hydrophobic pattern. If the substrate is then brought into contact with biological cells dispersed in a growth medium, cell adhesion proteins can adsorb strongly to the hydrophilic part of the surface. The cells follow suit, and thus a pattern is generated, with the cells following exactly the pat-

Figure 23.20 Neuronal cells on a sample carrying a microstructured PS brush. The cells avoid the areas covered by the brush and grow only on the bare SiO2. Large micron-sized topographical features that are also present on this samples do not interfere with cell alignment.

467

468

23 Photochemical Strategies for the Preparation and Microstructuring

tern written into the surfaces and ignoring topological steps on the surface. Thus, polymer brushes may provide a very efficient tool for the micromanipulation of biological cells.

Summary

In this chapter, three pathways for the photochemical synthesis and micropatterning of polymer brushes were described. All routes are based on various combinations of thermal and photochemical surface-initiated polymerization from monolayers of polymerization initiators via a free radical mechanism, in combination with photolithographic patterning techniques using deep UV, near UV, X-rays and electron or ion beams as the energy source. In all cases, we have demonstrated that microstructured polymer monolayers can be obtained, but no effort was made to explore the limits of these pathways with regard to the minimum line width or the obtainable edge characteristics. Depending on the actual process chosen to generate the pattern, both the positive and negative images of the mask can be obtained. If ablation techniques are used to write the microstructure, the shaded areas will still carry the polymer, which provides the negative image of the mask. If the patterns are generated in situ via photoinitiation from the azo monolayers, the opposite situation is established as the polymer grows in the illuminated regions on the sample, yielding a positive image of the mask. Due to the chemical tethering of the chains within the layers, one can therefore use the same resist material and the same chemistry to generate both types of images simply by adjusting the wavelength and process used for patterning. In all cases, it was found that the patterns generated were very stable due to the chemical bonds between the monolayer chains and the surface: even rigorous extraction of the samples with good solvents in a Soxhlet apparatus did not destroy the pattern. A further advantage of the chemical attachment of the polymers to the surface in these patterns is to be seen in the fact that polymers with different chemical compositions and accordingly different chemical and physical properties can be attached to the surfaces in a spatially resolved manner. This opens the way to chemically microstructured surfaces that display functionalities in preselected areas in a somewhat three-dimensional thicker layer above the surface. All other areas can, for example, act as inert barriers separating the functional patches on the surface, or even providing walls in the sense of a micron-sized titer plate.

References

References 1 Introduction to Microlithography, 2. Ed. (Eds.:

2 3 4

5 6 7 8 9 10 11 12 13

L. F. Thompson, C. G. Willson, M. J. Bowden) American Chemical Society, Washington DC, 1994. C. M. Niemeyer, D. Blohm, Angewandte Chemie-International Edition 1999, 38, 2865. E. Southern, K. Mir, M. Shchepinov, Nature Genetics 1999, 21, 5. A. Ulman, An Introduction to Ultrathin Organic Films, Academic Press, New York, 1991. Y. N. Xia, G. M. Whitesides, Angew. Chem. Int. Ed. 1998, 37, 551. H. Ishida, C. H. Chiang, J. L. Koenig, Polymer 1982, 23, 251. S. Naviroj, J. L. Koenig, H. Ishida, J. Macromol. Sc. Phys. 1983, B22, 291. A. M. Zaper, J. L. Koenig, Polymer Comp. 1985, 6, 156. G. Oster, G. K. Oster, H. Moroson, J. Polym. Sci. 1959, 34, 671. S. Tazuke, H. Kimura, J. Polym. Sci. Part C – Polym. Lett. 1978, 16, 497. S. Tazuke, H. Kimura, Makromol. Chem. – Macromol. Chem. Phys. 1978, 179, 2603. B. Ranby, Z. M. Gao, A. Hult, P. Y. Zhang, Acs Symposium Series 1988, 364, 168. G. Tovar, S. Paul, W. Knoll, O. Prucker, J. RUhe, Supramolecular Sci. 1995, 2, 89.

14 O. Prucker, J. Habicht, I. J. Park, J. RUhe,

15 16 17 18 19 20 21 22 23

24

25 26

Materials Sci. Eng. C-Biomimetic and Supramolecular Syst. 1999, 8-9, 291. O. Prucker, M. Schimmel, G. Tovar, W. Knoll, J. RUhe, Adv. Mater. 1998, 10, 1073. O. Prucker, J. RUhe, Langmuir 1998, 14, 6893. O. Prucker, J. RUhe, Macromolecules 1998, 31, 592. O. Prucker, J. RUhe, Macromolecules 1998, 31, 602. J. RUhe, W. Knoll, J. Macromol. Sci. – Polym. Rev. 2002, C42, 91. M. K. Mishra, J. Macromol. Sci. – Rev. Macromol. Chem. Phys. 1982, C22, 409. F. Deschrij, G. Smets, J. Polym. Sci. A 1 – Polym. Chem. 1966, 4, 2201. P. Smith, A. M. Rosenberg, J. Am. Chem. Soc. 1959, 81, 2037. X. H. Chen, L. M. Tolbert, C. L. Henderson, D. W. Hess, J. RUhe, J. Vacuum Sci. Technol. B 2001, 19, 2013. J. RUhe, R. Yano, J. S. Lee, P. KOberle, W. Knoll, A. OffenhVusser, J. Biomater. Sci. -Polym. Ed. 1999, 10, 859. W. Knoll, M. Matsuzawa, A. OffenhVusser, J. RUhe, Isr. J. Chem. 1996, 36, 357. A. OffenhVusser, J. RUhe, W. Knoll, J. Vacuum Sci. Technol. A – Vacuum Surfaces and Films 1995, 13, 2606.

469

471

Index a ABA-type brushes, properties 159 ABC triblock copolymer 57 acetone solution, binary polymer brushes 434–437 activated monomer mechanism 91–93 active sites, solid surfaces 320 addition of persistent radical 53 addition of sacrificial initiator 53–54 adhesion, affected by polymer brushes 337– 341 adhesion force distributions, pNIPAAM brushes 396 adsorption process 73 AFM see atomic force microscopy agglomeration, stabilization against 341–343 AIBN see azobis-isobutyronitrile Alexander model 16–17 aligned polypeptide brush 88 alkaline earth metal ions interaction with PAA gels 258 interaction with PMAA brushes 257–260 alkanethiols 171 alkanethiol-stabilized cold nanoparticles 217–218 alkylsiloxanes 171 alloc see allyoxycarbonyl alloc-amide approach 99 allyoxycarbonyl 99 alternate polyelectrolyte deposition (APD) 193–194 aluminum, interaction with PMAA brushes 263 amine chain-end functionalized polystyrene 320 amine-initiated grafting polymerization 93– 94 amine-initiated polymerization 90–91 analysis, polymer brushes 189–212

anionic chains, polyelectrolyte brushes 233 anomalous small-angle x-ray scattering see ASAXS APD see alternate polyelectrolyte deposition aqueous environments, weak polyelectrolyte brushes 253–265 aqueous polymerization 106 aqueous solution, pNIPAAM polymerization 381–402 ASAXS, spherical polyelectrolyte brushes 239 atomic force microscopy 195 binary polymer brushes 435–436 height images 56 PMMA brushes 277 pNIPAAM 385–386, 389, 393–399 poly(n-isopropyl acrylamide) brushes 376– 377 spherical polyelectrolyte brushes 242 atom transfer radical polymerization 45, 47, 51–86, 106, 151, 153, 274–276, 371 controlled 53–54 pNIPAAM 388 ATRP see atom transfer radical polymerization ATRP initiator 157 aziridines, monomers 183 azobis-isobutyronitrile 133, 156, 454 initiators, synthesis 143 azo free radical initiator 41 azo initiators 156 surface-initiated polymerization 453–455

b BAA 73–77 BAM 198 BCH 215 synthesis 218–219 5-(bicycloheptenyl)-triethoxysilane see BCH

472

Index binary brushes, synthesis 38 binary mixed brushes 403–425 binary polymer brushes, local chain organization 427–440 binary polymer brush layer 38 biological cells, photostructured polymer brushes 467 biosensor applications, polymer brushes 451 biosurfaces 337–340 functionalized 348 block copolymer brushes 58, 151–165 on flat surfaces 54–55 rearrangement 160–164 synthesis 154–159 block copolymers, synthesis 18 block copolymer shells 62–63 blood proteins 5 bottle brush 105–117 synthetic route 108 bovine serum albumin, interaction with spherical polyelectrolyte brushes 244–245 branched architecture, polymers 167 brewster angle microscopy see BAM bridging, ion interaction with PMAA brushes 265 bromoacetic acid see BAA brushes see also polymer brushes block copolymer 58 copolymer 54–55, 151–165 peptide 88–89 polyelectrolyte 231–248 polypeptide 87–103 spherical 58, 61–63 synthesis 35–50 brush thickness 78, 141 determination 73 pH dependence 236 BSA see bovine serum albumin bulk analysis, in-situ investigations 207 bulk polymerization, surface-initiated 386– 390

c carbon black colloidal dispersions, stabilization 342–343 carbon black fillers, modification 63 carbonyl absorption region, PMAA brushes 261 carboxyisopropylacrylamide 339 carboxylate groups, PMAA brushes 256–265 carboxyl terminated poly-(2-vinylpyridine) 410–411

carboxyl terminated polystyrene 410–411 carpets, switchable polymer 441–448 cast films, PMMA brushes 280–281 cationic polyelectrolyte chains, polyelectrolyte brushes 233 cationic surface-initiated polymerization 119–128 cations, interaction with PMAA brushes 254–264 CdS nanoparticles, stabilization 342 cell growth control 339 chain, polymer 39 chain-end functionalized polystyrene 320, 323 chain growth, initition 19 chain length, dry PMMA brushes 279–282 chain propagation 77 chains charged 232 covalently grafting 36 grafted 80 polyelectrolyte 231 tethered 36 chain stretching, PMMA brushes 278 chain transfer polymerization 136 reversible 39 charged chains 232 charge density, polyelectrolyte brushes 297– 299 charged macromolecules, at surfaces 289 charged polymer brushes 289 charged surfaces, spherical polyelectrolyte brushes 242 charge regulation, high-density polymer brushes 282 checkerboard morphology, mixed polymer brushes 407 chemical bonding, polymer brush synthesis 36 chemical grafting 70 chemical patterning, at the nanoscale 403– 425 chemical patterning of surfaces 449–451 chemisorbed species 227 chemisorption 19 2-(4-chlorosulfonylphenyl)ethyltrichlorosilane 274–275 chromatographic separation 346–347 CIPAAM see carboxyisopropylacrylamide classification, ion interaction with PMAA brushes 264–265 cleaved polymer 156

Index coatings 3–6 see also surface coatings; thin films cold nanoparticles 217–218 collapse, PMAA brushes 253, 257–266 colloidal dispersions, stabilization 341–343 colloidal initiator, TEM image 59 colloidal particle, surface-initiated polymerization 205 colloids inorganic 58 multilayered 58–61 compatibilization see stabilization complex formation, PMAA brushes 249–272 computer hard disk 5 confinement, counterions 237 conjugated polymer brushes, synthesis 377– 379 contact angle measurements 196 contact mode AFM, pNIPAAM brushes 397 contour length, spherical polyelectrolyte brushes 234–235 control, polymerization 214 controlled ATRP 53–54 controlled free radical polymerization 152 controlled polymerization techniques 41 controlled radical polymerizations see CRP coordination-insertion mechanism, ringopening polymerization 111 copolymer brushes 54–55, 151–165 diblock 154–158, 441–447 synthesis 154–159 triblock 158–159 copolymerization, high-density polymer brushes 282–283 copolymerization approach 95–98 copolymers, densely grafted 64 copper ions, interaction with PMAA brushes 260–263 core, near-monodisperse 234 core radius, spherical polyelectrolyte brushes 234–235 core-shell colloids 58–61 preparation 60 core-shell particles, stabilization 341–343 counterion condensation 238 counterions 235–237 confinement 237–238 correlation 238 coupling agent-coated particles 227 covalently attached initiators 120 covalently grafting chains, polymer brush synthesis 36 crosslinking, polymers 114

CRP 42–43 CTS see 2-(4chlorosulfonylphenyl)ethyltrichlorosilane 4-cyano-4-(azo-[4¢-cyano-(butyl)pentanoate])(l,10-dimercaptodecyl)pentanoate 144 4-cyano-4-(azo-[4¢-cyano-(butyl)pentanoate])-pentanoic acid (14) 144 cyclic esters 107

d deactivator, persistent radical 53 deep UV ablation, polymer brushes 455–460 degrafted polymers 208 dehydration, ion interaction with PMAA brushes 264–265 delamination 8 dendrimers 170 preparation 171 dendritic macroinitiators 65 densely grafted copolymers 64 densely grafted polymer brushes 449–469 design, photoinitiators 135–137 desorption, film 8 dewetting 8 diastereomers, SAM 137 diblock copolymer brushes 154–158, 441– 447 rearrangement 160–164 diffusion control, initiation 84 dimethyl acrylamide, polymerization 460– 462 dimple morphology, mixed polymer brushes 404–407 dimple phase, binary polymer brushes 430 dip-coating 74 dip-pen nanolithography 381, 383, 396–399 direct ion-beam writing, microstructured polymer brushes 460 disorder, SAM 137 displacement, film 8 disulfide initiator 139 divalent cations, interaction with PMAA brushes 257–263 DNS chip 3 DPN see dip-pen nanolithography dry brushes, PMMA 279–282 dry state analysis, binary polymer brushes 433 DUV see deep UV

e EBCL see electron beam chemical lithography

473

474

Index electromechanical interferometry, PMMA brushes 281 electron beam chemical lithography 132 electronic surfaces, functionalized 350 electron microscopy, X-ray photoemission 414–416 electrostatic interaction 236 ellipsometry 146, 196 multiple-angle nulling 252–253 pNIPAAM 385 end-functionalized polymers 191 entropy, polymer brush formation 297 environment-dependent surface behavior, mixed polymer brushes 403–425 environment-responsive lithography 422 equilibrium polymerization 113 experimental verification, spherical polyelectrolyte brushes 236

f films aging 75 characterization 193–198 destruction 7–8 formation 194 growth 140, 142 organic 129 preparation 70 see also thin films; ultra-thin films film thickness 177 limitations 12 flat surfaces 52–57, 192–193 polymer brushes 189–212 flow behavior, spherical polyelectrolyte brushes 240–242 formation of patterns 200–201 free energy chains 15 polymer brush formation 297 free initiator 120, 156 free-radical photochemical initiators 134 free radical polymerization 70, 152–153 free radical polymerization of styrene 141 free radicals 133–134 FT-IR spectra, PAA brushes 300–302 FT-IR spectroscopy, pNIPAAM 385 FT-RAIRS 146 functional groups 71 polymers 23 functionalized particles, polymer brush 341– 343 functionalized silicon wafer 176

functionalized surfaces 71, 332–336, 345– 350, 450 functional materials, polymer brushes 22–24 functional organic thin films 129

g gas chromatography 216 GC-MS 216 glass, substrate 131 glass transition temperature, polymer films 279–281 glassy state, binary polymer brushes 434 glycidoxypropyl trimethoxysilane 38 mixed polymer brushes 409–411 glycidyl methacrylate 282–283 GMA see glycidyl methacrylate gold nanoparticles stabilization 342 synthesis 222 thiol-stabilized 221 gold substrates, surface-initiated polymerization 383–384 gold surfaces 40, 88, 105 planar 131–132 GPS see glycidoxypropyltrimethoxysilane gradient, of the polymerization initiator 292– 293 grafted chains 80 grafted polymer chains, isolation 220 grafting block copolymers 283 grafting density 41, 83, 142 IB 123 paa brushes 287–315, 308–310 PMMA brushes 278 polymer chains 58 polypeptide brush 88 spherical polyelectrolyte brushes 234–235 grafting-from approach 41–50, 52, 57–63, 129, 173–175, 333–334, 451–453 mixed polymer brushes 411–412 PMAA brushes 251 grafting from grafts 173–175 grafting-off approach 333–334 grafting polymerization 90, 93–94, 274–276, 317 chemical 70 multi-step 172 grafting-to approach 11, 37–41, 88–89, 332– 333 analysis methods 205–208 hyperbranched polymers 170 mixed polymer brushes 409–411 grafting to grafts 172–173

Index gravimetric methods, in-situ investigations 206 growth, termination 113

h halogen exchange, ATRP 77 hexylisocyanate 221 high-density polymer brushes 273–286 highly branched polymers 177 high molecular-weight polymer 114 high solubility, branched polymers 167 homogeneous polymerization 224 homopolymer shells, polar 61–62 hybrid nanoparticles, PtBuA-silica 180 hybrid nanostructures 119–128 hybrids, synthesis 170–172, 175–181 hydrodynamic diameter, polymer chains 124 hydrodynamic volume 180 hydrolysis, PtBA 304–306 hydroxy groups, initiators 110 hydroxyl ligands 221 n-(n-hydroxy-undecyldisulfanyl)-undecanol 145 hyperbranched polymer, synthesis 169 hyperbranched polymer brush synthesis 46 hyperbranched polymers 167–186

photochemical 134 polymerization 292–293, 383 sacrificial 46–47, 53–54 synthesis 144–145 initiator sites 106 inorganic colloids 58 in-situ investigations, particles 206–208 interfacial properties, branched polymers 168 intramolecular transesterification, PHEMA-gPLA 113 intrinsic viscosity, branched polymers 167 inverse-piezoelectric effect 101 investigating polymer brush systems 198– 204 ion beams, microstructured polymer brushes 460 ionic strength, polymer solution 296, 306– 308 isobutylene see IB ITO see indium tin oxide

k kinetics polymer brush formation 317–328 surface-initiated ATRP 374–375

i

l

IB 119, 122–125 indium tin oxide 373 conjugated polymer brushes 377–379 individual spherical brushes 61 infrared monitoring, polymer formation 216–217 infrared spectra, PMAA brushes 260–261 infrared spectroscopy 226 inimers 45, 174, 176, 349 initiating site, concentration 125 initiation efficiency 83 initiator/monomer single molecule see inimers initiator concentration 81 polymerization 226 initiator functionalized surfaces 71 initiator monolayer 27 initiators azo 453–455 covalently attached 120 coverage 54 disulfide 139 free 122–125 hydroxy groups 110 lifetime 192

lactide 105–117 Langmuir-Blodgett technique 6 Langmuir monolayer, synthesis 18 LASIP see living anionic surface-initiated polymerization layer-by-layer technique, polyelectrolyte complexes 265, 268–270 layered profile, binary polymer brushes 429– 430 LBL technique, polyelectrolyte complexes 265, 268–270 LCST see lower critical solution temperature light scattering, in-situ investigations 206 linear polymeric macroinitiators 64–65 linear polymers 177 lithographic imaging 202 lithography dip-pen nano- 381, 383, 396–399 environment-responsive 422 photo- 421 living anionic surface-initiated polymerization 206 living cationic surface-initiated polymerization 119–128 living free radical polymerization 153

475

476

Index living grafting polymerization 95 living polymerization technique, polymer brushes 119 living radical polymerization 274–276 local chain organization, binary polymer brushes 427–440 long-chain macromolecules 190 lower critical solution temperature pNIPAAM 388–389, 391–392 polymer brushes 338–339 LRP see living radical polymerization

m macrodipoles peptides 101 interaction 89 macroinitiators 69–86 ATRP 77 dendritic 65 linear 64–65 silica 125 synthesis 73–77, 121 macroion, correlation 238 macromolecules charged 289 long-chain 190 mark-houwink exponent 180 masks photoablation of polymer brushes 458 photopolymerization 464 materials functional 22–24 reagents 115 surface-initiated polymerization 120–121 membrane surfaces, functionalized 347–349 MeOH see methanol solution MePVP brushes 265–269 metal ions, interaction with PMAA brushes 256–263 metal-mediated polymerization 213 methanol solution, pNIPAAM polymerization 381–402 methyl methacrylate polymerization 106 surface-initiated ATRP 374–375 mica 61–63 micelles, block polymers 238 microactuators, mixed polymer brushes 403 microchannels, tunable 422–423 microelectronics, adhesion effects 340 microfluidic channels, functionalized 349 micromechanical analysis, binary polymer brushes 434, 437–439

micropatterned brushes 26 micropatterning, surfaces 449–451 microscopy atomic force see atomic force microscopy brewster angle see brewster angle microscopy electron see electron microscopy film properties 195–198 in-situ investigations 207 photoemission electron see photoemission electron microcopy scanning electron see SEM scanning probe see SPM X-ray photoemission electron see X-ray photoemission electron microscopy microstructed polymer brushes 24–28 microstructuring, densely grafted polymer brushes 449–469 microvalves, polymer brushes in membranes 348 miscibility, branched polymers 167 mixed polymer brushes 403–425 local chain organization 427–440 MMA see methyl methacrylate; see also micromechanical analysis molecular brushes 63 synthesis 64–65 molecular modeling 147 molecular parameters, polymers 182 molecular weight control 122 molecular weight measurements 147 molecular weights, graft polymer 126 molecules, polymer 1–3 monolayers characterization 137–138 self-assembled 129–130 monomer amount, alloc-amide approach 100 monomer conversion 126 monomer mechanism, activated 91–93 monomers 335 cyclic 183 types 174 monomer solution, preparation 115 monovalent cations, interaction with PMAA brushes 254–257 morphology, mixed polymer brushes 404– 408 multicomponent polymer brushes 201 multifunctional patterns 465–466 multilayer build-up, PMAA brushes 265–270 multilayered core-shell colloids 58–61 multiple-angle nulling ellipsometry, PMAA brushes 252–253

Index multi-stage growth, polymerization 142 multi-step grafting approach 172

n nanocomposite particles, core-shell 342 nanocomposites diameter 225 formulations 229 silica-poly(norbornene) 219–220 nanocontact printing see nCP nanofabrication, polymer brushes 349–350 nanoimprinting 26 nanomorphology, surface 164 nanomotors, switchable polymer carpets 441 nano-objects, motion 441–448 nanoparticles 133 characterization 204–205 hybrids 121 polymer brushes 189–212 silica 119–128, 218 surfaces, characterization 213–230 synthesis 217–218 thiol-stabilized 221 nanopattern formation 161 nanopatterning mixed polymer brushes 403–425 pNIPAAM brushes 396–399 polymerization initiators 383 nanoscale features, polymer brushes 52 nanostructures, hybrid 119–128 NCA-polymerization 90 nickel-mediated 93 N-carboxyanhydride see NCA nCP 132 NCSEOS 215 synthesis 218–219 near-edge x-ray absorption fine structure see NEXAFS near-monodisperse core 234 neuronal cells, photostructured polymer brushes 467 neutral brush regime 290–291 neutral polymer brushes 288 neutron reflectometry, PMMA brushes 281 NEXAFS spectra PAA brushes 303–304 mixed polymer brushes 414 nickel-mediated polymerization 92–93 NIPAAM see N-isopropylacrylamide N-isopropylacrylamide, polymerization 382, 384–400 nitroxide-mediated polymerization reaction scheme see NMP

NMP 44 NMR, in-situ investigations 207 nonfouling biosurfaces 339–340 nonlithographic methods 203–204 5-norbornen-2-yl(ethyl)ethoxydimethylsilane see NCSEOS norbornene coupling agents 226

o OB regime see osmotic brush regime n-octyltrichlorosilane 292–293 one-step reaction, polymerization 23 optical techniques, polymer films 194–195 organic coatings see thin films organic films see thin films organic-inorganic hybrid nanostructures, preparation 119–128 osmotic brush regime 288, 290–291, 309– 310 osmotic measurements 237 OTS see n-octyltrichlorosilane

p P2VP-COOH see carboxyl terminated poly(2-vinylpyridine) PAA interaction with alkaline earth metal ions 258 interaction with silver ions 256 PAA brushes, surface-anchored 287–315 paraffin oil 292 partial electron yield NEXAFS, PAA brushes 303–304 particles characterization 204–205 coated 227 polymer brushes 205–208 particle substrates 192–193 patterning 45, 204 mixed polymer brushes 420–424 nonlithographic 203 polymers 24 surfaces 449–451, 453–464 patterns formation 200–201 multifunctional 465–466 PDI see polydispersity PDMAA see poly(dimethyl acrylamide) PDMS see polydimethylsiloxane PEL brushes see polyelectrolyte brushes PEL-PEL complexes 265–270 PEMA see 1-propoxyethyl methacrylate PEN see poly(ethylene naphthalate)

477

478

Index peptide brushes, preparation 88–95 peptides, density 88 persistent radical 53 PET see poly(ethylene terephthalate) PGMA 73–77 image 75 pH, polymer solution 235, 254, 296 phase behavior, pNIPAAM 390–393 phase diagram, mixed polymer brushes 406– 408 phase segregation 200–201 mixed polymer brushes 412 PHEMA 105–117 PHEMA-g-PLA intramolecular transesterification 113 synthesis 109 photoablation, polymer brushes 455–460 photochemical strategies, densely grafted polymer brushes 449–469 photochemical surface-initiated polymerization 129 photodecomposition, initiators 460–462 photoemission electron microscopy, X-ray 414–416 photo-emulsion polymerization 232 photo-iniferters 135 photoinitiated polymerization 129 photoinitiated radical polymerization mechanisms 133–135 photoinitiators 136 photolithography 132 microstructured polymer brushes 453–464 mixed polymer brushes 421 photopatterning of substrates 132 photopolymerization 71, 462–464 photoresist layer, sacrificial 202 photosensitizers 134–135 photostructured polymer brushes, applications 466–468 physical attachment, polymer brushes 35 physisorbed species 227 physisorption, polymer brush synthesis 36 piezoelectric effect, inverse 101 planar brushes 232 planar gold, substrate 131–132 plasmon absorption, gold 221 PMA see polymethylacrylate PM-IRRAS see polarization modulator infrared reflection adsorption spectroscopy PMMA 106, 250–269, 374–375 graft polymerization 274–275 high-density brushes 277–282 mixed polymer brushes 412–424

PMMA-b-PGMA see poly(methyl methacrylate-b-glycidyl methacrylate) PNIPAM see poly(N-isopropyl acrylamide) PO see paraffin oil polar block copolymer shells 62–63 polar homopolymer shells 61–62 polarization modulator infrared reflection adsorption spectroscopy 195 poly(acrylic acid) see PAA poly(dimethyl acrylamide), photoablation 458–459 poly(ethylene naphthalate) 372, 375–376 poly(ethylene terephthalate) 372, 375–376 poly(glycidyl methacrylate) see PGMA poly(hydroxyethyl methacrylate) see PHEMA poly(methacrylic acid) see PMAA poly(methyl methacrylate) see PMMA poly(methyl methacrylate-b-glycidyl methacrylate) 443–447 poly(NIPAAM) see poly(Nisopropylacrylamide) poly(N-isopropylacrylamide) 338–339, 348 atomic force microscopy 376–377 poly(styrene-co-2,3,4,5,6-pentafluorostyrene) binary polymer brushes 431–432 mixed polymer brushes 412–424 poly(tert-butyl acrylate) 288–313 poly(tetrafluoroethylene) see PTFE poly(triphenylamine acrylate) 373, 377–379 poly(vinylidene difluoride) 337 poly-(2-vinylpyridine), carboxyl terminated 410–411 polydimethylsiloxane 37, 424 polydispersity 106, 224 polydispersity index 275 polyelectrolyte brushes 17, 231–248 spherical 240–246 weak 249–272 polyelectrolyte complexes, surface-attached 265–270 polyelectrolyte multilayers 265–270 polyelectrolyte thin films 289 polymer brushes applications 331–370 atom transfer radical polymerization 51– 86 characterization 193–198, 213–230 charged 289 chemically attached 36 conformation 199–200 conjugated 377–379 definition 51 densely grafted 449–469

Index formation 317–328 generation 21 high-density 273–286 microstructed 24–28 mixed 403–425 models 428–429 molecular 63–65 multicomponent 201 neutral 288 planar 232 regimes 199–200 stimulus-responsive 54, 381–402 surface-grafted hyperbranched 189–212 synthesis 18–22, 33–185, 332–336 taylor-made surfaces 1–31 theory 15–17 thickness 141 polymer brush functionalized particles 341– 343 polymer chains control 124 isolation 220 surface grafting 168 tethered 70 vinyl-terminated 39 polymer films characterization 193–198 formation 194 ultra-thin 279–282 polymeric macroinitiators 64–65 polymeric substrates, surface-initiated ATRP 375–377 polymerization 138–146 AlBN-type SAMs 135 amine-initiated 90–91, 93–94 anionic 199 aqueous 106 atom transfer radical 274–276 free radical 70, 152–153 grafting 90, 95 living radical 274–276 mechanisms 133–135, 201–204, 214 metal-mediated 213 nickel-mediated 92 photoinitiated 129 reactions 174 radical 130 ring-opening 105–117 self-condensing 175–181 styrene 141 surface-initiated 119–129, 199 systems, well-defined 125 thermal 460–462

uncontrolled 135 polymerization initiators gradient 292–293 attachment 71 nanopatterning 383 polymerization ratio, soluble polymers 179 polymer layer 69 polymers 1–3 degrafted 208 end-functionalized 191 formation 216–217 grafting 317 growth, disulfide initiator 140 hyperbranched 167–186 molecular parameters 182 molecular-weight 114 surface attached 10–13 stability test 220 tethering 87 polymer surface modification 427 polymethylacrylate, binary polymer brushes 431–432 polypeptide brushes aligned 88 synthesis 87–103 polypeptides, rigid rod-like 87 polystyrene carboxyl terminated 410–411 chain-end functionalized 320, 323 polystyrene brushes 78–84, 143 photoablation 457–458 surface coatings 343–345 polystyrene latex 234 polystyrene layer, dewetted 14 polystyrene resins 343 polyterfluorpolyether 5 post-polymerization analysis 209 primary layer, highly reactive 72 primary polymer layer approach 69 propagation reactions, NCA-polymerization 91 1-propoxyethyl methacrylate 282–283 proteins in solution, interaction with spherical polyelectrolyte brushes 243–246 PS see polystyrene PS-COOH see carboxyl terminated polystyrene PSF see poly(styrene-co-2,3,4,5,6pentafluorostyrene) PSF/PMMA mixed brushes 417 PSPS/P2VP mixed brushes 417–424 PSSNa brushes 265–269 PtBA see poly(tert-butyl acrylate)

479

480

Index PtBuA-silica hybrid nanoparticles 180 PTFE 39, 419–420 PTPAA see poly(triphenylamine acrylate) PVDF see poly(vinylidene difluoride) pyrolysis 216, 227

q QCM 196 quartz crystal microbalance see QCM

r radical initiator, azo free 41 radical polymerization 45–46, 130 mechanisms 133–135 radicals free 133–134 persistent 53 radius of gyration 16, 191 RAFT techniques 39–40, 151, 153 reflectance FT-IR spectroscopy, pNIPAAM 385, 387 reflection absorption infrared spectroscopy see FT-RAIRS regimes, polymer brush conformation 199– 200 resins, polymer brush functionalized particles 343 responsive surfaces mixed polymer brushes 403–425 stimulus 345–346, 381–402 temperature 338–339 reversible addition-fragmentation chain transfer see RAFT reversible switching 153 RG see radius of gyration rheology, in-situ investigations 207 RJhe’s initiator 41 ring-opening polymerization 41, 105–117 ripple morphology, mixed polymer brushes 404–407 ripple profile, binary polymer brushes 429– 430 ROP see ring-opening polymerization rubbery state, binary polymer brushes 434

s sacrificial initiator 53–54, free 122–125 sacrificial photoresist layer 202 salt concentration, polyelectrolyte solution 290 salted brush regime 236, 288, 290–291, 308 SAM see self-assembled monolayers

SB regime see salted brush regime scaling theories mixed polymer brushes 405–408 polyelectrolyte brushes 289, 310–314 weak polyelectrolyte brushes in aqueous solutions 257–258 scanning electron microscopy see SEM scanning near-field optical microscopy 207 scanning probe microscopy see SPM SCF see self-consistent field SCVP see self-condensing vinyl (co)polymerization SEC see size-exclusion chromatography segmental adsorption, polymer brush formation 317–328 self-assembled covalently attached initiators 120 self-assembled monolayers 129, 130, 135, 292, 295, 450 ITO surfaces 379 preparation 383 self-condensing vinyl (co)polymerization 175–181 self-consistent field approach, mixed polymer brushes 405–408, 416 self-limiting, brush density 191 SEM 195–196 semi-dilute brushes, PMMA 279 sensors mixed polymer brushes 403, 423 polymer brushes 451 separation of mixtures 346–349 SFA 196 shells block copolymer 62–63 homopolymer 61–62 silane coupling agents, surface-initiated polymerizations 225 silica, substrate 131 silica macroinitiators 125 silica nanoparticles 119–128 synthesis 218 silica-poly(norbornene) nanocomposites, synthesis 219–220 silica-polymer mixture, synthesis 219 silicon surface 131, 171 ATRP 374–375 silicon wafers 176 PMAA brushes 260–261 substrate 72 silver ions interaction with PAA gels 256 interaction with PMAA brushes 256

Index silver nanoparticles 157 SIP see surface-initiated polymerization size distribution, polymers 233 size-exclusion chromatography 96, 294–295 skyscraper approach, chemical patterning of surfaces 451 smart surfaces see stimulus-responsive surfaces SNOM see scanning near-field optical microscopy sodium ions, interaction with PMAA brushes 255–256 soft lithography 9 solid substrates, PAA brushes 287–315 solid supports, hyperbranched polymers 183 solid surfaces, active sites 320 solution 93–94 polymer growth 114 solution grafting 40 solvents tethering reactions 320–322 triblock copolymer brushes 163 solvent-selective ultrathin films 55 solvent state analysis, binary polymer brushes 434–437 SPB see spherical polyelectrolyte brushes spectroscopy FT-IR 385 infrared 146, 226 polymer films 194–195 reflectance FT-IR 385 reflection absorption infrared see FTRAIRS X-ray photoelectron see XPS spherical brushes 58, 61–63 individual 61 polyelectrolyte 231–248 spherical colloidal particle, surface-initiated polymerization 205 SPM 195–196 SPS see surface plasmon resonance spectrometry SRP brushes see stimulus-responsive polymer brushes SST see strong stretching approximation stability test 220 stabilization, colloidal particles 341–343 step-by-step procedure 198–199 steric hindrance, reduction 127 stimulus-responsive surfaces 345–346 stimulus-responsive polymer brushes 54–55, 381–402 streaming potential measurements 196

stripping process, chemical patterning of surfaces 450 strong polyelectrolytes 290 strong stretching approximation, mixed polymer brushes 404–408 strong system, PEL-PEL complexes 265–266 styrene, polymerization 138–143, 141 substrates 131–133 gold 383–384 particle 192–193 photopatterning 132 solid 287–315 substrate surface, synthesis technique 37 surface-anchored PAA brushes 287–315 surface-attached polyelectrolyte complexes 265–270 surface-attached polymers 10 applications 331–370 surface behavior, environment-dependent 403–425 surface coatings 343–345 techniques 6–10 surface coverage determination 73 films 89 surface energy, pNIPAAM brushes 396 surface force apparatus see SFA surface force measurements, pNIPAAM 393–396 surface-grafted hyperbranched polymers 167–186 surface grafting 168 surface hydrolysis, PtBA 304–306 surface-immobilization 155 surface-initiated approach, polymerization 214 surface-initiated atom transfer radical polymerization 275 from polymeric substrates 375–377 kinetics 374–375 surface-initiated polymerization 1, 20–21, 28, 119–129, 191, 199, 225, 333, 371, 373 azo initiators 453–455 bulk 386–390 gold substrates 383–384 surface-initiated ROP 107 surface-initiated vinyl (co)polymerization 175–181 surface plasmon resonance spectrometry 195 surfaces 105–117 control of properties 55 flat 52–57, 192–193 functionalized 38, 71, 345–350

481

482

Index gold 40 indium tin oxide 377–379 investigation 199–200 modification 41 modification and functionalization 332– 336 nanoparticle 213–230 PHEMA 115 planar 171 polymer brush synthesis 37 stimulus-responsive 345–346 tailor made 1–31 temperature-responsive 338–339 surface-sensitive techniques, summary 197 surface tensiometry 196 swelling behavior PMAA brushes 253, 277–279 weak polyelectrolyte brushes in aqueous environments 253–265 switchable binary polymer brushes 427–440 switchable polymer carpets 441–448 switchable surfaces 345–346 switches, mixed polymer brushes 403 synthesis 2D hybrids 170–171, 175–178 3D hybrids 178–181 block copolymer brushes 151–165 immobilized macroinitiators 121 macroinitiators 73–77 mixed polymer brushes 409–412 PHEMA-g-PLA 109 photoinitiators 135–137 polypeptide brushes 87–103 trichlorosilane ATRP initiator 373 triphenylamine acrylate monomer 373 synthesis techniques see grafting synthetic brushes, physical attachment 35 synthetic routes 179

formation 332–336 organic 4, 129 preparation 70 polyelectrolyte 289 thiol layer, surface-initiated polymerizations 225 thiol-stabilized gold nanoparticles 221 three-dimensional hybrids, synthesis 171, 178–181 three-regime kinetics, polymer brush formation 324 titration curve 235–236 TMEOS 215 TMEOS-coated nanoparticles 218–219 TMEOS silica-polymer mixture 219 toluene solution, binary polymer brushes 434–437 topological patterning of surfaces 449–451 total charge density, polyelectrolyte brushes 297–299 TPAA see triphenylamine acrylate transport mechanisms, switchable polymer carpets 441–447 triblock copolymer brushes 57, 158–159 rearrangement 162 tribology, affected by polymer brushes 341 trichlorosilane ATRP initiator, synthesis 373 trimethylethoxysilane see TMEOS triphenylamine acrylate monomer, synthesis 373 trivalent cations, interaction with PMAA brushes 263 tunable microchannels, mixed polymer brushes 422–423 two-dimensional arrangement, polymer brushes 451 two-dimensional hybrids, synthesis 170–171, 175–178 two-step reaction, polymerization 23

t tapping mode AFM, pNIPAAM brushes 398 temperature-responsive surfaces 338–339 TEMPO (2,2,6,6-tetramethyl-1-piperidyloxy) 191 termination of growth 113 tethered chains, polymer brush synthesis 36, 70 tethering, polymer grafting 87, 318, 320–322, 335 thermal analysis, in-situ investigations 206 thermal polymerization 460–462 thin films coatings 343–345

u ultrathin films 279–282 characterization 193–198 solvent-selective 55 stimuli-responsive 55 surface coatings 343–345 UV ablation, polymer brushes 455–460

v vinyl (co)polymerization 175–181 vinyl monomers 173 vinyl-terminated polymer chain 39 viscosity, sphere suspensions 240–241

Index

w wafer, functionalized 176 weak polyelectrolyte brushes 249–272 weak polyelectrolytes 290 weak system, PEL-PEL complexes 265–266 well-defined polymerization system 125 wettability, of a surface 343–345 wet thickness, PAA brushes 306–310

XPS 146 X-ray exposure, microstructured polymer brushes 459–460 X-ray photoelectron spectroscopy see XPS X-ray photoemission electron microscopy, mixed polymer brushes 414–416

z zeta potential, in-situ investigations 207

x XPEEM see X-ray photoemission electron microscopy

483