Bioalcohol Production: Biochemical Conversion of Lignocellulosic Biomass

Bioalcohol production

ß Woodhead Publishing Limited, 2010

Related titles: Biodiesel science and technology: From soil to oil (ISBN 978-1-84569-591-0) Biodiesel fuel is increasingly being used in diesel supplies worldwide, as it both provides an environmentally preferable option and improves diesel engine performance through greater lubricity. The development and use of appropriate feedstocks to avoid food/fuel concerns, and technological developments directed towards improved quality and capacity, are crucial to the environmental impact of biodiesel production and utilisation and to the future of the industry. This book provides a comprehensive and timely reference on the entire biodiesel production chain, from soil to oil, detailing the range and development of biodiesel feedstocks as well as state-of-the-art biodiesel production technology. Handbook of biofuels production: Processes and technologies (ISBN 978-1-84569-679-5) This book provides a comprehensive and systematic reference on the range of biomass conversion processes and technology. The initial section of the book covers the biofuels production chain and analysis of the environmental, social and economic issues surrounding biofuels production. Sections then follow on the entire range of chemical, biochemical and thermochemical biofuels production routes, with chapters reviewing in detail the development of individual processes, from principles and feedstocks, to batch and continuous processes and technology, and on to modelling and optimisation. Handbook of waste management and co-product recovery in food processing, Volume 2 (ISBN 978-1-84569-391-6) Food processors are under pressure, from both consumers and legislation, to reduce the amount of waste they produce and to consume water and energy more efficiently. Handbook of waste management and co-product recovery in food processing provides essential information about the major issues and technologies involved in waste coproduct valorisation, methods to reduce raw material waste and water and energy consumption, waste reduction in particular industry sectors and end waste management. Chapters in Volume 2 focus on the transformation of food co-products using microorganisms and enzymes, advanced methods to optimise food manufacturing, such as closed-loop factories, non-food uses of food waste co-product and commercialisation issues. Details of these and other Woodhead Publishing materials books can be obtained by: · visiting our web site at www.woodheadpublishing.com · contacting Customer Services (e-mail: [email protected]; fax: +44 (0) 1223 893694; tel.: +44 (0) 1223 891358 ext. 130; address: Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, UK) If you would like to receive information on forthcoming titles, please send your address details to: Francis Dodds (address, tel. and fax as above; email: francis.dodds@woodhead publishing.com). Please confirm which subject areas you are interested in.

ß Woodhead Publishing Limited, 2010

Woodhead Publishing Series in Energy: Number 3

Bioalcohol production Biochemical conversion of lignocellulosic biomass

Edited by Keith Waldron

ß Woodhead Publishing Limited, 2010

Published by Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, UK www.woodheadpublishing.com Woodhead Publishing India Private Limited, G-2, Vardaan House, 7/28 Ansari Road, Daryaganj, New Delhi ± 110002, India www.woodheadpublishingindia.com Published in North America by CRC Press LLC, 6000 Broken Sound Parkway, NW, Suite 300, Boca Raton, FL 33487, USA First published 2010, Woodhead Publishing Limited and CRC Press LLC ß Woodhead Publishing Limited, 2010 The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publishers cannot assume responsibility for the validity of all materials. Neither the authors nor the publishers, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. Library of Congress Cataloging in Publication Data A catalog record for this book is available from the Library of Congress. Woodhead Publishing Limited ISBN 978-1-84569-510-1 (book) Woodhead Publishing Limited ISBN 978-1-84569-961-1 (e-book) CRC Press ISBN 978-1-4398-0171-0 CRC Press order number: N10038 The publishers' policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acid-free and elemental chlorine-free practices. Furthermore, the publishers ensure that the text paper and cover board used have met acceptable environmental accreditation standards. Typeset by Godiva Publishing Services Limited, Coventry, West Midlands, UK Printed by TJ International Limited, Padstow, Cornwall, UK

ß Woodhead Publishing Limited, 2010

Contents

Contributor contact details

xiii

Woodhead Publishing Series in Energy

xvii

Preface

xix

Part I Pretreatment and fractionation processes for lignocellulose-to-bioalcohol production 1

Hydrothermal pretreatment of lignocellulosic biomass

3

1.1 1.2 1.3 1.4 1.5 1.6

Introduction Physical comminution Hydrothermal pretreatment (liquid hot water and steam) Conclusions Future trends References

3 4 5 16 17 18

2

Thermochemical pretreatment of lignocellulosic biomass

S. E W A N I C K and R. B U R A , University of Washington, USA

S. P. S. C H U N D A W A T , V. BALAN, L. DA C O S T A S O U S A and B. E. DALE, Michigan State University, USA

2.1 2.2 2.3 2.4 2.5 2.6

Introduction Why is pretreatment necessary for lignocellulosics? Types of chemical pretreatment Comparing effectiveness of leading pretreatments on corn stover and poplar Characteristics of an ideal pretreatment Conclusions

ß Woodhead Publishing Limited, 2010

24

24 25 32 41 47 58

vi

Contents

2.7 2.8

Acknowledgements References

3

Key features of pretreated lignocellulosic biomass solids and their impact on hydrolysis

R. K U M A R , Zymetis Inc., USA and C. E. W Y M A N , University of California, USA 3.1 3.2

3.5 3.6 3.7 3.8

Introduction Key substrate features controlling cellulose hydrolysis: crystallinity Key substrate features controlling cellulose hydrolysis: degree of polymerization (DP) Key substrate features controlling cellulose hydrolysis: hemicellulose and degree of hemicellulose acetylation Key substrate features controlling cellulose hydrolysis: lignin Conclusions Acknowledgements References

4

Solvent fractionation of lignocellulosic biomass

3.3 3.4

4.1 4.2 4.3 4.4 4.5 4.6

N. S A T H I T S U K S A N O H , Z. ZHU, J. R O L L I N and Y.-H. P. Z H A N G , Virginia Polytechnic Institute and State University, USA Introduction Lignocellulosic biomass Cellulose solvent-based lignocellulose pretreatment Future trends Sources of further information and advice References

59 60

73

73 75 84 88 91 97 101 101

122

122 123 128 135 136 136

Part II Hydrolysis (saccharification) processes for lignocellulose-to-bioalcohol production 5

Dilute and concentrated acid hydrolysis of lignocellulosic biomass

A. S H A H B A Z I and B. Z H A N G , North Carolina Agricultural and Technical State University, USA 5.1 5.2 5.3

Introduction Dilute acid hydrolysis Concentrated acid hydrolysis

ß Woodhead Publishing Limited, 2010

143

143 143 144

Contents

vii

5.4 5.5 5.6 5.7 5.8 5.9

Process and apparatus of acid pretreatment Ethanol production plants currently using acid hydrolysis Unit operations pertinent to the ethanol industry Future trends Sources of further information and advice References and further reading

145 149 152 155 156 157

6

Enzymatic hydrolysis of lignocellulosic biomass

159

6.1 6.2 6.3 6.4 6.5 6.6 6.7

Introduction Enzymatic hydrolysis mechanism Relative saccharification efficiencies Factors affecting hydrolysis efficiency Methods to improve enzymatic hydrolysis Future trends References

159 160 163 164 168 170 171

7

Development of cellulases to improve enzymatic hydrolysis of lignocellulosic biomass

178

Introduction Cellulase structure and function Development of cellulases Recent developments Issues in cellulase development Future trends References and further reading

178 179 187 191 192 193 194

M. B A L L E S T E R O S , CIEMAT, Spain

R. J. Q U I N L A N , S. T E T E R and F. X U , Novozymes Inc., USA

7.1 7.2 7.3 7.4 7.5 7.6 7.7

Part III Lignocellulose-to-bioalcohol fermentation and separation processes 8

Integrated hydrolysis, fermentation and co-fermentation of lignocellulosic biomass

P. M A N Z A N A R E S , CIEMAT, Spain 8.1 8.2 8.3 8.4

Introduction Biological processing of lignocellulose Feedstocks and pretreatments for simultaneous saccharification and fermentation (SSF)/consolidated bioprocessing (CBP) Microbial strains for simultaneous saccharification and fermentation (SSF)/consolidated bioprocessing (CBP)

ß Woodhead Publishing Limited, 2010

205 205 206 210 213

viii

Contents

8.5 8.6

Future trends References

9

Challenges in co-fermentation of lignocellulosederived sugars using baker's yeast

224

9.1 9.2 9.3 9.4 9.5 9.6

Introduction Transporter preferences Combining recombinant pathways Transcriptional regulation in mixed substrate fermentations Conclusion and future trends References

224 225 230 232 238 239

10

Separation and purification processes for lignocellulose-to-bioalcohol production

218 219

D. R U N Q U I S T , N. S. P A R A C H I N and B. H A H N - H AÈ G E R D A L , Lund University, Sweden

H.-J. H U A N G , S. R A M A S W A M Y and U. W. T S C H I R N E R , University of Minnesota, USA and B. V. R A M A R A O , State University of New York, USA 10.1 10.2 10.3 10.4 10.5 10.6 10.7 10.8

Introduction Azeotropic distillation (AD) Extractive distillation (ED) Extractive fermentation Separation by adsorption Membrane separation Conclusions References

246

246 249 250 255 258 261 269 269

Part IV Monitoring and modelling processes in lignocellulose-to-bioalcohol production 11

Analytical monitoring of pretreatment and hydrolysis processes in lignocellulose-tobioalcohol production

C. B E C K E R , L. N. S H A R M A and C. K. C H A M B L I S S , Baylor University, USA 11.1 11.2 11.3

Introduction Target analytes resulting from pretreatment and hydrolysis processes Detection strategies

ß Woodhead Publishing Limited, 2010

281

281 283 285

Contents 11.4 11.5 11.6 11.7 11.8 11.9

12

Preparation of biomass hydrolysates for analytical characterization Analysis of carbohydrates Analysis of lignocellulosic degradation products Alternative techniques for analysis of carbohydrates and degradation products Conclusion and future trends References

300 301 302

Online monitoring of fermentation processes in lignocellulose-to-bioalcohol production

315

A. E L I A S S O N L A N T Z and K. V. G E R N A E Y , Technical University of Denmark, Denmark and C. J. F R A N Z EÂ N and L. O L S S O N , Chalmers University of Technology, Sweden 12.1 12.2 12.3 12.4 12.5 12.6 12.7 12.8

13

Introduction Variables of interest to monitor in bioethanol production processes Sampling issues and overview of potential monitoring techniques Chromatographic techniques Spectroscopic methods Software sensors Conclusions References

Modelling hydrolysis and fermentation processes in lignocellulose-to-bioalcohol production

T. T S O U T S O S , Technical University of Crete, Greece 13.1 13.2 13.3 13.4 13.5 13.6 13.7 13.8 13.9 13.10

ix

Introduction Saccharification of lignocellulosic biomass by chemical/ enzymatic processes Fermentation by various microorganisms Simultaneous saccharification and fermentation (SSF) Environmental issues Successful examples Future trends Sources of further information and advice Acknowledgements References

ß Woodhead Publishing Limited, 2010

287 288 296

315 318 319 322 324 330 333 334

340 340 342 350 354 355 356 357 358 359 359

x

Contents

Part V Life cycle assessment of, and multiple products from, lignocellulose-to-bioalcohol production 14

Environmental life cycle assessment of lignocellulose-to-bioalcohol production

Y. Z H A N G , J. M C K E C H N I E and H. L. M A C L E A N , University of Toronto, Canada and S. S P A T A R I , Drexel University, USA 14.1 14.2 14.3 14.4 14.5 14.6 14.7 14.8

15

Introduction Life cycle assessment (LCA) of biofuels Life cycle assessment (LCA) of biochemical lignocellulosic alcohol production Comparison of lignocellulosic alcohol biofuel life cycle assessments (LCAs) with those of other fuels Comparison of life cycle studies of lignocellulosic bioalcohols with those of alternative biomass utilization Routes for environmental improvement Future trends References

Chemical production from lignocellulosic biomass: thermochemical, sugar and carboxylate platforms

A. D. S M I T H , M. L A N D O L L , M. F A L L S and M. T. H O L T Z A P P L E , Texas A&M University, USA

365

365 367 372 377 381 383 385 387

391

15.1 15.2 15.3 15.4 15.5 15.6 15.7 15.8

Introduction Lignocellulose feedstocks Thermochemical platform Sugar platform Carboxylate platform Conclusions Sources of further information and advice References

391 392 393 400 405 410 411 411

16

Production of longer-chain alcohols from lignocellulosic biomass: butanol, isopropanol and 2,3-butanediol

415

Introduction

415

A. M. L OÂ P E Z C O N T R E R A S , W. K U I T , M. A. J. S I E M E R I N K , S. W. M. K E N G E N , J. S P R I N G E R and P. A. M. C L A A S S E N , Wageningen University and Research Centre, The Netherlands 16.1

ß Woodhead Publishing Limited, 2010

16.2 16.3 16.4 16.5 16.6 16.7 16.8 16.9 16.10 16.11

Contents

xi

Characteristics and uses of butanol, acetone, isopropanol and 2,3-butanediol Production of butanol and isopropanol by clostridia Advances in the production of butanol and isopropanol Methods for biological production of 2,3-butanediol Advances in the production of 2,3-butanediol Other long chain alcohols that can be produced biologically Future trends Sources of further information and advice Acknowledgements References

416 420 429 441 442 445 447 448 449 450

Index

461

ß Woodhead Publishing Limited, 2010

Contributor contact details

Lansing, MI 48910 USA E-mail: [email protected]

(* = main contact)

Editor K. W. Waldron Institute of Food Research Norwich Research Park Colney Norwich NR4 7UA UK E-mail: [email protected]

Chapter 1 R. Bura College of Forest Resources University of Washington Bloedel Hall 344 Box 352100 Seattle, WA 98195-2100 USA E-mail: [email protected]

Chapter 2 S. P. S. Chundawat*, V. Balan, L. da Costa Sousa and B. E. Dale Department of Chemical Engineering & Materials Science Michigan State University Biomass Conversion Research Lab (BCRL) Great Lakes Bioenergy Research Center (GLBRC) 3900 Collins Road

Chapter 3 C. E. Wyman* Center for Environmental Research and Technology and Chemical and Environmental Engineering Department Bourns College of Engineering University of California 1084 Columbia Avenue Riverside, CA 92507 E-mail: [email protected] R. Kumar Zymetis, Inc. College Park, MD USA E-mail: rajeev.kumar.th07@ alum.dartmouth.org

Chapter 4 N. Sathitsuksanoh, Z. Zhu, J. Rollin and Y.-H. P Zhang Biological Systems Engineering Department Virginia Polytechnic Institute and State University 210-A Seitz Hall Blacksburg, VA 24061 USA

ß Woodhead Publishing Limited, 2010

xiv

Contributor contact details

N. Sathitsuksanoh and Y.-H. P. Zhang Institute for Critical Technology and Applied Science (ICTAS) Virginia Polytechnic Institute and State University Blacksburg, VA 24061 USA Y.-H. P. Zhang* DOE BioEnergy Science Center (BESC) Oak Ridge, TN 37831 USA E-mail: [email protected]

Chapter 5 A. Shahbazi* and B. Zhang Biological Engineering North Carolina Agricultural and Technical State University Greensboro, NC 27411 USA E-mail: [email protected]

Chapter 6 M. Ballesteros Biomass Unit Department of Energy CIEMAT Avda. Complutense, 22 28040 Madrid Spain E-mail: [email protected]

Chapter 7 R. J. Quinlan*, S. Teter and F. Xu Novozymes, Inc. 1445 Drew Avenue Davis, CA 95618 USA E-mail: [email protected] [email protected] [email protected]

Chapter 8 P. Manzanares Biomass Unit Renewable Energies Division Energy Department CIEMAT Avda. Complutense 22-Ed 85-T 28040-Madrid Spain E-mail: [email protected]

Chapter 9 D. Runquist, N. S. Parachin and B. Hahn-HaÈgerdal* Department of Applied Microbiology Lund University PO Box 124 SE-22100 Lund Sweden E-mail: [email protected]

Chapter 10 H.-J. Huang, S. Ramaswamy* and U. W. Tschirner Department of Bioproducts and Biosystems Engineering University of Minnesota Kaufert Laboratory 2004 Folwell Avenue Saint Paul, MN 55108 USA E-mail: [email protected] [email protected] [email protected] B.V. Ramarao Paper and Bioprocess Engineering State University of New York 310 Walters Hall Syracuse, NY 13210 USA E-mail: [email protected]

ß Woodhead Publishing Limited, 2010

Contributor contact details

Chapter 11

Chapter 14

C. Kevin Chambliss Department of Chemistry and Biochemistry Baylor University One Bear Place #97348 Waco, TX 76798-7348 USA E-mail: [email protected]

Y. Zhang, J. McKechnie and H. L. MacLean* Department of Civil Engineering University of Toronto 35 St George Street Toronto, ON Canada M5S 1A4 E-mail: [email protected] [email protected] [email protected]

Chapter 12 A. Eliasson Lantz Center for Microbial Biotechnology Department of Systems Biology Technical University of Denmark DK-2800 Kgs. Lyngby Denmark Krist V. Gernaey Department of Chemical and Biochemical Engineering Technical University of Denmark DK-2800 Kgs. Lyngby Denmark Carl Johan FranzeÂn and Lisbeth Olsson* Industrial Biotechnology Department of Chemical and Biological Engineering Chalmers University of Technology SE-412 96 Gothenburg Sweden E-mail: [email protected]

Chapter 13 T. Tsoutsos Department of Environmental Engineering Technical University of Crete University Campus GR 73100 Greece E-mail: [email protected]

xv

S. Spatari Civil, Architectural & Environmental Engineering Drexel University 3141 Chestnut Street Philadelphia, PA 19104 USA E-mail: [email protected]

Chapter 15 A. Smith*, M. Landoll, M. Falls and M. T. Holtzapple Department of Chemical Engineering Texas A&M University 200 Jack Brown Building College Station, TX 77843 USA E-mail: [email protected] [email protected]

Chapter 16 A. M. LoÂpez Contreras*, Wouter Kuit, Jan Springer and Pieternel A. M. Claassen Bioconversion Group Wageningen UR Bornse Weilanden 9 6708 WG Wageningen The Netherlands E-mail: [email protected]

ß Woodhead Publishing Limited, 2010

xvi

Contributor contact details

M. A. J. Siemerink and S. W. M. Kengen Laboratory of Microbiology Wageningen UR Dreijenplein 10 6703 HB Wageningen The Netherlands

ß Woodhead Publishing Limited, 2010

Woodhead Publishing Series in Energy

1 Generating power at high efficiency: Combined cycle technology for sustainable energy production Eric Jeffs 2 Advanced separation techniques for nuclear fuel reprocessing and radioactive waste treatment Edited by Kenneth L. Nash and Gregg J. Lumetta 3 Bioalcohol production: Biochemical conversion of lignocellulosic biomass Edited by K. W. Waldron 4 Understanding and mitigating ageing in nuclear power plants: Materials and operational aspects of plant life management (PLiM) Edited by Philip G. Tipping 5 Advanced power plant materials, design and technology Edited by Dermot Roddy 6 Stand-alone and hybrid wind energy systems: Technology, energy storage and applications Edited by J. K. Kaldellis 7 Biodiesel science and technology: From soil to oil Jan C. J. Bart, Natale Palmeri and Stefano Cavallaro 8 Developments and innovation in carbon dioxide (CO2) capture and storage technology Volume 1: Carbon dioxide (CO2) capture, transport and industrial applications Edited by M. Mercedes Maroto-Valer 9 Geologic repository systems for safe disposal of spent nuclear fuels and radioactive materials Edited by Joonhong Ahn and Mick Apted

ß Woodhead Publishing Limited, 2010

xviii Woodhead Publishing Series in Energy 10 Wind energy systems: Optimising design and construction for safe and reliable operation Edited by John Dalsgaard Sùrensen and Jens Nùrkñr Sùrensen 11 Solid oxide fuel cell technology: Principles, performance and operations Kevin Huang and John Bannister Goodenough 12 Handbook of advanced radioactive waste conditioning technologies Edited by Michael I. Ojovan 13 Nuclear reactor safety systems Edited by Dan Gabriel Cacuci 14 Materials for energy efficiency and thermal comfort in buildings Edited by Matthew R. Hall 15 Handbook of biofuels production: Processes and technology Edited by Rafael Luque, Juan Campelo and James Clark 16 Developments and innovation in carbon dioxide (CO2) capture and storage technology Volume 2: Carbon dioxide (CO2) storage and utilisation Edited by M. Mercedes Maroto-Valer 17 Oxy-fuel combustion for fossil-fuel power plants: Developments and applications for advanced CO2 capture Edited by Ligang Zheng

ß Woodhead Publishing Limited, 2010

Preface

The demand for renewable automotive fuels is increasing. This reflects the need to address challenges associated with climate change, the increasing demand for energy from the rapidly growing global human population, and the continued increase in oil prices. Of particular interest is the potential to exploit lignocellulosic biomass. Lignocellulose, the structural cell-wall component of many plants, is a potential source of fermentable sugars. Under the correct conditions these can be released by acid or enzymatic hydrolysis and then fermented to produce alcohols, including bioethanol and biobutanol. These processes are complex and currently uneconomic. Nevertheless, the potential to exploit lignocellulose, whether from bespoke crops-for-fuel, or from waste streams from food crops (e.g. straw), is becoming more attractive in the light of emerging legislation to replace automotive liquid fossil fuels. As a result, a large body of multidisciplinary research has developed over the last 25 years covering a wide range of topics from plant cell-wall biochemistry through to enzyme biotechnology, fermentation science and alcohol recovery. It is a challenge to comprehend the breadth of such knowledge, and also to understand the impact of modifications at one point of the process chain on activities downstream. The aim of Bioalcohol production: Biochemical conversion of lignocellulosic biomass is to provide a succinct and global overview of research and development in the `lignocellulose-to-bioalcohol' chain. The book comprises five parts: Part I: Pretreatment and fractionation processes for lignocellulose-tobioalcohol production Effective enzymatic hydrolysis of lignocellulose is considerably enhanced by a range of thermophysical pretreatments. It is generally accepted that without such pretreatments, the process of hydrolysis may be too inefficient for commercial exploitation. Chapter 1 provides an overview of the role of hydrothermal pretreatments including steam explosion and liquid hot water processes; Chapter 2 considers the role of thermochemical pretreatments; Chapter 3 evaluates key

ß Woodhead Publishing Limited, 2010

xx

Preface

features of pretreated lignocellulose in relation to hydrolysis and Chapter 4 reviews methods for solvent fractionation of solid and liquid components including lignin. Part II: Hydrolysis (saccharification) processes for lignocellulose-tobioalcohol production Hydrolysis of lignocellulose may be carried out both thermochemically and enzymatically. This section reviews the major advances in both these approaches. Chapter 5 considers the use of dilute and concentrated acid in the hydrolysis of lignocellulose, both in batch and continuous processing. This is followed by two chapters on enzymatic hydrolysis. Chapter 6 is a general review article on the topic, whilst Chapter 7 focuses specifically on the development of cellulases to improve enzymatic hydrolysis. Part III: Lignocellulose-to-bioalcohol fermentation and separation processes This section consists of several review chapters covering fermentation research and development, and approaches for alcohol recovery. Chapter 8 evaluates research to integrate enzymatic hydrolysis with fermentation, whilst Chapter 9 reviews the challenges associated with co-fermentation of both hexose and pentose sugars. Chapter 10 then comprehensively reviews physical approaches to recovering alcohol via distillation, adsorption and membrane separation approaches. Part IV: Monitoring and modelling processes in lignocellulose-to-bioalcohol production Because the whole lignocellulose-to-bioalcohol process is complex, and because different steps can have considerable impacts on downstream activities, monitoring and control are of immense importance. This section provides overviews of monitoring and modelling of bioalcohol production systems: Chapter 11 reviews the monitoring of pretreatment and hydrolysis processes; Chapter 12 considers analytical monitoring of fermentation processes; and Chapter 13 reviews the modelling of both hydrolysis and fermentation processes. Part V: Life cycle assessment of, and multiple products from, lignocelluloseto-bioalcohol production The final part of the book focuses on one of the biggest challenges of all, that of integrating the above technologies in order to produce bioalcohol in a way that is not only optimal economically, but also environmentally. Therefore this section evaluates the potential for producing multiple products from such processes, and takes a life cycle overview. Chapter 14 reviews life cycle assessment of, and multiple products from, lignocellulose exploitation; Chapter 15 explores the potential of alternative chemical production from lignocellulosic biomass, and finally Chapter 16 evaluates the possibilities of creating longer-chain alcohols. Keith Waldron

ß Woodhead Publishing Limited, 2010

1

Hydrothermal pretreatment of lignocellulosic biomass S . E W A N I C K a n d R . B U R A , University of Washington, USA

Abstract: Lignocellulosic biomass has long been recognized as a potential sustainable source of mixed sugars fermentation to biofuels and other biochemicals. The hydrothermal pretreatments (steam explosion and hot water pretreatment) are the most effective pretreatments for a variety of biomass types and have been shown to work effectively at a commercial scale. Here, we consider the technical maturity of the hydrothermal pretreatments by looking at the process history, describing the mode of reactions and analyzing the influence of pretreatment conditions on the physico-chemical properties of pretreated biomass. Finally, we compare the effectiveness of hydrothermal pretreatments and outline the remaining challenges associated with harnessing the pretreatment for production of biochemicals. Key words: hydrothermal pretreatment, steam explosion, liquid hot water pretreatment, lignocellulosic biomass.

1.1

Introduction

Processing of lignocellulosic biomass to ethanol consists of four major unit operations: pretreatment, hydrolysis, fermentation and product separation/ purification. Pretreatment, disruption or fractionation is an important tool in the biomass to ethanol conversion process and is required to alter the structure of lignocellulosic biomass to make cellulose more accessible to the enzymes that convert the carbohydrate polymers into fermentable sugars. The goal is to break the lignin seal and disrupt the crystalline structure of cellulose. Regardless of biomass type, the pretreatment has to separate the biomass into cellulose, hemicellulose and lignin with high recovery of all components in pure form to allow for economical feasibility, i.e., through the separation of individual cells or destructuration of the cell wall to loosen up complexes and allow for further separation of main polymers. It is apparent that an effective pretreatment method should be efficient on different types of lignocellulosic biomass, inexpensive (for both operating and capital costs), require a minimum of pre-pretreatment (preparation/handling) steps and affect a maximum recovery of all lignocellulosic components in usable form. In addition, if ethanol is the final product of biomass to ethanol

ß Woodhead Publishing Limited, 2010

4

Bioalcohol production

conversion, the effective pretreatment should ensure the maximum hemicellulose and cellulose recovery in hydrolyzable and fermentable form. Although many pretreatment processes have been studied (biological, physical, chemical and combination of these approaches) no process currently available can provide all of these desired outcomes on all lignocellulosic materials. However, hydrothermal pretreatment is the most effective pretreatment for a variety of biomass types and has been shown to work effectively at a commercial scale. In this review, we will first analyze the technical maturity of the hydrothermal pretreatment process by looking at the history and description of the process conditions. Then, we will describe the mode of hydrothermal reactions and the influence of pretreatment conditions on the physico-chemical properties of pretreated biomass. The final section offers a comparison of the hydrothermal pretreatments (steam and hydro) and outlines the remaining challenges associated with harnessing the pretreatment for production of biochemicals.

1.2

Physical comminution

Size reduction of lignocellulosic biomass is an important factor in any pretreatment process. Mechanical means can be used to reduce particle size sufficiently so that no further pretreatment is required prior to enzymatic hydrolysis, obviating usage of chemicals and associated concerns such as corrosion, recycling, neutralization and storage. However, high energy requirements for these processes mean that they are typically not economically feasible. As particle size decreases, crystallinity is reduced, which increases enzymatic digestibility. To significantly improve hydrolysis, treatments must reduce particle size to less than 50 m (Datta, 1981). However, due to the high energy cost of mechanical size reduction, comminution past 200 m is not generally economically feasible (Datta, 1981). Milling processes include dry, wet and vibratory ball milling (Millett et al., 1979; Kelsey and Shafizadeh, 1980; Fukazawa et al., 1982). Compression, hammer and disc milling are also used (Schell and Harwood, 1994). These vary widely in terms of particle size distribution, energy usage and efficacy on different feedstock types. For example, far more energy is required to mill hardwoods to a given size than agricultural residues using both hammer and knife milling (Cadoche and Lopez, 1989). In starch-to-ethanol bioconversion, wet and dry milling are the most cost effective pretreatments (Bothast and Schlicher, 2005). However, for lignocellulosic biomass, the energy demands of any physical size reduction process are high. Comminution is consequently limited to pre-pretreatment, to be followed by a chemical or thermal pretreatment process.

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

1.3

5

Hydrothermal pretreatment (liquid hot water and steam)

Physical pretreatment of lignocellulosic biomass is often inadequate in providing complete fractionation to a readily digestible and fermentable product. The cell structure of lignocellulosic biomass is by nature complex and difficult to penetrate, so fractionation requires chemical reactions in addition to physical restructuring. Pretreatments utilizing primarily steam or liquid water at high temperatures can efficiently convert biomass to a form which can be easily digested by enzymes by facilitating autohydrolysis reactions within the biomass. Processes utilizing hot water or steam as the primary chemical are known as hydrothermal pretreatments. The two forms of hydrothermal pretreatment utilize steam (steam explosion) and aqueous water (liquid hot water pretreatment). These processes are advantageous compared to chemical methods as they are regarded as safer ± equipment corrosion is reduced ± and more `environmentally friendly', as often, no chemicals are required (Allen et al., 2001; Laser et al., 2002). Although process residence times and temperatures are similar, liquid hot water pretreatment differs from steam explosion in that water is present as a liquid instead of a gas during pretreatment. As a result, there are differences in reactor configurations, solids consistency during and after pretreatment and concentration of reaction products. In the last 80 years, there has been great progress in the development of aqueous processes to break down all types of lignocellulosic biomass. From agricultural residues to hardwoods to softwoods, hydrothermal pretreatments have the potential to sustainably generate material which can be readily converted to ethanol.

1.3.1

Process history and description

Steam explosion Steam explosion has long been used as a means of deconstructing biomass for many purposes, from structural materials to paper to biochemicals. The first use of steam explosion to produce a commercial product was the masonite process. Developed in the 1920s, the process was used to produce a fiberboard building material (Mason, 1928) with very high yields and minimal energy usage (Overend et al., 1987; Kokta and Ahmed, 1998). However, the coarse, dark substrate, while suitable for fiberboard, was unsuitable for paper products. Asplund used a similar high temperature and pressure process with the addition of mechanical refining to produce fiber for board manufacturing (Asplund, 1953). Refining at high temperatures allowed fibers to be fully separated with very high yields, although there was no delignification. By addition of ammonia and SO2, it was possible to use the Asplund process to produce fibers for papermaking but the cost associated with use and recycling of these chemicals was prohibitive (Kokta and Ahmed, 1998).

ß Woodhead Publishing Limited, 2010

6

Bioalcohol production

Incorporating defibrillation with steam explosion, siropulping was developed as a batch process using a masonite gun equipped with a defibrillation nozzle aimed at a metal grating. To increase the force of biomass contacting the grate, pressure in the gun was increased to 4.8 MPa using CO2. Addition of nitrogen enabled a pressure of up to 13.8 MPa, far higher than the 1.6 MPa obtained using pure steam at 200 ëC (Mamers et al., 1979, 1981). Although costly to run, results from siropulping were promising. Increasing the pressure resulted in increased digestibility of the solids, indicating a mechanical disruption effect. However, attempts to reduce temperature and cooking time without affecting digestibility were unsuccessful, demonstrating the importance of chemical processes in the pretreatment (Puri and Mamers, 1983). The Masonite process was revisited in the 1980s as a means of fractionating lignocellulosic biomass for animal feed and for biofuel production. Process configurations ranged from application of the original batch method by the IOGEN Corporation to a continuous digester developed by STAKE Technology. Both methods are used currently for bioethanol production at pilot scale. At lab scale, steam explosion has been shown to be effective in pretreating a wide variety of biomass types, including agricultural residues (corn fiber, corn stover), hardwoods (poplar, willow) and softwoods (Douglas-fir, pine, spruce) È hgren et al., 2007; Glasser and Wright, 1998; Sassner et al., (Bura et al., 2002; O 2005; Boussaid et al., 2000; Ewanick et al., 2007; SoÈderstrom et al., 2003a). The STAKE continuous process utilizes a coaxial feeder to move a plug of biomass through a steam-injected reactor chamber for a set time period. It then exits the digester through a discharge screw and blow valve into a cyclone (Jollez et al., 1994; Heitz et al., 1991). The IOGEN batch method involves loading of biomass inside a cylindrical digester injected with steam. The temperature is held at a constant value (from 150 to 280 ëC) for anywhere from 10 seconds to 15 minutes. After the time has elapsed, a valve at the bottom of the digester is opened and the biomass released to a cyclonic collection vessel (DeLong, 1981; Foody, 1984). For both batch and continuous processes, as the pressure drops to atmospheric, water and steam inside the biomass expand and cause the `explosion' of the cell structure. Following pretreatment, substrates undergo hydrolysis, fermentation and distillation to ethanol. The IOGEN demonstration plant in Ottawa, ON, Canada is designed to produce 3 million liters of ethanol annually from wheat, barley and oat straw (Galbe et al., 2007). Steam explosion fractionates biomass to yield a liquid fraction rich in monomeric and oligomeric sugars and solid fraction made up of digestible cellulose and lignin. Under optimized conditions, relatively pure products in high yields can be achieved, such as highly digestible cellulose or high yields of solubilized hemicellulosic sugars. Following enzymatic hydrolysis of the solid fraction, the majority of the cellulose is commonly converted to glucose, leaving behind lignin. These solids have a high heating value and can be burned for process energy or converted to pellets, which can be sold to improve process

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

7

Table 1.1 Pretreatment conditions and results for different feedstocks undergoing steam explosion pretreatment. Xylose recovery is determined based on xylose in original material. Ethanol yields are calculated based on the theoretical yield of 100% conversion of all fermentable sugars in the raw biomass Biomass

Conditions

Wheat straw

180 ëC, 10 min, 0.9% w/w H2SO4 Corn stover 190 ëC, 1.5 min, 1% w/w H2SO4 Corn fiber 190 ëC, 5 min, 3% SO2 (w/w) Willow 200 ëC, 4 min, 0.5% H2SO4 (w/w) Poplar 205 ëC, 3 min, 1% SO2 (w/w) Lodgepole 200 ëC, 5 min, pine 4% SO2 (w/w) Spruce 215 ëC, 5 min, 2.4% SO2 (w/w)

Xylose recovery

Ethanol Reference yield (% of theoretical)

85%a

70%

Ballesteros et al. (2006)

90%

85%

Tucker et al. (2003)

50%a

89%

Bura et al. (2003)

72%

79%

Sassner et al. (2008)

65%

64%

De Bari et al. (2007)

73%

77%

Ewanick et al. (2007)

68%

68%

Stenberg et al. (2000)

a

total hemicellulosic sugar recovery

economics (Galbe et al., 2007). The hemicellulosic sugars contained in the water-soluble fraction have value as a fermentation substrate for ethanol production or a starting reactant for other products (Wayman, 1980). The process demonstrates versatility and robustness on many types of biomass. While softwoods were formerly thought to be not suitable for steam explosion (Saddler et al., 1993), it has been shown in recent years that softwoods including spruce and pine can be pretreated using steam explosion to provide high ethanol yields (Table 1.1) (Ewanick et al., 2007; Stenberg et al., 1998; Rudolf et al., 2005). Since steam explosion can be used to pretreat the widest range of lignocellulosic biomass, it has been shown to be a robust process capable of generating a variety of products based on the pretreatment conditions and feedstock chosen. Liquid hot water Liquid hot water was used in the pulp and paper industry as early as the 1930s as an extraction method to remove hemicelluloses from wood prior to pulping (Richter, 1956). This was most commonly used for the production of dissolving pulp, where very pure cellulose (>90%) was desired (Ragauskas et al., 2006). Hot water pretreatment for subsequent enzymatic hydrolysis of cellulose was first developed to provide a carbon source for fermentative protein production (Bobleter and Pape, 1968; Bobleter et al., 1976). The process was further

ß Woodhead Publishing Limited, 2010

8

Bioalcohol production

developed under the names of hydrothermolysis (Bobleter and Pape, 1968; Bobleter et al., 1976), aquasolv (Allen et al., 1996), aqueous/steam aqueous fractionation (Jollez et al., 1994) and uncatalyzed solvolysis (Mok and Antal Jr, 1992). Depending on the circulation direction of hot water relative to the biomass, three different process configurations are commonly employed. In co-current or batch reactors, biomass and water are heated and held at temperature together for the desired time. Counter current reactors move water and biomass in opposite directions, while in flow-through reactors hot water flows over a stationary bed of biomass. In each configuration, the hot water dissolves biomass components, particularly hemicelluloses (Mosier et al., 2005a). Flowthrough and counter current systems generally provide higher hemicellulose sugar yields and cellulose digestibility, while batch systems require less water and energy to operate (Yang and Wyman, 2004). Operating conditions for all configurations range from 140 to 240 ëC for 0±20 minutes (Mok and Antal Jr, 1992). Operating at pilot scale in Denmark, the Integrated Biomass Utilization System (IBUS) is a continuous counter-current liquid hot water pretreatment system currently used to produce ethanol from wheat straw (Larsen et al., 2008). Optimum conditions for maximum yield of ethanol from the solid material were found by Petersen et al. to be 195 ëC for 6±12 minutes. At these conditions, 70% of hemicellulose is recovered, and of the 93±94% of recovered cellulose in the solids, 89% is converted to ethanol. At present, only the pretreated solids are converted to ethanol. The pentose content in the liquid stream is such that a pentose-fermenting organism is required before use of this stream is economical (Petersen et al., 2009). Prior to hydrolysis and fermentation, the pretreated cellulose is washed to remove inhibitors and residual hemicellulosic sugars (Larsen et al., 2008). The IBUS and numerous recent lab and pilot scale hot water studies utilize primarily hardwoods and agricultural residues including wheat straw, aspen, sugar cane bagasse and corn fiber, (Allen et al., 2001; Laser et al., 2002; Van Walsum et al., 1996; Mosier et al., 2005b).

1.3.2

Feedstock characteristics

Pretreatment conditions and raw biomass species greatly affect the final product of pretreatment. However, other factors play an important role as well, though they are much less well understood. Within a given type of biomass, differences include seasonal changes in chemical composition, ash content, and age. For wood, younger trees are more easily fractionated than older ones (Saddler et al., 1993; Ramos et al., 1992b), as is material derived from the more permeable sapwood compared to denser heartwood (Brownell and Saddler, 1987). Particle size, as well as the timing of harvest and storage prior to pretreatment, can have a major role in determining the efficacy of pretreatment. The moisture content in

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

9

any type of biomass is naturally highest immediately after harvest and can dramatically decrease during storage. Particle moisture content affects both steam consumption and pretreatment efficacy. Raw biomass at a high moisture content (above 50%) requires a longer reaction time in order to heat the additional water inside the cells, and consumes up to 50% more steam compared to air dried biomass (Brownell et al., 1986). This is especially evident when particle sizes are large. Additionally, the outside of the particle may cook faster than the inside. As particle sizes decreases, pretreatment conditions may become too severe as the biomass heats quickly without the buffering effect of a slower heating, increasing hemicellulose degradation (Cullis et al., 2004; Ballesteros et al., 2000). The `pre-pretreatment' of raw biomass (storage conditions, moisture content and particle size) is important to the final product of pretreatment, and becomes increasingly important in commercial processes where large amounts of biomass are required. Further research in this area is required in order to determine the optimum storage and pre-pretreatment conditions to maximize yields for a given feedstock.

1.3.3

Method of action

Hydrothermal (steam explosion and liquid hot water) pretreatments utilize acid liberated from hemicellulose side chains and high temperatures to hydrolyze hemicellulose, cellulose and lignin. Some biomass requires the addition of mineral acids (SO2 or H2SO4) to achieve the same level of pretreatment, but the method of action in both cases is similar. In both catalyzed and autohydrolysis chemical reactions, physical rearrangements govern the breakdown of biomass from polymers to oligomers to monomers. In the context of bioethanol production, the benefit of any pretreatment is increased cellulose surface area available to cellulytic enzymes. This occurs by dissociation of lignin and hemicellulose from cellulose and reductions in crystallinity and particle size. Enzyme accessibility has been extensively studied and is governed by pore size, surface area, and a number of other factors that will not be covered in this chapter (Wong et al., 1988; Mooney et al., 1999; Saddler et al., 1999; Chandra et al., 2007). Hydrothermal reactions Biomass added to the steam or hot water reactor first undergoes hydrolysis of hemicellulose, the most labile of the three primary components of lignocellulose (Fengel and Wegener, 1989). Saturated steam or hot water condenses water present in the cells of the biomass. As a result, organic acids (acetic acid and uronic acid) are liberated by saponification of hemicellulosic uronic and acetyl groups. Water itself acts as an acid at high temperatures; at 220 ëC, water has a pH of 5.6 and an ion product of 10ÿ11 (Marshall and Franck, 1981). In fact, at

ß Woodhead Publishing Limited, 2010

10

Bioalcohol production

these conditions, it is thought that water has a greater effect than acetic or formic acid (Lynd, 1996). Liberated acids hydrolyze the hemicellulose glycosidic bonds to form low to intermediate weight water soluble oligomers. At increased residence time and temperature, these oligomers undergo further hydrolysis to yield monomers. Monosaccharides, particularly pentoses, are highly unstable in high temperature, acidic environments. Under these conditions, pentoses and hexoses undergo dehydration reactions to form furfural, hydroxymethylfurfural, levulinic acid, formic acid and other degradation products known to be inhibitory to fermentation (Ramos, 2003) and enzymatic hydrolysis (Cantarella et al., 2004). Fortunately, the rate of formation of these compounds is slower than the depolymerization of hemicellulose, and can often be controlled by use of lower temperature and shorter residence time (Stenberg et al., 1998) which preserve oligomeric sugars and prevent depolymerization to monomers (Laser et al., 2002). Liquid hot water pretreatments control the degradation of oligomers by maintaining a pH close to 4 (Mosier et al., 2005b). Removal of sugars as they are solubilized by means of a flow-through or countercurrent liquid hot water process also controls the amount of degradation. In addition to hemicellulose, cellulose is also modified during hydrothermal pretreatment. Glycosidic bonds are hydrolyzed, albeit at a lower rate than hemicellulosic acidolysis. As cellulose molecules are randomly hydrolyzed, the degree of polymerization (DP) decreases (Puls et al., 1985). As hemicellulose and cellulose are degraded, lignin is depolymerized simultaneously, though at a somewhat slower rate. Acid hydrolysis of primarily -O-4 ether bonds gives rise to a high free phenolic count in the lignin (Li et al., 2007; Shevchenko et al., 2001; Wood and Saddler, 1988). Almost immediately after acidolysis, the lignin repolymerizes by acid catalyzed condensation between the aromatic C6 or C5 and a carbonium ion (Li et al., 2007). While a small portion of the newly condensed lignin is soluble in the aqueous media, the majority is hydrophobic and migrates to the middle lamella and lumen. At higher temperatures, lignin becomes highly fluid and forms spherical droplets visible in the lumen (Michalowicz et al., 1991) and on the surface of fibers (Kristensen et al., 2008). The full effect of the chemical reactions and rearrangements that occur during hydrothermal pretreatment depends on the severity of pretreatment as well as the original feedstock characteristics. The resulting unique physical and chemical characteristics of pretreated biomass determine enzymatic digestibility, ease of fermentation and subsequent ethanol yields.

1.3.4

Pretreatment severity

Severity is determined by the relationship between three factors: residence time, temperature and acid concentration. Time and temperature are easily measured, whereas the acid concentration is much more difficult to quantify. In auto-

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

11

hydrolysis reactions with no added acid, many researchers make use of a severity factor calculated using time and temperature to determine relative severity of different combinations (Equation 1.1) (Overend et al., 1987). This severity factor was developed based on the assumption of first order kinetics and on earlier factors such as the H-factor (Vroom, 1957) and P factor (Brasch and Free, 1965).   1:1 log …R0 † ˆ log t
exp …TÿTref †=14:75 R0 is the severity parameter, time is t, Tref is a reference temperature for a base case (often 100 ëC) and T is the reaction temperature. This factor is a good estimate of pretreatment severity for uncatalyzed samples, but does not take into account the effect of acid. Chum et al., (1990) devised a combined severity factor, adding the effect of the pH of the biomass prior to pretreatment to the equation (Equation 1.2) (Chum et al., 1990).   1:2 log …R0 † ˆ log t
exp …TÿTref †=14:75 ÿ pH The assumption of first order kinetics in steam explosion is of limited utility, particularly at high severity. Neither severity parameter takes into account that the cellulose fraction produced under severe conditions is much more readily digestible, while the liquid hemicellulosic fraction produced under mild conditions is much more fermentable. However, both reaction ordinates are useful as a guide to compare the effect of different pretreatment conditions. Steam explosion When it was first developed, steam explosion of wood for fiberboard was an uncatalyzed process. Since the purpose was to break down wood into a structure which could be formed into a sheet, temperatures as high as 280 ëC were employed (Boehm, 1930). Use of similar conditions to pretreat biomass for bioethanol production would generate unacceptably high levels of inhibitory compounds. The use of milder pretreatment conditions is necessary to minimize the formation of these compounds. With some types of biomass, simply reducing the time and temperature of pretreatment is enough to generate a substrate which is readily digestible and fermentable. Hardwoods and agricultural residues typically have highly acetylated hemicellulose (4-O-methylglucuronoxylans) which allows the biomass to undergo `autohydrolysis' under steam explosion conditions and obviates the need for an acid catalyst (Nabarlatz et al., 2007). More recalcitrant feedstocks, such as softwood, lack acetylated hemicellulose and instead have primarily minimally acetylated glucomannans and galactoglucomannans (Ramos, 2003) and require addition of an acid catalyst prior to pretreatment. Acid addition lowers the severity of the conditions required to achieve optimal substrate characteristics and minimizes inhibitor

ß Woodhead Publishing Limited, 2010

12

Bioalcohol production

formation (Ramos, 2003). To maximize ethanol yields, two-stage steam pretreatments utilizing low severity conditions to hydrolyze hemicellulose followed by higher severity to generate digestible cellulose have been employed (SoÈderstrom et al., 2002, 2003b). While marginally higher ethanol yields were obtained compared to an equivalent one-stage process, the additional cost of a second steam pretreatment step would likely prevent adoption of this technique. SO2 and H2SO4 are the most commonly used acid catalysts. Catalysis with SO2 has been shown to provide a higher glucose yield after hydrolysis and lower inhibition of fermentation, while use of H2SO4 has been shown to enhance the recovery of hemicellulosic sugars (Tengborg et al., 1998). SO2 is introduced in gaseous form prior to steam pretreatment, preventing the addition of extra moisture that occurs by soaking in H2SO4. At low temperatures, SO2 is converted to sulfuric acid by oxidation in the presence of oxygen. At elevated temperature in the absence of oxygen, disproportionation reactions convert a third to a half of the SO2 to reduction products including elemental sulfur or thiosulfate (Brownell et al., 1988). In addition, some SO2 is lost to the vapour phase. As a result of these reactions, it is difficult to determine exactly how much acid contributes to the pretreatment. It has been shown that impregnation of SO2 at concentrations higher than 3% has no additional beneficial effect (Gregg and Saddler, 1996; Clark et al., 1989). H2SO4 is introduced by soaking or spraying biomass in a dilute acid solution prior to pretreatment. While sugar recovery after pretreatment is higher than when SO2 is used, generation of inhibitors is higher (Tengborg et al., 1998). Overall ethanol yields are therefore higher when SO2 is used, despite slightly lower sugar yields after pretreatment. The required severity for different types of biomass is highly variable. Table 1.1 lists a select number of feedstocks along with their respective pretreatment conditions and yields. Xylose recovery is a good indicator of overall sugar recovery, as xylose is one of the most labile carbohydrates present in lignocellulosics (Fengel and Wegener, 1989). Ethanol yields are typically stated as a percentage of the theoretical value. Depending on the fermentative organism, only certain sugars can be converted to ethanol and at a certain rate. For example, Saccharomyces cerevisiae converts 6-carbon sugars and produces ethanol at a ratio of 0.51 gram ethanol per gram of sugar. In general, increased severity is required to go from agricultural residues to hardwoods to softwoods, with a concurrent decrease in xylose recovery. Ethanol yields do not exhibit such a clear trend. As shown in Fig. 1.1, as severity increases, the concentration of hexoses and pentoses increase in solution, but subsequently drop off as degradation reactions begin to dominate. Figure 1.2 shows the effect of increased severity on the composition of the water insoluble fraction. At high severity, much of the hemicellulose is hydrolyzed, increasing the relative concentrations of cellulose and lignin. As hemicellulose is removed and cellulose and lignin undergo chemical reactions, the cellulose becomes more readily digested by cellulases, generating more fermentable sugars. The final ethanol yield, there-

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

13

1.1 Schematic representation of the effect of pretreatment severity on concentrations of soluble lignin, pentoses and hexoses in the water insoluble fraction.

1.2 Schematic representation of the relative amount of hemicellulose, lignin and cellulose in the water insoluble fraction as severity increases.

fore, depends on the fractionation of fermentable sugars between liquid and solid fractions along with the digestibility of the solids and level of sugar degradation in the liquid. Liquid hot water While steam explosion is highly sensitive to pretreatment severity, liquid hot water is much less affected by changes to time and temperature. With many

ß Woodhead Publishing Limited, 2010

14

Bioalcohol production

feedstocks and process configurations, final sugar recoveries and subsequent ethanol yields are independent of pretreatment severity (Mok and Antal Jr, 1992). Particularly at low consistency, time and temperature have minimal effect on sugar recovery and overall ethanol yield. An approximate representation of the relative polymer degradation as severity increases is shown in Fig. 1.1 for hemicellulose, lignin and cellulose. Increasing severity leads to an increase in monomeric subunits in solution, then a decrease as monomers are degraded. The rate of both production and degradation of monomers is highly dependent on the process conditions as well as the feedstock and determines the resulting physical and chemical characteristics of the substrate.

1.3.5

Physical and chemical characteristics of pretreated biomass

Steam explosion The characteristics of steam exploded biomass are highly dependent on the severity of the pretreatment conditions. Different levels of time, temperature and catalyst concentration on the same raw feedstock will result in radically different pretreated substrates. In addition, the rate of decompression of the reactor at the end of the residence time can vary, depending on how the contents of the reactor are brought to atmospheric pressure ± whether by a gradual bleeding of the pressure or a sudden release of pressure resulting in an `explosion'. Both methods have the same macroscopic effect ± saturated water present inside the cells vaporizes and expands, separating the fibers. The microscopic effect of decompression is dependent upon the temperature differential. Even at relatively low process temperatures (180±200 ëC), lignin is past its glass transition temperature and exhibits thermoplastic behavior. Fibers separate at the middle lamella to yield a large number of bundled fibers, some individual fibers and minimal broken fibers. At moderate temperatures (200 ëC) free fibers dominate, while at higher temperatures (above 230 ëC) fiber damage and lignin repolymerization reactions result in mostly fiber fragments fused together by lignin (Schultz et al., 1983; Biermann et al., 1987). Temperatures required for optimal fiber separation vary by feedstock, with hardwoods requiring lower temperatures for fiber separation than softwoods. The effect of the rate of pressure drop on fiber properties has been shown to be significant but varies depending on the type of biomass and severity of the pretreatment. Brownell and Saddler (1987) showed that the overall ethanol yield of aspen steam pretreated with a gradual bleeding of pressure yielded the same amount of ethanol as material which underwent explosive decompression. This effect may be biomass dependent ± studies on eucalyptus have shown that explosive decompression prevents the steam pretreated solids from drying up when steam is slowly released (Ramos et al., 1992a). Another potential benefit

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

15

of bleeding off excess pressure is a reduction in volatile inhibitors (Foody, 1984). Cellulose crystallinity also changes as severity increases. As lignin is reorganized and hemicellulose solubilized, the molecular tensions holding crystalline cellulose together are weakened. This allows formerly amorphous cellulose to be incorporated into the crystalline structure, leading to an increase in the overall crystallinity (Shevchenko et al., 2001; Carrasco et al., 1994; Josefsson et al., 2001). However, this change may be dependent on process parameters, as other researchers have found that there is no change in crystallinity (Dekker and Wallis, 1983). Cellulose degree of polymerization (DP) also decreases; with the DP decreasing more rapidly after most of the hemicellulose is depolymerized. The DP tends towards the limiting degree of polymerization (LDOP) of approximately 150±250 DP. Once the limit is reached the average cellulose molecule will not reduce further in size (KraÈssig, 1993). The colour of both insoluble and soluble fractions darkens from light to dark brown as pretreatment severity increases (Sun et al., 2005; Negro et al., 2003). This is thought to be related to the breakdown of lignin and wood extractives. Lignin condensation could activate tannins and flavonoids towards condensation by lysing of protecting groups (sugars). Reaction with furfural and hydroxymethyl furfural, may also be responsible for these colour changes (Negro et al., 2003). While colour is not important for bioethanol production, it can be a qualitative measure of pretreatment severity. As severity increases, relative lignin content in the solid fraction increases as cellulose is solubilized. The lignin generated from repolymerization reactions is intermediate between native lignin in raw biomass and heavily condensed Klason lignin formed from acid pretreatment (Montane et al., 1998). Furfural and other degradation products react with lignin to form new C±C bonds, giving rise to condensation products known as pseudolignin (Klemola and Nyman, 1966). Addition of supplemental acid is thought to increase the molecular weight of condensation products and reduce solubility in organic solvents (Wright, 1988). Increased condensed lignin in the solid fraction can also reduce enzymatic digestibility by adsorbing proteins and inhibiting hydrolysis (Lu et al., 2002). Insoluble hemicellulose decreases dramatically as severity increases. The amount of hemicellulosic monomers and oligomers in the water soluble fraction concurrently increases as solubilization of the hemicellulose increases, then decreases as these sugars are degraded (Puls et al., 1985; Kabel et al., 2007). Pentose sugars (arabinose and xylose) degrade by dehydration reactions to furfural, hexoses (glucose, mannose and galactose) to 5-hydroxymethyl furfural (HMF). Furfural and HMF are further degraded to formic and levulinic acids (Cantarella et al., 2004). These compounds, along with lignin degradation products and acetic acid, are inhibitory to fermentative microorganisms.

ß Woodhead Publishing Limited, 2010

16

Bioalcohol production

Liquid hot water Liquid hot water processes are often run as a flow-through process where liquid is continually passed over biomass. This allows for continuous removal and cooling of the dissolved components, but results in a very dilute sugar stream (0.6±5.8 g/l) (Mosier et al., 2005a). Co-current or batch processes result in a higher concentration of sugar in the liquid stream but are more sensitive to process time and temperature. Mok and Antal Jr (1992) found that for 10 different herbaceous and hardwood species treated using a batch process for 0±15 minutes at 200±230 ëC, 100% of hemicellulose was dissolved with an average of 90% recovery of monomeric sugars. Solids concentration influences the final product, particularly in a batch system where increasing solids increases the acid concentration and subsequent degradation reactions. Laser et al. found that an increase from 1% to 5% solids resulted in a 97% reduction in ethanol yield after SSF although xylan yields remained high (81%) (Laser et al., 2002).

1.4

Conclusions

Steam explosion pretreatment is a robust means of fractionating lignocellulosic biomass. However, maximization of ethanol yield from enzymatic hydrolysis and microbial fermentation requires a compromise. Conditions yielding highly digestible cellulose will generate high levels of fermentation inhibitors, while milder conditions improve yields of fermentable sugars but fail to generate digestible cellulose. It is nearly impossible to maximize both enzymatic digestibility and fermentability. For a low value product like ethanol, the mixed stream resulting from a compromise severity is acceptable. The process can be adapted to fractionate highly pure components as well, such as furfural and HMF, or a clean hemicellulosic sugar solution. Liquid hot water pretreatments, particularly at high consistency, are capable of producing hydrolysates with low levels of inhibitors and highly digestible solids, but dilute liquid streams prevent efficient fermentation. Since acid catalysts are not required, problems with corrosion and chemical recycling and disposal are eliminated. In addition, while the process is highly effective on agricultural residues and some hardwoods, it has not yet been shown to be effective on softwood feedstocks. Table 1.2 compares the two processes of steam and hot water pretreatments. Both are carried out at similar temperatures and reaction times, so using the reaction ordinate in Equation 1.1, the relative severity of pretreatments carried out using the same conditions with different amounts of water present would be the same. It has been shown, however, that sugar recoveries and overall ethanol yields differ (Jacobsen and Wyman, 2002). The consistency during the reaction plays an important role. At higher consistency, as in steam explosion (>50%),

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

17

Table 1.2 Comparison of steam and liquid hot water pretreatments Pretreatment characteristic Solids consistency during pretreatment Readily digestible fiber Sugar concentration in water soluble fraction Pentose recovery Fermentative inhibitor formation Works on softwood Effective on hardwoods, agricultural residues Water usage

Steam explosion

Liquid hot water

High Yes High Low High Yes Yes Low

Low Yes Low High Low No Yes High

the relative acid concentration is much higher, leading to increased sugar degradation. At the lower consistencies (1±10%) used in liquid hot water treatments, the pH is higher and sugar recovery and solids yields are higher while the production of fermentation inhibitors is greatly reduced (Allen et al., 2001; Laser et al., 2002; Jacobsen and Wyman, 2002).

1.5

Future trends

At its present stage of development, the technology for processes based on steam treatment is mature enough at pilot plant and demonstration level, if the final product is ethanol. Several laboratory scale steam guns are operational at the national or university research labs such as National Renewable Energy Laboratories (NREL), University of British Columbia, Lund University, Virginia Tech University, and University of Washington. Currently, SunOpta Inc. (www.sunopta.com) is producing steam guns for commercial purposes. However, one of the problems associated with commercialization of the biomass to ethanol process is feedstock availability and cost. A future bioethanol facility able to utilize multiple feedstock sources would be at an advantage in that it would have more consistent supplies of raw material and would be better positioned to find this raw material at lower cost. In addition, it is likely that a future conversion facility will process biomass with impurities from harvesting and chipping units, such as soil, branches and needles. Although considerable research has been done in converting uniform, homogeneous feedstocks to ethanol (corn stover, wheat straw and hybrid poplar, pine among others), not much attention has been paid to the `pre-pretreatment' of lignocellulosics and the effects of biomass physical characteristics on pretreatment and ultimately, the overall ethanol yield. Additional efforts are required in understanding the effects of particle size, thickness, moisture content, biomass `freshness' and `purity', and in wood, for example, the effect on bark, needles or branches on the fractionation process. In addition, for the lignocellulosic biomass to ethanol

ß Woodhead Publishing Limited, 2010

18

Bioalcohol production

process to be economically feasible, it has to produce high value co-products. The process itself should be relatively simple. Although the phenomena involved in the steam-aqueous pretreatment of biomass to ethanol are well understood, not much research has been done in the fractionation of cellulose, hemicellulose and lignin into pure fractions for the green polymers industry. Compounds of interest include polyols (xylitol and arabitol), ethylene and propylene glycols, furfural and levulinic acid, among others (Werpy and Petersen, 2004). Ultimately, a greater fundamental understanding of the chemical and physical effects that occur during hydrothermal pretreatment, along with improved understanding of the physico-chemical structure of pretreated biomass, are essential for utilizing steam-aqueous pretreatment as a pretreatment process for biochemicals production.

1.6

References

Allen, S G, Spencer, M J and Antal, M J (1996), `Semi-chemical pulping using the Aquasolv process', Papers of the American Chemical Society, 211, 66. Allen, S G, Schulman, D, Lichwa, J, Antal Jr, M J, Laser, M and Lynd, L R (2001), `A comparison between hot liquid water and steam fractionation of corn fiber', Industrial and Engineering Chemistry Research, 40, 2934±2941. Asplund, A (1953), Sven. Papperstidn., 56, 550. Ballesteros, I, Oliva, J M, Navarro, A A, Gonzalez, A, Carrasco, J and Ballesteros, M (2000), `Effect of chip size on steam explosion pretreatment of softwood', Appl. Biochem. Biotechnol., 84-6, 97±110. Ballesteros, I, Negro, M J, Oliva, J M, CabanÄas, A, Manzanares, P and Ballesteros, M (2006), `Ethanol production from steam-explosion pretreated wheat straw', Appl. Biochem. Biotechnol., 130, 496±508. Biermann, C J, McGinnis, G D and Schultz, T P (1987), `Scanning electron microscopy of mixed hardwoods subjected to various pretreatment processes', J. Agric. Food Chem., 35, 713±716. Bobleter, O and Pape, G (1968), `Hydrothermal decomposition of glucose', Monatsh. Chem, 99, 1560±1567. Bobleter, O, Niesner, R and Rohr, M (1976), `The hydrothermal degradation of cellulosic matter to sugars and their fermentative conversion to protein', J. Appl. Polym. Sci., 20, 2083±2093. Boehm, R M (1930), `The Masonite process', Ind. Eng. Chem., 22, 493±497. Bothast, R J and Schlicher, M A (2005), `Biotechnological processes for conversion of corn into ethanol', Appl. Microbiol. Biotechnol., 67, 19±25. Boussaid, A, Esteghlalian, A R, Gregg, D J, Lee, K H and Saddler, J N (2000), `Steam pretreatment of Douglas-fir wood chips ± Can conditions for optimum hemicellulose recovery still provide adequate access for efficient enzymatic hydrolysis?', Appl. Biochem. Biotechnol., 84-6, 693±705. Brasch, D J and Free, K W (1965), `Prehydrolysis-kraft pulping of Pinus radiata grown in New Zealand', Tappi J., 48, 245±248. Brownell, H H and Saddler, J N (1987), `Steam pretreatment of lignocellulosic material for enhanced enzymatic hydrolysis', Biotechnol. Bioeng., 29, 228±235.

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

19

Brownell, H H, Yu, E K C and Saddler, J N (1986), `Steam-explosion pretreatment of wood ± Effect of chip size, acid, moisture-content and pressure-drop', Biotechnol. Bioeng., 28, 792±801. Brownell, H H, Schwald, W, Smaridge, T and Saddler, J N (1988), `Steam pretreatment of wood for enzymatic hydrolysis-chemical and physical changes involved', In: Saddler, J N, Stevens, D and Hages, D. (eds) Proceedings of the IEA Workshop on Bioconversion of Lignocellulosics, Forintek Canada. pp. A-1. Bura, R, Mansfield, S D, Saddler, J N and Bothast, R J (2002), `SO2-catalyzed steam explosion of corn fiber for ethanol production', Appl. Biochem. Biotechnol., 98±100, 59±72. Bura, R, Bothast, R J, Mansfield, S D and Saddler, J N (2003), `Optimization of SO2catalyzed steam pretreatment of corn fiber for ethanol production', Appl. Biochem. Biotechnol., 106, 319±335. Cadoche, L and Lopez, G D (1989), `Assessment of size reduction as a preliminary step in the production of ethanol from lignocellulosic wastes', Biological Wastes, 30, 153±157. Cantarella, M, Cantarella, L, Gallifuoco, A, Spera, A and Alfani, F (2004), `Effect of inhibitors released during steam-explosion treatment of poplar wood on subsequent enzymatic hydrolysis and SSF', Biotechnol. Prog., 20, 200±206. Carrasco, J E, Saiz, M C, Navarro, A, Soriano, P, Saez, F and Martinez, J M (1994), `Effects of dilute acid and steam explosion pretreatments on the cellulose structure and kinetics of cellulosic fraction hydrolysis by dilute acids in lignocellulosic materials', Appl. Biochem. Biotechnol., 45, 23±34. Chandra, R P, Bura, R, Mabee, W E, Berlin, A, Pan, X and Saddler, J N (2007), `Substrate pretreatment: the key to effective enzymatic hydrolysis of lignocellulosics?', Adv. Biochem. Eng/Biotechnol., 108, 67±93. Chum, H, Johnson, D K, Black, S and Overend, R (1990), `Pretreatment-catalyst effects and the combined severity parameter', Appl. Biochem. Biotechnol., 24, 1±14. Clark, T A, Mackie, K L, Dare, P H and McDonald, A G (1989), `Steam explosion of the softwood Pinus Radiata with sulphur dioxide addition. II. Process characterisation', J. Wood Chem. Technol., 9, 135±166. Cullis, I F, Saddler, J N and Mansfield, S D (2004), `Effect of initial moisture content and chip size on the bioconversion efficiency of softwood lignocellulosics', Biotechnol. Bioeng., 85, 413±421. Datta, R (1981), `Energy requirements for lignocellulose pretreatment processes', Process Biochemistry, 16, 16±19. De Bari, I, Nanna, F and Braccio, G (2007), `SO2-catalyzed steam fractionation of aspen chips for bioethanol production: optimization of the catalyst impregnation', Ind. Eng. Chem. Res, 46, 7711±7720. Dekker, R F H and Wallis, A F A (1983), `Enzymatic saccharification of sugarcane bagasse pretreated by autohydrolysis-steam explosion', Biotechnol. Bioeng., 25, 3027±3048. DeLong, E A (1981), Method of rendering lignin separable from cellulose and hemicellulose in lignocellulosic material and the product so produced, Canadian patent 1096374. Ewanick, S M, Bura, R and Saddler, J N (2007), `Acid-catalyzed steam pretreatment of lodgepole pine and subsequent enzymatic hydrolysis and fermentation to ethanol', Biotechnol. Bioeng., 98, 737±746. Fengel, D. and Wegener, G. (1989), Wood: chemistry, ultrastructure, reactions, Walter de Gruyter, Berlin.

ß Woodhead Publishing Limited, 2010

20

Bioalcohol production

Foody, P (1984), Method for obtaining superior yields of accessible cellulose and hemicellulose from lignocellulosic materials, 116305, 8th edn, Canada. Fukazawa, K, Revol, J F, Jurasek, L and Goring, D A I (1982), `Relationship between ball milling and the susceptibility of wood to digestion by cellulase', Wood Sci. Technol., 16, 279±285. Galbe, M., Sassner, P., Wingren, A. and Zacchi, G. (2007), `Process engineering economics of bioethanol production', in Biofuels, ed. L. Olsson, Springer, Berlin, 303±327. Glasser, W G and Wright, R S (1998), `Steam-assisted biomass fractionation. II. Fractionation behavior of various biomass resources', Biomass Bioenerg., 14, 219±235. Gregg, D and Saddler, J N (1996), `A techno-economic assessment of the pretreatment and fractionation steps of a biomass-to-ethanol process', Appl. Biochem. Biotechnol., 57, 711±727. Heitz, M, Capek-Menard, E, Koeberle, P G, Gagne, J, Chornet, E, Overend, R P, Taylor, J D and YU, E (1991), `Fractionation of Populus tremuloides at the pilot plant scale: optimization of steam pretreatment conditions using the STAKE II technology', Bioresour. Technol., 35, 23±32. Jacobsen, S E and Wyman, C E (2002), `Xylose monomer and oligomer yields for uncatalyzed hydrolysis of sugarcane bagasse hemicellulose at varying solids concentration', Ind. Eng. Chem. Res, 41, 1454±1461. Jollez, P, Chornet, E and Overend, R P (1994), `Steam-aqueous fractionation of sugar cane bagasse: an optimization study of process conditions at the pilot plant level', in Advances in Thermomechanical Biomass Conversion, ed. A.V. Bridgwater, Blackie Academic and Professional, London, 1659±1669. Josefsson, T, Lennholm, H and Gellerstedt, G (2001), `Changes in cellulose supramolecular structure and molecular weight distribution during steam explosion of aspen wood', Cellulose, 8, 289±296. Kabel, M A, Bos, G, Zeevalking, J, Voragen, A G J and Schols, H A (2007), `Effect of pretreatment severity on xylan solubility and enzymatic breakdown of the remaining cellulose from wheat straw', Bioresour. Technol., 98, 2034±2042. Kelsey, R G and Shafizadeh, F (1980), `Enhancement of cellulose accessibility and enzymatic hydrolysis by simultaneous wet milling', Biotechnol. Bioeng., 22, 1025± 1036. Klemola, A and Nyman, G A (1966), `Steam hydrolysis of birchwood: the effect of hydrolysis on the hemicellulose, cellulose, and on the lignin of birchwood', Paperi ja Puu, 48, 595±603. Kokta, B V and Ahmed, A (1998), `Steam explosion pulping' in Environmentally Friendly Technologies for the Pulp and Paper Industry, eds. R.A. Young and M. Akhtar, J. Wiley, New York, 191±214. KraÈssig, H A (1993), Cellulose: Structure, Accessibility and Reactivity, Gordon and Breach, Yverdon, Switzerland. Kristensen, J B, Thygesen, L G, Felby, C, Jorgensen, H and Elder, T (2008), `Cell-wall structural changes in wheat straw pretreated for bioethanol production', Biotechnol. Biofuels, 1, 5. Larsen, J, Petersen, M é, Thirup, L, Li, H W and Iversen, F K (2008), `The IBUS process-lignocellulosic bioethanol close to a commercial reality', Chem. Eng. Technol., 31, 765±772. Laser, M, Schulman, D, Allen, S G, Lichwa, J, Antal, M J and Lynd, L R (2002), `A comparison of liquid hot water and steam pretreatments of sugar cane bagasse for bioconversion to ethanol', Bioresour. Technol., 81, 33±44.

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

21

Li, J, Henriksson, G and Gellerstedt, G (2007), `Lignin depolymerization/ repolymerization and its critical role for delignification of aspen wood by steam explosion', Bioresour. Technol., 98, 3061±3068. Lu, Y, Yang, B, Gregg, D, Saddler, J N and Mansfield, S D (2002), `Cellulase adsorption and an evaluation of enzyme recycle during hydrolysis of steam-exploded softwood residues', Appl. Biochem. Biotechnol., 98, 641±654. Lynd, L R (1996), `Overview and evaluation of fuel ethanol from cellulosic biomass: Technology, economics, the environment, and policy', Annu. Rev. Energy Environ., 21, 403±465. Mamers, H, Menz, D and Yuritta, J P (1979), `Explosion pulping of annual and fast growing plants', Appita, 33, 201±205. Mamers, H, Yuritta, J P and Menz, D J (1981), `Explosion pulping of bagasse and wheat straw', Tappi J., 64, 93±96. Marshall, W M and Franck, E U (1981), `Ion product of water substance, 0±1000 ëC, 1± 10,000 bars: new international formulation and its background', J. Phys. Chem. Ref. Data, 10, 295±304. Mason, W H (1928), Apparatus for and explosion fibration of lignocellulose material, 1,655,618 edn, USA. Michalowicz, G, Toussaint, B and Vignon, M R (1991), `Ultrastructural changes in poplar cell wall during steam explosion treatment', Holzforschung, 45, 175±179. Millett, M A, Effland, M J and Caulfield, D F (1979), `Influence of fine grinding on the hydrolysis of cellulosic materials-acid vs enzymic', Adv. Chem. Ser, 181, 71±89. Mok, W S L and Antal Jr, M J (1992), `Uncatalyzed solvolysis of whole biomass hemicellulose by hot compressed liquid water', Ind. Eng. Chem. Res., 31, 1157± 1161. MontaneÂ, D, Salvado, J, Farriol, X, Vidal, P, Jollez, P and Chornet, E (1998), `Polysaccharides from biomass via thermomechanical process', in Polysaccharides: structural diversity and functional versatility, ed. S. Dumitriu, Marcel Dekker, New York, 1069±1085. Mooney, C A, Mansfield, S D, Beatson, R P and Saddler, J N (1999), `The effect of fiber characteristics on hydrolysis and cellulase accessibility to softwood substrates', Enzyme Microb. Technol., 25, 644±650. Mosier, N, Wyman, C, Dale, B, Elander, R, Lee, Y Y, Holtzapple, M and Ladisch, M (2005a), `Features of promising technologies for pretreatment of lignocellulosic biomass', Bioresour. Technol., 96, 673±686. Mosier, N S, Hendrickson, R, Brewer, M, Ho, N, Sedlak, M, Dreshel, R, Welch, G, Dien, B S, Aden, A and Ladisch, M R (2005b), `Industrial scale-up of pH-controlled liquid hot water pretreatment of corn fiber for fuel ethanol production', Appl. Biochem. Biotechnol., 125, 77±97. Nabarlatz, D, EbringerovaÂ, A and MontaneÂ, D (2007), `Autohydrolysis of agricultural byproducts for the production of xylo-oligosaccharides', Carbohydr. Polym., 69, 20±28. Negro, M J, Manzanares, P, Oliva, J M, Ballesteros, I and Ballesteros, M (2003), `Changes in various physical/chemical parameters of Pinus pinaster wood after steam explosion pretreatment', Biomass Bioenerg., 25, 301±308. È hgren, K, Bura, R, Saddler, J and Zacchi, G (2007), `Effect of hemicellulose and lignin O removal on enzymatic hydrolysis of steam pretreated corn stover', Bioresour. Technol., 98, 2503±2510. Overend, R P, Chornet, E and Gascoigne, J A (1987), `Fractionation of lignocellulosics by steam-aqueous pretreatments [and Discussion]', Phil. Trans. R. Soc. Lond., 321, 523±536.

ß Woodhead Publishing Limited, 2010

22

Bioalcohol production

Petersen, M é, Larsen, J and Thomsen, M H (2009), `Optimization of hydrothermal pretreatment of wheat straw for production of bioethanol at low water consumption without addition of chemicals', Biomass Bioenergy, 33(5), 834±840. Puls, J, Poutanen, K, KoÈrner, H U and Viikari, L (1985), `Biotechnical utilization of wood carbohydrates after steaming pretreatment', Appl. Microbiol. Biotechnol., 22, 416± 423. Puri, V P and Mamers, H (1983), `Explosive pretreatment of lignocellulosic residues with high-pressure carbon dioxide for the production of fermentation substrates', Biotechnol. Bioeng., 25, 3149±3161. Ragauskas, A J, Nagy, M, Kim, D H, Eckert, C A, Hallett, J P and Liotta, C L (2006), `From wood to fuels: Integrating biofuels and pulp production', Industrial Biotechnology, 2, 55±65. Ramos, L P (2003), `The chemistry involved in the steam treatment of lignocellulosic materials', Quim. Nova, 26, 863±871. Ramos, L P, Breuil, C, Kushner, D J and Saddler, J N (1992a), `Steam pretreatment conditions for effective enzymatic hydrolysis and recovery yields of Eucalyptus viminalis wood chips', Holzforschung, 46, 149±154. Ramos, L P, Breuil, C and Saddler, J N (1992b), `Comparison of steam pretreatment of eucalyptus, aspen, and spruce wood chips and their enzymatic hydrolysis', Appl. Biochem. Biotechnol., 34/35, 37±47. Richter, G A (1956), `Some aspects of prehydrolysis pulping', Tappi J., 39, 193±210. Rudolf, A, Alkasrawi, M, Zacchi, G and Liden, G (2005), `A comparison between batch and fed-batch simultaneous saccharification and fermentation of steam pretreated spruce', Enzyme Microb. Technol., 37, 195±204. Saddler, J N, Ramos, L P and Breuil, C (1993), `Steam pretreatment of lignocellulosic residues' in Bioconversion of Forest and Agricultural Plant Residues, ed. J.N. Saddler, CABI, Great Britain, pp. 73±91. Saddler, J N, Mooney, C A, Mansfield, S D and Beatson, R P (1999), `Influence of fiber characteristics on the cellulase accessibility to softwoods', Abstr. Pap. Am. Chem. Soc., 217, U264±U265. Sassner, P, Galbe, M and Zacchi, G (2005), `Steam pretreatment of Salix with and without SO2 impregnation for production of bioethanol', Appl. Biochem. Biotechnol., 124, 1101±1117. Sassner, P, MaÊrtensson, C G, Galbe, M and Zacchi, G (2008), `Steam pretreatment of H2SO4-impregnated Salix for the production of bioethanol', Bioresour. Technol., 99, 137±145. Schell, D J and Harwood, C (1994), `Milling of lignocellulosic biomass', Appl. Biochem. Biotechnol., 45, 159±168. Schultz, T P, Biermann, C J and McGinnis, G D (1983), `Steam explosion of mixed hardwood chips as a biomass pretreatment', Ind. Eng. Chem. Prod. Res. Dev., 22, 344±348. Shevchenko, S M, Chang, K, Dick, D G, Gregg, D J and Saddler, J N (2001), `Structure and properties of lignin in softwoods after SO2-catalyzed steam explosion and enzymatic hydrolysis', Cell Chem. Technol., 35, 487±502. SoÈderstrom, J, Pilcher, L, Galbe, M and Zacchi, G (2002), `Two-step steam pretreatment of softwood with SO2 impregnation for ethanol production', Appl. Biochem. Biotechnol., 98±100, 5±21. SoÈderstrom, J, Pilcher, L, Galbe, M and Zacchi, G (2003a), `Combined use of H2SO4 and SO2 impregnation for steam pretreatment of spruce in ethanol production', Appl. Biochem. Biotechnol., 105, 127±140.

ß Woodhead Publishing Limited, 2010

Hydrothermal pretreatment of lignocellulosic biomass

23

SoÈderstrom, J, Pilcher, L, Galbe, M and Zacchi, G (2003b), `Two-step steam pretreatment of softwood by dilute H2SO4 impregnation for ethanol production', Biomass Bioenergy, 24, 475±486. Stenberg, K, Tengborg, C, Galbe, M and Zacchi, G (1998), `Optimization of steam pretreatment of SO2-impregnated mixed softwoods for ethanol production', J. Chem. Technol. Biotechnol., 71, 299±308. Stenberg, K, Bollok, M, Reczey, K, Galbe, M and Zacchi, G (2000), `Effect of substrate and cellulase concentration on simultaneous saccharification and fermentation of steam-pretreated softwood for ethanol production', Biotechnol. Bioeng., 68, 204± 210. Sun, X F, Xu, F, Sun, R C, Fowler, P and Baird, M S (2005), `Characteristics of degraded cellulose obtained from steam-exploded wheat straw', Carbohydr. Res., 340, 97± 106. Tengborg, C, Stenberg, K, Galbe, M, Zacchi, G, Larsson, S, Palmqvist, E and HahnHaÈgerdal, B (1998), `Comparison of SO2 and H2SO4 impregnation of softwood prior to steam pretreatment on ethanol production', Appl. Biochem. Biotechnol., 702, 3±15. Tucker, M P, Kim, K H, Newman, M M and Nguyen, Q A (2003), `Effects of temperature and moisture on dilute-acid steam explosion pretreatment of corn stover and cellulase enzyme digestibility', Appl. Biochem. Biotechnol., 105±108, 165±177. Van Walsum, G P, Allen, S G, Spencer, M J, Laser, M S, Antal, M J and Lynd, L R (1996), `Conversion of lignocellulosics pretreated with liquid hot water to ethanol', Appl. Biochem. Biotechnol., 57, 157±170. Vroom, K E (1957), `The ``H'' factor: a means of expressing cooking times and temperatures as a single variable', Pulp Paper Mag. Can., 58, 228±231. Wayman, M (1980), `Alcohol from cellulosics: The autohydrolysis-extraction process', In: Anonymous Proceedings of the IV International Symposium on Alcohols Fuel Technology, October 5±8. Werpy, T and Petersen, G (2004), `Top Value Added Chemicals from Biomass. Volume I: Results of Screening for Potential Candidates from Sugars and Synthesis Gas', DOE/GO-102004-1992. Wong, K K Y, Deverell, K F, Mackie, K L, Clark, T A and Donaldson, L A (1988), `The relationship between fiber porosity and cellulose digestibility in steam-exploded Pinus radiata', Biotechnol. Bioeng., 31, 447±456. Wood, T M and Saddler, J N (1988), `Increasing the availability of cellulose in biomass materials', Meth. Enzymol., 160, 3±11. Wright, J D (1988), `Ethanol from biomass by enzymatic hydrolysis', Chem. Eng. Prog., 84, 62±74. Yang, B and Wyman, C E (2004), `Effect of xylan and lignin removal by batch and flowthrough pretreatment on the enzymatic digestibility of corn stover cellulose', Biotechnol. Bioeng., 86, 88±98.

ß Woodhead Publishing Limited, 2010

2

Thermochemical pretreatment of lignocellulosic biomass S. P. S. CHUNDAWAT, V. BALAN, L. DA COSTA S O U S A a n d B . E . D A L E , Michigan State University, USA

Abstract: The development of an economically viable and environmentally sustainable bio-based chemical industry has been largely impeded by the native recalcitrance of lignocellulosic feedstock to thermochemical (e.g., chemical pretreatment) and biological processing (e.g., enzymatic hydrolysis and fermentation). This chapter explores various thermochemical pretreatments that enhance enzymatic digestibility of lignocellulosics by solubilizing, hydrolyzing and chemically modifying individual cell-wall components like lignin, hemicellulose and cellulose. Substrate and pretreatment related factors that influence the effectiveness of the entire biorefinery process, based on the ultimate enzymatic digestibility and fermentability of the treated biomass, are closely examined from an economic and environmental point of view. Key words: thermochemical pretreatment, plant cell wall recalcitrance, enzymatic hydrolysis, cellulosic ethanol.

2.1

Introduction

The need for alternative liquid transportation fuels is imperative due to increasing demand from rapidly industrializing nations, dwindling petroleum and natural gas supplies, exclusive monopoly by politically volatile countries and detrimental effects of carbon dioxide emissions from fossil fuels on climate change. The sustainable development of economically viable bio-fuels from lignocellulosic biomass is one suitable alternative to the impending global energy crisis. However, lignocellulosic biomass is naturally recalcitrant to biological degradation due to several inherent characteristics of plant cell walls (Himmel et al., 2007). The enzymatic digestibility of native and pretreated lignocellulosics depends primarily on two types of factors: (a) substrate-related and (b) enzymerelated (Mansfield et al., 1999; Zhang and Lynd, 2004). Substrate-related factors are typically cellulose crystallinity and degree of polymerization (Fan et al., 1980; Stalbrand et al., 1998; Klein and Snodgrass, 1993; Ramos et al., 1993), hemicellulose side-chain branching (Chang and Holtzapple, 2000), lignin monomer type and content (Besle et al., 1994),

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

25

coumarate and ferulate cross-linking (Besle et al., 1994; Laureano-Perez et al., 2005), cell wall porosity (Ishizawa et al., 2007; Stone and Scallan, 1969; White and Brown, 1981; Thompson et al., 1992) and biomass particle size (Zadrazil and Puniya, 1995). Enzyme-related characteristics that influence hydrolysis are enzyme specific activity (Zhang et al., 2006b), enzyme synergy (Reese et al., 1950; Rosgaard et al., 2007; Bhat and Hazlewood, 2001; Gow and Wood, 1988; Wood, 1968; Wood and McCrae, 1986; Wood et al., 1989; Bhat and Bhat, 1997), enzyme inactivation due to non-productive protein binding to lignin and cellulose during hydrolysis (Eriksson et al., 2002a, 2002b; Yang et al., 2006; VaÈljamaÈe et al., 1999; Holtzapple et al., 1994; De La Rosa et al., 1994), and inhibition due to end-products or pretreatment degradation compounds (Gusakov and Sinitsyn, 1992; Selig et al., 2007). Depending on the pretreatment catalyst and solvent used, it is possible to solubilize, hydrolyze and physically separate cellulose, hemicellulose and lignin (Weil et al., 1994). Several types of pretreatment chemicals/solvents have been used, such as, concentrated acid (Goldstein and Easter, 1992), dilute acid (Saha and Bothast, 1999), caustic soda (Koullas et al., 1993), sulfur dioxide (Clark and Mackie, 1987), hydrogen peroxide (Gould, 1984), steam (FernaÂndez-BolanÄos et al., 2001), liquid ammonia (Dale et al., 1996), alkali-peroxide (Schmidt and Thomsen, 1998), lime (Kaar and Holtzapple, 2000), liquid hot water (Laser et al., 2002), carbon dioxide (Dale and Moreira, 1982), and several other organic solvents (Chum et al., 1988). Thermochemical pretreatment of lignocellulosic biomass is known to enhance the yield of fermentable sugars during enzymatic hydrolysis 3±10fold, depending on the nature of the substrate and type of pretreatment. This review closely examines the leading thermochemical pretreatments that have been tested on lignocellulosic feedstocks, such as, agricultural residues (e.g. corn stover) and energy crops (e.g. poplar). Substrate and pretreatment related factors affecting the viability of the process, based on the extent of enzymatic digestibility and ethanologenic fermentability of treated biomass hydrolyzate, are also examined.

2.2

Why is pretreatment necessary for lignocellulosics?

Thermochemical pretreatments (like ammonia fiber expansion or AFEX) help reduce lignocellulose recalcitrance by unwinding and leaching the tightly woven cell wall ultra structure (Chundawat et al., 2006; 2007; 2008b; Chundawat, 2009; Donohoe et al., 2008). The actual mechanism of `unwinding' the cell wall is unique to each pretreatment, closely dependent on the pretreament chemistry and nature of the substrate. Most flow-through-based pretreatments (employing high liquid solvent to solid biomass loadings) physically extract lignin and

ß Woodhead Publishing Limited, 2010

26

Bioalcohol production

hemicellulose into separate liquid process streams, hence improving cellulase accessibility to residual cellulose (Liu and Wyman et al., 2005; Chandra et al., È hgren et al., 2007; Wu and Lee, 1997; 2007; FernaÂndez-BolanÄos et al., 2001; O Donohoe et al., 2008; Kristensen et al., 2008). However, certain pretreatments like AFEX that do not physically extract lignin and hemicellulose are thought to modify the cell wall ultra-structure through a mechanism that is currently not well understood (Chundawat, 2009). Before exploring what parameters are responsible for an effective pretreatment, it would be necessary to closely examine the cell wall architecture in order to overcome their native recalcitrance.

2.2.1

Native plant cell wall recalcitrance

The primary constituents of lignocellulosic biomass are 30±50% cellulose (glucose polymer), 15±35% hemicellulose (hetero-sugar polymer), 10±30% lignin (phenyl propanoid polymer) and other minor constituents that include proteins (3±10%), lipids (1±5%), soluble sugars (1±5%) and minerals (5±10%) (Pauly and Keegstra, 2008). Cellulose is a linear homo-polysaccharide that consists of glucose (D-glucopyranose) units linked together by -(1-4) glycosidic linkages ( -D-glucan). The degree of polymerization (DP) of cellulose varies depending on its source. Cellulose from Avicel (PH-101) has 20-fold lower DP compared to untreated corn stover (300 vs. 7000, respectively) (Wyman et al., 2006; Kumar et al., 2009; Marx-Figini, 1969). The adjacent glucan chains form an elementary microfibril (3±5 nm diameter) of water-insoluble aggregates of varying length and width in the primary cell wall of corn stover (Ding and Himmel, 2006). These aggregates contain ordered (crystalline) and less-ordered (amorphous) regions of cellulose (Fengel and Wegener, 1989). The lattice forces responsible for maintaining the crystalline regions are a result of extensive inter- and intramolecular hydrogen bonding. It is this solid, crystalline morphology of cellulose that results in the slow saccharification kinetics largely due to steric hindrance of accessible glucan chains that are hidden below each other (Zhou et al., 2009). The cellulose polymorph typically seen in higher plants is I , which is a more tightly packed crystal structure, compared to other cellulose polymorphs due to differences in hydrogen bonding patterns (O'Sullivan, 1997). Chemical treatment of cellulose with sodium hydroxide or anhydrous liquid ammonia can modify the native crystal structure of cellulose I to cellulose II and III, respectively (Wada et al., 2008). The rate of enzymatic hydrolysis of cellulose is known to be closely dependent on its crystal structure and varies in the following order for various cellulose allomorphs, based on ease of saccharification; Amorphous > Cellulose III > Cellulose II ~ Cellulose I (Weimer et al., 1991; Igarashi et al., 2007). However, more work is needed to better understand the transformations between the various polymorphs during chemical pretreatments

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

27

and the effect of different enzyme systems (free vs. complexed) on cellulose allomorph digestion kinetics. Hemicelluloses are complex hetero-polysaccharides whose chemical composition, unlike cellulose, varies between cell tissues and plant species (Ebringerova, 2006; Ebringerova and Heinze, 2000). These polysaccharides are formed by a wide variety of sugar building blocks including pentoses (xylose, arabinose), hexoses (glucose, mannose, galactose) and uronic acids (4O-methyl-glucuronic, galacturonic acids) (Fengel and Wegener, 1989). Generally, hemicelluloses fall into two classes: (a) un-branched chains such as -(1-4)-linked xylans or mannans; and (b) branched chains such as (1-4)-linked galactoglucomannans and arabinoxylans (O2/O3 xylosyl linkages). The most abundant hemicelluloses found in agricultural residues like corn stover are arabinoxylans and arabino-(glucurono)-xylans (Buranov and Mazza, 2008; Ebringerova and Heinze, 2000). The ratio of pentosans (xylose, arabinose) and acidic sugars (uronic acids) varies considerably between tissues, but typically corn stover cell walls contain at least 20±25% xylan, 4±5% arabinan and 3±5% uronic acids. Lignin is a phenyl-propanoid-based macromolecular network that is primarily formed through the free-radical polymerization of p-hydroxy cinnamyl alcohol units of varying methoxyl content (Palonen, 2004). The chemical structure of lignin is complex and is largely based on its three phenolic precursors: coniferyl alcohol, sinapyl alcohol, and p-coumaryl alcohol. The ratios and absolute amounts of these precursors vary significantly between species, phenotypes, organs (leaf, stem, sheath), tissues (xylem, sclerenchyma, parenchyma, epidermis) and hence tremendously affect the physicochemical nature of lignocellulosic cell walls (Besle et al., 1994). Lignification of cell walls helps provide strength to plant tissues preventing collapse of waterconducting elements and provides defense against various pathogens (fungi and bacteria). These three cell wall components are organized together in a complex matrix (model schematic shown in Fig. 2.1) depending on the plant clade (monocots vs. dicots), type of cell (sclerenchyma vs. parenchyma cells) and type of cell wall (primary vs. secondary). Several simplified models of the primary cell wall have been presented over the years (McCann and Carpita, 2008; Carpita and Gibeaut, 1993), with little work on understanding the ultra-structural architecture of secondary cell walls (Bidlack et al., 1992; Ruel et al., 2006). Secondary cell walls (e.g., xylem vessels, sclerenchyma cells) are more recalcitrant to enzymatic hydrolysis than primary cell walls. Therefore, a better understanding of the chemical composition and ultra-structure architecture of the secondary wall before/after thermochemical pretreatment would help identify the fundamental rate-limiting steps to cell wall deconstruction.

ß Woodhead Publishing Limited, 2010

28

Bioalcohol production

2.1 Untreated grass cell wall structural components depicted by the intertwined matrix of cellulose, hemicellulose and lignin. Crystalline (straight, parallel lines) and amorphous cellulose (curved, wavy lines) microfibrils are composed of -1,4 linked glucan chains. Hemicellulose (thick, short lines) with no side-chains is hydrogen-bonded to cellulose, while branches with sidechains (of arabinosyl, acetyl, glucuronyl groups shown as circles and squares) help in cross-linking cellulose microfibrils. Lignin carbohydrate complexes (between arabinose and ferulic/coumaric acids) further restrict access of enzymes to cell wall polysaccharides. Cross-linked lignin monomers (consisting of hydroxyphenyl, syringyl, guaicyl units) constitute the core-lignin fraction.

2.2.2

Effect of thermochemical pretreatment on cell wall composition and ultra-structure

Most chemical pretreatments modify cell wall ultra-structures through certain physicochemical modifications, depending on the pretreatment chemistry and type of biomass, which ultimately helps enhance the enzymatic digestibility. The primary goal of any pretreatment is to enhance the rate of enzymatic hydrolysis, which is typically achieved through improving enzyme accessibility to the cell wall polysaccharides by removal of lignin and/or hemicellulose. Hemicellulose and lignin form physical barriers within the cell wall that limit enzymatic accessibility to cellulose. Most pretreatments target lignincarbohydrate complex (LCC) linkages between lignin and hemicellulose to help physically extract the amorphous matrix components while revealing the underlying cellulosic fibrils (Koshijima and Watanabe, 2003; Koshijima et al., 1989). One of the common LCC linkages is the ester bonds between arabinose sidechains and ferulic acid (Lapierre et al., 2001; Saulnier and Thibault, 1999).

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

29

While, deacetylation helps improve accessibility of debranched hemicellulose to endoxylanases (Kong et al., 1992; Chang and Holtzapple, 2000), acidic pretreatments also hydrolyze polysaccharide-based glycosidic linkages to form gluco- and xylo-oligomers depending on the severity of pretreatment (Yang and Wyman, 2008). Certain pretreatments modify the crystallinity and accessibility of cellulose to enzymes/microbes (Laureano-Perez et al., 2005; Weimer et al., 1991). Lignin-based ether linkages are prone to cleavage during acidic pretreatments (Shevchenko et al., 2001). While, alkaline pretreatments like wetalkali oxidation and ammoniation are also thought to chemically alter lignin structure, but the chemistries are not well understood (Sewalt et al., 1997; Gould, 1984; Klinke et al., 2002). Certain pretreatments physically extract lignin along with the liquid stream, primarily depending on the solvent to biomass loading employed (Yang and Wyman, 2004; Papatheofanous et al., 1996; Kim and Lee, 2005a). Chemical characterization of pretreated lignocellulosics is relatively better understood compared to the effect of thermochemical pretreatment on the three-dimensional ultra-structural network of cellulose-hemicellulose-lignin. There have been efforts in recent years to study the effect of dilute acid and steam explosion-based pretreatments on ultra-structural modifications within lignocellulosic grass cell walls using high resolution transmission/scanning electron (TEM, SEM) and atomic force microscopy (AFM)-based techniques (Donohoe et al., 2008; Selig et al., 2007; Kristensen et al., 2008). Dilute-acid pretreatment has been shown to cause lignin coalescence into droplets that migrate out of corn stover cell walls. The migrating droplets redeposit within and outside the cell walls at the end of pretreatment, ranging in size from 5 nm to 10 m (Donohoe et al., 2008). These spherical droplets are composed essentially of lignin and were found to severely inhibit enzymatic activity (Selig et al., 2007). Lignin re-localization away from cellulose microfibrils helps improve enzyme accessibility, however, complete lignin removal might cause extensive collapse of microfibrils. Extensive pretreatment may hence impede accessibility of cellulose microfibrils to enzymes (Weimer et al., 1986; Chou, 1986). Alkaline extraction and delignification of spruce-based softwood result in considerable nanostructural changes of the hemicellulose-lignin matrix surrounding cellulose microfibrils depending on the severity of pretreatment (Jungnikl et al., 2008). Alkali pulping typically results in swelling of cellulose crystallites as seen via AFM imaging (Fahlen and Salmen, 2003) and causes significant modification of cell wall morphology. Treating hard woods with liquid ammonia has shown to result in extensive fiber defibrillation with no significant extraction of lignin/ hemicellulose (O'Connor, 1972). Recent work on ammonia pretreatment (AFEX) has shown a unique mechanism by which the plant cell walls are modified resulting in enhanced enzymatic digestibility and fermentability (Chundawat, 2009; Lau and Dale; 2009). Localization of lignin and hemicellulose residues around cellulosic microfibrils seems to be an important rate-limiting step to the hydrolysis of plant cell

ß Woodhead Publishing Limited, 2010

30

Bioalcohol production

walls. This would suggest that engineering plant cell walls with artificial LCC linkages (and/or `zipper' lignin containing ester linkages) that can be easily cleaved during pretreatment would help reduce thermochemical severity to maximize digestibility (Mansfield, 2009). However, the final hurdle for rapid cell wall hydrolysis would be the breakdown of crystalline cellulose. This is essentially due to the spatially coupled organization of densely, packed glucan chains in cellulose microfibrils that results in steric obstruction of cellulases (Zhou et al. 2009). Despite the tremendous advances in high-resolution imaging of acid-treated plant cell walls (Singh et al., 2009; Donohoe et al., 2008), there is still much to be learned on understanding the rate-limiting steps that affect cell wall hydrolysis, especially for alkaline-based pretreatments (like AFEX).

2.2.3

Glycosyl hydrolases necessary for saccharification are dependent on type of pretreatment, thermochemical severity and cell wall composition

There are essentially three classes of cellulase enzymes (typically extracellular fungal enzymes) that have been extensively studied over the past five decades (Bhat and Hazlewood, 2001; Bayer et al., 1998); namely: (a) Cellobiohydrolases (CBH) or exoglucanases that act at the ends (reducing or non-reducing) of cellulose, processively cleaving cellobiose from the glucan chain ends. These enzymes typically belong to glycosyl hydrolase (GH) family 6, 7 and 9. (b) Endoglucanases (EG) act randomly to hydrolyze easily accessible, interior -1,4-glucan linkages of the cellulose chain, breaking it into smaller units and providing more `ends' for the exo-enzymes to act on. These enzymes typically belong to GH families 5, 7, 9, 12, 45 and 61. (c) -Glucosidases which hydrolyze cellobiose and short chain oligosaccharides into monomeric glucose units. These enzymes typically belong to GH family 1 and 3. Though, there have been several studies to optimize cellulase mixtures for hydrolyzing model cellulosic substrates like avicel and bacterial microcrystalline cellulose (Baker et al., 1998; Boisset et al., 2001), very little work has been conducted on pretreated lignocellulosic biomass. Baker and colleagues (1998) found that a ternary combination of CBH I: CBH II: EG I at 60 : 20 : 20 (0.4 mg/g glucan total enzyme loading) gave the highest glucan conversion for crystalline cellulose (i.e. sigmacell). Boisset and colleagues (2001) found a ternary combination of Cel7A: Cel6A: Cel 45A (native Humicola enzymes, analogous to CBH I, CBH II and EG I, respectively) at 68.75:30:1.25 (100 mg/g glucan total enzyme loading) gave the highest glucan conversion on crystalline cellulose. These two findings, among many others, indicate that the complex nature of synergy between similar endo- and exo-enzymes is dependent on the

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

31

total protein loading, substrate characteristics and individual protein binding properties to substrate active sites. Rosgaard et al. (2007) have studied the synergistic hydrolysis of acid, steam and hot-water pretreated barley straw using quaternary mixtures of purified CBH I, CBH II, EG I and EG II. The optimal ratio of the four enzymes required for higher severity (acid impregnation-steam explosion) pretreated straw was different from the lower severity pretreated straws (water impregnation-steam explosion and hot water extraction). The higher severity pretreated straw, required lesser amount of EG I (and greater amount of CBH II, with nearly constant CBH I) compared to lower severity material. The optimal ratio of CBH I: CBH II: EG I (Trichoderma enzymes) for high and low severity steam explosion pretreated barley straw was found to be 50:25:25 and 45:20:35, respectively. This is not surprising considering that acid pretreatment results in removal of amorphous cellulose giving a more crystalline polymer that requires lesser synergy between cellobiohydrolase (CBH I) and endoglucanases (EG I) (VaÈljamaÈe et al., 1999). Rosgaard et al. also found significant interaction effects between EG (EG I or EG II) with CBH I in the hydrolysis of lower severity pretreatments, possibly due to cross-activity of EG I on xylan (Gao et al., 2010). The complete degradation of side-chain decorated hemicelluloses requires the concerted action of several enzymes (Saha, 2003; Bhat and Hazlewood, 2001). The important classes of hemicellulose degrading enzymes are as follows: (a) Endoxylanases (EX) hydrolyze interior -1,4-xylosidic linkages of the xylan backbone. These enzymes typically belong to GH family 10 (acidic pI, high molecular weight) and 11 (basic pI, low molecular weight). (b) -Xylosidase ( X) hydrolyzes xylobiose dimers and short chain xylooligosaccharides to xylose. These enzymes typically belong to GH family 3. (c) -Arabinofuranosidase hydrolyzes terminal non-reducing -arabinofuranose from arabinoxylan side-chains. These enzymes typically belong to GH family 51, 54 and 62. (d) -Glucuronidase releases glucuronic acid from glucuronoxylan sidechains. These enzymes typically belong to GH family 67. (e) Acetyl xylan esterases hydrolyzes acetylated ester linkages from the xylan backbone. These enzymes typically belong to carbohydrate esterase family 1. (f) Phenolic acid esterases hydrolyze feruloyl and p-coumaroyl ester bonds in lignin-hemicellulose complexes. These enzymes typically belong to carbohydrate esterase family 1. Purified accessory hemicellulase synergy studies have been conducted for low severity, hot-water pretreated corn stover (Selig et al., 2008b; Knoshaug et al., 2008). It was found that glucan conversions to cellobiose (enzyme loading = 15 and 50 mg CBH I/g glucan) by CBH I increased by 13±84% in the presence of a suitable acetyl xylan and/or feruloyl esterase (enzyme loading = 2.5 mg/g

ß Woodhead Publishing Limited, 2010

32

Bioalcohol production

glucan). The greatest synergistic enhancements to glucan and xylan conversion were observed at lower cellobiohydrolase loading (15 mg/g glucan). Despite, significant removal of hemicellulose during dilute acid pretreatment, hemicellulases have been shown to significantly enhance the rate of hydrolysis (Selig et al., 2008a). This would suggest a significant enhancement in the hydrolysis È hgren et al., 2007) and alkaline (e.g. AFEX)rates for lower acid severity (O based treatments (Chundawat et al., 2006), containing a significant fraction of unhydrolyzed hemicellulose via supplementation of suitable hemicellulases and other accessory enzymes (Selig et al., 2008b). Recent findings for AFEX treated corn stover have shown synergistic interactions between cellulases (CBH I, CBH II and EG I) and hemicellulases (EX, X), as function of total enzyme loading, that maximize both glucan and xylan digestibility by several folds (Gao et al., 2010). Most commercially produced enzymes (typically from Trichoderma) do not have the correct ratio and total amounts of critical enzymes necessary to hydrolyze pretreated lignocellulosic biomass, since there has been no evolutionary pressure on microbes to grow on pretreated lignocellulosics (pretreatment also modifies the cell wall ultra-structure) to produce the optimum ratios of cellulases and hemicellulases (Rosgaard et al., 2007; Eriksson et al., 2002b). This is one of the reasons that crude protein mixtures should be optimized for maximizing sugar yield from pretreated biomass (Chundawat et al., 2008a). This would eventually lead to the construction of multigene expression systems that produce a balanced set of necessary enzymes for a particular feedstock-pretreatment combination. Several studies have looked at optimizing commercially available crude enzyme mixtures on acid pretreated (Berlin et al., 2007), steam exploded and organosolv treated lignocellulosic biomass (Berlin et al., 2005; 2006; Gusakov et al, 2007). Optimizing crude enzyme mixtures for acid pretreated corn stover helped reduce total protein loading by two-fold for equivalent hydrolysis yields (Berlin et al., 2007). Several covalent lignin-polysaccharide bonds are not entirely cleaved during alkali pulping and require accessory enzymes to help cleave these linkages, improving pulp digestibility (Buchert, 1992). Studies with steam exploded wheat straw have found synergistic interactions between crude cellulases and accessory enzymes (xylanase and esterases) (Tabka et al., 2006). However, it would be difficult to study the interaction of glycosyl hydrolases on pretreated lignocellulosics using crude enzyme mixtures. Purification of individual enzyme components to high purity has helped elucidate the role of cellulase and hemicellulase synergy during hydrolysis of acid and ammonia pretreated biomass (Selig et al., 2008b; Gao et al., 2010).

2.3

Types of chemical pretreatment

Pretreatments can be classified based on the nature of cell wall disruption (i.e. physical or chemical) during the process into four categories (Sun and Cheng,

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

33

2002; Mosier et al., 2005b; da Costa Sousa et al., 2009); (a) physical, (b) solvent fractionation, (c) chemical and, (d) biological-based pretreatments.

2.3.1

Physical pretreatment

Physical pretreatments disrupt the lignocellulose structure with little or no chemical modifications to the individual cell wall components. Some typical methods for physical pretreatment include biomass particle size reduction through communition, including dry, wet, vibratory and compression-based ball milling (Millett et al., 1975; Sidiras and Koukios, 1989; Tassinari et al., 1980; Alvo and Belkacemi, 1997). Physical treatments help enhance enzyme digestibility by increasing the cell wall accessible surface area to volume ratio. In some cases, extensive ball milling results in decrystallization of cellulose and reduction in degree of polymerization (Fan et al., 1981). Particle size reduction alone is not a sufficient pretreatment to significantly enhance the rate of enzymatic hydrolysis. The cost of size reduction also increases exponentially as the desired particle size decreases, making the process economically unfeasible in a commercial scenario (McMillan, 1994). However, some sort of minimal particle size reduction is necessary prior to most thermochemical pretreatments to improve material handling during processing.

2.3.2

Solvent fractionation-based pretreatment

Solvent fractionation using various solvents that can selectively solubilize cellulose, hemicellulose and lignin can also be used to pretreat lignocellulosic biomass. Cellulosic solvents like hydrazine, dimethyl sulfoxide (DMSO) and concentrated mineral acids can disrupt hydrogen bonding between cellulose microfibrils to solubilize crystalline cellulose (Heinze and Koschella, 2005). Solvent fractionation-based pretreatments may be further classified into the following three categories, based on the solvent system used. Organosolv process Historically, the organosolv process was investigated largely from the perspective of paper production (via pulping) from hardwoods. Organosolv process includes extracting lignin from lignocellulosic biomass using organic solvents like aromatic alcohols (phenols) or aliphatic alcohols (e.g. ethylene glycol, methanol, ethanol, butanol, glycerol) typically with an acidic catalyst (Sidiras and Koukios, 2004; Pan et al., 2006; 2007; Sun and Chen, 2008). The effect of organosolv-based process parameters (e.g., solvent composition, temperature, liquid to solid loading) on fractionation of the major components in woody biomass has been extensively studied over the years (X Zhao et al., 2009). Typical ethanol-based organosolv pretreatment parameters have the following

ß Woodhead Publishing Limited, 2010

34

Bioalcohol production

ranges; temperatures (90±120 ëC for grasses and 155±220 ëC for hard/soft woods), reaction times (25±100 min), acid catalyst loading (0.5±2%), and ethanol concentrations (25±75% v/v). Compared to aliphatic alcohols, aromatic alcohols were found to be more effective in solubilizing lignin from different feedstocks (Lee et al., 1987). In addition to alcohols, various amines have been used to delignify lignocellulosic biomass, but amines also tend to solubilize substantial amounts of carbohydrates in addition to lignin. An ethanol-based organosolv process has been commercialized by Lignol Innovation Corporation (Vancouver, Canada) to separate lignin, hemicellulose and cellulose fractions from woody biomass (Arato et al., 2005). The insoluble cellulose fraction is subjected to enzymatic hydrolysis to produce fermentable sugars, while the liquor from the organosolv step can be further processed to recover lignin, furfural, xylose, acetic acid and lipophylic extractive-based fractions. Ethanol used in the process is distilled and reused. The lignin fraction recovered can be used as an additive binder among other applications. In the case of poplar, about 88% of the total cellulose is recoverable as monomeric glucose after 100 hours of enzymatic hydrolysis of the residual solid fraction. Fractionation using phosphoric acid Fractionation of lignocelluloses to amorphous cellulose, hemicellulose, lignin and acetic acid using concentrated phosphoric acid, acetone and water-based mixtures is novel solvent fractionation-based pretreatment (Zhang et al., 2007). Phosphoric acid-based pretreatments may be carried out at moderate reaction conditions (i.e. 50 ëC, 1 bar), which minimizes acid catalyzed degradation of polysaccharides. The process gave 98% and 89% recovery yield for glucose and xylose, respectively, upon fractionation of corn stover. The selective fractionation of cell wall components is possible due to significant difference in solubility of cellulose, hemicellulose and lignin in phosphoric acid-acetone-water mixtures. The organic solvents are easily recoverable due to difference in volatility with respect to phosphoric acid (Zhang et al., 2006a; 2007; Moxyley et al., 2008). Despite several advantages associated with this pretreatment method, which include enhanced rate of amorphous cellulose hydrolysis and low pretreatment utility costs (e.g. 50 ëC vs. 100±200 ëC for other thermochemical treatments); there are several technical challenges (e.g., solvent cost and recovery, xylose recovery and fermentation) that need to be addressed to make this process commercially viable. Other issues include phosphoric acid/sugar separation, acid recovery, and acid re-concentration. Ionic liquid-based fractionation Ionic liquids (IL) are non-volatile solvents, under atmospheric conditions, composed exclusively of ions held together by coulombic forces. The first reported

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

35

IL that could solubilize cellulose was back in 1934 using molten N-ethylpyridinium chloride, in the presence of nitrogen-containing bases (Graenacher, 1934). However, the application of IL as lignocellulose pretreatment catalyst has gained momentum in recent years (Swatloski et al., 2002). Cellulose-dissolving IL usually contains anions of chloride, formate, acetate or alkyl phosphonate, since these ions form strong hydrogen bonds with cellulose. During a typical IL pretreatment of lignocellulosics, the biomass added to ionic liquids (ratio of 1 : 10±15, wt/wt) is heated (50±150 ëC) to solubilize the cellulosic, hemicellulosic and lignin components (Singh et al., 2009). The un-dissolved residue is filtered from solution and anti-solvents (e.g., water/methanol/ethanol) are added to the solution to recover the solubilized cellulose (as amorphous cellulose). The mass recovery of lignin and hemicellulose is currently unknown for corn stover (Singh et al., 2009). IL can be recovered from the anti-solvent by flash distillation and re-used in the process (Joglekar et al., 2007; Rinaldi et al., 2008, Dadi et al., 2006; 2007). Typically, ionic liquids that incorporate anions are found to be more effective in solubilizing cellulose, due to stronger hydrogen bonding between the anion and cellulose hydroxyl groups. On the other hand, IL with non-coordinating anions and long-chain substitution are poor cellulose solvents. IL with chloride anions appear to be the most effective solvents, while IL with aromatic side chains require higher temperatures to solubilize cellulose, due to their higher melting points and viscosities (H Zhao et al., 2009). While most of the reports on IL-based pretreatments are on processed cellulose (e.g., Avicel, Sigmacell), there are recent publications that have pretreated woody biomass (Kilpelainen, 2007; Lee et al., 2008). The solubilization efficiency of hard woods in IL was found to vary in the following order; ball-milled wood powder > sawdust > thermomechanical pulp > wood chip. Wood flour solubilization was tested using [Emim]+[CH3COO]± (1-ethyl-3methylimidazolium acetate)-based IL as a function of pretreatment time and temperature. Increasing residence time and IL treatment temperature caused a significant decrease in cellulose crystallinity; while improving lignin solubility (20±60% solubilized) and enhancing regenerated biomass enzymatic digestibility. Greater than 90% glucan digestibility was achieved using commercial enzymes after 24 hours of hydrolysis. There are several concerns that need to be addressed to facilitate commercialization of IL-based pretreatments. These include cost of ionic liquids, recovery of IL after pretreatment, hemicellulose recovery yield and enzymatic/microbial compatibility of IL (Turner et al., 2003).

2.3.3

Chemical pretreatment

Acidic-based pretreatments Most acidic-based pretreatments (e.g., dilute acid, steam explosion and liquid hot water treatment) have very similar chemistries but vary in thermochemical

ß Woodhead Publishing Limited, 2010

36

Bioalcohol production

severity. The acidic catalyst is either supplemented externally (as mineral acids) or is formed due to hydrolysis of hemicellulose acetyl linkages (forming acetic acid) and degradation of polysaccharides/lignin to short-chain aliphatic acids and phenolic acids. Water itself behaves as an acid at high temperatures typically employed in most pretreatments (Weil et al., 1998). The acid is responsible for the hydrolytic cleavage and removal of hemicellulose and lignin, hence improving accessibility of residual cellulose to glycosyl hydrolases. The extent of removal of hemicellulose and lignin varies considerably between various acidic pretreatments depending on the severity of pretreatment. The severity of acidic pretreatments can be quantified using a simple relationship which helps couple the effect of time (t), temperature (T) and acidity (pH) into a single fudge factor known as the `combined severity factor (S)' (Chum et al., 1990). log…S† ˆ log…R0 † ÿ pH   …T ÿ 100† R0 ˆ t  exp 14:75 Higher acid pretreatment severities, referring to longer residence times, higher temperatures and lower pH values, result in extensive hydrolysis of hemicellulose to monomeric sugars. However, higher pretreatment severity also results in the degradation of lignin and hemicellulose derivatives to compounds like furfural, 5-hydroxymethylfurfural, levulinic acid, phenolic acids/aldehydes and other aliphatic acids that can severely inhibit downstream biological processing (Klinke et al., 2004). Most acidic-based pretreatments look to optimize severity by trading-off between maximizing cellulose enzymatic digestibility, hemicellulose acid catalyzed hydrolysis to xylose and minimizing formation of biological inhibitors. Dilute-acid pretreatment (using either sulfuric, hydrochloric or nitric acid) was one of the first pretreatment methods implemented to pretreat lignocellulosic biomass in order to produce ethanol (Ruttan, 1909). Traditionally, dilute acid-based hydrolysis of lignocellulosic biomass has been used to manufacture furfural through hydrolysis and dehydration of hemicelluloses (Zeitsch, 2000). The remaining cellulosic fraction was typically hydrolyzed to monomeric glucose using concentrated acid (10±30%) prior to fermentation (Harris and Begliner, 1946). However, with the development of genetically modified microbes that can co-ferment glucose and xylose it became necessary to prevent extensive degradation of hemicellulose to furfural to increase ethanol yields. This has resulted in the development of dilute-acid-based pretreatments, typically using sulfuric acid, carried out at lower acid concentrations (0.05±5%) and temperatures (160±220 ëC) (Kim et al., 2001; Torget et al., 1992). Diluteacid pretreatment can be carried out in a stationary batch mode or flow-through continuous mode (Lloyd and Wyman, 2005; Lee et al., 1999).

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

37

Steam explosion is one of the most widely implemented pretreatment methodologies to pretreat several different types of lignocellulosic biomass (e.g., agricultural residues, hardwoods and softwoods). Steam explosion is typically carried out by rapidly heating lignocellulosic biomass with high pressure saturated steam with/without external chemical addition (typically acids) for certain duration of time (ranging from a few seconds to minutes), at a fixed temperature (160±290 ëC) prior to explosively releasing the pressure (Ballesteros et al., 2006; Ballesteros et al., 2000; Chandra et al., 2007). There have been several reported variants of the steam explosion process. These variants typically incorporate pre-impregnation with an acid (like sulfuric acid, sulfur dioxide) or alkali (like sodium hydroxide, ammonia) prior to steam explosion (Chen et al., 2005; Playne, 1984; Brownell and Saddler, 1984; Hsu, 1996). Liquid hot water pretreatment (also known as hydrothermal or aquasolv treatment) is a low severity acid-based pretreatment that uses water at high pressures (>5 MPa) to maintain the liquid state at elevated temperatures (160± 230 ëC) to pretreat lignocellulosics (Bobleter et al., 1981; Mosier et al., 2005b; Weil et al., 1998). The liquid water is contacted with the biomass in three possible modes; co-current, counter-current and flow-through (Mosier et al., 2005a,b,c). Recent variants to the liquid hot water pretreatment process allows for better pH control (ranging from pH 4±7) that limits non-specific degradation of polysaccharides (Mosier et al., 2005b). Carbonic acid, formed by dissolved carbon dioxide in water, has also been used as a pretreatment catalyst (van Walsum et al., 2007). In addition to mineral acids (like sulphuric acid, nitric acid), organic acids (carbonic, acetic, succinic, fumaric, maleic, citric acid) have been used (at 50 mM concentration) as pretreatment catalysts (Mosier et al., 2001; 2002). It was reported that maleic and fumaric acids possess superior selectivity for the production of fermentable sugars from cellulose than sulfuric acid (Kootstra et al., 2009). Both sulfuric and maleic/fumaric acid have the ability to hydrolyze -(1-4)-glycosidic linkages. But, at 150 ëC sulfuric acid degrades glucose/arabinose to undesirable byproducts, while maleic/fumaric acid does not. For example, corn stover treated at high solids loading (150±200 g/liter) by sulfuric acid results in 30% degradation of xylose. Maleic acid gave ~95% monomeric xylose yield with trace amounts of furfural, along with 90% glucose yield after enzymatic hydrolysis of the xylan depleted substrate (15 FPU cellulase/g glucan). The resulting unconditioned hydrolyzate was fermented by recombinant S. cerevisiae giving 87% theoretical ethanol yield. Since organic acids are weaker acids than mineral acids, protons are only partially dissociated under comparable reaction conditions. This could be one possible reason for the selective catalytic activity of organic acids compared to mineral acids. One of the major concerns is the recovery of organic acids. Suitable economic analyses have to be carried out in order to determine the feasibility of using di-carboxylic acids vs. mineral acids within a cellulosic biorefinery (Lu and Mosier, 2007; 2008).

ß Woodhead Publishing Limited, 2010

38

Bioalcohol production

Alkali-based pretreatments Alkali-based pretreatments typically differ in the type and thermodynamic state of the catalyst-solvent systems employed. Alkalis used for pretreatment typically include calcium hydroxide, ammonia and sodium hydroxide. Most alkali pretreatments help lower recalcitrance of lignocellulosics through saponification (or ammonolysis in the presence of liquid ammonia) of hemicellulose acetyl and lignin-carbohydrate complex linkages (Chang and Holtzapple, 2000; LaureanoPerez et al., 2005; Weimer et al., 1986). The extraction of lignin and hemicellulose from the biomass during most alkaline pretreatments also helps reduce non-specific binding of enzymes during cellulose enzymatic hydrolysis (Kim and Lee, 2005a). Ammonia fiber expansion (AFEX) is a unique alkaline pretreatment method that uses concentrated liquid ammonia-water mixtures to pretreat lignocellulosics (Dale et al., 1996; Dale and Moreira, 1982). AFEX is a low temperature process (60±140 ëC) that is carried out by adding liquid ammonia (0.3±2 kg ammonia per kg dry weight biomass) to pre-wet biomass (0.05±1 kg water per kg dry weight biomass) in a pressurized reactor that is cooked for 5±45 minutes before explosively releasing the ammonia. AFEX is a dry to dry process, unlike most other pretreatments, resulting in no separate liquid stream at the end of the pretreatment. The volatility of ammonia allows for easy recovery and reuse within a continuous process (Sendich et al., 2008). AFEX results in de-acetylation and cleavage of lignin-carbohydrate complexes through base catalyzed hydrolysis (Laureano-Perez et al., 2006). The extent of ammonolysis of ester linkages within the cell wall after AFEX is poorly understood (Chou, 1986; Chundawat, 2009; Weimer et al., 1986). Liquid ammonia is also thought to decrystallize cellulose (typically via formation of a more swollen crystal structure known as cellulose IIII from native cellulose I ) as shown by X-ray diffraction studies (Lewin and Roldan, 1971; Wada et al., 2006). Interestingly, cellulose III has been shown to have 3±5-fold higher rate of enzymatic hydrolysis compared to native cellulose I (isolated from Cladophora algae cell walls). The extent of cellulose III formation during AFEX (and related ammonia-based pretreatments) treatment of lignocellulosic biomass is being currently investigated (Chundawat, 2009). On-going research also indicates significant modification of the cell wall ultra and macro-structure during AFEX that results in enhanced enzymatic accessibility (unpublished data). An improved fundamental understanding of the mechanism of ammonia-based pretreatments would allow novel process modifications to further enhance the overall rate of enzymatic hydrolysis and fermentation of plant cell walls. There have been several variants of the ammonia treatment process that have been reported in the literature; such as supercritical ammonia treatment, ammonia recycle percolation (ARP), soaking in aqueous ammonia (SAA) and ammoniaperoxide pretreatment (Weimer et al., 1986; Kim and Lee, 1996; 2005a; 2005b;

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

39

Kim et al., 2003; Hennessey et al., 2007). Most of these pretreatments vary in the thermodynamic state of ammonia-water mixtures for varying ammonia concentrations. ARP pretreatment is carried out in a flow-through, recycle mode by percolating ammoniacal solutions (5±15% concentration) through a column reactor packed with biomass under high pressure (2±3 MPa) and temperatures (160±180 ëC). On the other hand, calcium hydroxide or lime-based pretreatments are typically carried out at 50±120 ëC, for 1 hr to 4 weeks using 0.05±0.5 g Ca(OH)2/g biomass and 2±10 g water/g biomass. Typically, lime treatment is also conducted in the presence of air or pressurized oxygen to facilitate lignin degradation and removal (Kim and Holtzapple et al., 2005; 2006). Oxidative decomposition of lignin in poplar was shown to be facilitated by high pressure oxygen supplementation during conventional lime pretreatments (Sierra, et al., 2009). A significant amount of lime is consumed in the process at elevated temperatures. However, the recovery of the solubilized catalyst in the liquor is possible (after neutralization to calcium carbonate) via integration to a suitable lime kiln technology (Kaar and Holtzapple, 2000). Unlike AFEX, there is a significant extraction of lignin and hemicellulose in other alkaline pretreatments like ARP and lime-based pretreatments. Oxidative pretreatments Ozone has also been used an oxidizing agent to break down lignin within lignocellulosic biomass to help increase enzymatic digestibility (Puri, 1983; Quesada et al., 1999; Garcia-Cubero et al., 2009). Traditionally, ozone had been used to bleach pulps replacing chlorine as the oxidant (Roncero et al., 2003). Ozonolysis, typically conducted at atmospheric conditions, has been used in combination with other alkaline pretreatments to extract lignin from forage residues and hence enhance enzymatic digestibility (Akin and Morrison, 1988; Ben-Ghedalia and Miron, 1981; Morrison et al., 1991). Ozone reacts selectively with lignin-based aromatics producing several types of degradation products (primarily aromatic aldehydes and aliphatic organic acids) that have been found to inhibit biological processing of ozone treated biomass (Quesada et al., 1997). Ozonolysis-based intermediate degradation products (like free radicals) are responsible for unwanted side-reactions, like polysaccharide degradation (Jablonsky et al., 2004; Ragnar et al., 1999; Shatalov and Pereira, 2007). There is a lack of ozonolysis pretreatment-based data for various agricultural residues (Silverstein et al., 2007). The possible inhibitory effects of various ozonolysis degradation products are currently poorly understood. However, the biggest hurdle for an economically feasible ozonolysis-based pretreatment is the cost of ozone. Alkaline wet oxidation is another form of oxidative pretreatment carried out typically under alkaline conditions at high temperatures (170±220 ëC) using pressurized air/oxygen or hydrogen peroxide as the oxidant (Martin and Thomsen, 2007; Varga et al., 2003; McGinnis et al., 1983). Alkalis typically

ß Woodhead Publishing Limited, 2010

40

Bioalcohol production

employed for wet oxidation include sodium carbonate (0.02±0.05 g Na2CO3/g dry biomass) and calcium hydroxide to help solubilize the hemicellulose and lignin fraction more effectively (Klinke et al., 2002; Martin and Thomsen, 2007; Sierra et al., 2009). The advantage of an alkaline medium during treatment at elevated temperatures is to minimize formation of various furan-based degradation products.

2.3.4

Biological pretreatments

Most chemical pretreatments are carried out at high temperature/pressure-based reaction conditions, and thus require significant capital investment (Eggeman and Elander, 2005). High thermochemical severity also results in the formation of several degradation products that are inhibitory to downstream biological processing (i.e. enzymatic hydrolysis and fermentation). Contrary to chemicalbased pretreatments, biological pretreatments (using enzymes or microbes as pretreatment catalysts) would consume lesser energy since they are carried out under milder reaction conditions (Lee, 1997; Keller et al., 2003; Lee et al., 2007). Interestingly, termites enhance digestibility of lignocellulosic biomass to gut microbiota via exposure to low temperature based oxidative alkaline pretreatment like conditions within the hind-gut (Brune, 1998). However, most industrially relevant biological treatment processes are slow (reaction time ranges between several hours to months) and have low-throughput. Biological pretreatments are typically carried out by inoculating the biomass with fungal spores (e.g., white rot basidiomycetes and even certain actinomycetes) or externally supplementing accessory enzymes (e.g., ferulic acid esterases and hemicellulases). White-rot fungi have been found to be effective to degrade lignin while minimizing polysaccharide consumption (Sun and Cheng, 2002; Kerem et al., 1992). During microbial pretreatments, a substantial amount of delignification can take place, possibly at the expense of polysaccharide consumption depending on the microbial source (Kerem et al., 1992). White-rot microbes typically secrete lignin peroxidases, along with various types of glycosyl hydrolases, that cleave the C±C lignin backbone in the presence of hydrogen peroxide. Other enzymes involved in aerobically catalyzed lignin degradation include Mn-dependent peroxidases, laccases (monophenol oxidase) and superoxide dismutase (Leonowicz et al., 1999). Details about the lignin degrading microbial systems have been summarized elsewhere (Lee, 1997; Leonowicz et al., 1999). In recent years, the use of ferulic acid esterases and other accessory enzymes as catalysts to cleave LCC linkages within plant cell walls to help reduce thermochemical severity during subsequent pretreatment and/or to maximize recovery of valuable phenolic by-products have been explored (Anderson et al., 2005; Akin et al., 2006). Several articles on microbial pretreatments using solid state fermentation (SSF) have been published for both grasses and hard woods (MendoncËa et al., 2008; Lee

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

41

et al., 2008). The effect of substrate moisture content, inorganic salt concentration, culture time on lignin degradation, solids recovery, and availability of carbohydrate on the biological pretreatment process have been explored. The solid state cultivation of cotton stalks using Phanerochaete chrysosporium, at 75% moisture content without salts was the preferable pretreatment condition resulting in 28% lignin degradation, 71% solids recovery and 42% availability of carbohydrates over a period of 14 days (Shi et al., 2008). Another approach is to use minimally treated mushroom spent straw (MSS) as a feedstock for downstream thermochemical and biological processing (Balan et al., 2008). When MSS was used without any pretreatment, the glucan digestibility was around 40%, using standard commercial cellulases (15 FPU/g glucan); and 20% for untreated rice straw. Further thermochemical pretreatment is necessary for getting higher sugar yields from MSS. By using microbial pretreated biomass as a feedstock for thermochemical pretreatments; it might be possible to lower processing costs by reducing pretreatment severity and minimizing chemical usage while obtaining higher overall hydrolysis yields. Unfortunately, there have been no detailed economic studies comparing the feasibility of scaling-up and integrating biological-based pretreatments within a conventional cellulosic biorefinery.

2.4

Comparing effectiveness of leading pretreatments on corn stover and poplar

A Consortium for Applied Fundamentals and Innovation (CAFI) for improving understanding and comparing leading pretreatment technologies was formed in 2000. The CAFI project has allowed a systematic comparison of pretreatments on common feedstocks (e.g., corn stover, poplar, switchgrass) using comparable standard methodologies. Some of the leading pretreatments that are currently part of the CAFI project include dilute acid, steam explosion (with sulfur dioxide), controlled pH liquid hot water, ammonia fiber expansion (AFEX), ammonia recycle percolation (ARP) and lime-based pretreatments. Table 2.1 summarizes the optimum pretreatment conditions; inclusive of reaction time, catalyst and water loading, for each pretreatment that resulted in optimal enzymatic hydrolysis yields for corn stover and poplar (Adapted from È hgren et al., 2005; 2007; Kumar et al., 2009; Wyman et al., 2005; 2006). Most O acidic pretreatments (except ARP) typically required high temperatures (170± 200 ëC) to effectively pretreat corn stover. However, for lignin-rich poplar all pretreatments required much higher temperatures (>150 ëC) and chemical loading for maximizing enzymatic digestibility. AFEX employs lower temperatures (90±160 ëC) and water loadings (0.6±3.2 g/g dry biomass) compared to other pretreatments. Ongoing process improvements have further allowed significant reduction in ammonia usage during AFEX (Sendich et al., 2008). Some of the common effects of CAFI-based pretreatments on the physicochemical properties of corn stover and poplar are shown in Table 2.2. AFEX and

ß Woodhead Publishing Limited, 2010

Table 2.1 Optimum pretreatment conditions for several leading CAFI-based thermochemical pretreatments that maximize enzymatic digestibility of pretreated corn stover and poplar (data for poplar shown in parentheses)

ß Woodhead Publishing Limited, 2010

Pretreatment type

Temperature (ëC)

Reaction time (min)

Catalyst

Catalyst loading (g/g dry biomass)

Water loading (g/g dry biomass)

Dilute acid Sulfur dioxide

160 (190) 190 (190)

20 (1.1) 5 (5)

Sulfuric acid Sulfur dioxide

0.015 (0.02) 0.03 (0.03)

3 (3.3) 4 (4)

Controlled pH/ hot water AFEX

190 (200)

15 (10)

±

±

5.25 (5.67)

90 (180)

5 (10)

Ammonia

1 (2)

0.6 (2.3)

ARP

170 (185)

10 (27.5)

Ammonia

0.5 (0.55)

2.8 (3.2)

Lime

55 (160)

4 weeks (120)

Calcium hydroxide

0.07 (0.2)

10 (1.6)

Specific notes

In batch mode using a parr reactor Soaked overnight in 3% acid solution prior to treatment In flowthrough mode (treated poplar washed with hot water) Liquid ammonia added to moist biomass prior to heating reactor Flowthrough mode using 5 ml/min of ammoniacal solution, 15% w/w w/wo purging with air (purged with oxygen at 200 psi, 39% solids)

Adapted from Úhgren et al. (2005, 2007); Wyman et al. (2005, 2009). Note: lime pretreatment for poplar was performed in the presence of pressurized O2.

Table 2.2 Prominent physicochemical effects of various leading thermochemical pretreatments on corn stover/poplar (data for poplar shown in parentheses). The cellulose crystallinity index for both untreated corn stover and poplar is 50 units each. The degree of polymerization for untreated corn stover and poplar cellulose is 7000 and 3500, respectively. The percent acetyl removed value for AFEX corn stover (marked by *) is lower than expected based on ongoing investigations (unpublished data; Chundawat, 2009) ß Woodhead Publishing Limited, 2010

Pretreatment type

Dilute acid Sulfur dioxide-based steam explosion Controlled pH liquid hot water Ammonia fiber expansion (AFEX) Ammonia recycle percolation (ARP) Lime

Cellulose crystalinity index

% cellulose removed

% hemicellulose removed

% lignin removed

Cellulose degree of polymerization

% acetyl groups removed

53 (51) nd (56)

5±10 (10±15) 3±5 (1±5)

70±75 (90±95) 40 (90±95)

18 (nd) 40±50 (nd)

2700 (500) 3000 (650)

55 (90) 55 (80±85)

45 (54) 36 (48) 26 (50)

5±10 (1±5) 0 1±5 (5±10)

40 (55±60) 0 50±60 (30±35)

nd (nd) 0 75±85 (40)

5600 (1800) 6600 (2700) 4600 (3200)

55 (70±75) 30±35* (70±80) 85±90 (85)

56 (55)

1±3 (1±3)

30±35 (3±5)

55±60 (50)

3200 (1600)

90±95 (95)

Adapted from Kim and Holtzapple (2005, 2006); Kim and Lee (2005a); Laureano-Perez et al. (2005); Lloyd andWyman (2005); Mosier et al. (2005a); Úhgren et al. (2005, 2007); Teymouri et al. (2005); Wyman et al. (2009); and Kumar et al. (2009). nd = not determined or unknown. Note: lime pretreatment for poplar was performed in the presence of pressurized O2.

44

Bioalcohol production

ARP are seen to cause a significant reduction in the crystallinity of corn stover compared to other pretreatments. This is probably due to modification of cellulose I to III during AFEX or due to extraction of lignin and hemicellulose during ARP (Lewin and Roldan, 1971; Kim and Lee, 2005a). However, measurement of crystallinity by XRD (X-ray diffraction) is compounded by presence of residual amorphous lignin and hemicellulose (Park et al., 2009). Acidic pretreatments that typically hydrolyze and extract the amorphous cellulose and hemicellulose components result in increasing the crystallinity as measured via XRD. The degree of polymerization of cellulose is reduced significantly after most pretreatments except AFEX (Kumar et al., 2009). Unlike other pretreatments, AFEX is a dry-to-dry process with no secondary liquid stream being generated at the end of the treatment. Most pretreatments (except AFEX) result in significant removal of cellulose (1±10%), hemicellulose (30± 95%) and lignin (20±85%). The solubilized monomeric and oligomeric sugars in the liquid stream for acidic-based pretreatments are sensitive to thermally induced decomposition that results in the formation of furans and other degradation products (Chen et al., 2006). Almost all pretreatments resulted in extensive deacetylation of hemicellulose ranging between 30 and 95% (basis is theoretical acetyl content). The effect of thermochemical pretreatments on corn stover/poplar glucan and xylan hydrolysis yields at the end of pretreatment (Stage A) and enzymatic hydrolysis (Stage B) (using 15 FPU/g glucan cellulase loading) is shown in Table 2.3 (Adapted from Wyman et al., 2005; 2009). Most pretreatments perform reasonably well on corn stover with AFEX resulting in the highest total glucan (96%) and xylan conversion (91%). A significant amount of hemicellulose is solubilized as oligomeric sugars (20±60%) for most treatments in the liquid stream after pretreatment. The sugar oligomers must be hydrolyzed to monomeric xylose through acid or enzyme catalyzed hydrolysis. Unlike other treatments, AFEX retains all the hemicellulose in the biomass at the end of pretreatment which is hydrolyzed by hemicellulases during enzymatic hydrolysis resulting in much higher monomeric sugar yields. However, for poplar most acidic pretreatments perform better than alkaline pretreatments except oxidative lime treatment. Steam explosion catalyzed by sulfur dioxide resulted in higher glucan yields compared to dilute sulfuric acid treatment for poplar. Unlike corn stover, a significant amount of hemicellulose was hydrolyzed to monomeric sugars for poplar during acidic-based pretreatments. This is probably due to the higher pretreatment severity employed for poplar compared to corn stover. Lime pretreatment was the only alkaline pretreatment that performed well on poplar. This is likely because pressurized oxygen was used during lime pretreatment that resulted in a more effective oxidatively catalyzed delignification of the cell wall compared to other alkaline treatments. Supplementation of xylanases was found to play a critical role in the hydrolysis of alkali pretreated poplar giving much higher

ß Woodhead Publishing Limited, 2010

Table 2.3 Effect of leading thermochemical pretreatments on corn stover (I) and poplar (II) glucan and xylan percent conversions (based on composition of untreated biomass) at the end of pretreatment (Stage A) and enzymatic hydrolysis (Stage B) (enzymatic hydrolysis conducted at 15 FPU/g glucan cellulase loading) % Glucan conversion for pretreated corn stover Pretreatment Stage A Enzymatic Stage B ß Woodhead Publishing Limited, 2010

Pretreatment type Dilute acid Controlled pH hot water AFEX ARP Lime

Monomeric

Oligomeric

Monomeric

Oligomeric

Monomeric

Oligomeric

Total

6 0 0 0 0

0 5 0 0 1

86 85 96 90 92

0 0 0 0 1

92 86 96 90 92

0 5 0 0 1

92 91 96 90 94

% Xylan conversion for pretreated corn stover Pretreatment Stage A* Enzymatic Stage B*

Dilute acid Controlled pH hot water AFEX ARP Lime

Total glucan conversion (A+B)

Total xylan conversion (A*+B*)

Monomeric

Oligomeric

Monomeric

Oligomeric

Monomeric

Oligomeric

Total

82 2 0 0 1

2 55 0 47 23

8 24 77 41 52

0 0 14 0 0

91 26 77 41 52

2 55 14 47 23

93 81 91 88 76

Table 2.3 Continued % Glucan conversion for pretreated poplar Pretreatment Stage A Enzymatic Stage B Pretreatment type ß Woodhead Publishing Limited, 2010

SO2 steam explosion Dilute acid (Sunds) Controlled pH hot water AFEX ARP Lime

Monomeric

Oligomeric

Monomeric

Oligomeric

Monomeric

Oligomeric

Total

3 24 0 0 0 0

0 0 2 0 1 0

97 63 54 53 49 90

0 0 0 0 0 5

100 87 54 53 49 90

0 0 2 0 1 5

100 87 56 53 49 96

% Xylan conversion for pretreated poplar Pretreatment Stage A* Enzymatic Stage B* Pretreatment type SO2 steam explosion Dilute acid (Sunds) Controlled pH hot water AFEX ARP Lime

Total glucan conversion (A+B)

Total xylan conversion (A*+B*)

Monomeric

Oligomeric

Monomeric

Oligomeric

Monomeric

Oligomeric

Total

54 63 4 0 0 0

20 0 54 0 37 5

9 9 38 52 31 65

0 0 0 0 1 8

64 72 42 52 31 65

20 0 54 0 38 12

84 72 96 52 69 78

Adapted from Wyman et al. (2005, 2009). Note: lime pretreatment for poplar was performed in the presence of pressurized O2.

Thermochemical pretreatment of lignocellulosic biomass

47

glucan and xylan conversions than has been reported in Table 2.3 (Kumar and Wyman, 2009a; 2009b). Comparison of glucose and xylose fermentability (by Saccharomyces cerevisiae LNH-ST 424A) of enzymatic hydrolyzates from various pretreatments for poplar results in varying ethanol metabolic yields (ranging from 85 to 100%) depending on the type of pretreatment (Lu et al., 2009). However, a more rigorous comparison between various leading pretreatment technologies is required to compare fermentability of the liquid extract (this is typically ignored) and solid fractions resulting from the pretreatment with and without detoxification at industrially relevant conditions (e.g., high solids loading to produce >4% w/w ethanol concentration). Recent findings comparing AFEX and dilute acid pretreated corn stover have shown that some amount of detoxification might be necessary for acidic-based pretreatments to prevent microbial inhibition (Lau et al., 2008b). While AFEX treated corn stover was found to be easily fermentable with no necessity for substrate water washing, detoxification or external nutrient supplementation (Lau et al., 2008a; 2008b; Lau and Dale, 2009). There is a significantly higher amount of potentially inhibitory degradation products (e.g., organic acids and phenolics) formed during dilute acid pretreatment compared to AFEX (Chundawat et al., 2007; 2008b). Without any detoxification, E. coli KO11 was unable to grow on acid treated corn stover hydrolyzate while Saccharomyces cerevisiae LNH-ST 424A performed poorly compared to growth on AFEX stover hydrolyzate (Lau et al., 2008b).

2.5

Characteristics of an ideal pretreatment

2.5.1

Factors affecting viability of any pretreatment technology

The efficacy of any pretreatment technology should not be evaluated exclusively based on the enzymatic digestibility of the pretreated solid biomass fraction (typically ignoring the liquid fraction). Some authors defend that this evaluation should be performed using a holistic perspective spanning technological, economical and environmental factors (da Costa Sousa et al., 2009). In this context, there are two important decisions to make before analyzing the efficacy of any pretreatment: 1) decide what factors are important to take into consideration, and 2) decide where to place the mass/energy boundaries to evaluate the role of each factor on the system. The factors that are important to take into consideration should not only comprise overall mass and energy balances for the process (inclusive of product yields, concentration and reaction rates), but should also encompass secondary factors that contribute to the economic and environmental impacts of the pretreatment. These secondary factors need to be analyzed within the boundaries of the study, which should be placed at the

ß Woodhead Publishing Limited, 2010

48

Bioalcohol production

radius of influence of the pretreatment, and possibly beyond the realm of biorefinery activity. From Figs 2.2, 2.3 and 2.4 it is possible to see how pretreatment can influence both pre- and post-processing steps such as, biomass type, harvesting conditions, milling, enzyme type and amount requirement, hydrolyzate preconditioning, microbial fermentation, by-product utilization, waste residue handling, and ethanol recovery. This makes pretreatment the central unit operation in a biorefinery with all-pervasive technological and economic implications on other processing steps. This kind of a cradle-to-grave assessment comparing pretreatments has been largely absent, until recently. The CAFI project was the first attempt towards this goal, allowing a systematic comparison of pretreatments (e.g., dilute-acid, AFEX) on common feedstocks (e.g., corn stover, poplar) using consistent analytical methods (e.g., same batch of commercial enzyme formulations). The technical and economic performance of all CAFI pretreatments for corn stover and poplar has been published in recent years (Wyman et al., 2005; 2009). However, CAFI considered only a few of the pretreatment technologies available for their study (dilute acid, liquid hot water, AFEX, ARP and limebased pretreatments). In a crude attempt to have a general idea about the state-of-the-art of the most important pretreatment technologies available today, Tables 2.4 and 2.5 were compiled using information available in the literature. These tables list some of the important factors to take into consideration when comparing pretreatment technologies. In this case, it is important to note that the pretreatment processes that were involved in CAFI project show data that have a comparable basis. However, this is not always true for other pretreatment technologies, since experiments were carried out using different feedstocks, enzyme mixtures, enzyme loadings and microbial fermentation conditions. The goal was to give a general perspective of the requirements of each pretreatment and study their effect on downstream biological processing. From Table 2.4, it is possible to find useful information about each type of pretreatment, such as the type of chemicals and the typical quantities used in the process, hazard information about these chemicals, their cost (ICIS, 2006), chemical recovery requirements, process conditions (temperature, pressure, residence time). Another aspect that should be taken into consideration is the `generality' of the pretreatment. Generality of a pretreatment refers to how the pretreatment performs on a variety of feedstocks, such as hardwoods, softwoods and grasses, in the dry or wet form. This is especially important when operational flexibility of a biorefinery is required in the presence of variable feedstocks. The effect of pretreatment method on enzymatic hydrolysis performance can be also analyzed from the typical glucan and xylan conversions (to monomeric and oligomeric sugars), enzyme loading and hydrolysis time. The pretreatment conditions and enzymatic hydrolysis performance are dependent on the type of

ß Woodhead Publishing Limited, 2010

ß Woodhead Publishing Limited, 2010

2.2 Material and energy flow balance for a generic pretreatment-based biorefinery unit operations from `cradle' (e.g., biomass cultivation) to `grave' (e.g., lignin residue and/or ethanol combustion).

ß Woodhead Publishing Limited, 2010

2.3 Detailed mass balance for generic unit operations involved in a second generation lignocellulosic biorefinery incorporating thermochemical pretreatment, enzymatic hydrolysis and microbial fermentation.

ß Woodhead Publishing Limited, 2010

2.4 Pretreatment is the center of the biorefinery `universe'. The figure depicts the strong inter-dependence between thermochemical pretreatment with other upstream/downstream operations and process-product economic viability.

Table 2.4 Important pretreatment-related parameters (e.g., chemical usage, hazards, cost of chemical, catalyst recovery, effluent waste, pressure/ temperature/reaction time) for biological and thermochemical pretreatments Pretreatment categorya

ß Woodhead Publishing Limited, 2010

Fractionation Organosolv (++)

Phosphoric acid (+++) Ionic liquids (++)

Chemical Dilute-acid (+++)

Steam explosion (+++)

Pretreatment-related parameters Chemical

Chemical usage/ Mg biomass

C2H5OH, 4:1 C2H5OH H2O (+ acid 2:1 H2O or base)

Chemical hazardsb

Chemicals costc $/Mg

H:2, F:3, R:0

$1300 (C2H5OH), $95 (H2SO4)

$450 (H3PO4) $1390 (CH3COCH3)

H3PO4, 13.5:1 H3PO4, H:2, F:0, H2O, 19:1 CH3COCH3 R:0/H:1, F:3, R:0 CH3COCH3 24:1 H2O Ionic liquids, H2O, CH3OH, C2H5OH

H2O, H2SO4

H2O, H2SO4 or SO2

10:1 Ionic liquid

N/A

0.03:1 H2SO4/ H:3, F:0, 4:1 H2O R:2, O:W

0.005:1 H2SO4/ 0.03:1 SO2

H:3, F:0, R:2, O:W/ H:2, F:0, R:0

Chemical recovery processes

% Catalyst recovery

Solid effluent

Temperature (ëC)

Pressure (psi)

Residence time

References

Precipitation/Filtration/ Distillation + Waste water treatment/Water recovery

97%

Yes

90±220

200±300

25±100 min

ICIS, 2006; Pan et al., 2006; Sidiras and Koukios, 2004; Sun and Chen, 2008; X Zhao et al., 2009

Distillation/Flash Separation + Waste water treatment/Water recovery

N/A

Yes

50

±

60±80 min

ICIS, 2006; H. Li et al., 2009; Zhang et al., 2007

$45000 (Ionic liquid)d

Active carbon adsorption + N/A Diethyl ether washing + Distillation +Chloroform/ methanol washing + activated alumina adsorption + Distillation + Waste water treatment/ Water recovery

Yes

100±150

±

0.5±2 hours

Dadi et al., 2007; ICIS, 2006; Q. Li et al., 2009; Wasserscheid and Haumann, 2006

$95 (H2SO4)

Acid neutralization (e.g. with lime) + Waste water treatment/Water recovery

N/R

Yes

160±220

30±220

2±30 min

Eggeman and Elander, 2005; ICIS, 2006; Lu et al., 2009; Mohagheghi et al., 1992; Schell et al., 2003; Wyman et al., 2005

$95 (H2SO4) $230 (SO2)

Acid neutralization (e.g. with lime) + Waste water treatment/Water recovery

N/R

Yes

160±290

200±350

5±15 min

ICIS, 2006; Li and Chen, 2008; H. Li et al., 2009; Varga et al., 2004; Wang et al., 2009

ß Woodhead Publishing Limited, 2010

Liquid hot water (++)

H2O

6:1 H2O

H:0, F:0, R:0

$0.50 (H2O)e

Waste water treatment/ Water recovery

AFEX (++)

H2O, NH3

1:1 NH3/ 0.6:1 H2O

H:3, F:1, R:0

$280 (NH3)

ARP (++)

H2O, NH3 0.47:1 NH3/ 2.7:1 H2O

H:3, F:1, R:0

$280 (NH3)

Lime (++)

Ca(OH)2, air/O2 (optional)

0.75:1 Ca(OH)2/ 10:1 H2O

H:1, F:0, R:1

Ozonolysis (+++)

H2O, O3

0.027:1 O3/ 0.3:1 H2O

Not rated by NFPA

Alkaline wet oxidation (++)

H2O, O2 H2O2, Na2CO3

1.2 MPa O2, H:3, F:0, 0.03:1 R:0, O:OX/ Na2CO3, H:1, F:1, R:2 15:1 H2O

Biological Fungi or bacteria (+)

N/R

N/R

N/A

N/R

Yes

160±230

350±400

15±20 min

Ballesteros et al., 2002; Eggeman and Elander, 2005; ICIS, 2006; Laser et al., 2002; Lu et al., 2009; Mosier et al., 2005a

Distillation/Condensation/ 97±99% Water Quenching/ Compression/ Water recovery

No

60±140

200±600

5±15 min

Alizadeh et al., 2005; Eggeman and Elander, 2005; ICIS, 2006; Lau et al., 2008b; Lu et al., 2009; Sendich et al., 2008

Distillation/Condensation/ Water Quenching/ Compression/ Water recovery

N/A

No

160±180

300±400

10±20 min

Eggeman and Elander, 2005; ICIS, 2006; Kim et al., 2003; Kim and Lee, 2005a; Lu et al., 2009

$180 (Ca(OH)2)

Neutralize base with CO2/ N/A Regenerate with lime kiln technology/Water recovery

Yes

25±150

0±200

1±8 weeks

Eggeman and Elander, 2005; ICIS, 2006; Kim and Holtzapple, 2005; Lu et al., 2009

N/Af

Recompression of the non- N/A reacted ozone to the feed line/Water recovery

No

25

±

2±3 hours

Garcia-Cubero et al., 2009; Quesada et al., 1999; Sun and Cheng, 2002

$820 (H2O2) $495 (Na2CO3)

Base neutralization/ Waste water treatment/ Water recovery

N/A

No

170±220

45±175

15 min

N/A

Waste water treatment/ Water recovery

N/R

No

20±25

±

14±23 days

Notes: Generality of pretreatment: High (+++), Medium (++), Low (+). Based on NFPA standards: H, Health; F, Flammability; R, Reactivity; O, Other hazard information (0 ± No special hazard, 4 ± Severe Hazard,W ± Reactivity with water, OX ± Oxidizer). c 2006 price by ICIS. d Current best case scenario projection for a generic ionic liquid. e Price of water typically ranges between $0.25 and $1/Mg. f Ozone was considered to be produced in-house. Cost of production not available. g N/A, N/R and Mg stand for Not available, Not required and Megagram (equivalent to 1metric tonne), respectively. a

b

ICIS, 2006; Klinke et al., 2003; Martin et al., 2008; Varga et al., 2003

Balan et al., 2008; Taniguchi et al., 2005; Watanabe, 2007

Table 2.5 Enzymatic hydrolysis-fermentation related parameter ranges (e.g., glucan and xylan hydrolysis yields) comprising various cellulosic substrates (i.e., grasses/straws and hardwoods/softwoods) and overall process economics for various biological and thermochemical pretreatments

ß Woodhead Publishing Limited, 2010

Pretreatment category

Enzymatic hydrolysis-related parametersb Total Total glucose xylose yield yield (%) (%)

Fractionation Organosolv Phosphoric acid Ionic liquids

Chemical Dilute acid

Steam explosion

Fermentation-related parametersb

Economicsa

Cellulase Hydrolysis Final Ethanol Washing Detoxification requireand nutrient loading/ time ethanol yieldc per g (hrs) concen(%) ment suppleglucan tration mentation (%)

References

Total fixed capital $/gal

85±100

N/A

20 FPU

48

3.7

99.5 (G)

Yes

Yes

N/A

>90

80

25 FPU

24

4.75

93 (G)

Yes

Yes

N/A

55±97

N/A

N/A

6±24

N/A

86 (G)

Yes

Yes

N/A

85±95

70±95

15 FPU

72

5.7

86 (G)

Yes

Yes

1.48 X

85±100 85±95

25 FPU

72

4.7

86 (G, X)

Yes

Yes

N/A

ICIS, 2006; Pan et al., 2006; Sidiras and Koukios, 2004; Sun and Chen, 2008; X Zhao et al., 2009 ICIS, 2006; H. Li et al., 2009; Zhang et al., 2007 Dadi et al., 2007; ICIS, 2006; Q. Li et al., 2009; Wasserscheid and Haumann, 2006 Eggeman and Elander, 2005; ICIS, 2006; Lu et al., 2009; Mohagheghi et al., 1992; Schell et al., 2003; Wyman et al., 2005, 2009 ICIS, 2006; Li and Chen, 2008; H. Li et al., 2009; Varga et al., 2004; Wang et al., 2009; Wyman et al., 2009; Lu et al., 2009

ß Woodhead Publishing Limited, 2010

Liquid hot water

55±90

80±95

20 FPU

24

2.9

82.7 (G, X)

No

Yes

1.82 X

AFEX

55±100 55±95

15 FPU

72±168

4

88.5 (G, X)

No

No

1.48 X

ARP

50±90

70±90

15 FPU

12±72

2

98.6 (G,X)

Yes

No

1.82 X

>90

80

15 FPU

72±168

4

100 (G, X)

Yes

No

1.33 X

Ozonolysis

80±90

N/A

29 FPU

48

N/A

N/A

N/A

N/A

N/A

Alkaline wet oxidation

70±80

50±55

73 FPU

24

5.2

83 (G)

Yes

No

N/A

40

N/A

15 FPU

72

N/A

N/A

Yes

No

N/A

Lime (with O2)

Biological Fungi or bacteria

Ballesteros et al., 2002; Eggeman and Elander, 2005; ICIS, 2006; Laser et al., 2002; Lu et al., 2009; Mosier et al., 2005b; Wyman et al., 2005, 2009 Alizadeh et al., 2005; Eggeman and Elander, 2005; ICIS, 2006; Lau et al., 2008a,b; Lu et al., 2009; Sendich et al., 2008; Wyman et al., 2005, 2009 Eggeman and Elander, 2005; ICIS, 2006; Kim et al., 2003; Kim and Lee, 2005a; Lu et al., 2009; Wyman et al., 2005, 2009 Eggeman and Elander, 2005; ICIS, 2006; Kim and Holtzapple, 2005; Lu et al., 2009; Wyman et al., 2005, 2009 Garcia-Cubero et al., 2009; Quesada et al., 1999; Sun and Cheng, 2002 ICIS, 2006; Klinke et al., 2003; Martin et al., 2008; Varga et al., 2003 Balan et al., 2008; Taniguchi et al., 2005; Watanabe, 2007

Notes: Lime pretreatment for poplar was performed in the presence of O2. a Ideal fixed capital given by X = $2.51/gal of ethanol. b Enzymatic hydrolysis/fermentation data is for corn stover, poplar and other related agricultural residues/energy crops. Data for CAFI pretreatments is based on corn stover and/or poplar. c Ethanol metabolic yield is either based on initial glucose (G), xylose (X) or both sugars (G/X).

56

Bioalcohol production

feedstock used. In Table 2.5, the data shown for the various pretreatments can correspond to different types of feedstock (e.g., corn stover, poplar and other lignocellulosics) and may not be always comparable. However, the data refer mostly to the best case scenario for each pretreatment. Fermentation is also affected by the choice of pretreatment, since the type of chemical modifications occurring in the plant cell wall can dictate the formation of potential inhibitors to microbial growth. The detoxification step is usually done by increasing the pH of the hydrolyzate to 9±11 at high temperature (~90 ëC) (Saha et al., 2005) using lime. This process will produce an insoluble waste residue (comprising various lignin and furan-based inhibitory compounds) that needs to be considered for the environmental and economical evaluation of the complete system. In addition to this fact, the extractive nature of some pretreatments promotes the removal of several micronutrients which could be used by microbes during fermentation (Lau and Dale, 2009). For this reason, in some processes it will be necessary to supplement the fermentation media with additional nutrients, which also impact the overall economics of the process. To further extend this analysis, it would be necessary to expand the boundaries of these studies to pre- and post-processing steps within a lignocellulosic biorefinery. This would help assess both economical and environmental impacts caused by the biorefinery for different pretreatment technologies. To complete this goal, it is necessary to integrate land management, feedstock processing, waste handling, distribution chains, animal feed production, etc., when evaluating the biorefinery process. In this context, life cycle analysis (LCA) is considered to be a powerful tool that could be used to help evaluate and design this system in a sustainable fashion (Kim and Dale, 2005; 2006).

2.5.2

Mass and energy balances

It is critical to carry out detailed material and energy balances over the entire biorefinery related operations (as shown in Figs 2.2 and 2.3) to estimate the process economic and environmental costs. Economic analysis would help determine high-impact process parameters that would significantly influence the feasibility of the operation. The interplay of various parameters like type of feedstock (prairie grass vs. hardwood), extent of biomass milling, total hemicellulase loading, hexose/pentose co-fermentation and co-product (residual protein) recovery with the pretreatment options can be better explored from this holistic perspective. For example, depending on the type of feedstock (corn stover vs. poplar) the severity of thermochemical pretreatment and total enzyme requirement would vary drastically. The type of pretreatment used (dilute acid vs. AFEX) would also influence detoxification and microbial fermentation strategies. It will be possible to estimate production costs for ethanol when

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

57

producing useful co-products such as animal feed using low severity pretreatments like AFEX (Bals et al., 2007; Carolan et al., 2007). The strategies for biomass collection, handling and transportation would also likely be influenced by the type of pretreatment technology. It would be easier to predict the actual economic potential of various biorefinery scenarios using detailed material and energy balances integrated with financial models (Eggeman and Elander, 2005; Sendich et al., 2008; Wooley et al., 1999).

2.5.3

Product yield, reaction rate, product concentration

The three important criteria that are typically used to compare various pretreatments are overall product yield (per unit mass of feedstock), rate of reaction and final product concentration. These parameters are typically found to most significantly influence the final minimum ethanol selling price (MESP) for all economic models (McAloon et al., 2000; Eggeman and Elander, 2005). In general, these factors will influence the equipment size and fixed cost investments for a biorefinery. More specifically, high yield is important to minimize feedstock use, which is considered to be the highest variable cost in a biorefinery (Aden et al., 2002). The minimization of feedstock usage is also necessary to maintain sustainability of the bioeconomy (compare to `bottom of the barrel' processing arguments in petroleum refineries), providing as much product (e.g. bioethanol) per acre of land as possible. To compare effectiveness of various pretreatments in terms of product yield, reaction rates and product concentration, it is important to have identical experimental setups. Most of the time, it is impossible to perform a fair comparison between pretreatment technologies, since the information available in the literature, comes from completely different experimental bases. Monomeric sugar yields during enzymatic hydrolysis are closely dependent on total solids loading, enzyme activity and inhibition. The solids to liquid ratio employed during pretreatment would affect the total solids loading employed during hydrolysis. Most pretreatments employ large quantities of water (5±10 kg water per kg biomass) that result in slurries at the end of pretreatment (which is typically washed and separated from the solid fraction that is used for hydrolysis experiments). Depending on the severity of acid-based treatments the lignin and hemicellulose are selectively solubilized from the solid glucan rich fraction. Most of the time, the solubilized hemicellulose is not easily fermentable to ethanol due to presence of inhibitors such as furfural and 5-hydroxymethylfurfural (Klinke et al., 2004), which results in lower overall ethanol yield. However, this situation opens the possibility for utilization of this stream for other applications, such as chemical conversion of sugars into valuable products (Zhao et al., 2007).

ß Woodhead Publishing Limited, 2010

58

2.5.4

Bioalcohol production

Economic and environmental feasibility

It would also be critical to conduct life cycle analysis (LCA) around various biorefinery scenarios taking in consideration the utilization of different feedstocks, agricultural practices, and technologies to produce biofuels and useful co-products (Kim and Dale, 2005). Certain feedstocks have higher biomass yields (per hectare) and might need lesser water and fertilizer inputs, like Miscanthus. However, the severity of thermochemical pretreatment necessary to achieve equivalent ethanol yields might negate the advantage of using miscanthus over corn stover as the biorefinery feedstock (Murnen et al., 2007). It therefore becomes pertinent to evaluate the non-tangible benefits of using certain feedstocks and technologies in a biorefinery, which have traditionally been ignored. It is also useful to evaluate biorefinery operations from a `bottom-of-thebarrel' processing outlook. Some parameters that contribute to the overall environmental and economical impact have to do with waste generation and disposal. For this reason, technologies that are able to utilize all lignocellulosic feedstock components should ideally be favored. Currently, 20±30% of the lignin from processed biomass is being burnt to generate energy. However, with further improvements in lignin utilization chemistry and technology it might be possible to recover useful by-products as feedstock for the petroleum-based polymer industry (Shevchenko et al., 1999; Lora and Glasser, 2002). In addition, processes that minimize water and chemical utilization should be favored, since it is important to minimize wastewater treatment and disposal of the nonrecycled chemicals for economical and environmental reasons. The quantity and nature of the chemical used during pretreatment will also dictate the investment in safety equipment as well as handling and disposal infrastructures. The use of different pretreatment chemicals will naturally generate variable environmental and economical issues to handle. Finally, one other interesting factor that should be considered for evaluating pretreatments is the potential of treated biomass to be used as animal feed (Weimer et al., 2003; Carolan et al., 2007). Most of the land in the United States is used to produce animal feed and thus the use of pretreated lignocellulosic biomass to feed ruminant animals can contribute to an important reduction of land use for animal feed production (Carolan et al., 2007). From this point of view, more land will be available to produce lignocellulosic biomass for biofuels, with minimal impact on the food chain.

2.6

Conclusions

Most studies on pretreatments have looked to optimize pretreatment parameters through maximization of hydrolysis (and/or fermentation) product yields. It is important to determine the chemical and ultra-structural modifications

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

59

incorporated within the cell wall during pretreatment to identify the key limiting factors that result in recalcitrance of native lignocellulosic biomass to bioprocessing. Further advances in this field would help in the development of novel pretreatment technologies that are cost effective and environmentally friendly. Another area of research that has lagged behind in this field of research has been in trying to predict the rate of enzymatic hydrolysis for pretreated lignocellulosic biomass (Laureano-Perez et al., 2005). Both substrate and enzyme related factors would be expected to contribute to the saccharification process. The rate of hydrolysis of a heterogenous, insoluble lignocellulosic substrate can be classified into two phases primarily: (a) the initial rate of hydrolysis that typically is dependent on the gross enzyme accessibility of the substrates. This initial rate of hydrolysis can be well correlated to the porosity of the biomass depending on the type of pretreatment (Grethlein, 1985); (b) the terminal rate of hydrolysis is dependent on several factors ranging from cellulose degree of polymerization, hemicellulose side-chain branching, nonproductive protein binding to lignin and enzyme inhibition by pretreatment and hydrolysis degradation products. There is a need to coherently incorporate these parameters for both phases of hydrolysis when developing saccharification kinetic models (Zhou et al., 2009). Consolidated bioprocessing (CBP) is a novel processing strategy for lignocellulosic biomass which is looking to consolidate the four biologicallymediated steps; of cellulase production, cellulose hydrolysis, hexose and pentose fermentation into a single processing step. This would help significantly reduce processing costs for a cellulosic biorefinery. There is no native microbial system that is currently able to perform CBP. Efforts are currently underway to develop a suitable CBP microbe. It would be of interest to study the effect of various pretreatment methodologies on the performance of CBP-based microbial systems. Cleavage of LCC linkages and extraction of hemicellulose-lignin residues seem to be the most important ultra-structural modification that takes place during most chemical pretreatments. However, in order to further enhance the rate of enzymatic hydrolysis of cell walls it would be necessary to modify the crystal structure of cellulose as well. The solid crystalline morphology of cellulose seems to be the ultimate rate-limiting step for efficient hydrolysis of plant cell walls (Zhou et al. 2009). Pretreatments that can further reduce the crystallinity of cellulose (either through formation of cellulose III or via conversion to amorphous cellulose), in a cost-effective manner, would have a significant advantage over existing pretreatment technologies.

2.7

Acknowledgements

The authors would like to thank CAFI team members for providing necessary data and making important suggestions during the course of preparation of this chapter. Some of the data for Tables 2.1, 2.2 and 2.3 were adapted from

ß Woodhead Publishing Limited, 2010

60

Bioalcohol production

previously published CAFI related paper by Kim and Holtzapple, 2005; 2006; Kim and Lee, 2005a; Laureano-Perez et al., 2005; Lloyd and Wyman, 2005; Mosier et al., 2005a; Teymouri et al., 2005; Wyman et al., 2009; Kumar et al., 2009; Kumar and Wyman, 2009a,b. This work was partly funded by DOE Great Lakes Bioenergy Research Center (www.greatlakesbioenergy.org) supported by the US Department of Energy, Office of Science, Office of Biological and Environmental Research, through Cooperative Agreement DE-FC0207ER64494 between The Board of Regents of the University of Wisconsin System and the US Department of Energy.

2.8

References

Aden A, Ruth M, Ibsen K, Jechura J, Neeves K, Sheehan J and Wallace B 2002 Lignocellulosic biomass to ethanol process design and economics utilizing cocurrent dilute acid prehydrolysis and enzymatic hydrolysis for corn stover. Technical Report (Golden: National Renewable Energy Laboratory (NREL)) pp. 1±154. Akin D E and Morrison W H 1988 Ozone treatment of forage: structure and digestibility of different lignified cell walls. Crop Science 28 337±42. Akin D E, Morrison W H, Rigsby L L, Barton F E, Himmelsbach D S and Hicks K B 2006 Corn stover fractions and bioenergy: chemical composition, structure, and response to enzyme pretreatment. Appl Biochem Biotechnol 129 104±16. Alizadeh H, Teymouri F, Gilbert T I and Dale B E 2005 Pretreatment of switchgrass by ammonia fiber explosion (AFEX). Appl Biochem Biotechnol 124 1133±41. Alvo P and Belkacemi K 1997 Enzymatic saccharification of milled timothy (Phleum pratense L.) and alfalfa (Medicago sativa L.). Biores Technol 61 185±98. Anderson W F, Peterson J, Akin D E and Morrison W H 2005 Enzyme pretreatment of grass lignocellulose for potential high-value co-products and an improved fermentable substrate. Twenty-Sixth Symposium on Biotechnology for Fuels and Chemicals, pp. 303±10. Arato C, Pye E K and Gjennestad G 2005 The lignol approach to biorefining of woody biomass to produce ethanol and chemicals. Appl Biochem Biotechnol 121±124 871± 82. Baker J O, Ehrman C I, Adney W S, Thomas S R and Himmel M 1998 Hydrolysis of cellulose using ternary mixtures of purified cellulases. Appl Biochem Biotechnol 70±72 395±403. Balan V, da Costa Sousa L, Chundawat S P S, Vismeh R, Jones A D and Dale B E 2008 Mushroom spent straw: a potential substrate for an ethanol-based biorefinery. J Ind Microbiol Biotechnol 35 293±301. Ballesteros I, Oliva J M, Navarro A A, GonzaÂlez A, Carrasco J and Ballesteros M 2000 Effect of chip size on steam explosion pretreatment of softwood. Appl Biochem Biotechnol 84±86 97±110. Ballesteros I, Oliva J M, Negro M J, Manzanares P and Ballesteros M 2002 Ethanol production from olive oil extraction residue pretreated with hot water. Appl Biochem Biotechnol 98±100 717±32. Ballesteros I, Negro M J, Oliva J M, Cabanas A, Manzanares P and Ballesteros M 2006 Ethanol production from steam-explosion pretreated wheat straw. Appl Biochem Biotechnol 130 496±508.

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

61

Bals B, Teachworth L, Dale B E and Balan V 2007 Extraction of proteins from switchgrass using aqueous ammonia within an integrated biorefinery. Appl Biochem Biotechnol 143 187±98. Bayer E A, Chanzy H, Lamed R and Shoham Y 1998 Cellulose, cellulases and cellulosomes. Curr Op Struct Biol 8 548±57. Ben-Ghedalia D and Miron J J 1981 Effects of sodium hydroxide, ozone and sulphur dioxide on the composition and in-vitro digestibility of wheat straw. J Sci Food Agric 32 224±8. Berlin A, Gilkes N, Kilburn D, Bura R, Markov A, Skomarovsky A, Okunev O, Gusakov A, Maximenko V, Gregg D, Sinitsyn A and Saddler J 2005 Evaluation of novel fungal cellulase preparations for ability to hydrolyze softwood substrates ± evidence for the role of accessory enzymes. Enzyme and Microbial Technology 37 175±84. Berlin A, Gilkes N, Kilburn D, Maximenko V, Bura R, Markov M, Skomarovsky A, Gusakov A, Sinitsyn A, Okunev O, Solovieva I and Saddler J N 2006. Evaluation of cellulase preparations for hydrolysis of hardwood substrates. Appl Biochem Biotechnol 130(1±3) 528±45. Berlin A, Maximenko V, Gilkes N and Saddler J N 2007 Optimization of enzyme complexes for lignocellulose hydrolysis. Biotechnol Bioeng 97(2) 287±96. Besle J, Cornu A and Jouany J 1994 Roles of structural phenylpropanoids in forage cell wall digestion. J Sci Food Agric 64 171±90. Bhat M and Bhat S 1997 Cellulose degrading enzymes and their potential industrial applications. Biotechnol Adv 15 583±620. Bhat M and Hazlewood G 2001 Enzymology and other Characteristics of Cellulases and Xylanases (Oxon, UK: CABI Publishing). Bidlack J, Malone M and Benson R 1992 Molecular structure and component integration of secondary cell walls in plants. Proc Oklahoma Acad Sci 72 51±56. Bobleter O, Binder H, Concin R and Burtscher E 1981 Energy from biomass, ed W Palz et al. (London: Applied Science Publishers) pp. 554±62. Boisset C, Petrequin C, Chanzy H, Henrissat B and Schulein M 2001 Optimized mixtures of recombinant Humicola insolens cellulases for the biodegradation of crystalline cellulose. Biotechnol Bioeng 72 (3) 339±45. Brownell H H and Saddler J N 1984 Steam explosion pretreatment for enzymatic hydrolysis. In: Biotechnology and Bioengineering Symposium 14, pp. 55±68. Brune A 1998 Termite guts: the world's smallest bioreactors. Trends in Biotechnology 16 16±21. Buchert J, Ranua M, Kantelinen A, Viikari L. 1992 The role of two Trichoderma reesei xylanases in the bleaching of kraft pulp. Appl Microb Biotechnol 37 825±9. Buranov A and Mazza G 2008 Lignin in straw of herbaceous crops. Ind Crops Prod 28 237±59. Carolan J E, Joshi S V and Dale B E 2007 Technical and financial feasibility analysis of distributed bioprocessing using regional biomass pre-processing centers. J Agri Food Ind Org 5 1±29. Carpita N and Gibeaut D 1993 Structural models of primary-cell walls in flowering plants ± Consistency of molecular-structure with the physical properties of the walls during growth. Plant J 3 1±30. Chandra R, Bura R, Mabee W, Berlin A, Pan X and Saddler J 2007 Substrate pretreatment: the key to effective enzymatic hydrolysis of lignocellulosics? Adv Biochem Eng Biotechnol 108 67±93. Chang V and Holtzapple M 2000 Fundamental factors affecting biomass enzymatic reactivity. Appl Biochem Biotechnol 84±86 5±37.

ß Woodhead Publishing Limited, 2010

62

Bioalcohol production

Chen H, Liu L, Yang X and Li Z 2005 New process of maize stalk amination treatment by steam explosion. Biomass Bioenergy 28 411±17. Chen S, Mowery R, Castleberry V, van Walsum G and Chambliss C 2006 High performance liquid chromatography method for simultaneous determination of aliphatic acid, aromatic acid and neutral degradation products in biomass pretreatment hydrolyzates. J Chromat A 1104 54±61. Chou Y 1986 Supercritical ammonia pretreatment of lignocellulosic materials. Biotechnol Bioeng Symp 17 19±32. Chum H, Johnson D, Black S, Baker J, Grohmann K, Sarkanen K, Wallace K and Schroeder H 1988 Organosolv pretreatment for enzymatic hydrolysis of poplars: (1) enzyme hydrolysis of cellulosic residues. Biotechnol Bioeng 31 643±9. Chum H L, Johnson D K, Black S K and Overend R P 1990 Pretreatment-catalyst effects and the combined severity parameter. Appl Biochem Biotechnol 24/25 1±14. Chundawat S P S 2009 Ultra-structural and physicochemical modifications within ammonia treated lignocellulosic cell walls and their influence on enzyme digestibility. PhD dissertation (Michigan State University, East Lansing). Chundawat S, Balan V and Dale B 2006 Effect of particle size based separation of milled corn stover on AFEX pretreatment and enzymatic digestibility. Biotechnol Bioeng 96 219±31. Chundawat S P S, Balan V, Jones A D, Sousa L, Lau M W and Dale B E 2007 Towards a fundamental understanding of ammonia fiber expansion (AFEX) pretreatment and its effect on enzymatic hydrolysis and fermentation of corn stover. In: 29th Symposium on Biotechnology for Fuels and Chemicals (Denver, Colorado). Chundawat S P S, Balan V, Dale B E 2008a High-throughput microplate technique for enzymatic hydrolysis of lignocellulosic biomass. Biotechnol Bioeng 99(6) 1281± 94. Chundawat S P S, Garlock R, Humpula J, Askeland P, Elder T, Sharma L, Chambliss K, Jones A D, Vismeh R, Agarwal U, Sousa L, Balan V and Dale B E 2008b Effect of AFEX pretreatment on corn stover cell wall ultra-structure and its related degradation reactions. In: 30th Symposium on Biotechnology for Fuels and Chemicals (New Orleans, Louisiana). Clark D and Mackie K 1987 Steam explosion of the softwood Pinus radiata with sulphur dioxide addition: (1) Process optimization. J Wood Chem Technol 7 373±403. da Costa Sousa L, Chundawat S P, Balan V and Dale B E 2009 `Cradle-to-grave' assessment of existing lignocellulose pretreatment technologies. Curr Opin Biotechnol 20 339±47. Dadi A P, Varanasi S and Schall C A 2006 Enhancement of cellulose saccharification kinetics using an ionic liquid pretreatment step. Biotechnol Bioeng 95 904±10. Dadi A P, Schall C A and Varanasi S 2007 Mitigation of cellulose recalcitrance to enzymatic hydrolysis by ionic liquid pretreatment. Appl Biochem Biotechnol 137± 140 407±21. Dale B and Moreira M 1982 A freeze-explosion technique for increasing cellulose hydrolysis. Biotechnol Bioeng Symp 12 31±43. Dale B, Leong C, Pham T, Esquivel V, Rios I and Latimer V 1996 Hydrolysis at low enzyme levels: application of the AFEX process. Biores Technol 56 111±16. De La Rosa L, Reshamwala S, Latimer V, Shawky B, Dale B and Stuart E 1994 Integrated production of ethanol fuel and protein from coastal Bermuda grass. Appl Biochem Biotech 45±46 483±97. Ding S and Himmel M 2006 The maize primary cell wall microfibril: a new model derived from direct visualization. J Agric Food Chem 54 597±606.

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

63

Donohoe B, Decker S, Tucker M, Himmel M and Vinzant T 2008 Visualizing lignin coalescence and migration through maize cell walls following thermochemical pretreatment. Biotechnol Bioeng (10.1002/bit.21959). Ebringerova A 2006 Structural diversity and application potential of hemicelluloses. Macromol Symp 232 1±12. Ebringerova A and Heinze T 2000 Xylan and xylan derivatives ± biopolymers with valuable properties: naturally occurring xylans structures, isolation procedures and properties. Macromol Rapid Commun 21 542±56. Eggeman T and Elander R T 2005 Process and economic analysis of pretreatment technologies. Biores Technol 96 2019±25. Eriksson T, Borjesson J and Tjerneld F 2002a Mechanism of surfactant effect in enzymatic hydrolysis of lignocelluloses. Enzyme Microb Technol 31 353±64. Eriksson T, Karlsson J and Tjerneld F 2002b A model explaining declining rate in hydrolysis of lignocellulose substrates with cellobiohydrolase I (Cel7A) and endoglucanase I (Cel7B) of Trichoderma reesei. Appl Biochem Biotechnol 101 41±60. Fahlen J and Salmen L 2003 Cross-sectional structure of the secondary wall of fibers as affected by processing. J Mat Sci 38 119±26. Fan LT, Lee Y and Beardmore D R 1980 The influence of major structural features of cellulose on rate of enzymatic hydrolysis. Biotechnol Bioeng 22 177±90. Fan L T, Lee Y H and Beardmore D R 1981 The influence of major structural features of cellulose on rate of enzymatic hydrolysis. Biotechnol Bioeng 23 419±24. Fengel D and Wegener G 1989 Wood: Chemistry, Ultrastructure and Reactions (Berlin: Walter de Gruyter). FernaÂndez-BolanÄos J, FelizoÂn B, Heredia A, RodrõÂguez R, GuilleÂn R and JimeÂnez A 2001 Steam-explosion of olive stones: hemicellulose solubilization and enhancement of enzymatic hydrolysis of cellulose. Biores Technol 79 53±61. Gao D, Chundawat S P S, Krishnan C, Balan V and Dale B E 2010 Mixture optimization of six core glycosyl hydrolases for maximizing saccharification of ammonia fiber expansion (AFEX) pretreated corn stover. Biores Technol 101, 2770±81. Garcia-Cubero M T, Gonzalez-Benito G, Indacoechea I, Coca M and Bolado S 2009 Effect of ozonolysis pretreatment on enzymatic digestibility of wheat and rye straw. Biores Technol 100 1608±13. Goldstein I and Easter J 1992 An improved process for converting cellulose to ethanol. Tappi 75 135±40. Gould J 1984 Alkaline peroxide delignification of agricultural residues to enhance enzymatic saccharification. Biotechnol Bioeng 26 46±52. Gow L and Wood T 1988 Breakdown of crystalline cellulose by synergistic action between cellulase components from Clostridium thermocellum and Trichoderma koningii. FEMS Microbiol Lett 50 247±52. Graenacher C 1934 Cellulose solution. US patent 1,943,176. Grethlein H E 1985 The effect of pore size distribution on the rate of enzymatic hydrolysis of cellulose substrates. Bio/Technology 3 155±60. Gusakov A and Sinitsyn A 1992 A theoretical analysis of cellulase product inhibition: effect of cellulase binding constant, enzyme/substrate ratio, and beta-glucosidase activity on the inhibition pattern. Biotechnol Bioeng 40 663±71. Gusakov A V, Salanovich T N, Antonov A I, Ustinov B B, Okunev O N, Burlingame R, Emalfarb M, Baez M, Sinitsyn A P 2007 Design of highly efficient cellulase mixtures for enzymatic hydrolysis of cellulose. Biotechnol Bioeng 97 1028±38. Harris E E and Begliner E 1946 Hydrolysis of wood. Treatment with sulfuric acid in a stationary digester. Ind Eng Chem 37 12±23.

ß Woodhead Publishing Limited, 2010

64

Bioalcohol production

Heinze T and Koschella A 2005 Solvents applied in the field of cellulose chemistry: a mini review. PolõÂmeros 15 84±90. Hennessey S M, Friend J, Dunson J B, Tucker M P, Elander R T and Hames B 2007 Integration of alternative feedstreams for biomass treatment and utilization (USA: DuPont, USA). Himmel M, Ding S, Johnson D, Adney W, Nimlos M, Brady J and Foust T 2007 Biomass recalcitrance: engineering plants and enzymes for biofuels production. Science 315 804±7. Holtzapple M T, Ripley E P and Nikolaou M 1994 Saccharification, fermentation, and protein recovery from low-temperature AFEX-treated coastal Bermuda grass. Appl Biochem Biotech 44 1122±31. Hsu T A 1996 Handbook on Bioethanol, Production and Utilization, ed C E Wyman (Washington, DC: Taylor and Francis). ICIS 2006 ICIS Chemical Business. Igarashi K, Wada M and Samejima M 2007 Activation of crystalline cellulose to cellulose III results in efficient hydrolysis by cellobiohydrolase. FEBS Journal 274 1785±92. Ishizawa C, Davis M, Schell D and Johnson D 2007 Porosity and its effect on the digestibility of dilute sulfuric acid pretreated corn stover. J Agric Food Chem 55 2575±81. Jablonsky M, VrsÏka M and KatusÏcÏaÂk S 2004 Cellulose protectors for improving ozone bleaching ± review. Wood Res 49 71±86. Joglekar H G, Rahman I and Kulkarni B D 2007 The path ahead for ionic liquids. Chem Eng Technol 30, 819±28. Jungnikl K, Paris O, Fratzl P and Burgert I 2008 The implication of chemical extraction treatments on the cell wall nanostructure of softwood. Cellulose 15 407±18. Kaar W E and Holtzapple M T 2000 Using lime pretreatment to facilitate the enzymatic hydrolysis of corn stover. Biomass Bioenergy 18 189±99. Keller F A, Hamilton J E and Nguyen Q A 2003 Microbial pretreatment of biomass: potential for reducing severity of thermo chemical biomass pretreatment. Appl Biochem Biotechnol 105±108 27±41. Kerem Z, Friesem D and Hadar Y 1992 Lignocellulose degradation during solid-state fermentation: Pleurotus ostreatus versus Phanerochaete chrysosporium. Appl Environ Microbiol 58 1121±7. Kilpelainen I, Xie H, King A, Granstrom M, Heikkinen S and Argypropoulos D S 2007 Dissolution of wood in ionic liquid. J Agric Food Chem 55 9142±8. Kim J S, Lee Y Y and Torget R W 2001 Cellulose hydrolysis under extremely low sulfuric acid and high-temperature conditions. Appl Biochem Biotechnol 91±93 331±40. Kim S and Dale B E 2005 Life cycle assessment of various cropping systems utilized for producing biofuels: bioethanol and biodiesel. Biomass Bioenergy 29 426±39. Kim S and Dale B E 2006 Ethanol fuels: E10 or E85 ± life cycle perspectives. Int J Life Cycle Assessment 11 117±21. Kim S and Holtzapple M T 2005 Lime pretreatment and enzymatic hydrolysis of corn stover. Biores Technol 96 1994±2006. Kim S and Holtzapple M T 2006 Delignification kinetics of corn stover in lime pretreatment. Biores Technol 97 778±85. Kim S B and Lee Y Y 1996 Fractionation of herbaceous biomass by ammonia-hydrogen peroxide percolation treatment. Appl Biochem Biotechnol 57/58 147±56. Kim T H and Lee Y Y 2005a Pretreatment and fractionation of corn stover by ammonia recycle percolation process. Biores Technol 96 2007±13.

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

65

Kim T H and Lee Y Y 2005b Pretreatment of corn stover by soaking in aqueous ammonia. Appl Biochem Biotechnol 124 1119±31. Kim T H, Kim J S, Sunwoo C and Lee Y Y 2003 Pretreatment of corn stover by aqueous ammonia. Biores Technol 90 39±47. Klein G and Snodgrass W 1993 Cellulose (London: Academic Press). Klinke H B, Ahring B K, Schmidt A S and Thomsen A B 2002 Characterization of degradation products from alkaline wet oxidation of wheat straw. Biores Technol 82 15±26. Klinke H B, Olsson L, Thomsen A B and Ahring B K 2003 Potential inhibitors from wet oxidation of wheat straw and their effect on ethanol production of Saccharomyces cerevisiae: wet oxidation and fermentation by yeast. Biotechnol Bioeng 81 738±47. Klinke H B, Thomsen A B and Ahring B K 2004 Inhibition of ethanol-producing yeast and bacteria by degradation products produced during pre-treatment of biomass. Appl Microbiol Biotechnol 66 10±26. Knoshaug E P, Selig M J, Baker J O, Decker S R, Himmel M E and Adney W M 2008 Heterologous expression of two ferulic acid esterases from Penicillium funiculosum. Appl Biochem Biotechnol 146 79±87. Kong F, Engler C and Soltes E 1992 Effects of cell-wall acetate, xylan backbone, and lignin on enzymatic hydrolysis of aspen wood. Appl Biochem Biotechnol 34±35 23±35. Kootstra A M, Mosier N S, Scott E L and Beeftink H H 2009 Differential effects of mineral and organic acids on the kinetics of arabinose degradation under lignocellulose pretreatment conditions. Biochem Eng J 43 92±7. Koshijima T and Watanabe T 2003 Association between Lignin and Carbohydrates in Wood and other Plant Tissues (Berlin: Springer-Verlag). Koshijima T, Watanabe T and Yaku F 1989 Structure and properties of the lignin carbohydrate complex polymer as an amphipathic substance. ACS Symposium Series 397 (American Chemical Society). Koullas D, Christakopoulos P, Kekos D, Koukios E and Macris B 1993 Effect of alkali delignification on wheat straw saccharification by Fusarium oxysporum cellulases. Biomass Bioenergy 4 9±13. Kristensen J, Thygesen L, Felby C, Jùrgensen H and Elder T 2008 Cell-wall structural changes in wheat straw pretreated for bioethanol production. Biotechnology for Biofuels 15. Kumar R and Wyman C E 2009a Effect of enzyme supplementation at moderate cellulase loadings on initial glucose and xylose release from corn stover solids pretreated by leading technologies. Biotechnol Bioeng 102 457±67. Kumar R and Wyman C E 2009b Effects of cellulase and xylanase enzymes on the deconstruction of solids from pretreatment of poplar by leading technologies. Biotechnol Prog 25 302±14. Kumar R, Mago G, Balan V and Wyman C E 2009 Physical and chemical characterizations of corn stover and poplar solids resulting from leading pretreatment technologies. Biores Technol 100 3948±62. Lapierre C, Pollet B, Ralet M-C and Saulnier L 2001 The phenolic fraction of maize bran: evidence for lignin-heteroxylan association. Phytochemistry 57 765±72. Laser M, Schulman D, Allen S G, Lichwa J, Antal M J and Lynd L R 2002 A comparison of liquid hot water and steam pretreatments of sugar cane bagasse for bioconversion to ethanol. Biores Technol 81 33±44. Lau M W and Dale B E 2009 Cellulosic ethanol production from AFEX-treated corn stover using Saccharomyces cerevisiae 424A(LNH-ST). Proc Natl Acad Sci 106 1368±73.

ß Woodhead Publishing Limited, 2010

66

Bioalcohol production

Lau M W, Dale B E and Balan V 2008a Ethanolic fermentation of hydrolyzates from ammonia fiber expansion (AFEX) treated corn stover and distillers grain without detoxification and external nutrient supplementation. Biotechnol Bioeng 99 529±39. Lau M W, Young J M, Balan V, Chundawat S P S, Bals B D, Gao D, Garlock R J and Dale B E 2008b A rigorous comparison of AFEX and dilute-acid pretreatment using recombinant ethanologens. In: 30th Symposium on Biotechnology for Fuels and Chemicals (New Orleans, Louisiana). Laureano-Perez L, Teymouri F, Alizadeh H and Dale B 2005 Understanding factors that limit enzymatic hydrolysis of biomass: characterization of pretreated corn stover. Appl Biochem Biotechnol 124 1081±99. Laureano-Perez L, Dale B E, O'Dwyer J P and Holtzapple M T 2006 Statistical correlation of spectroscopic analysis and enzymatic hydrolysis of poplar samples. Biotechnol Prog 22 835±41. Lee J 1997 Biological conversion of lignocellulosic biomass to ethanol. J Biotechnol 56 1±24. Lee J-W, Gwak1 K-S, Park J-Y, Park M-J, Choi D-H, Kwon M and Choi I-G 2007 Biological pretreatment of softwood Pinus densiflora by three white rot fungi. J Microbiol 45 455±91. Lee J-W, Kim HY, Koo B-W, Choi D-H, Kwon M and Choi I-G 2008 Enzymatic saccharification of biologically pretreated Pinus densiflora using enzymes from brown rot fungi. J Biosci Bioeng 106 162±7. Lee S-H, Doherty T V, Linhardt R J and Dordick J S 2008 Ionic liquid-mediated selective extraction of lignin from wood leading to enhanced enzymatic cellulose hydrolysis. Biotechnol Bioeng DOI 10.1002/bit.22179. Lee Y H, Robinson C W and Moo-Young M 1987 Evaluation of Organosolv process for the fraction and modification of corn stover for bioconversion. Biotechnol Bioeng 29 572±81. Lee Y Y, Iyer P and Torget R W 1999 Dilute acid hydrolysis of lignocellulosic biomass. Adv Biochem Eng Biotechnol 65 93. Leonowicz A, Matuszewska A, Luterek J, Ziegenhagen D, Wojta-Wasilewska M, NamSeok C, Hofrichter M, Rogalski J 1999 Biodegradation of lignin by white rot fungi. Fungal Genetics and Biology 27 175±85. Lewin M and Roldan L G 1971 The effect of liquid anhydrous ammonia in the structure and morphology of cotton cellulose. J Polym Sci Part C ± Polym Symp 213±29. Li H and Chen H 2008 Detoxification of steam-exploded corn straw produced by an industrial-scale reactor. Process Biochem 43 1447±51. Li H, Kim N J, Jiang M, Kang J W and Chang H N 2009 Simultaneous saccharification and fermentation of lignocellulosic residues pretreated with phosphoric acidacetone for bioethanol production. Biores Technol 100 3245±51. Li Q, He Y C, Xian M, Jun G, Xu X, Yang J M and Li L Z 2009 Improving enzymatic hydrolysis of wheat straw using ionic liquid 1-ethyl-3-methyl imidazolium diethyl phosphate pretreatment. Biores Technol 100 3570±5. Liu C and Wyman C 2005 Partial flow of compressed-hot water through corn stover to enhance hemicellulose sugar recovery and enzymatic digestibility of cellulose. Biores Technol 96 1978±85. Lloyd T A and Wyman C E 2005 Combined sugar yields for dilute sulfuric acid pretreatment of corn stover followed by enzymatic hydrolysis of the remaining solids. Biores Technol 96 1967±77. Lora J H and Glasser W G 2002 Recent industrial applications of lignin: a sustainable alternative to nonrenewable materials. J Polym Environ 10 39±48.

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

67

Lu Y and Mosier N S 2007 Biomimetic catalysis for hemicellulose hydrolysis in corn stover. Biotechnol Prog 23 116±23. Lu Y and Mosier N S 2008 Kinetic modeling analysis of maleic acid-catalyzed hemicellulose hydrolysis in corn stover. Biotechnol Bioeng 101 1170±81. Lu Y L, Warner R, Sedlak M, Ho N and Mosier N S 2009 Comparison of glucose/xylose cofermentation of poplar hydrolyzates processed by different pretreatment technologies. Biotechnol Prog 25 349±56. Mansfield S D 2009 Solutions for dissolution ± engineering cell walls for deconstruction. Curr Opin Biotechnol 20 286±94. Mansfield S D, Mooney C and Saddler J N 1999 Substrate and enzyme characteristics that limit cellulose hydrolysis. Biotech Prog 15 804. Martin C and Thomsen A B 2007 Wet oxidation pretreatment of lignocellulosic residues of sugarcane, rice, cassava and peanuts for ethanol production. J Chem Technol Biotechnol 82 174±81. Martin C, Thomsen M H, Hauggaard-Nielsen H and Belinda-Thomsen A 2008 Wet oxidation pretreatment, enzymatic hydrolysis and simultaneous saccharification and fermentation of clover-ryegrass mixtures. Biores Technology 99 8777±82. Marx-Figini M 1969 On the biosynthesis of cellulose in higher and lower plants. Journal of Polymer Science Part C: Polymer Symposia 28 57±67. McAloon A J, Taylor F, Yee W C, Ibsen K and Wooley R 2000 Determining the Cost of Producing Ethanol from Corn Starch and Lignocellulosic Feedstocks (Golden: National Renewable Energy Laboratory) pp. 1±44. McCann M and Carpita N 2008 Designing the deconstruction of plant cell walls. Curr Opin Plant Biol 11 314±20. McGinnis G D, Wilson W W, Prince S E and Chen C C 1983 Conversion of biomass into chemicals with high-temperature wet oxidation. Ind Eng Chem Prod Res Dev 22 633±6. McMillan J D 1994 Pretreatment of lignocellulosic biomass. In: M E Himmel, J O Baker and R P Overend (eds) ACS Symp Ser 566 pp. 292±324. MendoncËa R T, Jara J E, GonzaÂlez V, Elissetche J P, Freer J 2008 Evaluation of the white-rot fungi Ganoderma australe and Ceriporiopsis subvermispora in biotechnological applications. J Ind Microbiol Biotechnol 35 1323±30. Millett M A, Baker A J and Satter L D 1975 Pretreatments to enhance chemical, enzymatic, and microbiological attack of cellulosic materials. Biotechnol Bioeng 5, 193±219. Mohagheghi A, Tucker M, Grohmann K and Wyman C 1992 High solids simultaneous saccharification and fermentation of pretreated wheat straw to ethanol. Appl Biochem Biotechnol 33 67±81. Morrison W H, III, Scheijen M A and Boon J J 1991 Pyrolysis mass spectrometry of coastal Bermudagrass (Cynodon dactylon (L.) Pers.) and `Kentucky-3 l' tall fescue (Festuca arundinacea Schreb.) cell walls and their residues after ozonolysis and base hydrolysis. Anim Feed Sci Technol 32 17±26. Mosier N S, Sarikaya A, Ladisch C M and Ladisch M R 2001 Characterization of dicarboxylic acids for cellulose hydrolysis. Biotechnol Prog 17 474±80. Mosier N S, Ladisch C M and Ladisch M R 2002 Characterization of acid catalytic domains for cellulose hydrolysis and glucose degradation. Biotechnol Bioeng 79 610±18. Mosier N, Hendrickson R, Ho N, Sedlak M and Ladisch M R 2005a Optimization of pH controlled liquid hot water pretreatment of corn stover. Biores Technol 96 1986±93. Mosier N, Wyman C, Dale B, Elander R, Lee Y Y, Holtzapple M and Ladisch M 2005b

ß Woodhead Publishing Limited, 2010

68

Bioalcohol production

Features of promising technologies for pretreatment of lignocellulosic biomass. Biores Technol 96 673±86. Mosier N S, Hendrickson R, Brewer M, Ho N, Sedlak M, Dreshel R, Welch G, Dien B S, Aden A and Ladisch M R 2005c Industrial scale-up of pH-controlled liquid hot water pretreatment of corn fiber for fuel ethanol production. Appl Biochem Biotechnol 125 77±97. Moxyley G, Zhu Z and Zhang Y-H P 2008 Efficient sugar release by the cellulose solvent-based lignocellulose fractionation technology and enzymatic cellulose hydrolysis. J Agric Food Chem 56 7885±90. Murnen H K, Balan V, Chundawat S P S, Bals B, Sousa L D and Dale B E 2007 Optimization of ammonia fiber expansion (AFEX) pretreatment and enzymatic hydrolysis of Miscanthus x giganteus to fermentable sugars. Biotechnol Prog 23 846±50. O'Connor J 1972 Ammonia explosion pulping. Tappi 55 353±8. O'Sullivan A 1997 Cellulose: the structure slowly unravels. Cellulose 4 173±207. È hgren K, Galbe M and Zacchi G 2005 Optimization of steam pretreatment of SO2O impregnated corn stover for fuel ethanol production. Appl Biochem Biotechnol 124 1055±67. È hgren K, Bura R, Saddler J and Zacchi G 2007 Effect of hemicellulose and lignin O removal on enzymatic hydrolysis of steam pretreated corn stover. Biores Technol 98 2503±10. Palonen H 2004 Role of Lignin in the Enzymatic Hydrolysis of Lignocellulose (Espoo: Helsinki University of Technology) p. 84. Pan X, Gilkes N, Kadla J, Pye K, Saka S, Gregg D, Ehara K, Xie D, Lam D and Saddler J 2006 Bioconversion of hybrid poplar to ethanol and co-products using an Organosolv fractionation process: optimization of process yields. Biotechnol Bioeng 94 851±61. Pan X, Xie D, Kang K-Y, Yoon S-L and Saddler J N 2007 Effect of Organosolv ethanol pretreatment variables on physical characteristics of hybrid poplar substrates. Appl Biochem Biotechnol 136±140 367±77. Papatheofanous M G, Billa E, Koullas D P, Monties B and Koukios E G 1996 Two-stage acid-catalyzed fractionation of lignocellulosic biomass in aqueous ethanol systems at low temperatures. Biores Technol 54 305±10. Park S, Johnson D, Ishizawa C, Parilla P and Davis M 2009 Measuring the crystallinity index of cellulose by solid state 13C nuclear magnetic resonance. Cellulose 16 641± 7. Pauly M and Keegstra K 2008 Cell wall carbohydrates and their modifications as a resource for biofuels. Plant J 54 559±68. Playne M J 1984 Increased digestibility of bagasses by pretreatment with alkalis and steam explosion. Biotechnol Bioeng 26 426±33. Puri V P 1983 Ozone pretreatment to increase digestibility of lignocelluloses. Biotechnol Lett 5 773±6. Quesada J, Rubio M and Gomez D 1997 Chemical characterization of ozonated lignin solutions from corn (Zea mays) stalk and poplar (Populus deltoides) wood by capillary gas chromatography. J High Res Chromatog 20 565±8. Quesada J, Rubio M and Gomez D 1999 Ozonation of lignin rich solid fractions from corn stalks. J Wood Chem Technol 19 115±37. Ragnar M, Tord E and TorbjoÈrn R 1999 Radical formation in ozone reactions with lignin and carbohydrate model compounds. Holzforschung 53 292±8. Ramos L, Nazhad M and Saddler J N 1993 Effect of enzymatic hydrolysis on the

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

69

morphology and fine structure of pretreated cellulosic residues. Enzyme Microb Technol 15 821±31. Reese E, Siu R and Levinson H 1950 Biological degradation of soluble derivatives and its relationship to the mechanism of cellulose hydrolysis. J Bacteriol 59 485±97. Rinaldi R, Palkovits R and Schuth F 2008 Depolymerization of cellulose using solid catalyst in ionic liquids. Ang Chem 47 8047±50. Roncero M B, Torres A L, Colom J F and Vidal T 2003 TCF bleaching of wheat straw pulp using ozone and xylanase. Part A: paper quality assessment. Biores Technol 87 305±14. Rosgaard L, Pedersen S, Langston J, Akerhielm D, Cherry J and Meyer A 2007 Evaluation of minimal Trichoderma reesei cellulase mixtures on differently pretreated barley straw substrates. Biotechnol Prog 23 1270±6. Ruel K, Chevalier-billosta V, Guillemin F, Sierra J B and Joseleau J-P 2006 The wood cell wall at the ultrastructural scale ± formation and topochemical organization. Maderas: Ciencia y Technologia 8 107±16. Ruttan R F 1909 Ethyl alcohol from sawdust and other wood waste. Soc Chem Ind 28 1290±4. Saha B C 2003 Hemicellulose bioconversion. J Ind Microb Biotech 30(5) 279±91. Saha B and Bothast R 1999 Pretreatment and enzymatic saccharification of corn fiber. Appl Biochem Biotechnol 76 65±77. Saha B C, Iten L B, Cotta M A and Wu Y V 2005 Dilute acid pretreatment, enzymatic saccharification, and fermentation of rice hulls to ethanol. Biotechnol Prog 21 816± 22. Saulnier L and Thibault J-F 1999 Ferulic acid and diferulic acids as components of sugarbeet pectins and maize bran heteroxylans. J Sci Food Agric 79 396±402. Schell D J, Farmer J, Newman M and McMillan J D 2003 Dilute-sulfuric acid pretreatment of corn stover in pilot-scale reactor: investigation of yields, kinetics, and enzymatic digestibilities of solids. Appl Biochem Biotechnol 105±108 69±85. Schmidt A and Thomsen A 1998 Optimization of wet oxidation pretreatment of wheat straw. Biores Technol 64 139±51. Selig M, Viamajala S, Decker S, Tucker M, Himmel M and Vinzant T 2007 Deposition of lignin droplets produced during dilute acid pretreatment of maize stems retards enzymatic hydrolysis of cellulose. Biotechnol Prog 23 1333±9. Selig M, Knoshaug E, Adney W, Himmel M and Decker S 2008a Synergistic enhancement of cellobiohydrolase performance on pretreated corn stover by addition of xylanase and esterase activities. Biores Technol 99 4997±5005. Selig M, Knoshaug E, Decker S, Baker J, Himmel M and Adney W 2008b Heterologous expression of Aspergillus niger -D-xylosidase (XlnD): characterization on lignocellulosic substrates. Appl Biochem Biotechnol 146 57±68. Sendich E, Laser M, Kim S, Alizadeh H, Laureano-Perez L, Dale B and Lynd L 2008 Recent process improvements for the ammonia fiber expansion (AFEX) process and resulting reductions in minimum ethanol selling price. Biores Technol 99 8429±35. Sewalt V, Glasser W and Beauchemin K 1997 Lignin impact on fiber degradation. 3. Reversal of inhibition of enzymatic hydrolysis by chemical modification of lignin and additives. J Agric Food Chem 45 1823±8. Shatalov A A and Pereira H 2007 Polysaccharide degradation during ozone-based TCF bleaching of non-wood organosolv pulps. Carbohyd Polym 67 275±81. Shevchenko S M, Beatson R P and Saddler J N 1999 The nature of lignin from steam explosion enzymatic hydrolysis of softwood ± structural features and possible uses. Appl Biochem Biotechnol 79 867±76.

ß Woodhead Publishing Limited, 2010

70

Bioalcohol production

Shevchenko S M, Chang K, Dick D G, Gregg D J and Saddler J N 2001 Structure and properties of lignin in softwoods after SO2-catalyzed steam explosion and enzymatic hydrolysis. Cell Chem Technol 35 487±502. Shi J, Chinn M S and Sharma-Shivappa R R 2008 Microbial pretreatment of cotton stalks by solid state cultivation of Phanerochaete chrysosporium. Biores Technol 99 6556±64. Sidiras D and Koukios E 1989 Acid saccharification of ball milled straw. Biomass 19 289±306. Sidiras D and Koukios E 2004 Simulation of acid-catalysed organosolv fractionation of wheat straw. Biores Technol 94 91±8. Sierra R, Granda C and Holtzapple M T 2009 Short-term lime pretreatment of poplar wood. Biotechnol Prog 25 323±32. Silverstein R A, Chen Y, Sharma-Shivappa R R, Boyette M D and Osborne J 2007 A comparison of chemical pretreatment methods for improving saccharification of cotton stalks. Biores Technol 98 3000±11. Singh S, Simmons B A and Vogel K P 2009 Visualization of biomass solubilization and cellulose regeneration during ionic liquid pretreatment of switchgrass. Biotechnol Bioeng 104 68±75. Stalbrand H, Mansfield S, Saddler J, Kilburn D, Warren R and Gilkes N 1998 Analysis of molecular size distributions of cellulose molecules during hydrolysis of cellulose by recombinant Cellulomonas fimi beta-1,4-glucanases. Appl Environ Microbiol 64 2374±9. Stone J E and Scallan A M 1969 The effect of component removal upon the porous structure of the cell wall of wood. Part III. A comparison between sulphite and kraft processes. Pap Mag Can 69 T219±T40. Sun F and Chen H 2008 Organosolv pretreatment by crude glycerol from oleochemicals industry for enzymatic hydrolysis of wheat straw. Biores Technol 99 5474±9. Sun Y and Cheng J 2002 Hydrolysis of lignocellulosic materials for ethanol production: a review. Biores Technol 83 1±11. Swatloski R P, Spear S K, Holbrey J D and Rogers R D 2002 Dissolution of cellulose with ionic liquids. J Am Chem Soc 124 4974±5. Tabka M G, Herpoel-Gimbert I, Monod F, Asther M and Sigoillot J C 2006. Enzymatic saccharification of wheat straw for bioethanol production by a combined cellulase, xylanase and feruloyl esterase treatment. Enzyme Microbial Technol 39 897±902. Taniguchi M, Suzuki H, Watanabe D, Sakai K, Hoshino K and Tanaka T 2005 Evaluation of pretreatment with Pleurotus ostreatus for enzymatic hydrolysis of rice straw. J Biosci Bioeng 100 637±43. Tassinari T, Macy C and Spano L 1980 Energy requirements and process design considerations in compression-milling pretreatment of cellulosic wastes for enzymatic hydrolysis. Biotech Bioeng 22 1689±705. Teymouri F, Laureano-Perez L, Alizadeh H and Dale B E 2005 Optimization of the ammonia fiber explosion (AFEX) treatment parameters for enzymatic hydrolysis of corn stover. Biores Technol 96 2014±18. Thompson D, Chen H and Grethlein H 1992 Comparison of pretreatment methods on the basis of available surface area. Biores Technol 39 155±63. Torget R, Himmel M and Grohmann K 1992 Dilute acid pretreatment of two short rotation herbaceous crops. Appl Biochem Biotechnol 34±35 115±23. Turner M B, Spear S K, Huddleston J G, Holbrey J D and Rogers R D 2003 Ionic liquid salt-induced inactivation and unfolding of cellulase from Trichoderma reesei. Green Chem 5 443±7.

ß Woodhead Publishing Limited, 2010

Thermochemical pretreatment of lignocellulosic biomass

71

VaÈljamaÈe P, Sild V, Nutt A, Pettersson G and Johansson G 1999 Acid hydrolysis of bacterial cellulose reveals different modes of synergistic action between cellobiohydrolase I and endoglucanase I. Eur J Biochem 266 327±34. van Walsum G P, Garcia-Gil M, Chen S-F, Chambliss K 2007 Effect of dissolved carbon dioxide on accumulation of organic acids in liquid hot water pretreated biomass hydrolyzates. Appl Biochem Biotechnol 136±160 301±11. Varga E, ReÂczey K and Zacchi G 2004 Optimization of steam pretreatment of corn stover to enhance enzymatic digestibility. Appl Biochem Biotechnol 114 509±23. Varga E, Schmidt A S, ReÂczey K and Thomsen A B 2003 Pretreatment of corn stover using wet oxidation to enhance enzymatic digestibility. Appl Biochem Biotechnol 104 37±50. Wada M, Nishiyama Y and Langan P 2006 X-ray structure of ammonia-cellulose I: new insights into the conversion of cellulose I to cellulose III. Macromolecules 39 2947± 52. Wada M, Nishiyama Y, Chanzy H, Forsyth T and Langan P 2008 The structure of celluloses. Powder Diffraction 23 92±95. Wang K, Jiang J X, Xu F and Sun R C 2009 Influence of steaming pressure on steam explosion pretreatment of Lespedeza stalks (Lespedeza crytobotrya): Part 1. Characteristics of degraded cellulose. Polymer Degradation and Stability 94 1379± 88. Wasserscheid P and Haumann M 2006 Catalyst Separation, Recovery and Recycling, ed D J Cole-Hamilton and R P Tooze (Berlin: Springer) pp. 182±213. Watanabe T 2007 Trends in biorefinery and pretreatments of lignocellulosics by white rot fungi. Mokuzai Gakkaishi 53 1±13. Weil J, Westgate P, Kohlmann K and Ladisch M 1994 Cellulose pretreatments of lignocellulosic substrates. Enzyme Microb Technol 16 1002±4. Weil J R, Sarikaya A, Rau S-L, Goetz J, Ladisch C M, Brewer M, Hendrickson R and Ladisch M R 1998 Pretreatment of corn fiber by pressure cooking in water. Appl Biochem Biotechnol 73 1±17. Weimer P, Chou Y, Weston W and Chase D 1986 Effect of supercritical ammonia on the physical and chemical structure of ground wood. Biotechnol Bioeng Symp 17 5±18. Weimer P J, French A D and Calamari T A 1991 Differential fermentation of cellulose allomorphs by ruminal cellulolytic bacteria. Appl Environ Microbiol 57 3101±6. Weimer P J, Mertens D R, Ponnampalam E, Severin B F and Dale B E 2003 FIBEXtreated rice straw as a feed ingredient for lactating dairy cows. Anim Feed Sci Technol 103 41±50. White A and Brown M 1981 Enzymatic hydrolysis of cellulose: visual characterization of the process. Proc Natl Acad Sci USA 78 1047±51. Wood T 1968 Cellulolytic enzymes system of Trichoderma koningii. Separation of components attacking native cotton. Biochem J 109 217±29. Wood T and McCrae S 1986 The cellulase of Penicillium pinophilum. Synergism between enzyme components in solubilizing cellulose with special reference to the involvement of two immunologically distinct cellobiohydrolases. Biochem J 234 93±9. Wood T, McCrae S and Bhat K 1989 The mechanism of fungal cellulase action. Synergism between enzyme components of Penicillium pinophilum cellulase in solubilizing hydrogen-bond ordered cellulose. Biochem J 260 37±43. Wooley R, Ruth M F, Glassner D and Sheehan J 1999 Process design and costing of bioethanol technology: A tool for determining the status and direction of research and development. Biotechnol Prog 15 794±803.

ß Woodhead Publishing Limited, 2010

72

Bioalcohol production

Wu Z and Lee Y 1997 Ammonia recycled percolation as a complementary pretreatment to the dilute-acid process. Appl Biochem Biotechnol 63±65 21±34. Wyman C E, Dale B E, Elander R T, Holtzapple M T, Ladisch M R and Lee Y Y 2005 Comparative sugar recovery data from laboratory scale application of leading pretreatment technologies to corn stover. Biores Technol 96 2026±32. Wyman C, Kumar R, Dale B, Elander R, Holtzapple M T, Ladisch M, Lee Y, Moniruzzaman M and Saddler J 2006 Comparative data for enzymatic digestion of corn stover and poplar wood after pretreatment by leading technologies. In: AIChE Annual Meeting (San Francisco, California). Wyman C E, Dale B E, Elander R T, Holtzapple M T, Ladisch M R, Lee Y Y, Mitchinson C and Saddler J N 2009 Comparative sugar recovery and fermentation data following pretreatment of poplar wood by leading technologies. Biotechnol Prog 25 333±9. Yang B and Wyman C E 2004 Effect of xylan and lignin removal by batch and flowthrough pretreatment on the enzymatic digestibility of corn stover cellulose. Biotechnol Bioeng 86 88±95. Yang B and Wyman C E 2008 Characterization of the degree of polymerization of xylooligomers produced by flowthrough hydrolysis of pure xylan and corn stover with water. Biores Technol 99 5756±62. Yang B, Willies D and Wyman C 2006 Changes in the enzymatic hydrolysis rate of avicel cellulose with conversion. Biotech Bioeng 94 1122±8. Zadrazil F and Puniya A 1995 Studies on the effect of particle size on solid-state fermentation of sugarcane bagasse into animal feed using white-rot fungi. Biores Technol 54 85±7. Zeitsch K J 2000 The Chemistry and Technology of Furfural and its many By-products, vol 13 (New York: Elsevier). Zhang Y-H P and Lynd R 2004 Towards an aggregated understanding of enzymatic hydrolysis of cellulose: Noncomplexed cellulase systems. Biotech Bioeng 88 797± 824. Zhang Y-H P, Cui J B, Lynd L R and Kuang L R 2006a A transition from cellulose swelling to cellulose dissolution by o-phosphoric acid: evidence from enzymatic hydrolysis and supramolecular structure. Biomacromolecules 7 644-8. Zhang Y-H P, Himmel M and Mielenz J 2006b Outlook for cellulase improvement: screening and selection strategies. Biotechnol Adv 24 452±81. Zhang Y-H P, Ding S-Y, Mielenz J R, Cui J-B, Elander R T, Laser M, Himmel M E, McMillan J R and Lynd L R 2007 Fractionating recalcitrant lignocellulose at modest reaction conditions. Biotech Bioeng 97 214±23. Zhao H, Holladay J E, Brown H and Zhang Z C 2007 Metal chlorides in ionic liquid solvents convert sugars to 5-hydroxymethylfurfural. Science 316 1597±600. Zhao H, Jones C L, Baker G A, Xia S, Olubajo O and Person V N 2009 Regenerating cellulose from ionic liquids for accelerated enzymatic hydrolysis. J Biotechnol 139 47±54. Zhao X, Cheng K and Liu D 2009 Organosolv pretreatment of lignocellulosic biomass for enzymatic hydrolysis. Appl Microbiol Biotechnol 82 815±27. Zhou W, SchuÈttler H-B, Hao Z and Xu Y 2009 Cellulose hydrolysis in evolving substrate morphologies I: a general modeling formalism. Biotech Bioeng 104 261±74.

ß Woodhead Publishing Limited, 2010

3

Key features of pretreated lignocelluloses biomass solids and their impact on hydrolysis R . K U M A R , Zymetis, Inc., USA and C . E . W Y M A N , University of California, USA

Abstract: Prior to biological conversion of lignocellulosic biomass to ethanol or other products, natural barriers developed to protect plants must be overcome to realize efficient enzymatic hydrolysis, and a few pretreatment technologies are effective in inexpensively accomplishing this task through heating with chemicals. Over the years, changes in a number of structural and compositional attributes of biomass have been postulated to explain how pretreatment enhances enzymatic hydrolysis performance, but the complexity of biomass has always confounded development of a unified theory that can unequivocally predict how pretreated biomass solids will respond to enzymes. However, sugar release can be viewed as ultimately governed by two factors: 1) access of enzymes to cellulose and hemicellulose and 2) the effectiveness of enzymes attached to the surface in breaking down these carbohydrate chains to sugars and/or their oligomers. In this review, this perspective of enzyme access and effectiveness is applied to findings reported in the literature to provide a framework for understanding how various features in pretreated biomass solids could affect deconstruction of cellulose and hemicellulose to sugars and their yields. Key words: cellulase, cellulose, hemicelluloses, biomass, adsorption, accessibility, effectiveness, hydrolysis.

3.1

Introduction

Biological conversion of cellulosic biomass such as agricultural (e.g., corn stover) and forestry residues (e.g., sawdust) and herbaceous (e.g., switchgrass) and woody (e.g., poplar wood) energy crops into ethanol and other products offers the high yields to products vital to economic success, the potential for very low costs, and important strategic, environmental, and economic benefits (Farrell et al., 2006; Gomez et al., 2008; Lynd et al., 1991, 1996, 1999; Ragauskas et al., 2006; Schubert, 2006; Tilman et al., 2006; Wyman, 1999, 2003; Zhang, 2008). However, cellulosic materials have developed a natural resistance to biological attack to assure survival (Dhugga, 2007; Himmel et al., 2007), and a pretreatment step must be employed to overcome this resistance to high sugar yields (Chandra et al., 2007; Grethlein, 1984; Lynd et al., 2008;

ß Woodhead Publishing Limited, 2010

74

Bioalcohol production

Mosier et al., 2005; Sun and Cheng, 2002; Weil et al., 1994; Yang and Wyman, 2008). Dilute sulfuric acid is a leading option, but one economic study has projected it to be the most expensive single step in biomass conversion (Wooley et al., 1999). A few other pretreatment technologies based on heating biomass with ammonia, pH buffers, lime, or sulfur dioxide give similar cost and performance to dilute sulfuric acid (Mosier et al., 2005; Wyman et al., 2005a, 2005b), but lower cost options are still needed. In addition, the significant repercussions of pretreatment for other processing steps must be fully considered in the choice of pretreatment (Yang and Wyman, 2008). A more complete understanding of fundamental mechanisms responsible for pretreatment effectiveness would help accelerate development of lower cost approaches and improve their integration into the overall process (Wyman, 2007). Studies have attributed the effectiveness of pretreatment in improving enzymatic digestibility of biomass to increasing surface area and porosity (Chandra et al., 2008; Grethlein, 1984, 1985; Ishizawa et al., 2007; Mooney et al., 1997, 1998; Tanaka et al., 1988; Thompson et al., 1992; Wong et al., 1988; Zeng et al., 2007), removal of hemicellulose and lignin (Grohmann et al., 1986; Liu and Wyman, 2005; Pan et al., 2005; Yang and Wyman, 2004; Zhu et al., 2005), and reductions in cellulose crystallinity and the degree of polymerization (Chang and Holtzapple, 2000; Knappert et al., 1980; Puri, 1984; Yoshida et al., 2008). However, due to the complexity of biomass structures, changes in absolute cellulose crystallinity are difficult to determine accurately (Chang and Holtzapple, 2000; Puri, 1984; Puri and Pearce, 1986; Sun and Cheng, 2002). Interactions among other physical and chemical features of pretreated biomass make it difficult to isolate these variables and precisely determine which features have the greatest impact (Kumar et al., 2009; Lynd, 1996; Mansfield et al., 1999; Zhang and Lynd, 2004a). On top of that, enzymatic saccharification of cellulose is a heterogeneous reaction that requires successive adsorption of multiple enzymes on the surface for hydrolysis to occur (Kumar and Wyman, 2008; Lee and Fan, 1979; Ryu and Lee, 1982). These enzymes attain equilibrium with the substrate within an hour or two of incubation (Karlsson et al., 1999; Kumar and Wyman, 2009b; Lynd, 1996), with the amount of adsorbed enzymes not changing significantly over the course of hydrolysis (Eriksson et al., 2002b; Medve et al., 1998; Xu et al., 2008; Yu et al., 1995), especially for lignocellulosic biomass. Cellulase adsorption is generally quantified by fitting parameters to the Langmuir isotherm equation (Lynd et al., 2002; Walker and Wilson, 1991; Zhang and Lynd, 2004b), even though arguments have been made that this approach is oversimplified (Lynd et al., 2002). Thus, the role of pretreatment in rendering biomass digestible for enzymes is unfortunately still ambiguous and not well understood. In light of the complexity of biomass and the action of enzymes, we believe that enzymatic hydrolysis of cellulose and hemicellulose in pretreated biomass can be better viewed from the perspectives of the impact of substrate, enzyme,

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

75

and environmental chemical and physical factors on: 1) the accessibility of cellulose to enzymes, which is generally determined by the amount of enzyme adsorbed on cellulose in biomass, and 2) the effectiveness of the enzymes once they attach to cellulose. The emphasis of this review will be on identifying key substrate aspects that can impact enzyme adsorption or effectiveness or both based on information reported in the literature. Less detailed consideration will be given to important enzyme characteristics and physical parameters that likely impact these two potentially governing factors.

3.2

Key substrate features controlling cellulose hydrolysis: crystallinity

3.2.1

Accessibility

Enzymes are reported to rapidly hydrolyze amorphous cellulose to cellobiose and glucose, while the hydrolysis of crystalline cellulose is much slower, with the conclusion that the rate depends on cellulose crystallinity (Bertran and Dale, 1985; Ghose and Bisaria, 1979; Wood et al., 1989). The ordered structure of crystalline cellulose would impact the ability of cellulase to access cellulose based on the concept that a layer of cellulose must be removed before enzymes can reach layers (Fan et al., 1980; Lee and Fan, 1983; VaÈljamaÈe et al., 1999) and active sites lying underneath (Kongruang and Penner, 2004; Kongruang et al., 2004; Teeri, 1997; Zhang and Lynd, 2005), and studies that report rates slowing with increasing cellulose crystallinity are consistent with this hypothesis (Fan et al., 1981; Sasaki et al., 1979; Sinitsyn et al., 1991). However, others have observed the opposite effect to be true: hydrolysis increases with crystallinity (Grethlein, 1985; Puri, 1984), though the results for real biomass may be misinterpreted because removal of amorphous lignin and/or hemicellulose would increase biomass crystallinity and enhance digestibility. Furthermore, if the hydrolysis rates are much slower for crystalline regions, a classical question arises as to why crystallinity does not increase over the course of cellulose hydrolysis as a result of more rapid removal of amorphous cellulose (Ooshima et al., 1983; Paralikar and Betrabet, 1977). However, no significant change in crystallinity has been measured over the course of cellulose hydrolysis (Boisset et al., 1999; Chen et al., 2007; Lenz et al., 1990; Puls and Wood, 1991). In addition, in some cases, cellulose crystallinity was considered to have no effect on hydrolysis rates (Converse, 1993; Gharpuray et al., 1981; Kim and Holtzapple, 2006; Mansfield et al., 1999; Puri, 1984; Puri and Pearce, 1986; Rivers and Emert, 1988a, 1988b). The following points may help address this conundrum and understand the mechanism better. First, recent studies suggest that cellulases not only function as an hydrolytic agent but can simultaneously disrupt the cellulose structure to a significant extent (Himmel et al., 1999; Mansfield and Meder, 2003; Sinnott,

ß Woodhead Publishing Limited, 2010

76

Bioalcohol production

1998; Wang et al., 2008; Xiao et al., 2001). Thus, during hydrolysis, the action(s) of individual monocomponent enzymes are likely offset by concurrent modification by complementing enzymes (Mansfield and Meder, 2003). Second, for real biomass, crystallinity should not be confused with absolute cellulose crystallinity as real biomass has amorphous components other than cellulose (Kim and Holtzapple, 2006; Kumar et al., 2009). Third, use of high enzyme loadings to determine the impact of biomass features and other factors on hydrolysis may lead to misinterpretation by saturating the substrate. Fourth, almost all the characterization methods require a treatment before analysis such as drying, coating, etc., which may disturb the structure of the biomass. Nonetheless, better understanding of cellulases functioning at micro level and advanced analytical tools would help. Cellulase adsorption could be a useful measure of changes in cellulose accessibility with crystallinity. The enzyme adsorption capacity of amorphous cellulose is much greater than for crystalline material, leading one to expect amorphous regions to have greater hydrolysis rates and yields (> 50 times) than for crystalline areas (Hong et al., 2007; Jeoh et al., 2007; Lynd, 1996; MeunierGoddik and Penner, 1999; Ooshima et al., 1983; Pinto et al., 2006; Ryu and Lee, 1986; Sinitsyn et al., 1991; Zhang and Lynd, 2004b, 2005). Cellulase adsorption capacity is generally quantified based on the following Langmuir equation: ‰CEŠ ˆ

‰St Š‰Ef Š Kd ‡ ‰Ef Š

in which ‰CEŠ is the amount of adsorbed enzyme in mg/g, ‰Ef Š the free enzyme concentration in mg/ml,  the maximum adsorption capacity in mg/mg substrate, ‰St Š the substrate concentration in mg/ml, and Kd the equilibrium constant for the ratio ‰CŠ‰EŠ=‰CEŠ in mg of enzyme/ml. Representative values of the Langmuir parameters are summarized in Table 3.1 for a number of lignocellulosic materials reported in the literature; these parameters have been reviewed elsewhere for pure cellulose (Lynd et al., 2002; Walker and Wilson, 1991; Zhang and Lynd, 2004b). Unfortunately, most of the studies cited in Table 3.1 did not report cellulose/ biomass crystallinity, so a clear trend could not be seen between crystallinity and adsorption capacity. But as reported elsewhere (Kumar, 2008; Kumar et al., 2009) and discussed in the following sections, cellulose accessibility (as determined by cellulase adsorption) for a real biomass cannot be solely and clearly correlated with crystallinity. For example, Ooshima et al. reported that increasing pretreatment temperatures from 180 ëC to 220 ëC increased the adsorption capacity by almost six times, as shown in Table 3.1. This result would not be expected for pure cellulose because increased severity should remove more of the amorphous cellulose (Lenz et al., 1990; VaÈljamaÈe et al., 1999) and increase crystallinity, thereby making cellulose less accessible, as shown by Jeoh and corworkers (Jeoh et al., 2007). However, the opposite was

ß Woodhead Publishing Limited, 2010

Table 3.1 Langmuir parameters for lignocellulosic substrates for various enzymes and proteins

ß Woodhead Publishing Limited, 2010

Substrate/source

Enzyme/ Protein/ Brand name

Max. Ads. Capacity , mg/g subs.

Affinity A, ml/mg protein

Ads. Strength R ˆ   A, ml/g sub.

Birch, steam exploded Birch, steam exploded and alkali extracted

Celluclast 2L

214 237

2.1 2.8

42.8 663.6

Wheat straw, unpretreated but cut, milled, and sieved < 0.177 mm Wheat straw, NaOH pretreated

Trichoderma QM9414

8.34 (15 ëC)

7.19

60.0

71.46 (15 ëC)

2.27

162.3

Delignified rice straw, lignin < 5%

Trichoderma reesei, D1-6

256

0.16

41.0

Goel and Ramachandran (1983)

Hardwood, dilute acid pretreated at 180 ëC

Trichoderma reesei, GC 123, Genencor

14.1 (40 ëC)

12.5

176.3

Ooshima et al. (1990)

30.5 (40 ëC)

4.25

129.6

80.6 (40 ëC)

1.82

146.7

171.3

0.78

133.6

162.4

0.59

95.8

Hardwood, dilute acid pretreated at 200 ëC Hardwood, dilute acid pretreated at 220 ëC Douglas Fir, steam exploded Douglas Fir, steam exploded alkali extracted

Celluclast 2L + BetagNovozyme 188

Reference

Lee et al. (1994)

Estrada et al. (1988)

Lu (2002)

Table 3.1 Continued

ß Woodhead Publishing Limited, 2010

Substrate/source

Enzyme/ Protein/ Brand name

Max. Ads. Capacity , mg/g subs.

Affinity A, ml/mg protein

Corn stover, dilute acid pretreated at 190 ëC

Ads. Strength R ˆ   A, ml/g sub.

Cellulase, CPN from Iogen Beta-g, Novozyme 188

60

nd

10

nd

Willow, steam pretreated

Celluclast 2L

471

0.29

Spruce, steam pretreated

CBHI EGII

nd nd

5.5 (4 ëC) 1 (4 ëC)

Creeping wild ryegrass, dilute acid pretreated

Cellulase/ Celluclast 1.5 L Beta-g/ Novozyme 188

42.5

0.6

25.5

60.6

0.65

39.4

Corn stover, dilute acid pretreated at 140 ëC

Cellulase, Sp. CP Beta-g/ Novozyme 188 BSA

210 (4 ëC) 150 (50 ëC) 130 (4 ëC) 140 (50 ëC) 30 (4 ëC) 130 (50 ëC)

6.36 0.221 0.09 0.09 0.078 0.175

1335.0 33.2 11.7 12.6 2.34 22.75

Reference

Kadam et al. (2004)

136.6

Galbe et al. (1990) Palonen et al. (2004b) Zheng et al. (2007)

Willies (2007)

Ethanol pretreated lodgepole pine (EPLP), lignin 14.5%

ß Woodhead Publishing Limited, 2010

Spezyme CP

60.35

3.17

190

Celluclast

87.69

3.48

310

Steam exploded lodgepole pine (SELP), lignin  45.6%

Celluclast

101.05

1.48

150

Hardwood, dilute acid pretreated at 220 ëC Corn stover solids AFEXa ARP C. pH D. acid Lime SO2

C. thermocellum

317 (60 ëC)

344

10.9E+4

Bernardez et al. (1993)

Genencor Spezyme CP (@ 4 ëC)

99.7 113.8 101.7 90.7 133.6 124.8

1.86 46.2 0.77 2.49 0.88 0.90

185.4 5257.6 78.3 225.8 100.0 112.3

Kumar and Wyman (2009b)

Genencor Spezyme CP (@ 4 ëC)

107.4 113.5 56.2 170.9 195.2 150.8 142.2

0.21 0.11 0.43 0.94 0.08 0.09 1.14

23.0 13.1 23.9 159 15.6 14.5 161.0

Kumar and Wyman (2009a)

Poplar solids AFEX ARP C.pH D.acid Flowthrough Lime SO2 a

Tu et al. (2007)

Pretreatment type ; AFEX ± ammonia fiber expansion; ARP ± ammonia recycled percolation; C.pH ± controlled pH; D. acid ± dilute acid.

80

Bioalcohol production

3.1 Effect of cellulose crystallinity on maximum cellulose adsorption capacity (Lee et al., 1980). SSA is the specific surface area.

observed because increasing temperature (severity) not only makes possible changes in cellulose crystal structure but removes hemicellulose (Grethlein, 1985). In addition, cellulose DP is reduced (Knappert et al., 1980; Kumar et al., 2009), and lignin-hemicellulose/cellulose bonds are no doubt ruptured (Gupta et al., 2008; Kumar et al., 2009). For pure cellulosic substrates, Lee and coworkers (Lee et al., 1982) reported a decline in adsorption capacity of cellulose for complete cellulases1 with increasing crystallinity, as shown in Fig. 3.1. Although the specific surface area (SSA in Fig. 3.1) did not increase as crystallinity dropped, typically SSA and crystallinity are related to solids produced by mechanical pretreatments. Ooshima et al. documented similar cellulase adsorption patterns at 5 ëC for cellulose of varying crystallinity prepared by enzymatic digestion of Avicel cellulose for different times (Ooshima et al., 1983). Similarly, Hoshino et al. showed that purified exo and endo cellulases of Irpex lecteus had an inverse correlation between cellulose crystallinity and the maximum amount of protein adsorbed, as shown in Fig. 3.2 (Hoshino and Kanda, 1997; Hoshino et al., 1992). In a kinetic study, Ryu et al. demonstrated an increase in adsorption kinetic parameters with a drop in crystallinity (Ryu and Lee, 1986). In another study, Sinitsyn and coworkers (1991) reported an inverse correlation between crystallinity of pure cellulose and the adsorption of peroxidase and chymotrypsin proteins on cellulose. However, for baggase, protein adsorption was shown to increase with 1. Throughout the chapter complete cellulose(s) refers to the crude mixture containing two cellobiohydrolases (CBHI and CBHII), five endoglucanases (EG I to EG V), and a -glucosidase, unless otherwise stated.

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

81

3.2 Effect of cellulose crystallinity on maximum adsorption capacity and equilibrium constants for exo and endocellulase (Hoshino et al., 1997). Wmax is the maximum amount of enzyme adsorbed and K is the equilibrium constant.

delignification and a reduction in crystallinity index (CrI). For sodium hydroxide pretreated wheat straw, Estrada and coworkers (Estrada et al., 1988) found an inverse correlation between adsorption parameters and crystallinity. A study of cellulose binding domains and cellulose interaction showed greater adsorption of binding domains to amorphous than to crystalline cellulose (Pinto et al., 2006). Recently, Joeh and coworkers (Jeoh et al., 2007) revealed that crystallinity greatly reduces adsorption of Cellobiohydrolase I (Cel7A; CBHI), leading to a decreased extent of hydrolysis. Furthermore, air drying of dilute acid pretreated corn stover resulted in a decrease in the extent of CBHI adsorption, probably due to `hornification' of fibers (Esteghlalian et al., 2001) and/or increased crystallinity due to drying (Weimer et al., 1995). Different cellulase components have different adsorption capacities and activities (Lynd, 1996; Zhang and Lynd, 2004b), as shown in Table 3.2, where Avicel is highly crystalline (CrI = ~60%) and has a shorter cellulose chain length (DP ~ 300) than filter paper (CrI = ~40%; DP ~ 750±2800) (Zhang et al., 2006). Endoglucanse-I, which attacks and adsorbs preferentially on amorphous cellulose, was measured to have an average adsorption capacity and activity greater than for CBH-I on both types of cellulose studied. A similar pattern for Endoglucanase I (EGI, Cel7B) was reported by Ding and Xu (2004), but Klyosov (1982) observed that the adsorption capacity of endoglucanses from Trichoderma reesei did not depend on cellulose crystallinity. Yet, contrary to the numerous studies mentioned above, working with pure cellulose and lignocellulosic substrates, Goel and Ramachandran (1983) found no correlation

ß Woodhead Publishing Limited, 2010

Table 3.2 Adsorption capacity for cellulase components and their activity on cellulose substrates Parameter

ß Woodhead Publishing Limited, 2010

Maximum adsorption capacity (mg/g or mol/g)

Substrate

Avg.

Filter paper

Temperature/enzyme component

Temperature/enzyme component

Temp (ëC)

CBH-I

Temp (ëC)

EG-I

Temp (ëC)

20 25 4 20 40 30

69 70 48 51.8 40 63

30

126

50

Avg. Specific activity (mol glucose Equiv./mg/min)

Avicel

57 50 45 40 30

0.065 0.04 0.012 0.019 0.034

126 50 45 40 30

0.045 0.17 0.0046 0.196 0.104

*The data shown above were adapted from Lynd et al. (2002) and Zhang and Lynd (2004b).

CBH-I 0.17

Temp (ëC) 50

0.17 50 50 40

0.08 0.22 0.0046 0.102

EG-I 0.17

0.17 50 50 40

0.18 1.2 0.0023 0.461

Pretreated lignocelluloses biomass solids

83

between crystallinity and adsorption of cellulase enzymes activities. Furthermore, Banka et al. showed that adsorption of a non-hydrolytic protein designated Fibril Forming Protein (FFP) from Trichoderma reesei increased with crystallinity (Banka and Mishra, 2002).

3.2.2

Effectiveness

In addition to accessibility, cellulose crystallinity would likely impact the effectiveness of adsorbed cellulase components. The literature shows that cellulose crystallinity affects the synergism between cellulase components (Henrissat, 1994; Henrissat et al., 1985; Hoshino and Kanda, 1997; Hoshino et al., 1997; Kanda et al., 1980; Murashima et al., 2002; Nidetzky et al., 1993; Tarantili et al., 1996; VaÈljamaÈe, 2002; VaÈljamaÈe et al., 1999; Zhang and Lynd, 2004b). Hoshino et al. found increased synergism between CBHI and Endoglucanase II (EGII) from T. ressei with increased crystallinity and determined the highest synergism between Cellobiohydrolase II (CBHII, Cel6B) and EGII to be for a crystallinity index ~ 1.0. In another study, Igarashi and coworkers showed that the nature of the crystalline cellulose polymorph also affected hydrolytic activity of adsorbed Cel7A; for example, the maximum cellulase adsorption capacity on cellulose I was approximately 1.5 times that for cellulose I, although the rate of cellobiose generation from cellulose I was lower than that from cellulose I (Igarashi et al., 2006a, 2006b, 2007). Moreover, Mizutani et al. (2002) and Gama and Mota (1997) showed that the beneficial impact of surfactant on saccharification is influenced by crystallinity for pure cellulose. However, there is evidence that the presence of surfactants helps reduce unproductive adsorption of enzymes not only on lignin but on cellulose as well (Eriksson et al., 2002a; Kumar and Wyman, 2009d; Ooshima et al., 1986) and thus enhances their effective activity. Furthermore, several `restart studies' with pure microcrystalline cellulose have shown that unproductive binding of enzymes is one of the main reasons for the slow down of hydrolysis rate over hydrolysis time (Kumar and Wyman, 2009c; Ooshima et al., 1991; Yang et al., 2006). Besides, Ma and coworkers in a recent study have shown that irreversibly surface bound CBHI loses up to 70% of its activity in just 10 minutes (Ma et al., 2008). On a different note, Gruno et al. reported that endproduct inhibition of cellulase was higher for crystalline cellulose than amorphous (Gruno et al., 2004). Therefore, it appears that crystallinity impacts enzyme effectiveness. A more limited literature indicates that the processivity of the dominant enzyme of the Trichoderma system, Cel7A (CBHI) is affected by cellulose crystallinity. A rough estimate of processivity as measured in terms of the ratio of cellobiose to glucose released from bacterial microcrystalline cellulose (BMCC, CrI ~ > 85 %) and amorphous cellulose was reported to be 23 and 14, respectively, by Ossowski and coworkers (von Ossowski et al., 2003). In another

ß Woodhead Publishing Limited, 2010

84

Bioalcohol production

study, these processivity measures for Trichoderma reesei Cel7A were reported to be 88  10, 42  10, and 34  2:0 cellobiose units for bacterial cellulose (BC, CrI ~ 88), bacterial microcrystalline cellulose (BMCC, CrI ~ 92), and endoglucanse-pretreated bacterial cellulose (unknown CrI), respectively (Kipper et al., 2005). Although the study offered no explanation, the low processivity values for BMCC and endoglucanase treated BC could be due to hindrance by solitary chains left after erosion of the surface by the enzyme, unproductively adsorbed enzymes, and/or the nature of the substrate (Henrissat, 1998; Nutt et al., 1998; VaÈljamaÈe et al., 1999). Overall, because of the importance of enzyme action to hydrolysis rates, further studies are needed to elucidate the role of crystallinity on the processivity of Cel7A and other processive or pseudoprocessive enzymes from various micro-organisms (Horn et al., 2006). Per theoretical models, it is believed that crystalline cellulose chains have some structural irregularities that provide attack sites for endoglucanses (Lynd et al., 2002; Teeri, 1997). However, the literature leads us to believe that some cellulase components bound to this amorphous fraction of cellulose, which would be smaller in size for highly crystalline cellulose, may not be very active due to their large size, compared to the length of the amorphous fraction of cellulose microfibrils. In particular, the 35±40 cellobiose lattices occupied per bound cellulase molecule (Hong et al., 2007; Zhang and Lynd, 2004b) and/or their slow catalytic reaction rates due to the low processivity of CBHI (von Ossowski et al., 2003) limit the effectiveness of endo-glucanases by covering their preferential active sites. Incomplete hydrolysis of amorphous cellulose by Cel7A reported by Joeh et al. (2007) and competition among cellulase components during adsorption (Kyriacou et al., 1989; Nidetzky et al., 1996; Nieves et al., 1991; Ryu et al., 1984), mostly studied for microcrystalline cellulose, strongly support this idea. Furthermore, a nonlinear correlation was observed between CBHII adsorption and activity on filter paper (CrI ~ 40%) by Nidetzky et al. (1994a), and EGI was found to be more active compared to CBHI and CBHII, even though the maximum binding capacity of EGI was roughly equal to that for CBHI and much lower than for CBHII. In another study, Eriksson et al. found that although Cel7A (CBHI) unproductively adsorbed on steam-pretreated spruce (SPS), it desorbed back into solution when supplemented with Cel7B (EGI), enhancing conversion, perhaps due to preferential attack of amorphous cellulose by EGI releasing unproductively adsorbed CBHI (Eriksson et al., 2002b).

3.3

Key substrate features controlling cellulose hydrolysis: degree of polymerization (DP)

Several studies and literature reviews discuss the change in DP of insoluble and soluble cellulose during and after hydrolysis by complete cellulase mixtures or mono-components (Cao and Tan, 2002; Converse, 1993; Eremeeva et al., 2001;

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

85

Hilden et al., 2005; Kanda et al., 1976; Kleman-Leyer et al., 1996; Mansfield and Meder, 2003; Mansfield et al., 1999; Martinez et al., 1997; Pala et al., 2007; Zhang and Lynd, 2004b, 2005). However, the understanding of the impact of cellulose chain length on hydrolysis is still limited, and questions about cellulose DP and what role, if any, cellulose chain length plays in cellulose hydrolysis are still unanswered. Among the very few studies on this subject, Puri and coworker (Puri, 1984; Puri and Pearce, 1986) showed that a reduction in cellulose DP improved hydrolysis, but a lack of data on the effect on surface area and other substrate features makes conclusions of this study inconclusive. In another study, Knappert and coworkers developed a qualitative relationship between cellulose DP and digestibility (Knappert et al., 1980). Sinistyn and coresearchers showed that reduction in DP of cotton linters by -irradiation, while keeping crystallinity index (CI) constant, had a negligible impact on hydrolysis rates (Sinitsyn et al., 1991). A recent kinetic study by Zhang and Lynd indicated that a decrease in cellulose DP had a less effect on accelerating hydrolysis rates than increasing the accessibility of -glycosidic bonds as measured by the maximum amount of cellulase adsorbed on cellulose (Zhang and Lynd, 2006). However, the possibility of how cellulose chain length (DP) may affect accessibility was not discussed.

3.3.1

Accessibility

Given the typically large amount of CBHI in cellulase (>65%) and its catalytic site preferences (Beldman et al., 1985; Christina Divne, 1998; Nidetzky et al., 1994b; Teeri, 1997; Teeri et al., 1995), one could conclude that DP reduction should improve hydrolysis effectiveness by making more reducing chain ends available to CBHI (and non-reducing ends available to CBHII, which is about 20% of the total cellulase protein), with the result that lowering DP would be a promising target to enhance cellulose accessibility. However, cellulose crystallinity and DP appear to be closely correlated for mechanical pretreatments and the majority of thermochemical methods such as steam explosion and dilute acid, making it difficult to differentiate which controls (Chandra et al., 2007; Kumar et al., 2009). For example, mechanical pretreatments such as ball milling generally reduce both crystallinity and DP (Caulfield and Moore, 1974; Lee et al., 1982; Oh and Kim, 1987; Schwanninger et al., 2004; Sinitsyn et al., 1991) but could also affect lignin structure in real biomass, making it difficult to isolate the effect of just cellulose DP on enzyme adsorption and hydrolysis. Increasing thermochemical pretreatment severity removes a substantial portion of the amorphous region (Kumar et al., 2009), increasing substrate crystallinity, but the cellulose chain length may also drop to the level off degree of polymerization (LODP) (HaÊkansson and Ahlgren, 2005; Heitz et al., 1987; Kumar et al., 2009; Martinez et al., 1997; Millett et al., 1954; Treimanis et al., 1998). The relationship of viscosity average degree of polymerization (DPv) to

ß Woodhead Publishing Limited, 2010

86

Bioalcohol production

3.3 Effect of pretreatment severity on cellulose degree of polymerization (DPv) as measured by the viscosity method for corn stover solids prepared by leading pretreatment technologies. The line is shown to help follow the trend but is not fit to the data. AFEX ± ammonia fiber expansion, ARP ± ammonia recycled percolation, DA ± dilute acid, CpH ± controlled pH, SO2 ± sulfur dioxide (Kumar, 2008; Kumar et al., 2009).

pretreatment severity2 (log R0) is shown in Fig. 3.3 for corn stover solids prepared by leading pretreatment technologies that all employ heating with chemicals (Mosier et al., 2005; Wyman et al., 2005b), and DP drops with severity for almost all of these options. However, crystallinity3 can also be related to DP for several pretreatments, as shown in Fig. 3.4, clouding the interpretation of this data due to the drop in cellulase adsorption with increasing crystallinity discussed before. In another study, Engstrom and coworkers found that pulp's accessibility and reactivity for the viscose process increased significantly following treatment with monocomponent endoglucanases, which also resulted in DP reduction; however, similar results, at a comparable DP level, were not observed when pulp was treated with acid (Engstrom et al., 2006). Although the information on the effect of cellulose DP on cellulase adsorption is limited, Kaplan and coworkers (Kaplan et al., 1970) showed a significant drop in cellulase adsorption and associated lower hydrolysis of altered cellulose following photochemical degradation, which was probably due to a decrease

2. Severity factor, defined as R0 ˆ t
exp‰…TH ÿ TR †=14:75Š, includes only time and temperature parameters; however, all of these pretreatments except controlled pH (CpH) utilize different chemicals at various concentrations. 3. Crystallinity values were adopted from Laureano-Perez et al. (2005).

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

87

3.4 Crystallinity vs. cellulose viscosity degree of polymerization for corn stover solids prepared by leading pretreatment technologies. AFEX ± ammonia fiber expansion, ARP ± ammonia recycled percolation, DA ± dilute acid, CpH ± controlled pH, SO2 ± sulfur dioxide (Kumar, 2008).

in cellulose DP and some ring opening for weathered cotton cellulose. Yet, it is a well known quoted fact that 80% of fungal cellulase protein (CBHI and CBHII) preferably attacks chain ends (Carrard and Linder, 1999; Henrissat et al., 1985; Teeri et al., 1995), but unfortunately almost nothing has been done to conclusively show the impact of cellulose DP on cellulase adsorption.

3.3.2

Effectiveness

Theoretically, the lower the DP, the more reducing and non-reducing ends are available, and one would expect that more CBHI/II would be able to work at one time while making it easier for endoglucanases to act. For soluble cellulose, Nidetzky et al. found that the initial degradation velocity of cello-oligosaccharides by CBHI increased with DP below cellohexose and then remained constant for higher DP (Nidetzky et al., 1994b). Similar effects of DP for soluble cellodextrins on CBHII and EGI activity are reviewed elsewhere (Zhang and Lynd, 2004b). Furthermore, a decrease in -glucosidase activity with increasing DP has been reported (Lee and Fan, 1980; Wilson et al., 1994). However, to the authors' knowledge, no information is available on the effect of insoluble cellulose DP on the catalytic efficiency of cellulase except that higher DP could result in higher synergy between CBHI and EGI (Henrissat, 1994; Okazaki et al., 1981; Okazaki and Moo-Young, 1978; Zhang and Lynd, 2006). Furthermore, cellulose DP may affect the processivity index, with full processivity of

ß Woodhead Publishing Limited, 2010

88

Bioalcohol production

CBHI possibly not realized for short chains (Gupta and Lee, 2009; VaÈljamaÈe et al., 1999). Overall, studies of the effect of DP and crystallinity on enzymatic digestibility demonstrated that the susceptibility of pretreated substrates to enzymatic hydrolysis could not be easily predicted from the differences in their cellulose DP and crystallinity (Puri, 1984; Ramos et al., 1993), likely due to the complexity of real cellulosic substrates.

3.4

Key substrate features controlling cellulose hydrolysis: hemicellulose and degree of hemicellulose acetylation

3.4.1

Accessibility

It has been postulated that hemicellulose impedes access to cellulose by forming a sheath around glucan chains (Berlin et al., 2007; Ding and Himmel, 2006; Himmel et al., 2007; Jeoh et al., 2007; Kumar and Wyman, 2009f, 2009g; Selig et al., 2008; Yoshida et al., 2008), and several studies showed a direct relationship between cellulose digestion and hemicellulose removal (Allen et al., 2001; Grohmann et al., 1986; Ishizawa et al., 2007; Jeoh et al., 2005; Kabel et al., 2007; Kim et al., 2001; Palonen et al., 2004a; Um et al., 2003; Yang and Wyman, 2004; Zhu et al., 2005), with some even concluding that lignin removal is not necessary for good cellulose conversion (Clark et al., 1989; Grohmann et al., 1986). However, some substrates required high temperatures for the same degree of hemicellulose removal to be effective, suggesting that hemicellulose removal is not the only factor impacting digestibility (Torget et al., 1991; Yang et al., 2004). In addition, some reports do not postulate any role for hemicellulose removal in changing cellulose digestibility (Fan et al., 1982; Millett et al., 1975; Tsao et al., 1978). Unfortunately, hemicellulose alteration can also disrupt other biomass components (Chum et al., 1988; Grethlein, 1984; 1985; Iyer and Lee, 1999; Kumar et al., 2009; Maloney et al., 1985), making it challenging to draw firm conclusions about the degree to which it controls access of enzymes to cellulose. In addition, some contend that hemicellulose may actually be a marker related to disruption of the far less soluble lignin and that lignin disruption could be the key to greater digestion (Liu and Wyman, 2003, 2004a, 2004b, 2005; Yang and Wyman, 2004). Less attention has been given to how the degree of acetylation of the substrate impacts cellulose digestion. Hemicellulose chains are extensively acetylated in many types of biomass, and deacetylation was reported to enhance cellulose digestibility significantly, with some differences noted in the degree of removal needed (Kim and Holtzapple, 2005; Kumar and Wyman, 2009e; Lemos et al., 2000; Wood and McCrae, 1986). Removing hemicellulose also removes acetyl groups (Kabel et al., 2007; Maloney et al., 1985) and usually alters the form of lignin (Ooshima et al., 1990; Selig et al., 2007) left on the material, making it

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

89

difficult to isolate which factor was most influential in improving performance. One study showed that this effect appeared to become less important beyond removal of 75% of the acetyl groups (Grohmann et al., 1989), while other studies demonstrated continual improvements up to full removal (Kong et al., 1992; Kumar and Wyman, 2009e, 2009f). Grohmann and coworkers showed that removing acetyl esters from aspen wood and wheat straw made them 5 to 7 times more digestible, and Kong and coworkers (1992) observed a major effect of removing the acetyl content of aspen wood on cellulose digestibility even though lignin and polysaccharides were left in place. Consistent with this, Kumar and Wyman observed a significant enhancement in glucan (from 17 to ~40%) and xylan (from 6% to ~30%) digestibility by selective removal of acetyl groups from corn stover (Kumar and Wyman, 2009f), and > 60% of glucan and xylan digestion was realized with further supplementation of xylanase to cellulase. However, Chang and Holtzapple (2000) applied similar methods to poplar wood as above but showed that removal of acetyl bonds is less important than reduction in crystallinity and/or removal of lignin. Unfortunately, it is still debatable whether hemicellulose removal or the breakdown of cross-linked network of polysaccharides and bonds among them is responsible for enhanced digestion of cellulose in pretreated biomass. For example, Weimer and coworkers (2000) suggested that intimate association of xylan and cellulose does not inhibit the bio-degradability of polysaccharides. Furthermore, from a more applied perspective, some pretreatments such as Ammonia Fiber Expansion (AFEX) produce highly digestible cellulose without removing much hemicellulose (Dale et al., 1996; Teymouri et al., 2005; Vlasenko et al., 1997) but remove acetyl groups and probably other side chains from xylan, disrupting linkages among carbohydrates and lignin to a significant extent (Chundawat et al., 2007; Kumar et al., 2009). Although the role of acetyl groups and other side chains removal may seem limited, it is pretty clear that removal of these side chains during pretreatment would surely result in the reduction of enzyme requirements and enhance both xylan and glucan digestions as well (Fernandes et al., 1999; Kumar and Wyman, 2009f; Selig et al., 2008). Although its role in enhancing cellulose digestion is ambiguous, hemicellulose/xylan removal during pretreatment may be desirable for economic and technical reasons such as higher recovery of xylose and less need for hemicellulose degrading and accessory enzymes (Hespell et al., 1997; Knauf and Moniruzzaman, 2004; Kumar and Wyman, 2009g; Merino Sandra and Cherry, 2007). In addition, in a recent study we showed that removing hemicellulose during pretreatment can reduce cellulase/xylanase inhibition by soluble xylooligomers generated during enzymatic hydrolysis (Kumar and Wyman, 2009e, 2009f). Similarly, Kim et al. showed that the effluent exiting from Ammonia Recycled Percolation (ARP) pretreatment of corn stover; containing mostly xylooligomers, soluble lignin, and sugar and lignin degradation products, inhibited cellulase and microbial activity significantly (Kim et al., 2006).

ß Woodhead Publishing Limited, 2010

90

Bioalcohol production

Furthermore, Suh and Choi showed that xylooligomers inhibited endo-xylanase action (Suh and Choi, 1996). We believe that even slight branching of hemicellulose and its acetylated network can interfere with cellulase access to cellulose (Karlsson et al., 2002; Pan et al., 2006; Samios et al., 1997; Yu et al., 2003), but this is difficult to prove in that direct information on the effect of acetylation and hemicellulose on cellulose accessibility is scarce. However, Jeoh and coworkers (2005, 2007) recently reported increased cellulose accessibility, as measured by the adsorption of fluorescent labeled Cel7A (CBHI), and an increase in hydrolysis with the extent of xylan removal. It is also reported in several recent studies that supplementation of cellulase with xylanase, which should selectively only remove xylan, not only enhanced xylan conversion but glucan digestion as well. In addition, the linear relationship generally found between xylan and glucan digestion basically indicates that xylan removal affects cellulose accessibility (Berlin et al., 2007; Beukes et al., 2008; GarcõÂa-Aparicio et al., 2007; Gupta et al., 2008; Kumar and Wyman, 2009f, 2009g; Murashima et al., 2003; Selig et al., 2008). Hemicellulose deposition on cellulose during pretreatment (Gray et al., 2003, 2007; Kumar and Wyman, 2009f; Linder et al., 2003; Nagle et al., 2002) could also reduce the amount of cellulose available for cellulase action. Pan et al. in a study suggested (Pan et al., 2006) that acetyl groups in pulp may restrict cellulase accessibility to cellulose by inhibiting productive binding through increasing the diameter of cellulose and/or changing its hydrophobicity. Selective deacetylation of corn stover by the Kong et al. method (Kong et al., 1992) enhanced CBHI adsorption significantly more than delignification, increased the initial rate, and produced greater digestibility of cellulose and xylan as well, indicating increased cellulose accessibility (Kumar and Wyman, 2009b, 2009f). However, not much information is available in the literature to clarify whether selective hemicellulose removal and/or deacetylation impacts cellulase adsorption/accessibility, and further study is needed to understand the impact of xylooligomers on cellulase (xylanase) adsorption.

3.4.2

Effectiveness

For enzymatic hydrolysis of lignocellulosics, deacetylation and removal of other side chains may indirectly affect cellulase effectiveness through removing bonds/linkages to xylose that xylanase could not otherwise hydrolyze, thereby making xylanase more effective (Anand and Vithayathil, 1996; Fernandes et al., 1999; Glasser et al., 1995; Grabber et al., 1998a; Grohmann et al., 1989; Kormelink and Voragen, 1992; Mitchell et al., 1990; Rivard et al., 1992; Shallom and Shoham, 2003; Suh and Choi, 1996; Tenkanen et al., 1996; Wood and McCrae, 1986), which in turn increases cellulose digestibility (GarcõÂaAparicio et al., 2007; Kumar and Wyman, 2009a, 2009b; Murashima et al., 2003; Tabka et al., 2006; Yu et al., 2003). Although the effect of xylan removal

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

91

on cellulase efficiency is not yet known, it presumably affects the processive action of Cel7A by binding cellulase unproductively (Chernoglazov et al., 1988; Tenkanen et al., 1995) and, as discussed earlier, xylan oligomers, released during hydrolysis and pretreatments, strongly inhibit enzymes activity (Kim et al., 2006; Kumar and Wyman, 2009c, 2009e, 2009g; Suh and Choi, 1996). Although the direct effect of acetyl groups on cellulase effectiveness, however, may not yet be clear, they certainly affect xylanase effectiveness, as shown by Kumar and Wyman (Kumar and Wyman, 2009b, 2009f). Some literature reports further lead us to believe that acetylated/substituted xylooligomers should be much more inhibitory to enzymes effectiveness than just plain xylooligomers (Kumar and Wyman, 2009e; Suh and Choi, 1996), as removal of acetyl groups/ substitution from soluble xylooligomers by means of hydrolytic action of accessory enzymes such as acetyl xylan esterase and L-arabinofuranosidase would facilitate break down of xylooligomers by xylanase and beta-xylosidase and consequently would have lesser impact on cellulase action. Thus, more work is needed to clarify whether hemicellulose removal and deacetylation impact the accessibility of cellulase to cellulose or the effectiveness of cellulase on cellulose or both.

3.5

Key substrate features controlling cellulose hydrolysis: lignin

3.5.1

Accessibility

Lignin binds cellulosic fibers together in a composite structure with excellent properties but also shields cellulose from accessibility to enzymes (Wyman et al., 2005b). Various studies reported cellulose hydrolysis was improved with increasing lignin removal, although differences were reported in the degree of lignin removal needed (Converse, 1993; Grethlein, 1984; Yang et al., 2002; Yang and Wyman, 2004). Besides the degree of lignin removal, the ratio of syringyl to guaiacyl moieties in the lignin was considered to significantly influence digestibility (Yamamoto et al., 1990). Overall, the protective lignin sheath is thought to present a major impediment to enzymatic hydrolysis of cellulose in pretreated biomass by restricting enzyme accessibility to cellulose (Chandra et al., 2007; Chapple et al., 2007; Mansfield et al., 1999; Pan et al., 2005; Saddler et al., 1982; Taniguchi et al., 2005). The majority of studies in the literature have reported that enzymatic conversion of polysaccharides is enhanced by delignification of hardwood/softwood and lignocellulosics (Chang and Holtzapple, 2000; Cunningham et al., 1981; Gharpuray et al., 1981; Kabeya et al., 1993; Koullas et al., 1993; Liao et al., 2005; Morrison, 1983; Sawada et al., 1995; Schwald et al., 1988a; Stinson and Ham, 1995; Sudo et al., 1976; Yu et al., 1998); however, others found none or a negative correlation between lignin content/removal and digestibility of residual cellulose

ß Woodhead Publishing Limited, 2010

92

Bioalcohol production

(Draude et al., 2001; Jeoh et al., 2005; Kim et al., 2001; Saddler et al., 1982; Wong et al., 1988). Overall, the exact role of lignin in limiting hydrolysis has been difficult to define. One of its most significant effects is on fiber swelling and the resulting influence on cellulose accessibility (Mooney et al., 1998; Nelson and Oliver, 1971). For example, Yuldashev et al. observed that the amount of cellulase on the surface of cotton stalks (cellulose ± 44%, lignin ± 26.4%) was lower than for milled cotton stalks (cellulose ± 92%, lignin ± 0.6%), leading to a drop in conversion; however, lignin did not inactivate free or bound enzyme (Yuldashev et al., 1993). In another study, Ishihara and coworkers determined that lignin slows down enzyme adsorption but does not restrict carbohydrate conversion for steamed shirakamba wood (Ishihara et al., 1991). Limited delignification of wheat straw by sodium hydroxide, though not selective, was shown to increase cellulase adsorption by Estrada et al. (1988). Conversely, Mooney et al. concluded that the proportion of lignin does not influence cellulase adsorption for four different types of pulp that differed in lignin content (Mooney et al., 1997). Although several studies suggested that lignin removal/or lignin content does not affect cellulase adsorption on cellulose/biomass significantly (Eriksson et al., 2002a; Lu et al., 2002; Mooney et al., 1998), it has rarely been shown experimentally whether selective lignin removal affects cellulase adsorption. For the first time, Kumar and Wyman showed that selective removal of lignin from corn stover did not significantly increase cellulase accessibility to cellulose, as measured by purified Cel7A adsorption. Instead, lignin removal appeared to more directly affect xylan accessibility, which in turn affected cellulose accessibility, as evidenced by a much higher increase in xylan digestion than glucan and a linear relation between the percentage increase in xylan and glucan conversions (Kumar and Wyman, 2009b). Consistent with this hypothesis, in another study, we found that lignin removal by the acid chlorite method from biomass solids pretreated with high pH pretreatments, which leave most of the xylan in place, resulted in much higher enhancement of glucan and, especially, xylan digestibility, compared to low pH pretreatments such as dilute acid and SO2 steam explosion, which are known for their effectiveness in removing most of the hemicellulose during pretreatment (Kumar and Wyman, È hgren et al. also found a negligible impact of delignification on glucan 2009a). O È hgren et al., 2007). Furthermore, digestibility of steam exploded corn stover (O consistent with the above findings, Selig and coworkers reported that lignin appears to have a more direct impact on xylan than glucan accessibility by purified cellulase and xylanase activities, which in turn occludes glucan accessibility (Selig et al., 2009). Several studies in previous years reported that lignin removal affects hemicellulose more than glucan hydrolysis (Beveridge and Richards, 1975; Ford, 1983; Mes-Hartree et al., 1987; Morrison, 1983; Prabhu and Maheshwari, 1999; Teixeira et al., 1999). For example, Chang and coworkers applied lime pretreatment to effectively remove lignin from switch-

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

93

grass with a 5 and 21 times increase in glucan and xylan digestibility, respectively (Chang et al., 1997). On a different note, Mes-Hartree and coworkers employing biologically delignified aspen wood (BDA; 44% lignin removal) and steamed aspen wood for cellulase production showed that Trichoderma harzianum produced a low level of cellulase and gave significantly lower sugar yields for BDA than steamed aspen wood, because the latter had fewer pentosans than BDA and delignification did not result in enhanced cellulose accessibility (Mes-Hartree et al., 1987). Lignin has been claimed to depolymerize, dissolve, repolymerize, and then precipitate during pretreatment by hemicellulose hydrolysis, although no doubt in a different morphology that could change its impact on cellulose digestion (Donohoe et al., 2008; Li et al., 2007; Ramos et al., 1993; Schell et al., 1991; Schwald et al., 1988b; Selig et al., 2007; Shevchenko et al., 1999; Yang and Wyman, 2004). In addition, there is evidence that the high solubility of hemicellulose could aid in taking lignin into solution despite the low solubility of the later (Gray et al., 2003, 2007), but that the lignin would fall back onto the biomass once it breaks free from hemicellulose and polymerizes to low solubility compounds (Liu and Wyman, 2003, 2004a, 2004b, 2005; Yang et al., 2004). The removal/disruption of lignin may not only increase accessibility of xylan and cellulose, though indirectly, but also make more cellulase and other enzymes available to act (Kumar and Wyman, 2009d; Yang and Wyman, 2004). Because lignin is physically and chemically resistant to attack by enzymes, irreversiblly absorbs cellulase (and other enzymes), and acts as a impenetrable barrier to cellulase, its presence limits xylan/cellulose accessibility (Kumar and Wyman, 2009a, 2009b; Lu et al., 2002). Adsorption of enzymes/proteins on lignin has been shown to follow a Langmuir isotherm, with typical parameters shown in Table 3.3. The unproductive binding of protein to lignin is dependent on the source and its preparation (Kumar and Wyman, 2009a, 2009b; Ooshima et al., 1990; Sutcliffe and Saddler, 1986) and could likely be reduced by using additives (Boerjesson et al., 2007; Eriksson et al., 2002a; Sewalt et al., 1997a; Tu et al., 2007; Yang and Wyman, 2006). For example, as shown in Table 3.3, we found that lignin residues enzymatically extracted from corn stover and poplar solids, prepared by leading pretreatment options, had different cellulase adsorption capacities and affinities. Surprisingly, lignin prepared by dilute acid pretreatment, at least for poplar, had a low cellulase adsorption capacity but lignin prepared with AFEX pretreatment was found to have the least. It appears that chemicals/reagents used in pretreatment significantly affect lignin characteristics as there was no direct relationship found between pretreatment temperature/severity (log R0; includes time and temperature only) and adsorption parameters. However, Ooshima et al. applied dilute acid pretreatment of hardwood to show a decline in adsorption capacity with an increase in temperature due to shrinking and agglomeration of lignin (Ooshima et al., 1990). Similar observations of lignin melting and its

ß Woodhead Publishing Limited, 2010

Table 3.3 Langmuir parameters for enzyme/protein adsorption on lignin

ß Woodhead Publishing Limited, 2010

Substrate/source

Enzyme/ Protein/ Brand name

Larch lignin

EGI EGII EGI EGII Cellulase GC 123, Genencor

Beech lignin Lignin residue/180 ëC* Lignin residue/200 ëC Lignin residue/220 ëC EL1 EL2

Celluclast 1.5 L Beta-g/Novo 188 Cellulase, Sp. CP Beta-g/Novo188

Alkali lignin EL3

Bovine Serum Albumin CBHI/CBHI-CD/ EGII/EGII-CD

Max. Ads. Capacity , mg/g subs.

Affinity A, ml/mg protein

Ads. Strength R ˆ   A, ml/g sub.

± ± ± ± 100

0.09 0.11 0.03 0.03 0.41

± ± ± ± 40.8

66.6 12.3 86.1 173.5 590 (4 ëC) 790 (50 ëC) 170 (4 ëC) 130 (50 ëC) 180 (4 ëC) 280 (50 ëC) ±

0.66 0.81 0.51 0.75 0.06 0.18 0.45 0.86 0.64 0.91 1.7/0.0/ 0.6/0.2 0.6/0.0/

43.6 9.93 43.9 129.8 37.8 140 76.5 112 115 255 ±

±

±

Reference

Chernoglazov et al. (1988)

Ooshima et al. (1990)

Zheng et al. (2007) Willies (2007)

Palonen et al. (2004b)

0.2/0.0 2

ß Woodhead Publishing Limited, 2010

Corn stover-enzyme lignin AFEXa ARP C. pH D. acid Lime SO2 Poplar-enzyme lignin2 AFEX ARP D. acid FT Lime SO2 Alkali-lignin

Cellulase, Spezyme CP

38.7 41.6 63.6 53.0 64.9 67.5

2.99 10.70 0.60 0.68 2.69 6.39

116.0 445.0 36.2 174.5 37.8 431.5

Kumar and Wyman (2009b)

Cellulase, Spezyme CP

56.8 92.1 74.0 112.8 126.9 83.7 ±

2.14 0.59 0.29 0.67 0.11 0.25 11.8 (pH 4.0)/ 8.9 (pH 9.0)

121.8 54.8 21.2 75.9 14.3 21.0 ±

Kumar and Wyman (2009a)

Xylanase, Pulpzyme HC

Ryu and Kim (1998)

* Lignin was obtained from dilute acid pretreated hardwood prepared at three different temperatures. There is no information if the remaining protein was completely dislodged from lignin surface. 1. Lignin was obtained after complete enzymatic hydrolysis of carbohydrate part of dilute acid pretreated creeping wild rye grass. There is no information if the protein left on lignin was removed before adsorption studies. 2. Lignin was obtained after complete enzymatic hydrolysis of carbohydrate part (< 15% of carbohydrate left in substrate) of dilute acid pretreated corn stover.The protein remaining on lignin residue was dislodged by protease treatment. 3. Lignin was obtained after complete enzymatic hydrolysis of carbohydrate part of steam pretreated softwood. It was reported that 5.5% protein was left adsorbed on lignin after washing. a

Pretreatment type; AFEX ± ammonia fiber expansion; ARP ± ammonia recycled percolation; C.pH ± controlled pH; D. acid ± dilute acid.

96

Bioalcohol production

relocation are affirmed by others as well (Donaldson et al., 1988; Donohoe et al., 2008; Michalowicz et al., 1991; Selig et al., 2007). In a recent study, Selig et al. explained that droplets of lignin, formed during high temperature dilute acid or water only pretreatment, migrate to the cell wall, and may deposit on the cellulose surface to impede cellulase adsorption on cellulose (Selig et al., 2007). Lignin removal is expensive, and it is not clear whether lignin removal or disruption of its tight association with carbohydrates is more important. Grabber and coworkers suggested that inhibition of fungal hydrolases is not affected by lignin composition (Grabber et al., 1997); however, lignin concentration and its cross-linking with feruloylated xylans greatly affect degradability of cell wall (Grabber, 2005; Grabber et al., 1998b). Yet, a negative impact of lignin concentration on cell wall digestibility of tobacco stems was observed by Sewalt and coworkers (1997b) in another study.

3.5.2

Effectiveness

Although lignin's effect on hydrolysis is not entirely clear, lignin removal is technically and economically advantageous prior to cellulose saccharification because unproductive binding to lignin reduces enzyme availability, thereby limiting cellulase effectiveness (Berlin et al., 2005, 2006; Excoffier et al., 1991; Jùrgensen and Olsson, 2006; Kumar and Wyman, 2009a; Mandels and Reese, 1965; Selig et al., 2007; Sewalt et al., 1997a; Wu and Lee, 1997; Yang and Wyman, 2006), lignin breakdown products are likely to be inhibitory to fermentation and cellulase effectiveness (Hartley et al., 1976; Kaya et al., 1999; Lynd, 1996), and lignin increases viscosities (Berson et al., 2006; Fan et al., 2003) at the higher solid loadings needed commercially (Wingren et al., 2003), requiring more energy and negatively affecting cellulase effectiveness (Jùrgensen et al., 2007; Nutor and Converse, 1991; Pimenova and Hanley, 2003; VaÈljamaÈe et al., 2001). Furthermore, lignin and its derivatives were also reported to precipitate and bond with protein (Kawamoto et al., 1992; Makkar et al., 1987). In addition, during pretreatment, some soluble lignin depolymerization and degradation compounds may form, and these compounds, though their impact on cellulase adsorption is not known, may severely inhibit enzyme effectiveness (Excoffier et al., 1991; GarcõÂa-Aparicio et al., 2006; Kaya et al., 1999; Paul et al., 2003; Selig et al., 2007; Weil et al., 2002). Literature studies suggest that lignin droplets deposited on cellulose may interact with water, as one study shows that hydrophobic surfaces at a macroscopic level do not repel but attract water (van Oss, 1995), and form a boundary layer impeding cellulase movement (Donohoe et al., 2008; Matthews et al., 2006; Selig et al., 2007). Unproductive cellulase adsorption on lignin is hypothetically considered due to hydrophobic interactions (Bai et al., 2008; Kongruang et al., 2003; Tilton et al., 1991). In some studies, the extent of hydrolysis and the amount of free enzyme have been reported to increase with increased cellulase hydrophilicity (Kajiuchi

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

97

et al., 1993; Park et al., 2002), because proteins are highly hydrophobic due to clusters of closely located non-polar residues on their surface (Andreaus et al., 1999; Halder et al., 2005; Karlsson et al., 2005; Reinikainen et al., 1995; Suvajittanont et al., 2000) and tend to adsorb strongly on hydrophobic surfaces (Kongruang et al., 2003; van Oss, 1995). Furthermore, protein attachment to highly hydrophobic surfaces results in conformational changes and consequently irreversible adsorption and deactivation (Borjesson et al., 2007; Kajiuchi et al., 1993; Palonen, 2004; Park et al., 2002). In addition, lignin linkages with cellulose (Jin et al., 2006; Karlsson and Westermark, 1996; Kotelnikova et al., 1993) presumably impact the processive action of cellulase. Although lignin may reduce the active amount of enzyme available for cellulose hydrolysis, its relationship to effectiveness of adsorbed cellulase still needs further study.

3.6

Conclusions

Overall, it can be concluded that literature reports on enzymatic hydrolysis of cellulose can be viewed in terms of two key factors, cellulase accessibility to cellulose and cellulase effectiveness. For example, several studies have shown a strong correlation between rates/extent of hydrolysis and enzyme adsorption (Beltrame et al., 1982; Ding et al., 2000; Hogan et al., 1990; Karlsson et al., 1999; Klyosov, 1986; Kotiranta et al., 1999; Lee and Fan, 1979, 1982; Mansfield et al., 1999; Medve et al., 1998; Mooney et al., 1999; Nidetzky and Steiner, 1993; Sakata et al., 1985; Sethi et al., 1998; Watson et al., 2002; Yang et al., 2006), and we recently observed an almost linear relationship between the maximum protein adsorption capacity of cellulase on solids and the hydrolysis rate and yield in a study with corn stover and poplar solids prepared by promising pretreatment technologies (Kumar, 2008; Kumar and Wyman, 2009b; Mosier et al., 2005). Although cellulase accessibility to cellulose appears to be affected more by xylan removal than lignin removal, cellulase adsorption and its efficacy cannot be related to a solitary substrate feature or two for lignocellulosics. As summarized in Table 3.4, other substrate features may also have a significant impact on the two factors hypothesized to primarily control hydrolysis; however, the extent of their impact may either be lower than xylan/lignin removal or unclear due to their interdependence with lignin/xylan removal. For example, cellulose crystallinity appears to significantly impact accessibility, at least as suggested for cellulase adsorption data for pure cellulose and for pretreatments using reagents such as phosphoric acid that generate amorphous cellulose (Zhang et al., 2007). However, conventional methods used to determine biomass crystallinity may suggest otherwise. In addition, even for other thermochemical pretreatments, the reagents used in combination with heat not only disrupt lignin-carbohydrate linkages but change hydrogen bonds among cellulose chains

ß Woodhead Publishing Limited, 2010

98

Bioalcohol production

Table 3.4 A summary of how primary substrate features are hypothesized to impact cellulase accessibility to cellulose and cellulase effectiveness with impact ranking Substrate features

Cellulase accessibility to cellulose (Impact rankinga)

Cellulase effectiveness (Impact ranking)

Acetyl groups

Yes (04)

Cellulose crystallinity Cellulose DP

Small but noticeable effect (02) Yesb Inconclusive

Xylan content Lignin removal

A significant impact (10) Appears negligible (0.5)

Yesb Largely inconclusive but some impact (01) A major impact (06) A significant impact (08)

a

Ranking was based on 0 to 10, where 10 stands for the highest impact on the feature noted and zero for negligible impact. b Ranking was not given due to lack of convincing resolution in literature.

as well (Chundawat et al., 2007; He et al., 2008; Kumar et al., 2009). Consequently pretreated lignocellulosic solids, in most cases, have much higher cellulose accessibility (Kumar and Wyman, 2009a, 2009b), resulting in higher digestibility than pure cellulose such as Avicel (Kumar and Wyman, 2009e; Lloyd and Wyman, 2005). Thus, the role of crystallinity in cellulose accessibility remains unclear. For example, although the origins are different and cellulase effectiveness may differ, bacterial cellulose (BC; CrI ~ 60 to 70%) and bacterial microcrystalline cellulose (BMCC CrI ~> 85%) both have similar or higher crystallinity but much higher accessibility than microcrystalline cellulose Avicel (CrI ~ 50 to 60%) (Hong et al., 2007; Zhang and Lynd, 2004b). The literature also suggests that lignin does not directly limit glucan accessibility but greatly restricts xylan accessibility which in turn limits glucan accessibility, as shown by a simplified conceptual model in Fig. 3.5. According to this model, lignin is strongly linked to xylan but also has bonds to glucan, whereas xylan is more strongly linked to glucan than lignin and functions as a filler or spacer between lignin and glucan layers. Therefore, either xylan or lignin removal should enhance saccharification, but because xylan removal directly impacts glucan chain accessibility, removing xylan should be more advantageous than removing lignin. In addition to direct impact on enzyme accessibility to glucan, xylan removal has some additional advantages; for example, xylan removal should result in 1) reduced enzyme inhibition by xylooligomers and 2) reduced requirements for xylanases and other auxiliary enzymes for xylan debranching. However, lignin removal exposes more xylan, resulting in the need for additional xylan degrading and auxiliary enzymes to expose glucan to cellulose and makes more enzymes available for hydrolysis due to reduced unproductive binding. In addition, removing lignin during pretreatment could have a big impact on process

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

99

3.5 A simplified conceptual model of biomass structure.

economics by lowering mixing requirements in fermentation and making lignin available for other uses, provided lignin removal costs are low. Ammonia fiber expansion (AFEX) pretreatment is unique in that although AFEX removes little lignin or xylan, it still gives good digestibility, at least for non woody biomass. This anomaly could be attributed to disruption of lignincarbohydrate linkages (LCC) (Chundawat et al., 2007; Kumar et al., 2009; Laureano-Perez et al., 2005; Venkatesh et al., 2009) and lignin alteration resulting in reduced affinity for enzymes (Kumar and Wyman, 2009a, 2009b). Thus, based on cellulose accessibility (Kumar and Wyman, 2009a, 2009b) and hydrolysis data with AFEX (Sendich et al., 2008; Venkatesh et al., 2009), it could be concluded that LCC disruption is the most important requirement for an effective pretreatment, as shown in Fig. 3.6, with spacer (xylan)/lignin removal merely a way to accomplish this goal. Overall, altering the substrate through reducing substrate hemicelluloses, lignin, and acetyl contents; crystallinity; and degree of polymerization can particularly affect accessibility of enzymes to cellulose. However, although changes in the substrate can be necessary to realize good enzyme effectiveness, they may not be sufficient because of the importance of the nature of the numerous cellulase components and chemical and physical environmental factors to performance. For example, once cellulase protein adsorbs on the surface, its catalytic efficacy may further be dictated by physical parameters such as pH, temperature, ionic strength, and the presence of inhibitors (Andreaus et al., 1999; Kumar and Wyman, 2008; Panagiotou and Olsson, 2007; Reinikainen et al., 1995; Tengborg et al., 2001) as well as factors related to the substrate and enzyme.

ß Woodhead Publishing Limited, 2010

100

Bioalcohol production

3.6 A schematic decision tree of pretreatment effectiveness.

Although the focus of this review is on how modifications in biomass affect enzymatic hydrolysis, cellulase components molar ratios and their concentrations may affect their adsorption and effectiveness due to synergistic action (Beukes et al., 2008; Gupta et al., 2008; Murnen et al., 2007; Selig et al., 2008), and supplementation of cellulase with other enzymes such as -glucosidase/ xylosidase, xylanase, and debranching enzymes may also enhance cellulase adsorption/effectiveness, depending on substrate and pretreatment type (Girard and Converse, 1993; Huang and Penner, 1991; VaÈljamaÈe et al., 2001). The physical and chemical environment, substrate loadings (Kumar and Wyman, 2008; Stutzenberger and Lintz, 1986; Xiao et al., 2004), sugars (Kristensen et al., 2009; Kumar and Wyman, 2008; Todorovic et al., 1987; Wendorf et al., 2004), their oligomers (GarcõÂa-Aparicio et al., 2006), sugar degradation products (Kaya et al., 1999; Sineiro et al., 1997), chemical compounds (Eriksson et al., 2002a; Park et al., 1992), additives (Kim et al., 1988; Moloney and Coughlan, 1983), temperature (Golovchenko et al., 1992; Reinikainen et al., 1995), pH (Gerber et al., 1997; Kim and Hong, 2000), ionic strength (Azevedo et al., 2000; Sakata et al., 1985), and agitation (Azevedo et al., 2000; O'Neill et al., 2007; Sakata et al., 1985) have all been hypothesized to play roles in influencing

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

101

enzyme accessibility and effectiveness. On this basis, a concerted effort is needed to better understand fundamental physical and chemical features of lignocellulosic biomass that limit its deconstruction and the organization and interaction among biomass components that constitute a barrier to access by enzymes to breakdown carbohydrates into fermentable sugars. Such an understanding of factors that control the interactions of substrates and enzymes would be invaluable in identifying pathways to lower cost advanced coupled pretreatment and enzymatic hydrolysis systems. Because new accurate data are critical to meaningfully assess promising advances in plant, microbial, and enzymatic systems, improved analytical methods must also be developed to fully characterize biomass composition and its structure and characterize interactions among biomass and various chemical treatments, as well as with deconstruction and hydrolysis enzymes.

3.7

Acknowledgements

Support by the US Department of Energy Office of the Biomass Program (contract DE-FG36-04GO14017) and the National Institute of Standards and Technology (award 60NANB1D0064) made this research possible. We are also grateful to the Center for Environmental Research and Technology of the Bourns College of Engineering at the University of California, Riverside and the Thayer School of Engineering at Dartmouth College for providing key equipment and facilities. The corresponding author is grateful to the Ford Motor Company for funding the Chair in Environmental Engineering at the Center for Environmental Research and Technology at the Bourns College of Engineering at UCR that augments support for many projects such as this.

3.8

References

Allen SG, Schulman D, Lichwa J, Antal MJ, Jennings E, Elander R. 2001. A comparison of aqueous and dilute-acid single-temperature pretreatment of yellow poplar sawdust. Ind Eng Chem Res 40(10): 2352±2361. Anand L, Vithayathil PJ. 1996. Xylan-degrading enzymes from the thermophilic fungus Humicola lanuginosa (Griffon and Maublanc) Bunce: action pattern of xylanase and [beta]-glucosidase on xylans, xylooligomers and arabinoxylooligomers. J Ferment Bioeng 81(6): 511±517. Andreaus J, Azevedo H, Cavaco-Paulo A. 1999. Effects of temperature on the cellulose binding ability of cellulase enzymes. J Mol Catalysis B: Enzymatic 7: 233±239. Azevedo H, Bishop D, Cavaco-Paulo A. 2000. Effects of agitation level on the adsorption, desorption, and activities on cotton fabrics of full length and core domains of EGV (Humicola insolens) and CenA (Cellulomonas fimi). Enzyme Microb Technol 27(3±5): 325±329. Bai G, Goncalves C, Gama FM, Bastos M. 2008. Self-aggregation of hydrophobically modified dextrin and their interaction with surfactant. Thermochimica Acta 467(1± 2): 54±62.

ß Woodhead Publishing Limited, 2010

102

Bioalcohol production

Banka RR, Mishra S. 2002. Adsorption properties of the fibril forming protein from Trichoderma reesei. Enzyme Microb Technol 31(6): 784±793. Beldman G, Searle-Van Leeuwen MF, Rombouts FM, Voragen FG. 1985. The cellulase of Trichoderma viride. Purification, characterization and comparison of all detectable endoglucanases, exoglucanases and beta-glucosidases. Eur J Biochem 146(2): 301±308. Beltrame PL, Carniti P, Focher B, Marzetti A, Cattaneo M. 1982. Cotton cellulose: enzyme adsorption and enzymatic hydrolysis. J Appl Polym Sci 27: 3493±3502. Berlin A, Gilkes N, Kurabi A, Bura R, Tu M, Kilburn D, Saddler J. 2005. Weak ligninbinding enzymes: a novel approach to improve activity of cellulases for hydrolysis of lignocellulosics. Appl Biochem Biotechnol 121±124: 163±170. Berlin A, Balakshin M, Gilkes N, Kadla J, Maximenko V, Kubo S, Saddler J. 2006. Inhibition of cellulase, xylanase and beta-glucosidase activities by softwood lignin preparations. J Biotechnol 125(2): 182±209. Berlin A, Maximenko V, Gilkes N, Saddler J. 2007. Optimization of enzyme complexes for lignocellulose hydrolysis. Biotechnol Bioeng 97(2): 287±296. Bernardez TD, Lyford K, Hogsett DA, Lynd LR. 1993. Adsorption of Clostridium thermocellum cellulases onto pretreated mixed hardwood, Avicel, and lignin. Biotechnol Bioeng 42(7): 899±907. Berson R, Dasari R, Hanley T. 2006. Modeling of a continuous pretreatment reactor using computational fluid dynamics. Twenty-Seventh Symposium on Biotechnology for Fuels and Chemicals, pp. 621±630. Bertran MS, Dale BE. 1985. Enzymatic hydrolysis and recrystallization behavior of initially amorphous cellulose. Biotechnol Bioeng 27(2): 177±181. Beukes N, Chan H, Doi RH, Pletschke BI. 2008. Synergistic associations between Clostridium cellulovorans enzymes XynA, ManA and EngE against sugarcane bagasse. Enzyme Microb Technol 42(6): 492±498. Beveridge RJ, Richards GN. 1975. Digestion of polysaccharide constituents of tropical pasture herbage in the bovine rumen. VI. Investigation of the digestion of cell-wall polysaccharides of spear grass and of cotton cellulose by viscometry and by x-ray diffraction. Carbohyd Res 43(1): 163±172. Boerjesson J, Peterson R, Tjerneld F. 2007. Enhanced enzymatic conversion of softwood lignocellulose by poly(ethylene glycol) addition. Enzyme Microb Technol 40(4): 754±762. Boisset C, Chanzy H, Henrissat B, Lamed R, Shoham Y, Bayer EA. 1999. Digestion of crystalline cellulose substrates by the clostridium thermocellum cellulosome: structural and morphological aspects. Biochem J 340 (Pt 3): 829±835. Borjesson J, Engqvist M, Sipos B, Tjerneld F. 2007. Effect of poly(ethylene glycol) on enzymatic hydrolysis and adsorption of cellulase enzymes to pretreated lignocellulose. Enzyme Microb Technol 41(1±2): 186±195. Cao Y, Tan H. 2002. Effects of cellulase on the modification of cellulose. Carbohydr Res 337(14): 1291±1296. Carrard G, Linder M. 1999. Widely different off rates of two closely related cellulosebinding domains from Trichoderma reesei. Eur J Biochem 262(3): 637±643. Caulfield DF, Moore WE. 1974. Effect of varying crystallinity of cellulose on enzymic hydrolysis. Wood Science 6(4): 375±379. Chandra RP, Bura R, Mabee WE, Berlin A, Pan X, Saddler JN. 2007. Substrate pretreatment: the key to effective enzymatic hydrolysis of lignocellulosics? Adv Biochem Eng Biotechnol 108: 67±93. Chandra RP, Esteghlalian AR, Saddler JN. 2008. Assessing substrate accessibility to

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

103

enzymatic hydrolysis by cellulases. Hu TQ, editor: Blackwell Publishing Ltd., Oxford, pp. 60±80. Chang VS, Burr B, Holtzapple MT. 1997. Lime pretreatment of switchgrass. Appl Biochem Biotechnol 63±65: 3±19. Chang VS, Holtzapple MT. 2000. Fundamental factors affecting biomass enzymatic reactivity. Appl Biochem Biotechnol 84±86: 5±37. Chapple C, Ladisch M, Meilan R. 2007. Loosening lignin's grip on biofuel production. Nat Biotech 25(7): 746±748. Chen Y, Stipanovic A, Winter W, Wilson D, Kim Y-J. 2007. Effect of digestion by pure cellulases on crystallinity and average chain length for bacterial and microcrystalline celluloses. Cellulose 14(4): 283±293. Chernoglazov VM, Ermolova OV, Klesov AA. 1988. Adsorption of high-purity endo-1,4b-glucanases from Trichoderma reesei on components of lignocellulosic materials: cellulose, lignin, and xylan. Enzyme Microb Technol 10(8): 503±507. Christina Divne JSh, Tuula T. Teeri and T. Alwyn Jones. 1998. High-resolution crystal structures reveal how a cellulose chain is bound in the 50 A long tunnel of cellobiohydrolase I from Trichoderma reesei. J Mol Biol 275: 309±325. Chum HL, Johnson DK, Black SK, Baker J, Grohmann K, Sarkanen KV, Wallace K, Schroeder HA. 1988. Organosolv pretreatment for enzymatic hydrolysis of poplar. I. Enzyme hydrolysis of cellulosic residues. Biotechnol Bioeng 31: 643±649. Chundawat SP, Venkatesh B, Dale BE. 2007. Effect of particle size based separation of milled corn stover on AFEX pretreatment and enzymatic digestibility. Biotechnol Bioeng 96(2): 219±231. Clark TA, Mackie KL, Dare PH, McDonald AG. 1989. Steam explosion of the softwood Pinus radiata with sulfur dioxide addition. II. Process characterization. J Wood Chem Technol 9(2): 135±166. Converse AO. 1993. Substrate factors limiting enzymatic hydrolysis. In: Saddler JN, editor. Bioconversion of Forest and Agricultural Plant Residues. Wallingford: CAB International, pp. 93±106. Cunningham RL, Detroy RW, Bagby MO, Baker FL. 1981. Modifications of wheat straw to enhance cellulose saccharification by enzymic hydrolysis. Trans Illinois State Acad Sci 74(3±4): 67±75. Dale BE, Leong CK, Pham TK, Esquivel VM, Rios I, Latimer VM. 1996. Hydrolysis of lignocellulosics at low enzyme levels: application of the AFEX process. Biores Technol 56: 111±116. Dhugga KS. 2007. Maize biomass yield and composition for biofuels. Crop Sci 47(6): 2211±2227. Ding H, Xu F. 2004. Productive cellulase adsorption on cellulose. ACS Symposium Series 889 (Lignocellulose Biodegradation): 154±169. Ding H, Vlasenko E, Shoemaker C, Shoemaker S. 2000. Absorption of endoglucanase I and cellobiohydrolase I of Trichoderma reesei during hydrolysis of microcrystalline cellulose. ACS Symposium Series 769 (Glycosyl Hydrolases for Biomass Conversion): 131±143. Ding S-Y, Himmel ME. 2006. The maize primary cell wall microfibril: a new model derived from direct visualization. J Agric Food Chem 54(3): 597±606. Donaldson LA, Wong KKY, Mackie KL. 1988. Ultrastructure of steam-exploded wood. Wood Sci Technol 22(2): 103±114. Donohoe BS, Decker SR, Tucker MP, Himmel ME, Vinzant TB. 2008. Visualizing lignin coalescence and migration through maize cell walls following thermochemical pretreatment. Biotechnol Bioeng 101(5): 913±925.

ß Woodhead Publishing Limited, 2010

104

Bioalcohol production

Draude KM, Kurniawan CB, Duff SJB. 2001. Effect of oxygen delignification on the rate and extent of enzymatic hydrolysis of lignocellulosic material. Biores Technol 79(2): 113±120. Engstrom AC, Ek M, Henriksson G. 2006. Improved accessibility and reactivity of dissolving pulp for the viscose process: pretreatment with monocomponent endoglucanase. Biomacromolecules 7(6): 2027±2031. Eremeeva T, Bikova T, Eisimonte M, Viesturs U, Treimanis A. 2001. Fractionation and molecular characteristics of cellulose during enzymatic hydrolysis. Cellulose (Dordrecht, Netherlands) 8(1): 69±79. Eriksson T, Borjesson J, Tjerneld F. 2002a. Mechanism of surfactant effect in enzymatic hydrolysis of lignocellulose. Enzyme Microb Technol 31(3): 353±364. Eriksson T, Karlsson J, Tjerneld F. 2002b. A model explaining declining rate in hydrolysis of lignocellulose substrates with cellobiohydrolase I (cel7A) and endoglucanase I (cel7B) of Trichoderma reesei. Appl Biochem Biotechnol 101(1): 41±60. Esteghlalian AR, Bilodeau M, Mansfield SD, Saddler JN. 2001. Do enzymic hydrolyzability and Simons' stain reflect the changes in the accessibility of lignocellulosic substrates to cellulase enzymes? Biotechnol Prog 17(6): 1049±1054. Estrada P, Acebal C, Castillon MP, Mata I, Romero D. 1988. Adsorption of cellulase from Trichoderma reesei on wheat straw. Biotechnol Appl Biochem 10(1): 49±58. Excoffier G, Toussaint B, Vignon MR. 1991. Saccharification of steam-exploded poplar wood. Biotechnol Bioeng 38(11): 1308±1317. Fan LT, Lee Y-H, Beardmore DH. 1980. Major chemical and physical features of cellulosic materials as substrates for enzymic hydrolysis. Adv Biochem Eng 14: 101±117. Fan LT, Gharpuray MM, Lee Y. 1981. Evaluation of pretreatments for enzymatic conversion of agricultural residues. Biotechnol Bioeng Symp 11: 29±45. Fan LT, Lee Y-H, Gharpuray M. 1982. The nature of lignocellulosics and their pretreatments for enzymatic hydrolysis. Adv Biochem Eng 23: 157±187. Fan Z, South C, Lyford K, Munsie J, van Walsum P, Lynd L. 2003. Conversion of paper sludge to ethanol in a semicontinuous solids-fed reactor. Bioprocess Biosystems Engineering 26(2): 93±101. Farrell AE, Plevin RJ, Turner BT, Jones AD, O'Hare M, Kammen DM. 2006. Ethanol can contribute to energy and environmental goals. Science 311(5760): 506±508. Fernandes AC, Fontes CMGA, Gilbert HJ, Hazlewood GP, Fernandes TH, Ferreira LMA. 1999. Homologous xylanases from Clostridium thermocellum: evidence for bifunctional activity, synergism between xylanase catalytic modules and the presence of xylan-binding domains in enzyme complexes. Biochem J 342(1): 105±110. Ford CW. 1983. Effect of particle size and delignification on the rate of digestion of hemicellulose and cellulose by cellulase in mature pangola grass stems. Aust J Agric Res 34(3): 241±248. Galbe M, Eklund R, Zacchi G. 1990. Adsorption of cellulases on steam-pretreated willow. Appl Biochem Biotechnol 24±25: 87±101. Gama FM, Mota M. 1997. Enzymic hydrolysis of cellulose. (II): X-ray photoelectron spectroscopy studies on cellulase adsorption. Effect of the surfactant Tween 85. Biocatalysis and Biotransformation 15(3): 237±250. GarcõÂa-Aparicio M, Ballesteros M, Manzanares P, Ballesteros I, GonzaÂlez A, Jose Negro M. 2007. Xylanase contribution to the efficiency of cellulose enzymatic hydrolysis of barley straw. Appl Biochem Biotechnol 137±140(1): 353±365. GarcõÂa-Aparicio MP, Ballesteros I, Gonzalez A, Oliva JM, Ballesteros M, Negro MJ.

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

105

2006. Effect of inhibitors released during steam-explosion pretreatment of barley straw on enzymatic hydrolysis. Appl Biochem Biotechnol 129±132: 278±288. Gerber P, Joyce T, Heitmann J, Siika-Aho M, Buchert J. 1997. Adsorption of a Trichoderma reesei endoglucanase and cellobiohydrolase onto bleached Kraft fibres. Cellulose 4(4): 255±268. Gharpuray MM, Lee YH, Fan LT. 1981. Pretreatment of wheat straw for cellulose hydrolysis. Proceedings of the 11th Annual Biochemical Engineering Symposium: 1±10. Ghose TK, Bisaria VS. 1979. Studies on the mechanism of enzymatic hydrolysis of cellulosic substances. Biotechnol Bioeng 21(1): 131±146. Girard DJ, Converse AO. 1993. Recovery of cellulase from lignaceous hydrolysis residue. Appl Biochem Biotechnol 39±40: 521±533. Glasser WG, Ravindran G, Jain RK, Samaranayake G, Todd J. 1995. Comparative enzyme biodegradability of xylan, cellulose, and starch derivatives. Biotechnol Prog 11: 552±557. Goel SC, Ramachandran KB. 1983. Studies on the adsorption of cellulase on lignocellulosics. J Ferment Technol 61(3): 281±286. Golovchenko NP, Kataeva IA, Akimenko VK. 1992. Analysis of pH-dependent adsorption of Clostridium thermocellum endoglucanase on cellulose. Biokhimiya (Moscow) 57(7): 1031±1038. Gomez LD, Steele-King CG, McQueen-Mason SJ. 2008. Sustainable liquid biofuels from biomass: the writing's on the walls. New Phytologist 178(3): 473±485. Grabber JH. 2005. How do lignin composition, structure, and cross-linking affect degradability? A review of cell wall model studies. Crop Sci 45(3): 820±831. Grabber JH, Ralph J, Hatfield RD, Quideau S. 1997. p-Hydroxyphenyl, guaiacyl, and syringyl lignins have similar inhibitory effects on wall degradability. J Agric Food Chem 45(7): 2530±2532. Grabber JH, Hatfield RD, Ralph J. 1998a. Diferulate cross-links impede the enzymatic degradation of non-lignified maize walls. J Sci Food Agric 77(2): 193±200. Grabber JH, Ralph J, Hatfield RD. 1998b. Ferulate cross-links limit the enzymatic degradation of synthetically lignified primary walls of maize. J Agric Food Chem 46(7): 2609±2614. Gray MC, Converse AO, Wyman CE. 2003. Sugar monomer and oligomer solubility. Data and predictions for application to biomass hydrolysis. Appl Biochem Biotechnol 105±108: 179±193. Gray MC, Converse AO, Wyman CE. 2007. Solubilities of oligomer mixtures produced by the hydrolysis of xylans and corn stover in water at 180 C. Ind Eng Chem Res 46(8): 2383±2391. Grethlein HE. 1984. Pretreatment for enhanced hydrolysis of cellulosic biomass. Biotechnol Adv 2(1): 43±62. Grethlein HE. 1985. Effect of pore size distribution on the rate of enzymatic hydrolysis of cellulosic substrates. Bio/Technology 3(2): 155±160. Grohmann K, Torget R, Himmel M. 1986. Optimization of dilute acid pretreatment of biomass. Biotechnol Bioeng Symp 15: 59±80. Grohmann K, Mitchell DJ, Himmel ME, Dale BE, Schroeder HA. 1989. The role of ester groups in resistance of plant cell wall polysaccharides to enzymic hydrolysis. Appl Biochem Biotechnol 20±21: 45±61. Gruno M, VaÈljamaÈe P, Pettersson G, Johansson G. 2004. Inhibition of the Trichoderma reesei cellulases by cellobiose is strongly dependent on the nature of the substrate. Biotechnol Bioeng 86(5): 503±511.

ß Woodhead Publishing Limited, 2010

106

Bioalcohol production

Gupta R, Lee YY. 2009. Mechanism of cellulase reaction on pure cellulosic substrates. Biotechnol Bioeng 102(6): 1570±1581. Gupta R, Kim T, Lee Y. 2008. Substrate dependency and effect of xylanase supplementation on enzymatic hydrolysis of ammonia-treated biomass. Appl Biochem Biotechnol 148(1): 59±70. HaÊkansson H, Ahlgren P. 2005. Acid hydrolysis of some industrial pulps: effect of hydrolysis conditions and raw material. Cellulose 12(2): 177±183. Halder E, Chattoraj DK, Das KP. 2005. Adsorption of biopolymers at hydrophilic cellulose-water interface. Biopolymers 77(5): 286±295. Hartley RD, Jones EC, Wood TM. 1976. Carbohydrates and carbohydrate esters of ferulic acid released from cell walls of Lolium multiflorum by treatment with cellulolytic enzymes. Phytochemistry 15(2): 305±307. He Y, Pang Y, Liu Y, Li X, Wang K. 2008. Physicochemical characterization of rice straw pretreated with sodium hydroxide in the solid state for enhancing biogas production. Energy Fuels 22(4): 2775±2781. Heitz M, Carrasco F, Rubio M, Brown A, Chornet E, Overend RP. 1987. Physicochemical characterization of lignocellulosic substrates pretreated via autohydrolysis: an application to tropical woods. Biomass 13(4): 255±273. Henrissat B. 1994. Cellulases and their interaction with cellulose. Cellulose 1: 169±196. Henrissat B. 1998. Enzymic cellulose degradation. Cellulose Com 5(2): 84±90. Henrissat B, Driguez H, Viet C, Schuelein M. 1985. Synergism of cellulases from Trichoderma reesei in the degradation of cellulose. Bio/Technology 3(8): 722±726. Hespell RB, O'Bryan PJ, Moniruzzaman M, Bothast RJ. 1997. Hydrolysis by commercial enzyme mixtures of AFEX-treated corn fiber and isolated xylans. Appl Biochem Biotechnol 62(1): 87±97. Hilden L, VaÈljamaÈe P, Johansson G. 2005. Surface character of pulp fibres studied using endoglucanases. J Biotechnol 118(4): 386±397. Himmel ME, Ding SY, Johnson DK, Adney WS, Nimlos MR, Brady JW, Foust TD. 2007. Biomass recalcitrance: engineering plants and enzymes for biofuels production. Science 315(5813): 804±807. Himmel ME, Ruth MF, Wyman CE. 1999. Cellulase for commodity products from cellulosic biomass. Curr Opin Biotechnol 10(4): 358±364. Hogan CM, Mes-Hartree M, Saddler JN, Kushner DJ. 1990. Assessment of methods to determine minimal cellulase concentrations for efficient hydrolysis of cellulose. Appl Microbiol Biotechnol 32(5): 614±620. Hong J, Ye X, Zhang YH. 2007. Quantitative determination of cellulose accessibility to cellulase based on adsorption of a nonhydrolytic fusion protein containing CBM and GFP with its applications. Langmuir 23(25): 12535±12540. Horn SJ, Sikorski P, Cederkvist JB, Vaaje-Kolstad G, Sorlie M, Synstad B, Vriend G, Varum KM, Eijsink VG. 2006. Costs and benefits of processivity in enzymatic degradation of recalcitrant polysaccharides. Proc Natl Acad Sci USA 103(48): 18089±18094. Hoshino E, Kanda T. 1997. Scope and mechanism of cellulase action on different cellulosic substrates. Oyo Toshitsu Kagaku 44(1): 87±104. Hoshino E, Kanda T, Sasaki Y, Nisizawa K. 1992. Adsorption mode of exo- and endocellulases from Irpex lacteus (Polyporus tulipiferae) on cellulose with different crystallinities. J Biochem (Tokyo) 111(5): 600±605. Hoshino E, Shiroishi M, Amano Y, Nomura M, Kanda T. 1997. Synergistic actions of exo-type cellulases in the hydrolysis of cellulose with different crystallinities. J Ferment Bioeng 84(4): 300±306.

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

107

Huang X, Penner MH. 1991. Apparent substrate inhibition of the Trichoderma reesei cellulase system. J Agric Food Chem 39(11): 2096±2100. Igarashi K, Wada M, Hori R, Samejima M. 2006a. Surface density of cellobiohydrolase on crystalline celluloses. A critical parameter to evaluate enzymatic kinetics at a solid-liquid interface. Febs J 273(13): 2869±2878. Igarashi K, Wada M, Samejima M. 2006b. Enzymatic kinetics at a solid-liquid interface: hydrolysis of crystalline celluloses by cellobiohydrolase. Cellulose Com 13(4): 173±177. Igarashi K, Wada M, Samejima M. 2007. Activation of crystalline cellulose to cellulose III(I) results in efficient hydrolysis by cellobiohydrolase. Febs J 274(7): 1785±1792. Ishihara M, Uemura S, Hayashi N, Jellison J, Shimizu K. 1991. Adsorption and desorption of cellulase components during enzymatic hydrolysis of steamed shirakamba (Betula platyphylla Skatchev) wood. J Ferment Bioeng 72(2): 96±100. Ishizawa CI, Davis MF, Schell DF, Johnson DK. 2007. Porosity and its effect on the digestibility of dilute sulfuric acid pretreated corn stover. J Agric Food Chem 55(7): 2575±2581. Iyer PV, Lee YY. 1999. Product inhibition in simultaneous saccharification and fermentation of cellulose into lactic acid. Biotechnol Lett 21(5): 371±373. Jeoh T, Johnson DK, Adney WS, Himmel ME. 2005. Measuring cellulase accessibility of dilute-acid pretreated corn stover. Preprints of Symposia ± American Chemical Society, Division of Fuel Chemistry 50(2): 673±674. Jeoh T, Ishizawa CI, Davis MF, Himmel ME, Adney WS, Johnson DK. 2007. Cellulase digestibility of pretreated biomass is limited by cellulose accessibility. Biotechnol Bioeng 98(1): 112±122. Jin Z, Katsumata KS, Bach T, Lam T, Iiyama K. 2006. Covalent linkages between cellulose and lignin in cell walls of coniferous and nonconiferous woods. Biopolymers 83(2): 103±110. Jùrgensen H, Olsson L. 2006. Production of cellulases by Penicillium brasilianum IBT 20888 ± effect of substrate on hydrolytic performance. Enzyme Microb Technol 38(3±4): 381±390. Jùrgensen H, Vibe-Pedersen J, Larsen J, Felby C. 2007. Liquefaction of lignocellulose at high-solids concentrations. Biotechnol Bioeng 96(5): 862±870. Kabel MA, Bos G, Zeevalking J, Voragen AG, Schols HA. 2007. Effect of pretreatment severity on xylan solubility and enzymatic breakdown of the remaining cellulose from wheat straw. Biores Technol 98(10): 2034±2042. Kabeya H, Kubo T, Hosokawa J, Akamatsu I. 1993. Studies on the pretreatments for effective enzymic hydrolysis of cellulosic materials. IV. Chemical and physical pretreatments for enhancement of the enzymic hydrolysis of thermomechanical pulp. Shikoku Kogyo Gijutsu Shikensho Kenkyu Hokoku 24: 42±90. Kadam KL, Rydholm EC, McMillan JD. 2004. Development and validation of a kinetic model for enzymatic saccharification of lignocellulosic biomass. Biotechnol Prog 20(3): 698±705. Kajiuchi T, Park JW, Moon HY. 1993. Adsorption control of cellulase onto cellulose by modification with amphiphilic copolymer. J Chem Eng Japan 26(1): 28±33. Kanda T, Wakabayashi K, Nisizawa K. 1976. Synergistic action of two different types of endo-cellulase components from Irpex lacteus (Polyporus tulipiferae) in the hydrolysis of some insoluble celluloses. J Biochem (Tokyo) 79(5): 997±1005. Kanda T, Wakabayashi K, Nisizawa K. 1980. Modes of action of exo- and endocellulases in the degradation of celluloses I and II. J Biochem (Tokyo) 87(6): 1635± 1639.

ß Woodhead Publishing Limited, 2010

108

Bioalcohol production

Kaplan AM, Mandels M, Pillion E, Greenberger M. 1970. Resistance of weathered cotton cellulose to cellulase action. Appl Microbiol 20(1): 85±93. Karlsson J, Medve J, Tjerneld F. 1999. Hydrolysis of steam-pretreated lignocellulose: synergism and adsorption for cellobiohydrolase I and endoglucanase II of Trichoderma reesei. Appl Biochem Biotechnol 82: 243±258. Karlsson J, Momcilovic D, Wittgren B, Schulein M, Tjerneld F, Brinkmalm G. 2002. Enzymatic degradation of carboxymethyl cellulose hydrolyzed by the endoglucanases Cel5A, Cel7B, and Cel45A from Humicola insolens and Cel7B, Cel12A and Cel45Acore from Trichoderma reesei. Biopolymers 63(1): 32±40. Karlsson M, Ekeroth J, Elwing H, Carlsson U. 2005. Reduction of irreversible protein adsorption on solid surfaces by protein engineering for increased stability. J Biol Chem 280(27): 25558±25564. Karlsson O, Westermark U. 1996. Evidence for chemical bonds between lignin and cellulose in kraft pulps. J Pulp Paper Sci 22(10): J397±J401. Kawamoto H, Nakatsubo F, Murakami K. 1992. Protein-adsorbing capacities of lignin samples. Mokuzai Gakkaishi 38(1): 81±84. Kaya F, Heitmann JA, Jr., Joyce TW. 1999. Effect of dissolved lignin and related compounds on the enzymic hydrolysis of cellulose model compound. Cell Chem Technol 33(3±4): 203±213. Kim DW, Hong YG. 2000. Ionic strength effect on adsorption of cellobiohydrolases I and II on microcrystalline cellulose. Biotechnol Lett 22(16): 1337±1342. Kim DW, Yang JH, Jeong YK. 1988. Adsorption of cellulase from Trichoderma viride on microcrystalline cellulose. Appl Microbiol Biotechnol 28(2): 148±154. Kim S, Holtzapple MT. 2005. Lime pretreatment and enzymatic hydrolysis of corn stover. Biores Technol 96(18): 1994±2006. Kim S, Holtzapple MT. 2006. Effect of structural features on enzyme digestibility of corn stover. Biores Technol 97(4): 583±591. Kim SB, Um BH, Park SC. 2001. Effect of pretreatment reagent and hydrogen peroxide on enzymatic hydrolysis of oak in percolation process. Appl Biochem Biotechnol 91±93: 81±94. Kim TH, Lee YY, Sunwoo C, Kim JS. 2006. Pretreatment of corn stover by low-liquid ammonia recycle percolation process. Appl Biochem Biotechnol 133(1): 41±57. Kipper K, VaÈljamaÈe P, Johansson G. 2005. Processive action of cellobiohydrolase Cel7A from Trichoderma reesei is revealed as 'burst' kinetics on fluorescent polymeric model substrates. Biochem J 385(Pt 2): 527±535. Kleman-Leyer KM, Siika-Aho M, Teeri TT, Kirk TK. 1996. The cellulases endoglucanase I and cellobiohydrolase II of Trichoderma reesei act synergistically to solubilize native cotton cellulose but not to decrease its molecular size. Appl Environ Microbiol 62(8): 2883±2887. Klyosov AA. 1982. The regularities for formation and conversion of intermediate cellooligosaccharides and cellobiose in the course of the enzymatic hydrolysis of the insoluble cellulose. Biokhimiya 47: 608±618. Klyosov AA. 1986. Role of the activity and adsorption of cellulases in the efficiency of the enzymatic hydrolysis of amorphous and crystalline cellulose. Biochemistry 25: 540±542. Knappert D, Grethlein H, Converse A. 1980. Partial acid hydrolysis of cellulosic materials as a pretreatment for enzymatic hydrolysis. Biotechnol Bioeng 22: 1449± 1463. Knauf M, Moniruzzaman M. 2004. Lignocellulosic biomass processing: a perspective. Int Sugar J 106(1263): 147±150.

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

109

Kong R, Engler CR, Soltes EJ. 1992. Effects of cell-wall acetate xylan backbone and lignin on enzymatic hydrolysis of aspen wood. Appl Biochem Biotechnol 34: 23±35. Kongruang S, Penner MH. 2004. Borohydride reactivity of cellulose reducing ends. Carbohyd Polym 58(2): 131±138. Kongruang S, Bothwell MK, McGuire J, Zhou M, Haugland RP. 2003. Assaying the activities of Thermomonospora fusca E5 and Trichoderma reesei CBHI cellulase bound to polystyrene. Enzyme Microb Technol 32(5): 539±545. Kongruang S, Han M, Breton C, Penner M. 2004. Quantitative analysis of cellulosereducing ends. Appl Biochem Biotechnol 113(1): 213±231. Kormelink FJM, Voragen AGJ. 1992. Combined action of xylan-degrading and accessory enzymes on different glucurono-arabino xylans. Prog Biotechnol 7 (Xylans Xylanases): 415±418. Kotelnikova NE, Shashilov AA, Yongfa H. 1993. Effect of the presence of lignin on the structure and reactivity to hydrolysis of lignincarbohydrate complexes of poplar wood obtained by sulfate pulping. Wood Sci Technol 27(4): 263±269. Kotiranta P, Karlsson J, Siika-Aho M, Medve J, Viikari L, Tjerneld F, Tenkanen M. 1999. Adsorption and activity of Trichoderma reesei cellobiohydrolase I, endoglucanase II, and the corresponding core proteins on steam pretreated willow. Appl Biochem Biotechnol 81(2): 81±90. Koullas DP, Christakopoulos PF, Kekos D, Koukios EG, Macris BJ. 1993. Effect of alkali delignification on wheat straw saccharification by fusarium oxysporum cellulases. Biomass Bioenergy 4(1): 9±13. Kristensen J, Felby C, Jùrgensen H. 2009. Yield-determining factors in high-solids enzymatic hydrolysis of lignocellulose. Biotechnol Biofuels 2(1): 11. Kumar R. 2008. PhD Thesis: Enzymatic Hydrolysis of Cellulosic Biomass Solids Prepared by Leading Pretreatments and Identification of Key Features Governing Performance. Hanover, NH, USA: Thayer School of Engineering, Dartmouth College. Kumar R, Wyman CE. 2008. An improved method to directly estimate cellulase adsorption on biomass solids. Enzyme Microb Technol 42(5): 426±433. Kumar R, Wyman CE. 2009a. Access of cellulase to cellulose and lignin for poplar solids produced by leading pretreatment technologies. Biotechnol Prog 25(3): 807±819. Kumar R, Wyman CE. 2009b. Cellulase adsorption and relationship to features of corn stover solids produced by leading pretreatments. Biotechnol Bioeng 103(2): 252± 267. Kumar R, Wyman CE. 2009c. Does change in accessibility with conversion depend on both the substrate and pretreatment technology? Biores Technol 100(18): 4193± 4202. Kumar R, Wyman CE. 2009d. Effect of additives on the digestibility of corn stover solids following pretreatment by leading technologies. Biotechnol Bioeng 102(6): 1544± 1557. Kumar R, Wyman CE. 2009e. Effect of enzyme supplementation at moderate cellulase loadings on initial glucose and xylose release from corn stover solids pretreated by leading technologies. Biotechnol Bioeng 102(2): 457±467. Kumar R, Wyman CE. 2009f. Effect of xylanase supplementation of cellulase on digestion of corn stover solids prepared by leading pretreatment technologies. Biores Technol 100(18): 4203±4213. Kumar R, Wyman CE. 2009g. Effects of cellulase and xylanase enzymes on the deconstruction of solids from pretreatment of poplar by leading technologies. Biotechnol Prog 25(2): 302±314.

ß Woodhead Publishing Limited, 2010

110

Bioalcohol production

Kumar R, Mago G, Balan V, Wyman CE. 2009. Physical and chemical characterizations of corn stover and poplar solids resulting from leading pretreatment technologies. Biores Technol 100(17): 3948±3962. Kyriacou A, Neufeld RJ, MacKenzie CR. 1989. Reversibility and competition in the adsorption of Trichoderma reesei cellulase components. Biotechnol Bioeng 33(5): 631±637. Laureano-Perez L, Teymouri F, Alizadeh H, Dale BE. 2005. Understanding factors that limit enzymatic hydrolysis of biomass: characterization of pretreated corn stover. Appl Biochem Biotechnol 121±124: 1081±1099. Lee D, Yu AHC, Wong KKY, Saddler JN. 1994. Evaluation of the enzymatic susceptibility of cellulosic substrates using specific hydrolysis rates and enzyme adsorption. Appl Biochem Biotechnol 45/46: 407±415. Lee SB, Shin H S, Ryu DDY. 1982. Adsoprtion of cellulase on cellulose: effect of physiochemical properties of cellulose on adsoprtion and rate of hydrolysis. Biotechnol Bioeng 24: 2137±2153. Lee Y-H, Fan LT. 1979. Some experimental observations on the adsorption of cellulase onto cellulose and on the behavior of adsorbed cellulase. Proc 9th Ann Biochem Eng Symp: 13±22. Lee Y-H, Fan LT. 1980. Properties and mode of action of cellulase. Adv Biochem Eng 17: 101±129. Lee Y-H, Fan LT. 1982. Kinetic studies of enzymatic hydrolysis of insoluble cellulose: analysis of the initial rates. Biotechnol Bioeng 24: 2383±2406. Lee Y-H, Fan LT. 1983. Kinetic studies of enzymatic hydrolysis of insoluble cellulose: (II). Analysis of extended hydrolysis times. Biotechnol Bioeng 25: 939±966. Lemos MA, Teixeira JA, Mota M, Gama FM. 2000. A simple method to separate cellulose-binding domains of fungal cellulases after digestion by a protease. Biotechnol Lett 22: 703±707. Lenz J, Esterbauer H, Sattler W, Schurz J, Wrentschur E. 1990. Changes of structure and morphology of regenerated cellulose caused by acid and enzymatic hydrolysis. J Appl Polym Sci 41(5±6): 1315±1326. Li J, Henriksson G, Gellerstedt G. 2007. Lignin depolymerization/repolymerization and its critical role for delignification of aspen wood by steam explosion. Biores Technol 98(16): 3061±3068. Liao W, Wen Z, Hurley S, Liu Y, Liu C, Chen S. 2005. Effects of hemicellulose and lignin on enzymatic hydrolysis of cellulose from dairy manure. Appl Biochem Biotechnol 121±124: 1017±1030. Linder A, Bergman R, Bodin A, Gatenholm P. 2003. Mechanism of assembly of xylan onto cellulose surfaces. Langmuir 19(12): 5072±5077. Liu C, Wyman CE. 2003. The effect of flow rate of compressed hot water on xylan, lignin, and total mass removal from corn stover. Ind Eng Chem Res 42(21): 5409± 5416. Liu C, Wyman CE. 2004a. Effect of the flow rate of a very dilute sulfuric acid on xylan, lignin, and total mass removal from corn stover. Ind Eng Chem Res 43(11): 2781± 2788. Liu C, Wyman CE. 2004b. Impact of fluid velocity on hot water only pretreatment of corn stover in a flowthrough reactor. Appl Biochem Biotechnol 113±116: 977±987. Liu C, Wyman CE. 2005. Partial flow of compressed-hot water through corn stover to enhance hemicellulose sugar recovery and enzymatic digestibility of cellulose. Biores Technol 96(18): 1978±1985. Lloyd TA, Wyman CE. 2005. Combined sugar yields for dilute sulfuric acid pretreatment

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

111

of corn stover followed by enzymatic hydrolysis of the remaining solids. Biores Technol 96(18): 1967±1977. Lu Y, Yang B, Gregg D, Saddler JN, Mansfield SD. 2002. Cellulase adsorption and an evaluation of enzyme recycle during hydrolysis of steam-exploded softwood residues. Appl Biochem Biotechnol 98±100 (Biotechnology for Fuels and Chemicals): 641±654. Lynd LR. 1996. Overview and evaluation of fuel ethanol from cellulosic biomass: technology, economics, the environment, and policy. Ann Rev Energy Environ 21: 403±465. Lynd LR, Cushman JH, Nichols RJ, Wyman CE. 1991. Fuel ethanol from cellulosic biomass. Science (Washington, DC, United States) 251(4999): 1318±1323. Lynd LR, Elander RT, Wyman CE. 1996. Likely features and costs of mature biomass ethanol technology. Appl Biochem Biotechnol 57/58 (Seventeenth Symposium on Biotechnology for Fuels and Chemicals, 1995): 741±761. Lynd LR, Wyman CE, Gerngross TU. 1999. Biocommodity engineering. Biotechnol Prog 15(5): 777±793. Lynd LR, Weimer PJ, van Zyl WH, Pretorius IS. 2002. Microbial cellulose utilization: fundamentals and biotechnology. Microbiol Mol Biol Rev 66(3): 506±577. Lynd LR, Laser MS, Bransby D, Dale BE, Davison B, Hamilton R, Himmel M, Keller M, McMillan JD, Sheehan J and others. 2008. How biotech can transform biofuels. Nat Biotechnol 26(2): 169±172. Ma A, Hu Q, Qu Y, Bai Z, Liu W, Zhuang G. 2008. The enzymatic hydrolysis rate of cellulose decreases with irreversible adsorption of cellobiohydrolase I. Enzyme Microb Technol 42(7): 543±547. Makkar HPS, Dawra RK, Singh B. 1987. Protein precipitation assay for quantitation of tannins: determination of protein in tannin-protein complex. Analyt Biochem 166(2): 435±439. Maloney MT, Chapman TW, Baker AJ. 1985. Dilute acid hydrolysis of paper birch: kinetics studies of xylan and acetyl-group hydrolysis. Biotechnol Bioeng 27(3): 355±361. Mandels M, Reese ET. 1965. Inhibition of cellulases. Ann Rev Phytopathol 3: 85±102. Mansfield SD, Meder R. 2003. Cellulose hydrolysis ± the role of monocomponent cellulases in crystalline cellulose degradation. Cellulose 10(2): 159±169. Mansfield SD, Mooney C, Saddler JN. 1999. Substrate and enzyme characteristics that limit cellulose hydrolysis. Bio Prog 15(5): 804±816. Martinez JM, Reguant J, Montero MA, Montane D, Salvado J, Farriol X. 1997. Hydrolytic pretreatment of softwood and almond shells. Degree of polymerization and enzymatic digestibility of the cellulose fraction. Ind Eng Chem Res 36(3): 688±696. Matthews JF, Skopec CE, Mason PE, Zuccato P, Torget RW, Sugiyama J, Himmel ME, Brady JW. 2006. Computer simulation studies of microcrystalline cellulose I[beta]. Carbohyd Res 341(1): 138±152. Medve J, Karlsson J, Lee D, Tjerneld F. 1998. Hydrolysis of microcrystalline cellulose by cellobiohydrolase I and endoglucanase II from Trichoderma reesei: adsorption, sugar production pattern, and synergism of the enzymes. Biotechnol Bioeng 59(5): 621±634. Merino Sandra T, Cherry J. 2007. Progress and challenges in enzyme development for biomass utilization. Adv Biochem Eng/Biotechnol 108: 95±120. Mes-Hartree M, Yu EKC, Reid ID, Saddler JN. 1987. Suitability of aspenwood biologically delignified with Pheblia tremellosus for fermentation to ethanol or butanediol. Appl Microbiol Biotechnol 26(2): 120±125.

ß Woodhead Publishing Limited, 2010

112

Bioalcohol production

Meunier-Goddik L, Penner MH. 1999. Enzyme-catalyzed saccharification of model celluloses in the presence of lignacious residues. J Agric Food Chem 47(1): 346±351. Michalowicz G, Toussaint B, Vignon MR. 1991. Ultrastructural changes in poplar cell wall during steam explosion treatment. Holzforschung 45(3): 175±179. Millett MA, Moore WE, Saeman JF. 1954. Preparation and properties of hydrocelluloses. J Ind Eng Chem 46: 1493±1497. Millett MA, Baker AJ, Satter LD. 1975. Pretreatments to enhance chemical, enzymatic, and microbiological attack of cellulosic materials. Biotechnol Bioeng Symp 5: 193± 219. Mitchell DJ, Grohmann K, Himmel ME, Dale BE, Schroeder HA. 1990. Effect of the degree of acetylation on the enzymic digestion of acetylated xylans. J Wood Chem Technol 10(1): 111±121. Mizutani C, Sethumadhavan K, Howley P, Bertoniere N. 2002. Effect of a nonionic surfactant on Trichoderma cellulase treatments of regenerated cellulose and cotton yarns. Cellulose 9(1): 83±89. Moloney A, Coughlan MP. 1983. Sorption of Talaromyces emersonii cellulase on cellulosic substrates. Biotechnol Bioeng 25(1): 271±280. Mooney CA, Mansfield SD, Tuohy MG, Saddler JN. 1997. The effect of lignin content on cellulose accessibility and enzymic hydrolysis of softwood pulps. Biological Sciences Symposium, San Francisco, pp. 259±265. Mooney CA, Mansfield SD, Touhy MG, Saddler JN. 1998. The effect of initial pore volume and lignin content on the enzymatic hydrolysis of softwoods. Biores Technol 64(2): 113±119. Mooney CA, Mansfield SD, Beatson RP, Saddler JN. 1999. The effect of fiber characteristics on hydrolysis and cellulase accessibility to softwood substrates. Enzyme Microb Technol 25(8±9): 644±650. Morrison IM. 1983. The effect of physical and chemical treatments on the degradation of wheat and barley straws by rumen liquor-pepsin and pepsin-cellulase systems. J Sci Food Agric 34(12): 1323±1329. Mosier N, Wyman C, Dale B, Elander R, Lee YY, Holtzapple MT, Ladisch M. 2005. Features of promising technologies for pretreatment of lignocellulosic biomass. Biores Technol 96(6): 673±686. Murashima K, Kosugi A, Doi RH. 2002. Synergistic effects on crystalline cellulose degradation between cellulosomal cellulases from Clostridium cellulovorans. J Bacteriol 184(18): 5088±5095. Murashima K, Kosugi A, Doi RH. 2003. Synergistic effects of cellulosomal xylanase and cellulases from Clostridium cellulovorans on plant cell wall degradation. J Bacteriol 185(5): 1518±1524. Murnen HK, Balan V, Chundawat SPS, Bals B, Da Costa Sousa L, Dale BE. 2007. Optimization of ammonia fiber expansion (AFEX) pretreatment and enzymatic hydrolysis of Miscanthus x giganteus to fermentable sugars. Biotechnol Prog 23(4): 846±850. Nagle NJ, Elander RT, Newman MM, Rohrback BT, Ruiz RO, Torget RW. 2002. Efficacy of a hot washing process for pretreated yellow poplar to enhance bioethanol production. Biotechnol Prog 18(4): 734±738. Nelson R, Oliver DW. 1971. Study of cellulose structure and its relation to reactivity. J Polym Sci 36: 305±320. Nidetzky B, Steiner W. 1993. A new approach for modeling cellulase-cellulose adsorption and the kinetics of the enzymatic hydrolysis of microcrystalline cellulose. Biotechnol Bioeng 42: 469±479.

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

113

Nidetzky B, Hayn M, Macarron R, Steiner W. 1993. Synergism of Trichoderma reesei cellulases while degrading different celluloses. Biotechnol Lett 15(1): 71±76. Nidetzky B, Steiner W, Claeyssens M. 1994a. Cellulose hydrolysis by the cellulases from Trichoderma reesei: adsorptions of two cellobiohydrolases, two endocellulases and their core proteins on filter paper and their relation to hydrolysis. Biochem J 303 (Pt 3): 817±823. Nidetzky B, Zachariae W, Gercken G, Hayn M, Steiner W. 1994b. Hydrolysis of cellooligosaccharides by Trichoderma reesei cellobiohydrolases: experimental data and kinetic modeling. Enzyme Microb Technol 16(1): 43±52. Nidetzky B, Claeyssens M, Steiner W. 1996. Cellulose degradation by the major cellulases from Trichoderma reesei: synergistic interaction and competition for binding sites on cellulose. Biotechnology in the Pulp and Paper Industry: Recent Advances in Applied and Fundamental Research, Proceedings of the International Conference on Biotechnology in the Pulp and Paper Industry, 11±15 June 1995, Vienna, pp. 537±542. Nieves RA, Ellis RP, Todd RJ, Johnson TJA, Grohmann K, Himmel ME. 1991. Visualization of Trichoderma reesei cellobiohydrolase I and endoglucanase I on aspen cellulose by using monoclonal antibody-colloidal gold conjugates. Appl Environ Microbiol 57(11): 3163±3170. Nutor JRK, Converse AO. 1991. The effect of enzyme and substrate levels on the specific hydrolysis rate of pretreated poplar wood. Appl Biochem Biotechnol 28±29: 757± 772. Nutt A, Sild V, Pettersson G, Johansson G. 1998. Progress curves. A mean for functional classification of cellulases. Eur J Biochem 258(1): 200±206. O'Neill A, Araujo R, Casal M, Guebitz G, Cavaco-Paulo A. 2007. Effect of the agitation on the adsorption and hydrolytic efficiency of cutinases on polyethylene terephthalate fibres. Enzyme Microb Technol 40(7): 1801±1805. Oh KD, Kim C. 1987. A study on enzymic hydrolysis of cellulose in an attrition bioreactor. Korean J Chem Eng 4(2): 105±112. È hgren K, Bura R, Saddler J, Zacchi G. 2007. Effect of hemicellulose and lignin removal O on enzymatic hydrolysis of steam pretreated corn stover. Biores Technol 98(13): 2503±2510. Okazaki M, Moo-Young M. 1978. Kinetics of enzymic hydrolysis of cellulose: analytical description of a mechanistic model. Biotechnol Bioeng 20(5): 637±663. Okazaki M, Miura Y, Moo-Young M. 1981. Synergistic effect of enzymic hydrolysis of cellulose. Adv Biotechnol [Proc 6th Int Ferment Symp] 2: 3±8. Ooshima H, Sakata M, Harano Y. 1983. Adsorption of cellulase from Trichoderma viride on cellulose. Biotechnol Bioeng 25(12): 3103±3114. Ooshima H, Sakata M, Harano Y. 1986. Enhancement of enzymatic hydrolysis of cellulose by surfactant. Biotechnol Bioeng 28(11): 1727±1734. Ooshima H, Burns DS, Converse AO. 1990. Adsorption of cellulase from Trichoderma reesei on cellulose and lignacious residue in wood pretreated by dilute sulfuric acid with explosive decompression. Biotechnol Bioeng 36(5): 446±452. Ooshima H, Kurakake M, Kato J, Harano Y. 1991. Enzymic activity of cellulase adsorbed on cellulose and its change during hydrolysis. Appl Biochem Biotechnol 31(3): 253± 266. Pala H, Mota M, Gama FM. 2007. Enzymatic depolymerisation of cellulose. Carbohyd Polym 68(1): 101±108. Palonen H. 2004. PhD Thesis: Role of Lignin in Enzymatic Hydrolysis of Lignocellulose. Helsinki: University of Helsinki, Finland.

ß Woodhead Publishing Limited, 2010

114

Bioalcohol production

Palonen H, Thomsen AB, Tenkanen M, Schmidt AS, Viikari L. 2004a. Evaluation of wet oxidation pretreatment for enzymatic hydrolysis of softwood. Appl Biochem Biotechnol 117(1): 1±17. Palonen H, Tjerneld F, Zacchi G, Tenkanen M. 2004b. Adsorption of Trichoderma reesei CBH I and EG II and their catalytic domains on steam pretreated softwood and isolated lignin. J Biotechnol 107(1): 65±72. Pan X, Xie D, Gilkes N, Gregg DJ, Saddler JN. 2005. Strategies to enhance the enzymatic hydrolysis of pretreated softwood with high residual lignin content. Appl Biochem Biotechnol 121±124: 1069±1079. Pan X, Gilkes N, Saddler JN. 2006. Effect of acetyl groups on enzymatic hydrolysis of cellulosic substrates. Holzforschung 60: 398±401. Panagiotou G, Olsson L. 2007. Effect of compounds released during pretreatment of wheat straw on microbial growth and enzymatic hydrolysis rates. Biotechnol Bioeng 96(2): 250±258. Paralikar KM, Betrabet SM. 1977. Electron-Diffraction Technique for Determination of Cellulose Crystallinity. J Appl Polym Sci 21(4): 899±903. Park J-W, Park K, Song H, Shin H. 2002. Saccharification and adsorption characteristics of modified cellulases with hydrophilic/hydrophobic copolymers. J Biotechnol 93(3): 203±208. Park JW, Takahata Y, Kajiuchi T, Akehata T. 1992. Effects of nonionic surfactant on enzymic hydrolysis of used newspaper. Biotechnol Bioeng 39(1): 117±120. Paul SS, Kamra DN, Sastry VRB, Sahu NP, Kumar A. 2003. Effect of phenolic monomers on biomass and hydrolytic enzyme activities of an anaerobic fungus isolated from wild nil gai (Boselaphus tragocamelus). Lett Appl Microbiol 36(6): 377±381. Pimenova NV, Hanley TR. 2003. Measurement of rheological properties of corn stover suspensions. Appl Biochem Biotechnol 105±108: 383±392. Pinto R, Carvalho J, Mota M, Gama M. 2006. Large-scale production of cellulosebinding domains. Adsorption studies using CBD-FITC conjugates. Cellulose (Dordrecht, Netherlands) 13(5): 557±569. Prabhu KA, Maheshwari R. 1999. Biochemical properties of xylanases from a thermophilic fungus, Melanocarpus albomyces, and their action on plant cell walls. J Biosci (Bangalore, India) 24(4): 461±470. Puls J, Wood TM. 1991. The degradation pattern of cellulose by extracellular cellulases of aerobic and anaerobic microorganisms. Biores Technol 36(1): 15±19. Puri VP. 1984. Effect of crystallinity and degree of polymerization of cellulose on enzymic saccharification. Biotechnol Bioeng 26(10): 1219±1222. Puri VP, Pearce GR. 1986. Alkali-explosion pretreatment of straw and bagasse for enzymic hydrolysis. Biotechnol Bioeng 28(4): 480±485. Ragauskas AJ, Williams CK, Davison BH, Britovsek G, Cairney J, Eckert CA, Frederick WJ, Jr., Hallett JP, Leak DJ, Liotta CL and others. 2006. The path forward for biofuels and biomaterials. Science 311(5760): 484±489. Ramos LP, Breuil C, Saddler JN. 1993. The use of enzyme recycling and the influence of sugar accumulation on cellulose hydrolysis by Trichoderma cellulases. Enzyme Microb Technol 15(1): 19±25. Reinikainen T, Teleman O, Teeri TT. 1995. Effects of pH and high ionic strength on the adsorption and activity of native and mutated cellobiohydrolase I from Trichoderma reesei. Proteins 22(4): 392±403. Rivard CJ, Adney WS, Himmel ME, Mitchell DJ, Vinzant TB, Grohmann K, Moens L, Chum H. 1992. Effects of natural polymer acetylation on the anaerobic

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

115

bioconversion to methane and carbon dioxide. Appl Biochem Biotechnol 34±35: 725±736. Rivers DB, Emert GH. 1988a. Factors affecting the enzymatic hydrolysis of bagasse and rice straw. Biol Wastes 26: 85±95. Rivers DB, Emert GH. 1988b. Factors affecting the enzymatic hydrolysis of municipalsolid-waste components. Biotechnol Bioeng 31(3): 278±281. Ryu DDY, Lee SB. 1982. Enzymic hydrolysis of cellulose: effects of structural properties of cellulose on hydrolysis kinetics. Enzyme Eng 6: 325±333. Ryu DDY, Lee SB. 1986. Enzymic hydrolysis of cellulose: determination of kinetic parameters. Chem Eng Commun 45(1±6): 119±134. Ryu DDY, Kim C, Mandels M. 1984. Competitive adsorption of cellulase components and its significance in a synergistic mechanism. Biotechnol Bioeng 26(5): 488±496. Ryu K, Kim Y. 1998. Adsorption of a xylanase purified from Pulpzyme HC onto alkalilignin and crystalline cellulose. Biotechnol Lett 20(10): 987±990. Saddler JN, Brownell HH, Clermont LP, Levitin N. 1982. Enzymic hydrolysis of cellulose and various pretreated wood fractions. Biotechnol Bioeng 24(6): 1389± 1402. Sakata M, Ooshima H, Harano Y. 1985. Effects of agitation on enzymatic saccharification of cellulose. Biotechnol Lett 7(9): 689±694. Samios E, Dart RK, Dawkins JV. 1997. Preparation, characterization and biodegradation studies on cellulose acetates with varying degrees of substitution. Polymer 38(12): 3045±3054. Sasaki T, Tanaka T, Nanbu N, Sato Y, Kainuma K. 1979. Correlation between x-ray diffraction measurements of cellulose crystalline structure and the susceptibility to microbial cellulase. Biotechnol Bioeng 21(6): 1031±1042. Sawada T, Nakamura Y, Kobayashi F, Kuwahara M, Watanabe T. 1995. Effects of fungal pretreatment and steam explosion pretreatment on enzymic saccharification of plant biomass. Biotechnol Bioeng 48(6): 719±724. Schell DJ, Torget R, Power A, Walter PJ, Grohmann K, Hinman ND. 1991. A technical and economic analysis of acid-catalyzed steam explosion and dilute sulfuric acid pretreatments using wheat straw or aspen wood chips. Appl Biochem Biotechnol 28±29: 87±97. Schubert C. 2006. Can biofuels finally take center stage? Nat Biotechnol 24(7): 777± 784. Schwald W, Brownell HH, Saddler JN. 1988a. Enzymatic hydrolysis of steam treated aspen wood: influence of partial hemicellulose and lignin removal prior to pretreatment. J Wood Chem Technol 8(4): 543±560. Schwald W, Chan M, Brownell HH, Saddler JN. 1988b. Influence of hemicellulose and lignin on the enzymic hydrolysis of wood. FEMS Symposium 43 (Biochem Genet Cell Degrad): 303±314. Schwanninger M, Rodrigues JC, Pereira H, Hinterstoisser B. 2004. Effects of short-time vibratory ball milling on the shape of FT-IR spectra of wood and cellulose. Vibrational Spectroscopy 36(1): 23±40. Selig MJ, Viamajala S, Decker SR, Tucker MP, Himmel ME, Vinzant TB. 2007. Deposition of lignin droplets produced during dilute acid pretreatment of maize stems retards enzymatic hydrolysis of cellulose. Biotechnol Prog 23(6): 1333± 1339. Selig MJ, Knoshaug EP, Adney WS, Himmel ME, Decker SR. 2008. Synergistic enhancement of cellobiohydrolase performance on pretreated corn stover by addition of xylanase and esterase activities. Biores Technol 99(11): 4997±5005.

ß Woodhead Publishing Limited, 2010

116

Bioalcohol production

Selig MJ, Vinzant T, Himmel M, Decker S. 2009. The effect of lignin removal by alkaline peroxide pretreatment on the susceptibility of corn stover to purified cellulolytic and xylanolytic enzymes. Appl Biochem Biotechnol 155(1): 94±103. Sendich E, Laser M, Kim S, Alizadeh H, Laureano-Perez L, Dale B, Lynd L. 2008. Recent process improvements for the ammonia fiber expansion (AFEX) process and resulting reductions in minimum ethanol selling price. Biores Technol 99(17): 8429±8435. Sethi B, Mishra S, Bisaria VS. 1998. Adsorption characteristics of cellulases from a constitutive mutant of Trichoderma reesei. J Ferment Bioeng 86(2): 233±235. Sewalt VJH, Glasser WG, Beauchemin KA. 1997a. Lignin impact on fiber degradation. 3. Reversal of inhibition of enzymic hydrolysis by chemical modification of lignin and by additives. J Agric Food Chem 45(5): 1823±1828. Sewalt VJH, Ni W, Jung HG, Dixon RA. 1997b. Lignin impact on fiber degradation: increased enzymic digestibility of genetically engineered tobacco (Nicotiana tabacum) stems reduced in lignin content. J Agric Food Chem 45(5): 1977±1983. Shallom D, Shoham Y. 2003. Microbial hemicellulases. Curr Opin Microbiol 6(3): 219± 228. Shevchenko SM, Beatson RP, Saddler JN. 1999. The nature of lignin from steam explosion/ enzymatic hydrolysis of softwood: structural features and possible uses: scientific note. Appl Biochem Biotechnol 77±79: 867±876. Sineiro J, Dominguez H, Nunez MJ, Lema JM. 1997. Inhibition of cellulase activity by sunflower polyphenols. Biotechnol Lett 19(6): 521±524. Sinitsyn A, Gusakov A, Vlasenko E. 1991. Effect of structural and physico-chemical features of cellulosic substrates on the efficiency of enzymatic hydrolysis. Appl Biochem Biotechnol 30(1): 43±59. Sinnott ML. 1998. The cellobiohydrolases of Trichoderma reesei: a review of indirect and direct evidence that their function is not just glycosidic bond hydrolysis. Biochem Soc Trans 26(2): 160±164. Stinson JA, Ham RK. 1995. Effect of lignin on the anaerobic decomposition of cellulose as determined through the use of a biochemical methane potential method. Environ Sci Technol 29(9): 2305±2310. Stutzenberger F, Lintz G. 1986. Hydrolysis products inhibit adsorption of Trichoderma reesei C30 cellulases to protein-extracted lucerne fibres. Enzyme Microb Technol 8(6): 341±344. Sudo K, Matsumura Y, Shimizu K. 1976. Enzymic hydrolysis of woods. I. Effect of delignification on the hydrolysis of woods by Trichoderma viride cellulase. Mokuzai Gakkaishi 22(12): 670±676. Suh J-H, Choi Y-J. 1996. Synergism among endo-xylanase, -xylosidase, and acetyl xylan esterase from Bacillus stearothermophilus. J Microbiol Biotechnol 6(3): 173± 178. Sun Y, Cheng J. 2002. Hydrolysis of lignocellulosic materials for ethanol production: a review. Biores Technol 83: 1±11. Sutcliffe R, Saddler JN. 1986. The role of lignin in the adsorption of cellulases during enzymic treatment of lignocellulosic material. Biotechnol Bioeng Symp 17: 749± 762. Suvajittanont W, Bothwell MK, McGuire J. 2000. Adsorption of Trichoderma reesei CBHI cellulase on silanized silica. Biotechnol Bioeng 69(6): 688±692. Tabka MG, Herpoel-Gimbert I, Monod F, Asther M, Sigoillot JC. 2006. Enzymatic saccharification of wheat straw for bioethanol production by a combined cellulase xylanase and feruloyl esterase treatment. Enzyme Microb Technol 39(4): 897±902.

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

117

Tanaka M, Ikesaka M, Matsuno R, Converse AO. 1988. Effect of pore size in substrate and diffusion of enzyme on hydrolysis of cellulosic materials with cellulases. Biotechnol Bioeng 32(5): 698±706. Taniguchi M, Suzuki H, Watanabe D, Sakai K, Hoshino K, Tanaka T. 2005. Evaluation of pretreatment with Pleurotus ostreatus for enzymatic hydrolysis of rice straw. J Biosci Bioeng 100(6): 637±643. Tarantili PA, Koullas DP, Christakopoulos P, Kekos D, Koukios EG, Macris BJ. 1996. Cross-synergism in enzymic hydrolysis of lignocellulosics: Mathematical correlations according to a hyperbolic model. Biomass Bioenergy 10(4): 213±219. Teeri TT. 1997. Crystalline cellulose degradation: new insight into the function of cellobiohydrolases. Trends Biotechnol 15(5): 160±167. Teeri TT, Koivula A, Linder M, Reinikainen T, Ruohonen L, Srisodsuk M, Claeyssens M, Jones TA. 1995. Modes of action of two Trichoderma reesei cellobiohydrolases. Prog Biotechnol 10 (Carbohydrate Bioengineering): 211±224. Teixeira L, Linden J, Schroeder H. 1999. Alkaline and peracetic acid pretreatments of biomass for ethanol production. Appl Biochem Biotechnol 77(1): 19±34. Tengborg C, Galbe M, Zacchi G. 2001. Influence of enzyme loading and physical parameters on the enzymatic hydrolysis of steam-pretreated softwood. Biotechnol Prog 17(1): 110±117. Tenkanen M, Buchert J, Viikari L. 1995. Binding of hemicellulases on isolated polysaccharide substrates. Enzyme Microb Technol 17(6): 499±505. Tenkanen M, Siika-aho M, Hausalo T, Puls J, Viikari L. 1996. Synergism of xylanolytic enzymes of Trichoderma reesei in the degradation of acetyl-4-Omethylglucuronoxylan. Biotechnology in the Pulp and Paper Industry: Recent Advances in Applied and Fundamental Research, Proceedings of the 6th International Conference on Biotechnology in the Pulp and Paper Industry, Vienna, 11±15 June, 1995: 503±508. Teymouri F, Laureano-Perez L, Alizadeh H, Dale BE. 2005. Optimization of the ammonia fiber explosion (AFEX) treatment parameters for enzymatic hydrolysis of corn stover. Biores Technol 96(18): 2014±2018. Thompson DN, Chen H-C, Grethlein HE. 1992. Comparison of pretreatment methods on the basis of available surface area. Biores Technol 39: 155±163. Tilman D, Hill J, Lehman C. 2006. Carbon-negative biofuels from low-input highdiversity grassland biomass. Science 314(5805): 1598±1600. Tilton RD, Robertson CR, Gast AP. 1991. Manipulation of hydrophobic interactions in protein adsorption. Langmuir 7(11): 2710±2718. Todorovic R, Grujic S, Matavulj M. 1987. Effect of reaction end-products on the activity of cellulolytic enzymes and xylanase of Trichoderma harzianum. Microbio Lett 36(143±144): 113±119. Torget R, Walter P, Himmel M, Grohmann K. 1991. Dilute-acid pretreatment of corn residues and short-rotation woody crops. Appl Biochem Biotechnol 28±29(1): 75± 86. Treimanis A, Leite M, Bikova T, Eisimonte M, Eremeeva T, Viesturs U. 1998. The preparation of LODP-cellulose, its chemical properties and enzymic hydrolysis. Advances in Lignocellulosics Chemistry for Ecologically Friendly Pulping and Bleaching Technologies, 5th European Workshop on Lignocellulosics and Pulp, University of Aveiro, Aveiro, Port., 30 Aug±2 Sept, 1998: 7±9. Tsao GT, Ladisch M, Ladisch C, Hsu TA, Dale B, Chou T. 1978. Fermentation substrates from cellulosic materials: production of fermentable sugars from cellulosic materials. Ann Rep Ferment Proc 2: 1±21.

ß Woodhead Publishing Limited, 2010

118

Bioalcohol production

Tu M, Chandra RP, Saddler JN. 2007. Recycling cellulases during the hydrolysis of steam exploded and ethanol pretreated lodgepole pine. Biotechnol Prog 23(5): 1130± 1137. Um B-H, Karim MN, Henk LL. 2003. Effect of sulfuric and phosphoric acid pretreatments on enzymatic hydrolysis of corn stover. Appl Biochem Biotechnol 105±108: 115±125. VaÈljamaÈe P. 2002. PhD Thesis: The Kinetics of Cellulose Enzymatic Hydrolysis: Implications of the Synergism Between Enzymes. Uppsala: Uppsala University. VaÈljamaÈe P, Sild V, Nutt A, Pettersson G, Johansson G. 1999. Acid hydrolysis of bacterial cellulose reveals different modes of synergistic action between cellobiohydrolase I and endoglucanase I. Eur J Biochem 266(2): 327±334. VaÈljamaÈe P, Pettersson G, Johansson G. 2001. Mechanism of substrate inhibition in cellulose synergistic degradation. Eur J Biochem 268(16): 4520±4526. van Oss CJ. 1995. Hydrophobicity of biosurfaces ± origin, quantitative determination and interaction energies. Colloids and Surfaces, B: Biointerfaces 5(3/4): 91±110. Venkatesh B, Leonardo da Costa S, Shishir PSC, Derek M, Lekh NS, Chambliss CK, Bruce ED. 2009. Enzymatic digestibility and pretreatment degradation products of AFEX-treated hardwoods Populus (Populus nigra). Biotechnol Prog 25(2): 365± 375. Vlasenko EY, Ding H, Labavitch JM, Shoemaker SP. 1997. Enzymic hydrolysis of pretreated rice straw. Biores Technol 59(2 & 3): 109±119. von Ossowski I, Stahlberg J, Koivula A, Piens K, Becker D, Boer H, Harle R, Harris M, Divne C, Mahdi S et al. 2003. Engineering the exo-loop of Trichoderma reesei cellobiohydrolase, Cel7A. A comparison with Phanerochaete chrysosporium Cel7D. J Mol Biol 333(4): 817±829. Walker LP, Wilson DB. 1991. Enzymatic hydrolysis of cellulose: an overview. Biores Technol 36(1): 3±14. Wang L, Zhang Y, Gao P. 2008. A novel function for the cellulose binding module of cellobiohydrolase I. Science in China Series C: Life Sciences 51(7): 620±629. Watson DL, Wilson DB, Walker LP. 2002. Synergism in binary mixtures of Thermobifida fusca cellulases Cel6B, Cel9A, and Cel5A on BMCC and Avicel. Appl Biochem Biotechnol 101(2): 97±111. Weil J, Westgate P, Kohlmann K, Ladisch MR. 1994. Cellulose pretreatments of lignocellulosic substrates. Enzyme Microb Technol 16(11): 1002±1004. Weil JR, Dien B, Bothast R, Hendrickson R, Mosier NS, Ladisch MR. 2002. Removal of fermentation inhibitors formed during pretreatment of biomass by polymeric adsorbents. Ind Eng Chem Res 41(24): 6132±6138. Weimer PJ, Hackney JM, French AD. 1995. Effects of chemical treatments and heating on the crystallinity of celluloses and their implications for evaluating the effect of crystallinity on cellulose biodegradation. Biotechnol Bioeng 48(2): 169±173. Weimer PJ, Hackney JM, Jung HJ, Hatfield RD. 2000. Fermentation of a bacterial cellulose/xylan composite by mixed ruminal microflora: implications for the role of polysaccharide matrix interactions in plant cell wall biodegradability. J Agric Food Chem 48(5): 1727±1733. Wendorf JR, Radke CJ, Blanch HW. 2004. Reduced protein adsorption at solid interfaces by sugar excipients. Biotechnol Bioeng 87(5): 565±573. Willies D. 2007. Masters Thesis: An Investigation into the Adsorption of Enzymes and BSA Protein onto Lignocellulosic Biomass Fractions and the Benefit to Hydrolysis of Non-Catalytic Additives. Hanover, NH, USA: Thayer School of Engineering, Dartmouth College.

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

119

Wilson CA, McCrae SI, Wood TM. 1994. Characterisation of a -glucosidase from the anaerobic rumen fungus Neocallimastix frontalis with particular reference to attack on cello-oligosaccharides. J Biotechnol 37(3): 217±227. Wingren A, Galbe M, Zacchi G. 2003. Techno-economic evaluation of producing ethanol from softwood: comparison of SSF and SHF and identification of bottlenecks. Biotechnol Prog 19(4): 1109±1117. Wong KKY, Deverell KF, Mackie KL, Clark TA, Donaldson LA. 1988. The relationship between fiber-porosity and cellulose digestibility in steam-exploded Pinus radiata. Biotechnol Bioeng 31(5): 447±456. Wood TM, McCrae SI. 1986. The effect of acetyl groups on the hydrolysis of ryegrass cell walls by xylanase and cellulase from Trichoderma koningii. Phytochemistry 25(5): 1053±1055. Wood TM, McCrae SI, Bhat KM. 1989. The mechanism of fungal cellulase action. Synergism between enzyme components of Penicillium pinophilum cellulase in solubilizing hydrogen-bond ordered cellulose. Biochem J 260: 37±43. Wooley R, Ruth M, Sheehan J, Ibsen K, Majdeski H, Galvez A. 1999. Lignocellulosic biomass to ethanol process design and economics utilizing co-current dilute acid prehydrolysis and enzymatic hydrolysis: current and futuristic scenarios. NREL Report No. TP-580-26157 (July). Wu Z, Lee YY. 1997. Ammonia recycled percolation as a complementary pretreatment to the dilute-acid process. Appl Biochem Biotechnol 63±65: 21±34. Wyman CE. 1999. Biomass ethanol: technical progress, opportunities, and commercial challenges. Ann Rev Energy Environ 24: 189±226. Wyman CE. 2003. Potential synergies and challenges in refining cellulosic biomass to fuels, chemicals, and power. Biotechnol Prog 19(2): 254±262. Wyman CE. 2007. What is (and is not) vital to advancing cellulosic ethanol. Trends Biotechnol 25(4): 153±157. Wyman CE, Dale BE, Elander RT, Holtzapple MT, Ladisch MR, Lee YY. 2005a. Comparative sugar recovery data from laboratory scale application of leading pretreatment technologies to corn stover. Biores Technol 96(18): 2026±2032. Wyman CE, Dale BE, Elander RT, Holtzapple MT, Ladisch MR, Lee YY. 2005b. Coordinated development of leading biomass pretreatment technologies. Biores Technol 96: 1959±1966. Xiao Z, Gao P, Qu Y, Wang T. 2001. Cellulose-binding domain of endoglucanase III from Trichoderma reesei disrupting the structure of cellulose. Biotechnol Lett 23(9): 711±715. Xiao Z, Zhang X, Gregg DJ, Saddler JN. 2004. Effects of sugar inhibition on cellulases and beta-glucosidase during enzymatic hydrolysis of softwood substrates. Appl Biochem Biotechnol 113±116: 1115±1126. Xu F, Ding H, Osborn D, Tejirian A, Brown K, Albano W, Sheehy N, Langston J. 2008. Partition of enzymes between the solvent and insoluble substrate during the hydrolysis of lignocellulose by cellulases. J Mol Catalysis B: Enzymatic 51(1±2): 42±48. Yamamoto K, Fujii T, Sudo K, Shimizu K. 1990. Changes in cell wall structure and enzymic hydrolysis of steam exploded bamboo and bamboo grass. Baiomasu Henkan Keikaku Kenkyu Hokoku 22: 20±34. Yang B, Wyman CE. 2004. Effect of xylan and lignin removal by batch and flowthrough pretreatment on the enzymatic digestibility of corn stover cellulose. Biotechnol Bioeng 86(1): 88±95. Yang B, Wyman CE. 2006. BSA treatment to enhance enzymatic hydrolysis of cellulose

ß Woodhead Publishing Limited, 2010

120

Bioalcohol production

in lignin containing substrates. Biotechnol Bioeng 94(4): 611±617. Yang B, Wyman CE. 2008. Pretreatment: the key to unlocking low-cost cellulosic ethanol. Biofuels, Bioproducts and Biorefining 2(1): 26±40. Yang B, Boussaid A, Mansfield SD, Gregg DJ, Saddler JN. 2002. Fast and efficient alkaline peroxide treatment to enhance the enzymatic digestibility of steamexploded softwood substrates. Biotechnol Bioeng 77(6): 678±684. Yang B, Gray MC, Liu C, Lloyd TA, Stuhler SL, Converse AO, Wyman CE. 2004. Unconventional relationships for hemicellulose hydrolysis and subsequent cellulose digestion. ACS Symposium Series 889 (Lignocellulose Biodegradation): 100±125. Yang B, Willies DM, Wyman CE. 2006. Changes in the enzymatic hydrolysis rate of Avicel cellulose with conversion. Biotechnol Bioeng 94(6): 1122±1128. Yoshida M, Liu Y, Uchida S, Kawarada K, Ukagami Y, Ichinose H, Kaneko S, Fukuda K. 2008. Effects of cellulose crystallinity, hemicellulose, and lignin on the enzymatic hydrolysis of Miscanthus sinensis to monosaccharides. Bioscience, Biotechnology, and Biochemistry 72(3): 805±810. Yu AHC, Lee D, Saddler JN. 1995. Adsorption and desorption of cellulase components during the hydrolysis of a steam-exploded birch substrate. Biotechnol Appl Biochem 21(2): 203±216. Yu H-S, Huang X-Y, Wang Y. 1998. Enhancement in enzymic hydrolysis of straw. 16±19 June 1998, Vancouver, BC, pp. C269±C272. Yu P, McKinnon JJ, Maenz DD, Olkowski AA, Racz VJ, Christensen DA. 2003. Enzymic release of reducing sugars from oat hulls by cellulase, as influenced by Aspergillus ferulic acid esterase and Trichoderma xylanase. J Agric Food Chem 51(1): 218± 223. Yuldashev BT, Rabinovich ML, Rakhimov MM. 1993. Comparative study of cellulase behavior on the cellulose and lignocellulose surface during enzymic hydrolysis. Prikladnaya Biokhimiya i Mikrobiologiya 29(2): 233±243. Zeng M, Mosier NS, Huang CP, Sherman DM, Ladisch MR. 2007. Microscopic examination of changes of plant cell structure in corn stover due to hot water pretreatment and enzymatic hydrolysis. Biotechnol Bioeng 97(2): 265±278. Zhang Y-HP. 2008. Reviving the carbohydrate economy via multi-product lignocellulose biorefineries. J Ind Microbiol Biotechnol 35(5): 367±375. Zhang Y-HP, Lynd LR. 2004a. Toward an aggregated understanding of enzymatic hydrolysis of cellulose: noncomplexed cellulase systems. Biotechnol Bioeng 88(7): 797±824. Zhang Y-HP, Lynd LR. 2004b. Toward an aggregated understanding of enzymatic hydrolysis of cellulose: noncomplexed cellulase systems. Biotechnol Bioeng 88(7): 797±824. Zhang Y-HP, Lynd LR. 2005. Determination of the number-average degree of polymerization of cellodextrins and cellulose with application to enzymatic hydrolysis. Biomacromolecules 6(3): 1510±1515. Zhang Y-HP, Lynd LR. 2006. A functionally based model for hydrolysis of cellulose by fungal cellulase. Biotechnol Bioeng 94(5): 888±898. Zhang Y-HP, Himmel ME, Mielenz JR. 2006. Outlook for cellulase improvement: Screening and selection strategies. Biotechnol Adv 24(5): 452±481. Zhang Y-HP, Ding S-Y, Mielenz JR, Cui J-B, Elander RT, Laser M, Himmel ME, McMillan JR, Lynd LR. 2007. Fractionating recalcitrant lignocellulose at modest reaction conditions. Biotechnol Bioeng 97(2): 214±223.

ß Woodhead Publishing Limited, 2010

Pretreated lignocelluloses biomass solids

121

Zheng Y, Pan Z, Zhang R, Labavitch J, Wang D, Teter S. 2007. Adsorption characteristics of cellulase and -glucosidase to lignin, cellulose and pretreated creeping wild ryegrass. June, Minneapolis, MN, pp 1±13. Zhu Y, Lee YY, Elander RT. 2005. Optimization of dilute-acid pretreatment of corn stover using a high-solids percolation reactor. Appl Biochem Biotechnol 121±124: 1045±1054.

ß Woodhead Publishing Limited, 2010

4

Solvent fractionation of lignocellulosic biomass N . S A T H I T S U K S A N O H , Z . Z H U , J . R O L L I N and Y . - H . P . Z H A N G , Virginia Polytechnic Institute and State University, USA

Abstract: Effectively breaking lignocellulose recalcitrance, releasing the locked polymeric sugars and coutilization of lignocellulose components is the largest technical and economical challenge for the emerging bioeconomy. Cellulose solvent- and organic solvent-based lignocellulose fractionation (COSLIF), has shown to effectively fractionate lignocellulose under modest reaction conditions. The resulting solids, amorphous cellulose, from corn stover, switchgrass, common reed, poplar, and hemp herds, showed high glucan digestibilities and fast enzymatic hydrolysis rates. Utilization of coproducts fractionated from lignocellulose (e.g., lignin, acetic acid, and hemicellulose) would greatly increase overall potential revenues of the future biorefineries. Key words: biomass, cellulosic ethanol, cellulose solvent, COSLIF, enzymatic cellulose hydrolysis, hemicellulose, lignin, lignocellulose fractionation.

4.1

Introduction

Concerns about the accumulation of greenhouse gases and the depletion of cheap fossil fuels (e.g., oil and natural gas) are motivating the use of sustainable primary energy sources, such as biomass, solar energy, nuclear energy, wind energy, tidal energy, and so on. Energy used for transportation usually costs more than stationary energy. Transportation fuels must have high energy density (MJ/kg or MJ/L) and the ability to generate high power density (W/kg) in a compact space (Zhang 2008a). Currently, affordable transportation is available through the combination of liquid fuels (gasoline and diesel) and internal combustion engines. In the future, the majority of automobile powertrain systems will rely on the hydrogen/electricity system because of its high energy utilization efficiency, minimal pollutant emissions, and diverse sustainable primary energy sources (Zhang 2008a; 2008b). However, this transition will take a relatively long time.

ß Woodhead Publishing Limited, 2010

Solvent fractionation of lignocellulosic biomass

123

Liquid ethanol is a near-term alternative liquid transportation fuel. Currently, most ethanol is produced from soluble sugars made from sugarcane and corn kernels, but competition between food demand and biofuels production has been a cause of soaring food prices. In addition, even complete conversion of the limited supply of corn kernels to first generation biofuels would replace only a small fraction of current transportation fuel demand. The production of second generation biofuels from more abundant, non-food lignocellulosic biomass is a more sustainable option. Advanced biofuels produced this way would provide many benefits, such as promoting rural economies, enhancing energy security, increasing process energy efficiency (output/input), and decreasing greenhouse gas emissions. Typical lignocellulosic biomass includes: · · · ·

agricultural residues (e.g., wheat and rice straws, corn stover, bagasse, etc.) forest residues (e.g., wood chips and sawdust) dedicated bioenergy crops (e.g., switchgrass, poplar, etc.) industrial and municipal solid wastes.

Overcoming lignocellulosic biomass recalcitrance followed by enzymatic hydrolysis of reactive polymeric carbohydrates (i.e., cost-efficient liberation of fermentable sugars from biomass) is perhaps the most challenging technical and economic barrier to biorefinery success (Fortman et al. 2008; Lynd et al. 2008; Zhang 2008c). Pretreatment is among the most costly steps in biochemical conversion of biomass (Eggeman and Elander 2005; Wyman et al. 2005b), accounting for up to 40% of the total processing cost (Lynd 1996; Lynd et al. 2005). Also, it affects the costs of other operations including size reduction prior to pretreatment and enzymatic hydrolysis and fermentation after pretreatment. Pretreatment can also strongly influence downstream costs involving detoxification (if inhibitors are generated), enzymatic hydrolysis rate and enzyme loading, mixing power, product concentration, product purification, power generation, waste treatment demands, and other process variables (Wyman et al. 2005b; Zhang 2008c). The main purpose of this chapter is to describe a new technology that uses a combination of a cellulose solvent (concentrated phosphoric acid) and an organic solvent for fractionating lignocellulose components (cellulose, hemicellulose, acetic acid and lignin) under modest reaction conditions. Potential applications of the isolated lignocellulose components are briefly discussed.

4.2

Lignocellulosic biomass

Lignocellulosic biomass is cheaper than crude oil, natural gas, or corn kernels on the basis of energy content (Lynd et al. 2008; Zhang 2008c; Zhang and Lynd 2008). The key challenge is cost-effective release of the locked polymeric carbohydrates to fermentable soluble sugars (e.g., glucose and xylose). As

ß Woodhead Publishing Limited, 2010

124

Bioalcohol production

opposed to the relatively mature thermochemical conversion of biomass to syngas, biological conversion of biomass to fermentable sugars has great potential for reductions in processing costs and capital investment. It is believed that with intensive research and development efforts, biological conversion will become the predominant pathway for biomass utilization. Lignocellulosic biomass is a natural composite containing three main biopolymers (cellulose, hemicellulose, and lignin) that are intertwined chemically and physically. The complex structure of biomass makes chemical and biological degradation difficult. Many factors affect biomass recalcitrance, including substrate accessibility to cellulase, degree of polymerization, cellulose crystallinity, and lignin and hemicellulose contents (Himmel et al. 2007; Mansfield et al. 1999; Zhang and Lynd 2004). Two major factors are (1) low cellulose accessibility to cellulase (CAC) that hinders the enzymes from working efficiently and (2) lignin and hemicellulose on the surface of cellulose that blocks cellulase from accessing cellulose efficiently (Hong et al. 2007; Moxley et al. 2008; Zhang et al. 2006a; 2007; Zhang and Lynd 2004; 2006).

4.2.1

Cellulose and enzymatic cellulose hydrolysis

Cellulose is the most abundant component of lignocellulosic biomass, comprising around 30±50% of its dry weight. Cellulose is a homopolysaccharide made of anhydroglucopyranose linked by -1,4-glucosidic linkages, with a degree of polymerization (DP) ranging from several hundred to more than 10 000 (Zhang and Lynd 2004). Highly ordered hydrogen bonds and van der Waal's forces among nearby anhydroglucose units result in the formation of crystalline microfibrils (Zhang 2008c; Zhang and Lynd 2004). The microfibrils form further microfibrils, which constitute the basic framework of the plant cell walls and provide rigidity and strength. These crystalline cellulose chain bundles have very low substrate accessibility to large-size cellulases (Hong et al. 2007; 2008; Zhang and Lynd 2004), resulting in low hydrolysis rates occurring on crystalline cellulose (Zhang et al. 2006a). Cellulose hydrolysis requires endoglucanase, cellobiohydrolases, and glucosidase to work together (Zhang et al. 2006b; Zhang and Lynd 2004). Endoglucanase is responsible for cleaving accessible -glucosidic bonds of the cellulose chain randomly and creating more ends for the action of cellobiohydrolases. Cellobiohydrolases hydrolyze the ends of cellulose chains and release soluble cellobiose into the aqueous phase. The release of soluble sugars from solid substrate to the aqueous phase is rate-limiting for the whole hydrolysis process (Zhang and Lynd 2004; 2006). -Glucosidase cleaves cellobiose to form glucose, which reduces cellobiose inhibition for endoglucanase and cellobiohydrolase. In general, cellobiohydrolase is more sensitive to product inhibition than endoglucanase (Zhang and Lynd 2004). Enzymatic cellulose hydrolysis is a very complicated process, involving heterogeneous

ß Woodhead Publishing Limited, 2010

Solvent fractionation of lignocellulosic biomass

125

substrate properties as well as synergy and competition among several enzymes on limited accessible solid surfaces. A functionally based mathematic model (Zhang and Lynd 2006) has been developed to correlate disparate phenomena into an aggregated system and give some useful predictions for lignocellulose pretreatment and cellulase engineering. For example, the simulation results clearly suggest that the optimal ratio of cellobiohydrolase to endoglucanase for maximal synergy is a dynamic rather than constant value, depending on substrate properties (degree of polymerization and substrate accessibility), enzyme loadings, and reaction time (Zhang and Lynd 2006). One of the model's most important predictions is that increasing substrate accessibility to cellulase is the most important factor for increasing enzymatic hydrolysis rates (Zhang and Lynd 2006). This insight was the driving force for the development of a new technology called cellulose solvent- and organic solvent-based lignocellulose fractionation (COSLIF) (Moxley et al. 2008; Zhang et al. 2007). Quantitative determination of cellulose accessibility to cellulase (CAC) is valuable for investigating complicated enzymatic cellulose hydrolysis mechanisms. None of the previous methods for measuring cellulose accessibility ± nitrogen adsorption-based Brunauer-Emmett-Teller (BET), size exclusion chromatography, small angle X-ray scattering (SAXS), microscopy ± is perfectly applied in the enzymatic cellulose hydrolysis process because: (1) enzymatic cellulose hydrolysis occurs on the surface of hydrated solid matter in the aqueous phase (i.e., dried cellulosic samples have completely different supramolecular structures from hydrated samples) (Zhang and Lynd 2004); (2) cellulases are large-size molecules; and (3) cellulase is preferentially adsorbed on the 110 face of cellulose fibers that cellulase can hydrolyze (Chanzy et al. 1984; Hong et al. 2007; Lehtio et al. 2003). Small-size molecule adsorption methods, such as BET, water vapor sorption, alkali swelling, or the exchange of H to D atoms with D2O, result in an overestimation of CAC (Hong et al. 2007; Zhang and Lynd 2004). Cellulase-size exclusion chromatography can neither differentiate the effective cellulose surface for adsorption and hydrolysis nor account for the external surface (Hong et al. 2007). Maximum cellulase adsorption capacity has been suggested to represent CAC (Zhang and Lynd 2004) and the data in the literature have been summarized and used before (Zhang and Lynd 2004; 2006). However, it is relatively difficult to obtain reliable and accurate data based on the adsorption of active cellulases on cellulose because cellulase can hydrolyze substrate during the adsorption measurement, especially for easily-hydrolyzed cellulose (Steiner et al. 1988), resulting in rapid changes in substrate characteristics (Fan et al. 1980; Tanaka et al. 1986). Therefore, many adsorption studies have been conducted at a decreased temperature (e.g., 4 ëC) in order to minimize hydrolysis effects (Kyriacou et al. 1989; Medve et al. 1997; Ooshima et al. 1983; Reinikainen et al. 1995; Ryu et al. 1984; Steiner et al. 1988). The drawback of this strategy is that active cellulase adsorption at low temperatures may be significantly

ß Woodhead Publishing Limited, 2010

126

Bioalcohol production

different from that at hydrolysis temperatures (Kyriacou et al. 1988; 1989; Ooshima et al. 1983). Quantitative determination of cellulose accessibility to cellulase (m2/g cellulose) was established based on the Langmuir adsorption of a fusion protein containing a cellulose-binding module (CBM) and a green fluorescent protein (GFP) (Hong et al. 2007). One molecule of the recombinant fusion protein occupied 21.2 cellobiose lattices on the 110 face of bacterial cellulose nanofibers (Hong et al. 2007). The CAC values of several cellulosic materials ± regenerated amorphous cellulose (RAC), bacterial microcrystalline cellulose (BMCC), Whatman No. 1 filter paper, fibrous cellulose powder (CF1), and microcrystalline cellulose (Avicel FMC PH105) ± are 41.9, 33.5, 9.76, 4.53, and 2.38 m2/g, respectively. The CAC value of RAC made from Avicel is 17.6-fold larger than that of Avicel (Hong et al. 2007). The fastest hydrolysis of amorphous cellulose was observed for 10 g RAC/L at an enzyme loading of 15 filter paper units (FPUs) per gram of cellulose within three hours (Zhang et al. 2006a).

4.2.2

Hemicellulose

Hemicellulose, the second most abundant polysaccharide in lignocellulosic biomass, is a heteropolymer containing primarily pentoses (xylose), as well as some hexoses (e..g., glucose and mannose). The main role of hemicellulose is to interact with cellulose and lignin to cross-link the cellulose microfibrils to the lignin matrix. Since hemicellulose is a small-size branched polysaccharide with DP ranging from ~200 to 400, it is more vulnerable than cellulose to catalysts and enzymes. For example, complete hydrolysis of hemicellulose to monomeric sugars can be implemented at 121 ëC for 1 hour in the presence of 1% sulfuric acid, but 4% sulfuric acid is needed for cellulosic fragments (Moxley and Zhang 2007). In general, dilute acid or alkali can efficiently hydrolyze hemicellulose, resulting in a disruption of the linkages among cellulose, hemicellulose, and lignin. Complete enzymatic hydrolysis of hemicellulose requires more enzymes than cellulose (more than three enzymes) to work together, due to the more complicated structure of hemicellulose. More information about these hemicellulases has been reviewed elsewhere (Shallom and Shoham 2003; Singh et al. 2003). Enzymatic hemicellulose hydrolysis is much faster than enzymatic cellulose hydrolysis, due to its amorphous structure and high turn-over number of hemicellulases. Hemicellulose has broad applications. Hemicellulose has been utilized in the form of plant gum in thickeners, adhesives, protective colloids, emulsifiers, stabilizers as well as biodegradable oxygen barrier films (Hartman et al. 2006; Kamm and Kamm 2004; Zhang 2008c). Oligosaccharides may provide a source of even higher value-added products, such as animal feed additives (Davis et al. 2002; Fernandez et al. 2002). This polymer's primary monomeric sugar, xylose, can be easily fermented to sugar alcohol, xylitol (Mussatto et al. 2005; Woodyer

ß Woodhead Publishing Limited, 2010

Solvent fractionation of lignocellulosic biomass

127

et al. 2005). Fufural, a xylose degradation product, can be used in the production of lubricants, coatings, adhesives, and furan resins. Effective fermentation of xylose to ethanol also could provide an unlimited market for xylose (Gray et al. 2006; Zhang et al. 1995).

4.2.3

Lignin

Lignin is the most abundant phenolic biopolymer found in nature. Low-quality industrial lignin is isolated by the paper pulp industry, but this material is usually burned for energy generation and chemical recycling. The research and development of lignin utilization is lagging because of limited supplies of high-quality lignin. Likewise, the lack of ready markets for high-quality lignin limits the motivation to conduct R&D pertaining to its isolation. On the laboratory scale, high-quality lignin has been demonstrated to work as a substitute for polymeric materials, such as phenolic powder resins, polyurethane, and polyisocyanuratre foams as well as epoxy resins. Because it is a good adsorbent and has excellent adhesive, rheological, and collidial properties, lignin can also be used as a partial replacement for phenolic binders for oriented-strand board production (Lora and Glasser 2002; Zhang 2008c). Lignin is a raw chemical precursor for DMSO, vanilla, phenol, and ethylene (Eckert et al. 2007; Lora and Glasser 2002; Reddy and Yang 2005) and can also be converted to value-added carbon fiber with a selling price ranging from ~$5±20/kg (Zhang 2008c). Broad potential lignin applications also exist for agricultural chemicals. For example, lignin, a biodegradable UVlight antioxidant absorbent, is appropriate for release-controlled pesticides and slow-release fertilizers containing ionically or organically bound nitrogen or other fertilizers. This use offers an important ecological benefit, as slowrelease fertilizers are crucial for decreasing non-point groundwater pollution. One of the largest-scale uses of lignin could be as a soil conditioner to aid in formation of humus. In the future, lignin could also be used for coating easily degradable biomass for carbon sequestration purposes. Pre-isolation of lignin prior to ethanol fermentation can provide a good feedstock for further conversion to fuel additives or for thermochemical conversion to synthetic diesel (Zhang 2008c). Nearly pure lignin can be processed much more easily and efficiently by thermochemical catalysis (e.g., pyrolysis or gasification) than crude lignocellulosic biomass. Lignin applications for high-end markets, such as polymer substitutions and carbon fiber, are expected to be pursued before low-end lignin markets. Although lignin-based products do not currently compete with products derived from petrochemicals, the technical and economic situations are changing rapidly.

ß Woodhead Publishing Limited, 2010

128

4.3

Bioalcohol production

Cellulose solvent-based lignocellulose pretreatment

A number of pretreatment technologies have been studied and developed over the years. Nearly all lignocellulose treatments can be divided into three categories: (1) physical methods, including dry milling, wet milling, irradiation, and microwave treatment; (2) chemical methods, using dilute acids (dilute H2SO4, H3PO4, HCl, acetic acid, formic acid/HCl), alkalis (NaOH, lime, ammonia, amine, etc.), organosolv, oxidizing agents (O3, NO, H2O2, NaClO2), steam explosion with or without catalysts, CO2 explosion, ammonia fiber explosion (AFEX), hot water, hot water with flow-through, supercritical fluid extraction (CO2, CO2/H2O, CO2/SO2, NH3, H2O) and so on; and (3) biological methods, such as the use of white rod fungi. Many current mainstream lignocellulose pretreatments (e.g., steam explosion, dilute acid, ammonia based pretreatment) cannot efficiently disrupt orderly hydrogen bonds among glucan chains in crystalline cellulose. Instead, they aim at removing hemicellulose or lignin. The pretreated biomass has relatively slow hydrolysis rates and modest cellulose digestibility (Wyman et al. 2005a; 2005b). Moreover, many pretreatment methods are feedstock-specific. For example, ammonia fiber explosion (AFEX) has been found to be relatively ineffective for pretreating woody biomass (Alizadeh et al. 2005; Chundawat et al. 2007; Dale et al. 1996; Murnen et al. 2007). Until now nearly all intensively studied pretreatments share one or more common shortcomings: (1) severe pretreatment conditions (except AFEX), resulting in sugar degradation and inhibitor formation (Klinke et al. 2004; Mussatto and Roberto 2004); (2) low or modest cellulose digestibility because of the presence of residual lignin and hemicellulose (Wyman et al. 2005a); (3) high cellulase loading required; (4) slow hydrolysis rate because a significant fraction of pretreated lignocellulose remains crystalline; (5) large utility consumption (Eggeman and Elander 2005); (6) huge capital investment because of poor economy of scale (Zhang 2008c); (7) low co-utilization of all the major components of lignocellulose (except organosolv) (Zhang 2008c; Zhang et al. 2007). Because hemicellulose is the most vulnerable component in lignocellulosic biomass, many of lignocellulose pretreatments are focused on hemicellulose removal. Such pretreatments include steam explosion and dilute acid. Although most hemicellulose can be removed efficiently in these processes, the condensed lignin on the crystalline cellulose surface still slows down hydrolysis rates and decreases cellulose digestibility. Subsequent lignin-targeting processes have been tested to remove this remaining lignin by using oxidative reagents, flowthrough, and organic solvents.

ß Woodhead Publishing Limited, 2010

Solvent fractionation of lignocellulosic biomass

129

Hydrolysis of pure crystalline cellulose has a relatively slow hydrolysis rate and low glucan digestibility (Zhang et al. 2006a; 2007), suggesting that the efficient removal of hemicellulose and lignin is not enough for high yield enzymatic cellulose hydrolysis. Therefore, there is an urgent need to find new pretreatment technologies that can overcome the recalcitrance of lignocellulose by targeting the most recalcitrant component ± cellulose.

4.3.1

Cellulose solvent-only lignocellulose pretreatment

The first attempt to overcome biomass recalcitrance by using cellulose solvents was taken by Professors Mike Ladisch and George Tsao in 1978 (Ladisch et al. 1978). After searching for a number of cellulose solvents, they found that Cadoxen, an alkali solution of CdO in aqueous ethylenediamine, can dissolve biomass. The resultant regenerated amorphous cellulose (RAC) from pure cellulose is hydrolyzed quickly by cellulase with a very high digestability, while the glucan digestibility of pretreated biomass is modest (Ladisch et al. 1978). A drawback of this technology is that Cadoxen is corrosive and toxic, and any remaining toxic components in the solid biomass may inhibit subsequent enzyme hydrolysis and fermentation steps. With the invention of ionic liquids that dissolve cellulose (Swatloski et al. 2002), several attempts have been made to pretreat biomass by using different cellulose solvents (Dadi et al. 2006; KilpelaÈinen et al. 2007; Zhu 2008). Enzymatic glucan digestibility of ionic-liquid pretreated biomass ranges widely (Dadi et al. 2006; KilpelaÈinen et al. 2007; Zhu 2008), suggesting that more research is needed to understand its mechanisms. Remaining hemicellulose and lignin fractions on the surface of cellulose could be an obstacle to efficient cellulose hydrolysis. In addition, efficient recycling of costly ionic solvents remains challenging.

4.3.2

Cellulose solvent- and organic solvent-based lignocellulose fractionation (COSLIF)

In order to deal effectively with two root causes of the recalcitrance of lignocellulose ± breaking up orderly hydrogen bonds in crystalline cellulose and removing lignin and hemicellulose from the surface of cellulose ± a novel pretreatment process has been developed, cellulose solvent- and organic solventbased lignocellulose fractionation (COSLIF) (Zhang et al. 2007). COSLIF not only fractionates lignocellulose components based on the significantly different solubility of cellulose, hemicellulose, and lignin in the cellulose solvent, organic solvent, and water, respectively, but also recycles the solvents due to a large difference in solvent volatility (Zhang et al. 2007). Figure 4.1 shows the overall flowchart of the COSLIF technology where concentrated phosphoric acid is used as the cellulose solvent and acetone as the

ß Woodhead Publishing Limited, 2010

ß Woodhead Publishing Limited, 2010

4.1 Flowchart of cellulose solvent- and organic solvent-based lignocellulose fractionation (COSLIF) with recycling of concentrated phosphoric acid and acetone.

Solvent fractionation of lignocellulosic biomass

131

organic solvent (Zhang et al. 2007). The key design principles of COSLIF are (1) de-crystallization of the cellulose fibers (i.e., more cellulose accessibility so that cellulase can work on the substrate more efficiently), (2) removal of partial lignin and hemicellulose from cellulose (i.e., fewer substrate obstacles to the enzymes, so that cellulase can access the substrate more efficiently), and (3) modest reaction conditions (i.e., a decrease in sugar degradation, less inhibitor formation, lower utility consumption, and less capital investment). After testing five lignocellulosic feedstocks: corn stover, switchgrass, hurds of industrial hemps, common reed, and hybrid poplar, phosphoric acid was discovered to destroy biomass structure efficiently only beyond the critical concentration (i.e., ~83%); the reaction time ranges from 45 to 60 min, depending on the type of biomass. Five different well-pretreated biomass types have similar hydrolysis performance at an enzyme loading of 15 FPUs of cellulase and 30 units of -glucosidase per gram of glucan. The glucan digestibilities were around 90% at hour 12 and ~94±97% at hour 24 (Fig. 4.2(a)±(e)). One of the most widely studied pretreatments has been dilute acid (DA) (Bernardez et al. 1993; Grethlein 1985; Ooshima et al. 1990; Schell et al. 2003). This process is usually conducted at high temperatures and high pressures catalyzed by a dilute acid (often sulfuric acid). Dilute acid at high temperatures removes acid-labile hemicellulose. By doing so, the linkages among cellulose, hemicellulose, and lignin are disrupted (Converse 1993; Kumar and Wyman 2008; Lloyd and Wyman 2005; Moxley and Zhang 2007; Schell et al. 2003). In a comparison study between the COSLIF and DA technologies, glucan, hemicellulose, and lignin contents of COSLIF-pretreated and DA-pretreated corn stover were examined (Zhu et al. 2009). The results are shown in Table 4.1. As is evident from Table 4.1, COSLIF can remove more lignin than DA, while retaining more glucan and hemicellulose. The higher solid sugar retention by COSLIF is attractive because this allows a more rapid release of fermentable sugars during the enzymatic hydrolysis step. Figure 4.2(f) presents the different hydrolysis profiles for the same corn stover pretreated by COSLIF and dilute acid. The glucan digestibility of the COSLIF-pretreated corn stover reached more than 90% at hour 12 and 97% at hour 24. In contrast, the DA-pretreated corn stove had much slower hydrolysis rates, and its final digestibility was 84% at hour 72. Table 4.1 Comparison of the effects of COSLIF and DA on biomass composition

Glucan content Hemicellulose content Lignin content

COSLIF

Dilute acid (DA)

58:2  2:5% 6:2  0:3% 19:7  0:3%

53:7  1:5% 3:4  0:2% 30:3  0:7%

ß Woodhead Publishing Limited, 2010

132

Bioalcohol production

4.2 Hydrolysis curves for different feedstocks at 15 FPUs of cellulase/g glucan and 50 ëC. Pretreatment conditions: (a) corn stover (85% H3PO4, 50 ëC and 40 min), (b) switchgrass (85% H3PO4, 50 ëC and 40 min), (c) common reed (85% H3PO4, 50 ëC and 60 min), (d) industrial hemp hurds (85% H3PO4, 50 ëC and 60 min), (e) poplar (85% H3PO4, 50 ëC and 60 min) and (f) corn stover (COSLIF, 85% H 3 PO 4 at 50 ëC for 60 min) vs. dilute acid (DA) pretreatment of the same feedstocks (1.4% H2SO4, 165ëC, and 8 min).

ß Woodhead Publishing Limited, 2010

ß Woodhead Publishing Limited, 2010

4.3 Mass balance for switchgrass via the COSLIF technology and enzymatic hydrolysis at 50 ëC with 10 g/L glucan, 15 FPUs of cellulase and 30 units of -glucosidase per gram of glucan.

134

Bioalcohol production

Figure 4.3 presents the material balance of swithgrass pretreated by the COSLIF technology and enzymatic cellulose hydrolysis at an enzyme loading of 15 FPUs of cellulase and 10 units of -glucosidase per gram of glucan. The overall glucose and xylose yields were calculated to be 85% and 63%, respectively. With technological improvements (e.g., a supplementary hemicellulase for enzymatic hemicellulose hydrolysis, optimization of reaction conditions, pre-extraction of water soluble sugars before the pretreatment, and washing methods and conditions such as solvent temperatures and flow rates), higher xylose recovery yields are expected without the sacrifice of glucose yields.

4.3.3

Supramolecular structures

The supramolecular structural changes in industrial hemp hurds before and after the different pretreatments are shown by using a scanning electron microscope (SEM) (Fig. 4.4). The intact biomass presents its plant cell vascular bundles and its fibril structure under SEM (Fig. 4.4(a)). Modest pretreatment conditions (84.0% H3PO4, 50 ëC and 30 min) open larger holes on the surface of plant cell walls by removing the easily digested lignocellulose fraction (i.e., hemicellulose) but the supramolecular fibril structure receives only partial damage (Fig. 4.4(b)). A well-treated lignocellulose sample (84.0% H3PO4, 50 ëC and 60 min) shows all fibrous structures of the lignocellulose completely disrupted (Fig. 4.4(c)). These images are completely different from those after treatments such as dilute acid and ammonia recycle percolation, which show residual structures even after very long treatment times (Kim and Lee 2005; Zeng et al. 2007). The dramatic changes in the supramolecular structures of corn stover before and after pretreatment are also shown by the atomic force microscopy (AFM) images in Fig. 4.5. Before pretreatment, the corn stover cell wall structures and elementary cellulose fibers are clearly identified (Fig. 4.5(a)). After pretreatment, no fibril structures are observed (Fig. 4.4(b)), indicating that concentrated phosphoric acid not only disrupts all the linkages among cellulose, hemicellulose, and lignin, but also breaks up the orderly hydrogen bonds among glucan chains. The much faster hydrolysis rates and higher glucan digestibility

4.4 SEM micrographs of hurds of industrial hemps. (a) intact biomass, (b) modestly treated biomass, and (c) well-treated biomass.

ß Woodhead Publishing Limited, 2010

Solvent fractionation of lignocellulosic biomass

135

4.5 AFM images for corn stover. (a) intact biomass and (b) well-pretreated biomass.

of the COSLIF-pretreated corn stover are attributed to more effective biomass structure destruction (Figs 4.4 and 4.5).

4.4

Future trends

Lignocellulose fractionation based on the different solubilities of lignocellulose components in different solvents is a new concept. The COSLIF technology, which exploits this idea, is in its infancy (Moxley et al. 2008; Sathitsuksanoh et al. 2009; 2010; Zhang et al. 2007). COSLIF pretreatment has several advantages, such as high glucan digestibility, a fast hydrolysis rate, low cellulase use, nearly feedstock-independent pretreatment results, greater potential revenues from co-products (acetic acid and lignin or even hemicellulose), and minimal formation of inhibitors. Several challenges remain, however, such as the high ratios of cellulose solvent and organic solvent to biomass, which may result in high processing costs for efficient recycling of both solvents and possibly high capital investment. Therefore, further studies of the COSLIF technology will be focused on: · decreasing cellulose solvent use per unit biomass by finding better cellulose solvents, · decreasing organic solvent use per unit biomass by using better organic solvents and efficient washing methods, · efficiently recycling both solvents through flashing, distillation or fractionation distillation, · identifying suitable solid/liquid unit operations, · efficiently regenerating the cellulose solvent, · characterizing the properties of isolated lignin,

ß Woodhead Publishing Limited, 2010

136 · · · ·

Bioalcohol production

developing new applications for relatively pure lignin, studying the feasibility of cellulase recycling, conducting economic analysis based on an ASPEN-Plus model, and validating technology feasibility with a pilot plant.

Substantial progress can be made in these areas, and the principles of lignocellulose fractionation have potential applications in lignocellulose-based biorefineries. In the short term, cellulosic ethanol production based on celluloserich wastes from existing industries, such as corn fiber from corn ethanol biorefineries, wheat hull from flour processing facilities, and sawdust from lumber manufacturers, is more attractive, since integrated biorefineries could not only solve solid waste disposal problems but also produce value-added products such as biofuels. Much smaller biorefineries that utilize cellulosic waste from on-site manufacturers could be profitable due to the large savings in feedstock costs (~$30±90/ton of biomass, i.e., $0.35±1.00 per gallon of cellulosic ethanol). The applicability of this nearly feedstock-independent COSLIF technology for biomass residues from local manufacturers could provide great opportunities to build profitable small-size biorefineries (i.e., 100 tons of biomass per day) that could produce ~2.8 million gallons of cellulosic ethanol per year, as well as acetic acid as a value-added co-product. In the long term, full utilization of lignocellulose components other than carbohydrates, such as lignin, will be extremely important for the bioeconomy.

4.5

Sources of further information and advice

· Classic reviews of cellulose solvents (Fengel and Wegener 1984; Jayme and Lang 1963; Pereira et al. 1988; Zhang and Lynd 2003). · Ionic liquids as a cellulose solvent (ElSeoud et al. 2007; Swatloski et al. 2002). · The cellulose solvent-only lignocellulose pretreatment (Ladisch et al. 1978). · The COSLIF technology (Moxley et al. 2008; Zhang et al. 2007).

4.6

References

Alizadeh H, Teymouri F, Gilbert TI, Dale BE. 2005. Pretreatment of switchgrass by ammonia fiber explosion (AFEX). Appl. Biochem. Biotechnol. 121/124: 1133± 1141. Bernardez TD, Lyford K, Hogsett DA, Lynd LR. 1993. Adsorption of Clostridium thermocellum cellulase onto pretreated mixed hardwood, Avicel and Lignin. Biotechnol. Bioeng. 43: 899±907. Chanzy H, Henrissat B, Vuong R. 1984. Colloidal gold labeling of 1,4- -glucan cellobiohydrolase adsorbed on cellulose substrates. FEBS Lett. 172: 193±196. Chundawat S, Venkatesh B, Dale B. 2007. Effect of particle size based separation of milled corn stover on AFEX pretreatment and enzymatic digestibility. Biotechnol. Bioeng. 96(2): 219±231.

ß Woodhead Publishing Limited, 2010

Solvent fractionation of lignocellulosic biomass

137

Converse AO. 1993. Substrate factors limiting enzymatic hydrolysis. In: Saddler JN, editor. Bioconversion of Forest and Agricultural Plant Residues. Oxford: CAB International, pp. 93±106. Dadi AP, Varanasi S, Schall CA. 2006. Enhancement of cellulose saccharification kinetics using an ionic liquid pretreatment step. Biotechnol. Bioeng. 95(5): 904± 910. Dale BE, Leong CK, Pham TK, Esquivel VM, Rios I, Latimer VM. 1996. Hydrolysis of lignocellulosics at low enzyme levels: application of the AFEX process. Biores. Technol. 56: 111±116. Davis ME, Maxwell CV, Brown DC, de Rodas BZ, Johnson ZB, Kegley EB, Hellwig DH, Dvorak RA. 2002. Effect of dietary mannan oligosaccharides and(or) pharmacological additions of copper sulfate on growth performance and immunocompetence of weanling and growing/finishing pigs. J. Anim. Sci. 80: 2887±2894. Eckert C, Liottaabc C, Ragauskasb A, Hallettac J, Kitchensac C, Hillac E, Draucker L. 2007. Tunable solvents for fine chemicals from the biorefinery. Green Chemistry 9: 545±548. Eggeman T, Elander RT. 2005. Process and economic analysis of pretreatment technologies. Biores. Technol. 96: 2019±2025. ElSeoud OA, Koschella A, Fidale LC, Dorn S, Heinze T. 2007. Applications of ionic liquids in carbohydrate chemistry: a window of opportunities. Biomacromolecules 8(9): 2629±2647. Fan LT, Lee Y-H, Beardmore DH. 1980. Mechanism of the enzymatic hydrolysis of cellulose: effects of major structural features of cellulose on enzymatic hydrolysis. Biotechnol. Bioeng. 22: 177±199. Fengel D, Wegener G. 1984. Wood: Chemistry, Ultrastructure, Reactions. Berlin: Walter de Gruyter & Co. Fernandez F, Hinton M, Van Gils B. 2002. Dietary mannan-oligosaccharides and their effect on chicken caecal microflora in relation to Salmonella Enteritidis colonization. Avian Pathol. 31: 49±58. Fortman JL, Chhabra S, Mukhopadhyay A, Chou H, Lee TS, Steen E, Keasling JD. 2008. Biofuel alternatives to ethanol: pumping the microbial well. Trends Biotechnol. 26(7): 375±381. Gray KA, Zhao L, Emptage M. 2006. Bioethanol. Curr. Opin. Chem. Biol. 10(2): 141± 146. Grethlein HE. 1985. The effect of pore size distribution on the rate of enzymatic hydrolysis of cellulose substrates. Bio/Technol. 3: 155±160. Hartman J, Albertsson A-C, Lindblad MS, SjoÈberg J. 2006. Oxygen barrier materials from renewable sources: material properties of softwood hemicellulose-based films. J. Appl. Polym. Sci. 100(4): 2985±2991. Himmel ME, Ding S-Y, Johnson DK, Adney WS, Nimlos MR, Brady JW, Foust TD. 2007. Biomass recalcitrance: engineering plants and enzymes for biofuels production. Science 315(5813): 804±807. Hong J, Ye X, Zhang Y-HP. 2007. Quantitative determination of cellulose accessibility to cellulase based on adsorption of a nonhydrolytic fusion protein containing CBM and GFP with its applications. Langmuir 23(25): 12535±12540. Hong J, Ye X, Wang Y, Zhang Y-HP. 2008. Bioseparation of recombinant cellulose binding module-protein by affinity adsorption on an ultra-high-capacity cellulosic adsorbent. Anal. Chem. Acta 621: 193±199. Jayme G, Lang F. 1963. Cellulose solvents. Methods Carbohydr. Chem. 3: 75±83.

ß Woodhead Publishing Limited, 2010

138

Bioalcohol production

Kamm B, Kamm M. 2004. Principles of biorefineries. Appl. Microbiol. Biotechnol. 64: 137±145. KilpelaÈinen I, Xie H, King A, Granstrom M, Heikkinen S, Argyropoulos D. 2007. Dissolution of wood in ionic liquids. J. Agric. Food Chem. 55: 9142±9148. Kim TH, Lee YY. 2005. Pretreatment and fractionation of corn stover by ammonia recycle percolation process. Biores. Technol. 96: 2007±2013. Klinke HB, Thomsen AB, Ahring BK. 2004. Inhibition of ethanol-producing yeast and bacteria by degradation products produced during pre-treatment of biomass. Appl. Microbiol. Biotechnol. 66: 10±26. Kumar R, Wyman CE. 2008. An improved method to directly estimate cellulase adsorption on biomass solids. Enzyme Microb. Technol. 42(5): 426±433. Kyriacou A, Neufeld RJ, MacKenzie CR. 1988. Effect of physical parameters on the adsorption characteristics of fractionated Trichoderma reesei cellulase components. Enzyme Microb. Technol. 10: 675±681. Kyriacou A, Neufeld RJ, MacKenzie CR. 1989. Reversibility and competition in the adsorption of Trichoderma reesei cellulase components. Biotechnol. Bioeng. 33: 631±637. Ladisch MR, Ladisch CM, Tsao GT. 1978. Cellulose to sugars: new path gives quantitative yield. Science 201: 743±745. Lehtio J, Sugiyama J, Gustavsson M, Fransson L, Linder M, Teeri TT. 2003. The binding specificity and affinity determinants of family 1 and family 3 cellulose binding modules. Proc. Natl. Acad. Sci. USA 100: 484±489. Lloyd TA, Wyman CE. 2005. Combined sugar yields for dilute sulfuric acid pretreatment of corn stover followed by enzymatic hydrolysis of the remaining solids. Biores. Technol. 96: 1967±1977. Lora JH, Glasser WG. 2002. Recent industrial applications of lignin: a sustainable alternative to nonrenewable materials. J. Polym. Environ. 10: 39±48. Lynd LR. 1996. Overview and evaluation of fuel ethanol from cellulosic biomass: technology, economics, the environment, and policy. Annu. Rev. Energy Env. 21: 403±465. Lynd LR, van Zyl WH, McBride JE, Laser M. 2005. Consolidated bioprocessing of cellulosic biomass: an update. Curr. Opin. Biotechnol. 16: 577±583. Lynd LR, Laser MS, Bransby D, Dale BE, Davison B, Hamilton R, Himmel M, Keller M, McMillan JD, Sheehan J and others. 2008. How biotech can transform biofuels. Nat. Biotechnol. 26(2): 169±172. Mansfield SD, Mooney C, Saddler JN. 1999. Substrates and enzyme characteristics that limit cellulose hydrolysis. Biotechnol. Prog. 15: 804±816. Medve J, Stahlberg J, Tjerneld F. 1997. Isotherms for adsorption of cellobiohydrolase I and II from Trichoderma reesei on microcrystalline cellulose. Appl. Biochem. Biotechnol. 66: 39±56. Moxley G, Zhang Y-HP. 2007. More accurate determination of acid-labile carbohydrate composition in lignocellulose by modified quantitative saccharification. Energy Fuels 21: 3684±3688. Moxley GM, Zhu Z, Zhang Y-HP. 2008. Efficient sugar release by the cellulose solvent based lignocellulose fractionation technology and enzymatic cellulose hydrolysis. J. Agric. Food Chem. 56 (17): 7885±7890. Murnen HK, Balan V, Chundawat SPS, Bals B, daCostaSousa L, Dale BE. 2007. Optimization of ammonia fiber expansion (AFEX) pretreatment and enzymatic hydrolysis of Miscanthus x giganteus to fermentable sugars. Biotechnol. Prog. 23(4): 846±850.

ß Woodhead Publishing Limited, 2010

Solvent fractionation of lignocellulosic biomass

139

Mussatto SI, Roberto IC. 2004. Alternative for detoxification of diluted acid lignocellulosic hydrolyzates for use in fermenative processes: a review. Biores. Technol. 93: 1±10. Mussatto SI, Dragone G, Roberto IC. 2005. Kinetic behavior of Candida guilliermondii yeast during xylitol production from brewer's spent grain hemicellulosic hydrolysate. Biotechnol. Prog. 21(4): 1352±1356. Ooshima H, Sakata M, Harano Y. 1983. Adsorption of cellulase from Trichoderma viride on cellulose. Biotechnol. Bioeng. 25: 3103±3114. Ooshima H, Burns DS, Converse AO. 1990. Adsorption of cellulase from Trichoderma reesei on cellulose and lignacious residue in wood pretreated by dilute sulfuric acid with explosive decompression. Biotechnol. Bioeng. 36: 446±452. Pereira AN, Mobedshahi M, Ladisch MR. 1988. Preparation of cellodextrins. Methods Enzymol. 160: 26±43. Reddy N, Yang Y. 2005. Biofibers from agricultural byproducts for industrial applications. Trends Biotechnol. 23: 22±27. Reinikainen T, Teleman O, Teeri TT. 1995. Effects of pH and high ionic strength on the adsorption and activity of native and mutated cellobiohydrolase I from Trichoderma reesei. Proteins 22: 392±403. Ryu DDY, Kim C, Mandels M. 1984. Competitive adsorption of cellulase components and its significance in synergistic mechanism. Biotechnol. Bioeng. 26: 488±496. Sathitsuksanoh N, Zhu ZG, Templeton N, Rollin JA, Harvey SP, Zhang Y-HP. 2009. Saccharification of a potential bioenergy crop, Phragmites australis (Common Reed), by lignocellulose fractionation followed by enzymatic hydrolysis at decreased cellulase loadings. Ind. Eng. Chem. Res. 48(13): 6441±6447. Sathitsuksanoh N, Zhu Z, Ho T-J, Bai M-D, Zhang Y-HP. 2010. Bamboo saccharification through cellulose solvent-based biomass pretreatment followed by enzymatic hydrolysis at ultra-low cellulase loadings. Biores. Technol. DOI: 10.1016/ j.biortech.2009.09.081. Schell DJ, Farmer J, Newman M, McMillan JD. 2003. Dilute-sulfuric acid pretreatment of corn stover in pilot-scale reactor: investigation of yields, kinetics, and enzymatic digestibilities of solids. Appl. Biochem. Biotechnol. 105/108: 69±85. Shallom D, Shoham Y. 2003. Microbial hemicellulases. Curr. Opin. Microbiol. 6: 219± 228. Singh S, Madlala AM, Prior BA. 2003. Thermomyces lanuginosus: properties of strains and their hemicellulases. FEMS Microbiol. Lett. 27: 3±16. Steiner W, Sattler W, Esterbauer H. 1988. Adsorption of Trichoderma reesei cellulase on cellulose: experimental data and their analysis by different equations. Biotechnol. Bioeng. 32: 853±865. Swatloski RP, Spear SK, Holbrey JD, Rogers RD. 2002. Dissolution of cellulose with ionic liquids. J. Am. Chem. Soc. 124: 4974±4975. Tanaka M, Nakamura H, Taniguchi M, Morita T, Matsuno R, Kamikubo T. 1986. Elucidation of adsorption processes of cellulases during hydrolysis of crystalline cellulose. Appl. Microbiol. Biotechnol. 23: 263±268. Woodyer R, Simurdiak M, van der Donk WA, Zhao H. 2005. Heterologous expression, purification, and characterization of a highly active xylose reductase from Neurospora crassa. Appl. Environ. Microbiol. 71(3): 1642±1647. Wyman CE, Dale BE, Elander RT, Holtzapple MT, Ladisch MR, Lee YY. 2005a. Comparative sugar recovery data from laboratory scale application of leading pretreatment technologies to corn stover. Biores. Technol. 96: 2026±2032. Wyman CE, Dale BE, Elander RT, Holtzapple MT, Ladisch MR, Lee YY. 2005b.

ß Woodhead Publishing Limited, 2010

140

Bioalcohol production

Coordinated development of leading biomass pretreatment technologies. Biores. Technol. 96: 1959±1966. Zeng M, Mosier NS, Huang C-P, Sherman DM, Ladisch MR. 2007. Microscopic examination of changes of plant cell structure in corn stover due to hot water pretreatment and enzymatic hydrolysis. Biotechnol. Bioeng. 97(2): 265±278. Zhang M, Eddy C, Deanda K, Finkestein M, Picataggio S. 1995. Metabolic engineering of a pentose metabolism pathway in ethanologenic Zymomonas Mobilis. Science 267: 240±243. Zhang Y-HP. 2008a. Driving tomorrow by sugars. Public Service Review: Science & Technology 1: 126. Zhang Y-HP. 2008b. A sweet solution. Public Service Review: Science & Technology 1: 150. Zhang Y-HP. 2008c. Reviving the carbohydrate economy via multi-product biorefineries. J. Ind. Microbiol. Biotechnol. 35(5): 367±375. Zhang Y-HP, Lynd LR. 2003. Cellodextrin preparation by mixed-acid hydrolysis and chromatographic separation. Anal. Biochem. 322: 225±232. Zhang Y-HP, Lynd LR. 2004. Toward an aggregated understanding of enzymatic hydrolysis of cellulose: noncomplexed cellulase systems. Biotechnol. Bioeng. 88: 797±824. Zhang Y-HP, Lynd LR. 2006. A functionally-based model for hydrolysis of cellulose by fungal cellulase. Biotechnol. Bioeng. 94: 888±898. Zhang Y-HP, Lynd LR. 2008. New directions and leap forward concepts: consolidated bioprocessing. In: Himmel ME, editor. Biomass Recalcitrance: Deconstructing the Plant Cell Wall for Bioenergy. Oxford: Blackwell Publishing. Zhang Y-HP, Cui J-B, Lynd LR, Kuang LR. 2006a. A transition from cellulose swelling to cellulose dissolution by o-phosphoric acid: evidences from enzymatic hydrolysis and supramolecular structure. Biomacromolecules 7(2): 644±648. Zhang Y-HP, Himmel M, Mielenz JR. 2006b. Outlook for cellulase improvement: screening and selection strategies. Biotechnol. Adv. 24(5): 452±481. Zhang Y-HP, Ding S-Y, Mielenz JR, Elander R, Laser M, Himmel M, McMillan JD, Lynd LR. 2007. Fractionating recalcitrant lignocellulose at modest reaction conditions. Biotechnol. Bioeng. 97(2): 214±223. Zhu S. 2008. Use of ionic liquids for the efficient utilization of lignocellulosic materials. J. Chem. Technol. Biotechnol. 83(6): 777±779. Zhu ZG, Sathitsuksanoh N, Vinzant T, Schell DJ, McMillan JD, Zhang Y-HP. 2009. Comparative study of corn stover pretreated by dilute acid and cellulose solventbased lignocellulose fractionation: enzymatic hydrolysis, supramolecular structure, and substrate accessibility. Biotechnol. Bioeng. 103(4): 715±724.

ß Woodhead Publishing Limited, 2010

5

Dilute and concentrated acid hydrolysis of lignocellulosic biomass A . S H A H B A Z I and B . Z H A N G , North Carolina Agricultural and Technical State University, USA

Abstract: Utilizing lignocellulosic biomass as a feedstock will allow us to significantly expand ethanol production capacity. The use of acid hydrolysis for the conversion of lignocellulosic feedstock to monomer sugars is a process that has been studied for the last 100 years. Generally, there are two types of acid hydrolysis: dilute and concentrated, each having unique properties and effects on biomass, and each having advantages and disadvantages in terms of economics. This chapter reviews dilute acid hydrolysis, concentrated acid hydrolysis, inhibitors of ethanol production, apparatus used in acid pretreatment, ethanol production plants currently using acid hydrolysis, and unit operations pertinent to the ethanol industry. Key words: lignocellulosic biomass, dilute acid hydrolysis, concentrated acid hydrolysis, inhibitors, apparatus of acid pretreatment, ethanol production plants.

5.1

Introduction

The use of acid hydrolysis for the conversion of cellulose to glucose is a process that has been studied for the last 100 years. This technology is unique in that it enables widely available cellulosic materials (biomass) to be converted into sugar in an economical manner, thus providing an inexpensive raw material for fermentation into ethanol. This method has become an alternative for commercial use due to the low cost of biomass. Despite new advancements in technology, unique system optimizations in acid hydrolysis (which are dependent on location and feedstock) are still needed in order for ethanol to compete with petroleum, as well as keep up with the world's growing energy demands. Generally, there are two types of acid hydrolysis: dilute and concentrated, each having unique properties and effects on biomass, and each having advantages and disadvantages in terms of economics.

5.2

Dilute acid hydrolysis

Dilute acid (0.5±1.0% sulfuric acid) pretreatment at moderate temperatures (140±190 ëC) can effectively remove and recover most of the hemicellulose as

ß Woodhead Publishing Limited, 2010

144

Bioalcohol production

dissolved sugars. Under this method, glucose yields from cellulose increase with hemicellulose removal to almost 100% (Knappert et al., 1981). Dilute acid hydrolysis consists of two chemical reactions. One reaction converts cellulosic materials to sugar and the other converts sugars into other chemicals, many of which inhibit the growth of downstream fermentation microbes. The same conditions that cause the first reaction to occur simultaneously cause overdegradation of sugars and lignin, creating inhibitory compounds such as organic acids, furans, and phenols. During the pretreatment process, lignin is disrupted and partial lignin is dissolved, increasing cellulose susceptibility to enzymes (Yang and Wyman, 2004). The addition of acid produces higher hemicellulose sugar yields (Garrote et al., 1999; Parajo et al., 2004) than pretreatment without acid, in a process known as `autohydrolysis' (Heitz et al., 1991; Saddler et al., 1993). Similar to the use of dilute sulfuric acid, the use of sulfur dioxide (SO2) enhances glucose yields (Mackie et al., 1985). However, many prefer dilute sulfuric acid pretreatment because it is cheap and produces hemicellulose yields of up to 90% (Hsu, 1996).

5.2.1

Flow-through acid pretreatment

The use of very dilute sulfuric acid (appromixiately 0.07%) in a flow-through reactor configuration is more effective at acid levels under 0.1% than the typical dilute acid pretreatment, which is usually performed at acid levels of 0.7±3.0%. In one study, the more reactive hemicellulose in yellow poplar was hydrolyzed at lower temperatures in a countercurrent flow-through pretreatment.

5.3

Concentrated acid hydrolysis

Concentrated acid hydrolysis (about 70% acid content) uses a low temperature (100 ëF) and low pressure. The rate of cellulose recovery from the initial pretreatment process and the conversion rate of cellulose to glucose under this process are much higher (90%) than with dilute acid hydrolysis. One concentrated acid hydrolysis model was developed by USDA and further refined by Purdue University and the Tennessee Valley Authority (Energy Efficiency and Renewable Energy web site). The TVA concentrated acid process is conducted in two stages. In the first stage, corn stover is combined with dilute (10%) sulfuric acid, and heated to 100 ëC for 2±6 hours in the first (or hemicellulose) hydrolysis reactor. The low temperatures and pressures minimize the degradation of sugars. To recover the sugars, the hydrolyzed material in the first reactor is soaked in water and then drained. This liquid contains hemicellulose sugar. In order to recover the cellulose, the hemicellulose-free substrate is then dried and soaked in a 30±40% concentration of sulfuric acid for 1±4 hr as a precellulose hydrolysis step. This material is then dehydrated, causing the acid

ß Woodhead Publishing Limited, 2010

Dilute and concentrated acid hydrolysis

145

concentration in the material to increase to 70%. After reacting in another vessel for 1±4 hr at 100 ëC, the reactor contents are filtered to remove solids and recover the sugar and acid. The sugar/acid solution from the second stage is then recycled to the first stage to provide the acid for the first stage hydrolysis. The sugars from the second stage cellulose hydrolysis are thus recovered in the liquid from the first stage hydrolysis. The primary advantage of the concentrated process is the high sugar recovery efficiency, which can be on the order of more than 90% of both hemicellulose and cellulose sugars (Badger, 2002). The low temperatures and pressures employed also allow the use of relatively low-cost materials, such as fiberglass tanks and piping. The weaknesses, compared to other processes, are in its relatively slow rate of conversion, and the fact that more economical and efficient acid recovery systems need to be developed. Unless acid is removed, large quantities of lime must be used to neutralize the sugar solution, which requires the disposal of salts. This adds to the cost and makes the end product more expensive.

5.3.1

Concentrated phosphoric acid

A pretreatment process has been developed that uses a highly volatile organic solvent (acetone), together with concentrated phosphoric acid as a non-volatile cellulose solvent, and water to fractionate lignocellulose into amorphous cellulose, hemicellulose, lignin, and acetic acid under moderate reaction conditions (50 ëC and atmospheric pressure). One study utilizing this process achieved its highest sugar yields after enzymatic hydrolysis at an enzyme loading rate of 15 filter paper units of cellulase and 60 IU of beta-glucosidase per gram of glucan. This was due to minimal sugar degradation during fractionation, as well as the high enzymatic cellulose digestibility (97% in 24 h) that occurred during the hydrolysis step (Zhang et al., 2007) (Fig. 5.1).

5.4

Process and apparatus of acid pretreatment

5.4.1

Co-current reactor

Co-current/batch reactors are typical of dilute acid pretreatment. In a co-current pretreatment reactor, biomass liquid slurry passes through heat exchangers, is heated to the desired temperature (140±180 ëC), and held at temperature for 5±20 min as the slurry passes through an insulated plug-flow, snake-coil (heat exchanger 2). The slurry is cooled and the heat is recovered by counter-current heat exchanger 1 with the incoming slurry (Fig. 5.2).

5.4.2

Batch reactor

The batch reactor has been used for pretreatment at 140±210 ëC, for example, in the 1 L Parr high pressure reactor (Parr Instruments, Moline, IL). For a typical

ß Woodhead Publishing Limited, 2010

146

Bioalcohol production

5.1 Flowchart of cellulose-solvent and organic-solvent lignocellulose fractionation with recycling of concentrated phosphoric acid and acetone.

run, 50 grams of feedstock and 500 ml of dilute sulfuric acid were placed inside the cylinder. The cylinder was then sealed and purged with nitrogen gas at a flow rate of 80 ml/minute in order to remove air and to prevent secondary reactions, such as thermal cracking and repolymerization. The reactor was heated to the desired temperature. After completing the reaction, the cylinder was cooled down by running cooling water through the coils inside the reactor for 30 minutes. The contents were filtered to separate the water solution and the pretreated biomass. The water solution was collected into a sample bottle for later analysis. The glucose yields resulting from enzymatic hydrolysis after pretreatment with 1% dilute sulfuric acid pretreatment were 54.8% using the 1 L Parr reactor. A high temperature (180±190 ëC) pretreatment was more effective than a lower temperature (121 ëC) pretreatment. It has been recommended that further studies be carried out at a faster heating rate and shorter residence time for the pretreatment process (Zhang et al., 2009).

5.4.3

Percolation reactors

Percolation reactors could reduce degradation time for sugars. In some studies of this process, liquid was forced through a packed biomass bed in the percolation

ß Woodhead Publishing Limited, 2010

Dilute and concentrated acid hydrolysis

147

5.2 The co-current pretreatment method.

reactor (Zhu et al., 2004). The percolation process may enhance hemicellulose and lignin removal rates, and result in high yields of hemicellulose and even cellulose (Bobleter, 1994; Allen et al., 1996; Liu and Wyman, 2003) (Fig. 5.3). However, percolation or flow-through reactors are challenging to apply commercially, and the high amounts of water used require high energy inputs for both pretreatment and product recovery processes.

5.4.4

Pretreatment using a solvent extractor

The solvent extractor provides a fast and efficient way to extract various biomass samples for sugar analysis to determine sample viability for alcohol production (Dionex Application Note 363). In one study of this process, a Dionex ASE 350 Accelerated Solvent Extractor was used for dilute acid pretreatment of biomass below 190 ëC (Zhang et al., 2010) (Fig. 5.4). Approximately 2.0 g of ground biomass (composed of cattails) was placed into a tared 66 mL Dionium extraction cell containing a glass fiber filter. Then the appropriate number of 250 ml collection vials were weighed and placed onto the ASE system. The extractor passed 1% H2SO4 into the cell containing biomass, then the cell was heated to the desired temperature (140±190 ëC) in situ, and the desired temperature was maintained for 5±15 min.

ß Woodhead Publishing Limited, 2010

148

Bioalcohol production

5.3 Laboratory setup for percolation reactor system: 1) acid tank; 2) HPLC pump; 3) preheating coil; 4) percolation reactor; 5) thermometer; 6) gas chromatography oven; 7) heat exchanger; 8) product collecting tank; 9) three-way valve; 10) nitrogen gas.

5.4 A Dionex ASE 350 Accelerated Solvent Extractor.

ß Woodhead Publishing Limited, 2010

Dilute and concentrated acid hydrolysis

149

5.5 The flow-through pretreatment method.

5.4.5

Flow-through reactor

Flow-through technologies pass a fresh acid/water stream through a higher temperature zone and then through a lower temperature region to improve yield and reduce the exposure of sugars to severe conditions (Fig. 5.5). Temperatures of 140, 150, and 174 ëC were studied for the first stage, while the second stage was run at 170, 180, 190, 200, and 204 ëC. Times of 10, 15, and 20 min were used in each with sulfuric acid levels of 0.0735%, 0.4015%, and 0.735% by weight. Approximately 83.0±100% of the hemicellulose and 26.3±52.5% of the lignin were solubilized, with sugar monomers comprising 79.6±95.2% of the hemicellulose and oligemers comprising the remainder. The pretreated cellulose portion obtained up to 90% enzymatic digestibility (Torget et al., 1996, 1998, 1999). Although hemicellulose yields were excellent and lower acid loadings yielded highly digestible cellulose, the equipment configurations as well as the high ratio of liquid to solids used in the flow-through process require significant energy input, and offset these advantages.

5.5

Ethanol production plants currently using acid hydrolysis

5.5.1

Arkenol Inc.

Arkenol has developed and holds patents on various technologies, control strategies and computer programs, directed towards acid hydrolysis. Most ethanol

ß Woodhead Publishing Limited, 2010

ß Woodhead Publishing Limited, 2010

5.6 Simplified flow diagram showing conversion of cellulose/hemicellulose to mixed sugars using Arkenol's concentrated acid hydrolysis.

Dilute and concentrated acid hydrolysis

151

plants worldwide that use concentrated acid hydrolysis, such as Izumi biorefinery in Japan and BlueFire Ethanol in California, use the Arkenol model (Fig. 5.6). Arkenol uses a skid-mounted, 150 liter vessel made of 316SS and coated with a Kynar-like coating for acid resistance. The helical screw mixer is mounted on an orbit arm that sweeps the inner walls of the reactor to insure proper mixing. The direction and speed (e.g., up/down, clockwise/counterclockwise) of the orbit arm and the mixer are adjustable at the vessel's control panel. Feedstock is added through the hinged access port located on the reactor head. All physical parameters of the batch hydrolysis mixture are monitored. The unit is fully equipped with several thermoprobes, an IR temperature measurement device (to provide an independent measurement of the surface temperature of the bulk mass), motor amp transducers (to determine bulk turnover rates and to measure kinetic energy inputs due to mixing), and vacuum transducers. The reactor skid features an integral control panel for mixer control and for operation of the condenser unit, mounted above the reactor. A vacuum pump is used in conjunction with the cooling jacket to cool the mass during the exothermic decrystallization phase. In this phase, the condenser insures that any steam that is produced from the reaction is returned to the vessel. Upon completion of the reaction, the discharge shoot is opened manually to allow flow into the high viscosity, positive displacement pump and out to either the plate or frame filter units. Filtration is used to separate the acid/sugar mixture from the lignin solids available from the 1st and 2nd stage hydrolysis reactions. Filtration is also used to separate the gypsum formed during the acid neutralization step later in the process. The reaction vessel can be discharged into either of the two filtration systems by using the positive displacement pump. This is a vertical tower, single plate pilot unit that has approximately 2 sq. ft of area for filtration. Wash water is added to the vessel which is pressurized with air to facillitate washing of the cake. Using a roll cloth, the resulting cake may be ejected following washing. The filtrate, a mixture of acid and C5 and C6 sugars, is collected in 5-gallon vessels for storage and ultimate feed to the chromatographic separation system. The vertical-leaf plate and frame filter system is also used to process lignin-cake and gypsum co-product streams. Again using the positive displacement pump, the hydrolysate mixture is pumped through the channels in each leaf to fill the inner space between each leaf. When pressure is applied, via a hydraulic piston, the `juice' is pressed from the cake, reducing its overall moisture content to about 50%. In similar manner, the cake is washed, and then discharged into the bin below. The filtrate, here too, a mixture of acid and C5 and C6 sugars, is collected in 5-gallon vessels for storage and ultimate feed to the chromatographic separation system. This unit has 35 sq. ft of filter area and has the capacity to receive the entire contents of the decrystallization/hydrolysis reaction vessel. In commercial scale, the operation of the units is entirely automatic, from fill to discharge to wash to cake ejection.

ß Woodhead Publishing Limited, 2010

152

Bioalcohol production

At this point in the reaction, the chromatographic separation system makes use of unique resins to preferentially retard the flow of one component of the feed stream, and thus separate sugars from acids. Resins may be anionic or cationic and will produce different results, separating the components of the feed into streams, each with a unique concentration and purity. Arkenol has worked with Dow, Mitsubishi, Finex, Rohm & Haas and other suppliers of resin beads to optimize the conditions under which an economical acid/sugar separation may take place. Raffinate streams containing sugar at a typical 12±15% concentration are produced at purities of greater than 98%. Similar systems are found in the sugar industry throughout the world, where they are used to separate sugars (e.g., glucose, fructose) from molasses.

5.6

Unit operations pertinent to the ethanol industry

5.6.1

Feedstock preparation

Facilities engaged in feedstock preparation remove most non-cellulosic material from the waste stream through a series of mechanical and manual sorting. This removes any hazardous, non-processable, or recyclable material, creating a cellulosic-rich feedstock that will be needed for the chemical reaction during hydrolysis. The feedstock preparation system is designed as a traditional materials recovery facility. The MSW is brought in from collection vehicles onto the tipping floor where large items are removed. Conveyors move the remaining MSW to a mechanical separation system where the MSW is sorted by size, and the bulk of inorganic recyclable materials (primarily aluminum, ferrous and non-ferrous metals and plastics) are removed. The MSW then enters a shredder and passes underneath a series of magnetic and electrical separators, which remove the remaining aluminum and ferrous and non-ferrous material. Once the MSW waste stream is mostly clear of metals, glass, plastics and other non-cellulosic material, the feedstock is again shredded to a smaller size, dried, fluffed and moved to the hydrolyzation phase.

5.6.2

Hydrolyzation and cellulose conversion

The solid feedstock is transferred to bioreactor tanks, where it is mixed with concentrated sulfuric acid, at the proper acid/moisture ratio, and heated, thus depolymerizing and liberating the cellulose and hemicelluloses into fermentable sugars. Then the slurry enters a separator to separate the liquid from the solids. The liquid portion proceeds to an ion exclusion unit to separate the sulfuric acid and the sugar. The remaining solids are used as fuel in the plant's steam generator or sold as biofuel to coal fired power generators.

ß Woodhead Publishing Limited, 2010

Dilute and concentrated acid hydrolysis

5.6.3

153

Minimizing microbe inhibiting compounds created by hydrolysis

It is known that hydrolysis of lignocellulosic materials creates a wide range of compounds which are inhibitory to fermentation microorganisms. Based on their origin, the inhibitors are usually divided into three major groups: 1) weak acids such as acetate, 2) furan derivatives such as furfural and 5-hydroxymethylfurfural (5-HMF), and 3) phenolic compounds such as 4-hydroxybenzaldehyde (4-HB), vanillin, and syringaldehyde. These compounds penetrate the cell membrane of fermentation microorganisms and cause inhibition. Their degree of inhibition is different depending on the number of methyl groups that they possess. Also, different microorganisms differ in their tolerance to inhibitors. If the inhibitors are identified and the mechanisms of inhibition elucidated, fermentation can be improved by developing specific detoxification methods, either by choosing an adapted microorganism, or by optimizing the fermentation strategy. To better understand these compounds, we must understand where they came from. Acetate, one of the most abundant organic acids generated through pretreatments, is known to result from the hydrolysis of acetylxylan in hemicellulose. Xylose and other pentose sugars are liberated during the degradation of hemicellulose, and further degradation releases furfural. 5-HMF is the result of hexose degradation. 4-HB, vanillin, and syringaldehyde exert inhibitory effects on microbial ethanol production and are generated by the partial breakdown of lignin through the p-hydroxyphenyl residue, guaiacyl residue, and syringyl residue. Sakai et al. (2007) have investigated ethanol production by Corynebacterium glutamicum R under growth-arrested conditions and concluded that this method showed high tolerance to all organic acid, furan, and phenolic inhibitors. This tolerance was mainly due to growth-arrested conditions. The authors believe that use of C. glutamicum strain R-idhA-pcCRA723 may be developed into an efficient ethanol production method without the need for detoxification step. Mes-Hartree and Sadler (1983) pretreated wheat straw and aspen woodchips by steam explosion and concluded that the inhibitory substances primarily inhibited p-glucosidase component of the cellulose complex of Trichoderma harzianum E58. Furthermore, they state that the furan derivatives furfural and hydroxyl methyl furfural were not inhibitory at concentrations normally found in steam exploded lignocellulosic substrates. Genetically modified microorganisms such as E. coli and T. mattranii are reported to have the potential to utilize all sugars in the hemicellulose hydrolysates to produce ethanol without the need for prior detoxification (Klinke et al., 2004). Dilute acid hydrolysis has more of an effect on hemicellulose and lignin than on crystalline cellulose. A major portion of all inhibitory compounds are created through the over breakdown of hemicellulose and lignin. Since large amounts of

ß Woodhead Publishing Limited, 2010

154

Bioalcohol production

time, heat and pressure are used in dilute acid hydrolysis, it is the worst pretreatment in terms of the inhibitory compounds it creates. Concentrated acid breaks down sugars at a low temperature and pressure quickly and efficiently, and inhibitory compounds are minimal. Some methods for detoxifying inhibitory compounds from hydrolysates that have been investigated are described below (Palmqvist and Hahn-HaÈgerdal, II, 2000): · Treatment with the enzymes peroxidase and laccase, obtained from the ligninolytic fungus Trametes versicolor, has been shown to increase the maximum ethanol productivity of hemicellulose hydrolysate from two to three times. The laccase treatment led to selective and virtually complete removal of phenolic monomers (2.6 g/l in the crude hydrolysate) and phenolic acids. The detoxifying mechanism was suggested to be oxidative polymerization of low molecular weight phenolic compounds. · Filimentous soft-rot fungus Trichoderma reesei has been reported to degrade inhibitors in a hemicellulose hydrolysate, increasing maximum ethanol productivity by approximately three times. Acetic acid, furfural and benzoic acid derivatives were removed from the hydrolysate that was treated with T. reesei. · Roto-evaporation has been used to detoxify hydrolysates. The detoxification was attributed to a decrease in the concentration of acetic acid, furfural and vanillin by 54%, 100% and 29%, respectively, compared to the concentrations in the hydrolysate. · Ethyl acetate extraction removes up to 56% of acetic acid and completely depletes furfural, vanillin, and 4-hydroxybenzoic acid. It has also been shown to increase the glucose consumption rate in a hydrolysate. · When 0.5 mol/l NaHCO3 was used for extraction, in order to further fractionate the inhibitor containing ether extract, neither the ether phase nor the water phase inhibited fermentation. This indicated that the inhibitors are not alkali stable. Detoxification of lignocellulosic hydrolysates by alkali treatment, i.e. increasing the pH to 9 with Ca(OH)2 (over-liming) and readjustment to 5.5 with H2SO4, caused the large formation of precipitate. The ethanol productivity was increased. The detoxifying effect of alkali is due to both the precipitation of toxic components and to the instability of some inhibitors at high pH. A combination of sulphite and overliming has proven to be the most efficient method of detoxification (Palmqvist and Hahn-HaÈgerdal, II, 2000), but requires an additional step, more materials, and more man hours, so the most economical method would be a process that minimizes creation of sugar degradation products combined with a fermentation microorganism that resists the effects of any inhibitory compounds.

ß Woodhead Publishing Limited, 2010

Dilute and concentrated acid hydrolysis

5.6.4

155

Sugar/acid separation technologies

Arkenol Inc. has developed a chromatographic separation system that makes use of unique resins to preferentially retard the flow of one component of the feed stream. Resins may be anionic or cationic and will produce different results; separating the components of the feed into streams with unique concentration and purity. Dow, Mitsubishi, Finex, Rohm & Haas and other suppliers of resin beads optimize the conditions under which an economic acid/sugar separation may take place. Raffinate streams containing sugar at 12±15% concentrations are produced at purities of greater than 98%. The Japanese developed a concentrated sulfuric acid process that was commercialized in 1948 (JGC Corp.). The important feature of their process was the use of membranes to separate the sugar and acid in the product stream. The membrane separation achieved 80% acid recovery. Research and development based on the concentrated sulfuric acid process studied by USDA (which came to be known as the `Peoria Process') picked up again in the United States in the 1980s, particularly at Purdue University (National Renewable Energy Laboratory web site). Among the improvements added by Purdue researchers were: (1) recycling of dilute acid from the hydrolysis step for pretreatment and (2) improved recycling of sulfuric acid. Minimizing the use of sulfuric acid and recycling the acid cost-effectively are critical factors in the economic feasibility of the process. BlueFire Inc. has developed a technology that uses commercially available ion exchange resins to separate the acid from sugar. The separated sulfuric acid is recirculated and reconcentrated to the level required by the decrystallization and hydrolysis steps. The small quantity of acid left in the sugar solution is neutralized with lime to make hydrated gypsum, CaSO4
2H2O, an insoluble precipitate which is readily separated from the sugar solution. Izumi Inc. uses a sulfuric acid recovery system that is higher than 97% with reconcentration from 18% to 75%. Izumi technology is based on commercial membrane distillation and purification system supplied by Mitsui, which achieves significant operating cost savings over molecular sieve technology.

5.7

Future trends

Utilizing lignocellulosic biomass as a feedstock is seen as the next step towards significantly expanding ethanol production capacity. The issues associated with acid hydrolysis are still significant and diverse, depending on many factors such as biomass species, glucose yield, expensive construction material, acid recovery, toxicity, reactor design, reaction conditions (Lee et al., 1999; Taherzadeh and Karimi, 2007). The problems of biomass acid hydrolysis are regional, and should be solved independently. The solutions may come from the following aspects: development of various reactor configurations, optimization

ß Woodhead Publishing Limited, 2010

156

Bioalcohol production

of hydrolysis parameters, treatment of liquid feed, strains improvement, and development of novel fermentation technology. Another problem from downstream processing is utilization of lignin. In the near future, large quantities of lignin residue material will be available from biomass-to-ethanol processes and other biorefineries and associated processes. Because it is an inexpensive feedstock, lignin is commonly used by combustion to provide heat and/or power. In general, however, using lignin for recovery of heat and power is not as economical because it can be used as a source of aromatic compounds. Therefore, it is very important to explore emerging technologies such as pyrolysis and gasification of lignin to produce multiple products and obtain higher values. For example, liquefaction of lignin extracted from aspen wood resulted in a 90% yield of bio-oil, representing a significant added value (Zhang et al., 2008).

5.8

Sources of further information and advice

· US Department of Energy, Office of the Biomass Program http://www.eere.energy.gov/biomass · USDA Biofuels Research Program http://www.nal.usda.gov/ttic/biofuels.htm · National Bioenergy Center http://www1.eere.energy.gov/biomass/bioenergy_center.html · National Renewable Energy Laboratory http://www.nrel.gov/biomass/ · Japanese Concentrated Acid Hydrolysis Business Plan http://www.igloo.org/libraryservices/download-nocache/Library/subjects/ sciencet/sciencet~1/energy/abioetha · JGC Corporation http://www.jgc.co.jp/en/01newsinfo/2006/release/20060620.html · Energy Efficiency and Renewable Energy (EERE) http://www1.eere.energy.gov/biomass/printable_versions/concentrated_ acid.html · Izumi Biorefinery http://bluefireethanol.com/images/IZUMI_Status_2004_for_BlueFire_ 051606.pdf · BlueFire Ethanol http://bluefireethanol.com/ · Arkenol http://www.arkenol.com/Arkenol%20Inc/tech01.html · Masada Resource Group http://masadaonline.com/ · Arkenol Thechnology Valuation http://bluefireethanol.com/pdf/Booz-Allen_Hamilton_Valuation_Report_ for_Arkenol_051606.pdf

ß Woodhead Publishing Limited, 2010

Dilute and concentrated acid hydrolysis

5.9

157

References and further reading

Allen, S. G., Kam, L. C., Zemann, A. J., Antal, M. J., 1996. Fractionation of sugar cane with hot compressed liquid water. Ind Eng Chem Res 35: 2709±2715. Badger, P. C. 2002. Ethanol from cellulose: A general review. In: J. Janick and A. Whipkey (eds), Trends in New Crops and New Uses. ASHS Press, Alexandria, VA, pp. 17±21. Bobleter, O., 1994. Hydrothermal degradation of polymers derived from plants. Prog Polym Sci 19: 797±841. Dionex Application Note 363, Using Accelerated Solvent Extraction (ASE) in Alternative Fuel Research. Garrote, G., Dominguez, H., Parajo, J. C., 1999. Mild autohydrolysis: an environmentally friendly technology for xylooligosaccharide production from wood. J Chem Technol Biotechnol 74 (11): 1101±1109. Heitz, M., Capek-Menard, E., Korberle, P.G., Gange, J., Chornet, E., Overend, R. P., Taylor, J. D., Yu, E., 1991. Fractionation of populus tremuloides at the pilot plant scale: optimization of steam pretreatment conditions using the STAKE II technology. Biores Technol 25: 23±32. Hsu, T.-A., 1996. Pretreatment of biomass. In: Wyman, C.E. (ed.), Handbook on Bioethanol, Production and Utilization. Taylor & Francis, Washington, DC. Kinke, H. B., Thomsen, A. B., Ahring, B. K., 2004. Inhibition of ethanol-producing yeast and bacteria by degradation products produced during pre-treatment of biomass. Appl Microbiol Biotechnol 66: 10±26. Knappert, D. R., Grethlein, H. E., Converse, A. O., 1981. Partial acid hydrolysis of poplar wood as a pretreatment for enzymatic hydrolysis. Biotechnol Bioeng 11: 67±77. Lee, Y.Y., Iyer, P., Torget, R.W., 1999. Dilute-acid hydrolysis of lignocellulosic biomass. Adv Biochem Eng Biotechnol 65: 93. Liu, C., Wyman, C.E., 2003. The effect of flow rate of compressed hot water on xylan, lignin, and total mass removal from corn stover. Ind Eng Chem Res 42(21): 5409± 5416. Mackie, K. L., Brownell, H. H., West, K. L., Saddler, J. N., 1985. Effect of sulfur dioxide and sulphuric acid on steam explosion of aspenwood. J Wood Chem Technol 5(3): 405±425. Mes-Hartree, M., Saddler, J. N., 1983. The nature of inhibitory materials present in pretreated lignocellulosic substrates which inhibit the enzymatic hydrolysis of cellulose. Biotechnol Lett 5 (8): 531±536. Palmqvist, E., Hahn-HaÈgerdal, B., 2000. Fermentation of lignocellulosic hydrolysates. II: inhibitors and mechanisms of inhibition. Biores Technol 74(1): 25±33. Parajo, J. C., Garrote, G., Cruz, J. M., Dominguez, H., 2004. Production of xylooligo saccharides by autohydrolysis of lignocellulosic materials. Trends Food Sci Technol 15 (3±4): 115±120. Saddler, J. N., Ramos, L. P., Breuil, C., 1993. Steam pretreatment of lignocellulosic residues. In: Saddler, J. N. (ed.), Bioconversion of Forest and Agricultural Plant Residues. CAB International, Oxford, pp. 73±92. Sakai, S., Tsuchida, Y., Okino, S., Ichihashi, O., Kawaguchi, H., Watanabe, T., Inui, M., Yukawa, H., 2007. Effect of lignocelluloses-derived inhibitors on growth of and ethanol production by growth-arrested Corynebacterium glutamicum R. Appl Environ Microbiol 73(7): 2349±2353. Taherzadeh, M.J., Karimi, K., 2007. Acid-based hydrolysis processes for ethanol from lignocellulosic materials: a review. BioResources 2(3): 472±499.

ß Woodhead Publishing Limited, 2010

158

Bioalcohol production

Torget, R., Hatzis, C., Hayward, T. K., Hsu, T.-A., Philippidis, G. P., 1996. Optimization of reverse-flow, two-temperature, dilute-acid pretreatment to enhance biomass conversion to ethanol. Appl Biochem Biotechnol 57/58: 85±101. Torget, R. W., Kidam, K. L., Hsu, T.-A., Philippidis, G. P., Wyman, C. E., 1998. Prehydrolysis of lignocellulose. US Patent No. 5,705,369. Torget, R. W., Nagle, N., Jennings, E., Ibsen, K., Elander, R., 1999. Novel pilot scale reactor for the aqueous fractionation of hardwood for the improved production of ethanol. In: 21st Symposium on Biotechnology for Fuels and Chemicals. Wyman, C. E., 1999. Biomass ethanol: technical progress, opportunities, and commercial challenges. Ann Rev Energy Environ 24: 189±226. Yang, B., Wyman, C. E., 2004. Effect of xylan and lignin removal by batch and flow through pretreatment on the enzymatic digestibility of corn stover cellulose. Biotechnol Bioeng 86(1): 88±95. Zhang, B., Huang, H., Ramaswamy, S., 2008. Reaction kinetics of the hydrothermal treatment of lignin. Appl Biochem Biotechnol 147: 119±131. Zhang, B., Shahbazi, A., Wang, L., Diallo, O., 2009. Biofuels production with cattails from constructed wetlands. In: 31st Symposium on Biotechnology for Fuels and Chemicals, San Francisco, CA. Zhang, B., Shahbazi, A., Wang, L., Diallo, O., Whitmore, A., 2010. Dilute acid and hot water pretreatment of wheat straw using a solvent extractor. In: 32nd Symposium on Biotechnology for Fuels and Chemicals, Clearwater Beach, FL. Zhang, Y.-H.P., Ding, S.-Y., Mielenz, J. R., Cui, J.-.B., Elander, R. T., Laser, M., Himmel, M. E., McMillan, J. D., Lynd, L. R., 2007. Fractionating recalcitrant lignocellulose at modest reaction conditions. Biotechnol Bioeng 97(2): 214±223. Zhu, Y., Lee, Y.Y., Elander, R.T., 2004. Dilute-acid pretreatment of corn stover using a high-solids percolation reactor. Appl Biochem Biotechnol 117(2): 103±114.

ß Woodhead Publishing Limited, 2010

6

Enzymatic hydrolysis of lignocellulosic biomass M . B A L L E S T E R O S , CIEMAT, Spain

Abstract: Enzymatic hydrolysis technologies for producing ethanol from lignocellulosic materials offer much greater scope for advancements than those using acid catalysts. However, technical barriers for enzymatic hydrolysis to cost-effectively hydrolyze cellulose to glucose must be overcome. This chapter begins by dicussing the mechanisms of enzymatic cellulose hydrolysis and the synergistic interactions of cellulolytic enzyme components. It then reviews factors affecting the hydrolysis efficiency and research efforts to enhance saccharification efficiencies. Key words: biomass, cellulose, cellulases, enzymatic hydrolysis, enzymatic efficiency.

6.1

Introduction

Technology to produce ethanol from lignocellulosic biomass possesses great potential due to the widespread availability, large quantity, and relatively low cost of cellulosic materials. However, although progress in research during the past decades has shown that producing ethanol from cellulose processes is technically feasible, cost-effective technologies have not yet been achieved. Lignocellulosic feedstocks typically contain 55±76% dry weight of carbohydrates that are polymers of five and six carbon sugar units (Wyman, 1994, 1996). These carbohydrate polymers must be broken down to their respective monomers before microorganisms can complete the conversion to ethanol or other products. The biomass hydrolysis (the depolymerization of the biomass polysaccharides to fermentable sugars) is a crucial step in the overall process due to its relatively large contribution to the total cost of producing ethanol from lignocellulosic substrates (Nieves et al., 1998; Himmel et al., 1999; Galbe and Zacchi, 2002). Although cellulosic materials can be hydrolyzed using acid or enzymes, the hydrolysis must be performed via environmentally friendly and economically feasible technologies (Lynd et al., 2005). Acid hydrolysis of cellulose requires a relatively high temperature and produces degradation of glucose to hydroxymethylfurfural and other products, thus reducing the glucose yield that can be obtained by this process. The enzyme based application is advantageous over chemical treatments due to its higher conversion efficiency and lower byproduct production, the use of more moderate and non-corrosive operating

ß Woodhead Publishing Limited, 2010

160

Bioalcohol production

conditions and low process energy requirements (Bon and Ferrara, 2007). The National Renewable Energy Laboratory, in the USA, has estimated that future cost reduction could be four times greater for the enzyme process than for the dilute acid process (NREL, 1999). Nevertheless, biomass biodegradation is a highly complex multi-enzymatic process. The development of an efficient conversion process to obtain free sugars from cellulosic substrates by enzymatic hydrolysis remains a challenging goal since, due to the tough crystalline structure, the effectiveness of enzymes currently available is too low and several days are required to achieve good glucose yields. Enzymes are relative newcomers with respect to biomass-to-ethanol technologies. While the chemistry of sugar production from wood has almost 100 years of process development, enzymes for biomass hydrolysis only have 50 years of serious research efforts. The understanding of enzymatic hydrolysis of cellulose began in the South Pacific during World War II where a fungus that broke down cotton clothing and equipment was discovered. This organism, Trichoderma reesei, was found to produce cellulase enzymes. Since then, although many generations of cellulases have been developed through genetic modifications of the fungus strain, an effective cellulase enzyme system is not commercially available yet. Only recently, as a result of intensive research, costeffective catalysts have begun to emerge (Zhang et al., 2006). Reduction in the cost of cellulases can be achieved only by (i) concerted efforts which address several aspects of enzyme with improved hydrolytic characteristics such as binding affinity, catalytic activity and thermostability, and (ii) by the development of efficient technologies for saccharification which includes the use of better `enzyme cocktails' and conditions for hydrolysis. In this chapter the challenges for cost-effective cellulose depolimerization into fermentable sugars are addresed. The mechanisms of cellulose hydrolysis and factors affecting the hidrolysis efficiency are reviewed. Possible future advances in the area of cellulosic bioconversion are also discussed.

6.2

Enzymatic hydrolysis mechanism

In nature, lignocellulose is degraded by a battery of hydrolytic and oxidative enzymes produced by a variety of fungi and bacteria that are able to synergistically degrade the cellulose, hemicellulose and lignin (PeÂrez et al., 2002). Aerobic cellulose degraders, both bacterial and fungal, utilize cellulose through the production of substantial amounts of extracellular cellulase enzymes that are freely recoverable from culture supernatants (Wojtczak et al., 1987, Valaskova and Baldrian, 2006). Anaerobic bacteria degrade cellulose primarily via complexed cellulase systems called cellulosome that consists of multiple subunits that interact with each other synergistically and degrade cellulosic substrate (Bayer et al., 2004).

ß Woodhead Publishing Limited, 2010

Enzymatic hydrolysis of lignocellulosic biomass

161

6.1 Mechanism for cellulose breakdown.

Cellulases represent the primary family needed to depolymerize lignocellulosic substrates. Cellulose complex consists of endoglucanases (EG), cellobiohydrolases (CBH, also called exoglucanases), and -glucosidases (BG). EGs cleave glycosidic bonds within cellulose microfibrils, acting preferentially at amorphous (accessible) cellulose regions (Teeri and Koivula, 1995; Teeri, 1997). EGs fragment cellulose chains to generate reactive ends for CBHs, which act `processively' to degrade cellulose, including crystalline cellulose, from either the reducing (CBH I) or non-reducing (CBH II) ends, to generate mainly cellobiose (Koivula et al., 1998). These fundamentally different catalytic mechanisms allow different types of cellulases to interact synergistically (see Fig. 6.1). Cellobiose, at high concentrations, inhibits CBH activity. Thus BG, which converts cellobiose into glucose, is often required to reduce end-product inhibition in conditions where cellobiose accumulates. Although the structure and function of cellulase systems have been made the subject of intensive research over the last 30 years, obtaining an efficient enzymatic hydrolysis of cellulose remains a difficult goal. The heterogeneous and insoluble nature of lignocellulosc biomass represents a challenge for cellulase systems. Native cellulose undergoing an attack by cellulase exhibits extensive changes in physical properties prior to producing a measurable quantity of reducing sugar (Koivula et al., 1998). These changes include fragmentation, swelling, considerable loss in tensile strength, transverse cracking and lowering of the degree of polymerization (Fan et al., 1980). A general feature of most cellulases is a modular structure often including both catalytic (CD) and carbohydrate-binding modules (CBMs). The CBM binding to the cellulose surface, presumably facilitates cellulose hydrolysis by bringing the catalytic domain in close proximity to the insoluble cellulose (Teeri et al., 1998). Thus, the rate and extent of the enzymatic hydrolysis of lignocellulosic substrates are influenced not only by the catalytic performance of the enzymes but also by substrate characteristics.

ß Woodhead Publishing Limited, 2010

162

Bioalcohol production

Substrate features affecting enzymatic hydrolysis, in addition to concentration, include degree of polymerization, crystallinity, accesible area and the presence of hemicellulose and lignin. To make the enzymatic cellulose conversion more difficult, these factors depend on the particular substrate being used and, in addition, the substrate characteristics change as the reaction progresses. Therefore, it is important to enhance knowledge in biological and chemical features of plant biomass and translate these new findings into the conversion processes (Himmel et al., 2007). Currently, no definitive model of the cell wall has been established, particularly one that relates the cell wall composition to its mechanical properties. The architectural features of the primary cell wall is a fundamental framework of cellulose and cross-linking glucans are embedded in a second matrix of pectin polysaccharides (homogalacturonic acid regions with neutral sugar side chains). An additional independent network consists of the structural proteins and lignin (a complex network of aromatic compounds called phenylpropainoid) (Mutwil et al., 2008). Cellulose is insoluble and, despite its homogeneous chemical composition, exists in a number of both amorphous and crystalline regions where water is almost completely excluded. Cellulose and pectin networks are largely independent or only interact weakly through hydrogen bonding. Hemicellulose, a heterogeneous polymer built by pentoses (D-xylose, D-arabinose), hexoses (Dmannose, D-glucose, D-galactose) and sugars acids, interacts much more strongly with cellulose and makes the network more rigid (Laureano-PeÂrez et al., 2005). Lignin is covalently bonded to polysaccharides, thus reducing the accessible surface area of cellulose. The robust structure of lignocellulosic biomass makes cellulose hydrolysis more difficult than starch. In general, cellulases act slowly, and high enzyme loadings are currently needed to achieve reasonable rates and yields (Himmel et al., 1999; Wooley et al., 1999). In any process to obtain ethanol from lignocellulosic biomass a pretreatment step is needed to reduce recalcitrance of cellulose to hydrolysis. Generally, pretreatment increases enzymatic digestibility breaking down the macroscopic rigidity of biomass and decreasing the physical barriers to mass transport by depolymerizing and solublizing hemicellulose to monosacharrides and oligosaccharides (Himmel et al., 2007), breaking the lignin seal (Zhu et al., 2008) or disrupting the crystalline structure of cellulose (Fan et al., 1981). Cellulosic substrates resulting from pretreatment processes typically contain residual hemicellulose, that impedes access to cellulase components, and lignin, to which many cellulase components bind. Therefore, enzymatic activities distinct from those involved in cellulose degradation may be required for more complete cellulose hydrolysis. Commercial cellulase preparations generally contain other enzymatic activivities besides cellulases that improve cellulose hydrolysis (Berlin et al., 2006)

ß Woodhead Publishing Limited, 2010

Enzymatic hydrolysis of lignocellulosic biomass

6.3

163

Relative saccharification efficiencies

During cellulose hydrolysis, the combined action of endoglucanases and cellobiohydrolases produce changes in the solid substrate features (Zhang and Lynd, 2004) and in cellulose accessibility (Boisset et al., 2001; Wang et al., 2003) that results in rapid changes in hydrolysis rates. The rate and extent of the enzymatic hydrolysis of lignocellulosic substrates are highly influenced by enzyme loadings, hydrolysis time and physical and chemical characteristics of the pretreated substrate. Typically, the rate of cellulose hydrolysis by enzymes decreases rapidly with conversion, leading to decreased yields, long processing times, and high enzyme usage. During the course of enzymatic hydrolysis of a pretreated lignocellulosic substrate, after an initial rapid phase, the hydrolysis rate decreases rapidly. The reaction rate continuously declines as the conversion percentage of cellulose increases and, in most cases, as a result of an incomplete hydrolysis, a recalcitrant cellulosic residue remains (see Fig. 6.2). Many hypotheses have been presented to explain this observation, including enzyme inactivation due to thermal effects (Eriksson et al., 2002a,b), inhibition by hydrolysis products (Gusakov et al., 1985, 1996; Eriksson et al., 2002b; Holtzapple et al., 1990), formation of an inactive enzyme-substrate (lignin) complex cellulase (Sutcliffe and Saddler, 1986; Ooshima et al., 1990; Chang and Holtzapple, 2000; Draude et al., 2001; Converse et al., 2004), substrate transformation into a less digestible form (Zhang et al., 1999), and/or the heterogeneous structure of the

6.2 Time course of enzymatic hydrolysis of pretreated wheat straw at different enzyme concentrations: (u) 5 UPF/g cellulose, (n) 15 FPU/g cellulose and (s) 25 FPU/g cellulose (Garc|¨ a-Aparicio et al., 2004).

ß Woodhead Publishing Limited, 2010

164

Bioalcohol production

6.3 Effect of substrate concentration on enzymatic hydrolysis of pretreated barley straw (Garc|¨ a-Aparicio et al., 2007).

substrate (Nidetzky et al., 1994; Zhang et al., 1999). A number of studies have reported that the slow down in rate for cellulose hydrolysis is not due to a loss in cellulose reactivity but to the fact that enzyme performance is slowed down by obstacles that interfere with their path or by a loss in activity or processitivity, making them less effective (Breyer and Matthews, 2001; Zhou and Ingram, 2001). The specific rate of cellulose hydrolysis, at a given percentage of conversion, also decreases with increasing cellulose concentration (GarcõÂa-Aparicio et al., 2004; Cara et al., 2007; Hodge et al., 2008) (see Fig. 6.3). However, the economy of fuel ethanol industry based on lignocellulosic biomass is improved at higher substrate loadings due to reduced capital cost for smaller reaction vessels, decreased water usage and the higher sugar concentration after saccharification (Hodge et al., 2008). Hydrolysis process at high solid content will increase the ethanol concentration and facilitate the downstream process and product recovery, which have a significant effect on both capital and operating costs due to energy expenditure for distillation (Mohagheghi et al., 1992). Performing the saccharification reaction at high levels of insoluble solids creates a number of process-related problems associated with enzyme-substrate mixing and cellulase effectiveness such as sugar inhibition (Xiao et al., 2004). At increased levels of solids, the ability of the enzyme to reach the reaction site is reduced and, at the same time, sugar inhibition is higher due to the increasing difficulty in diffusion of sugars away from catalytic site.

6.4

Factors affecting hydrolysis efficiency

Although the factors that influence the efficiency of hydrolysis of lignocellulose have been revised by different authors (Fan et al., 1980, 1981; Wald et al., 1984;

ß Woodhead Publishing Limited, 2010

Enzymatic hydrolysis of lignocellulosic biomass

165

Table 6.1 Summary of factors affecting enzymatic hydrolysis of lignocelluloses Factor

References

Substrate features Lignin content Hemicellulose content Particle size Specific surface area Cellulose crystallinity and degree of polymerization Nature of enzyme Cellulase activity Reduction of costs Cocktail composition

Chang and Holtzapple (2000), Chandra et al. (2007) Tanaka et al. (1988), Varga et al. (2004) Draude et al. (2001) Grethlein and Converse (1991) Chang and Holtzapple (2000), Zhang and Lynd (2004) Mcfarland et al. (2007) Reeta et al. (2007), Tolan (2008) Rosgaard et al. (2007)

Zhu et al., 2008), to achieve a complete, rapid and efficient conversion of cellulosic substrates remains a challenging goal. Factors affecting hydrolysis efficiency can be classified into three main groups: the structure of the substrate, the nature of the enzymes and the enzyme±substrate interactions. A summary of factors that affect cellulose hydrolysis are shown in Table 6.1.

6.4.1

Structure of the substrate

The rate and extent of cellulose hydrolysis by cellulase enzymes are influenced by two main physical and chemical factors of the substrate: (i) the crystallinity of cellulose (Chang and Holtzapple, 2000) and its degree of polymerization (Zhang and Lynd, 2004) that reduce enzyme effectiveness, and (ii) the matrix polysaccharides and lignin coat the cellulose fibril that act as a physical barrier preventing enzymes from reaching the cellulose. One of the reasons for the difficulty in cellulose hydrolysis is attributed to the dual nature of the substrate, since the accessibility of the crystalline and amorphous regions is different. Although some discrepancies have been reported with regard to the effect of crystallinity of cellulose on enzymatic hydrolysis performance, it is broadly accepted that highly crystalline cellulose is less accessible to cellulase attack than the amorphous cellulose. It has been suggested that amorphous cellulose is hydrolyzed at the beginning of the process, resulting in an accumulation of crystalline regions and, therefore, the substrate recalcitrance increased as the enzymatic hydrolysis progresses (Puri, 1984, Bertran and Dale, 1986). Different authors have reported that biomass digestibility is enhanced with increasing lignin removal (Chang and Holtzapple, 2000; Draude et al., 2001).

ß Woodhead Publishing Limited, 2010

166

Bioalcohol production

The lignin content and type of lignin have a significant effect on the hydrolysis of lignocellulosic substrates as lignin acts as a physical barrier to the accesibility of cellulases (Mooney et al., 1998). Delignification causes biomass swelling and disruption of lignin structure and subsequently, an increase in internal surface area and pore volume (Fan et al., 1980). Lignin has been also implicated as a competitive cellulase adsorbent, which reduces the amount of cellulase available to catalyze cellulose hydrolysis (Ooshima et al., 1990). Accordingly, delignification reduces irreversible and non-productive adsorption of cellulase on lignin. It has been also suggested that residual lignin blocks the progress of cellulase down the cellulose chain (Zhang and Lynd, 2004). The role of the hemicellulose in preventing cellulose hydrolysis is less obvious although there is good evidence to support the action of hemicellulose as a barrier resticting access to cellulases (Tanaka et al., 1988; Grethlein and Converse, 1991; Varga et al., 2004). Many authors have demonstrated that solubilization of hemicellulose during pretreatment facilitates subsequent È hgren et al., 2007; PeÂrez et al., 2007; Musatto et hydrolysis by cellulases (O al., 2008). Further evidence has been shown employing cellulases in combination with hemicellulases to hydrolyze preteated substrates (Berlin et al., 2006; GarcõÂa-Aparicio et al., 2007; Mussatto et al., 2008).

6.4.2

Nature of the enzymes

Although a wide variety of bacteria and fungi produce cellulolytic enzymes able to hydrolyze cellulose (Das et al., 2007; Zeng et al., 2007), relatively few species produce high levels of extracellular cellulase capable of solubilizing crystalline cellulose extensively (Bhat and Bhat, 1997). Cellulases used for current industrial applications are mainly fungal, primarily due to efficiencies in fungal enzyme secretion. Microorganisms of the genus Trichoderma produce relatively large quantities of endo- -glucanase and exo- -glucanase, but only low levels of -glucosidase, while those of the genus Aspergillus produce relatively large quantities of endo- -glucanase and -glucosidase with low levels of exo- -glucanase production (Madamwar and Patel, 1992). Although commercially available T. reesei cellulase preparations, such as Novozymes' Celluclast 1.5 L and Genencor International's Spezyme, are widely used in applications such us detergent and textile industries, the purchase price is too high for the biomass to ethanol industry. Production cost of cellulases may be brought down by multifaceted approaches which includes the production of more efficient enzymes and mixtures of enzymes, the use of cheap lignocellulosic substrates for production of the enzyme (Reeta et al., 2007) and inhouse enzyme production (Tolan, 2008). Among the recent investments aiming to reduce the biocatalyst cost, it is worth mentioning the US project developed by NREL and Genencor and Novozymes (2000±2003) to reduce enzyme cost tenfold by means of finding

ß Woodhead Publishing Limited, 2010

Enzymatic hydrolysis of lignocellulosic biomass

167

more efficient enzymes and mixtures of enzymes through a combination of diversity mining, protein engineering and system optimization. Although it is reported that the enzyme cost has been lowered to 8±10 cents per liter (Mcfarland et al., 2007), further decrease is still necessary to reduce enzyme costs and, by extension, to deploy the biomass to ethanol technology. It is also necessary to reduce the enzyme-dose level required for biomass saccharification by improving the specific performance of the Trichoderma reesei mix. Although Trichoderma reesei is one of the best-studied fungal cellulolytic systems (Gruber et al., 1990), its complete inventory of enzyme associated with lignocellulosic biomass hydrolysis has not been elucidated. In the presence of available cellulose, T. reesei secretes a broad spectrum of cellulolytic enzyme activities, most notably cellobiohydrolases (CBHs) and endo-1,4-glucanases (EGs). In addition, it secretes over 30 other minor enzyme components that could play an important, and still undiscovered, role in facilitating the breakdown of crystalline cellulosic substrates. Recently, it has been demonstrated that the optimal ratios of cellulases secreted by T. reesei in submerged fermentation usually differs from the optimum cellulolytic activities needed to hydrolyze pretreated biomass (Rosgaard et al., 2007), indicating that tailor-made enzyme mixtures, with a careful combination of monocomponent enzymes, can promote both a more efficient enzymatic hydrolysis of lignocellulose and a more rational utilization of enzymes.

6.4.3

Enzyme±substrate interactions

Enzymatic hydrolysis of cellulose involves soluble enzymes working on insoluble substrate. It should be considered that hydrolysis of cellulose is a surface phenomenon and the complexity of the heterogeneous enzyme±substrate system is remarkable. Before the reaction itself, the enzyme has to be bound to the substrate, i.e. the process requires the adsorption of the biocatalyst onto the insoluble substrate prior to the hydrolysis (Mansfield et al., 1999). The susceptibility of the cellulose to enzymatic hydrolysis is largely determined by the accessibility of the binding sites to the enzyme, which determines the subsequent adsorption of enzyme onto the solid substrate. Hydrolysis is considered to be a slow process because the crystalline structure of the cellulose and the number of active binding sites available is limited (Himmel et al., 2007). Many of the fungal cellulases are modular proteins that contain a catalytic domain (CD) and a carbohydrate-binding module (CBM) (Henrissat and Bairoch, 1996; Bourne and Henrissat, 2001). CBMs in most fungal cellulases are located at either end of the catalytic domain and are connected to CD by an often heavily glucosylated linker (Ohmiya et al., 1997). The efficiency of the cellulase is enhanced by the presence of the CBM and it is clearly correlated with better hydrolytic activity towards insoluble cellulose (Linder and Teeri, 1997). CBMs allow cellulases to act on crystalline cellulose

ß Woodhead Publishing Limited, 2010

168

Bioalcohol production

by destabilizing the hydrogen bond structure of cellulose, making the polysaccharide chains more accessible to the catalytic domain (Quentin et al., 2003). Only by cooperation with non-catalytic specific binding modules are the enzymes able to disrupt the crystal surface at the solid-liquid interphase, to make single cellulose fibers accessible for hydrolysis. In general, it is known that a CBM in a cellulase molecule enhances the hydrolytic activity of a catalytic domain adjacent to the CBM by increasing the enzyme concentration on the surface of an insoluble substrate (Karita et al., 1996) or by supplying the catalytic module with a more easily degradable substrate, i.e., amorphous cellulose (Xiao et al., 2001; Linder and Teeri, 1997).

6.5

Methods to improve enzymatic hydrolysis

Significant enhancements in enzymatic hydrolysis yield and reductions in the cost of enzymes, and their application, are expected as a result of the extensive research developed in the last decades. Improvements in cellulase performance, the creation of an appropriate pretreated substrate amenable to hydrolysis, and an adequate process configuration are vital to increase overall sugar yields in lignocellulosic conversion.

6.5.1

Improvements in cellulase performance

Extensive research is being carried out on improving the hydrolytic efficiency of cellulolytic enzymes through cellulase engineering based on rational design or directed evolution for each cellulase component enzyme, as well as on the reconstitution of cellulase components (GonzaÂlez-Blasco et al., 2000; Cherry and Fidanstef, 2003; Wang et al., 2005; Johannes and Zhao, 2006). This involves screening of new enzyme-producing microorganisms, random mutagenesis of fungal strains and genetic engineering of individual enzymes (Schulein, 2000; Wilson, 2004). Among all possible strategies, optimization of the fundamental properties of the cellulases such as thermostability and end-product inhibition are very important in industrial applications. Recent progress in increasing stability of cellulases towards temperature has been reported by means of (i) screening naturally occurring proteins with thermostable properties (Bock et al., 1998); (ii) using molecular biology techniques to modify, and thus improve the thermostability of natural proteins (Cheng et al., 2002); and (iii) employing additives to protect proteins from high temperatures (George et al., 2001). Enzymes improvements to minimize end-product inhibition and alleviate non-specific binding of enzyme, and therefore reducing non-specific enzyme inhibition, are also crucial to increase reaction rates. Optimizing the cellulase mixture by altering components and their ratios is gaining increasing recognition as a way to enhance enzymatic hydrolysis.

ß Woodhead Publishing Limited, 2010

Enzymatic hydrolysis of lignocellulosic biomass

169

Several studies have reported that the synergism between cellulases is tightly connected with the type and ratio of the individual enzyme (Gusakov and Sinitsyn, 1992; Kim et al., 1998). The hydrolytic efficiency of a multi-enzyme mixture in the process of lignocellulose saccharification depends both on properties of individual enzymes and their ratio in the multi-enzyme cocktail (Berlin et al., 2007; Boisset et al., 2001). The multi-enzyme cocktail secreted by selected fungal strains may be non-optimal in a biotechnological process. Optimizing this mixture, therefore, becomes very important and should be modulated to increase the lignocellulose conversion.

6.5.2

Pretreated substrate characteristics

Although the properties of cellulase enzyme complex have a significant effect on how effectively a lignocellulosic substrate is hydrolyzed, the intrinsic structure/ composition of the substrate itself affects enzymatic hydrolyisis yields in a great extent (Chandra et al., 2007). Recent advances in knowing cell-wall chemical and physical structures, and their changes in the pretreatment step, will help to produce an `optimal substrate' for a more efficient breakdown by cellulases. High glucose concentration after hydrolysis is preferable for the fermentation process, thereby decreasing the energy demand in the subsequent distillation step. Therefore, to enhance the cellulose content of substrate, the non-cellulose components need to be decreased. Pretreatment is crucial for ensuring good sugar yields from enzymatic hydrolysis. Further research is needed to develop new processes to improve the recovery of hemicellulose and lignin after the extraction of cellulose in the pretreatment step. To reach the goal for producing cost-competitive ethanol from biomass, new findings from plant science and carbohydate chemistry must be translated and integrated into the conversion processes (Himmel et al., 2007). Other approaches to improve enzymatic hydrolysis efficiencies come from innovative process configurations such as immobilized cells and/or enzymes and fluidized-bed reactors (Padukone, 1996). The presence of solids in the hydrolysis vessel clearly presents a formidable challenge in efficient system design and incorporation of new bioreactor designs. Biomass to ethanol processing technology has exhibited a trend toward increasing consolidation over time. The first application of enzymatic hydrolysis was used in separate hydrolysis and fermentation (SHF) steps. In the SHF process, the substrate is pretreated and then subsequently hydrolyzed causing, especially at high solid concentration, end-product inhibition of enzymes. To overcome this problem, the two process steps can be performed simultaneously, i.e. simultaneous saccharification and fermentation (SSF). SSF has been widely demosnstrated to have significant economic advantages over SHF (Takagi et al., 1977; SoÈderstroÈm, 2005). Combining the two process steps also results in a lower capital cost, and the fact that the ethanol concentration is higher during

ß Woodhead Publishing Limited, 2010

170

Bioalcohol production

SSF than SHF reduces the risk of contamination (Wyman et al., 1992). The disadvantage, however, is that the cellulase enzyme and fermentation organism had to operate under the same conditions, decreasing the sugar and ethanol yields. The use of thermotolerant microorganisms in the SSF process could help to overcome this barrier. The fermentable sugars can also be separated by employing membrane reactors using ultrafiltration. Thus, the reaction can be significantly accelerated by continuous removal of the produced sugars. To improve the yield and rate of the enzymatic hydrolysis, membrane reactors with different configurations have been investigated (Ghan et al., 2002; Ohlson et al., 1984a,b; Yang et al., 2006). The membrane reactor, utilizing an ultrafiltration membrane with an appropriate molecular-weight cutoff, keeps the larger components in the reactor, while lowmolecular-weight molecules, such as glucose and cellobiose, pass through the membrane and leave the reactor as permeate. Operating enzymatic hydrolysis at higher initial substrate concentrations is also technically challenging because the initial viscosity of the material is very high, which makes mixing difficult and inadequate and the power consumption in stirred tank reactors becomes elevated. Moreover, the rheological properties of media change drastically during the course of the hydrolysis process. One interesting approach to promote intensive mass transfer, and therefore enhance enzymatic cellulose hydrolysis, is to stir the reaction mixture with ferromagnetic particles (Gusakov et al., 1996). However, the relatively high power consumption represents a potential drawback that may hinder the practical application of these advanced biorreactors. Future developments in bioreactor design could attempt to address some of the limitations of enzymatic hydrolysis of cellulose. However, progress in reactor design will require advances in understanding the fundamentals of enzymatic hydrolysis of cellulosic material.

6.6

Future trends

The technologies for converting lignocellulosic matter into ethanol via enzymatic hydrolyisis are becoming increasingly cost competitive as the technology advances. However, despite their promise, enzymatic cellulosic conversion technologies are still probably at least 8±10 years away from significant commercial expansion. The core technological barrier is cellulosic biomass recalcitrance to enzymatic attack. Despite the current information available on cellulase systems and on the structure of plant cell walls, application of this knowledge to cellulose degradation has met with limited success. This fact may be attributed to at least two factors: (i) the inherent complexity and heterogeneity of lignocellulosic materials, and (ii) the current limited understanding of the basic hydrolysis processes. Therefore, it is necessary to advance the knownledge of the molecular mechanisms underlying cellulose degradation and the

ß Woodhead Publishing Limited, 2010

Enzymatic hydrolysis of lignocellulosic biomass

171

development of new and superior enzymes with characteristics that comply with industrial requirements in terms of performance and cost. In terms of future prospects, to understand plant cell-wall chemical and physical structures, that is to say how they are synthesized and can be deconstructed, is crucial to design effective and economical methods to obtain fermentable sugars from lignocellulosic biomass. Tools must be developed to analyze biomass at the levels and specificities needed to understand how certain physical and chemical features of cell-wall inhibit to facilitate enzymatic interactions and subsequent saccharification. The development in recombiant DNA technology is enabling researchers to develop cost-effective ways of producing selected enzymes and unique enzyme combinations. Recently, some advances have been made in the identification of novel proteins, different from CBH, EG-BG, that play an important, and still undiscovered role in cellulose degradation. By applying novel biotechnology tools it will be feasible to identify, clone and express these novel proteins in widely used rot fungus (such as Trichoderma reesei) and create new `supercellulase'-producing fungi strains. Another recent approach is searching for highly active fibrolytic enzymes from the symbiotic cellulolytic systems of termites and ruminals. Its ability to digest cellulose has been adscribed to a large number of bacterial, fungal, yeasts and protozoal species, which are optimally adapted for the degradation of lignocellulosic material (Weimer, 1996; Varma et al., 1994). Currently, the nature potential of these microbes to hydrolyze cellulose may be explored by isolating pure microorganisms from this complex microbial community, which produces highly active cellulases, hemicellulases and feruloyl esterases (KoÈning et al., 2005). Using biotechnological methods it is now possible to characterize their cellulolytic activities and produce selected cellulolytic enzymes in large quantities by heterologous expression of the corresponding genes. Plant and aninmal expression systems are also being developed for the production of recombinant cellulolytic enzymes (Louime and Uckelmann, 2008). In the future, foreseen process scenarios will involve new generations of hydrolytic enzymes, functioning near their theoretical limits, and engineered energy crops with better compositional features to serve as improved substrates for these new generation enzymes. Indeed, it is entirely possible that the next generation of energy crops will contain the gene encoding enzymes necessary for self-deconstruction, activated before harvest or at the end of the growth cycle (Himmel et al., 2007).

6.7

References

Bayer E A, Belaich J P, Shoham Y and Lamed R (2004), `The cellulosomes: multienzyme machines for degradation of plant cell wall polysaccharides'. Annu Rev Microbiol, 58, 521±554.

ß Woodhead Publishing Limited, 2010

172

Bioalcohol production

Berlin A, Gilkes N, Kiburn D, Maximento V, Bura R and Markov A (2006), `Evaluation of cellulase preparations for hydrolysis of hardwood substrates', Appl Biochem Biotechnol, 130, 528±545. Berlin A, Maximenko V, Gilkes N and Saddler J (2007), `Optimization of enzyme complexes for lignocellulose hydrolysis', Biotechnol Bioeng, 97-2, 287±296. Bertran M S and Dale B E (1986), `Determination of cellulose accessibility by differential scanning calorimetry', J. Appl. Polym. Sci, 32, 4241±4253. Bhat M and Bhat S (1997), `Cellulose degrading enzymes and their potential industrial applications', Biotechnol. Adv, 15, 583±620. Bock J D, Yernool D A and Eveleigh E (1998), `Purification, Characterization, and Molecular Analysis of Thermostable Cellulases CelA and CelB from Thermotoga neapolitana', Appl Environ Microbiol, 64-12, 4774±4781. Boisset C, Petrequin C, Chanzy H, Henrissat B and Schulein M (2001), `Optimized mixtures of recombinant Humicola insolens cellulases for the biodegradation of crystalline cellulose', Biotechnol Bioeng, 72, 339±345. Bon E and Ferrara M A (2007), `Bioethanol production via enzymatic hydrolysis of cellulosic biomass' in The role of agriculture biotechnologies for production of bioenergy in developing countries, FAO seminar, Rome, http://www.fao.org/ biotech/seminaroct2007.htm Bourne Y and Henrissat B (2001), `Glycoside hydrolases and glycosyltransferases: families and functional modules', Curr Op Struct Biol, 11-5, 593±600. Breyer W A and Matthews B W (2001), `A structural basis for processivity', Protein Sci, 10-9, 1699±1711. Cara C, Moya M, Ballesteros I, Negro M J, GonzaÂlez A and Ruiz E (2007), `Influence of solid loading on enzymatic hydrolysis of pretreated olive tree biomass', Proc Biochem, 42-6, 1003±1009. Chandra R P, Bura R, Mabee W E, Berlin A, Pan X and Saddler J N (2007), `Substrate Pretreatment. The key to effective enzymatic hydrolysis of lignocelllulosic?', Adv Biochem Eng/Biotecnol, 10, 67±93. Chang V S and Holtzapple M T (2000), `Fundamental factors affecting biomass enzymatic reactivity', Appl Biochem Biotechnol, 63±65, 3±19. Cheng K J, Huang C T, Chen C C and Tang T Y (2002), `Enhancing protein thermostability', US Patent App. 10/087,699. Cherry J R and Fidanstef A L (2003), `Directed evolution of industrial enzymes: an update'. Curr Opin Biotechnol, 14, 438±443. Converse A O, Matsuno R, Tanaka M and Taniguchi M (2004), `A model of enzyme adsorption and hydrolysis of microcrystalline cellulose with slow deactivation of the adsorbed enzyme', Biotechnol Bioeng, 32-1, 38±45. Das M, Royer T V and Leff L G (2007), `Diversity of fungi, bacteria and actinomycetes on leaves decomposing in a steam', Appl Environ Microbiol, 73, 756±767. Draude K M, Kurniawan C B and Duff S J B (2001), `Effect of oxygen delignification on the rate and extent of enzymatic hydrolysis of lignocellulosic material', Bioresour Technol, 79, 113±120. Eriksson T, Karlsson J and Tjerneld F (2002a), `A model explaining declining rate in the hydrolysis of lignocellulosic substrates with cellobiohydrolase I (Cel7A) and endoglucanase (Cel7B) of Trichoderma reesei', Appl Biochem Biotechnol, 101, 41± 60. Eriksson T, Karlsson J and Tjerneld F (2002b), `Mechanism of surfactant effect in enzymatic hydrolysis of lignocellulose', Enzyme Microb Technol, 31, 353±364. Fan L T, Lee Y-H and Beardmore D H (1980), `Mechanism of the enzymatic hydrolysis

ß Woodhead Publishing Limited, 2010

Enzymatic hydrolysis of lignocellulosic biomass

173

of cellulose: Effects of major structural features of cellulose on enzymatic hydrolisis', Biotechnol Bioeng, 22, 177±199. Fan L T, Lee Y H and Beardmore D R (1981), `The influence of major structural features of cellulose on rate of enzymatic hydrolysis', Biotechnol Bioeng, 23, 419±424. Galbe M and Zacchi G (2002), `A review of the production of ethanol from softwood', Appl Microbiol Biotechnol, 59, 618±628. GarcõÂa-Aparicio M P, Negro M J, Ballesteros I, Ballesteros M, GonzaÂlez A, Manzanares P and Oliva J M (2004), in Van Swaaij W M P, FjaÈllstrom T, Helm P and Grassi A, Second World Biomass Conference: Biomass for energy, industry and climate protection, ETA-Florence, WIP-Munich, vol II, 1557±1560. GarcõÂa-Aparicio M P, Ballesteros M, Manzanares P, Ballesteros I, GonzaÂlez A and Negro M J (2007), `Xylanase contribution to the efficiency of cellulose enzymatic hydrolysis of barley straw', Appl Biochem Biotechnol, 136±140, 353±365. George S P, Ahmad A and Rao M B (2001), `A novel thermostable xylanase from Thermomonospora sp.: influence of additives on thermostability', Bioresour Technol, 78, 21±224. Ghan Q, Allen S J and Taylor G (2002), `Design and operation of an integrated membrane reactor for enzymatic cellulose hydrolysis', Biochem Eng J, 12, 223± 229. GonzaÂlez-Blasco G, Sanz-Aparicio J, GonzaÂlez B, Hermoso J A and Polaina J (2000), `Directed evolution of beta-glucosidase A from Paenabacillus polymyxa to thermal resistance', J Biol Chem, 275, 13708±13712. Grethlein H E and Converse A O (1991), `Common aspects of acid prehydrolysis and steam explosion for pretreating wood', Biores Technol, 36, 77±82. Gruber F, Visser J, Kubicek C P and de Graaff L H (1990), `Cloning of the Trichoderma reesei pyrG gene and its use as a homologous marker for a high-frequency transformation system', Curr Gen, 18-5, 447±451. Gusakov A V and Sinitsyn A P (1992), `A theoretical analysis of cellulase product inhibition: Effect of cellulase binding constant, enzyme/substrate ratio, and betaglucosidase activity on the inhibition pattern', Biotechnol Bioeng, 40-6, 663±671. Gusakov A V, Sinitsyn A P, Gerasimas V B, Savitskene R Y and Steponavichus Y Y (1985), `A product inhibition study of cellulases from Trichoderma longibranquiatum using dyed cellulose', J Biotechnol, 3, 167±174. Gusakov A V, Sinitsyn A P, Davidkyn I Y, Davidkyn V Y and Protas O (1996), `Enhancement of enzymatic hydrolysis using a novel type of bioreactor with intensive stirring induced by electromagnetic field', Appl Biochem Biotechnol, 56, 141±153. Henrissat B and Bairoch A (1996), `Updating the sequence-based classification of glycosyl hydrolases', Biochem J, 316, 695±696. Himmel M E, Ruth M F and Wyman C E (1999), `Cellulase for commodity products from cellulosic biomass', Curr. Opi Biotechnol, 10, 358±364. Himmel M E, Ding S-Y, Johnson D K, Adney W S, Nimlos M R, Brady J W and Foust T D (2007), `Biomass recalcitrance: engineering plants and enzymes for biofuels production', Science, 315, 804±807. Hodge D B, Karim M N, Schell D J and McMillan J D (2008), `Soluble and insoluble solids contributions to high-solids enzymatic hydrolysis of lignocellulose', Bioresour Technol, 99-18, 8940±8948. Holtzapple M T, Coganta M, Shu Y and Hendrickson C (1990), `Inhibition of Trichoderma reesei cellulase by sugars an solvents', Biotechnol Bioeng, 36, 275± 287.

ß Woodhead Publishing Limited, 2010

174

Bioalcohol production

Johannes T W and Zhao H (2006), `Directed evolution of enzymes and biosynthetic pathways', Curr Opin Microbiol, 9, 261±265. Karita S, Sakka K and Ohmiya K (1996), `Cellulose-binding domains confer an enhanced activity against insoluble cellulose to Ruminococcus albus endoglucanase IV', J Ferment Bioeng, 81, 553±556. Kim E, Irwin D C, Walker L P and Wilson D B (1998), `Factorial optimization of a sixcellulase mixture', Biotechnol Bioeng, 5-58, 494±501. Koivula A, Kinnari T, Harjunpaa V, Ruohonen L, Teleman A, Drakenberg T, Rouvinen J, Jones T A and Teeri T T (1998), `Tryptophan 272: an essential determinant of crystalline cellulose degradation by Trichoderma reesei cellobiohydrolase Cel6A', FEBS Lett, 429(3), 341±346. KoÈning H, Frohlich J and Hertel H (2005), `Diversity and lignocellulolytic activities of cultured micoorganisms', in KoÈning, H and Varma A, Intestinal Microorganisms Termites and other Invertebrates, Springer Verlag, Heidelberg, 271±301. Laureano-PeÂrez L, Teymouri F, Alizadeh H and Dale B E (2005), `Understanding factors that limit enzymatic hydrolysis of biomass', Appl Biochem Biotechnol, 121±124, 1081±1099. Linder M and Teeri T T (1997), `The roles and function of cellulose-binding domains', J. Biotechnol. 57, 15±28. Louime C and Uckelmann H (2008), `Cellulosic ethanol: securing the planet future energy needs', Int J Mol Sci, 9, 838±841. Lynd L R, van Zyl W H, McBride J E and Laser M (2005), `Consolidated bioprocessing of cellulosic biomass: an update', Curr Opin Biotechnol, 16, 331±346. Madamwar D and Patel S (1992), `Formation of cellulases by co-culturing of Trichoderma reesei and Aspergillus niger on cellulosic waste', World J Microbiol Biotechnol, 8-2, 183±186. Mansfield S D, Mooney C and Saddler J N (1999), `Substrate and enzyme characteristics that limit cellulose hydrolysis', Biotechnol Prog, 15, 804±816. Mcfarland K C, Hanshu D, Teter S, Vlasenko E, Feng X and Cherry J (2007), `Development of improved cellulase mixtures in a single production organism', ACS Symposium Series, 972, 19±31. Mohagheghi A, Tucker M, Grohmann K and Wyman C (1992), `High solids simultaneous saccharification and fermentation of pretreated wheat straw to ethanol', Appl Biotech Biotechnol, 33, 67±81. Mooney C A, Mansfield S D, Touhy M G and Saddler J N (1998), `The effect of initial pore volume and lignin content on th enzymatic hydrolysis of softwoods', Bioresour Technol, 64, 113±119. Mussatto S I, Fernandes M, Milagres A M F and Roberto I C (2008), `Effect of hemicellulose and lignin on enzymatic hydrolysis of cellulose from brewer's spent grain', Enzym Microbiol Technol, 43-2, 124±129. Mutwil M, Debolt S and Persson S (2008), `Cellulose synthesis: a complex complex', Curr Opin Plant Biol, 11, 252±257. Nidetzky B, Steiner W and Claeyssens M (1994), `Cellulose hydrolysis by the cellulases from Trichoderma reesei: adsorptions of two cellobiohydrolases, two endocellulases and their core proteins on filter paper and their relation to hydrolysis', Biochem J, 303, 817±823. Nieves R A, Ehrman C I, Adney W S, Elander R T and Himmel M E (1998), `Survey and analysis of commercial cellulase preparation suitable for biomass conversion to ethanol', World J Microbiol Biotechnol, 14, 301±304.

ß Woodhead Publishing Limited, 2010

Enzymatic hydrolysis of lignocellulosic biomass

175

NREL (1999), Bioethanol Multi-Year Technical Plan, National Renewable Energy Laboratory, Golden. CO, USA. È hgren K, Bura R, Saddler J and Zacchi G (2007), `Effect of hemicellulose and lignin O removal on enzymatic hydrolysis of steam pretreated corn stover', Bioresour Technol, 98, 2503±2510. Ohlson I, Tragardh G and Hahn-Hagerdal B (1984a), `Enzymatic hydrolysis of sodiumhydroxide-pretreated sallow in an ultrafiltration membrane reactor', Biotechnol Bioeng, 26, 647±653. Ohlson I, Tragardh G and Hahn-Hagerdal B (1984b), `Evaluation of UF and RO in a cellulose saccharification process', Desalination, 51, 93±101. Ohmiya K, Sakka K, Karita S and Kimura T (1997), `Structure of cellulases and their applications', Biotechnol Genet Eng Rev, 14, 365±414. Ooshima H, Burns D S and Converse A (1990), `Adsoprtion of cellulase from Trichoderma reesei on cellulose and lignacious residue in wood pretreated by dilute sulphuric acid with explosive decompression', Biotecnol Bioeng, 36, 446±452. Padukone, N (1996), `Advance process options for bioethanol production', in Wyman C E, Handbook on Bioethanol: Production and Utilization, Washington, DC. PeÂrez J, MunÄoz-Delgado J, de la Rubia T and Martinez J (2002), `Biodegradation and biological treatments of cellulose, hemicellulose and lignin: an overview', Inter Microbiol, 5, 53±63. PeÂrez J A, GonzaÂlez A, Oliva J M, Ballesteros I and Manzanares P (2007), `Effect of process variables on liquid hot water pre-treatment of wheat straw for bioconversion to fuel-ethanol in a batch reactor', J Chem Technol Biotechnol, 82, 929±938. Puri V T (1984), `Effect of crystallinity and degree of polymerization of cellulose on enzymatic saccharification', Biotechnol Bioeng, 26, 1219±1222. Quentin M, van der Valk H, van Dam J and de Jong E (2003), `Cellulose-binding domains: tools for innovation in cellulosic fiber production and modification', in Mansfield S and Saddler J (eds), Applications of Enzymes to Lignocellulosics. ACS Symposium Series 855, American Chemical Society, Washington, DC, 132±155. Reeta R S, Sukumaran R K and Panday A (2007), `Improved cellulase production by Trichoderma reesei RUT C30 under SSF through process optimization', Appl Biochem Biotechnol, 142-1, 60±70. Rosgaard L, Pedersen S, Langston J, Akerhielm D, Cherry J R, Meyer A S (2007), `Evaluation of minimal Trichoderma reesei cellulase mixtures on differently pretreated barley straw substrates', Biotechnol Prog, 23-6, 1270±1276. Schulein M (2000), `Protein engineering of cellulases', Biochem Biophys Acta, 1534, 239±252. SoÈderstroÈm J, Galbe M and Zacchi G (2005), `Separate versus simultaneous saccharification and fermentation of two-step steam pretreated softwood for ethanol production', J Wood Chem Technol, 25-3, 187±202. Sutcliffe R and Saddler J M (1986), `The role of the lignin in the adsorption of cellulases during enzymatic treatment of lignocellulosic material', Biotechnol Bioeng, 17, 749±762. Takagi M, Abe S, Suzuki S, Emert G H and Yata N (1977), `A method for production of alcohol direct from cellulose using cellulase and yeast', Process Biocon Symp, IIT Delhi, 551±571. Tanaka M, Ikesaka M and Matsuno R (1988), `Effect of pore size in substrate and diffusion of enzyme on hydrolysis of cellulosic material with cellulase', Biotechnol Bioeng, 32, 698±706.

ß Woodhead Publishing Limited, 2010

176

Bioalcohol production

Teeri T T (1997), `Crystalline cellulose degradation: new insight into the function of cellobiohydrolases', Trend Biotechnol, 15, 160±167. Teeri T T and Koivula A (1995), `Cellulose degradation by native and engineered fungal cellulases', Carbohydr Eur, 12, 28±33. Teeri T T, Koivula A, Linder M, Wohlfahrt G, Divne C, Jones T A (1998), `Trichoderma reesei cellobiohydrolases: why so efficient on crystalline cellulose?', Biochem Soc Trans, 26(2), 173±178. Tolan J S (2008), `Iogen's demonstration process for producing ethanol from cellulosic biomass', in Kamm B, Gruber P R and Kamm M, Biorefineries ± Industrial Processes and Products, Wiley-VCH Verlag GmbH. Valaskova V and Baldrian P (2006), `Degradation of cellulose and hemicelluloses by the brown rot fungus Piptoporus betulinus production of extracellular enzymes and characterization of the major cellulases', Microbiol, 152, 3613±3622. Varga E, Reczey K and Zacchi G (2004), `Optimization of steam pretreatment of corn stover to enhance enzymatic digestibility', Appl Biochem Biotecnol, 113±116, 509± 523. Varma M, Kolli B K, Paul J, Saxena S and KoÈnig H (1994), `Lignocellulosic degradation by microorganisms from termite hills and termite guts. A survey on the present state of the art', FEMS Microbiol Rev, 15, 9±28. Wald S, Wilke C R and Blanch H W (1984), `Kinetics of enzymatic hydrolysis of cellulose', Biotechnol Bioeng, 26, 225±230. Wang T, Liu X, Cheng G, Zhang Y and Gao P (2005), `Directed evolution for engineering pH profile of endoglucanase III from Trichoderma reesei, Biomol Eng, 22, 89±94. Wang W, Liu J, Chen G, Zhang Y and Gao P (2003), `Function of a low molecular weight peptide from Trichoderma pseudokoningii S38 during cellulose biodegradation', Curr Microbiol, 46, 371±379. Weimer P J (1996), `Why don't rumial bacteria digest cellulose faster?', J Dairy Sci, 79, 1496±1502. Wilson D B (2004), `Studies on Thermobifida fusca plant cell wall degrading enzymes', Chem Rec, 4, 72±82. Wojtczak G, Breuil C, Yamada J and Saddler J N (1987), `A comparison of the thermostability of cellulases from various thermophilic fungi', Appl Microbiol Biotechnol, 27-1, 82±87. Wooley R, Ruth M, Sheehan J, Ibsen K, Majdeski H and Galvez A (1999), `Lignocellulosic biomass to ethanol process design and economics utilizing cocurrent dilute acid prehydrolysis and enzymatic hydrolysis: current and futuristic scenarios', NREL Technical Report: NREL/TP-580±26157, National Renewable Energy Laboratory, Golden, CO, USA. Wyman C E (1994), `Ethanol from lignocellulosic biomass: technology, economics, and opportunities', Bioresource Technol, 50, 3±16. Wyman C E (1996), `Ethanol production from lignocellulosic biomass: overview', in Wyman C E, Handbook on Bioethanol: Production and Utilization, Washington, DC, Taylor & Francis,1±18. Wyman C E, Spindler D D and Grohmann K (1992), `Simultaneous saccharification and fermentation of several lignocellulosic feedstocks to fuel ethanol', Biomass Bioenergy, 5, 301±307. Xiao Z, Gao P-J, Qu Y-B and Wang T-H (2001), `Cellulose-binding domain of endoglucanase III from Trichoderma ressei', Biotechnol Lett, 23, 711±715.

ß Woodhead Publishing Limited, 2010

Enzymatic hydrolysis of lignocellulosic biomass

177

Xiao Z, Zhang X, Gregg D J and Saddler J N (2004), `Effects of sugar inhibition on cellulase and -glucosidase during enzymatic hydrolysis of softwood substrates', Appl Biochem Biotechnol, 113±116, 1115±1126. Yang S, Ding W Y and Chen H Z (2006), `Enzymatic hydrolysis of rice straw in a tubular reactor coupled with UF membrane', Process Biochem, 41, 721±725. Zeng Y H, Huang G, Xi H X, Wuang R, Huang D, Huang G and Li J (2007), `Microbial community succession and lignocellulose degradation during agricultural waste composting', Biodegrad, 18, 793±802. Zhang Y-H P, Himmel E E and Mielenz J R R (2006), `Outlook for cellulase improvement: screening and selection strategies', Biotechnol Adv, 24, 452±481. Zhang Y-H P and Lynd L R (2004), `Towards and aggregated understanding of enzymatic hydrolysis of cellulose: noncomplexed cellulase systems', Biotechnol Bioeng, 88, 1563±1569. Zhang S, Wolfgang D E and Wilson D B (1999), `Substrate heretogenety causes the nonlinear kinetics of insoluble cellulose hydrolysis', Biotechnol Bioeng, 66, 35±41. Zhou S and Ingram L O (2001), `Simultaneous saccharification and fermentation of amorphus cellulose to ethanol by recombinant Klebsiella oxytoca SZ21 without supplemental cellulase', Biotechnol Lett, 23, 1455±1462. Zhu L, O'Dwyer J P, Chang V S, Granda C B and Holtzapple M T (2008), `Structural features affecting biomass enzymatic digestibility', Bioresour Technol, 99, 3817± 3828.

ß Woodhead Publishing Limited, 2010

7

Development of cellulases to improve enzymatic hydrolysis of lignocellulosic biomass R . J . Q U I N L A N , S . T E T E R and F . X U , Novozymes, Inc., USA

Abstract: In the current thrust to commercialize cellulosic biomass-derived ethanol, of utmost importance is the development of cost-effective, highly active, storage and process stable enzymes suitable for the conversion of a broadly varied cellulosic feedstocks. The enzymes produced by cellulasesecreting fungi and bacteria provide the basis for the development of commercially viable enzymatic cellulosic bioethanol production. We discuss the approaches currently utilized in cellulase discovery and optimization, recent advances in the field, challenges to cellulase development and potential enhancements that will ultimately enable efficient enzymatic function under the process constraints imposed by the economics of cellulosic bioethanol production. Keywords: cellulase, biomass, second generation biofuel, protein engineering, fungi.

7.1

Introduction

In the current thrust to commercialize cellulosic biomass-derived ethanol, of utmost importance is the development of cost-effective, highly active, storage and process-stable enzymes suitable for the conversion of broadly varied cellulosic feedstocks. First generation bioethanol is currently in widespread production with the world's largest producers, the US and Brazil, utilizing almost entirely corn and sugarcane-based feedstock (Lin and Tanaka, 2006; Kumar et al., 2008, Renewable Fuels Association, 2009). For easily hydrolyzed feedstocks such as starch, straightforward enzymatic degradation is sufficient, and enzyme cost is not a limiting factor. However, as government mandates for biofuel production require growth beyond levels that can be readily obtained from starch or sugar sources, an ever greater proportion of that production must derive from cellulosic feedstock, and enabling technologies for second-generation bioethanol become imperative (US Department of Energy, 2006, 2008a; D'Aquino, 2007). The price of enzymes must, to most ethanol suppliers, remain a cost both independent of changes in the feedstock and fuel oil markets and sufficiently low as to enable ethanol to be produced as a viable alternative to conventional

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

179

petrochemical fuels over a broad range of fuel prices. Historically, the quantity of cellulases; the enzymes that catalyze cellulose hydrolysis, required for biomass conversion has been disproportionately large, and the associated cost has been a limiting-factor in cellulosic ethanol development (McAloon et al., 2000; Aden et al., 2002; Teter and Cherry, 2005). The goal, therefore, of both academic and industrial research in this area is the significant reduction of enzyme costs by optimizing output, efficiency and specificity of cellulase mixtures. While the enzymes produced by cellulase-secreting fungi and bacteria provide a foundation for the development of commercially viable enzymatic cellulosic bioethanol production, there are fundamental differences between the targets and optimal conditions for cellulose degradation in nature, and those of industrial ethanol production (Kumar et al., 2008; Zhang et al., 2006). Considerable enzyme optimization, modification and fundamental research are therefore necessary to understand the underlying mechanisms, kinetic obstacles and to identify potential enhancements that will ultimately enable efficient enzymatic function under the experimental and process constraints imposed by the economics of cellulosic bioethanol production.

7.2

Cellulase structure and function

7.2.1

Types of fungal cellulases

Canonical cellulase mixtures, including that of the widely studied, widely utilized, filamentous fungus, Trichoderma reesei, (Hypocrea jecorina) consist of several key constituents and a host of minor components (Table 7.1). The key constituents are the cellobiohydrolases (CBHs) and endoglucanases (EGs). CBHs are processive exocellulases, which hydrolyze crystalline cellulose to generate cellobiose ( -(1!4)-D-glucopyranosyl-D-glucopyranose) as their predominant product. EGs are non-processive endocellulases with catalytic preference for non-crystalline cellulose, which cleave a cellulose chain randomly along its length. The bulk of the protein, by mass, is made up of four proteins: cellobiohydrolases I and II (CBH-I, CBH-II), and endoglucanase I and II (EG-I, EGII). The most significant minor components include -glucosidase (cellobiase, BG), which hydrolyzes cellobiose to D-glucose (Davies and Henrissat, 1995). Indeed, synthetic minimal cellulase mixtures routinely contain the four principle enzymes, CBH-I, CBH-II, EG-I, EG-II in isolation, or for greater efficiency, the four key components supplemented with BG (Baker et al., 1998; Gusakov et al., 2007; Viikari et al., 2007). While these constitute the major fungal cellulase constituents, to date over 100 different families of glycohydrolases have been classified by sequence similarity (Henrissat, 1991; Henrissat and Bairoch, 1993, 1996; Coutinho and Henrissat, 1999; Cantarel et al., 2008). Among the cellulase-producing organisms, fungal cellulase mixtures are most likely to be commercially viable, as hyper-cellulolytic strains producing

ß Woodhead Publishing Limited, 2010

180

Bioalcohol production

Table 7.1 The major secreted proteins of T. reesei cultured on medium inducing cellulase production. Common name

Common acronym

CAZy classificationa

Stereochemistryb

Cellobiohydrolase-I Cellobiohydrolase -II endoglucanase-I endoglucanase -II endoglucanase -III endoglucanase -IV endoglucanase-V -glucosidase I -glucosidase II xyloglucanase xylanase-I xylanase -II xylanase -III xylanase -IV -xylosidase Mannanase Acetylxylan esterase -galactosidase Arabinofuranosidase-I Arabinofuranosidase-II Arabinofuranosidase-III Feruloyl esterase Cellulose binding protein-I Cellulose binding protein-II Swollenin Trypsin-like protease

CBH-I CBH-II EG-I EG-II EG-III EG-IV EG-V BGL-I BGL-II EGL6 Xyn-I Xyn-II Xyn-III Xyn-IV BXL-I Man-I AXE AG AF-I AF-II AF-III FAE CBP-I CBP-II SWO

cel7A cel6A cel7B cel5A cel12A cel61A cel45A cel3 cel1A cel74A GH11 GH11 GH10 GH5 GH3 Man5A axe1 Bga1 Abf1 Abf2 Abf3 Fae1 Cip1 Cip2 Swo Trp

retaining inverting retaining retaining retaining n.d. inverting retaining retaining inverting retaining retaining retaining retaining retaining retaining retaining retaining retaining retaining

Fraction of total secreted proteincd (%) 40±60e 12±20 5±10e 1±10e 0±2 0.2±0.8 <5 0.5±1.4 n.d. 1.3±2.4 0.25 0.1 n.d. 0.05±0.1 0±0.2 0.1±0.3 n.d. n.d. 0.2±0.4 0.1±0.2 0±0.3 n.d. n.d. n.d. n.d. 0±0.2

a

Henrissat and Davies, 1997; bCantarel et al., 2008; cFractional composition is highly-dependent on specific growth medium and on the specific strain of T. reesei examined; dHerpoel-Gimbert et al., 2008; eRosgaard et al., 2006; n.d.: not determined.

yields of 40±100 g/L secreted enzyme have been developed (Cherry and Fidantsef, 2003). Cellulase-producing species, however, are found in such diverse phylogeny as bacteria, archea, plants and insects. While these are not discussed extensively here, recent reviews highlight the important similarities and differences between these cellulases and those from fungal sources (Gilbert, 2007; Ding et al., 2008; Bayer et al., 2004). These non-fungal cellulases remain potentially useful for diversity-driven protein engineering for enhanced cellulases. Fungal cellulase producers include the so-called white-rot, brownrot and soft-rot fungi, differing widely in cellulase composition and in their mechanisms of lignin-degradation (Singh and Chen, 2008; Baldrian and ValaÂsÏaskovaÂ, 2007).

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

7.2.2

181

Cellulase structure

Commonly, cellulases have modular structure, containing a cellulose-binding module (CBM) and a core, catalytic domain (CD) linked via a flexible, frequently glycosylated linker. The occurrence and position of the CBMs vary with cellulase type and source organism (HildeÂn and Johansson, 2004; Davies and Henrissat, 1995; Boraston et al., 2004). CBMs are generally small peptides in fungal cellulases (4 kD), though much larger domains are reported for bacterial cellulases (20 kD). These domains function to ease phase transfer between the soluble phase and the insoluble substrate, thereby reducing a threedimensional search for cellulose ends to a two-dimensional exploration. CBMs are diverse; 52 unique families have been classified by structural and sequence homology (Cantarel et al., 2008). One of the best studied of these, CBM-I, features three to four aromatic residues in a planar orientation, facilitating binding to the hydrophobic patches of the cellulose surface. It is somewhat surprising that these critical binding-residues, in CBM-I and others, comprise the most terminal amino acids of the enzyme. The genetic advantages of modular structure include maintaining diversity despite limited gene numbers via unique combinations of CD and CBM, and varying stoichiometric ratios of CBMs to CDs. This permits tailoring binding specificity to enzymatic activity and facilitating post-translational modification via protease activity between CD and CBM (HildeÂn and Johansson, 2004; Martinez et al., 2008; Blake et al., 2006). The H-bonding pattern of cellulose alternates with the face of the cellobiose repeating unit; therefore CBM-binding must reflect this electrostatic surface. Recent molecular dynamics models have suggested that in the case of the T. reesei CBH-I CBM, three exposed binding residues are accompanied by a fourth, partially buried residue that is exposed by partial unfolding of the CBM during productive binding (Nimlos et al., 2007). Binding of the CBM liberates a short oligoglucan, facilitating its association with the catalytic Ê active site tunnel, into which domain. The catalytic domain possesses a 20±50 A the oligoglucan is threaded and hydrolyzed to its constituent cellobiose units (Fig. 7.1). The processivity of CBH catalysis is attributed to the closed active site tunnel surrounding the cellulose polymer. The active site contains binding subsites for multiple anhydroglucose units of a cellulose chain, which are, by convention, numbered by position relative to the scissile glycosidic bond, hydrolysis occurring between ‡1 and ÿ1 subsites. CBHs differ in the specific number of subsites, T. reesei CBH-I contains ten subsites, denoted ÿ7 to ‡3, whereas CBH-II contains five (Rabinovich et al., 2002; Davies and Henrissat, 1995; Vasella et al., 2002). Endoglucanases are structurally similar to their CBH counterparts, differing most prominently in the loops that delineate the active site tunnels in CBHs, which are truncated to generate an open active site cleft in EGs. These enzymes are believed to have specificity for regions of imperfect crystallinity, and thus

ß Woodhead Publishing Limited, 2010

182

Bioalcohol production

7.1 Mechanisms of cellulolysis. (a) Depiction of synergistic cellulase activity; (b) single displacement catalytic mechanism of CBH-II; (c) double displacement catalytic mechanism of CBH-I.

their activity is generally low on such substrates as microcrystalline cellulose and much higher on modified or synthetic cellulose substrates such as carboxymethylcellulose (CMC). The open active site clefts of endoglucanases allow non-processive cellulolysis and endo-activity on the cellulose polymer.

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

183

7.1 Continued

Prokaryotic cellulases, in brief, may differ fundamentally in their organization from those of fungal cellulases, and extensive heteroassociation of cellulases into complexes called cellulosomes is commonly observed for anaerobic bacteria (Fierobe et al., 2005; Mingardon et al., 2007a; Gilbert, 2007). Scaffoldins provide a structural framework for cellulases; cohesin and dockerin domains on the scaffoldin and the cellulase, respectively, provide interaction sites for the various cellulases. Cellulosomal organization is believed to enhance cellulose hydrolysis by facilitating enzyme synergy via proximity-effects (Mingardon et al., 2007b; Caspi et al., 2008). Additionally, fungal and bacterial cellulases may differ in localization. While fungal cellulases are secreted into the extracellular medium, prokaryotic cellulases may remain membraneassociated and thereby retain saccharification products at or near the cell surface.

ß Woodhead Publishing Limited, 2010

184

7.2.3

Bioalcohol production

Non-cellulase secreted factors

The secretome of T. reesei contains its key cellulase components in roughly the following composition: 45±60% CBH-I, 15% CBH-II, 12% EG-I, 9% EG-II, and <2% EG-III. A full complement of other proteins, both enzymatic and accessory, are often secreted along with the principal components, their relative composition dependent on the fungal species and the composition of the cellulosic material used as the carbon source in the growth medium (Olsson et al., 2003; Bashkirova et al., 2005; Berlin et al., 2005; Vanden Wymelenberg, 2009; JuhaÂsz et al., 2005). Non-cellulase, secreted factors include, but are not limited to, hemicellulases, GH61 proteins, swollenins, oxidoreductases, proteases, etc., and may total greater than 30 proteins. Many of these components of the T. reesei secretome are listed in Table 7.1, along with representative relative compositions. Enzymes enabling fungi to degrade hemicellulose and modified cellulose may be secreted in lower abundance; examples of these include -glucuronidase, arabinofuranosidase, acetyl xylan esterase, feruloyl esterase, pectin and glucuronan lyases, etc. (Lee et al., 2005; Benoit et al., 2008; Payasi et al., 2009). In the case of white-rot fungi, with Phanerochaete chrysosporium as the canonical example, enzymes responsible for oxidative delignification are also secreted. Examples include laccase, Mnand lignin-peroxidases (Kersten and Cullen, 2007; Banci et al., 1998). A number of proteases are also secreted by the fungi, the function of which may include the salvage of nitrogen, a resource likely to be limiting for organisms surviving on cellulose.

7.2.4

Cellulose degradation

The breakdown of pure cellulose to its repeating unit, cellobiose, is accomplished by the combined activities of the key cellulases, CBH-1, CBH-II and EG. These factors act synergistically, which is to say that their combined cellulolytic activity is greater than the sums of their respective individual activities. The chief mechanisms for synergism include exo-endo synergism, wherein endoglucanases depolymerize cellulose, generating additional ends for exo-activity and thereby enhancing exoglucanase activity, and exo-exo synergism, wherein CBH-I and CBH-II act at opposite ends of the cellulose chain, reducing and nonreducing, respectively, stripping cellulose layers and exposing additional ends for hydrolysis (Henrissat et al., 1985). Most enzyme mixtures are supplemented with exogenous -glucosidase, as the breakdown of cellulose to cellobiose in vitro produces high cellobiose concentrations, leading to significant product inhibition (Bezerra and Dias, 2005; Sternberg et al., 1977). Depending on the nature of the biomass pretreatment, cellulosic feedstocks are rarely pure crystalline cellulose. Residual hemicellulose is generally present, particularly in the case of alkali-based pretreatments, and further synergistic

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

185

enhancement is observed when hemicellulases (Table 7.1) and -xylosidase are present in the cellulase mixture.

7.2.5

Enzymatic mechanisms of cellulose degradation

The enzymatic mechanisms of several cellulases have been determined, based on biochemical data and on crystal structures solved in various states of ligandbinding and catalysis (Vasella et al., 2002). The specific catalytic mechanism of the cellulases involves nucleophilic attack by either an active site nucleophile, or by base-activated water, aided by one or two residues acting as the general acid/ base; the specific mechanism of nucleophilic attack differing between CBH-I and CBH-II (Fig. 7.1). CBH-I hydrolyzes the glycosidic bond in a doubledisplacement reaction with retention of configuration. CBH-II utilizes a singledisplacement mechanism resulting in inversion of configuration at the anomeric carbon. Like CBHs, specific EG families possess different catalytic stereochemistry, and these are summarized in Table 7.1. The exceptional case among cellulases are the family 4 glycohydrolases, bacterial cellulases for which an elimination mechanism has been reported (Chiba, 1997; Rabinovich et al., 2002; Vasella et al., 2002; Vocadlo and Davies, 2008; Mosier et al., 1999; Vivian et al., 2006). Elimination mechanisms are commonly observed for non-cellulase accessory enzymes including the pectin and glucuronan lyases. A generally overlooked, though well-studied, mechanism of cellulose breakdown is via oxidative degradation. Most commonly ascribed to the brown-rot fungi, wherein cellulase production differs significantly from T. reesei and well-described for lignolytic pathways of white-rot fungi, oxidative cleavage may play a significant role in cellulolysis in some organisms (Wood, 1994; Kersten and Cullen, 2007). The brown-rot fungus Postia placenta, for instance, has in its genome no canonical CBH-I or II, and may rely largely on extracellular Fenton chemistry for cellulosedegradation (Martinez et al., 2009). The enzymatic activities of secreted peroxidases, and the generation of superoxide or hydroxyl radicals by secreted oxidoreductases such as cellobiose dehydrogenase (CDH) have been shown to significantly reduce the degree of polymerization of cellulose in vitro (KrusaÊ et al., 2005, 2008; Mansfield et al., 1997). Support for these types of mechanisms also comes from analysis of the P. chrysosporium genome, which reveals a considerable fraction of oxidative enzymes among the putative extracellular enzymes (Kersten and Cullen, 2007). At issue in the exploitation of these mechanisms for commercial bioethanol production is the non-specific nature of radical-mediated cellulose modification, and the significant potential for protein damage by these radicals.

7.2.6

Kinetics of cellulolysis

The kinetic profile of cellulose degradation is, at once, both simple and perplexing. The time-dependence of enzymatic cellulose conversion to cellobiose/

ß Woodhead Publishing Limited, 2010

186

Bioalcohol production

7.2 Time-dependence of biomass saccharification by T. reesei cellulases. Fractional conversion of biomass to cellobiose and glucose is plotted, and the apparent biphasic nature of biomass saccharification is apparent. Hydrolysis by two concentrations of enzyme is shown. Solid line: high enzyme dose; dotted line: low enzyme dose.

glucose is fractal, and appears as a biphasic or even multiphasic curve (Fig. 7.2); a rapid initial rate followed by a second, slower rate asymptotically approaching a limit conversion (Kadam et al., 2004; Xu and Ding, 2007; VaÈljamaÈe et al., 2003; Kipper et al., 2005). The magnitude and rate of the initial phase is highly dependent on the enzyme concentration; however, the time-evolution of the following phase has a much smaller dependence on initial enzyme dose. The rate of this phase and the limiting extent of conversion depend heavily on the specific enzymes in the mixture. Several explanations for the heterogeneous hydrolysis have been proposed, and this is an area of active discussion in the literature (Zhang and Lynd, 2006). The proposed mechanisms include substrate heterogeneity and recalcitrance, inhibition, or loss of active enzyme. Cellulose is a heterogeneous substrate, its regions of variable degrees of crystallinity leading to differential rates of catalysis and increased recalcitrance as hydrolysis progresses. More significant complications arise from the even greater heterogeneity of lignocellulosic biomass feedstocks. (Ai-Zuhair, 2008). Loss of enzyme activity is additionally anticipated as inhibition arises with increasing product concentrations or from non-productive binding to lignin, hemicellulose or steric barriers to processive hydrolysis (Selig et al., 2007; Tu et al., 2007; ValaÂsÏaskova and Baldrian, 2006; Berlin et al., 2006). Decreasing hydrolytic rates may arise from enzyme denaturation via thermal unfolding or from shear interactions at the liquid-solid or liquid-air interface. Loss of enzyme activity is

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

187

observed during incubation under hydrolysis conditions, and the synergism between various cellulases amplifies the deleterious effects of loss of individual component activities. Non-specific protein adsorption to lignin is a generally observed phenomenon (Chandra et al., 2007), which may be exacerbated by the presence of CBMs designed to bind insoluble polymer. Non-productive binding to hemicellulose or to cellulose that has been modified, kinked or has become entangled with hemicellulose during substrate pretreatment, coupled with the slow dissociation rates expected for processive cellulases may similarly contribute to loss of activity (Ma et al., 2008; Jeoh et al., 2007). It is expected that these binding issues will become more significant as hydrolysis progresses and the non-substrate components represent a more substantial fraction of the residual biomass. Furthermore, from a kinetic perspective, in cases where there is a non-zero probability of non-productive, irreversible binding following each dissociation step, then complete loss of activity is expected at infinite time. The overall reduction in cellulolysis rate is likely to reflect the summation of all these contributing factors. Discussion of the kinetics of hydrolysis highlights some of the major challenges in the development of second generation bioethanol enzymes: an entire suite of enzyme activities must be employed to saccharify the highly heterogeneous and insoluble biomass substrate. While maintaining a minimal composition of constituent enzymes, all the requisite catalytic functionalities necessary to contend with the substrate heterogeneities exposed as hydrolysis progresses must be maintained throughout the saccharification, despite the significant enzyme stress imposed by the hydrolysis conditions.

7.3

Development of cellulases

In the 60 years of study since the isolation of various cellulase-secreting fungi and bacteria was accomplished in the 1940s, our understanding of cellulases has progressed to the point where detailed molecular models now exist for both cellulose structure, and the association with and progressive hydrolysis of that substrate by CBH-I (Nimlos et al., 2007; Zhong et al., 2008). Conversely, the practical application of that understanding to enhance saccharification of lignocellulosic substrate under conditions relevant to the commercial bioethanol industry has progressed only marginally during this time. The single greatest enhancement accomplished thus far has arguably been the supplementation of T. reesei fermentation broths with exogenous -glucosidase, thereby reducing the product inhibition that arises from high total solids saccharification (Sternberg et al., 1977). Much of the basic cellulase research performed has utilized synthetic substrates, minimal cellulase mixtures and low solids content in order to generate initial rate data for enzyme comparison. The practical realities of industrial biomass hydrolysis will undoubtedly little resemble these experimental conditions. High solids saccharifications will be performed on

ß Woodhead Publishing Limited, 2010

188

Bioalcohol production

diverse substrates, pushing the limits of conversion to maximally digest the available substrate. Delineation of specific enzyme performance and the optimization thereof under these conditions will ultimately be required. Even more critical to this endeavor may be the identification of the translational factors correlating data at these two experimental extremes. The efforts underway to obtain complete cellulosic biomass hydrolysis at high solids using minimal enzyme dosing can be broadly classified into two foci: screening natural diversity and in vitro protein engineering (Zhang et al., 2006; Schwartzkopff, 2008; SchuÈlein, 2000). The former category involves the utilization of fungal and prokaryotic libraries and the enlargement thereof by direct sampling of decaying cellulosic material for new microbial diversity, and by in silico sampling of genomic databases. The latter category seeks to improve the enzymatic and structural characteristics of specific cellulases, either by designbased mutagenesis or via in vitro evolutionary methods. These two approaches are ideally utilized in complement, wherein insights and comparisons gained from nature and from new cellulases guide protein engineering efforts, and in vitro experiments in turn suggest new candidates for library screening and ultimately to heterologous expression of improved variants. It is critically important, however, to note that cellulase development efforts cannot take place in isolation; process modeling and engineering must take place concurrently, such that the criteria utilized for enzyme improvement and discovery match those ultimately required by biorefineries.

7.3.1

Identifying novel cellulases

Attempts to identify and isolate novel cellulases, homologs of cellulases with different, novel or desired characteristics, and to identify novel trans-acting, cellulase-enhancing factors have included broadly surveying natural diversity. Fungal and bacterial libraries are screened for cellulase production by growth on appropriate media, and suitable candidates are identified (Toyama et al., 2008; Teter et al., 2006). Additional screening criteria may also be imposed; classic examples being thermo- or halophilicity, resistance to inhibitors found in or on substrates, and ability to metabolize specific biomass feedstock. Gene expression patterns reflect the growth conditions, and consequently proteomic and microarray analyses play a crucial role in discovery. This highlights yet another issue in cellulase development, the variety of cellulases and their differential expression to fit minor differences in cellulosic carbon source in situ cannot be reflected in the production of commercially relevant enzyme mixtures, which must, for the large part, be universal with respect to substrate, at least in the initial stages of second generation bioethanol development (Schwartzkopff, 2008). Factors identified in screening are then purified to homogeneity and biochemically characterized. Additionally, genome and secretome sequencing of existing and novel cellulose-degrading fungi and bacteria provide insight via

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

189

comparative genomics and sequence homologies. Each of these approaches serves to build our library of cellulases, and our libraries of gene and expression data, leading to the isolation of cellulases with such desired properties as thermal stability, varied pH optima, specific activity, reduced sensitivity to inhibition, etc.

7.3.2

Genomics and metagenomics approaches

In addition to the search for novel cellulases from both novel and cultured microbial sources, a concerted effort is underway to develop and provide genomic and metagenomic tools, data and support to the search for bioethanol enzymes and biomass feedstocks (Li et al., 2009; Rubin, 2008). Beyond the genome sequencing performed by industrial cellulase producers, the US Department of Energy Joint Genome Institute (JGI), its collaborating academic and industrial partners provide extensive genomics and metagenomics research and database support. The JGI's Plant Genome Program provides data for such current and future biomass feedstocks as corn, sorghum, loblolly pine and soybean, yielding insights into the genetic basis underlying plant cell wall biogenesis in these organisms. Of more direct impact to the development of cellulases is the Microbial Genome Program, and in particular the Fungal Genomics Program. Notable successes thus far include the publication of the T. reesei, Postia placenta and Phanerochaete chrysosporium genomes (Martinez et al., 2004, 2008, 2009). Expected in the near future are the sequences of hypercellulolytic mutants of T. reesei and a number of prokaryotic cellulosedegrading organisms. Ongoing metagenomics projects, those that focus on the sequencing of microbial colony DNA rather than isolated strains or organisms, include the recently completed termite hindgut (Warnecke et al., 2007), the cellulose-degrading and fermenting esophageal crop of the Amazonian stinkbird Hoatzin (Opisthocomus hoazin), and the symbiotic microbiota of the wooddegrading pacific shipworm (Bankia setacea). Through the Integrated Microbial Genome and Integrated Microbial Genome/Microbiome informatics portals (IMG, IMG/M) the array of publically available genome sequences is available for comparative analysis (US Department of Energy, 2006, 2008a,b).

7.3.3

Protein engineering

Protein engineering efforts to date have focused on coaxing T. reesei or other fungal cellulases to greater activity under the conditions prescribed by the predicted economics of bioethanol fermentation (Dumon et al., 2008; Qin et al., 2008; Hess, 2008). Historically, these conditions are exemplified by either conversion of suitable lignocellulosic substrate at 50 ëC and pH 5.0, optimal conditions for T. reesei cellulases or alternatively under conditions amenable to simultaneous saccharification/fermentation (SSF). Based on the relatively high

ß Woodhead Publishing Limited, 2010

190

Bioalcohol production

melting points of the T. reesei cellulases, ~60±65 ëC, thermal stability of the enzymes would not be expected to be a significant hurdle. Somewhat surprisingly, however, irreversible loss of activity is typically observed, an issue ascribed to the extensive incubation time required for complete cellulose conversion (Vlasenko et al., 2003; Voutilainen et al., 2007, 2008) and thus thermal stability is an active target for protein engineering. Another well-defined target for protein engineering is the amelioration of product-inhibition; both cellobiose-dependent inhibition of CBH (Ki typically ~5 mM) and glucosedependent inhibition of -glucosidase (Ki typically 1 mM). Under the high total cellulose conditions most likely to be economically viable, glucose concentrations are likely to be extremely high (for comparison, 80% conversion of 10% acid-pretreated corn stover yields greater than 100 mM glucose), particularly at late-stage hydrolysis. Considering that this stage of hydrolysis also represents the most recalcitrant of the residual substrate, production of cellobiose is slow; consequently a maximally efficient -glucosidase is an absolute requirement for complete conversion (SimoÄes et al., 2007). While rational enzyme design methods including site-directed mutagenesis and domain-swapping are well-established mechanisms for enzyme improvement, they are necessarily limited in scope and are time-intensive. Specific alterations in highly active enzymes to match residues or domains in thermostable homologs, or swapping and mutagenesis of CBDs, are common practice (Ito et al., 2008). Conversely, directed evolution is a tool that can achieve rapid improvement if a screen can be adequately devised and if suitably large libraries can be developed (Wang et al., 2005; Xia and Wang, 2009). Fluorogenic or chromogenic substrates, including 4-methyl-umbelliferyl- -D-lactopyranoside (MUL) and 4-methyl-umbelliferyl- -D-cellobioside (MUC), are commonly utilized in activity screening. Barriers to the effective utilization of this technique include the slow rate of cellulose hydrolysis on native cellulosic substrates, which in turn yields very slow results; and the more rapid results obtained using chromogenic substrates, the results of which accurately reflect neither performance on native, crystalline cellulose, nor performance on high solids (Kabel et al., 2006). Another significant obstacle to this technique is the difficulty in obtaining heterologous expression in more genetically pliable organisms such as E. coli, yeast or phage, with which large libraries can be easily generated, particularly given the extensive glycosylation common to fungal cellulases.

7.3.4

Fungal expression systems

At the intersection of the two broad research foci of protein optimization and diversity screening is fungal Molecular Biology. As screening identifies new factors and engineering provides optimized factors, these must ultimately be expressed in high yield at optimal ratios with other cellulase components (Mach

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

191

and Zeilinger, 2003; Stricker et al., 2008). Improved enzyme activity and hence lower enzyme dosing is important, but high-level expression is also required to allow adoption of new components and new mixtures. Multiple expressions are possible, but a single, high-yielding expression host for cellulase production may ultimately prove to be the most economically favored (McFarland et al., 2007). Developing this host requires advanced molecular genetics tools, selectable markers and promoter control. Furthermore, strain mutagenesis, already having produced hyper-cellulolytic strains including PC-3-7, RUT-C30 and CL847, represents a reliable method for enhancing protein yields and hence reducing cost, regardless of the outcome of protein improvement efforts.

7.3.5

Synthetic cellulase mixtures

In contrast to the conventional method of supplementing or enhancing components of a secreted cellulase mixture, a straightforward and direct method of cellulase development is the generation of artificial enzyme mixtures from purified or partially purified components (Boisset et al., 2001). This approach has the advantage of providing complete control over the cellulases present, tailoring their relative compositions to optimal ratios and enabling the use of components from multiple sources and thereby permitting the precise matching of enzyme mixtures to the substrate being hydrolyzed. This also has the benefit of avoiding complications arising from heterologous expression and posttranslational modification. The weaknesses inherent to this approach include the exclusion of minor cellulase components, and these may, in net, contribute significantly to overall hydrolysis or provide synergistic enhancement to the principal cellulases, and the higher costs associated with parallel fermentations and additional formulation.

7.4

Recent developments

Several significant recent advances have been made toward achieving higher conversion at lower enzyme doses. Chief among these are the discovery and development of novel EG (Baker et al., 2005), CBH-I (Larenas et al., 2005; Goedegebuur et al., 2008) and CBH-II (Heinzelman et al., 2009) enzymes with enhanced hydrolysis rates or thermostability. Major gains have been achieved by direct supplementation of T. reesei cellulase mixtures with exogenous glucosidase, xylanase and family 61 glycohydrolases (Selig et al., 2008; Merino and Cherry, 2007; Brown et al., 2008a,b; Karkehabadi et al., 2008; Koseki et al., 2008; Teter and Cherry, 2005). The relative concentration of residual hemicellulose varies with biomass pretreatment method. Clearly, as the concentration and variety of residual hemicellulose increase, the concentration and breadth of specificity of the hemicellulases that must be included in the cellulase mixture increase correspondingly. Surprisingly, however, addition of hemicellulases

ß Woodhead Publishing Limited, 2010

192

Bioalcohol production

may strongly enhance saccharification even in those cellulosic feedstocks with low hemicellulose content (Selig et al., 2008; Rosgaard et al., 2006; Gupta et al., 2008). Supplementation of the T. reesei cellulase mixture with family 61 glycohydrolases has been shown to augment hydrolysis on some biomass substrates, though the mechanism underlying this enhancement is unclear.

7.5

Issues in cellulase development

Several of the most critical issues in the development of highly active, low cost cellulases have been highlighted in previous sections. It is clear that many of these active questions involve obstacles inherent to the substrate, rather than to the cellulases per se. The principal task, therefore, is to discover or enhance the activity of enzymes necessary to overcome the complications provided by both the cellulosic substrate and the process conditions necessary for profitability (SimoÄes et al., 2007; Saleem et al., 2005). Regardless of the pretreatment method utilized, lignin, hemicellulose and byproducts specific to the pretreatment will be present in the cellulosic substrate and the enzyme mixture must be robust to those components. Significant and unique barriers to enzymatic saccharification will exist for any pretreatment technique that ultimately proves process-amenable, and the cellulase mixture must mirror the specific set of inhibitors inherent to the pretreatment. As the variety of potential commercially relevant biomass pretreatments has thus far remained broad, commercial cellulase development has been hindered by ambiguity surrounding the inhibitor and substrate spectrum likely to be encountered. It should also be noted that compositional variation in biomass is broad, depending heavily on the growth conditions year-to-year, and intuitively but not necessarily obviously, biomass of equivalent composition is not equivalently hydrolyzable, even following equivalent pretreatment. It has become clear that biomass digestibility depends not only on the substrate composition, but on the organization of those components. Regardless, if the limitation is imposed that cellulose conversion be quantitative, or nearly so, a fixed mass of lignocellulosic feedstock will, as hydrolysis progresses, become more recalcitrant, lignin and hemicellulose will comprise a greater fraction of residual biomass, and the concentration of soluble sugars will increase. The obvious challenges therefore include mitigation of non-productive binding to hemicellulose or lignin, reduction of product inhibition and facilitation of cellulase turnover on more crystalline substrate. Less apparent issues include side-chain cross-linking of cellulose and esterification of xylan potentially acting as roadblocks to efficient, processive hydrolyses. Methods for the enzymatic removal of these roadblocks must therefore be realized. Further challenges arise from the process conditions under which enzyme function is likely to be required ± of these, temperature, hydrolysis time and pH

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

193

have been discussed. Additionally, the physical properties of high solids including high viscosity, hydrophobic packing of lignin and cellulose, significant requirements for waters of hydration at the cellulose-water interface and for solvation of soluble saccharides lead to reduction of enzyme accessible area and diffusion rates. Consequently, the enzymes must be robust to shear resulting from the mixing necessary for extensive saccharification at high solids (Jùrgensen et al., 2007). Moreover, mutual reinforcement of these issues occurs as the high solids condition necessary for bioethanol production exacerbates the inhibitory effects that arise from the inhibitory contents of the substrate. Arguably the single greatest challenge in the development of cellulases for second generation bioethanol is uncertainty of the conditions under which the enzymes will be required to function and the pretreatment techniques that ultimately will be utilized. Enzyme improvement is an attainable goal when the targets and parameters required for enhancement are well-defined. The biomass substrate, the pretreatment technique and the process conditions under which the lignocellulosic feedstock is likely to be hydrolyzed are only now becoming clear, and significant advances in the field are anticipated.

7.6

Future trends

As the second generation bioethanol market matures, new processes are developed and not yet existing processes are improved, research in this field may shift focus in several different areas. First among them is the utilization of high-temperature hydrolysis, enabled by thermostable cellulases. Research is currently being performed with synthetic mixtures of cellulases from thermophilic and hyperthermophilic organisms (Viikari et al., 2007; Blumer-Schuette et al., 2008). Significant advantages to high temperature saccharification include faster catalytic rate, reduced risk of contamination from bacterial or fungal sources, and reduced substrate viscosity. Secondly, SSF processes are an area of considerable current research, and while these have largely been left for discussion elsewhere in this volume, the advantages of SSF are widely recognized. At present, this approach is limited by the cellulase performance under the suboptimal conditions, ~34 ëC and pH values above 5.5, required for fermentation by conventional S. cerevisiae strains. This, then, represents a well-defined challenge for both protein optimization and conventional molecular biology; development of enzymes with higher pH-optima and resistant to ethanoldenaturation and the parallel development of ethanologenic yeast or bacterial strains capable of C5 and C6 fermentation at higher temperatures. Consolidated bioprocessing is also an area of active research, and the development of ethanolinsensitive, active cellulases for heterologous expression in ethanologenic microorganisms is anticipated (Den Haan et al., 2007; van Zyl et al., 2007). Most significantly, as libraries of cellulases continue to grow and our understanding of the differences in biomass feedstock for various bioethanol

ß Woodhead Publishing Limited, 2010

194

Bioalcohol production

producers and even regional variations in biomass composition improves, it is possible that enzyme mixtures will be tailored empirically to maximize performance (Rosgaard et al., 2006). Once process parameters are in place and profitability has been demonstrated for the first second generation feedstocks, new biomass feedstocks will undoubtedly arise, each with its own protein design challenges. Anchored with an efficient, generic cellulase mixture, specific cellulases and enhancing factors may be utilized as supplements, maximizing saccharification yields and thereby minimizing costs for bioethanol producers exploiting a broad range of biomass feedstocks.

7.7

References and further reading

Aden, A, Ruth, M, Ibsen, K, Jechura, K, Neeves, K, Sheehan, J, Wallace, B, Montague, L, Slayton, A and Lukas, J 2002, Lignocellulosic Biomass to Ethanol Process Design and Economics Utilizing Co-Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis for Corn Stover, NREL/TP-510-32438, Ai-Zuhair, S 2008, `The effect of crystallinity of cellulose on the rate of reducing sugars production by heterogeneous enzymatic hydrolysis', Bioresource Technology, 99(10), 4078±4085. Baker, JO, Ehrman, CI, Adney, WS, Thomas, SR and Himmel, ME 1998, `Hydrolysis of cellulose using ternary mixtures of purified cellulases', Applied Biochemistry and Biotechnology, 70±72, 395±403. Baker, JO, McCarley, JR, Lovett, R, Yu, CH, Adney, WS, Rignall, TR, Vinzant, TB, Decker, SR, Sakon, J and Himmel, ME 2005, `Catalytically enhanced endocellulase Cel5A from Acidothermus cellulolyticus', Applied Biochemistry and Biotechnology, 121, 129±148. Baldrian, P and ValaÂsÏaskovaÂ, V 2007, `Degradation of cellulose by basidiomycetous fungi', FEMS Microbiology Reviews, 32, 501±521. Banci, L, Ciofi-Baffoni, S and Tien, M 1998, `Lignin and Mn peroxidase-catalyzed oxidation of phenolic lignin oligomers', Biochemistry, 38, 3205±3210. Bashkirova, E, Rey, MW and Berka, RM 2005, `Combination of suppression subtraction hybridization and microarray technologies to enumerate biomass-induced genes in the cellulolytic fungus Trichoderma reesei', in DK Arora and RM Berka (eds), Applied Mycology and Biotechnology Vol. 5, Genes and Genomics, pp. 275±299. Bayer, EA, Belaich, JP, Shoham, Y and Lamed, R 2004, `The cellulosomes: multienzyme machines for degradation of plant cell wall polysaccharides', Annual Review of Microbiology, 58, 521±554. Benoit, I, Danchin, EGJ, Bleichrodt, RJ, de Vries, RP 2008, `Biotechnological applications and potential of fungal feruloyl esterases based on prevalence, classification and biochemical diversity', Biotechnology Letters, 30, 387±396. Berlin, A, Gilkes, N, Kilburn, D, Bura, R, Markov, A, Skomarovsky, A, Okunev, O, Gusakov, A, Maximenko, V, Gregg, D, Sinitsyn, A and Saddler, J 2005, `Evaluation of novel fungal cellulase preparations for ability to hydrolyze softwood substrates ± evidence for the role of accessory enzymes' Enzyme and Microbial Technology, 37(2), 175±184. Berlin, A, Balakshin, M; Gilkes, N, Kadla, J, Maximenko, V, Kubo, S and Saddler, J 2006, `Inhibition of cellulase, xylanase and beta-glucosidase activities by softwood

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

195

lignin preparations', Journal of Biotechnology, 125(2), 198±209. Bezerra, RMF and Dias, AA 2005, `Enzymatic kinetic of cellulose hydrolysis ± inhibition by ethanol and cellobiose', Applied Biochemistry and Biotechnology, 126(1), pp. 49±59. Blake, AW, McCartney, L, Flint, JE, Bolam, DN, Boraston, AB, Gilbert, HJ and Knox, JP 2006, `Understanding the biological rationale for the diversity of cellulose-directed carbohydrate-binding modules in prokaryotic enzymes', Journal of Biology Chemistry, 281(39), 29321±29329. Blumer-Schuette, SE, Kataeva, I, Westpheling, J, Adams, MWW and Kelly, RM 2008, `Extremely thermophilic microorganisms for biomass conversion: status and prospects' Current Opinion in Biotechnology, 19, 210±219. Boisset, C, PeÂtrquin, C, Chanzy, H, Henrissat, B and SchuÈlein, M 2001, `Optimized mixtures of recombinant Humicola insolens cellulases for the biodegradation of crystalline cellulose' Biotechnology and Bioengineering, 72(3), 339±345. Boraston, AB, Bolam, DN, Gilbert, HJ and Davies, GJ 2004, `Carbohydrate-binding modules: fine-tuning polysaccharide recognition', Biochemical Journal, 382, 769781. Brown, K, Lopez de Leon, A and Ding, H 2008a, Polypeptides having cellulolytic enhancing activity and polynucleotides encoding same, US Patent 800298. Brown, K, Harris, P, Zaretsky, E, Re, E, Vlasenko, E, McFarland, KC and Lopez de Leon, A 2008b, Polypeptide from a cellulolytic fungus having cellulolytic enhancing activity, US Patent 7361495. Cantarel, BL, Coutinho, PM, Rancurel, C, Bernard, T, Lombard, V and Henrissat, B 2008, `The Carbohydrate-Active EnZymes database (CAZy): an expert resource for Glycogenomics', Nucleic Acids Research, 37, D233±238. Caspi, J, Irwin, D, Lamed, R, Li, YC, Fierobe, HP, Wilson, DB and Bayer, EA 2008, `Conversion of Thermobifida fusca free exoglucanases into cellulosomal components: comparative impact on cellulose- degrading activity', Journal of Biotechnology, 135(4), 351±357. Chandra, RP, Bura, R, Mabee, WE, Berlin, A, Pan, X and Saddler, JN 2007, `Substrate pretreatment: the key to effective enzymatic hydrolysis of lignocellulosics', in L Olssen (ed.), Advances in Biochemical Engineering/Biotechnology, SpringerVerlag, Berlin, Heidelberg, pp. 67±93. Cherry, J and Fidantsef, A 2003, `Directed evolution of industrial enzymes: an update', Current Opinion in Biotechnology, 14, 438±443. Chiba, S 1997, `Molecular mechanism in -glucosidase and glucoamylase', Bioscience, Biotechnology and Biochemistry, 61, 1233±1239. Coutinho, P and Henrissat, B 1999, `Carbohydrate-active enzymes: an integrated database approach', in HJ Gilbert, G Davies, B Henrissat and B Svensson (eds.), Recent Advances in Carbohydrate Bioengineering, pp. 3±14. D'Aquino, R 2007, Cellulosic ethanol ± tomorrow's sustainable energy source, AICHE update, http://www.aiche.org/uploadedFiles/Energy_Website/Publications/ 070308_Cellulosic_Ethanol.pdf Davies, G and Henrissat, B 1995, `Structures and mechanisms of glycosyl hydrolases', Structure, 3, 853±859. Den Haan, R, Rose, SH, Lynd, LR and van Zyl, WH 2007, `Hydrolysis and fermentation of amorphous cellulose by recombinant Saccharomyces cerevisiae', Metabolic Engineering, 9(1), 87±94. Ding, SY, Xu, Q, Crowley, M, Zeng, Y, Nimlos, M, Lamed, R, Bayer, EA and Himmel,

ß Woodhead Publishing Limited, 2010

196

Bioalcohol production

ME 2008, `A biophysical perspective on the cellulosome: new opportunities for biomass conversion' Current Opinion in Biotechnology, 19(3), 218±227. Dumon, C, Varvak, A, Wall, MA, Flint, JE, Lewis, RJ, Lakey, JH, Morland, C, Luginbuhl, P, Healey, S, Todaro, T, DeSantis, G, Sun, M, Parra-Gessert, L, Tan, XQ, Weiner, DP and Gilbert, HJ 2008, `Engineering hyperthermostability into a GH11 xylanase is mediated by subtle changes to protein structure', Journal of Biology Chemistry, 283(33), 22557±22564. Fierobe, HP, Mingardon, F, Mechaly, A, Belaich, A, Rincon, MT, Pages, S, Lamed, R, Tardif, C, Belaich, JP and Bayer, EA 2005, `Action of designer cellulosomes on homogeneous versus complex substrates ± Controlled incorporation of three distinct enzymes into a defined trifunctional scaffolding', Journal of Biological Chemistry, 280(16), 16325±16334. Gilbert, HJ 2007, `Cellulosomes: microbial nanomachines that display plasticity in quaternary structure', Molecular Microbiology, 63(6), 1568±1576. Goedegebuur, F, Gualfetti, P, Mitchinson, C and Neefe, P 2008, CBH1 homologs and variant CBH1 cellulases, US Patent 7452707. Gupta, R, Kim, TH and Lee, YY 2008, `Substrate dependency and effect of xylanase supplementation on enzymatic hydrolysis of ammonia-treated biomass', Applied Biochemistry and Biotechnology, 148(1±3), 59±70. Gusakov, AV, Salanovich, TN, Antonov, AI, Ustinov, BB, Okunev, ON, Burlingame, R, Emalfarb, M, Baez, M and Sinitsyn, AP 2007, `Design of highly efficient cellulase mixtures for enzymatic hydrolysis of cellulose', Biotechnology and Bioengineering, 97(5), 1028±1038. Heinzelman, P, Snow, CD, Wu, I, Nguyen, C, Villalobos, A, Govindarajan, S, Minshull, J and Arnold, FH 2009, `A family of thermostable fungal cellulases created by structure-guided recombination', Proceedings of the National Academies of Science of the USA, 106(14), 5610±5615. Henrissat, B 1991, `A classification of glycosyl hydrolases based on amino-acid sequence similarities', Biochemical Journal, 280, 309±316. Henrissat, B and Bairoch, A 1993, `New families in the classification of glycosyl hydrolases based on amino-acid sequence similarities', Biochemical Journal, 293, 781±788. Henrissat, B and Bairoch, A 1996, `Updating the sequence-based classification of glycosyl hydrolases' Biochemistry Journal, 316, 695±696. Henrissat, B and Davies, G, 1997, `Structural and sequence-based classification of glycoside hydrolases', Current Opinion in Structural Biology, 7(5), 637±44. Henrissat, B, Driguez, H, Viet, C and SchuÈlein, M 1985, `Synergism of cellulases from Trichoderma reesei in the degradation of cellulose', Bio/Technol, 3(8), 722±726. Herpoel-Gimbert, I, Margeot, A, Dolla, A, Jan, G, MolleÂ, D, Lignon, S, Mathis, H, Sigoillot, JC, Monot, F and Asther, M 2008, `Comparative secretome analysis of two Trichoderma reesei RUT-C30 and CL847 hypersecretory strains', Biotechnology for Biofuels, 1, 18. Hess, M 2008, `Thermoacidophilic proteins for biofuel production', Trends in Microbiology, 16(9), 414±419. HildeÂn, L and Johansson, G 2004, `Recent developments on cellulases and carbohydratebinding modules with cellulose affinity', Biotechnology Letters, 26(22), 1683± 1693. Ito, J, Fujita,Y, Ueda, M, Fukuda, H and Kondo, A 2008, `Improvement of cellulosedegrading ability of a yeast strain displaying Trichoderma reesei endoglucanase II by recombination of cellulose-binding domains', Biotechnology Progress, 20(3),

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

197

688±691. Jeoh, T, Ishizawa, CI, Davis, MF, Himmel, ME, Adney, WS and Johnson, DK 2007, `Cellulase digestibility of pretreated biomass is limited by cellulose accessibility', Biotechnology and Bioengineering, 98(1), 112±122. Jùrgensen, H, Vibe-Pedersen, J, Larsen, J and Felby, C 2007, `Liquefaction of lignocellulose at high-solids concentrations', Biotechnology and Bioengineering, 96(5), 862±870. JuhaÂsz, T, Szengyel, Z, ReÂczey, K, Siika-Aho, M and Viikari, L 2005, `Characterization of cellulases and hemicellulases produced by Trichoderma reesei on various carbon sources', Process Biochemistry, 40(11), 3519±3525. Kabel, MA, van der Maarel, MJEC, Klip, G, Voragen, AGJ and Schols, HA 2006, `Standard assays do not predict the efficiency of commercial cellulase preparations towards plant materials', Biotechnology and Bioengineering, 93(1), 56±63. Kadam, KL, Rydholm, EC and McMillan, JD 2004, `Development and validation of a kinetic model for enzymatic saccharification of lignocellulosic biomass', Biotechnology Progress, 20, 698±705. Karkehabadi, S, Hansson, H, Kim, S, Piens, K, Mitchinson, C and Sandgren, M 2008, `The first structure of a glycoside hydrolase family 61 member, Cel61B from Hypocrea jecorina, at 1.6 angstrom resolution', Journal of Molecular Biology, 383(1), 144±154. Kersten, P and Cullen, D 2007, `Extracellular oxidative systems of the lignin-degrading Basidiomycete Phanerochaete chrysosporium', Fungal Genetics and Biology, 44, 77±87. Kipper, K, VaÈljamaÈe, P and Johansson, G 2005, `Processive action of cellobiohydrolase Cel7A from Trichoderma reesei is revealed as Ãburst¨ kinetics on fluorescent polymeric model substrates', Biochemical Journal, 385, 527±535. Koseki T, Mese, Y, Fushinobu, S, Masaki, K, Fujii, T, Ito, K, Shiono, Y, Murayama, T and Iefuji, H 2008, `Biochemical characterization of a glycoside hydrolase family 61 endoglucanase from Aspergillus kawachii', Applied Microbiology and Biotechnology, 77(6), 1279±1285. KrusaÊ, M, Henriksson, G, Johansson, G, Reitberger, T and Lennholm, H 2005, `Oxidative cellulose degradation by cellobiose dehydrogenase from Phanerochaete chrysosporium: a model compound study', Holzforschung, 59, 263±268. KrusaÊ, M, Lennholm, H and Henriksson, G 2008, `Pre-treatment of cellulose by cellobiose dehydrogenase increases the degradation rate by hydrolytic cellulases', Cellulose Chemistry and Technology, 41, 105±111. Kumar, R, Singh, S and Singh, OV 2008, `Bioconversion of lignocellulosic biomass: biochemical and molecular perspectives', Journal of Industrial Microbiology and Biotechnology, 35(5), 377±391. Larenas, EA, Goedegebuur, F, Gualfetti, P and Mitchinson, C 2005, Variant Humicola grisea CBH1.1, US Patent 2005/0037459. Lee, SS, Yu, S and Withers, SG 2005, `Mechanism of action of exo-acting -1,4-glucan lyase: a glycoside hydrolase family 31 enzyme', Biologia, Bratislava, 60, Suppl. 16, 137±148. Li, LL, McCorkle, SR, Monchy, S, Taghavi, S and van der Lelie, D 2009, `Bioprospecting metagenomes: glycosyl hydrolases for converting biomass', Biotechnology for Biofuels, 2, 10. Lin, Y and Tanaka, S 2006, `Ethanol fermentation from biomass resources: current state and prospects', Applied Microbiology and Biotechnology, 69, 627±642. Ma, AZ, Hu, Q, Qu, YB, Bai, ZH, Liu, WF and Zhuang, GQ 2008, `The enzymatic

ß Woodhead Publishing Limited, 2010

198

Bioalcohol production

hydrolysis rate of cellulose decreases with irreversible adsorption of cellobiohydrolase I', Enzyme and Microbial Technology, 42(7), 543±547. Mach, RL and Zeilinger, S 2003, `Regulation of gene expression in industrial fungi: Trichoderma' Applied Microbiology and Biotechnology, 60, 515±522. Mansfield, SD, DeJong, E and Saddler, JN 1997, `Cellobiose dehydrogenase, an active agent in cellulose depolymerization', Applied and Environmental Microbiology, 63(10), 3804±3809. Martinez, D, Larrondo, LF, Putnam, N, Gelpke, MD, Huang, K, Chapman, J, Helfenbein, KG, Ramaiya, P, Detter, JC, Larimer, F, Coutinho, PM, Henrissat, B, Berka, R, Cullen, D and Rokhsar, D 2004, `Genome sequence of the lignocellulose degrading fungus Phanerochaete chrysosporium strain RP78', Nature Biotechnology, 22(6), 695±700. Martinez, D, Berka, RM, Henrissat, B, Saloheimo, M, Arvas, M, Baker, SE, Chapman, J, Chertkov, O, Coutinho, PM, Cullen, D, Danchin, EGJ, Grigoriev, IV, Harris, P, Jackson, M, Kubicek, CP, Han, CS, Ho, I, Larrondo, LF, de Leon, AL, Magnuson, JK, Merino, S, Misra, M, Nelson, B, Putnam, N, Robbertse, B, Salamov, AA, Schmoll, M, Terry, A, Thayer, N, Westerholm-Parvinen, A, Schoch, CL, Yao, J, Barabote, R, Nelson, MA, Detter, C, Bruce, D, Kuske, CR, Xie, G, Richardson, P, Rokhsar, DS, Lucas, SM, Rubin, EM;, Dunn-Coleman, N, Ward, M and Brettin, TS 2008, `Genome sequencing and analysis of the biomass-degrading fungus Trichoderma reesei (syn. Hypocrea jecorina)', Nature Biotechnology, 26(5), 553±560. Martinez, D, Challacombe, J, Morgenstern, I, Hibbett, D, Schmoll, M, Kubicek, CP, Ferreira, P, Ruiz-Duenas, FJ, Martinez, AT, Kersten, P, Hammel, KE, Vanden Wymelenberg, A, Gaskell, J, Lindquist, E, Sabat, G, BonDurant, SS, Larrondo, LF, Canessa, P, Vicuna, R, Yadav, J, Doddapaneni, H, Subramanian, V, Pisabarro, AG, Lavin, JL, Oguiza, JA, Master, E, Henrissat, B, Coutinho, PM, Harris, P, Magnuson, JK, Baker, SE, Bruno K, Kenealy, W, Hoegger, PJ, KuÃes, U, Ramaiya, P, Lucas, S, Salamov, A, Shapiro, H, Tu, H, Chee, CL, Misra, M, Xie, G, Teter, S, Yaver, D, James, T, Mokrejs, M, Pospisek, M, Grigoriev, IV, Brettin, T, Rokhsar, D, Berka, R and Cullen, D 2009, `Genome, transcriptome, and secretome analysis of wood decay fungus Postia placenta supports unique mechanisms of lignocellulose conversion', Proceedings of the National Academy of Sciences of the USA, 106(6), 1954±1959. McAloon, A, Taylor, F, Yee, W, Ibsen, K and Wooley, R 2000, Determining the Cost of Producing Ethanol from Corn Starch and Lignocellulosic Feedstocks, US Department of Agriculture and US Department of Energy McFarland, KC, Ding, H, Teter, S, Vlasenko, E, Xu, F and Cherry, J 2007, `Development of improved cellulase mixtures in a single production organism', in G Eggleston and JR Vercellotti (eds), ACS Symposium Series, 972, Industrial applications of enzymes based on carbohydrate-based material, Oxford University Press, New York, pp. 19±31. Merino, S and Cherry, J 2007, `Progress and Challenges in Enzyme Development for Biomass Utilization', in L Olssen and BK Ahring (eds), Advanced Biochemical Engineering and Biotechnology, 108, 95±120. Mingardon, F, Chanal, A, LoÂpez-Contreras, AM, Dray, C, Bayer, EA and Fierobe, HP 2007a, `Incorporation of fungal cellulases in bacterial minicellulosomes yields viable, synergistically acting cellulolytic complexes, Applied and Environmental Microbiology, 73(12), 3822±3832.

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

199

Mingardon, F, Chanal, A, Tardif, C, Bayer, EA and Fierobe, HP 2007b, `Exploration of new geometries in cellulosome-like chimeras', Applied and Environmental Microbiology, 73(22), 7138±7149. Mosier, NS, Hall, P, Ladisch, CM and Ladisch, MR 1999, `Reaction kinetics, molecular action, and mechanisms of cellulolytic proteins', Advances in Biochemical Engineering and Biotechnology, 65, 23±40. Nimlos, MR, Matthews, JF, Crowley, MF, Walker, RC, Chukkapalli, G, Brady, JV, Adney, WS, Clearyl, JM, Zhong, LH and Himmel, ME 2007, `Molecular modeling suggests induced fit of Family I carbohydrate-binding modules with a broken-chain cellulose surface', Protein Engineering Design and Selection, 20(4), 179±187. Olsson, L, Christensen, TMIE, Hansen, KP and Palmqvist, EA 2003, `Influence of the carbon source on production of cellulases, hemicellulases and pectinases by Trichoderma reesei Rut C-30', Enzyme and Microbial Technology, 33(5), 612±619. Payasi, A, Sanwal, R and Sanwal, GG 2009, `Microbial pectate lyases: characterization and enzymological properties', World Journal of Microbiology and Biotechnology, 25, 1±4. Qin, YQ, Wei, XM, Song, X and Qu, YB 2008, `Engineering endoglucanase II from Trichoderma reesei to improve the catalytic efficiency at a higher pH optimum', Journal of Biotechnology, 135(2), 190±195. Rabinovich, ML, Melnik, MS and Bolobova, AV 2002, `The structure and mechanism of action of cellulolytic enzymes', Biochemistry (Moscow), 67(8), 850±871. Renewable Fuels Association 2009, US fuel ethanol industry biorefineries and capacity, Renewable Fuels Association, Rosgaard, L, Pedersen, S, Langston, J, Akerhielm, D, Cherry, JR and Meyer, AS 2006, `Evaluation of minimal Trichoderma reesei cellulase mixtures on differently pretreated barley straw substrates', Biotechnology Progress, 22, 493±498. Rubin, E 2008, `Genomics of cellulosic biofuels', Nature, 454(7206), 841±845. Saleem, M, Rashid, MH, Jabbar, A, Perveen, R, Khalid, AM and Rajoka, MI 2005, `Kinetic and thermodynamic properties of an immobilized endoglucanase from Arachniotus citrinus', Process Biochemistry, 40(2), 849±855. SchuÈlein, M 2000, `Protein engineering of cellulases', Biochimica Biophysica Acta, 1543, 239±252. Schwartzkopff, F 2008, Forget Corn: Mushrooms May Hold Key to Energy Crisis, Bloomberg, . Selig, MJ, Viamajala, S, Decker, SR, Tucker, MP, Himmel, ME and Vinzant, TB 2007, `Deposition of lignin droplets produced during dilute acid pretreatment of maize stems retards enzymatic hydrolysis of cellulose', Biotechnology Progress, 23, 1333±1339. Selig, MJ, Knoshaug, EP, Adney, WS, Himmel, ME and Decker, SR 2008, `Synergistic enhancement of cellobiohydrolase performance on pretreated corn stover by addition of xylanase and esterase activities', Bioresource Technology, 99(11), 4997±5005. SimoÄes, R, Leite, R, Gomes, E and da Silva, R 2007, `Characterization and comparison of thermostability of purified beta-glucosidases from a mesophilic Aureobasidium pullulans and a thermophilic Thermoascus aurantiacus', Process Biochemistry, 42(7), 1101±1106. Singh, D and Chen, S 2008, `The white-rot fungus Phanerochaete chrysosporium: conditions for the production of lignin-degrading enzymes', Applied Microbiology and Biotechnology, 81, 399±417.

ß Woodhead Publishing Limited, 2010

200

Bioalcohol production

Sternberg, D, Vijayakumar, P and Reese, E 1977, ` -Glucosidase: microbial production and effect on enzymatic hydrolysis of cellulose', Canadian Journal of Microbiology, 23, 139. Stricker, A, Mach, RL and de Graaff, LH 2008, `Regulation of transcription of cellulasesand hemicellulases-encoding genes in Aspergillus niger and Hypocrea jecorina (Trichoderma reesei)', Applied Microbiology and Biotechnology, 78, 211±220. Teter, S and Cherry, J 2005, Improving Cellulose Hydrolysis with New Cellulase Compositions, AICHE annual meeting, . Teter, S, Lamsa, M, Cherry, J and Ward, C 2006, Methods for transforming and expression screening of filamentous fungal cells with a DNA library, US Patent Application 20060172379. Toyama, H, Nakano, M, Satake, Y and Toyama, N 2008, `Rapid isolation of the Trichoderma strain with higher degrading ability of a filter paper and superior proliferation characteristics using avicel plates and the double-layer selection medium', Applied Biochemistry and Biotechnology, 145(1±3), 23±28. Tu, M, Chandra, RP and Saddler, JN, 2007, `Evaluating the distribution of cellulases and the recycling of free cellulases during the hydrolysis of lignocellulosic substrates', Biotechnology Progress, 23, 398±406. US Department of Energy 2006, Breaking the Biological Barriers to Cellulosic Ethanol: A Joint Research Agenda, DOE/SC-0095 . US Department of Energy 2008a, Joint Genome Institute 2008 Progress Report, LBNL1597E, CSO 15866, . US Department of Energy 2008b, BioEnergy Research Centers: an overview of the science, DOE/SC-0104, < http://genomicsgtl.energy.gov/centers/brcbrochure.pdf>. ValaÂsÏaskovaÂ, V and Baldrian, P 2006, `Estimation of bound and free fractions of lignocellulose- degrading enzymes of wood-rotting fungi Pleurotus ostreatus, Trametes versicolor and Piptoporus betulinus', Research in Microbiology, 157(2), 119±124. VaÈljamaÈe, P, Kipper, K, Pettersson, G and Johansson G 2003, `Synergistic cellulose hydrolysis can be described in terms of fractal-like kinetics', Biotechnology and Bioengineering, 84(2), 254±257. Vanden Wymelenberg, A, Gaskell, J, Mozuch, M, Kersten, P, Sabat, G, Martinez, D and Cullen, D 2009, `Transcriptome and secretome analysis of Phanerochaete chrysosporium reveal complex patterns of gene expression' Applied Environmental Microbiology, 75(12), 4058±4068. van Zyl, WH, Lynd, LR, den Haan, R and McBride, JE 2007, `Consolidated bioprocessing for bioethanol production using Saccharomyces cerevisiae', Advanced Biochemical Engineering and Biotechnology, 108, 205235 Vasella, A, Davies, GJ and Bohm, M 2002, `Glycosidase mechanisms', Current Opinion in Chemical Biology, 6, 619±629. Vivian, L, Yip, Y and Withers, SG 2006, `Family 4 glycoside hydrolases are special: the first ±elimination mechanism amongst glycoside hydrolases', Biocatalysis and Biotransformation, 24, 167±176. Viikari, L, Alauranen, M, Puranen, T, VehmaanperaÈ, J and Siika-aho, M 2007, `Thermostable enzymes in lignocellulose hydrolysis', in L Olssen (ed.), Advances in Biochemical Engineering/Biotechnology, Springer-Verlag, Berlin, Heidelberg, vol. 108, pp. 121±145.

ß Woodhead Publishing Limited, 2010

Development of cellulases to improve enzymatic hydrolysis

201

Vlasenko, E, Xu, F and Cherry, J 2003, `Thermostability, substrate specificity and hydrolysis of cellulose by endoglucanases from families 5, 6, and 45 of glycoside hydrolases, 25th Symposium on Biotechnology for Fuels and Chemicals, Breckenridge, CO. Vocadlo, DJ and Davies, GJ 2008, `Mechanistic insights into glycosidase chemistry', Current Opinion in Chemical Biology, 12(5), 539±555. Voutilainen, SP, Boer, H, Linder, MB, Puranen, T, Rouvinen, J, Vehmaanpera, J and Koivula, A 2007, `Heterologous expression of Melanocarpus albomyces cellobiohydrolase Cel7B, and random mutagenesis to improve its thermostability', Enzyme and Microbial Technology, 41(3), 234±243. Voutilainen, SP, Puranen, T, Siika-Aho, M, Lappalainen, A, Alapuranen, M, Kallio, J, Hooman, S, Viikri, L, VehmaanperaÈ, J and Koivula, A 2008, `Cloning, expression, and characterization of novel thermostable family 7 cellobiohydrolases', Biotechnology and Bioengineering, 101(3), 515±528. Wang, T, Liu, X, Yu, Q, Zhang, X, Qu, Y, Gao, P and Wang, T 2005, `Directed evolution for engineering pH profile of endoglucanase III from Trichoderma reesei' Biomolecular Engineering, 22(1±3), 89±94. Warnecke, F and Hess, M 2009, `A perspective: metatranscriptomics as a tool for the discovery of novel biocatalysts', Journal of Biotechnology, 142(1), 91±95. Warnecke, F, LuginbuÈhl, P, Ivanova, N, Ghassemian, M, Richardson, TH, Stege, JT, Cayouette, M, McHardy, AC, Djordjevic, G, Aboushadi, N, Sorek, R, Tringe, SG, Podar, M, Martin, HG, Kunin, V, Dalevi, D, Madejska, J, Kirton, E, Platt, D, Szeto, E, Salamov, A, Barry, K, Mikhailova, N, Kyrpides, NC, Matson, EG, Ottesen, EA, Zhang, X, HernaÂndez, M, Murillo, C, Acosta, LG, Rigoutsos, I, Tamayo, G, Green, BD, Chang, C, Rubin, EM, Mathur, EJ, Robertson, DE, Hugenholtz, P and Leadbetter, JR 2007, `Metagenomics and functional analysis of hindgut microbiota of a wood-feeding higher termite', Nature, 450(7169), 560±569. Wood, P 1994, `Pathways for production of Fenton's reagent by wood-rotting fungi', FEMS Microbiology Reviews, 13, 313±320. Xia, T and Wang, Q 2009, `Directed evolution of Streptomyces lividans xylanase B toward enhanced thermal and alkaline pH stability', World Journal of Microbiology and Biotechnology, 25(1), 93±100. Xu, F and Ding, H 2007, `A new kinetic model for heterogeneous (or spatially confined) enzymatic catalysis: contributions from the fractal and jamming (overcrowding) effects', Applied Catalysis A: General, 317, 70±81. Zhang, YHP, Himmel, ME and Mielenz, JR 2006, `Outlook for cellulase improvement: Screening and selection strategies', Biotechnology Advances, 24(5), 452±481. Zhang, YHP and Lynd, LR 2006, `A functionally based model for hydrolysis of cellulose by fungal cellulase', Biotechnology and Bioengineering, 94(5), 888±898. Zhong, L, Matthews, JF, Crowley, MF, Rignall, T, TaloÂn, C, Cleary, JM, Walker, RC, Chukkapalli, G, McCabe, C, Nimlos, MR, Brooks, CL, Himmel, ME and Brady, JW 2008, `Interactions of the complete cellobiohydrolase I from Trichoderma reesei with microcrystalline cellulose 1 ', Cellulose, 15, 261±273.

ß Woodhead Publishing Limited, 2010

8

Integrated hydrolysis, fermentation and co-fermentation of lignocellulosic biomass P . M A N Z A N A R E S , CIEMAT, Spain

Abstract: This chapter examines the most important process configurations for bioethanol production from lignocellulosic biomass, focusing on integrated configurations as simultaneous saccharification and fermentation, its variant with co-fermentation and consolidated bioprocessing. Firstly, basics of process configurations are reviewed. Secondly, the influence of lignocellulose structure and different biomass pretreatments on process performance is discussed. Then, the chapter deals with the main options in the use of specific microbial strains for the different processes, ending with some future research trends on the subject. Key words: process integration, simultaneous saccharification and fermentation, co-fermentation, consolidated bioprocessing, lignocellulosic biomass.

8.1

Introduction

Ethanol production from lignocellulose is receiving major research attention due to its important potential for conversion to sugars and fuels. Its role on diversifying current production based on starchy or sugar-based biomass appears to be a key factor to boost implementation of bioethanol in the current fuel market. For fermentation of lignocellulosic materials, cellulose has to be degraded into glucose (hydrolysis or saccharification) using acid or enzymes. Although extensive work has been performed in the use of dilute acid to hydrolyze the cellulose polymer with good results (Torget et al., 2000; SoÈderstroÈm et al., 2002; Lloyd and Wyman, 2005), enzymatic hydrolysis has demonstrated better performance for subsequent fermentation because no degradation components of glucose are formed and a lower process temperature is required. The use of enzymatic hydrolysis in biomass to ethanol processes provides opportunities to improve this technology so that ethanol from biomass is competitive when compared to other liquid fuels (Wyman, 1999). In this respect, during the last years intensive research has been focused on the improvement of cellulolytic enzymes performance (specific activity, thermostability, combination of different enzymatic activities, etc.) aimed at reducing cellulases cost (Yang and Wyman, 2007). The high price of cellulolytic enzymes has often been

ß Woodhead Publishing Limited, 2010

206

Bioalcohol production

identified as a significant bottleneck in the scaling up of the enzyme-based biomass to ethanol processes (Wyman, 2001). When enzymatic hydrolysis is involved in the bioprocessing of biomass to ethanol, a pretreatment step is vital to effectively prepare cellulose for enzyme action and ensure high sugar yields. Hence, pretreatment has frequently been highlighted as one of the most influential, though costly process step, having a major influence on both prior (e.g., size reduction) and subsequent operations (e.g., enzymatic hydrolysis and fermentation) (Lynd, 1996). Numerous studies on bioprocesses for the production of ethanol from lignocellulosic biomass based on enzymatic hydrolysis are being carried out worldwide (Ballesteros et al., 2001; Hamelinck et al., 2005; GaÂspaÂr et al., 2007; SaÂnchez and Cardona, 2008). In spite of the large variety of existing and incompletely developed technologies, it is generally recognized that an opportunity to reduce production costs and improve feasibility of industrial application exists through the integration of the different operations within the whole process of ethanol production. This can be achieved through the development of integrated bioprocesses that combine different steps into one single unit. It is gaining more and more interest due to the advantages related to its application in the case of ethanol production: reduction of energy costs and decrease in the size and number of process units (Cardona and SaÂnchez, 2007). The most prominent and probably the most developed process configuration is that of simultaneous saccharification and fermentation (SSF), which can be considered a first effort in process integration in fuel-ethanol production. In SSF the enzymatic degradation of cellulose is combined with the fermentation of glucose obtained from hydrolysis. A further step in the integration pathway would be the inclusion of the pentose fermentation, which has been developed in the variant of SSF called simultaneous saccharification and co-fermentation (SSFC). And finally, the most integrated process configuration is represented in consolidated bioprocessing (CBP), in which only one microbial community is employed for both the production of hydrolytic enzymes and fermentation. This chapter aims to provide a succinct overview of the state of the art in the integrated process configurations for ethanol production from lignocellulose biomass. Fundamentals of SSF and CBP technologies are reviewed initially, while the influence of lignocellulose structure and different pretreatments on process performance is discussed in the second section. The third section deals with research advances in the use of specific microbial strains for the different processes and the chapter ends with some future research trends on the subject.

8.2

Biological processing of lignocellulose

Three main different process configurations can be applied to convert biomass to ethanol using cellulolytic enzymes (Fig. 8.1). Firstly, separated hydrolysis and fermentation (SHF), in which both steps are carried out in different reactors

ß Woodhead Publishing Limited, 2010

Integrated hydrolysis, fermentation and co-fermentation

207

8.1 Main process configurations in enzyme-based ethanol production from lignocellulosic biomass.

under optimal conditions. Hydrolysis of the pretreated substrate is carried out at an optimum enzyme temperature (about 50 ëC) and once completed, the resulting cellulose hydrolysate is fermented to ethanol at optimum temperature for fermenting yeast (32±35 ëC). Apart from the advantage of using optimal conditions for both process steps, SHF provides easier recycling of yeast in the fermentation reactor (Olsson et al., 2005). In spite of these advantages, the implementation of SSF, in which enzymatic hydrolysis of cellulose is performed with simultaneous fermentation of the sugars, has been reported as one of the most important advances related to the overall process of ethanol production from lignocellulosic feedstocks (Cardona and SaÂnchez, 2007). In the SSF process, the stages are virtually the same as in the SHF, except that both are performed in the same reactor. The concept of SSF was first defined by Takagi et al. (1977) and since then has been extensively applied to the bioconversion of lignocellulose. The key advantage of SSF is that the presence of yeast together with the cellulolytic enzyme complex reduces the accumulation of sugars within the reactor ± thereby increasing yield and saccharification rate with respect to SHF (Wyman and Hinman, 1990). An improved enzymatic hydrolysis yield allows a reduction in the enzyme dose, thereby contributing to a decrease in the final process cost. Moreover, SSF offers a simpler and easier operation and a reduced requirement for capital equipment since a single fermentor is used for the entire process ± thereby further curbing the investment costs. In SSF, the ethanol is present in the culture medium, which

ß Woodhead Publishing Limited, 2010

208

Bioalcohol production

causes the mixture to be less vulnerable to invasion by undesired microorganisms (Wyman, 1994). Nevertheless, it is also important to control possible inhibition effects of ethanol concentration on microorganisms. One main drawback of the SSF concerns the different optimal conditions, mainly pH and temperature, for hydrolysis and fermentation. One of the solutions proposed, based on the use of thermostable yeast strains capable to produce ethanol at higher temperatures is discussed below. A higher level of integration is found in the variation of SSF ± the so-called simultaneous saccharification and co-fermentation (SSCF). In SSCF, the use of mixed cultures or a recombinant microorganism able to ferment both type of sugars achieves co-fermentation of hexoses and pentoses. SSCF using mixed microorganism cultures of different compatibles species was studied in an initial stage (Olsson and Hahn-HaÈgerdal, 1996; Chandrakant and Bisaria, 1998). This type of configuration requires that both fermenting microorganisms are compatible in terms of operating pH and temperature. Moreover, problems related to the low growth rate of pentose fermenting yeasts in comparison to hexosefermenting yeast may arise, so resulting in more elevated conversion of hexoses than pentoses to ethanol. A variant of co-fermentation, consisting of using a single microorganism capable of assimilating both hexoses and pentoses, seems to be a most promising integration alternative. The use of genetically modified yeast or bacteria allows implementing processes resulting in high conversion and ethanol yields by using all the sugars contained in lignocellulosic biomass. The microorganisms most commonly modified for this purpose are Saccharomyces cerevisiae and Zymomonas mobilis, to which genes allowing assimilation of pentoses have been introduced, or ethanologenic bacteria like Escherichia coli, to which genes encoding the metabolic pathway for production of ethanol have been introduced. For example, this type of application has been recently reported by Kim et al. (2008), who tested the simultaneous saccharification and co-fermentation (SSCF) on pretreated barley hulls using recombinant E. coli (KO11), resulting in on overall ethanol yield of 89.4 % based on the glucan and xylan content of the substrate. Finally, the consolidated bioprocessing (CBP), known also as direct microbial conversion (DMC), in which only one microbial community is employed for both the production of cellulases and fermentation, i.e, cellulase production, cellulose hydrolysis and fermentation are carried out in a single step. In this case, a dedicated process for production of cellulases is not required, which makes CBP a highly integrated configuration (Lynd et al., 2005). According to Lynd, CBP offers the potential for lower cost and higher efficiency than processes featuring dedicated cellulase production (SHF, SSF and SSCF). This is principally due to the avoidance of costs associated with enzyme manufacturing. Moreover, the possibility of achieving higher hydrolysis rates through the use of thermophilic organisms and/or complexed cellulase systems contributes

ß Woodhead Publishing Limited, 2010

Integrated hydrolysis, fermentation and co-fermentation

209

8.2 Simplified scheme of consolidated bioprocessing concept. A) Native cellulolytic strategy. B) Recombinant cellulolytic strategy. HC, hemicellulose; C, cellulose; L, lignin. Processing pathway for a thermoanaerobic microorganism (e.g. Clostridium thermocellum) is indicated by continuous lines. Processing pathway for an ethanologenic microorganism (e.g. Zymomonas mobilis) is indicated by dotted lines. Reprinted from Cardona and Sa¨nchez (2007).

to the expectation of a reduction in reactor volume and capital investment in CBP. Lynd et al. (2002) pointed out that the feasibility of a CBP process would be established when a microorganism or microbial consortium could be developed according to one of two strategies (Fig. 8.2). As described by SaÂnchez and Cardona (2008), the first of these, termed `native cellulolytic strategy' (labeled `A' in Fig. 8.2) involves the improvement of the fermentative properties of a good producer of cellulases. The approach focuses on engineering microorganisms which already display a high native cellulolytic activity, in order to improve their production of ethanol through an increase in their alcohol yield. Such modifications include reducing or eliminating the production of byproducts as acetic acid or lactate, and increasing the ethanol tolerance and titres. The second approach, termed the `recombinant cellulolytic strategy' (labeled `B' in Fig. 8.2) seeks to transform a microorganism with good fermentative properties (high ethanol yields and tolerances) into a good producer of cellulases such that it is capable of utilizing cellulose within a CBP configuration. It is recognized that each strategy involves considerable uncertainty (Lynd et al., 2005), and they both merit investigation since they could be advantageous for different products. In the opinion of Yan and Wyman (2007), the CBP process configuration offers a potential pathway to low-cost ethanol through the use of a single step, but still faces important challenges in achieving high selectivity and yields using real substrates.

ß Woodhead Publishing Limited, 2010

210

8.3

Bioalcohol production

Feedstocks and pretreatments for simultaneous saccharification and fermentation (SSF)/ consolidated bioprocessing (CBP)

It is generally acknowledged that lignocellulose biomass can play an important role in the diversification of raw materials for fuel-ethanol production, which is currently based on food crops as sugarcane, corn or wheat. The strong dependency of fuel ethanol industry on food materials, together with the strategies adopted in many countries to expand its production and use in the near term, is leading to an intense debate food vs. fuel. In this context, the use of lignocellulosic materials, which are available at large scale at relatively low cost are envisaged as the most promising alternative feedstock for bioethanol production in the near future, so avoiding interference in the food market. According to SaÂnchez and Cardona (2008), prospective lignocellulosic materials for ethanol production can be divided into six main groups: crop residues (cane bagasse, corn stover, wheat straw, rice straw, rice hulls, barley straw, sweet sorghum bagasse, olive stones and pulp), hardwood (aspen and poplar), softwood (pine, spruce), cellulose wastes (newsprint, waste office paper, recycled paper sludge), herbaceous biomass (alfalfa hay, switchgrass, reed canary grass, coastal Bermudagrass, thymothy grass), and municipal solid wastes. During the last 20 years different research groups have tested most of these lignocelulosic feedstocks as substrates for ethanol production. For example, a summary of recent SSF studies on different raw materials can be found in the paper published by Olofsson et al. (2008). Results show that in most cases acceptable ethanol yields can be obtained (from 60 to 94% of theoretical, depending on the material and conditions) in a large variety of feedstocks from different sources (hardwoods, softwoods, waste paper, etc.), which highlights the potential for the biological conversion of lignocellulosic biomass to ethanol. The main limitation to the use of lignocellulose for ethanol production by biological processes such as SSF or CBP concerns the recalcitrant nature of this type of material, creating a strong barrier to cellulolytic enzymes. Cellulosic biomass is a heterogeneous complex of carbohydrate polymers (cellulose and hemicellulose) and lignin, a complex polymer of phenylpropane units. The specific structure of cellulose favors the ordering of the polymer chains into tightly packed highly crystalline structures that are resistant to depolymerization. Moreover, the carbohydrate polymers are tightly bound to lignin mainly by hydrogen bonds but also by some covalent bonds, so contributing to the recalcitrance of cellulosic biomass to hydrolysis (Carpita and Gibeaut, 1993; Brett and Waldron, 1996). It is necessary to overcome the physical and chemical barriers of cellulosic feedstocks in order to render them amenable to enzymatic hydrolysis. Therefore, process designs for biologically converting cellulosic materials nearly always include a pretreatment step.

ß Woodhead Publishing Limited, 2010

Integrated hydrolysis, fermentation and co-fermentation

211

In general, pretreatment greatly enhances enzyme-mediated hydrolysis in processes such as SSF. Nevertheless, it is apparent that the nature of the substrate and pretreatment method used greatly influences the effectiveness of the enzyme action. Different pretreatments induce different changes in lignocellulose. Furthermore, development of an ideal pretreatment process is difficult, given that lignocellulosic biomass is derived from a wide range of sources (Kim and Dale, 2004; Dien et al., 2007; SaÂnchez and Cardona, 2008). Pretreatment strategies have been categorized into biological, physical and chemical processes or a combination of these approaches. A comprehensive review of the pretreatment field is out of the scope of this work, and the reader is referred to dedicated chapters in this book or to some publications describing the diversity and mode of action of pretreatment processes (Jacobsen et al., 2000; Guo et al., 2008; Chandra et al., 2007; Hendriks and Zeeman, 2009). An example of coordinated development of leading biomass pretreatment technologies is the work performed by a consortium of researchers [Biomass Refining Consortium for applied Fundamental and Innovation (CAFI)] on a well-characterized single feedstock, namely corn stover (Wyman et al., 2005). Among the tested process technologies, Moiser et al. (2005) pointed out that the steam-explosion (SE), liquid hot water (LHW), dilute acid (DA), lime and ammonia fiber explosion (AFEX) treatment methods are good candidates as cost-effective pretreatments. The authors reviewed the fundamental modes of action for these pretreatment methods and reported on how they cause dissimilar effects on the chemical/physical structure of the lignocellulosic biomass, so resulting in differences in the performance of the pretreated materials in transformation processes as SSF. Based on the outcome of this work, Table 8.1 shows the effect of various pretreatment methods on the chemical/physical structure of Table 8.1 Effect of various pretreatment methods on the chemical/physical structure of lignocellulosic biomass. Adapted from Moiser et al. (2005) Increases Decrystallizes accessible cellulose surface area Uncatalyzed steam explosion Liquid hot water Dilute acid Flow-through acid AFEX Lime

‡ ‡ ‡ ‡ ‡ ‡

Removes Removes hemilignin cellulose ‡

ND ‡ ND

‡ ‡ ‡ ÿ ÿ

‡ Major effect ÿ Minor effect ND: Not determined

ß Woodhead Publishing Limited, 2010

Alters lignin structure ÿ

ÿ ÿ ‡ ‡

ÿ ‡ ‡ ÿ ‡

212

Bioalcohol production

lignocellulose biomass. In SE, DA and LHW some cellulose depolymerization occurs, while in the AFEX method cellulose decrystallizes to some extent. Solubilization of hemicelluloses is also a common fate in pretreatment technologies although the rate varies depending on the type of process. For example, SE, DA and LHW have been reported to solubilize 80±100% of the hemicellulose of raw material, while in pretreatments as AFEX or lime solubilization may vary from 0 to 60% depending on the raw material and pretreatment conditions (Lynd et al., 2002). Removal of hemicellulose increases the mean pore size of the substrate and therefore increases the extent for cellulose hydrolysis. The lignin content and the type of lignin also have a significant effect on the hydrolysis of various cellulosic substrates in processes such as SSF, since lignin acts as both a physical barrier, restricting access of cellulases to cellulose, and as an attractant to cellulases, resulting in non-productive binding (Chandra et al., 2007). Therefore, varying the pretreatment method or conditions employed to improve cellulose or hemicellulose yields will clearly affect the lignin content of substrate. AFEX pretreatment removes lignin from raw material. However, a large proportion of lignin remains in the solid phase after steam pretreatment or dilute acid pretreatment. Lignin redistribution is thought to explain why the latter pretreatments are effective processes for SSF even though the lignin is not actually removed. It is thought that lignin melts during pretreatment and coalesces upon cooling such that its properties are altered (Lynd et al., 2002). The treatment is also thought to result in a decrease in the adsorption capacity of lignin to cellulases, and in consequence in an increase in the adsorptive capacity of cellulose. As can be seen in Table 8.1, all selected methods result in an increased surface area accessible to cellulase enzymes (porosity) in relation to untreated material. Increased porosity results from a combination of hemicellulose solubilization, lignin solubilization, and lignin redistribution with the relative importance of these factors differing greatly among different pretreatment processes. The significant correlation between initial pore volume (porosity) of lignocellulose and their extent of hydrolysis has been recently reported by Chandra et al. (2007) and Zeng et al. (2007). It has been proposed that the efficacy of cellulases is improved when the pores are large enough to accommodate all the components of cellulase systems. The significance of pretreatment in processes featuring enzymatic hydrolysis as SSF or CBP is emphasized by Chandra et al. (2007), who pointed out that the effectiveness of pretreatment affects both the upstream selection of biomass, the efficiency of recovery of the cellulose, hemicellulose and lignin components, and the chemical and morphological characteristics of the resulting cellulosic component, which governs subsequent hydrolysis and fermentation steps in biological transformation processes. In general, optimal pretreatment conditions in an SSF process do not necessarily differ much from those of an SHF processes utilizing lignocellulosic

ß Woodhead Publishing Limited, 2010

Integrated hydrolysis, fermentation and co-fermentation

213

biomass. However, if the pretreatment used generates a liquid fraction, as e.g. in LHW, SE or dilute acid, several compounds present in pretreatment hydrolysates may affect the saccharification and fermentation steps. The liquid fraction is composed mainly of hemicellulose-derived components, either in monomeric or oligomeric form and, depending on the pretreatment conditions, a series of inhibitory compounds (furaldehydes, weak acids and phenolics) derived from sugar degradation, de-acetylation of hemicellulose and lignin breakdown, respectively. In SSF, some of these compounds are converted by the fermenting organisms, which would explain the higher reported ethanol yields in SSF compared to SHF (Tenborg et al., 2001; SoÈderstroÈm et al., 2005). In relation to CBP, the properties of pretreated feedstocks are of central importance in defining criteria for developing organisms for use in biomass conversion processes. In this context, Lynd et al. (2002) made a series of key observations in relation to the abilities required for developed organisms, i.e, to hydrolyze cellulose, to hydrolyze insoluble hemicellulose, to convert all sugars derived from hemicellulose during pretreatment, or to remain metabolically active in the presence of inhibitory compounds generated during pretreatment. In contrast, it is stated that it is not necessary that organisms degrade lignin, although modifications that decrease cellulase binding to lignin are valuable.

8.4

Microbial strains for simultaneous saccharification and fermentation (SSF)/ consolidated bioprocessing (CBP)

8.4.1

Simultaneous saccharification and fermentation

Historically, the most commonly used microorganism for ethanol fermentation has been the yeast Saccharomyces cerevisiae, especially in the sucrose-based ethanol industry. This microorganism produces ethanol at a high yield and also has a high ethanol tolerance (Alexandre et al., 1994). In addition, it has proven to be robust to inhibitory compounds present in hydrolysates of lignocellulosic biomass (Olsson and Hahn-HaÈgerdal, 1993). However, wild type S. cerevisiae has a limitation, being unable to metabolize xylose, which is the main hemicellulose-derived sugar from hardwood and agricultural residues. A first approach to solve this problem could be the use of a natural xylose-fermenting yeast, such as Pichia stipitis and Candida shehatae in SSF. However, their tolerance to inhibitory compounds in undetoxified hydrolysates is rather low and, in addition, a well-controlled supply of oxygen is required to effectively ferment xylose (Olofsson et al., 2008). To overcome this drawback, research has been carried out using genetic engineering tools to achieve several improvement of P. stipitis growth in anaerobic conditions (TomaÂs-Pejo et al., 2008). In addition, the last few years have seen significant efforts in the development of engineered bacteria and yeasts for fermentation of both hexoses and pentoses as

ß Woodhead Publishing Limited, 2010

214

Bioalcohol production

most promising alternative (see below in discussion on microorganisms for SSCF). The use of S. cerevisiae in SSF copes with the disadvantage of different optimal conditions for hydrolysis and fermentation since cellulases work optimally at 40±50 ëC whereas the fermentation of hexoses by S. cerevisiae is best carried out at 30±32 ëC. To allow fermentation at temperatures closer to the optimal enzymatic activity, the use of thermotolerant yeasts appears to be an attractive alternative. In early stages of the research on this topic, Szczodrak and Targonski (1988) selected several strains belonging to the genera Saccharomyces, Kluyveromyces and Fabospora in view of their capacity to ferment glucose, galactose and mannose at 40, 43 and 46 ëC, respectively. Also BolloÂk and ReÂczey (2000) evaluated five different Kluyveromyces strains at different temperatures; K. marxianus Y01070 proved to be the best thermotolerant strain. Later work has been performed on ethanol production from various lignocellulosic materials by SSF with thermotolerant strains. Substrates such as wastes from the papermaking industry (KaÂdaÂr et al., 2004; Ballesteros et al., 2002), sugar cane leaves (Krishna et al., 2001), poplar and eucalyptus wood, sorghum bagasse and wheat straw (Ballesteros et al., 2004, 2006) have been successfully tested in SSF at 42±43 ëC with selected thermotolerant strains of K. marxianus and K. fragilis. Interestingly, the search of thermotolerant strains for SSF at high temperatures also applies to S. cerevisiae. Recently Araque et al. (2008) have dealt with the selection of S. cerevisiae strains able to ferment sugars obtained from lignocellulosic material at temperatures of 35±45 ëC with high ethanol yield. Selected strains showing an ethanol yield of at least 70% at 40 ëC on maintenance medium with 20 g/l glucose were submitted to two thermal acclimatization treatments. They were then tested in SSF at 40 ëC on organosolv-pretreated pine biomass. A good performance of selected strain IR2-9 in SSF was demonstrated by the higher ethanol yield attained in comparison to that using the control yeast. Apart from providing more compatible conditions of hydrolysis and fermentation with the use of thermotolerant yeast for ethanol production, several other benefits including energy savings through a reduction in cooling cost and significant decreased risk of contamination have been shown. Nevertheless, high temperatures have been proven to enhance the inhibitory effect of ethanol, which in turn may result in the need to use higher cell concentrations in order to achieve required results. Undoubtedly, the isolation and selection of microorganisms adapted to these hard conditions is an interesting issue that merits further research (Cardona and SaÂnchez, 2007).

8.4.2

Simultaneous saccharification and co-fermentation

In the case of microorganisms for application in the SSCF process, significant progress has been made in recent years in the field of xylose fermentation.

ß Woodhead Publishing Limited, 2010

Integrated hydrolysis, fermentation and co-fermentation

215

Several microorganisms have been genetically engineered to produce ethanol from mixed-sugar substrates by using two different approaches: (a) to divert carbon flow from native fermentation products to ethanol in efficient mixedsugar utilizers such as Esherichia, Erwinia, and Klebsiella; and (b) to introduce the pentose-utilizing capability in the efficient ethanol producers such as Saccharomyces and Zymomonas (Saha, 2004). E. coli was early identified as promising bacterium for genetic modification, since it can grow on a wide range of carbon sources and presents quite good ethanol tolerance (Dien et al., 2003). Notwithstanding, it shows low ethanol yield because sugars are most efficiently converted to organic acids (acetic or lactic acid) instead of ethanol. Thus, main efforts have been focused to redirecting fluxes to ethanol in this bacterium. The most successful approach has been the introduction in E. coli of Z. mobilis genes for alcohol dehydrogenase and pyruvate decarboxilase (Ingram et al., 1991, 1998). A recombinant strain E. coli K011, which has these two enzymes overexpressed, has shown good results in ethanol production from several types of lignocellulosic subtrates including corn cobs (Dien et al., 1997), barley hulls (Kim et al., 2008) and rice hulls (Moniruzzaman and Ingram, 1998). The strain E.coli LY01 derived from K011, has shown higher tolerance to inhibitors present in lignocellulosic hydrolysates (Brandon et al., 2008). An important disadvantage of using E. coli for SSFC processes with cellulases concerns the optimal pH of 6.5, which is not compatible with the optimal pH for cellulolytic enzymes. Moreover, its potential lack of public acceptance, because of the existence of some pathogen strains, makes it less suitable for SSF at industrial scale (TomaÂs-Pejo et al., 2008). Apart from E. coli, xylose can be assimilated by the engineered bacterium Zymomonas mobilis. Whereas Z. mobilis may become an important ethanolproducing microorganism from glucose-based feedstocks, its substrate utilization range is restricted to the fermentation of glucose, sucrose, and fructose. As such, the wild type lacks the essential xylose assimilation and pentose metabolism pathways but it has been successfully engineered with a xylose-metabolizing pathway from E. coli. Significant efforts have been made at National Renewable Energy Laboratory (NREL) in genetic engineering of Z. mobilis (Mohagheghi et al., 1998, 2002; Zhang, 2007), resulting in the development of several recombinant strains showing capability to co-ferment glucose, xylose, and arabinose. SSCF process using Z. mobilis recombinant strains has been demonstrated in the case of ethanol production from yellow poplar hardwood (McMillan et al., 1999) and sugar cane bagasse (Texeira et al., 2000). However, toxic substances present in pretreated hydrolysates or solid biomass substrates substantially affect the ethanol-producing efficiency of these organisms. Hence, the lignocellulosic biomass hydrolysate fermentability still needs to be improved for successful application of recombinant strains. On the other hand, significant efforts have been made in recent decades to design recombinant xylose and arabinose fermenting strains of S. cerevisiae.

ß Woodhead Publishing Limited, 2010

216

Bioalcohol production

8.3 Metabolic pathway for xylose utilization. A) XR-XDH pathway, B) XI pathway.

Since this yeast cannot utilize xylose but can ferment its isomer D-xylulose, different approaches aimed at introducing a heterologous pathway converting xylose to xylulose from naturally pentose-utilizing bacteria and fungi have been tested (Fig. 8.3). So, main strategies have been the construction of recombinant strains by introduction of gene encoding for xylose isomerase (XI) from bacteria or by introduction of genes encoding for xylose reductase (XR) and xylitol dehydrogenase (XDH) from fungi. Also the endogenous gene encoding xylulokinase (XK) has to be overexpressed to obtain significant xylose fermentation. S. cerevisiae has the genes encoding the reductive-oxidative xylose pathway enzymes in its genome although they are expressed at too low levels to allow xylose utilization. An important drawback to overcome when using xylose fermenting S. cerevisiae is that the transport of xylose occurs through hexose transporters, although with lower affinity than for hexose sugars. Therefore, various metabolic engineering efforts involving recombinant S. cerevisiae have led to improvements in the initial rate of xylose consumption (Chu et al., 2007). An advanced metabolic engineering strategy would be the expression of heterologous xylose transporters to further increase xylose utilization in xylosefermenting S. cerevisiae strains. In a publication by Hahn-HaÈgerdal et al. (2007) a summary of fermentation results of recombinant S. cerevisiae strains is presented, including experiments in batch and continuous culture as well as in different aeration conditions. In

ß Woodhead Publishing Limited, 2010

Integrated hydrolysis, fermentation and co-fermentation

217

spite of the extensive research performed in laboratory yeast strains in recent years, there are relatively few studies about the performance of these strains on È hgren lignocellulosic hydrolysates. For example, in a recent work published by O et al. (2006), the whole steam-pretreated slurry of corn stover was submitted to SSF with the recombinant S .cerevisiae strain TMB3400, previously grown in prehydrolysate. Results showed co-fermentation of glucose and xylose with an overall ethanol yield of 64% of theoretical at 5% WIS loading. Among the reasons underlying the scarcity of references on industrial applications of engineered S. cerevisiae strains, the lack of robustness required for industrial application, low tolerance to lignocellulose hydrolysates and strain instability have been reported (Hahn-HaÈgerdal et al., 2007). On the other hand, the use of genetically engineered bacteria or yeast in industrial biotechnology could give rise to some controversy in the public due to the potential health and environmental risk associated with the use of GMOs (Tamis et al., 2009). To overcome this drawback, it would be necessary to guarantee the compliance of the existing biotechnology regulatory frameworks and regulations ensuring that proposed uses of engineered organisms in the open environment received appropriate scrutiny and risk assessment (Glass, 1991; EU, 2001).

8.4.3

Consolidated bioprocessing

Regarding the development of microorganisms for cellulose conversion via CBP, although important advances have been achieved during the last decade, up to date microorganisms with the whole combination of properties required are not available. As described above in Section 8.2, microorganism development can be pursued according to two strategies. The native cellulolytic strategy involves microorganism with high cellulolytic activity to improve ethanol production (yield and tolerance). Among cellulolytic microorganisms, some thermophilic anaerobes belonging to the genera Clostridium are of particular interest. According to Wyman (1994), in most of the studies about CBP, Clostridium thermocellum is used for enzyme production, cellulose hydrolysis and glucose fermentation, whereas C. thermosaccharolyticum allows the co-fermentation of glucose and pentoses to ethanol. In particular, the CPB process using C. thermocellum has been proven to provide higher substrate conversion than system using cellulases from T. reseei and S. cerevisiae. Among the features impacting the feasibility of CBP using the native cellulolytic strategy, ethanol tolerance appears as a key microorganism feature (Lynd et al., 2005). Studies carried out on ethanol tolerance of C. thermocellum have led to the selection of strains being able to grow at ethanol concentrations over 60 g/l, which means a good perspective for application to lignocellulose conversion processes (Strobel and Lynn, 2004). Other thermophiles such as Thermoanaerobacterium thermosaccharolyticum, also submitted to metabolic engineering studies for application in CBP, are more

ß Woodhead Publishing Limited, 2010

218

Bioalcohol production

sensitive to organic acids and their salts than to ethanol and so, the removal of organics allows improved ethanol yields. The recombinant cellulolytic strategy involves the genetic modification of a microorganism that produces ethanol at high yield so it is able to use cellulose in a CBP configuration. The primary objective is to engineer a heterologous cellulase system that enables growth and fermentation on a pretreated lignocellulosic substrate. Heterologous expression of cellulases has been studied primarily in bacterial hosts as engineered E. coli, K. oxytoca and Z. mobilis and the yeast S. cerevisiae. For example, two endoglucanases of E. chrysanthemi and a cellulase of Acetobacter xylinum have been successfully expressed in E. coli (Zhou et al., 1999; Okamoto et al., 1994). Similarly, cloning strategy in K. oxytoca resulted in a recombinant strain that produced carboxymethylcellulase (CMCase) activity (Zhou and Ingram, 2001). Regarding the expression of saccharolytic enzymes in S. cerevisiae, the expression of cellulases, xylanases and amylases on S. cerevisiae strains has been described (Fujita et al., 2004). The work demonstrates that these recombinant strains can achieve significant hydrolysis of amorphous cellulose and a variety of lignocellulosic substrates. Notwithstanding, it is also highlighted that apart from the number of genes expressed, it is even more important to attain a high-level expression of these genes (van Zyl et al., 2007). In the papers by Lynd et al. (2002, 2005), a comprehensive study of metabolic engineering approaches in native and recombinant strategies in bacteria and yeast can be found. Moreover, a detailed review of the current status of saccharolytic enzyme (cellulases and hemicellulases) expression in S. cerevisiae for CBP has been published (van Zyl et al., 2007).

8.5

Future trends

It seems to be widely accepted that future large-scale production of ethanol will most certainly have to be based on production from lignocellulosic materials. In lignocellulose biomass to ethanol production technology, several key factors for enhancing competitiveness have been identified, depending on the process configuration. In SSF, future overall performance depends strongly on development of cheaper and more efficient enzymes for saccharification. Research priorities such as increasing specific activity, thermal stability and hydrolysis rates, reducing cost by a factor of 20 to 30, reducing non-specific binding to lignin and improving the combination of cellulase, hemicellulase and other enzyme activities to maximize the sugar yield have been identified (SaÂnchez and Cardona, 2008; Yan and Wyman, 2007). On the other hand, a high substrate loading and hence a high insoluble solids concentration, is crucial for the economy of the SSF process. Strategies to allow the use of high substrate

ß Woodhead Publishing Limited, 2010

Integrated hydrolysis, fermentation and co-fermentation

219

content as fed-batch approach or the inclusion of a pre-hydrolysis step, which decreases the viscosity of the medium at the start of fermentation, show great potential to attain higher ethanol concentrations and consequently to improve production costs. One of the major challenges is to optimize the integration of process engineering, fermentation technology, and enzyme engineering within the SSF process. Regarding the co-fermentation of hexoses and pentoses, efforts will be oriented to the selection and development of microbial strains with improved characteristics of stability, sugar assimilation efficiency, tolerance to ethanol and resistance to inhibitors. Studies performed up to now on metabolic engineering strategies for pentose fermentation have resulted in the selection of laboratory strains with proven properties such as cell biomass yield and stability under well-defined conditions but not under industrial conditions. Thus, the major future challenge remains to translate the knowledge acquired from laboratory strains to industrial production strains that provide the fermentation capacity and robustness required for industrial processes. In the search for efficiency in fuel ethanol production, the most significant outcomes will probably come from the implementation of CBP through the development of tailored recombinant microorganisms. Similarly to metabolic engineering for pentose fermentation, essentially all work carried out in organism development for CBP has involved laboratory strains and therefore, much of this work has to be transferred to industrial strains. Lynd et al. (2002) described in depth the strengths and challenges associated with the twoorganism development strategies for CBP (native and recombinant cellulolytic strategies), emphasizing the key role of biotechnology to progress toward application of both paradigms. And finally, it is also important to consider that biomass to ethanol R&D must move beyond enhancing conversion technologies alone and, for example, use biotechnology tools to re-design the feedstock for specific products. In addition, biotechnology opens the door for future success by being useful in an integrated product design strategy; for example, where feedstock and bioconversion can both be designed to allow optimal interaction in the system. Notwithstanding, such integrated approaches, require broad scientific coordination, managed teamwork, and priority for R&D support funding. An integrated cross-discipline strategy will be vital to make large enough technical and economic breakthroughs for biomass utilization to contribute to any future sustainable energy platform. There is no doubt that it is an enormous challenge to be faced during the coming years.

8.6

References

Alexandre H, Rousseaux I, Charpentier C (1994) `Ethanol adaptation mechanisms in Saccharomyces cerevisiae', Biotechnol Appl Biochem, 20, 173±183.

ß Woodhead Publishing Limited, 2010

220

Bioalcohol production

Araque E, Parra C, RodrõÂguez M, Freer J, Baeza, J (2008) `Selection of thermotolerant yeast strains Saccharomyces cerevisiae for bioethanol production', Enzyme Microb Technol, 43(2), 120±123. Ballesteros I, Oliva J M, Saez F, Ballesteros M (2001) `Ethanol production from lignocellulosic byproducts of olive oil extraction', Appl Biochem Biotechnol ± Part A Enzyme Engineering and Biotechnology, 91±93, 237±252. Ballesteros I, Oliva J M, Negro M J, Manzanares P, Ballesteros M (2002) `Ethanol production from paper material using a simultaneous saccharification and fermentation system in a fed-batch basis', World J Microb Biot, 18(6), 559±561. Ballesteros M, Oliva J M, Negro M J, Manzanares P, Ballesteros I (2004) `Ethanol from lignocellulosic materials by a simultaneous saccharification and fermentation process (SFS) with Kluyveromyces marxianus CECT 10875', Process Biochem 39, 1843±1848. Ballesteros I, Negro M J, Oliva J M, CabanÄas A, Manzanares P, Ballesteros M (2006) `Ethanol production from steam-explosion pretreated wheat straw', Appl Biochem Biotechnol, 129±132, 496±508. BolloÂk M and ReÂczey K (2000) `Screening of different Kluyveromyces strains for simultaneous saccharification and fermentation', Acta Aliment. Hung. 29, 59±70. Brandon S K, Eiteman M A, Patel K, Richbourg M M, Miller D J, Anderson W F, Peterson J D (2008) `Hydrolysis of Tifton 85 bermudagrass in a pressurized batch hot water reactor', J Chem Technol Biotechnol, 83, 505±512. Brett C and Waldron K (1996) `The molecular component of the cell wall'. In: Physiology and Biochemistry of Plant Cell Walls. Black, M. and Charlwood, B. (eds), Chapman & Hall, 2nd edition, London, 4±44. Cardona C A and SaÂnchez, O J (2007) `Fuel ethanol production: Process design trends and integration opportunities', Bioresour Technol, 98, 2415±2457. Carpita N C and Gibeaut D M (1993) `Structural models of primary-cell walls in flowering plants ± consistency of molecular structure with the physical properties of the walls during growth', Plant J, 3, 1±30. Chandra R P, Bura R, Mabee W E, Berlin A, Pan X, Saddler J N (2007) `Substrate pretreatment: the key to effective enzymatic hydolysis of lignocellulosics?', Adv Biochem Engin/Biotechnol, 108, 67±93. Chandrakant P and Bisaria V S (1998) `Simultaneous bioconversion of cellulose and hemicellulase to ethanol', Crit Rev Biotechnol, 18(4), 295±331. Chu B C H and Lee H (2007) `Genetic improvement of Saccharomyces cerevisiae for xylose fermentation', Biotechnol Adv, 25, 425±441. Dien B S, Hespell R B, Ingram L O, Bothast R J (1997) `Conversion of corn milling fibrous co-products into ethanol by recombinant Escherichia coli strains K011 and SL4', World J Microbiol Biotechnol, 13, 619±625. Dien B S, Cotta M A, Jeffries T W (2003) `Bacteria engineered for fuel ethanol production: current status', Appl Biochem Biotechnol, 63 (2003) 258±266. Dien B S, Cotta M A, Jung H-J G, Vogel K P (2007) `Developing herbaceous energy crops as feedstocks for bioethanol production', ACS National Meeting Book of Abstracts, 1p. EU (2001) `Directive 2001/18/EC of the European Parliament and of the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms', Official Journal of the European Communities, L 106/1, 17.4.2001. Fujita Y, Ito J, Ueda M, Fukuda H, Kondo A (2004) `Synergistic saccharification and direct fermentation to ethanol of amorphous cellulose by use of an engineered yeast strain codisplaying the types of cellulolytic enzyme', Appl Environ Microbiol, 70,

ß Woodhead Publishing Limited, 2010

Integrated hydrolysis, fermentation and co-fermentation

221

5407±5414. GaÂspaÂr M, KaÂlmaÂn G, ReÂczey K (2007) `Corn fiber as a raw material for hemicellulose and ethanol production', Process Biochem 42 (7), 1135±1139. Glass D J (1991), `Impact of government regulation on commercial biotechnology', in Ono R D, The Business of Biotechnology: From the Bench to the Street, Butterworth-Heinemann, Stoneham, MA, 169±198. Guo G L, Chen W H, Chen W H, Men L C, Hwang W S (2008) `Characterization of dilute acid pretreatment of silvergrass for ethanol production', Bioresour Technol, 99 (14), 6046±6053. Hahn-HaÈgerdal B, Karhumaa K, Jepsson M, Gorwa-Grauslund M F (2007) `Metabolic engineering for pentose utilization in Saccharomyces cerevisiae', Adv Biochem Eng/Biotechnol, 108, 147±177. Hamelinck C N, van Hooijdonk G, Faaij A P C (2005) `Ethanol from lignocellulosic biomass: techno-economic performance in short-, middle-, and long-term', Biomass Bioen 22, 384±410. Hendriks A T W M and Zeeman G (2009) `Pretreatments to enhance the digestibility of lignocellulose biomass', Bioresour Technol, 100, 10±18. Ingram L O, Alterthum F, Conway T, Ohta K, Shanmugam K T (1991) `Ethanol production by Escherichia Coli strains co-expressing Zymomonas Pdc and Adh genes', US Patent No. 5,000,000. Ingram L O, Gomez P F, Moniruzzaman M, Wood B E, Yomano L P, York S W (1998) `Metabolic engineering of bacteria for ethanol production', Biotechnol Bioeng, 58 205±214. Jacobsen E A, Ballance G M, Thomsen K K (2000) `Cellulose and hemicellulose hydrolysis models for application to current and novel pretreatment processes', Appl Biochem Biotechnol, 84±86, 81±96. KaÂdaÂr Z, Szengyel Z, ReÂczey K (2004) `Simultaneous saccharification and fermentation (SSF) of industrial wastes for the production of ethanol', Ind Crop Prod, 20, 103± 110. Kim S and Dale B E (2004) `Global potential bioethanol production from wasted crops and crop residues', Biomass Bioen 26 (4), 361±375. Kim T H, Taylor F, Hicks K B (2008) `Bioethanol production from barley hull using SAA (soaking in aqueous ammonia) pretreatment', Bioresour Technol, 99, 5694±5702. Krishna S H, Reddy T J, Chowdary G V (2001) `Simultaneous saccharification and fermentation of lignocellulosic wastes to ethanol using a thermotolerant yeast', Bioresour Technol, 77, 193±196. Lloyd T A and Wyman C E (2005) `Combined sugar yields for dilute sulfuric acid pretreatment of corn stover followed by enzymatic hydrolysis of the remaining solids', Bioresour Technol, 96, 1967±1977. Lynd L R (1996) `Overview and evaluation of fuel ethanol from cellulosic biomass: technology, economics, the environment and policy', Annu Rev Energ Env, 21, 403±465. Lynd L R, Weimer P J, van Zyl W H, Pretorius I S (2002) `Microbial cellulose utilization: fundamentals and biotechnology', Microbiol Mol Biol R, 66 (3), 506±577. Lynd L R, van Zyl W H, McBride J E, Laser M (2005) `Consolidated bioprocessing of cellulosic biomass: an update', Curr Opin Biotech, 16, 577±583. McMillan J D, Newman M M, Templeton D W, Mohagheghi A (1999) `Simultaneous saccharification and cofermentation of dilute-acid pretreated yellow poplar hardwood to ethanol using xylose-fermenting Zymomonas mobilis', Appl Biochem Biotechnol ± Part A Enzyme Engineering and Biotechnology, 77±79, 649±665.

ß Woodhead Publishing Limited, 2010

222

Bioalcohol production

Mohagheghi A, Evans K, Finkelstein M, Zhang M (1998) `Cofermentation of glucose, xylose, and arabinose by mixed cultures of two genetically engineered Zymomonas mobilis strains', Appl Biochem Biotechnol ± Part A Enzyme Engineering and Biotechnology 70±72, 285±299. Mohagheghi A, Evans K, Chou Y-C, Zhang M (2002) `Cofermentation of glucose, xylose, and arabinose by genomic dna-integrated xylose/arabinose fermenting strain of Zymomonas mobilis AX101', Appl Biochem Biotechnol ± Part A Enzyme Engineering and Biotechnology, 98±100, 885±898. Moiser N, Wyman C E, Dale B, Elander R, Lee Y Y, Holtzapple M, Ladisch M (2005) `Features of promising technologies for pretreatment of lignocellulosic biomass', Bioresour Technol, 96, 673±686. Moniruzzaman M and Ingram L O (1998) `Ethanol production from dilute acid hydrolysate of rice hulls using genetically engineered Escherichia coli', Biotechnol Lett, 20 (10), 943±947. È hgren K, Bengtsson O, Gorwa-Grauslund M F, Galbe M, Hahn-Hagerdal B, Zacchi G O (2006) `Simultaneous saccharification and co-fermentation of glucose and xylose in steam-pretreated corn stover at high fiber content with Saccharomyces cerevisiae TMB3400', J Biotechnol, 126 (4), 488±498. Okamoto T S, Yamano S, Ikeaga H, Nakmura K (1994) `Cloning of the Acetobacter xylimun cellulase gene and its expression in Escherichia coli and Zyomonas mobilis', Appl Microbiol Biotechnol 42, 563±568. Olofsson K, Bertilsson M, LideÂn G (2008) `A short review on SSF ± an interesting process option for ethanol production from lignocellulosic feedstocks', Biotechnology for Biofuels, 1(7), 1±14. Olsson L and Hahn-HaÈgerdal B (1993) `Fermentative performance of bacteria and yeast in lignocellulose hydrolysates', Process Biochem, 28, 249±257. Olsson L and Hahn-HaÈgerdal B (1996) `Fermentation of lignocellulosic hydrolysates for ethanol production', Enzyme Microb Technol, 18, 312±331. Olsson L, Jorgensen H, Krogh K B R, Roca C (2005) `Bioethanol production from lignocellulosic material', in Dimitriu S, Polysaccharides: Structural Diversity and Functional Versatility, Marcel Dekker, New York, 957±993. Saha B C (2004) `Lignocellulose biodegradation and applications in biotechnology', ACS Sym Ser. 889, 2±34. SaÂnchez O J and Cardona C A (2008) `Trends in biotechnological production of fuel ethanol from different feedstocks', Bioresour Technol, 99, 5270±5295. SoÈderstroÈm J, Pilcher L, Galbe M, Zacchi G (2002) `Two-step steam pretreatment of softwood with SO2 impregnation for ethanol production', Appl Microbiol Biotechnol 98±100, 5±21. SoÈderstroÈm J, Galbe M, Zacchi G, (2005) `Separate versus simultaneous saccharification and fermentation of two-step pretreated softwood for ethanol production', J Wood Chem Technol, 25, 187±202. Strobel H P and Lynn B (2004) `Proteomic analysis of ethanol sensitivity in Clostridium thermocellum', in Proceeding of the National Meeting of he American Society of Microbiology, New Orleans, LA. Szczodrak J and Targonski Z (1988) `Selection of thermotolerant yeast strains for simultaneous saccharification and fermentation of cellulose', Biotechnol Bioeng, 31, 300±303. Tagaki M E, Abe S, Suzuki S, Emert G H, Yata N (1977) `A method for production of alcohol direct from cellulose using cellulase and yeast', in Ghose T K, Proc Bioconversion Sym, IIT, New Delhi, India, 551±571.

ß Woodhead Publishing Limited, 2010

Integrated hydrolysis, fermentation and co-fermentation

223

Tamis, W L M, van Dommelen A, de Snoo G R (2009) `Lack of transparency on environmental risks of genetically modified micro-organisms in industrial biotechnology', Journal of Cleaner Production 17 (6) 581±592. Teixeira L C, Linden J C, Schroeder H A (2000) `Simultaneous saccharification and cofermentation of peracetic acid-pretreated biomass', Appl Biochem Biotechnol ± Part A Enzyme Engineering and Biotechnology, 84±86, 111±127. Tengborg C, Galbe M, Zacchi G (2001) `Reduced inhibition of enzymatic hydrolysis of steam-pretreated softwood', Enzyme Microb Technol, 28 (9±10), 835±844. TomaÂs-Pejo E, Oliva J M, Ballesteros M (2008) `Realistic approach for full-scale bioethanol production from lignocellulose: a review', J Sci Ind Res, 100(6), 1122± 1131. Torget R W, Kim J S, Lee Y Y (2000) `Fundamental aspects of dilute acid hydrolysis/ fractionation kinetics of hardwood carbohydrates. I. Cellulose hydrolysis', Ind Eng Chem Res, 39, 2817. van Zyl W H, Lynd L R, den Haan R, McBride J E (2007) `Consolidated bioprocessing for bioethanol production using Saccharomyces cerevisiae', Adv Biochem Engin/ Biotechnol, 108, 205±235. Wyman C E (1994) `Ethanol from lignocellulosic biomass: technology, economics and opportunities', Bioresour Technol, 50, 3±16. Wyman C E (1999) `Biomass ethanol: technical progress, opportunities and commercial challenges', Annu Rev Energ Env 24, 19±226 Wyman C E (2001) `Twenty years of trials, tribulations and research progress in bioethanol technology ± selected key events along the way', Appl Biochem Biotechnol, 91±3, 5±21. Wyman C E and Hinman N D (1990) `Ethanol: fundamentals of production from renewable feedstocks and use as a transportation fuel', Appl Biochem Biotechnol, 24/25, 735±742. Wyman C E, Dale B E, Elander R T, Holtzapple M, Ladisch M R, Lee Y Y (2005) `Comparative sugar recovery data from laboratory scale application of leading pretreatment technologies to corn stover', Bioresour Technol 96 (18), 2026±2032. Yan B and Wyman C E (2007) `Biotechnology for cellulosic ethanol', APBN, 11(9), 555± 563. Zeng M, Moiser N S, Huang C P, Sherman D M, Ladisch M R (2007) `Microscopic examination of changes of plant cell structure in corn stover due to cellulase activity and hot water pretreatment', Biotechnol Bioeng, 97 (2), 265±278. Zhang M (2007) `Engineered Zymomonas ferments five-carbon sugars', Ind Bioprocessing, 29(7), 6. Zhou S and Ingram L O (2001) `Simultaneous saccharification and fermentation of amorphous cellulose to ethanol by recombinant Klebsiella oxytoca SZ21 without supplemental cellulase', Biotechnol Lett, 23, 1455±1462. Zhou S, Yomano A Z, Saleh A Z, Davis, F C, Aldrich H C, Ingram L O (1999) `Enhancement of expression and apparent secretion of Erwinia chrysanthemi endoglucanase (encoded by celZ) in Escherichia coli', Appl Environ Microbiol, 65, 2439±2445.

ß Woodhead Publishing Limited, 2010

9

Challenges in co-fermentation of lignocellulosesderived sugars using baker's yeast D . R U N Q U I S T , N . S . P A R A C H I N and B . H A H N È G E R D A L , Lund University, Sweden HA

Abstract: Lignocellulosic biomass presents an attractive but challenging substrate for bioethanol production. Compared to starch and sucrose based processes, utilization of lignocellulose is complicated by more elaborate pretreatment of the biomass and the presence of toxic inhibitory compounds. In addition, lignocellulosic biomass is composed of several types of sugar polymers, which gives rise to a complex blend of hexose and pentose monomer sugars in the feed medium. Brewer's yeast, Saccharomyces cerevisiae, has been metabolically engineered to utilize the pentose sugars xylose and arabinose; however, successful co-fermentation of all sugars derived from the lignocellulosic biomass is even less straightforward. Mixed substrate fermentation may lead to sequential utilization of sugars, increased byproduct formation and longer overall fermentation times. In this chapter, challenges relating specifically to co-fermentation of lignocelluloses-derived sugars are discussed. Preferential utilization of certain sugars will be analyzed from the perspective of transport and transcriptional control. Practical solutions to improve co-fermentation of lignocelluloses-derived sugars are discussed and suggestions for further improvements are given. Key words: ethanol, lignocelluloses, fermentation, xylose, arabinose, Saccharomyces cerevisiae.

9.1

Introduction

Utilization of lignocellulosic biomass for microbial production of fuel ethanol has been researched since the early 1980s (Jeffries, 1985; Skoog and HahnHaÈgerdal, 1988). In comparison to ethanol production based on corn, wheat and sugar cane, lignocellulosic biomass can be found in residues from the agriculture and forest industry and its usage does not compete with food and feed production. Dedicated lignocellulose rich energy crops, such as switchgrass and elephant grass, can also be grown on marginal land unfit for cultivation of crops for human or animal consumption (Schmer et al., 2008). Consequently the price of lignocellulosic biomass is lower than for traditional crops, and the potential for cost-efficient ethanol production is higher (Galbe et al., 2007). Lignocellulose is, however, significantly harder to degrade than starch and requires

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

225

more severe pre-treatment of the raw material (Almeida et al., 2007; Galbe and Zacchi, 2007) and tailor-made process design solutions (Hahn-HaÈgerdal et al., 2006; Galbe et al., 2007; Olofsson et al., 2008a). In contrast to cellulose and starch, lignocellulose is composed of a mixture of hexose (glucose, mannose and galactose) and pentose (xylose and arabinose) sugars. Cost-competitive ethanol yields from lignocellulose therefore require fermentation of both hexose and pentose constituents (Galbe et al., 2007). Brewer's yeast, Saccharomyces cerevisiae, is the organism of choice for the mature starch based ethanol industry. This organism ferments hexose sugars efficiently but unfortunately lacks natural ability to utilize pentose sugars. Usage of S. cerevisiae in ethanol production, however, has several practical advantages compared to pentose fermenting yeast and bacteria (Hahn-HaÈgerdal et al., 2006). Significant metabolic engineering has therefore been applied to confer pentose-fermenting ability to S. cerevisiae and enable utilization of all lignocellulose constituents (Jeffries, 2006; Hahn-HaÈgerdal et al., 2007a, 2007b). In the current text we would like to bring attention to an aspect of mixed substrate fermentation of lignocelluloses-derived sugars that has not been extensively addressed. Compared to single sugar components, fermentation of mixed sugars presents unique challenges. Biochemical properties of the individual sugars control the rate, sequence and yield of the fermentation process. For instance, in a typical fermentation of lignocelluloses-derived sugars, glucose and mannose are the preferred substrates (Rudolf et al., 2005; Alkasrawi et al., 2006). Once glucose and mannose have been consumed, utilization of galactose and pentose sugars follows (Olofsson et al., 2008b; Rudolf et al., 2008). The sequential consumption pattern increases the overall fermentation time and lowers production rates. Co-fermentation of two or more sugar components may also increase byproduct formation and decrease product yields compared to individual sugar components (Karhumaa et al., 2006). In some cases the solution to these problems is technical and in others a metabolic engineering strategy is more suitable. In this chapter we summarize current knowledge on cofermentation of lignocellulose-derived sugars by S. cerevisiae.

9.2

Transporter preferences

Co-fermentation of two or more sugars is initially controlled by transport. The uptake rate of individual substrates is determined by kinetic parameters of the available transporters. Expression of the transporters, on the other hand, is regulated at the transcriptional level. The transporter kinetics and the transcriptional regulation thus reflect two levels of substrate uptake control. The difference may seem semantic; however, it is useful to address these concepts separately. In particular, different solutions should be pursued if the substrate uptake rate is limited by transport kinetics or transcriptional repression.

ß Woodhead Publishing Limited, 2010

226

9.2.1

Bioalcohol production

Uptake of hexose sugars

Among the sugars present in a lignocellulose hydrolysate, glucose, mannose and galactose represent hexose sugars that can be fermented by native S. cerevisiae (Galbe and Zacchi, 2007). Glucose and mannose are transported by the Hxt transport system, and are by practical significance utilized simultaneously (Rudolf et al., 2005; Alkasrawi et al., 2006). Galactose is transported by the Gal2 transporter and fermented to ethanol by S. cerevisiae (Kou et al., 1970). During co-fermentation with glucose however, the Gal2 permease is repressed and galactose is only consumed after depletion of glucose and mannose (Matern and Holzer, 1977; DeJuan and Lagunas, 1986; Horak and Wolf, 1997). The order of hexose uptake is thus primarily controlled at the transcriptional level.

9.2.2

Uptake of pentose sugars

In lignocellulose hydrolysate, xylose and arabinose constitute the pentose fraction (Galbe and Zacchi, 2007). Since S. cerevisiae is naturally incapable of utilizing pentoses, a specific transport system does not exist for these sugars. Xylose is taken up by Hxt transporters in the cell, albeit with a much lower affinity than for glucose (Table 9.1). The affinity of the Hxt transporters for xylose has been determined and depends on the particular transporter (KoÈtter and Ciriacy, 1993; Lee et al., 2002; Saloheimo et al., 2007). Generally the highaffinity glucose transporters have the highest affinity for xylose (Km = 100± 200 mM), whereas the low-affinity glucose transporters have very low affinity for xylose (Km = 1±1.5 M). Since high-affinity Hxt transporters are only expressed at low glucose concentrations, transcriptional regulation is important for xylose uptake. In contrast to xylose, arabinose is transported via the GAL2 permease in galactose grown cells (Kou et al., 1970). Analogous to galactose utilization, arabinose transport via Gal2 is dependent on induction of Gal2 expression. While basic induction results from growth on a non-repressive carbon source Table 9.1 Kinetic parameters of Hxt transporters for glucose (Reifenberger et al., 1997; Maier et al., 2002) and xylose (Saloheimo et al., 2007) uptake Glucose

Xylose

Transporter

Km (mM)

Vmax (nmol/min/ mg protein)

Km (mM)

Vmax (nmol/min/ mg protein)

Hxt1 Hxt2 Hxt4 Hxt7

90±130 60  10 6±9 1±2

690  26 450  20 160  8 186  7

880  8 260  130 170  120 130  9

750  94 340  10 190  23 110  7

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

227

(such as galactose), over-expression of Gal2 using constitutive glycolytic promoters did additionally improve arabinose uptake (Becker and Boles, 2003). During fermentation of a mixture of sugars, xylose and arabinose are consumed together with galactose, following depletion of glucose and mannose. It is thus clear that arabinose transport as well as xylose transport, is governed by both transcriptional and kinetic regulation. The latter is manifested in the generally poor affinity of Hxt and Gal2 transporters for pentose sugars, while transcriptional regulation suppresses transporters with the highest affinity for pentose sugars during glucose co-fermentation (Sedlak and Ho, 2004).

9.2.3

Improving transport of pentose sugars

We have seen that xylose and arabinose are poorly transported by native S. cerevisiae transporters. In addition to the low affinity for these substrates, transcriptional repression is evident during glucose co-fermentation (Reifenberger et al., 1997; Lee et al., 2002; Sedlak and Ho, 2004). There have thus been attempts to improve transport of pentose sugars in recombinant S. cerevisiae by expression of heterologous transporters. Arabinose transport and metabolism have been studied in natural arabinose fermenting yeast (Fonseca et al., 2007). In S. cerevisiae, however, arabinose transport has so far only been studied briefly (Kou et al., 1970; Becker and Boles, 2003). In contrast, xylose transport has been studied in much greater detail by over-expression of heterologous and endongenous xylose transporters in recombinant xylose-utilizing S. cerevisiae (Table 9.2) (Hamacher et al., 2002; GaÂrdoÂnyi et al., 2003b; Sedlak and Ho, 2004; Leandro et al., 2006, 2008; Saloheimo et al., 2007; Hector et al., 2008; Katahira et al., 2008; Runquist et al., 2008).

9.2.4

Expression of heterologous transporters

Expression of the heterologous xylose trasporters Hup1 (Chloella kessleri), Stp2/3 (Arabidopsis thaliana), XylE (Echerichia coli) and At5g59250 (Arabidopsis thaliana) failed to support xylose growth in a xylose-utilizing S. cerevisiae
HXT strain (Hamacher et al., 2002). Reconstitution of xylose growth in this background was, however, possible using single expression of endogenous high-affinity Hxt transporters (Hamacher et al., 2002; Sedlak and Ho, 2004; Saloheimo et al., 2007). Over-expression of Hxt transporters in a wild-type strain did, however, not increase xylose uptake or ethanol formation (Hamacher et al., 2002). Expression of the Trxlt1 transporter (Hypocrea jecorina) in a
HXT strain yielded weak xylose growth after adaptation (Saloheimo et al., 2007). The Gxf1 transporter (Candida intermedia) was able to support xylose growth in S. cerevisiae and displayed significantly improved affinity for xylose (Km = ~50 mM) (Leandro et al., 2006, 2008). Improved xylose uptake and

ß Woodhead Publishing Limited, 2010

Table 9.2 Heterologous xylose transporters expressed in Saccharomyces cerevisiae. Transporters have been characterized by the following: reconstruction of xylose growth in a
HXT background; transport kinetics and xylose utilization. N.d. not determined ß Woodhead Publishing Limited, 2010

Expression in S. cerevisiae Gene

Organism

Xylose growth

Transport kinetics

Xylose utilization

Reference

Trxlt1 Hup1 Stp2/3 XylE At5g59250 At5g59250 Sut1 GXF1

Hypocrea jecorina Chloella kessleri Arabidopsis thaliana Echerichia coli Arabidopsis thaliana Arabidopsis thaliana Pichia stipitis Candida intermedia

Yes, after adaptation No No No No Not tested Not tested Yes

N.d. N.d. N.d. N.d. N.d. N.d. N.d. Improved xylose affinity

Not improved Not improved Not improved Not improved Not improved Yes Yes Yes

GXS1

Candida intermedia

No

Improved xylose affinity

N.d.

Saloheimo et al. (2007) Hamacher et al. (2002) Hamacher et al. (2002) Hamacher et al. (2002) Hamacher et al. (2002) Hector et al. (2008) Katahira et al. (2008) Leandro et al. (2006), Runquist et al. (2008) Leandro et al. (2006), Leandro et al. (2008)

Challenges in co-fermentation

229

ethanol productivity was seen in a strain expressing the Gxf1 transporter (Fig. 9.1) (Runquist et al., 2008). Likewise expression of the Sut1 transporter (Picchia stipitis) and the putative transporter At5g59250 (Arabidopsis thaliana) was found to improve xylose and glucose utilization in S. cerevisiae (Hector et al., 2008; Katahira et al., 2008). The latter results stand in contrast to previous work where expression of At5g59250 failed to support growth in an
HXT strain (Hamacher et al., 2002). It is, however, possible that although sole expression of At5g59250 does not support growth, At5g59250 is able to increase xylose

9.1 Expression of Gxf1 transporter in xylose-utilizing recombinant S. cerevisiae (Runquist et al., 2008). Symbols: (l) strain expressing the Gxf1 transporter; ( ) reference strain. (a) Initial rate of D-C14 xylose uptake. (b) Aerobic growth in 4 g/L xylose.



ß Woodhead Publishing Limited, 2010

230

Bioalcohol production

uptake when expressed in strain expressing other (Hxt) transporters (Hamacher et al., 2002; Hector et al., 2008).

9.2.5

Is transport controlling the uptake rate of pentose sugars?

Recent developments have shown that expression of heterologous transporters have the potential to increase xylose uptake in recombinant xylose-utilizing S. cerevisiae (Hector et al., 2008; Katahira et al., 2008; Runquist et al., 2008). Some ambiguity, however, still remains regarding the cause of improvement. Although xylose transport in S. cerevisiae is poor, several studies have shown that transport is only limiting the xylose uptake rate at low xylose concentrations (KoÈtter and Ciriacy, 1993; GaÂrdoÂnyi et al., 2003a; Runquist et al., 2008). Among the heterologous transporters expressed in S. cerevisiae (Table 9.2), improved affinity for xylose has only been demonstrated for the Gxf1 tranporter (Leandro et al., 2006; Runquist et al., 2008). Xylose transport is, however, controlled by both transcriptional and enzyme kinetic restrictions. Reduced xylose transport has been demonstrated in glucose-repressed cells (Reifenberger et al., 1997; Lee et al., 2002). During glucose co-fermentation, or any condition when expression of high affinity Hxt transporters is repressed, over-expression of a xylose transporter may improve xylose uptake. Thus in terms of increased xylose transport, the relative contribution of improved affinity and increased expression of xylose transporters, needs further clarification in carefully designed systematic studies.

9.3

Combining recombinant pathways

Recombinant S. cerevisiae strains expressing xylose- and arabinose-utilizing pathways have been developed. Xylose metabolism has been realized using a reduction/oxidation pathway consisting of xylose reductase (XR) and xylitol dehydrogenase (XDH) (KoÈtter and Ciriacy, 1993; Ho et al., 1998; Eliasson et al., 2000) or an isomerase-based pathway consisting of xylose isomerase (XI) (Walfridsson et al., 1996; Kuyper et al., 2003) (Fig. 9.2). Arabinose in turn can be utilized via an isomerization pathway consisting of L-arabinose isomerase (AI), L-ribulokinase (RK), and L-ribulose-5-P 4-epimerase (RE) (Becker and Boles, 2003) or a reduction/oxidation-based pathway consisting of xylose reductase (XR), L-arabinitol 4-dehydrogenase (ADH), L-xylulose reductase (LXR) and xylitol dehydrogenase (XDH) (Fig. 9.2) (Richard et al., 2003; Verho et al., 2004). Although more research is needed primarily on arabinose utilization, strains expressing a combination of both xylose and arabinose pathways have been constructed (Karhumaa et al., 2006; Wisselink et al., 2007; Bettiga et al., 2008). In principle these strains are able to utilize xylose and arabinose separately; however, it has been shown that co-fermentation of the two sugars is restricted by some inherent difficulties.

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

231

9.2 Xylose and arabinose pathways expressed in recombinant Saccharomyces cerevisiae. Isomerase-based pathways are depicted using dashed lines (----) and the oxidation/reduction pathways are shown with solid lines (öö). XR, xylose reductase; XDH, xylitol dehydrogenase; XI, xylose isomerase; XK, xylulokinase; AI, arabinose isomerase; RK; ribulokinase; RE; ribulose-5-P 4epimerse; ADH, arabitol 4-dehydrogenase; LXR, L-xylulose reductase.

9.3.1

Enzyme cross-affinity

The combination of the two recombinant pathways for xylose and arabinose utilization in S. cerevisiae has raised the problem of cross-affinities of the involved enzymes (Fig. 9.2). In particular xylose reductase (XR) displays similar affinity for xylose and arabinose (Verduyn et al., 1985; Rizzi et al., 1988). Thus in S. cerevisiae strains combining the oxidation/reduction and isomerization pathways for xylose- and arabinose-utilization, the major fraction of the consumed arabinose was converted to arabitol (Karhumaa et al., 2006; Bettiga et al., 2008). Since an arabitol dehydrogenase is not a part of the isomerization-based pathway, arabitol represents a dead end that can not be further metabolized. One way to avoid arabitol formation from arabinose during xylose cofermentation is to use the xylose isomerase based pathway which does not include xylose reductase (XR). This approach first proved difficult since the extensive selection procedure required for arabinose utilization, drastically lowered the capacity of the same strain for xylose utilization (Wisselink et al., 2007). A later improved evolutionary engineering strategy was, however, successful in developing a xylose/arabinose co-fermenting strain based on xylose and arabinose isomerase pathways (Wisselink et al., 2009). It should also be possible to reduce arabitol formation by combining the oxidation/reduction pathways for xylose- and arabinose-utilization (Fig. 9.2). In this way arabitol would not constitute a dead end and would be further metabolized to xylitol and ethanol. An oxidation/reduction pathway has been

ß Woodhead Publishing Limited, 2010

232

Bioalcohol production

investigated for arabinose metabolism but so far only partially successful results have been observed (Richard et al., 2003; Verho et al., 2004). Efficient arabinose utilization through an oxidation/reduction-based pathway have, however, recently been observed in the laboratory of the authors using a suitable combination of enzymes. These results point towards complete co-fermentation of xylose and arabinose without any evolutionary engineering.

9.4

Transcriptional regulation in mixed substrate fermentations

In S. cerevisiae it is well established that utilization of glucose is preferred next to other sugars such as sucrose, maltose and galactose. Preferential uptake of glucose is realized by transcriptional repression of genes encoding competing pathways. This process has been termed glucose repression and a brief outline of the involved mechanisms is presented in this section (Gancedo, 1998; Santangelo, 2006). Reduced glucose repression has been attempted by genetic engineering to improve fermentation of other carbon sources in the presence of glucose (Ostergaard et al., 2000, 2001; Roca et al., 2004; Bro et al., 2005; Velagapudi et al., 2006). Despite some success under carefully controlled conditions, deletion or over-expression of individual regulatory proteins has yielded little effect at the physiological level (Table 9.3). Co-fermentation of glucose during xylose utilization has, on the other hand, been seen to increase xylose consumption at low glucose concentrations (Jeffries et al., 1985; È hgren et al., 2006; Olofsson et al., Meinander and Hahn-HaÈgerdal, 1997; O 2008b). The cause of this effect is not completely understood; however, it has been suggested that xylose is not recognized as a `fermentable' carbon source at the metabolic or transcriptional level (Jin et al., 2004; Souto-Maior et al., 2009). It has thus been hypothesized that signaling for xylose-utilization is not optimal and that trace amounts of glucose are able to induce activity and/or expression of relevant enzymes.

9.4.1

Glucose signaling

For the purpose of this review, glucose signaling is structured into three categories: 1. repression of pathways utilizing non-fermentable carbon sources; 2. regulation of HXT transporter genes in response to substrate concentration; and 3. the Ras2/cAMP/PKA pathway regulating major cellular processes such as glycolysis, ribosome synthesis, replication and growth (Fig. 9.3). This text does not attempt to extend or summarize current knowledge in the field. Rather a simplified view is presented that relates specifically to cofermentation of lignocellulose derived sugars.

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

233

Table 9.3 Physiological response to altered glucose signaling in Saccharomyces cerevisiae Regulatory element

Function

Physiological effect with gene deletion

References

Mig1

Glucose repression

Partial de-repression of SUC2

Olsson et al. (1997) Olsson et al. (1997), Klein et al. (1999) Klein et al. (1999)

Increased co-consumption of sucrose in mixture with glucose Reduced lag phase between galactose and glucose consumption Slightly improvement of xylose utilization

Roca et al. (2004)

Mig2

Glucose repression

Phenotype similar as wild type strain

Klein et al. (1999)

Gal6/ Gal80

Repressor of galactose genes

Increase galactose uptake rate and ethanol productivity in co-fermentation with glucose

Ostergaard et al. (2000, 2001)

Snf3 and Rgt2

Glucose sensing

Increased glucose consumption rate Does not influence overflow metabolism

Raghevendran (2005)

Grr1

Carbon catabolic repression

Reduced growth rate in glucose No significant difference in biomass and ethanol yield in glucose

Raghevendran et al. (2004), Westergaard et al. (2004)

Hap4

Activation of respiratory metabolism

Only activates respiratory response at specific growth rates

van Maris et al. (2001), Lascaris et al. (2004), Raghevendran et al. (2006)

Nsf1/ Ypl230w

Putative transcription factor involved in sugar metabolism

Growth rate not affected in either fermentable or nonfermentable carbon sources

Hlynialuk et al. (2008)

Ylr042c

Protein of unknown function

Improved xylose metabolism in recombinant strains

Bengtsson et al. (2008), Parachin et al. (2010)

ß Woodhead Publishing Limited, 2010

234

Bioalcohol production

9.3 Schematic illustration of glucose signaling in Saccharomyces cerevisiae (adapted from Gancedo, 1998; Carlson, 1999; Johnston and Kim, 2005; Santangelo, 2006). Abbreviations explained in text.

Glucose repression The term glucose repression is used to describe the transcriptional down regulation of metabolic pathways involved in the utilization of alternative carbon sources (galactose, maltose, sucrose, etc.). Glucose repression is thus a subset of the total cellular response to glucose, which includes profound changes in cellular metabolism and activation and repression of a multitude of genes (Carlson, 1999; Santangelo, 2006). The signaling pathways and mechanisms involved in these processes have been studied and several detailed reviews on the topic are available (Gancedo, 1998; Belinchon and Gancedo, 2003; Santangelo, 2006). The main conductor of glucose repression is arguably the transcription factor Mig1, which directly or indirectly controls the expression of hundreds of genes (Nehlin et al., 1991; Klein et al., 1998; Santangelo, 2006). Mig1 inhibits expression of several genes (GAL1, SUC2, etc.) by direct binding to carbon source responsive elements (SCREs) in the promoter region. Several activating transcription factors (Gal4, Mal63, Cat8, etc.) are also repressed by Mig1. The repressive activity of Mig1 is controlled by the Snf1 complex (Treitel et al., 1998; Carlson, 1999). At high glucose concentrations the activity of Snf1 is inhibited and Mig1 is active. At low glucose concentrations on the other hand, the Snf1 kinase is active and Mig1 is

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

235

inactivated by phosphorylation. The exact manner by which the Snf1 complex senses the glucose concentration (or uptake rate) is still debated. Regulation of HXT transporters Expression of HXT genes in S. cerevisiae is regulated by two glucose specific sensors in the plasma membrane, Snf3 and Rgt2 (Ozcan and Johnston, 1999; Moriya and Johnston, 2004; Kim and Johnston, 2006). Snf3 is activated by low sugar concentrations and Rgt2 is activated by high sugar concentrations. Expression of HXT genes is repressed by the transcription factor Rgt1 in combination with the co-repressors Mth1 and Std1. Upon glucose signaling by Snf3 or Rgt2, degradation of Mth1 and Std1 is induced and repression by Rgt1 is lifted. The ability of Rgt1 to repress transcription of HXT genes is further diminished by phosporylation by protein kinase A (PKA) when glucose is abundant (Kim and Johnston, 2006). Ras2/cAMP/PKA dependent regulation The Ras2/cAMP/PKA regulatory pathway affects many different aspects of cellular metabolism and the complete network is far from elucidated. The activity of PKA is induced by glucose through several overlapping sensing mechanisms involving the sensors Ras2 and Gpa2 and conversion of ATP to cAMP by adenylate cyclase (Cyr1) (Thevelein and de Winde, 1999; Rolland et al., 2002). PKA itself detects the energetic state of the cell by sensing the cAMP level. Upon induction, PKA phosphorylates a magnitude of proteins in the cell and controls basic cellular functions such as ribosome synthesis and replication (Thevelein and de Winde, 1999; Rolland et al., 2002). Specifically, the rate of glycolysis seems to be connected to the activity of PKA. Several glycolytic genes are controlled by the upstream activating sequences (UAS), which contain binding sites for the transcriptional activators Rap1 and Gcr1 (Baker, 1991; Sasaki et al., 2005). The Rap1 and Gcr1 transcription factors are in turn activated by phosphorylation by PKA. The Gcr1 recognition sequence is very specific for glycolytic genes. It has been shown that
GCR1 mutants grown on glucose have increased activity of the TCA cycle and significantly lower growth rates and expression of glycolytic enzymes (Lopez and Baker, 2000; Sasaki et al., 2005). The growth rate on non-fermentable carbon sources on the other hand was unchanged. In this context it is tempting to speculate that regulation of glycolysis during fermentation of xylose and arabinose may be sub-optimal.

9.4.2

Engineering the regulatory response of S. cerevisiae

When the function of a regulatory pathway is known, deletion or overexpression of the involved genes seems a straightforward approach to redirect

ß Woodhead Publishing Limited, 2010

236

Bioalcohol production

cellular metabolism. The molecular interactions of glucose repression indicate that if repression is lifted, expression of pathways involved in the metabolism of non-fermentable sugars is increased. Lifting glucose repression during cofermentation of glucose and non-fermentable carbon sources should thus lead to simultaneous, as apposed to sequential, uptake of sugars. In some instances such an approach has been successful, which supports the proposed mechanism (Table 9.3 and Fig. 9.3). On the other hand, for most regulatory proteins there is redundancy in biological function where deletion of one protein has a tendency of being compensated by others. Thus in many cases deletion of transcription factors have led to little or no physiological effect (Table 9.3). Drastic attenuation of glucose repression is expected by down-regulation of Mig1, the main regulator of glucose repression (Fig. 9.3). Indeed deletion of MIG1 reduced repression of invertase (Suc2) and slightly increased co-consumption of sucrose during mixed sucrose/glucose batch fermentation (Olsson et al., 1997). The total sugar consumption rate in the glucose/sucrose fermentation was, however, slower in the
MIG1 strain and surprisingly co-consumption of maltose was reduced in mixed sugar fermentations. Partial de-repression of the catabolic system was later attributed to the presence of another transcription factor, MIG2 (Wu and Trumbly, 1998). Double deletion of MIG1 and MIG2, however, did not influence results significantly. The effect of MIG1 deletion was also studied with respect to galactose utilization. Reduced lag phase between glucose and galactose consumption was observed in mixed fermentation (Klein et al., 1999). Moreover, when MIG1 deletion was combined with deletion of specific galactose repressors, Gal6 and Gal80, glucose repression was further reduced with resulting increase in galactose uptake and ethanol productivity/ yield (Ostergaard et al., 2000, 2001). Slight improvement of xylose/glucose cofermentation was also seen for the MIG1 deletion (Roca et al., 2004). Since xylose is not a natural substrate for S. cerevisiae, the cause of the improvement is in this case not clear. Although quantitative physiological changes are evident in
MIG1 mutants, the broad repressive effect of this regulator makes it a rather blunt target for redirecting cellular metabolism. Alteration of regulatory proteins with more specific functions has, on the other hand, led to very small physiological effects. Deletion of glucose sensors Snf3 and Rgt2 (Fig. 9.3) were hypothesized to interrupt signaling of high extra-cellular glucose concentrations and affect Crabtree metabolism. Deletion of both sensors did, however, not prevent overflow metabolism to ethanol at high glucose concentration (Raghevendran, 2005). In the same signaling pathway, the transport repressor Rgt1p is deactivated by Grr1p (Ozcan and Johnston, 1995; Ozcan et al., 1996). As expected, deletion of Grr1p changed the transcription of a large number of genes including hexose transporters and the glucose sensors Snf3 and Rgt2 (Westergaard et al., 2004). Regardless, in batch cultivation on glucose only minor phenotypic changes were observed (Westergaard et al., 2004). Another regulatory protein,

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

237

Hap4, has been described to activate transcription of respiratory genes and regulation of yeast respiratory capacity (Santangelo, 2006). Nevertheless, overexpression and deletion of Hap4 only slightly affected ethanol yield and critical dilution rate in aerobic batch and glucose limited chemostat (van Maris et al., 2001; Lascaris et al., 2004; Raghevendran et al., 2006). Finally when it comes to putative regulatory proteins, their effect on mixed sugar fermentation and catabolic repression is even harder to elucidate. For instance, YPL230w is an uncharacterized orf suggested to be involved in glucose de-repression (Hlynialuk et al., 2008). Supporting this idea, the regulator was localized to the nucleus when the yeast was grown in non-fermentable carbon sources but not on glucose. Deletion of YPL230w, however, did not affect growth rate on either fermentable or non-fermentable carbon sources (Hlynialuk et al., 2008). Another uncharacterized orf, YLR042C, showed significantly reduced expression in strains selected for improved xylose utilization (Bengtsson et al., 2008). A follow-up study demonstrated that the effect of YLR042c deletion was dependent both on xylose concentration and on the capability of the yeast to utilize xylose (Parachin et al., 2010). Reduced glucose-repression and increased co-fermentation of glucose and xylose were also observed for a
YLR042c strain.

9.4.3

Regulatory response to novel substrates

The regulatory response to xylose is interesting since this substrate is not naturally utilized by S. cerevisiae. Since xylose is a new substrate, it is fair to assume that S. cerevisiae lacks an evolutionary optimized regulatory response to this substrate. This problem becomes apparent in that quite extensive metabolic engineering has had to be applied to the central carbon metabolism to enable proficient xylose usage in S. cerevisiae. For instance, it was seen that the level of xylulokinase (XKS1) limited xylose uptake (Xue and Ho, 1990), and that the non-oxidative branch of the pentose phosphate pathway was insufficient (Johansson and Hahn-HaÈgerdal, 2002). When these limitations and others were abolished, the xylose growth rate increased by an order of magnitude compared to simple integration of heterologous genes of the initial xylose metabolism (Karhumaa et al., 2005). Furthermore it was shown in the previous section that engineering the regulatory response of S. cerevisiae, is at least partly successful in increasing the uptake rate of non-fermentable sugars (galactose and xylose) in the presence of glucose. These results also indicate that the genetic regulation of S. cerevisiae is not optimized for xylose or galactose utilization and/or cofermentation. The regulatory response to xylose has been studied on a genetic, protein, and metabolic level (Salusjarvi et al., 2003, 2006, 2008; Wahlbom et al., 2003; Jin et al., 2004; Bengtsson et al., 2008). More research in this area is, however, needed to enable an optimal response by S. cerevisiae to pentose sugars.

ß Woodhead Publishing Limited, 2010

238

Bioalcohol production

Crabtree-negative characteristics of xylose fermentation Several lines of evidence suggest that xylose is not recognized as a Crabtreepositive sugar. Transcriptional analysis of aerobic batch cultivation has clearly pointed to differences in expression of key enzymes during glucose and xylose metabolism (Jin et al., 2004; Salusjarvi et al., 2008). On xylose, expression of genes associated non-fermentable growth (HXK1, PYK2, ADH2, etc.) is induced. In general, transcriptional analysis has indicated an overall higher flux in gluconeogenesis and TCA cycle during xylose consumption. Also at the metabolic level, major differences exist between glucose and xylose consumption (Souto-Maior et al., 2009). In carbon-limited chemostat cultures at low dilution rates, the metabolic profile of xylose and glucose consumption is identical. Under nitrogen-limited conditions, glycerol production was observed from both glucose and xylose, but overflow metabolism in terms of ethanol production was only seen from glucose. Glycerol production is indicative of saturation of the cellular oxidizing capacity and is observed in anaerobic conditions or during overflow metabolism (Geertman et al., 2006; Vemuri et al., 2007). Thus during nitrogen-limited aerobic cultivation on glucose and xylose, saturation of NADH oxidizing capacity was evident. However, saturation of oxidizing capacity only triggered overflow metabolism in the case of glucose. Effect of glucose concentration on xylose fermentation Glucose inhibits xylose utilization by displaying hundred-fold higher affinities for available transporters (Table 9.1). Low levels of glucose have, however, been observed to increase the rate of xylose fermentation by S. cerevisiae. The effect was first reported for the naturally xylose consuming yeast Pachysolen tannophilus (Jeffries et al., 1985). In this study, periodic glucose addition increased the aerobic yield of ethanol from xylose. In batch cultivation of S. cerevisiae it has been seen that xylose consumption is increased during glucose/ xylose co-consumption at low xylose concentrations (Meinander and HahnHaÈgerdal, 1997). Under simultaneous saccharification and fermentation (SSF), when glucose is steadily released at low concentrations, the xylose uptake rate È hgren et al., 2006; Olofsson et al., 2008b). and ethanol yield are increased (O Similarly, improved co-consumption of glucose and xylose was observed when cellobiose was utilized by a S. cerevisiae strain expressing a cell surface glucosidase (Nakamura et al., 2008). It was thus possible to control the rate of glucose degradation by different -glucosidase expression levels, and consequently also the rate of xylose uptake.

9.5

Conclusion and future trends

In this chapter we have discussed genetic and biochemical interactions that govern co-fermentation of lignocellulose-derived sugars in recombinant S.

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

239

cerevisiae. The resulting picture makes evident that a few obstacles prevent optimal fermentative response to particular sugar mixtures (e.g., glucosegalactose or glucose-xylose). From the perspective of industrial ethanol production, the type of transporters expressed in the cell may prevent co-consumption of individual constituents in mixed fermentations (e.g., glucose-galactose). The low affinity of native transporters towards particular substrates may also limit the rate of fermentation, e.g. xylose and arabinose. For co-consumption of xylose and arabinose, lack of enzyme substrate specificity may also prevent efficient ethanol formation. Finally, engineering of the regulatory response to mixed substrates, holds the potential to alter cellular metabolism towards specific needs. On account of the redundancy of the substrate signaling pathways, we believe, however, that a systems biology approach is necessary for industrial exploitation of these systems (Westergaard et al., 2007; Dikicioglu et al., 2008). In addition, most reports on the regulatory response to different sugars have been quite mechanistic and fundamental. Thus it remains to be seen if laboratory experiments with glucose repression and signaling can be translated to an industrial setting using lignocellulose hydrolysate. In this sense over-expression of transporters (endogenous or heterologous) or process chemical solutions such as SSF, may provide better solutions in the short term perspective.

9.6

References

Alkasrawi, M., Rudolf, A., Liden, G. and Zacchi, G. (2006) Influence of strain and cultivation procedure on the performance of simultaneous saccharification and fermentation of steam pretreated spruce. Enzyme and Microbial Technology, 38, 279±286. Almeida, J. R. M., Modig, T., Petersson, A., Hahn-HaÈgerdal, B., Liden, G. and GorwaGrauslund, M. F. (2007) Increased tolerance and conversion of inhibitors in lignocellulosic hydrolysates by Saccharomyces cerevisiae. Journal of Chemical Technology and Biotechnology, 82, 340±349. Baker, H. V. (1991) GCR1 of Saccharomyces cerevisiae encodes a DNA-binding protein whose binding is abolished by mutations in the CTTCC sequence motif. Proceedings of the National Academy of Sciences of the United States of America, 88, 9443±9447. Becker, J. and Boles, E. (2003) A modified Saccharomyces cerevisiae strain that consumes L-arabinose and produces ethanol. Applied and Environmental Microbiology, 69, 4144±4150. Belinchon, M. M. and Gancedo, J. M. (2003) Xylose and some non-sugar carbon sources cause catabolite repression in Saccharomyces cerevisiae. Archives of Microbiology, 180, 293±297. Bengtsson, O., Jeppsson, M., Sonderegger, M., Parachin, N. S., Sauer, U., HahnHaÈgerdal, B. and Gorwa-Grauslund, M. F. (2008) Identification of common traits in improved xylose-growing Saccharomyces cerevisiae for inverse metabolic engineering. Yeast, 25, 835±847. Bettiga, M., Hahn-HaÈgerdal, B. and Gorwa-Grauslund, M. F. (2008) Comparing the xylose reductase/xylitol dehydrogenase and xylose isomerase pathways in arabinose

ß Woodhead Publishing Limited, 2010

240

Bioalcohol production

and xylose fermenting Saccharomyces cerevisiae strains. Biotechnology for Biofuels, 1, 16. Bro, C., Knudsen, S., Regenberg, B., Olsson, L. and Nielsen, J. (2005) Improvement of galactose uptake in Saccharomyces cerevisiae through overexpression of phosphoglucomutase: Example of transcript analysis as a tool in inverse metabolic engineering. Applied and Environmental Microbiology, 71, 6465±6472. Carlson, M. (1999) Glucose repression in yeast. Current Opinion in Microbiology, 2, 202±207. DeJuan, C. and Lagunas, R. (1986) Inactivation of the galactose transport system in Saccharomyces cerevisiae. FEBS Letters, 207, 258±261. Dikicioglu, D., Pir, P., Onsan, Z. I., Ulgen, K. O., Kirdar, B. and Oliver, S. G. (2008) Integration of metabolic modeling and phenotypic data in evaluation and improvement of ethanol production using respiration-deficient mutants of Saccharomyces cerevisiae. Applied and Environmental Microbiology, 74, 5809± 5816. Eliasson, A., Christensson, C., Wahlbom, C. F. and Hahn-HaÈgerdal, B. (2000) Anaerobic xylose fermentation by recombinant Saccharomyces cerevisiae carrying XYL1, XYL2, and XKS1 in mineral medium chemostat cultures. Applied and Environmental Microbiology, 66, 3381±3386. Fonseca, C., Romao, R., de Sousa, H. R., Hahn-HaÈgerdal, B. and Spencer-Martins, I. (2007) L-Arabinose transport and catabolism in yeast. FEBS Journal, 274, 3589± 3600. Galbe, M. and Zacchi, G. (2007) Pretreatment of lignocellulosic materials for efficient bioethanol production. Advances in Biochemical Engineering/Biotechnology, 108, 41±65. Galbe, M., Sassner, P., Wingren, A. and Zacchi, G. (2007) Process engineering economics of bioethanol production. Advances in Biochemical Engineering/ Biotechnology, 108, 303±327. Gancedo, J. M. (1998) Yeast carbon catabolite repression. Microbiology and Molecular Biology Reviews, 62, 334±361. GaÂrdoÂnyi, M., Jeppsson, M., Liden, G., Gorwa-Grausland, M. F. and Hahn-HaÈgerdal, B. (2003a) Control of xylose consumption by xylose transport in recombinant Saccharomyces cerevisiae. Biotechnology and Bioengineering, 82, 818±824. GaÂrdoÂnyi, M., Osterberg, M., Rodrigues, C., Spencer-Martins, I. and Hahn-HaÈgerdal, B. (2003b) High capacity xylose transport in Candida intermedia PYCC 4715. FEMS Yeast Research, 3, 45±52. Geertman, J. M. A., van Maris, A. J. A., van Dijken, J. P. and Pronk, J. T. (2006) Physiological and genetic engineering of cytosolic redox metabolism in Saccharomyces cerevisiae for improved glycerol production. Metabolic Engineering, 8, 532±542. Hahn-HaÈgerdal, B., Galbe, M., Gorwa-Grauslund, M. F., Liden, G. and Zacchi, G. (2006) Bio-ethanol ± the fuel of tomorrow from the residues of today. Trends in Biotechnology, 24, 549±556. Hahn-HaÈgerdal, B., Karhumaa, K., Fonseca, C., Spencer-Martins, I. and GorwaGrauslund, M. F. (2007a) Towards industrial pentose-fermenting yeast strains. Applied Microbiology and Biotechnology, 74, 937±953. Hahn-HaÈgerdal, B., Karhumaa, K., Jeppsson, M. and Gorwa-Grauslund, M. F. (2007b) Metabolic engineering for pentose utilization in Saccharomyces cerevisiae. Advances in Biochemical Engineering/Biotechnology, 108, 147±177. Hamacher, T., Becker, J., GaÂrdoÂnyi, M., Hahn-HaÈgerdal, B. and Boles, E. (2002)

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

241

Characterization of the xylose-transporting properties of yeast hexose transporters and their influence on xylose utilization. Microbiology, 148, 2783±2788. Hector, R. E., Qureshi, N., Hughes, S. R. and Cotta, M. A. (2008) Expression of a heterologous xylose transporter in a Saccharomyces cerevisiae strain engineered to utilize xylose improves aerobic xylose consumption. Applied Microbiology and Biotechnology, 80, 675±684. Hlynialuk, C., Schierholtz, R., Vernooy, A. and van der Merwe, G. (2008) Nsf1/Ypl230w participates in transcriptional activation during non-fermentative growth and in response to salt stress in Saccharomyces cerevisiae. Microbiology, 154, 2482±2491. Ho, N. W., Chen, Z. and Brainard, A. P. (1998) Genetically engineered Saccharomyces yeast capable of effective cofermentation of glucose and xylose. Applied Environmental Microbiology, 64, 1852±1859. Horak, J. and Wolf, D. H. (1997) Catabolite inactivation of the galactose transporter in the yeast Saccharomyces cerevisiae: Ubiquitination, endocytosis, and degradation in the vacuole. Journal of Bacteriology, 179, 1541±1549. Jeffries, T. W. (1985) Emerging technology for fermenting D-xylose. Trends in Biotechnology, 3, 208±212. Jeffries, T. W. (2006) Engineering yeasts for xylose metabolism. Current Opinion in Biotechnology, 17, 320±326. Jeffries, T. W., Fady, J. H. and Lightfoot, E. N. (1985) Effect of glucose supplements on the fermentation of xylose by Pachysolen tannophilus. Biotechnology and Bioengineering, 27, 171±176. Jin, Y. S., Laplaza, J. M. and Jeffries, T. W. (2004) Saccharomyces cerevisiae engineered for xylose metabolism exhibits a respiratory response. Applied and Environmental Microbiology, 70, 6816±6825. Johansson, B. and Hahn-HaÈgerdal, B. (2002) The non-oxidative pentose phosphate pathway controls the fermentation rate of xylulose but not of xylose in Saccharomyces cerevisiae TMB3001. FEMS Yeast Research, 2, 277±282. Johnston, M. and Kim, J. H. (2005) Glucose as a hormone: receptor-mediated glucose sensing in the yeast Saccharomyces cerevisiae. Biochemical Society Transactions, 33, 247±252. Karhumaa, K., Hahn-HaÈgerdal, B. and Gorwa-Grauslund, M. F. (2005) Investigation of limiting metabolic steps in the utilization of xylose by recombinant Saccharomyces cerevisiae using metabolic engineering. Yeast, 22, 359±368. Karhumaa, K., Wiedemann, B., Hahn-HaÈgerdal, B., Boles, E. and Gorwa-Grauslund, M. F. (2006) Co-utilization of L-arabinose and D-xylose by laboratory and industrial Saccharomyces cerevisiae strains. Microbial Cell Factories, 5, 18. Katahira, S., Ito, M., Takema, H., Fujita, Y., Tanino, T., Tanaka, T., Fukuda, H. and Kondo, A. (2008) Improvement of ethanol productivity during xylose and glucose co-fermentation by xylose-assimilating S. cerevisiae via expression of glucose transporter Sut1. Enzyme and Microbial Technology, 43, 115±119. Kim, J. H. and Johnston, M. (2006) Two glucose-sensing pathways converge on Rgt1 to regulate expression of glucose transporter genes in Saccharomyces cerevisiae. Journal of Biological Chemistry, 281, 26144±26149. Klein, C. J. L., Olsson, L. and Nielsen, J. (1998) Glucose control in Saccharomyces cerevisiae: the role of MIG1 in metabolic functions. Microbiology-Sgm, 144, 13± 24. Klein, C. J., Rasmussen, J. J., Ronnow, B., Olsson, L. and Nielsen, J. (1999) Investigation of the impact of MIG1 and MIG2 on the physiology of Saccharomyces cerevisiae. Journal of Biotechnology, 68, 197±212.

ß Woodhead Publishing Limited, 2010

242

Bioalcohol production

KoÈtter, P. and Ciriacy, M. (1993) Xylose fermentation by Saccharomyces cerevisiae. Applied Microbiology and Biotechnology, 38, 776±783. Kou, S. C., Christen and Cirillo, V. P. (1970) Galactose transport in Saccharomyces cerevisiae. 2. Characteristics of galactose uptake and exchange in galactokinaseless cells. Journal of Bacteriology, 103, 671±678. Kuyper, M., Harhangi, H. R., Stave, A. K., Winkler, A. A., Jetten, M. S. M., de Laat, W., den Ridder, J. J. J., Op den Camp, H. J. M., van Dijken, J. P. and Pronk, J. T. (2003) High-level functional expression of a fungal xylose isomerase: the key to efficient ethanolic fermentation of xylose by Saccharomyces cerevisiae? FEMS Yeast Research, 4, 69±78. Lascaris, R., Piwowarski, J., van der Spek, H., Teixeira de Mattos, J., Grivell, L. and Blom, J. (2004) Overexpression of HAP4 in glucose-derepressed yeast cells reveals respiratory control of glucose-regulated genes. Microbiology, 150, 929±934. Leandro, M. J., Goncalves, P. and Spencer-Martins, I. (2006) Two glucose/xylose transporter genes from the yeast Candida intermedia: first molecular characterization of a yeast xylose-H+ symporter. Biochemical Journal, 395, 543± 549. Leandro, M. J., Spencer-Martins, I. and Goncalves, P. (2008) The expression in Saccharomyces cerevisiae of a glucose/xylose symporter from Candida intermedia is affected by the presence of a glucose/xylose facilitator. Microbiology, 154, 1646± 1655. Lee, W. J., Kim, M. D., Ryu, Y. W., Bisson, L. F. and Seo, J. H. (2002) Kinetic studies on glucose and xylose transport in Saccharomyces cerevisiae. Applied Microbiology and Biotechnology, 60, 186±191. Lopez, M. C. and Baker, H. V. (2000) Understanding the growth phenotype of the yeast gcr1 mutant in terms of global genomic expression patterns. Journal of Bacteriology, 182, 4970±4978. Maier, A., Volker, B., Boles, E. and Fuhrmann, G. F. (2002) Characterisation of glucose transport in Saccharomyces cerevisiae with plasma membrane vesicles (countertransport) and intact cells (initial uptake) with single Hxt1, Hxt2, Hxt3, Hxt4, Hxt6, Hxt7 or Gal2 transporters. FEMS Yeast Research, 2, 539±550. Matern, H. and Holzer, H. (1977) Catabolite inactivation of the galactose uptake system in yeast. Journal of Biological Chemistry, 252, 6399±6402. Meinander, N. Q. and Hahn-HaÈgerdal, B. (1997) Influence of cosubstrate concentration on xylose conversion by recombinant, XYL1-expressing Saccharomyces cerevisiae: a comparison of different sugars and ethanol as cosubstrates. Applied Environmental Microbiology, 63, 1959±1964. Moriya, H. and Johnston, M. (2004) Glucose sensing and signaling in Saccharomyces cerevisiae through the Rgt2 glucose sensor and casein kinase I. Proceedings of the National Academy of Sciences of the United States of America, 101, 1572±1577. Nakamura, N., Yamada, R., Katahira, S., Tanaka, T., Fukuda, H. and Kondo, A. (2008) Effective xylose/cellobiose co-fermentation and ethanol production by xyloseassimilating S. cerevisiae via expression of beta-glucosidase on its cell surface. Enzyme and Microbial Technology, 43, 233±236. Nehlin, J. O., Carlberg, M. and Ronne, H. (1991) Control of yeast GAL genes by Mig1 repressor ± a transcriptional cascade in the glucose response. EMBO Journal, 10, 3373±3377. È hgren, K., Bengtsson, O., Gorwa-Grauslund, M. F., Galbe, M., Hahn-HaÈgerdal, B. and O Zacchi, G. (2006) Simultaneous saccharification and co-fermentation of glucose and xylose in steam-pretreated corn stover at high fiber content with

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

243

Saccharomyces cerevisiae TMB3400. Journal of Biotechnology, 126, 488±498. Olofsson, K., Bertilsson, M. and Liden, G. (2008a) A short review on SSF ± an interesting process option for ethanol production from lignocellulosic feedstocks. Biotechnology for Biofuels, 1, 7. Olofsson, K., Rudolf, A. and Liden, G. (2008b) Designing simultaneous saccharification and fermentation for improved xylose conversion by a recombinant strain of Saccharomyces cerevisiae. Journal of Biotechnology, 134, 112±120. Olsson, L., Larsen, M. E., Ronnow, B., Mikkelsen, J. D. and Nielsen, J. (1997) Silencing MIG1 in Saccharomyces cerevisiae: Effects of antisense MIG1 expression and MIG1 gene disruption. Applied and Environmental Microbiology, 63, 2366±2371. Ostergaard, S., Olsson, L., Johnston, M. and Nielsen, J. (2000) Increasing galactose consumption by Saccharomyces cerevisiae through metabolic engineering of the GAL gene regulatory network. Nature Biotechnology, 18, 1283±1286. Ostergaard, S., Walloe, K. O., Gomes, C. S. G., Olsson, L. and Nielsen, J. (2001) The impact of GAL6, GAL80, and MIG1 on glucose control of the GAL system in Saccharomyces cerevisiae. FEMS Yeast Research, 1, 47±55. Ozcan, S. and Johnston, M. (1995) Three different regulatory mechanisms enable yeast hexose transporter (HXT) genes to be induced by different levels of glucose. Molecular and Cellular Biology, 15, 1564±1572. Ozcan, S. and Johnston, M. (1999) Function and regulation of yeast hexose transporters. Microbiology and Molecular Biology Reviews, 63, 554±569. Ozcan, S., Leong, T. and Johnston, M. (1996) Rgt1p of Saccharomyces cerevisiae, a key regulator of glucose-induced genes, is both an activator and a repressor of transcription. Molecular and Cellular Biology, 16, 6419±26. Parachin, N. S., Bengtsson, O., Hanh-HaÈgerdal, B. and Gorwa-Grauslund, M. F (2010) Deletion of YLR042c improves ethanolic xylose fermentation by recombinant Saccharomyces cerevisiae. Manuscript submitted for publication. Raghevendran, V. (2005) Glucose regulation in Saccharomyces cerevisiae: a physiogenomic study. Ph.D Thesis. Raghevendran, V., Gombert, A. K., Christensen, B., KoÈtter, P. and Nielsen, J. (2004) Phenotypic characterization of glucose repression mutants of Saccharomyces cerevisiae using experiments with 13C-labelled glucose. Yeast, 21, 769±779. Raghevendran, V., Patil, K. R., Olsson, L. and Nielsen, J. (2006) Hap4 is not essential for activation of respiration at low specific growth rates in Saccharomyces cerevisiae. Journal of Biological Chemistry, 281, 12308±12314. Reifenberger, E., Boles, E. and Ciriacy, M. (1997) Kinetic characterization of individual hexose transporters of Saccharomyces cerevisiae and their relation to the triggering mechanisms of glucose repression. European Journal of Biochemistry, 245, 324± 333. Richard, P., Verho, R., Putkonen, M., Londesborough, J. and Penttila, M. (2003) Production of ethanol from L-arabinose by Saccharomyces cerevisiae containing a fungal L-arabinose pathway. FEMS Yeast Research, 3, 185±189. Rizzi, M., Erlemann, P., Buithanh, N. A. and Dellweg, H. (1988) Xylose fermentation by yeasts. 4. Purification and kinetic-studies of xylose reductase from Pichia stipitis. Applied Microbiology and Biotechnology, 29, 148±154. Roca, C., Haack, M. B. and Olsson, L. (2004) Engineering of carbon catabolite repression in recombinant xylose fermenting Saccharomyces cerevisiae. Applied Microbiology and Biotechnology, 63, 578±583. Rolland, F., Winderickx, J. and Thevelein, J. M. (2002) Glucose-sensing and -signalling mechanisms in yeast. FEMS Yeast Research, 2, 183±201.

ß Woodhead Publishing Limited, 2010

244

Bioalcohol production

Rudolf, A., Alkasrawi, M., Zacchi, G. and Liden, G. (2005) A comparison between batch and fed-batch simultaneous saccharification and fermentation of steam pretreated spruce. Enzyme and Microbial Technology, 37, 195±204. Rudolf, A., Baudel, H., Zacchi, G., Hahn-HaÈgerdal, B. and Liden, G. (2008) Simultaneous saccharification and fermentation of steam-pretreated bagasse using Saccharomyces cerevisiae TMB3400 and Pichia stipitis CBS6054. Biotechnology and Bioengineering, 99, 783±790. Runquist, D., Fonseca, C., RaÊdstroÈm, P., Spencer-Martins, I. and Hahn-HaÈgerdal, B. (2008) Expression of the Gxf1 transporter from Candida intermedia improves fermentation performance in recombinant xylose-utilizing Saccharomyces cerevisiae. Applied Microbiology and Biotechnology, 82, 123. Saloheimo, A., Rauta, J., Stasyk, O. V., Sibirny, A. A., Penttila, M. and Ruohonen, L. (2007) Xylose transport studies with xylose-utilizing Saccharomyces cerevisiae strains expressing heterologous and homologous permeases. Applied Microbiology and Biotechnology, 74, 1041±1052. Salusjarvi, L., Poutanen, M., Pitkanen, J. P., Koivistoinen, H., Aristidou, A., Kalkkinen, N., Ruohonen, L. and Penttila, M. (2003) Proteome analysis of recombinant xylosefermenting Saccharomyces cerevisiae. Yeast, 20, 295±314. Salusjarvi, L., Pitkanen, J. P., Aristidou, A., Ruohonen, L. and Penttila, M. (2006) Transcription analysis of recombinant Saccharomyces cerevisiae reveals novel responses to xylose. Applied Biochemistry and Biotechnology, 128, 237±261. Salusjarvi, L., Kankainen, M., Soliymani, R., Pitkanen, J. P., Penttila, M. and Ruohonen, L. (2008) Regulation of xylose metabolism in recombinant Saccharomyces cerevisiae. Microbial Cell Factories, 7, 16. Santangelo, G. M. (2006) Glucose signaling in Saccharomyces cerevisiae. Microbiology and Molecular Biology Reviews, 70, 253±282. Sasaki, H., Kishimoto, T., Mizuno, T., Shinzato, T. and Uemura, H. (2005) Expression of GCR1, the transcriptional activator of glycolytic enzyme genes in the yeast Saccharomyces cerevisiae, is positively autoregulated by Gcr1p. Yeast, 22, 305± 319. Schmer, M. R., Vogel, K. P., Mitchell, R. B. and Perrin, R. K. (2008) Net energy of cellulosic ethanol from switchgrass. Proceedings of the National Academy of Sciences of the United States of America, 105, 464±469. Sedlak, M. and Ho, N. W. (2004) Characterization of the effectiveness of hexose transporters for transporting xylose during glucose and xylose co-fermentation by a recombinant Saccharomyces yeast. Yeast, 21, 671±684. Skoog, K. and Hahn-HaÈgerdal, B. (1988) Xylose fermentation. Enzyme and Microbial Technology, 10, 66±80. Souto-Maior, A., Runquist, D. and Hahn-HaÈgerdal, B. (2009) Crabtree-negative characteristics of recombinant xylose-utilizing Saccharomyces cerevisiae. Journal of Biotechnology, 143, 119±123. Thevelein, J. M. and de Winde, J. H. (1999) Novel sensing mechanisms and targets for the cAMP-protein kinase A pathway in the yeast Saccharomyces cerevisiae. Molecular Microbiology, 33, 904±918. Treitel, M. A., Kuchin, S. and Carlson, M. (1998) Snf1 protein kinase regulates phosphorylation of the Mig1 repressor in Saccharomyces cerevisiae. Molecular and Cellular Biology, 18, 6273±6280. van Maris, A. J., Bakker, B. M., Brandt, M., Boorsma, A., Teixeira de Mattos, M. J., Grivell, L. A., Pronk, J. T. and Blom, J. (2001) Modulating the distribution of fluxes among respiration and fermentation by overexpression of HAP4 in Saccharomyces

ß Woodhead Publishing Limited, 2010

Challenges in co-fermentation

245

cerevisiae. FEMS Yeast Research, 1, 139±149. Velagapudi, V. R., Wittmann, C., Lengauer, T., Talwar, P. and Heinzle, E. (2006) Metabolic screening of Saccharomyces cerevisiae single knockout strains reveals unexpected mobilization of metabolic potential. Process Biochemistry, 41, 2170± 2179. Vemuri, G. N., Eiteman, M. A., McEwen, J. E., Olsson, L. and Nielsen, J. (2007) Increasing NADH oxidation reduces overflow metabolism in Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences of the United States of America, 104, 2402±2407. Verduyn, C., Vankleef, R., Frank, J., Schreuder, H., Vandijken, J. P. and Scheffers, W. A. (1985) Properties of the NAD(P)H-dependent xylose reductase from the xylosefermenting yeast Pichia stipitis. Biochemical Journal, 226, 669±677. Verho, R., Putkonen, M., Londesborough, J., Penttila, M. and Richard, P. (2004) A novel NADH-linked L-xylulose reductase in the L-arabinose catabolic pathway of yeast. Journal of Biological Chemistry, 279, 14746±14751. Wahlbom, C. F., Otero, R. R. C., van Zyl, W. H., Hahn-HaÈgerdal, B. and Jonsson, L. J. (2003) Molecular analysis of a Saccharomyces cerevisiae mutant with improved ability to utilize xylose shows enhanced expression of proteins involved in transport, initial xylose metabolism, and the pentose phosphate pathway. Applied and Environmental Microbiology, 69, 740±746. Walfridsson, M., Bao, X., Anderlund, M., Lilius, G., Bulow, L. and Hahn-HaÈgerdal, B. (1996) Ethanolic fermentation of xylose with Saccharomyces cerevisiae harboring the Thermus thermophilus xylA gene, which expresses an active xylose (glucose) isomerase. Applied and Environmental Microbiology, 62, 4648±4651. Westergaard, S. L., Bro, C., Olsson, L. and Nielsen, J. (2004) Elucidation of the role of Grr1p in glucose sensing by Saccharomyces cerevisiae through genome-wide transcription analysis. FEMS Yeast Research, 5, 193±204. Westergaard, S. L., Oliveira, A. P., Bro, C., Olsson, L. and Nielsen, J. (2007) A systems biology approach to study glucose repression in the yeast Saccharomyces cerevisiae. Biotechnology and Bioengineering, 96, 134±145. Wisselink, H. W., Toirkens, M. J., del Rosario Franco Berriel, M., Winkler, A. A., van Dijken, J. P., Pronk, J. T. and van Maris, A. J. (2007) Engineering of Saccharomyces cerevisiae for efficient anaerobic alcoholic fermentation of Larabinose. Applied and Environmental Microbiology, 73, 4881±4891. Wisselink, H. W., Toirkens, M. J., Wu, Q., Pronk, J. T. and van Maris, A. J. (2009) Novel evolutionary engineering approach for accelerated utilization of glucose, xylose and arabinose mixtures by engineered Saccharomyces cerevisiae. Applied and Environmental Microbiology, 75, 907±914. Wu, J. and Trumbly, R. J. (1998) Multiple regulatory proteins mediate repression and activation by interaction with the yeast Mig1 binding site. Yeast, 14, 985±1000. Xue, X. D. and Ho, N. W. Y. (1990) Xylulokinase activity in various yeasts including Saccharomyces cerevisiae containing the cloned xylulokinase gene. Applied Biochemistry and Biotechnology, 24-5, 193±199.

ß Woodhead Publishing Limited, 2010

10

Separation and purification processes for lignocellulose-to-bioalcohol production H . - J . H U A N G , S . R A M A S W A M Y and U . W . T S C H I R N E R , University of Minnesota, USA and B . V . R A M A R A O , State University of New York, USA

Abstract: The chapter begins by introducing the lignocellulosic biomass-toethanol biorefinery. It then reviews and discusses the separation methods and technologies in bioalcohol production, product separation and dehydration and removal of inhibitors in fermentation systems, including extractive distillation with ionic liquids and hyper-branched polymers, adsorption with molecular sieve and bio-based adsorbents, extractive fermentation, membrane pervaporation in bioreactors, and vacuum membrane distillation (VMD), etc. Key challenges and opportunities in separation technologies for the future biorefinery are also presented. Key words: ethanol, bioalcohol, biorefinery, separation and purification, extractive fermentation, membrane separation-bioreactor hybrid.

10.1

Introduction

Bioalcohol or bioethanol is expected to be one of the most significant liquid transportation fuels in future lignocellulosic biorefineries. Hence, this chapter will focus on separation technologies incorporating bioethanol as a principal product. Basically, there are two kinds of lignocellulosic biomass-to-ethanol biorefineries: first, basic lignocellulosic biomass-to-ethanol, and second, integrated lignocellulosic biorefineries including biomass-to-ethanol and other co-products.

10.1.1 Basic lignocellulosic biomass-to-ethanol biorefinery Cellulosic ethanol is a more promising alternative renewable bio-fuel in the future than today's starch based (corn) ethanol due to its higher net fossil fuel displacement potential as well as net greenhouse gas (GHG) emissions reductions. Lignocellulosic biomass as the feedstock for ethanol includes agricultural residues (corn stover, rice/wheat straws, sugarcane bagasse), herbaceous crops (alfalfa, switchgrass), wood (hardwoods, softwoods) and forestry wastes,

ß Woodhead Publishing Limited, 2010

Separation and purification processes

247

10.1 Overall process block diagram for a basic lignocellulose to ethanol biorefinery.

wastepaper, and other wastes such as municipal waste (Wyman, 1996; Huang et al., 2009a). As seen in other chapters in this book, the basic process for conversion of cellulosic biomass to fuel ethanol involves the following steps: feedstock handling (unwrapping, washing, shredding, etc.), pretreatment and hydrolysate conditioning (hemicellulose hydrolysis, hydrolysate overliming and neutralization, and solid-liquid separation, etc.), saccharification and co-fermentation, product separation and purification, wastewater treatment, product storage, lignin combustion for production of electricity and steam, and all other utilities (Fig. 10.1) (Aden et al., 2002).

10.1.2 Integrated lignocellulosic biorefinery (ILCB) Forests are an enormous source of lignocellulosic biomass, and the pulp mills represent an excellent existing platform for retrofitting in forest products industry. In recent years, however, the US forest products industry has encountered severe competition from overseas. Thus, the forest biorefinery based on the existing pulp mills, i.e. integrated lignocellulose biorefinery, was proposed to produce added fuel and chemicals, along with pulp and paper, in order to increase the overall revenue and profit (Amidon, 2006; Liu et al., 2006; Huang et al., 2008). The process of ILCB (Fig. 10.2) involves hemicellulose extraction prior to pulping, isolation of long and short fiber after pulping, hemicellulose conversion to ethanol, the short fiber (cellulose) conversion to ethanol, and the long fiber (cellulose) for production of paper and other fiber based materials such as bio-composites. Besides, lignin dissolved into black liquor after pulping

ß Woodhead Publishing Limited, 2010

248

Bioalcohol production

10.2 Process block diagram of an integrated forest biorefinery.

can be further gasified to produce syngas, which can be further synthesized to produce fuels and chemicals, and electricity and process steam (Huang et al., 2009b, 2010). ILCB, therefore, can make full use of all the biomass feedstock components to produce, in addition to pulp and paper and bio-alcohol, other value-added multiple co-products including various chemicals and energy (electricity and steam). This biorefinery producing multiple co-products similar to today's petroleum refinery offers enormous opportunity for the future bioeconomy based on sustainable use of renewable resources.

10.1.3 Separation and purification in lignocellulosic biorefineries This chapter aims to explore various potential separation methods and technologies, which might help reduce the overall ethanol production cost and improve the overall techno-economic feasibility of the biorefinery. Thus, the following sections will focus on variety of separation technologies in recovery and dehydration of ethanol and removal of fermentation inhibitors for increasing product yield. As the fermentation beer contains only 5±12 wt% ethanol, the separation of ethanol from this dilute aqueous solution is an energy-intensive process. In addition, ethanol forms an azeotrope at 95.6 wt% with water at 78.15 ëC, which makes it impossible to separate and dehydrate ethanol from the beer solution in a single distillation column. Hence, a two-step process is often applied: firstly, the beer is concentrated to around 92.4 wt% ethanol by ordinary distillation, then the resulting ethanol is further dehydrated to pure ethanol by using azeotropic distillation, extractive distillation, liquid-liquid extraction, adsorption, or mem-

ß Woodhead Publishing Limited, 2010

Separation and purification processes

249

brane pervaporation. In addition, in-situ removal of ethanol from fermentation broth (usually below 10 wt% ethanol) is used to eliminate the product ethanol inhibition and to assist in the pre-concentration process. In the following, the various separation technologies suitable for bio-alcohol production are described in detail.

10.2

Azeotropic distillation (AD)

Azeotropic distillation (AD) is a process to break azeotrope where another volatile component, called the entrainer, the solvent, or the mass separating agent (MSA), is added to form a new lower-boiling azeotrope that is heterogeneous. As illustrated in Fig. 10.3, the AD process comprises two distillation columns: an azeotropic column for dehydration of 92.4 wt% ethanol solution from the preconcentration step with the aid of entrainer, and a stripping column for separation of entrainer from the product stream. In the azeotropic column, ethanol product (>99 wt%) leaves the bottoms. The ternary azeotrope formed containing water vapor, entrainer, and small amounts of ethanol exit from the tops, and then enters a separator (called decanter), and splits into organic phase (ethanol-entrainer) and aqueous phase (water-entrainer) streams. The former is refluxed back into the first column, while the latter is processed in the stripping column for recovery of entrainer and ethanol (Lee and Pahl, 1985; Kovach III and Seider, 1987; Chianese and Zinnamosca, 1990; Luyben, 2006). The common entrainers used for separating the ethanol±water mixture by heterogeneous azeotropic distillation are benzene (Chianese and Zinnamosca, 1990; Wasylkiewicz et al., 2003), toluene (Partin, 1996; Feng et al., 2000),

10.3 Flow sheet of AD system for ethanol dehydration (Chianese and Zinnamosca, 1990; Luyben, 2006).

ß Woodhead Publishing Limited, 2010

250

Bioalcohol production

cyclohexane (Gomis et al., 2005), and a mixed solvent, e.g. a mixture of benzene and n-octane (Chianese and Zinnamosca, 1990). Health and safety concerns, such as carcinogenic effect of benzene, and the flammability of cyclohexane, must be considered during the selection of entrainers. The AD system described above is less utilized in the ethanol separation and purification due to its disadvantages of large capital cost, high-energy requirement, and health and safety concerns.

10.3

Extractive distillation (ED)

Extractive distillation (ED) is the vapor-liquid separation process with the addition of a miscible, high boiling or relatively non-volatile component (solvent), to increase the relative volatility of the components to be separated and hence increase the separation factor. This method is commonly applied in chemical industry to separate close boiling point or azeotropic mixtures. The added component as separating agent can be conventional liquid solvent, dissolved salt, a mixture of conventional liquid solvent and dissolved salt, ionic liquid, or hyper-branched polymer.

10.3.1 ED with conventional solvent In the typical ED process for ethanol dehydration (Fig. 10.4), a conventional liquid solvent of high-boiling point is introduced in the upper part above the feed. Ethylene glycol is an effective solvent in ED for dehydration of ethanol from the fermentation broth. With this solvent, anhydrous ethanol can be obtained in a column with only 18 theoretical stages, a low reflux ratio of 1.5, and a low solvent/feed ratio of 0.27 (Lee and Pahl, 1985). This process was shown to be competitive compared to azeotropic distillation (Maciel and Brito, 1995).

10.4 Extractive distillation.

ß Woodhead Publishing Limited, 2010

Separation and purification processes

251

10.5 Extractive distillation with gasoline for ethanol dehydration (Chianese and Zinnamosca, 1990).

Gasoline is a suitable solvent for ED of fermentation liquor to produce motor fuel ethanol (gasohol). In the gasohol production process (Fig. 10.5), the ethanol±water volatility is reversed in the presence of gasoline containing high percent of heavier hydrocarbon (C7 and C8) fractions, causing the mixture of ethanol and the heavier hydrocarbon (HH) fractions to exit at the bottom, and the mixture of water, lighter hydrocarbons (LH) and residual ethanol to be withdrawn overhead. The bottom stream is then mixed with the organic phase of the decanter, mainly consisting of LH, to provide the final gasohol product (Chianese and Zinnamosca, 1990).

10.3.2 ED with dissolved salt For the ethanol±water system, relatively small amounts of a dissolved salt can be added as mass separating agent in ED (Fig. 10.6) to significantly increase the relative volatility of the more volatile component of the mixture (LlanoRestrepo and Aguilar-Arias, 2003). The commonly tested dissolved salts in ED for ethanol dehydration are acetates such as potassium acetate (KAc) (Cook and Furter, 1968; Furter, 1972; Lynd and Grethlein, 1984) and sodium acetate (NaAc) (Furter, 1992) and haloids including CaCl2 (Barba et al., 1985; Pinto et al., 2000; Llano-Restrepo and Aguilar-Arias, 2003), NaCl, KCl, and KI (Pinto et al., 2000). Cook and Furter (1968) used a pilotscale bubble-cap tray column to study ED with KAc, and found that KAc is effective in breaking the ethanol±water azeotrope and the ED with dissolved salts is more efficient than with conventional liquid extractants in ethanol±water separation. Besides, a mixture of two or more salts, e.g. a 70/30 mixture of potassium and sodium acetate, can also be utilized in ED, with lower capital cost and energy consumption compared to azeotropic distillation with benzene and ED with ethylene glycol (Furter, 1992; Llano-Restrepo and Aguilar-Arias, 2003).

ß Woodhead Publishing Limited, 2010

252

Bioalcohol production

10.6 Extractive distillation with dissolved salt (Furter, 1972).

By comparison of the energy requirements between ED with CaCl2 and azeotropic distillation with benzene, pentane or diethyl ester, ED with ethylene glycol or gasoline, solvent extraction, and membrane pervaporation, Barba et al. (1985) found that ED with CaCl2 and membrane pervaporation have similar energy consumptions and they are much better than the other separation approaches in energy-savings. Pinto et al. (2000) also found, by process simulation using Aspen Plus that the saline ED with CaCl2 has lower energy consumption than conventional ED with ethylene glycol, which is in agreement with Barba et al. (1985). They also made comparison between NaCl, KCl, KI and CaCl2 as separating agents in ED for ethanol dehydration, and found that CaCl2 provides the largest salting out effect on ethanol. The saline ED processes can be further optimized to lower energy consumption and/or capital costs by considering process flowsheet improvement and heat integration, etc. For instance, Ligero and Ravagnani (2003) compared two different process systems of ED with KAc by simulation. The first system consists of an ED column where the fermentation beer was fed into a multiple effect evaporator and a spray dryer for recovery of salt, which is recycled to the column. The second system is composed of a conventional distillation, by which fermentation beer is firstly pre-concentrated, an ED column for further concentration of ethanol, and a single spray dryer for salt recovery. The results show that the second system has less energy consumption than the first. Lynd and Grethlein (1984) optimized a saline ED process system by heat integration for ethanol separation and dehydration. The process composed of a preconcentration column with intermediate heat pumps and optimal side-stream return, a saline ED column with KAc as an agent, a salt-concentrating evaporator, and a spray dryer was shown to have lower capital costs and much less energy consumption as compared with conventional separation methods. Briefly, the saline ED is a better process to obtain anhydrous ethanol from the

ß Woodhead Publishing Limited, 2010

Separation and purification processes

253

fermentation broth, due to the use of only one column, which requires the lowest energy consumption, and uses non-toxic solvent.

10.3.3 ED with mixture of conventional liquid solvent and dissolved salt The mixture of conventional liquid extractant (predominant fraction) and dissolved salt (a very small fraction) can also be used as mass separating agent in ED for ethanol purification. The commonly used dissolved salts in this ED process include NaCl, CaCl2, AlCl3, KNO3, and CH3COOK, etc. (Duan et al., 1980; Lei et al., 1982; Zhang et al., 1984). With combined use of ethylene glycol (conventional solvent) and one of the above mentioned dissolved salts, the relative volatilities of light components were found in the range of 1.9±4.2, with the decreasing order of salt effect: AlCl3 > CaCl2 > NaCl, and decreasing order of the effect of acidic roots: Ac± > Cl± > NO±3, where the volume ratio of ethanol solution to separating agent is 1.0, and the salt concentration is 0.2 g salt per ml of solvent. The vapor±liquid equilibria of three systems ± ethanol/water, ethanol/water/ethylene glycol, and ethanol/water/ethylene glycol±CaCl2, measured by Lei et al. (2002) showed that the ED with combined ethylene glycol and dissolved salt was more efficient in separating ethanol and water than with ethylene glycol only.

10.3.4 ED with hyper-branched polymers Non-volatile hyper-branched polymers, the highly branched macromolecules with a large number of functional groups, are also novel separating agents used in ED for ethanol dehydration. They can be produced by one-step reactions, representing economically favorable agents for large-scale industrial applications (Seiler et al., 2004). Unlike linear polymers, hyper-branched polymers have high selectivity and capacity, low viscosity, and good thermal stability. The commonly tested hyper-branched polymers as ED separating agents for ethanol dehydration are poly(ethylene glycol) (PEG) (Al-Amer Am, 2000), poly(acrylic acid) (PAA) (Al-Amer Am, 2000), and hyper-branched polyglycerol (PG) (Seiler et al., 2004). Al-Amer Am (2000) measured the vapor-liquid equilibrium (VLE) data of the ethanol±water-PEG and ethanol±water-PAA systems, and found that PEG at 10 wt% and PAA at 0.45 wt% can break the ethanol±water azeotrope to produce anhydrous ethanol. On the other hand, Seiler et al. (2004) found that the PG effect on the relative volatility of ethanol over water was the same order as that of the conventional entrainer 1,2-ethanediol, and the overall heat duty can be saved up to 19% for ED with PG, compared to the conventional ED process (Seiler et al., 2004).

ß Woodhead Publishing Limited, 2010

254

Bioalcohol production

10.3.5 ED with ionic liquid Over the last few years, Ionic liquids (ILs) have gained increasing attention for their high potential applications as entrainers in extractive distillation for separation of close boiling mixtures or azeotropes. ILs consisting of an organic cation and an inorganic anion are novel separating agents used in ED for separation of ethanol±water mixture (Arlt et al., 2001). They can greatly increase the relative volatility of ethanol over water, due to the salt effect similar to the conventional dissolved salts. They are promising separating agents for ED of ethanol±water mixture, due to their low melting point, negligible volatility leading to a decrease in VOC emissions, low viscosity, thermal stability, good solubility and lower corrosiveness than ordinary high melting salts. The process of ED with ILs is superior to the ED with combined use of liquid solvent and conventional salt in that the former has higher separation ability, easy operation, and no entrainment of the solvent into the overhead product of the column (Lei et al., 2005). The commercially available ILs suitable for the extractive distillation of ethanol from beer liquor, are 1-butyl-3-methylimidazolium tetrafluoroborate ([BMIM] + [BF4] ÿ), 1-ethyl-3-methylimidazolium tetrafluoroborate ([EMIM]+[BF4]±) and 1-butyl-3-methylimidazolium chloride ([BMIM]+[Cl]±). The comparison between these ILs as separating agent in the ED (Fig. 10.7) for ethanol dehydration made by Seiler et al. (2004) showed that these ILs remarkably enhance the relative volatility of ethanol to water, in the following decreasing order: [BMIM] +[Cl]± > [EMIM]+[BF4]± > [BMIM]+[BF4]±, and It was also found that the influence of [BMIM] +[Cl]± and [EMIM]+[BF4]± on the relative volatility is greater than that of the conventional separating agent 1,2ethanediol, and the overall heat duty can be saved up to 24% for the [EMIM]+[BF4]± process in comparison with the conventional ED process.

10.7 Extractive distillation using ionic liquid as nonvolatile entrainer (Seiler et al., 2004).

ß Woodhead Publishing Limited, 2010

Separation and purification processes

255

More recently, Zhao et al. (2006) measured the isobaric vapor±liquid equilibrium (VLE) data for ethanol±water systems containing ILs 1-methyl-3methylimidazolium dimethylphosphate ([MMIM][DMP]), 1-ethyl-3methylimidazolium diethylphosphate ([EMIM][DEP]), 1-butyl-3methylimidazolium bromide ([BMIM][Br]), 1-butyl-3-methylimidazolium chloride ([BMIM][Cl]) and 1-butyl-3-methylimidazolium hexafluorophosphate ([BMIM][PF6]) at atmospheric pressure. They found that all ILs studied have significant salting-out effects, which lead to an increase in the relative volatility of ethanol. They also found that the salting-out effect followed the decreasing order of [BMIM][Cl] > [BMIM][Br] > [BMIM][PF6] and [MMIM][DMP] > [EMIM][DEP]. Wang et al. (2007) studied on the IL 1-ethyl-3-methylimidazolium dimethylphosphate ([EMIM][DMP]) as an entrainer for ethanol± water separation. Results showed that the relative volatility of ethanol is enhanced and the ethanol±water azeotrope is eliminated using [EMIM][DMP] as entrainer. Ge et al. (2008) made comparisons betweens eight ionic liquids composed of an anion from [BF4]±, [N(CN)2]±, [Cl]± or [OAc]±, and a cation from [BMIM]+ or [EMIM]+ and found that [EMIM][Cl] and [EMIM][OAc] are the best and the second best ILs, respectively, in enhancement of the relative volatility of ethanol. In brief, ILs are a promising green (environmentally friendly), effective, and potentially energy-saving entrainers. Extractive distillation with ILs can play a very important role in efficient separation and dehydration of ethanol.

10.3.6 Comparisons of different separating agents for extractive distillation The advantages and disadvantages of extractive distillation with different separating agents are compared in Table 10.1. Briefly, compared to extractive distillation with liquid solvent, dissolved salt or the mixture of liquid solvent and dissolved salt, extractive distillation with IL or hyper-branched polymers has excellent separation efficiency and selectivity without pollution of distillate by separating agents, thus requiring less energy consumption. In addition, the recent development of halogen-free and hydrolysis-stable IL such as [BMIM] [octylsulfate] (ECOENGTM 418) brings some promise (Seiler et al., 2004). Therefore, extractive distillation with IL and hyper-branched polymers represent two most promising novel separation technologies for bioalcohol.

10.4

Extractive fermentation

Extractive fermentation, or liquid-liquid extraction-fermentation hybrid, is the process in which in-situ extraction is conducted to remove the product ethanol and other inhibitory compounds, thus eliminating inhibitions caused by ethanol and other inhibitors and hence increasing the ethanol yield.

ß Woodhead Publishing Limited, 2010

256

Bioalcohol production

Table 10.1 Comparisons of different agents for extractive distillation for ethanol dehydration MSA

Advantages

Disadvantages

Liquid solvent

Less energy consumption than azeotropic distillation; Flexible selection of the possible solvents.

High energy consumption due to its high solvent/ feed mass ratio (=5±8)

Dissolved salt

High production capacity and low energy assumption due to its smaller solvent/feed ratio; Won't contaminate the overhead product due to its nonvolatility; Environmentally friendly and no health and safety hazards.

Potential problems in dissolution, transport and recycling of salt, jamming and erosion of equipment.

Mixture of liquid Easy operation (like liquid solvent and solvent) and high separation dissolved salt ability (like dissolved salt).

Less availability of suitable salts; Potential corrosion of salts to the equipment; Possible contamination of the overhead product by liquid extractants.

Room temperature ionic liquid (IL)

Considerable reduction of required heat duties due to their nonvolatility, high selectivities and capacities, especially a larger variety of feasible IL regeneration options; Only one distillation column required, representing low energy consumption; Won't pollute the distillate; IL's properties (solubility, capacity, selectivity, viscosity and thermal stability) can be tailored.

IL containing halogen anions is expensive and has insufficient stability for hydrolysis for long-term applications; Small amounts of corrosive and toxic substance (HF) forms during the hydrolysis.

Hyperbranched polymers

Excellent separation efficiency and selectivity; Won't contaminate the top product; High thermal stability; Available in large quantity at low cost; Low solution viscosity; No corrosion and toxicity; Their chemical and physical properties can be tailored.

Limited hyper-branched polymers are commercially available as entrainers for extractive distillation; Lack of property data.

ß Woodhead Publishing Limited, 2010

Separation and purification processes

257

The suitable solvents for extractive fermentation must be nontoxic to microorganisms (i.e. biocompatible), in addition to the criteria of solvent selection for the conventional extraction such as high distribution coefficient and selectivity, low solubility in the aqueous phase, density different from that of the broth to ensure phase separation by gravity, low viscosity, large interfacial tension and low tendency to emulsify in the broth, high stability, and cheap cost, etc. (Weilnhammer and Blass, 1994). The potential biocompatible solvents investigated for in-situ extraction of ethanol from beer liquor are oleyl alcohol (Weilnhammer and Blass, 1994; Moritz and Duff, 1996), n-dodecanol (Gyamerah and Glover, 1996; Boluda et al., 2005), isoamyl acetate and iso-octyl alcohol (Koullas et al., 1999), and nonanoic acid (Boudreau and Hill, 2006), etc. In the continuous fermentation (Fig. 10.8) for producing ethanol with the thermophilic, anaerobic bacterium Clostridium thermohydrosulfuricum, oleyl alcohol was used for in-situ extraction of ethanol to eliminate the ethanol product inhibition. It is found that the ethanol yield of the extractive fermentation is two times that of conventional fermentations without in-situ extraction (Weilnhammer and Blass, 1994). In the simultaneous saccharification and extractive fermentation (SSEF) process for converting cellulose hydrolysate to ethanol, oleyl alcohol was also applied to continuously extract the ethanol product out of the bioreactor. The SSEF process could increase the ethanol productivity by 65% and greatly reduce the amount of water consumption, compared to non-extractive batch fed simultaneous saccharification and fermentation (SSF) (Moritz and Duff, 1996). The combination of increasing ethanol yield and decreasing water consumption leads to significant reduction in overall ethanol production cost. In a pilot-scale extractive fermentation for producing ethanol with in-situ product removal by n-dodecanol extraction and with recycling of the extracted broth raffinate, the fresh water consumption reduced by 78% (Gyamerah and Glover, 1996). Isoamyl acetate, iso-octyl alcohol, n-butyl acetate, dibutyl ether

10.8 Continuous fermentation with in situ extraction (Weilhammer and Blass, 1994).

ß Woodhead Publishing Limited, 2010

258

Bioalcohol production

and dibutyl oxalate were also tested as potential agents for in-situ extraction of ethanol. Isoamyl acetate and iso-octyl alcohol were found to be very good solvents with ethanol distribution coefficients of above 1, and separation factors in Bancroft coordinates of the order of 70 and 2000, respectively. More recently, the ethanol-producing process of extractive fermentation with in-situ extraction of ethanol from fermentation broth using the fatty acids (valeric acid, oleic acid and nonanoic acid), followed by a flash process, was examined (Boudreau and Hill, 2006). The combined nonanoic acid extraction with the flash process consumed 38% less thermal energy than the conventional distillation process. Most recently, several -branched alcohols in the 14±20 carbons range were found to have improved extractive performance to recover ethanol from aqueous solutions compared to commonly studied solvents such as oleyl alcohol and 1-dodecanol (Offeman et al., 2008). Solvent toxicity to commercial yeast commonly used in bioethanol production was also evaluated for these high carbon alcohols and several lower molecular weight alcohols. It was found that amongst the alcohols studied, those having 12 or fewer carbons were toxic or inhibitory to the yeast, whereas those having 14 or more carbons were non-toxic and non-inhibitory. In brief, the use of extractive fermentation leads to increase in ethanol yield and decrease in fresh water consumption. Coupling extractive fermentation with flash separation can also bring about significant reduction in energy consumption. However, this process requires careful selection of biocompatible extracting agents.

10.5

Separation by adsorption

Two categories of adsorption technologies can be applied for the ethanol±water separation: the vapor-phase adsorption of water from the vapor mixture out of pre-concentration distillation column, and the liquid-phase adsorption of water from the fermentation broth (Beery and Ladisch, 2001a).

10.5.1 Vapor-phase adsorption of water The most potential adsorbents applied for vapor-phase adsorption of water from ethanol±water mixtures include inorganic adsorbents such as molecular sieves (Carmo and Gubulin, 1997; Al-Asheh et al., 2004), lithium chloride (Beery and Ladisch, 2001b), silica gel (Beery and Ladisch, 2001b), activated alumina (Kondo et al., 1997), and bio-based adsorbents such as corn grits (Crawshaw and Hills, 1990; Beery and Ladisch, 2001b). Inorganic adsorbents Zeolites molecular sieve (type 3A and 4A) are widely applied as absorbents in separating ethanol±water mixture (Carmo et al., 2004). As the nominal pore size

ß Woodhead Publishing Limited, 2010

Separation and purification processes

259

of 3A zeolite molecular sieves is 3 Angstroms (0.30 nm), and the approximate molecular diameters of water and ethanol molecules are 0.28 nm and 0.44 nm, respectively, water molecules can penetrate the pores of the molecular sieve adsorbent while ethanol is retained (Carmo and Gubulin, 1997). Inorganic adsorbents such as molecular sieves, silica gel, and lithium chloride have been successfully used as dehydration desiccants in fermentation ethanol plants (Beery and Ladisch, 2001b). Especially, the molecular sieve dehydration technologies based on adsorption have been commercially used in ethanol plants for many years. For instance, the molecular sieve dehydration unit (MSDU) developed by Delta-T Company has been widely applied in fuel-grade ethanol plants in North America since 1992. The MSDU process is effective in breaking the ethanol±water azeotrope to produce anhydrous fuel ethanol. Delta-T UltraDry MSDUs can even produce alcohol with less than 100 ppm moisture for industrial applications (Delta-T Company, 2009). So far, the molecular sieve dehydration process has been optimized for saving energy through process synthesis and heat integration. But it still consumes a large amount of energy since its desorption/regeneration stage is operated at a relatively high temperature. As a mature and effective technology, however, the molecular sieve dehydration process will continue to have commercial application until another novel separation technology with higher efficiency and lower capital and operating costs becomes mature and commercial. Bio-based adsorbents There are two categories of bio-based adsorbents for adsorption of water for ethanol dehydration: starch-based adsorbents such as cornmeal, starch, corn grits and other grains, and lignocellulosic adsorbents such as rice/wheat straw, sugar cane bagasse, corn cobs, and wood chips (Ladisch and Tsao, 1982; Rakshit et al., 1993; Chang et al., 2006a). Starchy biomass adsorbents can selectively adsorb water in the vapor mixture to obtain anhydrous ethanol (99.5 wt% ethanol) (Ladisch and Dyck, 1979). The most investigated starchy adsorbents are corn grits, corn meals and starch, which have different mean particle diameters and different relative amounts of amylose and amylopectin. The adsorption experiments with these adsorbents at 90 ëC showed that the water selectivity over ethanol increases with the amylopectin/amylose ratio in starches (Crawshaw and Hills, 1990). The fixed bed adsorber packed with starchy adsorbents is the commonly used equipment for adsorption of water vapor to break the ethanol±water azeotrope to obtain anhydrous ethanol (Ladisch et al., 1984; Hu and Xie, 2001; Beery and Ladisch, 2001a; Chang et al., 2006b). For instance, after the vapors of 92.4 wt% ethanol from distillation were passed over a fixed bed of corn grits, almost all the water is adsorbed on corn grits and anhydrous ethanol is obtained (Beery and Ladisch, 2001a). This approach had already been applied in many bio-ethanol plants

ß Woodhead Publishing Limited, 2010

260

Bioalcohol production

(Ladisch et al., 1984). The adsorption capacity of water is dependent on the vapor superficial velocity flowing through the fixed bed, the bed temperature, and the particle size distribution of adsorbents. The adsorption capacity of water for corn meals was up to 0.14±0.025 g water g/g adsorbent (Hu and Xie, 2001). The adsorption temperature for corn meals is in the range of 82±100 ëC (Chang et al., 2006b). The typical adsorption temperature for starch-based adsorbents is around 90 ëC (Crawshaw and Hills, 1990; Chang et al., 2006b). On the other hand, desorption (regeneration), at 105 ëC, often operates in a fluidized bed, rather than a general fixed-bed, in order to efficiently control the bed channel and faster the operation. Cellulosic biomass can also selectively adsorb water in the vapor mixture to obtain anhydrous ethanol (99.5 wt% ethanol) (Ladisch and Dyck, 1979). The lignocellulosic adsorbents such as rice straw, bagasse and microcrystalline cellulose powder were effective for adsorption of water in the vapor mixture with 80±90% ethanol to produce anhydrous ethanol (Rakshit et al., 1993). The adsorption of water on lignocellulosic materials mainly depends on the hydroxyl groups of the carbohydrates and the lignin (Berthold et al., 1996). Natural corncobs, natural and activated palm stone (date pits) and oak were also explored as adsorbents (Al-Asheh et al., 2004). In addition, bleached wood pulp, oak sawdust, and kenaf core have been studied for dehydration of the preconcentrated ethanol solutions containing 90, 95, and 97 wt% ethanol in a thermal swing adsorption column, and the results demonstrate that water can be selectively adsorbed resulting in anhydrous ethanol (Benson and George, 2005).

10.5.2 Liquid-phase adsorption of water Molecular sieve (A-type zeolites), silica gel, starch-based and cellulosic materials, etc., can also be utilized as adsorbents for adsorption of liquid water from ethanol aqueous solution. Zeolite A has a high capacity and selectivity for separation of water from ethanol (Ruthven, 1984). The different combinations of starch-based and cellulosic materials, including white corn grits, -amylasemodified yellow corn grits, polysaccharide-based synthesized adsorbent, and slightly gelled polysaccharide-based synthesized adsorbent also have potential uses in liquid-phase adsorption of water. With starch-based adsorbents, 1±20% of liquid water can be removed from ethanol. The adsorption capacity of water increases with the water content in the ethanol±water mixture. Compared with inorganic adsorbents such as silica gel and molecular sieves, the starch-based adsorbents have lower non-equilibrium adsorption capacity at low water concentration of less than 10 wt%, but they have the similar non-equilibrium adsorption capacity to that of the inorganic adsorbents at the concentrations of above 10 wt%. The starchy adsorbents adsorb water by forming hydrogen bonds between the hydroxyl groups on the surface of the adsorbent and the water molecules (Beery and Ladisch, 2001a). -amylase is often introduced in the

ß Woodhead Publishing Limited, 2010

Separation and purification processes

261

starch-based materials to modify the porosity and surface properties of starch so as to enhance the water sorption properties (Beery and Ladisch, 2001b). It is reported that corn grits are the only bio-based adsorbents having been successfully applied in industry (Beery and Ladisch, 2001b).

10.5.3 Advantages and disadvantages of adsorption The vapor phase adsorption consumes lower energy than distillation, due to the requirement of only one-time vaporization (Hu and Xie, 2001). Zeolite molecular sieves are highly selective. They have good mechanical strength and stable adsorption performance, but they require high temperatures and/or low pressures to regenerate because of its strong adsorption of water, compared with bio-based adsorbents (Crawshaw and Hills, 1990). In addition, molecular sieves are more expensive than bio-based adsorbents. Bio-based adsorbents have lower separation capacity and less stable than molecular sieves, but their regeneration temperature is much lower. In some cases of using bio-based adsorbents for removal of water, the saturated adsorbents can be used directly as feedstock, and simply fresh adsorbents are used without regeneration step, decreasing a large amount of energy consumption.

10.6

Membrane separation

Membrane separation is the process for separating liquid or gas mixtures by a semi-permeable membrane. Due to its potential low energy consumption and the elimination of the need of entrainers, membrane separation has gained wide interest for a few decades. For some large-scale industrial applications, the energy consumption of the membrane-based separation can be one order of magnitude lower than that of traditional thermal separations (Koros, 2004). Especially, membrane pervaporation (PV), one of the most commonly used membrane separations, has been considered as one of the most effective and energy-efficient process for breaking azeotropic or close-boiling mixtures. Till now, the PV technique has been applied for ethanol dehydration in more than 100 plants around the world (Dong et al., 2006). Based on the solution-diffusion mechanism, the driving force of pervaporation is the gradient of the chemical potential between the feed and the permeate sides of the membrane. In the vacuum PV process, azeotropic mixture contacts the membrane at the feed (or retentate) side, where the retained retentate leaves the unit. The permeate side is connected to a vacuum pump to lower the partial pressure of pervaporated permeate. Most membranes are hydrophilic or water permselective due to water's smaller molecular size, while few membranes are hydrophobic or ethanol permselective. In terms of materials for membrane production there are three categories of membranes: inorganic, polymeric, and mixed matrix or composite membrane.

ß Woodhead Publishing Limited, 2010

262

Bioalcohol production

10.6.1 Hydrophilic membranes for ethanol dehydration by pervaporation Hydrophilic polymeric membranes A large number of hydrophilic polymeric PV membranes for ethanol dehydration have been investigated, including naturally occurring polymers and derivatives such as chitosan and sodium alginate (NAALG), and synthetic polymers such as poly(vinyl alcohol) (PVA), polysulfone (PSF), polyimides (PI), and polyeletrolytes, etc. (Chapman et al., 2008). Having excellent film-forming ability and easy chemical modification, chitosan, NaAlg, PLA, and other polyelectrolytes, which have been used as wall materials for microencapsulation of drugs for controlled release (H-J Huang et al., 2006), are also good polymers for fabricating pervaporation membranes. However, membranes prepared by using only one of these polymers are not stable enough or lack mechanical strength and stability in aqueous solutions due to their water solubility or swelling in water, which could lower the selectivity of water to ethanol in spite of increasing the permeation flux. On the other hand, even though these membranes allow a high permeation flux due to the high hydrophilicity of the polymers, it leads to low selectivity of water over ethanol (Uragami et al., 2005). Therefore, the Chitosan, NaAlg, PVA, or other polyelectrolytes membranes are often modified by cross-linking with other chemicals, blending or copolymerization with other polymers, in order to improve water insolubility, mechanical strength, membrane structure, and hence the membrane separation performance aiming at high selectivity and sufficient permeation flux. So far, there has been significant research on the membranes based on Chitosan, NaAlg, PVA, and other polyelectrolytes for ethanol dehydration. Chapman et al. (2008) tabulated about 85 membranes of these types and their performances (separation factors and flux) for ethanol dehydration from literature. From these and some other references (Lee et al., 2000; Shao and Huang, 2007; Zhang et al., 2007), it can be seen that most membranes based on these polymers including their modifications by cross-linking or blending with other polymers have either low separation factors (w/e < 100) or low flux (<0.5) for pervaporation of water from 10 wt% (or below) ethanol solutions. For instance, the hollow-fiber PVA-NaAlg composite membrane, supported by a polysulfone hollow-fiber ultrafiltration membrane, has a high separation factor of 384, but low permeation flux of 0.384 kg m±2 h±1 for dehydrating the beer feed of 90 wt% ethanol at 45 ëC (Dong et al., 2006). The glutaraldehyde crosslinked chitosan (CS-GA) membrane modified by the reaction of its surface carboxylation with anhydride groups of maleic anhydride (MA) to the CS-GA-MA membrane increased the separation factor and the flux from 105 and 0.25 kg/m2 h of the unmodified CS-GA membrane to 634 and 0.30 kg m±2 h±1, respectively, for the separation of 90 wt% ethanol/water mixture at 60 ëC (Zhang et al., 2007). The membrane modification led to a few times higher separation factor, but its flux is still low.

ß Woodhead Publishing Limited, 2010

Separation and purification processes

263

Polysulfone and polyimides are also good polymers for membrane having excellent mechanical properties, good heat and chemical resistances. The PSF membranes could be modified by sulfonation or sodium substitution leading to increase in both separation factor and flux of the resulting sulfonated PSF membranes (Chen et al., 2001) and sodium sulfonate PSF membranes (Hung et al., 2003). Some of these PSF-based membranes have both high separation factors (>200) and high flux (>0.6 kg m±2 h±1) (see Table 10.3). Polyimides have received increasing interests as materials for pervaporation membrane for ethanol dehydration. Based on the recent reviews and other articles (Chapman et al., 2008; Jiang, 2009; Qiu et al., 2009; Li and Lee, 2006), most polyimide membranes have been reported to have relatively lower performance. Specifically, most polyimide membranes have either low separation factors (<50) or low permeation flux (<0.3 kg m±2 h±1) for dehydration of more than 90 wt% ethanol aqueous solutions. The homogeneous CS/CMCNa (sodium carboxymethyl cellulose, 80 wt%) polyelectrolyte complex membrane has a good performance of J = 1.14 kg m±2 h±1 and w/e = 1062 in dehydrating 10 wt% water-ethanol at 70ëC (Zhao et al., 2009). It is worth to note that most membranes were tested with more than 5 wt% water in feed. However, to obtain ethanol product of more than 95 wt% purity, it is necessary to test the membrane performances at low water concentration (< 5%). Till now, there has been limited published literature in this area (Namboodiri and Vane, 2007) and more attention is required in the future to explore the pervaporation performance at low water content for membranes showing high selectivity and flux at more than 5 wt% water in feed. From the engineering point of view, it is ideal for a hydrophilic PV membrane to have both selectivity higher than 1000 for water over ethanol, and the overall permeation flux exceeding 1.00 kg m±2 h±1 (Shi et al., 1998). Unfortunately, however, most of the polymer-based membranes reported have either low flux or low selectivity. Table 10.2 lists some hydrophilic polymeric membranes with relatively good performances for ethanol dehydration. Hydrophilic inorganic membranes Hydrophilic inorganic membranes mainly include silica (van Veen et al., 2001; Sekulic et al., 2005; Sommer and Melin, 2005), zeolite A (Kondo et al., 1997; Okamoto et al., 2001; Sato and Nakane, 2007), zeolite X and zeolite Y (Kita et al., 2001), and mordenite (Navajas et al., 2007). Recently, Bowen and his colleagues (Bowen et al., 2004) and Caro and Noack (Caro and Noack, 2008) have made reviews on zeolite membranes. Inorganic pervaporation membranes have many advantages including their high separation factor and permeation flux (Shah et al., 2000), and excellent temperature stability and mechanical strength (Sommer and Melin, 2000). Tubular zeolite and silica membranes, for instances, are still stable at above 300 ëC and above 100 bar of the feed pressures. The

ß Woodhead Publishing Limited, 2010

264

Bioalcohol production

Table 10.2 PV performance of hydrophilic polymeric membranes for ethanol dehydration Membrane*

Flux Reference (kg mÿ2 hÿ1)

wt% of water in feed

T (ëC)

w/e

PAAHCl/PVA99

10

70

1360

0.9

PAAHCl/PVA99PVA88

10 2

70 60

2750 130

0.942 0.167

3 5

49.5 60

135 1000

0.15 2.00

5

80

350

0.65

Namboodiri and Vane (2007) Shi et al. (1998) Namboodiri and Vane (2007) Yeom et al. (2001) Shieh and Huang (1997) Lee et al. (1997)

10

60

1791

0.472

Zhang et al. (2007)

10

60

5469

0.424

10 10 10 5

25 45 40 60

600 1300 240 900

0.7±0.9 0.88 1.7 1.0

5

60

3500

1.63

Jiraratananon et al. (2002) Chen et al. (2001) Hung et al. (2003) Kim et al. (2000) Yanagishita et al. (1994) Karakane et al. (1991)

PVA-GA Chitosan/PAA composite H2SO4 cross-linked chitosan H2SO4 cross-linked chitosan Chitosan/HEC PSF Sodium sulfonate PSF PI/PSF composite PI-2080 aromatic polyimide PAA/polyion complex

* w/e = the separation factor of water over ethanol, PAAHCl = poly(allylamine hydrochloride), PVA99 = 99% hydrolyzed PVA, PVA88 = 88% hydrolyzed PVA, GA = glutaraldehyde, PI = polyimide, PAA = poly(acrylic acid), HEC = hydroxyethylcellulose, PSF = polysulfone. PAAHCl±PVA99: 60 wt% PAAHCl, 35 wt% PVA99, 5 wt% GA PAAHCl/PVA99±PVA88: 23.75 wt% PVA99, 23.75 wt% PVA88, 47.5 wt% PAAHCl, 5 wt% GA

tabular Zeolite NaA membrane module, having the pervaporation flux of ca. 2.35 kg m±2 h±1 and the separation factor of above 5000 for the solution of 95 wt% ethanol at 95 ëC, can be available at a low price (Kondo et al., 1997). The first commercial large-scale PV plant consisting of 16 membrane modules each containing 125 NaA zeolite membrane tubes with a total membrane area of about 5.8 m2, could produce 530 L/h of more than 99.8 wt% ethanol from 90 wt% ethanol at 120 ëC (Morigami et al., 2001). Table 10.3 lists the typical PV separation factors and the permeation flux for the hydrophilic zeolite membranes. The NaA zeolite membrane showed the highest water-selective permeation and the highest permeation flux in pervaporation (PV) toward water/ethanol liquid mixtures. Compared to polymeric membranes, the zeolite NaA membranes have advantages of high selectivity and high water permeation flux in entire feed range, non-swelling in water, and higher performance at higher temperature (stable at least up to 200 ëC) (Okamoto et al., 2001).

ß Woodhead Publishing Limited, 2010

Separation and purification processes

265

Table 10.3 PV performance of hydrophilic inorganic membranes for ethanol dehydration Type

Zeolite NaA Zeolite NaA Zeolite NaA Zeolite NaA Zeolite NaA Zeolite A Zeolite T Zeolite T NaX NaY Sillica Sillica Sillica Mordenite

wt% of water in feed 2 5 5 10 10 10.3 10.1 10 10 10 3.6 10 11 10

T (ëC)

w/e

40 95 75 75 75 70 70 75 75 75 70 80 70 150

>1,000 5,000 16,000 >5,000 10,000 18,000 1,000 900 170 130 350 800 160 139

Flux Reference (kg mÿ2 hÿ1) 1.0 2.35 1.10 5.60 2.15 1.12 0.91 1.10 1.91 1.59 1.6 1.00 2.0 0.16

Ahn et al. (2006) Kondo et al. (1997) Okamoto et al. (2001) Sato and Nakane (2007) Okamoto et al. (2001) Sommer and Melin (2005) Sommer and Melin (2005) Cui et al. (2004) Kita et al. (2001) Kita et al. (2001) van Veen et al. (2001) Sekulic et al. (2005) Sommer and Melin (2005) Navajas et al. (2007)

Where w/e is the separation factor of water over ethanol.

Hydrophilic mixed matrix or composite membranes To achieve a high ratio of membrane performance/cost, various inorganicpolymer mixed matrix or composite membranes such as polystyrenesulfonatealumina (Martin et al., 1995) and KA zeolite-incorporated cross-linked PVA multilayer mixed matrix membranes (MMMMs) (Z Huang et al., 2006) have recently been studied for pervaporation separation of ethanol±water mixtures. The separation factor of polystyrenesulfonate-alumina composite membranes was shown to be 400 (Martin et al., 1995). The separation performance of the MMMMs is better than that of multi-ply homogenous membranes (MHM) containing no zeolite. In addition, a series of three-layer zeolite-embedded PVA composite membranes have been successfully fabricated with loading of 20 wt% of different zeolites, including 3A, 4A, 5A, NaX, NaY, and silicalite, etc. The addition of zeolite resulted in decreasing the activation energies for water and ethanol, and hence increasing the separation selectivity (Z Huang et al., 2006). Compared to the pure polymeric membranes, the mixed matrix membranes can substantially increase permeability and selectivity (Caro and Noack, 2008). Nowadays, however, the separation factors and flux of the mixed matrix membranes for ethanol dehydration are still very low. For example, the separation performances of all the 25 mixed matrix membranes from literature mentioned in the review article (Chapman et al., 2008) are still poor, with all flux less than 0.55 kg m±2 h±1 and most below 0.30 kg m±2 h±1. Most recently, some researches investigated the novel nanocomposite pervaporation membranes, e.g. chitosan/

ß Woodhead Publishing Limited, 2010

266

Bioalcohol production

TiO2 nanocomposite pervaporation membranes for ethanol dehydration (Yang et al., 2009). Compared with pure CS membrane and CS/TiO2 blending membranes, CS/TiO2 nanocomposite membranes exhibited better pervaporation performance for ethanol dehydration.

10.6.2 Hydrophobic pervaporation membranes for ethanol removal from water The potential hydrophobic membranes for removal of ethanol from aqueous solution include the hydrophobic polymeric membranes poly(dimethyl siloxane) (PDMS) (Slater et al., 1990; Takegami et al., 1992; Chen et al., 1998) and Poly(1-trimethylsilyl-1-propyne) (PTMSP) (Nagase et al., 1991; GonzaÂlezVelasco et al., 2002), the hydrophobic zeolite membranes (Sano et al., 1994; Ikegami et al., 1997, 1999), and the composite membranes such as silicalitePDMS membranes (Vankelecom et al., 1995, 1997; Moermans et al., 2000). The ethanol±water separation factors of PDMS, PTMSP, composite membranes, and zeolite are reported to be in the range of 4.4±10.8, 9±26, 7±59, 12±106, respectively (Vane, 2005). However, the separation factors in some other cases might be outside these ranges. For example, the separation factor of ethanol over water was 218, when using a silicate zeolite membrane where ethanol (98.2 wt%) at permeate was continuously obtained from the fermentation broth of 20 wt% ethanol (Nomura et al., 2002). In general, the ethanol±water separation factors are largely ranked in the following increasing order: PDMS < PTMSP < composite membranes < zeolite membranes. A few of hydrophobic pervaporation membranes with relatively good performances for removal of ethanol from water are tabulated in Table 10.4. The commercially available PDMS membrane module was used in a continuous fermentation/membrane pervaporation system to concentrate ethanol, resulting in permeate of 20±23 wt% while 4±6 wt% level was retained (O'Brien and Craig, 1996). The polydimethylsiloxane-polystyrene interpenetrating polymer network (PDMS-PS IPN) supported membranes used for separation of ethanol from aqueous solutions has superior mechanical and film-forming properties to those of PDMS because of the addition of PS, which is more hydrophobic and of higher tensile strength than PDMS. The selectivity of these PDMS-PS membranes varied with the feed composition. Specifically, the membrane was more selective for ethanol for the feed of low ethanol percent, while for the feed of high ethanol percent it was more selective for water (Ruckenstein and Liang, 1996). Currently, hydrophobic zeolite membranes are already commercially available, while polymeric and composite membranes are still under study. In addition, zeolite membranes are more expensive than polymer membranes, but the former has higher separation factors and flux than the latter. Therefore, zeolite membranes may be more cost effective on per unit ethanol basis (Vane, 2005).

ß Woodhead Publishing Limited, 2010

Separation and purification processes

267

Table 10.4 PV performance of typical hydrophobic membranes for ethanol removal from water Membrane type*

wt% of water in feed

T Flux (ëC) (kg mÿ2 hÿ1)

e/w

Reference

Silicalite-1

4.65

30

0.6

62

Silicalite-1 Silicalite-1

5 5

30 30

0.76 0.29

58 120

Silicalite-1 Silicalite-1 Silicalite-1

5 5 10

60 60 30

1.8 1.13 0.56

89 93 100

Calcined silicalite-1 (MFI) Ozonicated silicalite-1 (MFI) Ge-ZSM-5 (Si/Ge=41)

16

75

2.0

38

Nomura et al. (1998) Sano et al. (1994) Matsuda et al. (1998) Lin et al. (2001) Lin et al. (2003) Nomura et al. (2001) Kuhn et al. (2009)

16

75

1.1

42

Kuhn et al. (2009)

5

30

0.22

47

5

60

0.16

70 6

22 65

0.51 0.33

10

40

0.56

10

87

0.66

10

25

0.91

10

33

1.2

6

30

0.50

B-ZSM-5 PDMS/silicalite-1 Nano-sized silicalite-1 (40%) in PDMS PDMS/MDMS (93% PDMS) BTDA/PDMS (66 wt%) with ODA PDMS/PVDF multi-layer SolSep 3360 (nanofiltration membrane) PTMSP

Bowen et al. (2002) 31 Bowen et al. (2003) 43.6 Jia et al. (1992) 14.6 Moermans et al. (2000) 10.6 Krea et al. (2004) 2.7

Lai et al. (1994)

31

Chang and Chang (2004) 7.0 Verhoef et al. (2008) 16.5 Volkov et al. (2004)

*Where e/w = the separation factor of ethanol over water, MDMS = 1,3-bis(3-aminopropyl) tetramethyldisolxane, BTDA = 3,30,4,40 -benzophenonetetracarboxylic dianhydride, ODA = 4,40 oxydianiline, PVDF = polyvinylidene fluoride.

10.6.3 Membrane pervaporation-bioreactor hybrid The fermentation broth in an ethanol-producing bioreactor often contains inhibitors including ethanol product, furfural, phenolics, and other chemicals, which lowers the ethanol yield. Therefore, the product ethanol and the other inhibitory compounds must be simultaneously removed from the fermentation broth. This can be done by combining fermentation with hydrophobic membrane pervaporation for in-situ removal of the inhibitors. The process can be operated

ß Woodhead Publishing Limited, 2010

268

Bioalcohol production

10.9 A complete membrane pervaporation-bioreactor hybrid process for ethanol fermentation and dehydration (Vane, 2005).

continuously and the recovered organic VOCs (ethanol, acetone, butanol, 2propanol) can be reused within other processes. To avoid fouling of the hydrophobic membrane, a microfiltration membrane is usually added before pervaporation. In addition, the ethanol-enriched solution, i.e. the permeate of the hydrophobic membrane, can be further dehydrated by a hydrophilic membrane to produce anhydrous ethanol. The complete process diagram is illustrated in Fig. 10.9. In general, the hydrophobic membranes described above can be applied in the bioreactor-membrane pervaporation hybrid system. However, more tests are needed, including examine of the by-products (e.g., acetic acid, succinic acid and glycerol) on the pervaporation performance. The fermentation coupled with ethanol removal by membrane pervaporation has been studied by some researchers (O'Brien and Craig, 1996; Lipnizki et al., 2000; Ikegami et al., 2002). The PDMS-PAN-PV membrane utilized in the separation of ethanol from the fermented mash by pervaporation showed the high flux 2.6±3.5 kg/m±2 h±1 at 3.0±6.2 wt% ethanol in feed and selectivity of more than 8 (Lewandowska and Kujawski, 2007). But the obtained permeated ethanol concentration of 10±21 wt% ethanol was still very low, meaning that selectivity of ethanol to water is not high enough. Hydrophobic zeolite membranes, specifically silicalite, could enrich ethanol solution up to 85% (v/v) from 10% (v/v) fermented ethanol (Ikegami et al., 1997). But, due to the effect of by-products such as succinic acid and glycerol, the permeate flux and the selectivity decreased significantly with elapse of fermentation time. To solve this problem, silicon rubber (KE-45)coated silicalite membrane was prepared for reliable production of highly concentrated bioethanol (Ikegami et al., 2003).

10.6.4 Vacuum membrane distillation (VMD) ± bioreactor hybrid Vacuum membrane distillation (VMD) is an appealing process for separation of aqueous mixtures. It is quite similar to pervaporation, with the only difference

ß Woodhead Publishing Limited, 2010

Separation and purification processes

269

being that the separation factor of VMD is not affected by the membrane used, but by the vapor-liquid equilibrium of the feed solution (Gostoli and Sarti, 1989). As membrane pervaporation-bioreactor hybrid, the VMD-bioreactor hybrid process is also suitable for separation of ethanol and the other inhibitory compounds from fermentation broths (Bausa and Marquardt, 2000; Gryta et al., 2000; Izquierdo-Gil and Jonsson, 2003). For instance, ethanol was produced in a membrane distillation bioreactor where volatile compounds (ethanol and other inhibitors) were separated from the fermentation broth by porous capillary polypropylene membranes, leading to increase in the productivity and the sugarto-ethanol conversion rate (Gryta et al., 2000). VMD is commercially competitive because of its high selectivity of ethanol over water, large flux, high thermal efficiency and low energy cost (Lei et al., 2005).

10.7

Conclusions

This chapter attempts to provide a critical review of the various separation technologies that can be used in future bioalcohol and integrated lignocellulosic biorefinery producing liquid fuels and other co-products. As described above, there are two key separation steps in the biorefinery that offer challenges and opportunities. First is the separation of fermentation inhibitors. The promising separation technologies for removal of inhibitors are the three in-situ detoxification hybrid processes including extractive fermentation, membrane pervaporation-bioreactor, and VMD-bioreactor, which can eliminate the inhibition of products and inhibitory compounds, increase the fermentation yield and productivity, and reduce (fresh) water consumption due to recycle. The second key separation challenge in biorefinery is the azeotropic nature of ethanol±water mixture posing challenges to remove the last amounts of water producing fuel grade ethanol. The promising technologies for breaking the ethanol±water azeotrope to obtain anhydrous ethanol, are the extractive distillation with ionic liquid and hyper-branched polymers, adsorption with molecular sieve and bio-based adsorbents, representing low energy consumption.

10.8

References

Aden A, Ruth M, Ibsen K, Jechura J, Neeves K, Sheehan J, Wallace B, Montague L, Slayton A, and Lukas J (2002), `Lignocellulosic biomass to ethanol process design and economics utilizing co-current dilute acid prehydrolysis and enzymatic hydrolysis for corn stover', NREL report TP-510-32438. Available at http:// www.nrel.gov/docs/fy02osti/32438.pdf. Ahn H, Lee H, Lee S B, and Lee Y (2006), `Pervaporation of an aqueous ethanol solution through hydrophilic zeolite membranes', Desalination, 193, 244±251. doi: 10.1016/ j.desal.2005.06.062. Al-Amer Am (2000), `Investigating polymeric entrainers for azeotropic distillation of the

ß Woodhead Publishing Limited, 2010

270

Bioalcohol production

ethanol/water and MTBE/methanol systems', Ind Eng Chem Res, 39(10), 3901± 3906. doi: 10.1021/ie0000515. Al-Asheh S, Ganat F, and Al-Lagtah N (2004), `Separation of ethanol±water mixtures using molecular sieves and biobased adsorbents', Chem Eng Res Design, 82(A7), 855±864. doi: 10.1205/0263876041596779. Amidon T (2006), `The Biorefinery in New York: Woody biomass into commercial ethanol', Pulp Pap Mag Can, 107(6), 47. Arlt M, Seiler M, Jork C, Schneider T (2001), DE Patent No. 10114734. Barba D, Brandani V, and Giacomo G (1985), `Hyperazeotropic ethanol salted-out by extractive distillation, Theoretical evaluation and experimental check', Chem Eng Sci, 40, 2287±2292. Bausa J and Marquardt W (2000), `Shortcut design methods for hybrid membranedistillation processes for the separation of nonideal multicomponent mixtures', Ind Eng Chem Res, 39, 1658±1672. doi: 10.1021/ie990703t. Beery K E and Ladisch M R (2001a), `Chemistry and properties of starch based desiccants', Enzyme Microb Technol, 28, 573±581. doi: 10.1016/S01410229(00)00345-8. Beery K E and Ladisch M R (2001b), `Adsorption of water from liquid-phase ethanol± water mixtures at room temperature using starch-based adsorbents', Ind Eng Chem Res, 40, 2112±2115. doi: 10.1021/ie0009381. Benson T J and George C E (2005), `Cellulose based adsorbent materials for the dehydration of ethanol using thermal swing adsorption', Chem Mat Sci, 11(1), 697± 701. 10.1007/s10450-005-6009-1. Berthold J, Rinaudo M, and Salmen L (1996), `Association of water to polar groups; estimations by an adsorption model for ligno-cellulosic materials', Colloids and Surfaces A: Physicochemical and Engineering Aspects, 112(2), 117±129. doi: 10.1016/0927-7757(95)03419-6. Boluda N, Gomis V, Ruiz F, and Bailador H (2005), `The influence of temperature on the liquid± liquid±solid equilibria of the ternary system water + ethanol + 1-dodecanol', Fluid Phase Equilibria, 235, 99±103. doi: 10.1016/j.fluid.2005.06.027. Boudreau T M and Hill G A (2006), `Improved ethanol±water separation using fatty acids', Process Biochem, 41, 980±983. doi: 10.1016/j.procbio.2005.11.006. Bowen T C, Li S, Tuan V A, Falconer J L, and Noble R D (2002), `Pervaporation of aqueous organic mixtures through Ge-ZSM-5 zeolite membranes', Desalination, 147, 327±329. doi: 10.1016/S0011-9164(02)00560-X. Bowen T C, Li S, Kalipcilar H, Falconer J L, and Noble R D (2003), `Pervaporation of organic/water mixtures through B-ZSM-5 zeolite membranes on monolith supports', J Membr Sci, 215, 235±247. doi: 10.1016/S0376-7388(02)00617-8 Bowen T C, Noble R D, and Falconer J L (2004), `Fundamentals and applications of pervaporation through zeolite membranes', J Membr Sci, 245, 1±33. doi: 10.1016/ j.memsci.2004.06. Carmo M J and Gubulin J C (1997), `Ethanol±water adsorption on commercial 3a zeolites: kinetic and thermodynamic data', Brazilian J Chem Eng, 14(3) 1±10. doi: 10.1590/S0104-66321997000300004. Carmo M J, Adeodato M G, Moreira A M, Parente Jr. E J S, and Vieira R S (2004), `Kinetic and Thermodynamic study on the liquid phase adsorption by starchy materials in the alcohol-water system', Adsorption, 10(3), 211±218. doi: 10.1023/ B:ADSO.0000046357.86218.16. Caro J and Noack M (2008), `Zeolite membranes ± recent developments and progress', Microporous and Mesoporous Materials, 115, 215±233. doi: 10.1016/

ß Woodhead Publishing Limited, 2010

Separation and purification processes

271

j.micromeso.2008.03.008. Chang C L and Chang M S (2004), 'Preparation of multi-layer silicone/PVDF composite membranes for pervaporation of ethanol aqueous solutions', J Membr Sci, 238(1± 2), 117±122. doi: 10.1016/j.memsci.2004.03.026. Chang H, Yuan X-G, Tian H, and Zeng A-W (2006a), `Experimental investigation and modeling of adsorption of water and ethanol on cornmeal in an ethanol±water binary vapor system', Chem Eng Technol, 29(4), 454±461. doi: 10.1002/ ceat.200500384. Chang H, Yuan X-G, Tian H, and Zeng A-W (2006b), `Experimental study on the adsorption of water and ethanol by cornmeal for ethanol dehydration', Ind Eng Chem Res, 45, 3916±3921. doi: 10.1021/ie051043x. Chapman P D, Oliveira T, Livingston A G, Li K (2008), 'Membranes for the dehydration of solvents by pervaporation', J Membr Sci, 318, 5±37. doi: 10.1016/ j.memsci.2008.02.061. Chen S H, Yu K C, Lin S S, Chang D J, and Liou R M (2001), `Pervaporation separation of water/ethanol mixture by sulfonated polysulfone membrane', J Membr Sci, 183, 29. doi: 10.1016/S0376-7388(00)00544-5. Chen X, Ping Z H, and Long Y C (1998), `Separation properties of alcohol-water mixture through silicalite-I-filled silicone rubber membranes by pervaporation', J Appl Polym Sci, 67, 629±636. doi: 10.1002/(SICI)1097-4628(19980124)67. Chianese A and Zinnamosca F (1990), `Ethanol dehydration by azeotropic distillation with a mixed-solvent entrainer', Chem Eng J, 43, 59±65. Cook R A and Furter W F (1968), `Extractive distillation employing a dissolved salt as separating agent', Can J Chem Eng, 46, 119±123. Crawshaw J P and Hills J H (1990), `Sorption of ethanol and water by starchy materials', Ind Eng Chem Res, 29, 307±309. doi: 10.1021/ie00098a023. Cui Y, Kita H, and Okamoto K I (2004), `Zeolite T membrane: preparation, characterization, pervaporation of water/organic liquid mixtures and acid stability', J Membr Sci, 236, 17±27. doi: 10.1016/j.memsci.2003.12.018. Delta-T Company (2009), http://www.deltatcorp.com/molsieve.aspx. Dong Y Q, Zhang L, Shen J N, Song M Y, and Chen H L (2006), `Preparation of poly(vinyl alcohol)-sodium alginate hollow-fiber composite membranes and pervaporation dehydration characterization of aqueous alcohol mixtures', Desalination, 193, 202±210. doi: 10.1016/j.desal.2005.08.023. Duan Z T, Lei L H, and Zhou R Q (1980), `Study on extractive distillation with salt(I)', Petrochem Technol (China), 9, 350±353. Feng G, Fan L T, and Friedler F (2000), `Synthesizing alternative sequences via a Pgraph-based approach in azeotropic distillation systems', Waste Management, 20, 639±643. doi: 10.1016/S0956-053X(00)00051-9. Furter W F (1992), `Extractive distillation by salt effect', Chem Eng Commun, 116, 35± 40. doi: 10.1080/00986449208936042. Furter W F (1972), `Extractive distillation by salt effect', in: Gould R F (eds), Extractive and Azeotropic Distillation, Advances in Chemistry Series, 115, 35±45. Ge Y, Zhang L, Yuan X, Geng W, and Ji J (2008), `Selection of ionic liquids as entrainers for separation of (water + ethanol)', J Chem Thermodynamics, 40, 1248±1252. doi: 10.1016/j.jct.2008.03.016. Gomis V, Font A, Pedraza R, and Saquete M D (2005), `Isobaric vapor±liquid and vapor± liquid±liquid equilibrium data for the system water + ethanol + cyclohexane', Fluid Phase Equilibria, 235, 7±10. doi: 10.1016/j.fluid.2005.07.015. GonzaÂlez-Velasco J R, Gonzalez-Marcos J A, and Lopez-Dehesa C (2002),

ß Woodhead Publishing Limited, 2010

272

Bioalcohol production

`Pervaporation of ethanol±water mixtures through poly(1-trimethylsilyl-1-propyne) (PTMSP) membranes', Desalination, 149, 61±65. doi: 10.1016/S00119164(02)00692-6. Gostoli C and Sarti G C (1989), `Separation of liquid mixtures by membrane distillation', J Membr Sci, 41 211±224. Gryta M, Morawski A W, and Tomaszewska M (2000), `Ethanol Production in membrane distillation bioreactor', Catalysis Today, 56, 159±165. doi: 10.1016/S09205861(99)00272-2. Gyamerah M and Glover J (1996), `Production of ethanol by continuous fermentation and liquid±liquid extraction', J Chem Tech Biotechnol, 66, 145±52. doi: 10.1002/ (SICI)1097-4660(199606)66. Hu X and Xie W (2001), `Fixed-Bed adsorption and fluidized-bed regeneration for breaking the azeotrope of ethanol and water', Sep Sci Technol, 36 (1), 125±136. Huang H-J, Chen X D, and Yuan W K (2006), `Microencapsulation based on emulsification for producing pharmaceutical products: a literature review', Dev Chem Eng Mineral Process, 14, 515±544. doi: 10.1002/apj.5500140318. Huang H-J, Ramaswamy S, Tschirner U W, and Ramarao B V (2008), `A review of separation technologies in current and future biorefineries', Sep Purif Technol, 62(1), 1±21. doi: 10.1016/j.seppur.2007.12.011. Huang H-J, Ramaswamy S, Al-Dajani W, Tschirnera U, and Cairncross R A (2009a), `Effect of biomass species and plant size on cellulosic ethanol: A comparative process and economic analysis', Biomass Bioenergy, 33(2), 234±246. doi: 10.1016/ j.biombioe.2008.05.007. Huang H-J, Lin W, Ramaswamy S, Tschirner U (2009b), `Process modeling of comprehensive integrated forest biorefinery ± an integrated approach', Appl Biochem Biotechnol, 154(1±3), 26±37. doi: 10.1007/s12010-008-8478-7. Huang H-J, Ramaswamy S, Al-Dajani W, Tschirner U (2010), `Process modeling and analysis of pulp mill-based integrated biorefinery with hemicellulose pre-extraction for ethanol production: a comparative study', Biores Technol, 101(2), 624±631. doi: 10.1016/j.biortech.2009.07.092. Huang Z, Guan H-M, Tan W L, Qiao X-Y, and Kulprathipanja S (2006), `Pervaporation study of aqueous ethanol solution through zeolite-incorporated multilayer poly(vinyl alcohol) membranes: Effect of zeolites', J Membr Sci, 276, 260±271. doi: 10.1016/j.memsci.2005.09.056. Hung M Y, Chen S H, Liou R M, Hsu C S, and Lai J Y (2003), `Pervaporation separation of a water/ethanol mixture by a sodium sulfonate polysulfone membrane', J Appl Polym Sci, 90, 3374±2283. doi: 10.1002/app.12947. Ikegami T, Yanagishita H, Kitamoto D, Haraya K, Nakane T, Matsuda H, Koura T, and Sano N (1997), `Production of highly concentrated ethanol in a coupled fermentation/pervaporation process using silicalite membranes', Biotechnol Tech, 11, 921±924. 10.1023/A:1018474603027. Ikegami T, Yanagishita H, Kitamoto D, Haraya K, Nakane T, Matsuda H, Koura N, and Sano T (1999), `Highly concentrated aqueous ethanol solutions by pervaporation using silicalite membrane ± improvement of ethanol selectivity by addition of sugars to ethanol solution', Biotechnol Lett, 21, 1037±1041. 10.1023/ A:1005627421221. Ikegami T, Yanagishita H, Kitamotoa T, Negishi H, Haraya K, and Sano T (2002), `Concentration of fermented ethanol by pervaporation using silicalite membranes coated with silicone rubber', Desalination, 149, 49±54. doi: 10.1016/S00119164(02)00690-2.

ß Woodhead Publishing Limited, 2010

Separation and purification processes

273

Ikegami T, Kitamoto D, Negishi H, Haraya K, Matsuda H, Nitanai Y, Koura N, Sano T, and Yanagish H (2003), `Drastic improvement of bioethanol recovery using a pervaporation separation technique employing a silicone rubber-coated silicalite membrane', J Chem Technol Biotechnol, 78:1006±1010. doi: 10.1002/jctb.897. Izquierdo-Gil M A and Jonsson G (2003), `Factors affecting flux and ethanol separation performance in vacuum membrane distillation (VMD)', J Membr Sci, 214, 113± 130. doi: 10.1016/S0376-7388(02)00540-9. Jia M D, Peinemann K V, and Behling R D (1992), `Preparation and characterization of thin-film zeolite-PDMS composite membranes', J Membr Sci, 73, 119. Jiang L Y, Wang Y, Chung T-S, Qiao X Y, and Lai J-Y (2009), `Polyimides membranes for pervaporation and biofuels separation', Prog Polym Sci, doi: 10.1016/ j.progpolymsci.2009.06.001 Jiraratananon R, Chanachai A, Huang R Y M, and Uttapap D (2002), `Pervaporation dehydration of ethanol±water mixtures with chitosan/hydroxyethylcellulose (CS/ HEC) composite membranes. 1. Effect of operating conditions', J Membr Sci, 195, 143. doi: 10.1016/S0376-7388(01)00563-4. Karakane H, Tsuyumoto M, Maeda Y, and Honda Z (1991), `Separation of WaterEthanol by Pervaporation through Polyion Complex Composite Membrane', J Appl Polym Sci, 42, 3229. Kim J H, Lee K H, and Kim S Y (2000), `Pervaporation separation of water from ethanol through polyimide composite membranes', J Membr Sci, 169, 81±93. doi: 10.1016/ S0376-7388(99)00335-X. Kita H, Fuchida K, Horita T, Asamura H, and Okamoto K I (2001), `Preparation of Faujasite membranes and their permeation properties', Sep Purif Technol, 25, 261± 268. Kondo M, Komori M, Kita H, and Okamoto K (1997), `Tubular type pervaporation module with zeolite NaA membrane', J Membr Sci, 133, 133±141. doi: 10.1016/ S0376-7388(97)00087-2. Koros W J (2004), `Evolving beyond the thermal age of separation processes: Membranes can lead the way', AIChE J, 50, 2326±2334. doi: 10.1002/aic.10330. Koullas D P, Umealu O S, and Koukios E G (1999), `Solvent selection for the extraction of ethanol from aqueous solutions', Sep Sci Technol, 34(11), 2153±2163. 10.1021/ bp00008a006. Kovach III J W and Seider W D (1987), `Heterogeneous azeotropic distillaion-homotopycontinuation methods', Comput Them Engng, 11(6), 593±605. Krea M, Roizard D, Moulai-Mostefa N, and Sacco D (2004), `New copolyimide membranes with high siloxane content designed to remove polar organics from water by pervaporation', J Membr Sci, 241, 55±64. doi: 10.1016/j.memsci.2004.03.040. Kuhn J, Sutanto S, Gascon J, Gross J, and Kapteijn F (2009), `Performance and stability of multi-channel MFI zeolite membranes detemplated by calcination and ozonication in ethanol/water pervaporation', J Membr Sci, 339, 261±274. doi: 10.1016/j.memsci.2009.05.006. Ladisch M R and Dyck K (1979), `Dehydration of ethanol: new approach gives positive energy balance', Science, 205, 898±900. doi: 10.1126/science.205.4409.898. Ladisch M R and Tsao G T (1982), `Vapor phase dehydration of aqueous alcohol mixtures', US Patent 4 345 973. Ladisch M R, Voloch M, Hong J, Bienkowski P, and Tsao G T (1984), `Cornmeal adsorber for dehydrating ethanol vapors', Ind Eng Chem Process Des Dev, 23, 437± 443. doi: 10.1021/i200026a005. Lai J Y, Li S H, and Lee K R (1994), `Permselectivities of polysiloxaneimide membrane

ß Woodhead Publishing Limited, 2010

274

Bioalcohol production

for aqueous ethanol mixture in pervaporation', J Membr Sci, 93, 273±82. Lee F-M and Pahl R H (1985), `Solvent screening study and conceptual extractive distillation process to produce anhydrous ethanol from fermentation broth', Ind Eng Chem Process Des Dev, 24, 168±172. doi: 10.1021/i200028a029. Lee K R, Teng M Y, Lee H H, and Lai J Y (2000), 'Dehydration of ethanol/water mixtures by pervaporation with composite membranes of polyacrylic acid and plasma-treated polycarbonate', J Membr Sci, 164, 13±23. doi: 10.1016/S0376-7388(99)00097-6. Lee Y M, Nam S Y, and Woo D J (1997), `Pervaporation of ionically surface crosslinked chitosan composite membranes for water±alcohol mixtures', J Membr Sci, 133, 103±110. doi: 10.1016/S0376-7388(97)00089-6. Lei Z, Chen B, and Ding Z (2005), Special Distillation Processes, Amsterdam, Elsevier. Lei L H, Duan Z T, and Xu Y F (1982), `Study on extractive distillation with salt(II)', Petrochem Technol (China), 11, 404±409. Lei Z, Wang H, Zhou R, and Duan Z (2002), `Influence of salt added to solvent on extractive distillation', Chem Eng J, 87, 149±156. doi: 10.1016/S13858947(01)00211-X. Lewandowska W and Kujawski W (2007), `Ethanol production from lactose in a fermentation/pervaporation system', J Food Eng, 79, 430±437. doi: 10.1016/ j.jfoodeng.2006.01.071. Li C L and Lee K R (2006), `Dehydration of ethanol/water mixtures by pervaporation using soluble polyimide membranes', Polym Int, 55, 505±512. doi: 10.1002/pi.1991 Ligero E L and Ravagnani T M K (2003), `Dehydration of ethanol with salt extractive distillation ± a comparative analysis between processes with salt recovery', Chem Eng Proc, 42(7), 543±552. doi: 10.1016/S0255-2701(02)00075-2. Lin X, Kita H, and Okamoto K (2001), `Silicalite membrane preparation, characterization, and separation performance', Ind Eng Chem Res, 40, 4069± 4078. doi: 10.1021/ie0101947. Lin X, Chen X, Kita H, and Okamoto K (2003), `Synthesis of silicalite tubular membranes by in situ crystallization', AIChE J, 49(1), 237±247. doi: 10.1002/ aic.690490122. Lipnizki F, Hausmanns S, Laufenberg G, Field R, and Kunz B (2000), `Use of pervaporation±bioreactor hybrid processes in biotechnology', Chem Eng Technol, 23, 569±577. doi: 10.1002/1521-4125(200007)23. Liu S, Amidon T E, Francis R C, Ramarao B V, Lai Y Z, and Scott, G M (2006), `From forest biomass to chemicals and energy', Ind Biotechnol, 2(2), 113±120. doi: 10.1089/ind.2006.2.113. Llano-Restrepo M and Aguilar-Arias J (2003), `Modeling and simulation of saline extractive distillation columns for the production of absolute ethanol', Computers Chem Eng, 27, 527±549. doi: 10.1016/S0098-1354(02)00215-6. Luyben W L (2006), `Control of a multiunit heterogeneous azeotropic distillation process', AIChE J, 52(2), 623±637. doi: 10.1002/aic.10650. Lynd L R and Grethlein H E (1984), `IHOSR/Extractive distillation for ethanol separation', Chem Eng Prog, 80, 59±62. Maciel M R W and Brito R P (1995), `Evaluation of the dynamic behavior of an extractive distillation column for dehydration of aqueous ethanol mixtures', Computers Chem Eng, 19, Suppl, pp. S405±S408. Martin C R, Aranda P, and Chen W J (1995), `Pervaporation separation of ethanol/water mixtures by polystyrenesulfonate/alumina composite membranes', J Membr Sci, 107, 199±207. doi: 10.1016/0376-7388(95)00066-4. Matsuda H, Yanagishita H, Kitamoto D, Nakane T, Haraya K, Koura N, and Sano T

ß Woodhead Publishing Limited, 2010

Separation and purification processes

275

(1998), `Preparation of silicalite pervaporation membrane with high ethanol permselectivity', Membrane, 23, 259. Moermans B, De Beuckelaer W, Vankelecom I F J, Ravishankar R, Martens J A, and Jacobs P A (2000), `Incorporation of nano-sized zeolites in membranes', Chem Commun, 2467±2468. doi: 10.1039/b007435g. Morigami Y, Kondo M, Abe J, Kita H, and Okamoto K (2001), `The first large-scale pervaporation plant using tubular-type module with zeolite NaA membrane', Sep Purif Technol, 25(1±3), 251±260. doi: 10.1016/S1383-5866(01)00109-5. Moritz J W and Duff S J B (1996), `Simultaneous saccharification and extractive fermentation of cellulosic substrates', Biotechnol Bioeng, 49, 504±511. 10.1002/ (SICI)1097-0290(19960305)49. Nagase Y, Takamura Y, and Matsui K (1991), `Chemical modification of poly(substituted-acetylene): V. Alkylsilylation of poly(1-trimethylsilyl-1-propyne) and improved liquid separating property at pervaporation', J Appl Polym Sci, 42, 185±190. doi: 10.1002/app.1991.070420122. Namboodiri and Vane L M (2007), `High permeability membranes for the dehydration of low water content ethanol by pervaporation', J Membr Sci, 306, 209±215. doi: 10.1016/j.memsci.2007.08.050. Navajas A, Mallada R, Tellez C, Coronas J, Menendez M, and Santamaria J (2007), `Study on the reproducibility of mordenite tubular membranes used in the dehydration of ethanol', J Membr Sci, 299, 166±173. doi: 10.1016/ j.memsci.2007.04.038. Nomura M, Bin T, and Nakao S (2002), `Selective ethanol extraction from fermentation broth using a silicalite membrane', Sep Purif Technol, 27(1), 59±66. Nomura M, Yamaguchi T, and Nakao S (1998), 'Ethanol/water transport through silicalite membranes', J Membr Sci, 144, 161±171. doi: 10.1016/S03767388(98)00043-X. Nomura M, Yamaguchi T, and Nakao S (2001), `Transport phenomena through intercrystalline and intracrystalline pathways of silicalite zeolite membranes', J Membr Sci, 187(1±2), 203±212. doi: 10.1016/S0376-7388(01)00350-7. O'Brien D J and Craig J C (1996), `Ethanol production in a continuous fermentation/ membrane pervaporation system', Appl Microbiol Biotechnol, 44, 699±704. doi: 10.1007/s002530050619. Offeman R D, Stephenson S K, Franqui D, Cline J L, Robertson G H, and Orts W J (2008), `Extraction of ethanol with higher alcohol solvents and their toxicity to yeast', Sep Purif Technol, 63(2), 444±451. doi: 10.1016/j.seppur.2008.06.005. Okamoto K I, Kita H, Horii K, and Tanaka K (2001), `Zeolite NaA membrane: preparation, single-gas permeation, and pervaporation and vapor permeation of water/organic liquid mixtures', Ind Eng Chem Res, 40, 163±175. doi: 10.1021/ ie0006007. Partin L R (1996), `A case study in applying pc software: Flowsheet synthesis', Proceedings of Chemputers 4: Computer Tools for the Chemical Process Industries, Houston, Texas. 11±13. Pinto R T P, Wolf-Maciel M R, Lintomen L (2000), `Saline extractive distillation process for ethanol purification', Computers Chem Eng, 24, 1689±1694. doi: 10.1016/ S0098-1354(00)00455-5. Qiu W, Kosuri M, Zhou F, and Koros W J (2009), `Dehydration of ethanol±water mixtures using asymmetric hollow fiber membranes from commercial polyimides', J Membr Sci, 327, 96±103. doi: 10.1016/j.memsci.2008.11.029. Rakshit S K, Ghosh P, and Bisaria V S (1993), `Ethanol separation by selective

ß Woodhead Publishing Limited, 2010

276

Bioalcohol production

adsorption of water', Bioprocess Biosys Eng, 8(5±6) 279±282. doi: 10.1007/ BF00369841. Ruckenstein E and Liang L (1996), `Pervaporation of ethanol±water mixtures through polydimethylsiloxane-polystyrene interpenetrating polymer network supported membranes', J Membr Sci, 114, 227±234. doi: 10.1016/0376-7388(95)00319-3. Ruthven D M (1984), Principles of Adsorption and Adsorption Processes, John Wiley, New York. Sano T, Yanagishita H, Kiyozumi Y, Mizukami F, and Haraya K (1994), `Separation of ethanol±water mixture by silicalite membrane on pervaporation', J Membr Sci, 95 (3), 221±228. Sato K and Nakane T (2007), `A high reproducible fabrication method for industrial production of high flux NaA zeolite membrane', J Membr Sci, 301, 151±161. doi: 10.1016/j.memsci.2007.06.010. Seiler M, Jork C, Kavarnou A, Arlt W, and Hirsch R (2004), `Separation of azeotropic mixtures using hyperbranched polymers or ionic liquids', AIChE J, 50(10), 2439± 2454. doi: 10.1002/aic.10249. Sekulic J, Elshof J E T, and Blank D H A (2005), `Separation mechanism in dehydration of water/organic binary liquids by pervaporation through microporous silica', J Membr Sci, 254, 267±274. Shah D, Kissick K, Ghorpade A, Hannah R, and Bhattacharyya D (2000), `Pervaporation of alcohol±water and dimethylformamide±water mixtures using hydrophilic zeolite NaA membranes: mechanisms and experimental results', J Membr Sci, 179, 185± 205. doi: 10.1016/S0376-7388(00)00515-9. Shao P and Huang R Y M (2007), 'Polymeric membrane pervaporation', J Membr Sci, 287, 162±179. doi: 10.1016/j.memsci.2006.10.043. Shi Y, Wang X, Chen G, Golemme G, Zhang S, and Drioli E (1998), `Preparation and characterization of high-performance dehydrating pervaporation alginate membranes', J Appl Polym Sci, 68, 959±968. Shieh J J and Huang R Y M (1997), `Pervaporation with chitosan membranes, II. Blend membranes of chitosan and polyacrylic acid and comparison of homogeneous and composite membrane based on polyelectrolyte complexes of chitosan and polyacrylic acid for the separation of ethanol±water mixtures', J Membr Sci, 127, 185±202. doi: 10.1016/S0376-7388(96)00279-7. Slater C S, Hickey P J, and Juricic F P (1990), `Pervaporation of aqueous ethanol mixtures through poly(dimethyl siloxane) membranes', Sep Sci Technol, 25, 1063±1077. Sommer S and Melin T (2000), `Zeolite membranes in chemical industry', The 2000 18th Annual Membrane Technology/Separations Planning Conference, Boston, Massachusetts, USA. Sommer S and Melin T (2005), `Performance evaluation of microporous inorganic membranes in the dehydration of industrial solvents', Chem Eng Process, 44(10), 1138±1156. doi: 10.1016/j.cep.2005.02.006. Takegami S, Yamada H, and Tsujii S (1992), `Pervaporation of ethanol/water mixtures using novel hydrophobicmembranes containing polydimethylsiloxane'. J Membr Sci, 75, 93±105. Uragami T, Matsugi H, and Miyata T (2005), `Pervaporation characteristics of organic± inorganic hybrid membranes composed of poly(vinyl alcohol-co-acrylic acid) and tetraethoxysilane forwater/ethanol separation', Macromolecules, 38, 8440±8446. doi: 10.1021/ma051450k. Vane L M (2005), `A review of pervaporation for product recovery from biomass fermentation processes', J Chem Technol Biotechnol, 80, 603±629. doi: 10.1002/jctb.1265.

ß Woodhead Publishing Limited, 2010

Separation and purification processes

277

Vankelecom I F J, Depre D, De Beukelaer S, and Uytterhoeven J B (1995), `Influence of zeolites in PDMS membranes: pervaporation of water/alcohol mixtures', J Phys Chem, 99, 13193±13197. doi: 10.1021/j100035a024. Vankelecom I F J, De Beukelaer S, and Uytterhoeven J B (1997), `Sorption and pervaporation of aroma compounds using zeolitefilled PDMS membranes', J Phys Chem B, 101, 5186±5190. doi: 10.1021/jp962273f. van Veen H M, van Delft Y C, Engelen C W R, and Pex P P A C (2001), `Dewatering of organics by pervaporation with silica membranes', Sep Purif Technol, 22±23, 361± 366. doi: 10.1016/S1383-5866(00)00119-2. Verhoef A, Figoli A, Leen Bram, Bettens B, Drioli E, and Bruggen B V (2008), `Performance of a nanofiltration membrane for removal of ethanol from aqueous solutions by pervaporation', Sep Purif Technol, 60, 54±63. doi: 10.1016/ j.seppur.2007.07.044. Volkov V V, Fadeev A G, Khotimsky V S, Litvinova E G, Selinskaya Y A, McMillan J D, and Kelley S S (2004), 'Effects of synthesis conditions on the pervaporation properties of poly[1-(trimethylsilyl)-1-propyne] useful for membrane bioreactors', J Appl Polym Sci, 91(4), 2271±2277. doi: 10.1002/app.13358. Wang J F, Li C X, Wang Z H, Li Z J, and Jiang Y B (2007), `Vapor pressure measurement for water, methanol, ethanol, and their binary mixtures in the presence of an ionic liquid 1-ethyl-3-methylimidazolium dimethylphosphate', Fluid Phase Equilibria, 255, 186±192. doi: 10.1016/j.fluid.2007.04.010. Wasylkiewicz S K, Kobylka L C, and Castillo F J L (2003), `Synthesis and design of heterogeneous separation systems with recycle streams', Chem Eng J, 92, 201±208. doi: 10.1016/S1385-8947(02)00256-5. Weilnhammer C and Blass E (1994), `Continuous fermentation with product recovery by in-situ extraction', Chem Eng Technol, 17, 365±373. doi: 10.1002/ceat.270170602. Wyman C E (1996), Ethanol production from lignocellulosic biomass: overview, in: Wyman C E (ed.), Handbook on Bioethanol: Product Ion and Utilization, Taylor & Francis, Washington DC, pp. 1±18. Yanagishita H, Maejima C, Kitamoto D, and Nakane T (1994), `Preparation of asymmetric polyimide membrane for water/ethanol separation in pervaporation by the phase inversion process', J Membr Sci, 86, 231. Yang D,Li J, Jiang Z, Lu L, and Chen X (2009), `Chitosan/TiO2 nanocomposite pervaporation membranes for ethanol dehydration', Chem Eng Sci, 64, 3130±3137. doi: 10.1016/j.ces.2009.03.042. Yeom C K, Lee S H, and Lee J M (2001), `Pervaporative permeations of homologous series of alcohol aqueous mixtures through a hydrophilic membrane', J Appl Polym Sci, 79, 703±713. doi: 10.1002/1097-4628(20010124)79. Zhang Q K, Qian W C, and Jian W J (1984), `Study on extractive distillation with salt (III)', Petrochem Technol (China), 13, 1±9. Zhang W, Li G W, Fang Y J, and Wang X P (2007), `Maleic anhydride surfacemodification of crosslinked chitosan membrane and its pervaporation performance', J Membr Sci, 295, 130±138. doi: 10.1016/j.memsci.2007.03.001. Zhao J, Dong C C, Li C X, Meng H, and Wang Z H (2006), `Isobaric vapor±liquid equilibria for ethanol±water system containing different ionic liquids at atmospheric pressure', Fluid Phase Equilibria, 242, 147±153. doi: 10.1016/j.fluid.2006.01.023. Zhao Q, Qian J, An Q, Gao C, Gui Z, and Jin H (2009), `Synthesis and characterization of soluble chitosan/sodium carboxymethyl cellulose polyelectrolyte complexes and the pervaporation dehydration of their homogeneous membranes', J Membr Sci, 333, 68±78. doi: 10.1016/j.memsci.2009.02.001.

ß Woodhead Publishing Limited, 2010

11

Analytical monitoring of pretreatment and hydrolysis processes in lignocellulose-tobioalcohol production C . B E C K E R , L . N . S H A R M A and C . K . C H A M B L I S S , Baylor University, USA

Abstract: This chapter presents common analytical technologies used to study and/or monitor sample composition during chemical pretreatment and enzymatic hydrolysis of lignocellulosic materials. Analytical determinations of carbohydrates and degradation products present in biomass hydrolysates are reviewed. Capabilities for universal and selective detection are highlighted along with critical evaluation of the presented analytical strategies. Key words: biomass, carbohydrates, degradation products, chemical pretreatment, enzymatic hydrolysis, analytical monitoring.

11.1

Introduction

Bioprocessing of lignocellulosic materials to produce ethanol has rapidly emerged at the frontier of energy and fuels research; largely as a result of diminishing fossil-fuel supplies (Dresselhaus and Thomas 2001, Lynd et al. 1991, Lynd 1996, Mielenz 2001). Essentially all bioprocess strategies under consideration for commercial-scale ethanol production consist of the following three steps: (1) pretreatment, (2) enzymatic hydrolysis, and (3) subsequent fermentation to produce ethanol (Lynd 1996, Mielenz 2001, Sun and Cheng 2002, Wyman et al. 2005). Pretreatment of lignocellulosic materials is designed to release sugars and improve the digestibility of cellulose. However, a variety of degradation products are also formed during pretreatment, many of which are inhibitory and/or toxic to downstream enzymatic and/or microbial steps (Lynd 1996, Sun and Cheng 2002, Klinke et al. 2004, Olsson and Hahn-HaÈgerdal 1996, Palmqvist and Hahn-HaÈgerdal 2000). Accordingly, comprehensive understanding and quantitation of potentially-fermentable sugars and degradation products formed during pretreatment and hydrolysis of lignocellulosic materials are paramount to optimizing production of ethanol from biomass. Quantitative monitoring of any bioprocess typically represents a challenging endeavor, especially considering the complexity of biological materials. While

ß Woodhead Publishing Limited, 2010

282

Bioalcohol production

some processes may require information on a single analyte, a near quantitative mass balance is often necessary. More importantly, such quantitative information can be irrelevant if the relationships between various analytes and process outcome are not realized. As a result, extensive interdisciplinary collaborations can be required to develop bioprocess-monitoring schemes that inform on both the analytes present and their overall effect on process outcome. Regrettably, the rate-limiting step in development of useful bioprocessmonitoring strategies is often encountered in the laboratories of the analytical chemist. In order to determine the relationship between various chemical components of a bioprocess and the outcome of that process, these components must first be identified. Only once they are identified is it possible to develop intuitive, hypothesis-driven experiments which reveal underlying component/ outcome relationships. Again, once relationships are established, analytical technologies must either be employed or developed that are capable of effectively monitoring the bioprocess in real time (i.e., online monitoring). By nature of their constant and long-term use, online monitoring techniques must be automated, robust, sensitive, and inexpensive. An ideal online monitoring approach would also be able to provide high information density while minimizing the ambiguity of analyte identification. That is, a single system to accurately monitor all bioprocess components of interest is preferable to multiple analysis steps. As might be expected, analytical technologies capable of meeting all or even most of the above criteria for monitoring the pretreatment and hydrolysis of lignocellulosic materials are simply not available. In fact, owing to the high complexity of biomass hydrolysates, even determining sample composition remains a challenge (Hames 2009). Current state-of-the-art for analytical monitoring of bioprocesses related to pretreatment and hydrolysis of lignocellulosic materials is typically restricted to off-line analyses. Compare, for example, the desired performance characteristics of an ideal analytical monitoring strategy that are currently met by online versus off-line analytical techniques (Table 11.1). Existing online monitoring techniques for pretreatment and hydrolysis of lignocellulosic materials are limited and only monitor the evolution of select monosaccharides and degradation products (Cheng and Chang 2007, Okatch and Torto 2003, Rumbold et al. 2002). Although efforts have been directed toward the development of rapid analyses using spectroscopy and chemometric models (Laureano-Perez et al. 2005, Hames et al. 2003), these methods have not been widely implemented in online approaches due to their lack of maturity and the variable nature of biomass hydrolysates. Although off-line techniques currently offer several advantages over an online approach (esp. in terms of information provided), accurately determining the composition of biomass hydrolysates remains a problem that is not straightforward.

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

283

Table 11.1 The relationship between current state-of-the-art for online and off-line monitoring techniques and the desired performance characteristics of an ideal analytical monitoring strategy Ideal performance characterstics Rapid Robust Cost effective Information rich High sensitivity Specificity Full automation/control

11.2

Current state-of-the-art Off-line

Online

± X ± X X X ±

X ± ± ± ± ± ±

Target analytes resulting from pretreatment and hydrolysis processes

A number of components inherent to biomass are released during pretreatment and hydrolysis including ash, lignin residue, and various inorganic salts. With the exception of lignin residue, these components are typically assessed in biomass prior to pretreatment, thus outside the scope of the current discussion. In addition, methods for assessing these components have been reviewed elsewhere (Hames 2009, T Ehrman 1996). Techniques described in this chapter primarily focus on methods for monitoring those analytes that arise directly as a result of the decomposition of lignocellulosic material during pretreatment and hydrolysis. Renewable lignocellulosic materials such as corn stover, poplar wood, switch grass, and sorghum represent sustainable biomass feedstocks with potential for economically viable ethanol production. Lignocellulosic biomass is composed of cellulose (40±50%), hemicellulose (25±35%), and lignin (15±20%). Cellulose is a polymer of -1,4 linked glucose units, and hemicellulose is a heteropolymer of hexose and pentose sugars. These two components of lignocellulose (i.e., cellulose and hemicellulose) represent the origin of potentially fermentable sugars that might be obtained during pretreatment and hydrolysis of a biomass feedstock (Lynd et al. 1991, Lynd 1996, Galbe and Zacchi 2002, Ladisch et al. 1979). Lignin, on the other hand, is composed predominantly of condensed, polymeric phenyl-propane units. Chemical pretreatment and enzymatic hydrolysis processes (as outlined in Fig. 11.1) are aimed at converting a solid biomass feedstock into fermentable sugars. Numerous degradation products also result from these processes as the hemicellulose and lignin components of lignocellulose are broken down (Gable and Zacchi 2002, Larsson et al. 1999a, Lynd 1996, Olsson and Hahn-HaÈgerdal 1996). Hemicellulose is hydrolyzed to

ß Woodhead Publishing Limited, 2010

284

Bioalcohol production

11.1 Typical scheme for production of bioethanol from lignocellulosic biomass.

hexose and pentose sugars, which may be further degraded to weak organic acids and various furanic products (Klinke et al. 2004, Hamelinck et al., 2005). The hydrolysis of lignin also results in a variety of aliphatic and aromatic degradation products (Klinke et al. 2004, Olsson and Hahn-HaÈgerdal 1996, Demirbas 2008, Hamelinck et al. 2005, Larsson et al. 1999a). The amount of fermentable sugars available for bioethanol production, as well as the abundance and variety of degradation products produced during pretreatment depends largely on the relationship between feedstock and pretreatment conditions. An ideal pretreatment and hydrolysis scheme will both maximize the production of fermentable sugars and minimize potentiallyharmful degradation products such that the overall cost of bioethanol is reduced (Lynd 1996, Sun and Cheng 2002, Mosier et al. 2005b, Taherzadeh and Karimi 2008). Currently, while there are many different bioprocesses under development for the production of bioethanol, no single process has emerged as a superior technology (Mielenz 2001, Hamelinck et al. 2005, Mosier et al. 2005b, Taherzadeh and Karimi 2008, Szczodrak and Fiedurek 1996, Gomez et al. 2008b). Accordingly, pretreatment hydrolysates (i.e., the pretreated biomass sample) contain varying levels of solubilized cellulose, sugars released during lignocellulose degradation, and numerous degradation products which vary as a function of feedstock type and pretreatment conditions (Klinke et al. 2004, Olsson, Hahn-HaÈgerdal 1996, Palmqvist and Hahn-HaÈgerdal 2000, Chen et al. 2007). Although many of these degradation products have the potential to be utilized as value added products (Ladisch et al. 1979, Anderson et al. 2005, Howard et al. 2003, Landucci et al. 1994), others are harmful to downstream enzymatic and microbial processes (Klinke et al. 2004, Olsson and Hahn-

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

285

HaÈgerdal 1996, Palmqvist and Hahn-HaÈgerdal 2000, Klinke et al. 2001, Luo et al. 2002, Delgenes et al. 1996, Yourchisin and van Walsum 2004). Thus, identification of these degradation products and the development of a basic understanding of their role in bioalcohol production are paramount to optimizing biomass-to-ethanol conversion efficiencies (Klinke et al. 2004, Palmqvist and Hahn-HaÈgerdal 2000).

11.3

Detection strategies

The analytical technique(s) selected to monitor target analytes depends largely on the specific requirements of a given analysis. Broadly, analytical detection can be divided into universal and selective monitoring strategies. Both universal and selective detection are viable options depending, respectively, on whether a discovery-based or targeted analytical approach is desired. Universal detection, including low wavelength ultraviolet (UV) detection, refractive index (RI) detection, thermal conductivity detection (TCD), flame ionization detection (FID) and evaporative light scattering detection (ELSD) exhibits a response for all analytes containing specific physical or chemical properties. For example, low wavelength UV detectors will exhibit a response to any analyte containing a chromophore which absorbs light in the UV region. On the other hand, selective detection systems such as mass spectrometry (MS), pulsed amperometric detection (PAD), fluorescence spectroscopy, and conductometric detection techniques are capable of targeting properties which are specific to a particular class of compounds (e.g., PAD detection of carbohydrates) or even individual analytes (e.g., MS analysis). To compare the utility of both universal and selective detection techniques, analyses of a biomass hydrolysate by both UV and MS detection are provided in Figs 11.2 and 11.3. Figure 11.2 contains chromatograms demonstrating analysis of a biomass hydrolysate resulting from dilute-acid pretreatment of a corn stover feedstock by high performance liquid chromatography with simultaneous ultraviolet spectroscopy and tandem mass spectrometry (HPLC-UV-MS/MS) detection. As might be inferred from these data, selective detection is most applicable in targeted analysis where sensitivity and analyte identification/ verification are primary objectives. To obtain the chromatogram in Fig. 11.2 (Trace B), several analytes were selected by their unique m/z, and then dissociated in the gas phase via high energy collisions. The m/z of dissociation products that were unique to each molecule was then monitored. Obviously, prior knowledge of sample composition and the behavior of molecules (i.e., dissociation patterns) during analysis is required to perform such an analysis. In contrast, analysis of the same sample by low-wavelength UV detection (Fig. 11.2, Trace A) reveals chromatograms that are significantly more complex than those obtained by selective MS analysis and likely more representative of the sample composition. Several of the chromatographic peaks in Trace A were

ß Woodhead Publishing Limited, 2010

286

Bioalcohol production

11.2 Selected retention window for the reverse-phase separation of degradation products from a corn-stover hydrolysate using: (a) Low wavelength UV (210 nm) detection and (b) MS detection via selected reaction monitoring. Peaks are identified as: (1) 3,4-dihydroxybenzoic acid, (2) 2,5dihydroxybenzoic acid, (3) 3,4-dihydroxybenzoic acid, (4) Salicylic acid, and (5) 4-hydroxybenzaldehyde. Eluted peaks denoted by an asterisk (*) in the UV chromatogram represent unidentified sample components.

11.3 Selected retention window for the reverse-phase separation of a standard solution containing aliphatic acids, aldehydes, and phenolic compounds using: (a) Low wavelength UV (210 nm) detection and (b) MS detection via selected reaction monitoring. Peaks are identified as: (1) 4-hydroxyacetophenone, (2) Caffeic acid, (3) Syringic acid, (4) Vanillin, and (5) 4-hydroxybenzoic acid.

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

287

tentatively identified by careful comparison of peak retention times in the sample to those of a calibration standard. However, a number of additional peaks remain unidentified in the UV chromatogram (indicated by asterisks). The observation of these peaks is useful to the analyst because they inform on the presence of additional components which should be further interrogated if they represent a substantial fraction of the sample's composition. A potential caveat to universal detection strategies is that important information may be lost or misrepresented when assessing complex samples. For example, Fig. 11.3 shows a region of a HPLC-UV-MS/MS chromatogram where multiple components coelute. As can be seen from Fig. 11.3, several low abundance analytes (peaks 3 and 4) remain undetected by low-wavelength UV detection (Trace A). Using only low-wavelength UV, the peak eluting at 34 minutes may be identified as either caffeic acid or syringic acid. Assessment of detector response at several wavelengths is commonly used to evaluate peak purity and enhance confidence in component identification (Chen et al., 2006). Although inspection of the data in Fig. 11.3 at multiple wavelengths would likely rule out syringic acid and confirm the presence of caffeic acid in the given example, results would be inconclusive if both analytes were present in the sample at concentrations that exhibit similar detector response. Both selective and universal detection strategies may be used for quantitative purposes. As with most universal detection techniques, the peak identification of the targeted analytes from UV detection relies primarily on retention time. Therefore quantitation can be challenging when multiple components are coeluting. For example, peaks 2, 3, and 4 (Fig. 11.2, Trace A) all require the use of complex deconvolution software to determine theoretical detector response in the absence of coeluting species. Moreover, although peaks 3 and 4 were not observed by UV detection (Fig. 11.3), the UV response of compounds 2 and 5 will be influenced by coeluting compounds 3 and 4 respectively, thus introducing error into their quantitative analysis. Selective detection offers the advantage of cleaner chromatograms and enhanced sensitivity for quantitative analysis, but is limited by the need for prior knowledge of sample composition.

11.4

Preparation of biomass hydrolysates for analytical characterization

Prior to analysis, biomass hydrolysates typically require some degree of sample preparation. Sample preparation is necessary to eliminate both soluble and insoluble materials that may interfere with a specified analysis. The type and extent of sample preparation depend on both the sample matrix and analyte of interest, and usually involves some combination of dilution, filtration, extraction, and pre-fractionation (Nogueira et al. 2005, Chen et al. 2006, Sanz and Martinez-Castro 2007, Sharma et al. 2009, Smith 2003, Soga and Serwe 2000). Analysis of carbohydrates present in relatively cleaner samples can involve only

ß Woodhead Publishing Limited, 2010

288

Bioalcohol production

dilution and filtration steps to prepare samples for analysis (Nogueira et al. 2005, Soga and Serwe 2000). However, the complex matrices typically encountered in pretreatment and hydrolysis samples have dictated that additional preparative steps for carbohydrate analysis be employed (e.g., liquid-liquid extractions and extraction and pre-fractionation on solid phase extraction (SPE) cartridges (Laureano-Perez et al. 2005, Matsumuto et al. 2005, Hameister and Kragl 2006, Jonsson and Mathiasson 2001, Davis 1998, Gilar et al. 2001, Schiller et al. 2002)). Similarly, analysis of lignocellulosic degradation products is preceded by multiple preparative steps including dilution, precipitation and filtration (to remove insoluble solid materials, carbohydrates as well as bulky bio-polymers), and extraction of the target analytes with a suitable organic solvent such as methyl tert-butyl ether (MTBE), methylene chloride, ethanol, or ethyl acetate (Chen et al. 2006, Sharma et al. 2009, Klinke et al. 2001, Luo et al. 2002, DiNardo and Larson 1994, Klinke et al. 2002). Although an exhaustive description of the various preparation techniques that have been employed for analysis of carbohydrates and lignocellulosic degradations products is beyond the aim of this chapter, the reader is referred to the references discussed in each section for more detailed information.

11.5

Analysis of carbohydrates

Several analytical methodologies may be employed to monitor carbohydrates existing in biomass hydrolysates. These methodologies range from simple colorimetric assays to more advanced chromatographic techniques. Selection of a specific technique will depend both on the technologies available to the practitioner and on what information is required from the analysis. For example, if the experimental objective is quantitation of glucose and the presence of other carbohydrates is irrelevant, the specificity of a simple enzymatic assay might be preferable over a more complex chromatographic approach. Alternatively, some form of chromatography would be required for quantitation of individual carbohydrates. A summary of the carbohydrates most frequently encountered in biomass hydrolysates is contained in Table 11.2. Glucose, which originates from degradation of hemicellulose and cellulose during pretreatment, and more substantially from enzymatic hydrolysis of cellulose, is perhaps the most ubiquitous monomeric carbohydrate in contemporary bioethanol literature owing to its well-studied conversion to ethanol via yeast fermentation. However, many other monomeric carbohydrates are present in biomass hydrolysates as a result of hemicellulose hydrolysis, which occurs during chemical pretreatment. In addition to monomeric sugars, many di- and oligosaccharides are formed during pretreatment, typically from incomplete enzymatic hydrolysis (Dien et al. 2006, Gomez et al. 2008a). Many of these additional carbohydrates (especially xylose) are substrates of various fermentation strategies including fermentation by

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

289

Table 11.2 Common sugars present in biomass hydrolysates of lignocellulosic materials Sources Hemicellulose Sugars Monomers Dimers Oligomers

Cellulose

(5-Carbon/unit)

(6-Carbon/unit)

(6-Carbon/unit)

Arabinose Xylose Xylobiose Xylotriose Xylotetrose Xylopentose Xylohexose

Galactose Mannose ± ± ± ± ±

Glucose ± Cellobiose Cellotriose Cellotetrose Cellopentose Cellohexose

alternative yeasts (Olsson and Hahn-HaÈgerdal 1996, Delgenes et al. 1996, Mosier et al. 2005a, Nichols et al. 2008) and bacterias (Olsson and Hahn-HaÈgerdal 1996, Delgenes et al. 1996, Ingram et al. 1999). Thus, techniques capable of quantifying the various carbohydrates present in a given biomass hydrolysate are paramount to realizing the full energy content of lignocellulosic materials.

11.5.1 Colorimetric carbohydrate analysis Colorimetric assays are relatively simple methodologies which may be utilized to monitor carbohydrates formed during hydrolysis of lignocellulosic materials. These assays provide a direct correlation between the absorbance (i.e. color intensity) of a carbohydrate-containing solution and the corresponding carbohydrate concentration. Colorimetric methods are fast and convenient procedures capable of directly determining the amount of carbohydrates released from lignocellulosic substrates (Irick et al. 1988, King and Garner 1947, Schwald et al. 1988). Enzymatic assays provide selective methods which are substrate specific, whereas chemical methods can provide both quasi-selective and universal carbohydrate determination. Enzyme assays Enzymatic methods for detection of carbohydrates are rapid, highly specific, and typically sensitive to low concentrations of analyte (ranging from low parts-permillion to parts-per-thousand concentrations) (Beach and Turner 1958, Comer 1956, Huggett and Nixon 1956, Marks 1996, Saifer and Gerstenfeld 1958). In addition, minimal sample preparation is required when compared to more extensive analysis techniques (Marks 1996, Berlin et al. 2006b, Chung et al. 1997). Glucose oxidase/peroxidase assay (Comer 1956, Huggett and Nixon

ß Woodhead Publishing Limited, 2010

290

Bioalcohol production

1956, Marks 1996, Saifer and Gerstenfeld 1958) is the most common example of enzymatic carbohydrate analysis applied to biomass hydrolysates (Chung et al. 1997, Breuil and Saddler 1985). This method provides a direct assessment of free glucose in a sample which is useful to determine both hemicellulose-derived glucose or to monitor glucose evolution during enzymatic hydrolysis of cellulose. Frequently, total glucose concentration of pretreated biomass is also desired (i.e., free and cellulose-bound glucose) and for such a requirement, cellulose can be quantitatively hydrolyzed to glucose prior to analysis (Berlin et al. 2006b, Chung et al. 1997, Breuil and Saddler 1985, Skomarovsky et al. 2006, Spindler et al. 1990). Some less common enzymatic methods for analysis of carbohydrates involve the use of alternative enzymes (Spackman and Cobb 2002, Viola and Davies 1992, Mitchell et al. 1998). Many of these enzymatic methods are commercially available as enzyme kits and are routinely applied for the analysis of carbohydrates in the food industry. These methods may be applied for the analysis of carbohydrates in biomass hydrolysate at the practitioner's discretion, taking care to be mindful of potential interferences from sample components which inhibit enzyme action. It has been found that the accuracy of glucose oxidase/peroxidase is influenced by compounds present during analysis (Schwald et al. 1988, Breuil and Saddler 1985, Fales et al. 1961). For example, Brueil and Saddler (1985) observed a significant underestimation (20±80%) of glucose concentration in steam-explosion-pretreated aspen wood hydrolysate containing lignocellulosic degradation products. Similar interferences may be expected for other enzymatic assays as well. Chemical assays Chemical methods for colorimetric carbohydrate analysis rely on unique chemistries between reducing sugars (i.e., sugars that form a free aldehyde or carbonyl group in basic solution which is subsequently susceptible to reduction by an oxidizing reagent) and various reagents to produce a colorimetric response. Several chemical methods including dinitrosalicylic acid (DNS) (Miller 1959), Somogyi-Nelson (S-M) (Nelson 1994, Somogyi 1945, 1952), Anthrone (Mishima et al. 2006), and the phenol sulfuric acid methods (DuBois et al. 1956, Fournier 2005, Mecozzi 2005) are all available for sugar analysis. DNS and S-M methods are the most commonly used chemical methods for analysis of reducing sugars (Skomarovsky et al. 2006, Mishima et al. 2006, Coward-Kelly et al. 2003, Green et al. 1989, Saha et al. 2005). While not all carbohydrates are reducing sugars, non-reducing sugars can be chemically converted to reducing sugars prior to analysis. For example, the Anthrone and phenol sulfuric acid methods have been applied to determine concentrations of both the reducing and non-reducing sugars in a sample (Skomarovsky et al. 2006, Mishima et al. 2006, Fournier 2005, Mecozzi 2005). In this approach, analysis of non-reducing sugars is achieved by conversion to reducing sugars under strongly acidic conditions,

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

291

and the concentration of non-reducing sugars is determined by difference, before and after conversion. Despite providing rapid, facile methods for carbohydrate analysis, chemical methods suffer from several caveats. For example, chemical methods do not provide information on individual carbohydrates; rather they provide an estimate of total sugars present in a given sample. Carbohydrate concentration is determined as glucose equivalents, which introduces error as calibration curves cannot be accurate without prior knowledge of sample composition (Schwald et al. 1988, Fales et al. 1961, Berlin et al. 2006a, Jeffries et al. 1998). In addition, solution absorbance during sugar analysis using the Anthrone method is influenced by halides (e.g., chloride and bromide ions) which are present in the sample matrix (Fales et al. 1961). Analysis of carbohydrates by the phenol sulphuric acid method is also found to be affected by ions such as Cu2+, Fe2+, Al3+, NO2ÿ, and NO3ÿ, and organic acids such as lactic acid, and tartaric acid present in the samples (Fujikawa et al. 1974). Such complications introduce additional uncertainty; especially in samples such as biomass hydrolysates where sample composition can vary drastically from batch to batch.

11.5.2 Gas chromatography Gas chromatography (GC) has frequently been employed for carbohydrate analysis. In contrast to colorimetric techniques, GC enables separation of carbohydrates (prior to detection) for individual interrogation. The separation in GC is based on the differential interaction of analytes in the gas phase and their partitioning to the stationary phase of the GC column. Hydrogen and other inert gases such as helium and nitrogen are typically used as a carrier gas to move the vaporized analytes through a GC column. The retention time of analytes depends both upon differences in their vapor pressures at a given temperature and how they interact with the stationary phase. Detection techniques utilized in gas chromatography Depending on the specificity of information required, gas chromatography methods can be coupled with universal or selective detectors. Flame ionization detection (FID) provides a universal detection technique for GC which has been utilized by several research groups for carbohydrate analysis (Dien et al. 2006, Adams et al. 1999, Black and Fox 1996, Cocchi et al. 2006, Sweeley et al. 1963). However, since analyte retention time is the only identification parameter in GC-FID, determined concentrations can be influenced by other sample components that coelute with the analyte of interest, resulting in either overestimation of analyte concentration or even a false-positive response when analyte is not present in the sample. Thus, great care must be taken to validate reported concentrations. Flame ionization detection also suffers from low

ß Woodhead Publishing Limited, 2010

292

Bioalcohol production

sensitivity (compared to MS detection) which can prove problematic when analyzing carbohydrates which occur in pretreatment hydrolysates at trace concentrations or when high dilution volumes are used. When available, mass spectrometry can provide a more attractive alternative to flame ionization detection. The use of GC-MS has been demonstrated for analysis of carbohydrates in bacterial cell walls (Black and Fox 1996, Fox 1999), foods and beverages (Cocchi et al. 2006, Cotte et al. 2003), and lignocellulosic materials (Bonaduce et al. 2007, Kamm et al. 2006, WillfoÈr et al. 2009). Gas chromatography-mass spectrometry (GC-MS) offers high sensitivity, and the MS detector can uniquely be operated in either universal (full scan experiments) or selective (i.e. tandem MS) detection modes. Full scan MS experiments provide data over a selected m/z range on any analyte ionized. Alternatively, tandem MS experiments employ a unique molecular transition for each analyte of interest, which allows for selective analysis even in the presence of coeluting components (Fox 1999, Kamm et al. 2006, Steinberg and Fox 1999, Medeiro and Simoneit 2007, Laine et al. 2002, Fox et al. 1998). In addition, tandem MS experiments offer higher sensitivity compared to full scan analysis since only targeted analytes are allowed to pass through the mass spectrometer to the detector. Carbohydrate derivatization A primary disadvantage of utilizing GC techniques is that carbohydrate derivatization is required prior to analysis. Carbohydrates are non-volatile species. Thus, they must be converted to volatile derivatives before being introduced to GC instrumentation. There are many methods available for derivatization of carbohydrates. Formation of acetyl, trifluoroacetyl (TFA), and trimethylsilyl (TMS) derivatives are commonly used methods to convert sugars to volatile derivatives (Sanz and Martinez-Castro 2007, Black and Fox 1996, MolnaÂr-Perl 2000, Sawardeker et al. 1965). However, when using TMS or other similar derivatization procedures, a simple monomeric carbohydrate can produce up to four chromatographic peaks if it possesses an anomeric center and exists in both pyranose and furanose forms (Fig. 11.4) (Black and Fox 1996, MolnaÂr-Perl 2000, Sawardeker et al. 1965, DeJongh et al. 1969, Knapp 1979). Incomplete derivatization can further complicate chromatograms. The presence of multiple peaks resulting from a single analyte is obviously problematic in situations where sample composition is already highly complex. One approach to circumvent the occurrence of multiple peaks from a single carbohydrate is to reduce sugars to their alditol derivatives, thus eliminating the anomeric center of the sugar and its ability to form multiple ring structures (Black and Fox 1996, Fox et al. 1998, Knapp 1979, Harley et al. 2002, Kim et al. 1967). However, when taking a reductive approach, one must be mindful that different carbohydrates can yield the same alditol. For example, the reduction of

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

293

11.4 Isomers of D-glucose that are detected during GC analysis.

D-arabinose and D-xylose both yield D-arabitol (Knapp 1979). Conversion of sugars to oxime derivatives prior to silylation with TMS has also been employed to simplify chromatograms obtained by GC analysis of carbohydrates (Adams et al. 1999, Black and Fox 1996, Cocchi et al. 2006, Kamm et al. 2006, MolnaÂrPerl 2000, Katona et al. 1999, MolnaÂr-Perl et al. 1984). In addition to chromatographic complications (i.e., extraneous peaks), chemical derivatization is a time-consuming process, often requiring as much or more time than the separation itself (Black and Fox 1996, Kamm et al. 2006, Agblevor et al. 2004, 2007, Starke et al. 2000). Although useful for research and discovery, the total length of analysis dictates that GC methods are less feasible candidates for online monitoring strategies. It is also relevant to note that errors may occur during analysis of larger (higher mass) carbohydrates owing to the thermally labile nature of carbohydrate oligomers (Raemy and Schweizer 1983).

11.5.3 High performance liquid chromatography High performance liquid chromatography (HPLC) is a common approach for separation and analysis of non-volatile compounds. Since the first application

ß Woodhead Publishing Limited, 2010

294

Bioalcohol production

towards carbohydrate analysis in biomass hydrolysates in the early 1980s, the use of HPLC methods to analyze sugars has been growing rapidly (C I Ehrman 1996, Black and Fox 1996, MolnaÂr-Perl 2000, Churms 1996, Davies and Hounsell 1996). Owing to the hydrophilic nature of sugars, traditional, reverse phase liquid chromatography (RPLC) is not a practical approach for carbohydrate analysis. Although RPLC is capable of separating dimers from trimers as well as individual dimers and trimers, the monomeric sugars most abundant in biomass hydrolysates are weakly retained and thus not easily resolved (MolnaÂrPerl 2000, Hicks 1998, Verzele et al. 1987). Ion-exchange techniques Ion-exchange techniques (e.g., ion-exchange chromatography and ligandexchange chromatography) are frequently employed for analysis of sugars in biomass pretreatment hydrolysates. A common approach utilizing ion exchange is carbohydrate separation on strong-base anion-exchange columns (Mosier et al. 2005a, Nichols et al. 2008, Sluiter et al. 2008, Lee 1990, Kaar et al. 1998, Cataldi et al. 2000, Guignard et al. 2005). Under alkaline conditions, carbohydrates exist as their oxyanion derivatives which can be separated by high-performance anion exchange chromatography (HPAEC) (C I Ehrman 1996, Zook and LaCours 1995). Alternatively, cation-exchange columns have been applied for ligandexchange chromatography separations which take advantage of facile generation of different on-column counter-ions that can alter selectivity depending on whether analysis is directed at monosaccharides (Baker and Himmel 1986, Noll et al. 1990) or at disaccharides and other polysaccharides (C I Ehrman 1996, Hicks 1998, Baker and Himmel 1986, Brobst and Scobell 1982). For example, optimal separation of monosaccharides can be achieved using, H+-, lead-, and strontium-loaded ion exchange columns (Baker and Himmel 1986, Noll et al. 1990, C I Ehrman 1996), while silver and calcium ions provide optimal separations for oligo- and polysaccharides (C I Ehrman 1996, Black and Fox 1996, Hicks 1998, Verzele et al. 1987, Noll et al. 1990, Brobst and Scobell 1982). Alternative HPLC separation techniques Several other separation techniques have also been employed for carbohydrate separation. A number of mixed-bed specialty columns have been developed which use proprietary resins designed to optimize various carbohydrate separations (Mosier et al. 2005a, Sluiter et al. 2008, Balan et al. 2009, Dadi et al. 2007). Normal-phase chromatography using amino-modified silica gel has also been used to achieve efficient separation of monomeric carbohydrates (Hicks 1998, Verzele et al. 1987). This approach demonstrated full or partial resolution of several monomeric sugars and provides the option to tailor selectivity through choice of the amino-modifier. The disadvantage associated with use of modified

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

295

silica-gel columns is that reducing sugars tend to form a glycosylamine bond with the amino group in the stationary phase; resulting in unstable column performance, poor reproducibility, and a significantly-reduced column life. Universal detection techniques for analysis via HPLC Both refractive index (RI) and evaporative light scattering (ELS) detectors are attractive universal detectors for HPLC analysis of sugars. RI detectors respond to the change in refractive index of a solution in the presence of analyte, and ELS detectors measure light scattered by analyte molecules. The use of RI detection for the analysis of carbohydrates in biomass hydrolysates is fairly prevalent in literature (Mosier et al. 2005a, Kaar et al. 1998, Balan et al. 2009, Dadi et al. 2007, Gomez et al. 2002). However, low sensitivity and incompatibility of RI detectors with analyses that require gradient separations can be serious limitations in some instances (Cataldi et al. 2000, Martens and Frankenberger 1991). ELS detection has also been successfully demonstrated for the analysis of carbohydrates in biomass hydrolysates and unlike RI detection, ELS detectors can be employed where gradient elution profiles are used (Agblevor et al. 2007, Ganzera and Stuppner 2005, Karlsson et al. 2005, Wei and Ding 2000). However, potential drawbacks with ELS detection techniques include non-linearity in detector response and diverse differences in sensitivity among target analytes present in biomass hydrolysates (Cataldi et al. 2000, Martens and Frankenberger 1991). Selective detection techniques for analysis via HPLC Pulsed amperometric detection (PAD) can provide quasi-selective detection of carbohydrates. The response in amperometric detection depends upon the reduction or oxidation current of analytes at a particular electrode potential. As only specific analyte types are oxidized or reduced at a given potential, there is less chance of interference from coeluting analytes. Although not specific to individual carbohydrates, PAD can be performed such that only carbohydrates are detected. The application of PAD to analyze sugars has been growing rapidly also owing to its facile coupling with ion exchange chromatography and other HPLC separations (Guignard et al. 2005, Zook and LaCours 1995, Wright and Wallis 1996, Sevcik et al. 2010, Lee et al. 2007, 2008, Kano et al. 1996, MarkoVarga et al. 1994). The compatibility of PAD for gradient analysis makes it an attractive alternative to RI detection and picomole detection limits have been reported, indicating PAD can be a sensitive technique (Cataldi et al. 2000, Davies and Hounsell 1996). However, it is important to note that during analyses of sugars in samples that have been both pretreated and hydrolyzed; the high sensitivity of PAD may require the incorporation of multiple dilutions in order to adjust the analyte concentration to within the pulsed amperometric detector's linear range.

ß Woodhead Publishing Limited, 2010

296

Bioalcohol production

Examples of selective detection of HPLC-separated carbohydrates by MS detection are infrequent in the literature. Although decoupled analysis using matrix assisted laser desorption ionization (MALDI) to ionize pre-fractioned carbohydrates has been reported (Burrell et al. 2007, Gholipour et al. 2008a, 2008b, Hashir et al. 2007), the extended analysis times and unreliable quantitation due to multiple peaks at low mass (<800 amu) has resulted in limited applicability of MALDI-MS techniques. Atmospheric-pressure electrospray ionization (ESI) provides a more compatible ionization source which can be directly coupled to HPLC systems. Although monitoring the negative ion spectra for carbohydrate determination has been demonstrated (Antonio et al. 2008, Cohen et al. 2004), the high pH conditions which promote formation of negative-ions tend to foul ESI sources. Alternative options for mass-selective detection include monitoring carbohydrates as their molecular adducts of H+, Na+, Li+, NH4+, Cs+ and Clÿ (Black and Fox 1996, Kamm et al. 2006, Guignard et al. 2005, Rogatsky et al. 2005, Richardson et al. 2001, Tolstikov and Fiehn 2002); however, the high salt content of eluents that is required to form these molecular adducts typically results in poor performance using ESI-MS. Derivatization strategies have also been employed which enable the detection of derivatized carbohydrates as their protonated adducts via ESI-MS (Ding et al. 2008). In addition, post-column ion suppression can be employed to enable ESI from eluents containing high salt concentration or at high pH. Efficient techniques to achieve ion suppression have employed ion-exchange membranes and microdialysis for mobile phases of high ionic strength, and electrolysis of hydroxide to water for mobile phases at high pH (Okatch and Torto 2003, Rumbold et al. 2002, Guignard et al. 2005).

11.6

Analysis of lignocellulosic degradation products

Despite their commercial importance (in terms of both their inhibitory effect during carbohydrate fermentations and their potential as value added products), the quantitative analysis of lignocellulosic degradation products has received relatively little attention (compared to carbohydrates) among the scientific community. A significant effort has been extended towards understanding inhibitory effects of known degradation products. For example, acetic acid, furfural and various aromatic aldehydes have been shown to be inhibitory (in varying degrees) to fermentation processes involving Sachharomyces ceriviciae and E. coli. (Klinke et al. 2001, Luo et al. 2002, Zaldivar et al. 2001). A number of degradation products have also been identified as value added products (Ladisch et al. 1979, Anderson et al. 2005, Howard et al. 2003, Landucci et al. 1994, Takeuchi et al. 2007). However, owing to the variable nature of biomass feedstocks and the wide variety of pretreatment processes under investigation, many degradation products observed by universal detection strategies can not be identified at a high level of confidence. The issue of high-confidence

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

297

11.5 The three primary residues of lignin.

identifications is further complicated by the fact that degradation products generated during pretreatment are sensitive to a number of pretreatment variables including selection of feedstock and pretreatment method, and the various conditions under which the pretreatment is performed (e.g., temperature, time, pressure, oxygen content, and pH) (Klinke et al. 2004, Olsson and HahnHaÈgerdal 1996, Palmqvist and Hahn-HaÈgerdal 2000, Demirbas 2008, Larsson et al. 1999a, Chen et al. 2007). It is generally believed that a majority of the unidentified degradation products produced originate from lignin decomposition. For example, a majority of identified aromatic degradation products have been linked directly to derivatives of para-hydroxyphenyl (H), guaiacyl (G), and syringyl (S) lignin residues (see, for example, Figs 11.5 and 11.6) (Klinke et al. 2004, Yu et al. 2008). However, it has also been demonstrated that sugar degradation products such as 5-hydroxymethyl furfural, furfural, and levulinic acid can be formed directly from degradation of hemicellulose-derived sugars under pretreatment conditions (Klinke et al. 2004, Palmqvist and Hahn-HaÈgerdal 2000, Demirbas 2008). Thus, novel methods capable of monitoring a variety of degradation products, as well as being able to identify when additional, unidentified degradation products are present are required.

11.6.1 Gas chromatography To date, the diversity of degradation products generated during pretreatment has traditionally required the use of multiple analytical techniques that are selected based on the chemical class of the target analytes. Both GC-FID and GC-MS methods have been reported for analysis of degradation products in biomass hydrolysates (Luo et al. 2002, DiNardo and Larson 1994, Klinke et al. 2002, Marko-Varga et al. 1994, Fenske et al. 1998, Larsson et al. 1999b, Niemela 1988, 1998, Niemela and Sjostrom 1986, Pecina et al. 1986). Similar to analysis of carbohydrates, degradation products typically require derivatization prior to analysis. TMS derivatization of carboxylic acids, aromatic aldehydes and phenolic compounds is a common approach of derivatization for GC analyses in

ß Woodhead Publishing Limited, 2010

298

Bioalcohol production

11.6 Representative degradation products originating from degradation of the guaiacyl alcohol residue of lignin.

biomass hydrolysate samples (Klinke et al. 2002, Fenske et al. 1998, Larsson et al. 1999b, Niemela 1988, Niemela and Sjostrom 1986, Pecina et al. 1986). Carboxylic acids and aldehydes have also been analyzed with a combined diazomethane and hydroxylamine derivatization method (DiNardo and Larson 1994). A significant complication to developing robust methodologies using GC techniques is that it is inherently difficult to develop derivatization strategies that enable quantitative monitoring for samples of unknown composition.

11.6.2 High performance liquid chromatography A number of HPLC methodologies have been implemented for analysis of degradation products present in biomass hydrolysates. As with GC, HPLC approaches also suffer from the complexity of biomass samples and have historically achieved poor chromatographic resolution which can be attributed to the presence of numerous components from different compound classes. In an effort to simplify chromatographic analysis, multiple chromatography strategies have been employed which vary depending on analyte class. Aliphatic acids are traditionally determined using high-performance anion-exchange chromatography with UV (Klinke et al. 2002, Gey 1987, Persson et al. 2002) or conductivity detection (Klinke et al. 2002, Persson et al. 2002, Bonn et al. 1988).

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

299

Alternatively, analysis of aromatic acids, aldehydes, furans, and other phenolic compounds have been accomplished using reversed-phase chromatography with refractive index (Luo et al. 2002, Yu et al. 2008, Burtscher et al. 1987, Guo et al. 2008, Kootstra et al. 2009), UV (Klinke et al. 2002, Pecina et al. 1986, Kootstra et al. 2009, Lapierre et al. 1983, Arnetoli et al. 2008, Bonn et al. 1984), or mass spectrometry detection (Klinke et al. 2002, Persson et al. 2002, Matta et al. 2003). More recently, RP-HPLC-UV methods for the simultaneous determination of aliphatic, aromatic and phenolic degradation products have also been demonstrated (Chen et al. 2006, Shui and Leong 2002) using non-linear gradient elution. Despite the utility of more comprehensive separations, unique determination of multiple analytes from different classes in biomass hydrolysates remains challenging (Chen et al. 2006). In an effort to enhance confidence in analyte identification, an RP-HPLC-ESI-UV-MS/MS method for selective analysis of degradation products from multiple compound classes has also been demonstrated (Sharma et al. 2009). This approach combines the benefits associated with both universal and selective detection strategies and enabled the simultaneous identification and quantitation of 39 select aliphatic, aromatic, and aldehyde degradation products in biomass hydrolysates (Sharma et al. 2009, Nichols et al. 2008). It is important to note that while coeluting species in the hydrolysate sample matrix do not interfere with detection of target analytes during MS/MS analysis, they can significantly affect their ionization efficiency at the ionization source. The influence of matrix components on ionization efficiency has been shown to result in both signal enhancement and signal reduction for a given analyte. This phenomenon is referred to as `matrix effects' (Tang and Kabarle 1993). Matrix effects commonly cause pre-established calibration curves to become unreliable, significantly affecting analyte quantitation (Tang and Kebarle 1993, Matuszewski et al. 1998, 2003, Taylor 2005). Several common approaches to compensate for matrix effects during quantitative analyses are available, including isotope-dilution, matrix-matched calibration, and standard addition methods (Matuszweski 2003, Ramirez et al. 2007, Vanderford and Snyder 2006). Isotope dilution methods, which employ labeled standards of analytes to spike the sample before analysis, are considered the most preferable approach to compensate for matrix effects. However, owing to the general absence of commercially available labeled standards and the cost-intensive nature of analysis, isotope dilution has limited applicability. Although matrix-matched calibration attempts to mimic or simulate the existing matrix in calibration standards (thus minimizing the influence of matrix interference during quantitative analysis), obtaining a clean matrix blank is difficult, if not impossible, for biomass hydrolysates. Despite being labor intensive and requiring larger sample volumes, single point or multi-point standard addition methods along with thorough sample cleanup are thus the most intuitive approach to effectively compensate for observed matrix effects (Sharma et al. 2009).

ß Woodhead Publishing Limited, 2010

300

11.7

Bioalcohol production

Alternative techniques for analysis of carbohydrates and degradation products

Several alternative analytical techniques including capillary electrophoresis (CE) (Honda et al. 1989, Horning et al. 1973, Oefner et al. 1992, El Rassi 1994, Suzuki and Honda 1998), Fourier transform-infrared (FT-IR) spectroscopy (Tarkosova and Kopikova 2000, Hashimoto and Kameoka 2000, Hashimoto et al. 2005, Blanco et al. 2004), and C-13 NMR (Duquesnoy et al. 2008) are available for the analysis of carbohydrates and degradation products in various matrices. However, these techniques are less commonly employed as compared to chromatographic techniques. CE methods have been increasingly applied to analyze carbohydrates in various biological matrices such as hydrolyzed plant materials (Dahlman et al. 2000, Huber et al. 1994) and food and beverages (Moreno et al. 2002, Morales-Cid et al. 2007, Khandurina 2008). Aliphatic and aromatic acids, aromatic aldehydes and ketones, and furanic compounds (i.e., the lignocellulosic degradation products present in biomass hydrolysates) have also been successfully analyzed with capillary electrophoresis (Khandurina 2008, Dupont et al. 2007, Law et al. 2007, Rivasseau et al. 2006, Iinuma and Herrmann 2003). Analysis of carbohydrates and degradation products by CE in biomass hydrolysates presents several challenges to the analytical community. The separation of analytes in capillary electrophoresis depends upon the differential migration of charged molecules in an electric field. Since most carbohydrates are neutral they must be first converted into charged species. Charging carbohydrates in CE buffer has been accomplished via complexation with inorganic metal cations or oxyacids (e.g., borate, stannate, mobibdate), or by inducing ionization under strong alkaline conditions (El Rassi 1994, 1997, Khandurina 2008, Oefner and Chiesa 1994). Lignocellulosic degradation products, which are primarily either weakly acidic or have a fairly high pKa, require derivatization strategies similar to carbohydrates in order to achieve separation employing CE methodologies (Khandurina 2008, Dupont et al. 2007, Law et al. 2007, Rivasseau et al. 2006, Iinuma and Herrmann 2003). Where separation via CE is feasible, development of suitable detection strategies remains a challenge. For example, most carbohydrates lack the strong chromophoric or fluoriphoric functional groups required for typical optical detectors that are used with CE separations. Although low-wavelength UV measurements are possible, the sensitivity and selectivity of these measurements are very low due to the low extinction coefficients associated with most carbohydrates (El Rassi 1994, Honda 1996). Thus they must be derivatized with suitable ultraviolet-absorbing or fluorescent chromophores by various pre- and post-column derivatization schemes prior to analysis (Khandurina 2008, Oefner and Chiesa 1994, Honda 1996). Detection of carbohydrates is also possible using refractometric and electrochemical detection techniques; however, the reported sensitivity of those

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

301

measurements is low (Khandurina 2008, Honda 1996). Despite the increasing application of CE methods in biological samples, the application of CE towards analysis of both carbohydrates and lignocellulosic degradation products in biomass hydrolysates can prove especially difficult owning to the complex nature of biomass-hydrolysate matrices. Two approaches have been taken to alleviate the complications with detection of carbohydrates during CE analysis. In one approach, coupling of CE with ESI-MS for detection of carbohydrates was demonstrated to provide efficient detection of CE-separated carbohydrates (Khandurina 2008, Iinuma and Herrmann 2003). Such a strategy would also likely be applicable to analysis of degradation products. However, a significant caveat to CE-MS analysis is the requirement of either MS-friendly buffers or advanced desalting techniques (Khandurina 2008, Klampfl 2006, Khandurina and Guttman 2005), thereby introducing additional complication into the analysis. A separate approach has also been taken using electrochemical detection. Since carbohydrates are weak acids, they are converted to their oxy-anions at high pH (>12) and these oxyanions have been directly detected using CE and pulsed amperometric detection (Khandurina 2008, Oefner and Chiesa 1994). It is important to note that neither of these strategies has been assessed for the analysis of analytes existing in biomass hydrolysates. Previous experience in our laboratory and other labs that routinely conduct biomass analyses has proven that many of the clever carbohydrate separations that have been successfully demonstrated in literature for relatively less-complex sample matrices have not proven to be applicable to analyses conducted on lignocellulosic hydrolysates.

11.8

Conclusion and future trends

Ideally, the future of bioprocess monitoring would be driven by techniques capable of providing not only information on the wide array of components present during the processing of lignocellulosic materials, but also information which alerts users to which components are having the greatest impact on the process itself. As described in this chapter, many strides have been made towards more completely understanding the composition of biomass hydrolysates. However, little progress has been made towards understanding how that composition ultimately effects the production of bioethanol. From an analytical perspective, the key to moving forward will be the development of rapid analysis techniques that can both handle the complexity of biomass sample matrices and keep pace with emerging multiplexed assessments of pretreatment, enzymatic hydrolysis, and fermentation steps in the bioconversion process. Achievement of this goal is likely to depend not only on the availability of novel analysis paradigms, but also on the availability of advanced statistical models and computational software promoting elucidation of the complex relationships that are likely to exist between sample composition and bioconversion efficiency.

ß Woodhead Publishing Limited, 2010

302

11.9

Bioalcohol production

References

Adams, M.A., Chen, Z., Landman, P. and Colmer, T.D. (1999) `Simultaneous determination by capillary gas chromatography of organic acids, sugars, and sugar alcohols in plant tissue extracts as their trimethylsilyl derivatives', Analytical Biochemistry, 266 (1), 77±84. DOI: 10.1006/abio.1998.2906. Agblevor, F.A., Murden, A. and Hames, B.R. (2004) `Improved method of analysis of biomass sugars using high-performance liquid chromatography', Biotechnology Letters, 26 (15), 1207±1210. DOI: 10.1023/B:BILE.0000036596.95796.42. Agblevor, F.A., Hames, B.R., Schell, D. and Chum, H. L. (2007) `Analysis of biomass sugars using a novel HPLC method', Applied Biochemistry and Biotechnology, 136 (3), 309±326. DOI: 10.1007/s12010-007-9028-4. Anderson, W.F., Peterson, J., Akin, D.E. and Morrison, W.H. (2005) `Enzymatic treatment of grass lignocellulose for potential high-value co-products and improved fermentable substrate', Applied Biochemistry and Biotechnology, 121±124, 303± 310. Antonio, C., Pinheiro, C., Chaves, M.M., Ricardo, C.P., Ortuno, M.F. and Thomas-Oates, J. (2008) `Analysis of carbohydrates in Lupinus albus stems on imposition of water deficit, using porous graphitic carbon liquid chromatography±electrospray ionization mass spectrometry', Journal of Chromatography A, 1187, 111±118. Arnetoli, M., Montegrossi, G., Buccianti, A. and Gonnelli, C. (2008) `Determination of organic acids in plants of Silene paradoxa L. by HPLC', Journal of Agricultural and Food Chemistry, 56 (3), 789±795. Baker, J.O. and Himmel, M.E. (1986) `Separation of sugar anomers by aqueous chromatography on calcium- and lead-form ion-exchange columns: application to anomeric analysis of enzyme reaction products', Journal of Chromatography A, 357 (1), 161±181. DOI: 10.1016/S0021-9673(01)95818-2. Balan, V., da Costa Sausa, L., Chundawat, S.P.S., Marshall, D., Sharma, L.N., Chambliss, C.K. and Dale, B.E. (2009) `Enzymatic digestibility and pretreatment degradation products for AFEX-treated Hardwood (Populus Nigra)', Biotechnology Progress, in press. DOI 10.1021/bp.160. Beach, E.F. and Turner, J.J. (1958) `An enzymatic method for glucose determination in body fluids', Clinical Chemistry, 4 (6), 462±475. Berlin, A., Balakshin, M., Gilkes, N., Kadla, J., Maximenko, V., Kubo, S. and Saddler, J. (2006a) `Inhibition of cellulase, xylanase and b-glucosidase activities by softwood lignin preparations', Journal of Biotechnology, 125, 198±209. Berlin, A., Maximenko, V., Bura, R., Kang, K., Gilkes, N. and Saddler, J. (2006b) `A rapid microassay to evaluate enzymatic hydrolysis of lignocellulosic substrates', Biotechnology and Bioengineering, 93 (5), 880±886. DOI: 10.1002/bit.20783. Black, G.E. and Fox, A. (1996), `Recent progress in the analysis of sugar monomers from complex matrices using chromatography in conjunction with mass spectrometry or stand-alone tandem mass spectrometry', Journal of Chromatography A, 720 (1±2), 51±60. DOI: 10.1016/0021-9673(95)00335-5. Blanco, M., Peinado, A.C. and Mas, J. (2004) `Analytical monitoring of alcoholic fermentation using NIR spectroscopy', Biotechnology and Bioengineering, 88 (4), 536±542. Bonaduce, I., Brecoulaki, H., Colombini, M.P., Lluveras, A., Restivo, V. and Ribechini, E. (2007) `Gas chromatographic±mass spectrometric characterization of plant gums in samples from painted works of art', Journal of Chromatography A, 1175, 275± 282.

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

303

Bonn, G., Pecina, R., Burtscher, E. and Bobleter, O. (1984) `Separation of wood degradation products by high-performance liquid chromatography', Journal of Chromatography A, 287 (1), 215±221. DOI: 10.1007/BF00473895. Bonn, G., Oefner, P.J. and Bobleter, O. (1988) `Analytical determination of organic acids formed during hydrothermal and organosolv degradation of lignocellulosic biomass'', Fresenius Z Anal Chem, 331 (1), 46±50. DOI: 10.1007/BF00376197. Breuil, C. and Saddler, J.N. (1985) `Limitations of using D-glucose oxidase peroxidase method for measuring glucose derived from lignocellulosic substrates', Biotechnology Letters, 7 (3), 191±196. DOI: 10.1007/BF01027818. Brobst, K.M. and Scobell, H.D. (1982), `Modern chromatographic methods for the analysis of carbohydrate mixtures', Starch, 34 (4), 117±121. DOI: 10.1002/ star.19820340405. Burrell, M.M., Earnshaw, C.J. and Clench, M.R. (2007) `Imaging matrix assisted laser desorption ionization mass spectrometry: a technique to map plant metabolites within tissues at high spatial resolution', Journal of Experimental Botany, 58 (4), 757±763. Burtscher, E., Bobleter, O., Schwald, W., Concin, R. and Binder, H. (1987) `Chromatographic analysis of biomass reaction products produced by hydrothermolysis of poplar wood', Journal of Chromatography A, 390 (2), 401± 412. Cataldi, T.R.I., Campa, C. and de Benedetto, G.E. (2000) `Carbohydrate analysis by highperformance anion-exchange chromatography with pulsed amperometric detection: the potential is still growing', Fresenius Z Anal Chem, 368 (8), 739±758. DOI: 10.1007/s002160000588. Chen, S.-F., Mowery, R.A., Castleberry, V.A., van Walsum, P.G. and Chambliss, C.K. (2006) `High-performance liquid chromatography method for simultaneous determination of aliphatic acid, aromatic acid and neutral degradation products in biomass pretreatment hydrolysates', Journal of Chromatography A, 1104 (1±2), 54±61. DOI: 10.1016/j.chroma.2005.11.136. Chen, S.-F., Mowery, R.A., Chambliss, C.K. and van Walsum, G.P. (2007), `Pseudo reaction kinetics of organic degradation products in dilute-acid-catalyzed corn stover pretreatment hydrolysates', Biotechnology and Bioengineering, 98, 1135± 1145. DOI: 10.1021/jf0700327. Cheng, C. and Chang, K.C. (2007) `Sampling, dilution, and loading device-coupled highperformance liquid chromatography method for successive on-line analyses of major carbohydrate products in immobilized cellulase hydrolysate of paper cellulose', Analytical Science, 23 (3), 305±310. Chung, Y., Bakalinsky, A. and Penner, M.H. (1997) `Analysis of biomass cellulose in simultaneous saccharification and fermentation processes', Applied Biochemistry and Biotechnology, 66, (3), 249±262. DOI: 10.1007/BF02785591. Churms, S.C. (1996) `Recent progress in carbohydrate separation by high-performance liquid chromatography based on hydrophilic interaction', Journal of Chromatography A, 720 (1±2), 75±91. DOI: 10.1016/0021-9673(95)00306-1. Cocchi, M., Durante, C., Grandi, M., Lambertini, P., Manzini, D. and Marchetti, A. (2006) `Simultaneous determination of sugars and organic acids in aged vinegars and chemometric data analysis', Talanta, 69 (5), 1166±1175. DOI: 10.1016/ j.talanta.2005.12.032. Cohen, A., SchagerloÈf, H., Nilsson, C., Melander, C., Tjerneld, F. and Gorton, L. (2004) `Liquid chromatography±mass spectrometry analysis of enzyme-hydrolysed carboxymethylcellulose for investigation of enzyme selectivity and substituent

ß Woodhead Publishing Limited, 2010

304

Bioalcohol production

pattern', Journal of Chromatography A, 1029 (1±2), 87±95. DOI: 10.1016/ j.chroma.2003.12.010. Comer, J.P. (1956) `Semiquantitative specific test paper for glucose in urine', Analytical Chemistry, 28 (11), 1748±1750. DOI: 10.1021/ac60119a030. Cotte, J.F., Casabianca, H., Chardon, S., Lheritier, J. and Grenier-Loustalot, M.F. (2003) `Application of carbohydrate analysis to verify honey authenticity', Journal of Chromatography A, 1021 (1), 145±155. DOI: 10.1016/j.chroma.2003.09.005. Coward-Kelly, G., Aiello-Mazzari, C., Kim, S., Granda, C. and Holtzapple, M. (2003) `Suggested improvements to the standard filter paper assay used to measure cellulase activity', Biotechnology and Bioengineering, 82 (6), 745±749. DOI: 10.1002/bit.10620. Dadi, A.P., Schall, C.A. and Varanasi, S. (2007) `Mitigation of cellulose recalcitrance to enzymatic hydrolysis by ionic liquid pretreatment', Applied Biochemistry Biotechnology, 137 (1), 136±140. DOI: 10.1007/s12010-007-9068-9. Dahlman, O., Jacobs, A., Liljenberg, A. and Olsson, A.I. (2000) `Analysis of carbohydrates in wood and pulps employing enzymatic hydrolysis and subsequent capillary zone electrophoresis', Journal of Chromatography A, 891 (1), 157±174. DOI: 10.1016/S0021-9673(00)00619-1. Davies, M.J. and Hounsell, E.F. (1996) `Carbohydrate chromatography: towards yoctomole sensitivity', Biomedical Chromatography, 10 (6), 285±289. DOI: 10.1002/(SICI)1099-0801(199611)10:6<285::AID-BMC616>3.0.CO;2-W. Davis, M.W. (1998) `A rapid modified method for compositional carbohydrate analysis of lignocellulosics by high pH anion-exchange chromatography with pulsed amperometric detection (HPAEC/PAD)', Journal of Wood Chemistry and Technology, 18 (2), 235±252. DeJongh, D.C., Radford, T., Hribar, J.D., Hanessian, S., Bieber, M., Dawson, G. and Sweeley, C.C. (1969) `Analysis of trimethylsilyl derivatives of carbohydrates by gas chromatography and mass spectrometry', Journal of the American Chemical Society, 91 (7), 1728±1740. Delgenes, J.P., Moletta, R. and Navarro, J.M. (1996) `Effects of lignocellulose degradation products on ethanol fermentations of glucose and xylose by Saccharomyces cerevisiae, Zymomonas mobilis, Pichia stipitis, and Candida shehatae', Enzyme Microbial Technology, 19, 220±225. Demirbas, A. (2008) `Products from lignocellulosic materials via degradation process', Energy Sources, 30 (1), 27±37. DOI: 10.1080/00908310600626705. Dien, B.S., Jungb, H.J.G., Vogel, K.P., Caslerd, M.D., Lamb, J.F.S., Iten, L., Mitchell, R.B. and Sarath, G. (2006) `Chemical composition and response to dilute-acid pretreatment and enzymatic saccharification of alfalfa, reed canarygrass, and switchgrass', Biomass and Bioenergy, 30 (10), 880±891. DOI: 10.1016/ j.biombioe.2006.02.004. DiNardo, R.C. and Larson, R.A. (1994) `Constituents of newsprint and municipal solid waste hydrolysates', Bioresource Technology, 48 (2), 135±141. Ding, C., Wang, L., Tian, C., Li, Y., Sun, Z., Wang, H., Suo, Y. and You, J. (2008) `LCDAD-ESI-MS characterization of carbohydrates using a new labeling reagent', Chromatographia, 68 (11±12), 893±902. DOI: 10.1365/s10337-008-0805-1 Dresselhaus, M.S. and Thomas, I.L. (2001) `Alternative energy technologies', Nature, 414, 332±337. DuBois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A. and Smith, F. (1956) `Colorimetric method for determination of sugars and related substances', Analytical Chemistry, 28 (3), 350±356.

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

305

Dupont, A., Egasse, C. and Morin, A.V., (2007) `Comprehensive characterisation of cellulose- and lignocellulose degradation products in aged papers: capillary zone electrophoresis of low-molar mass organic acids, carbohydrates, and aromatic lignin derivatives', Carbohydrate Polymers, 68, 1±16. Duquesnoy, E., Castola, V. and Casanova, J. (2008) `Identification and quantitative determination of carbohydrates in ethanolic extracts of two conifers using 13C NMR spectroscopy', Carbohydrate Research, 343, 893±902. Ehrman, C.I. (1996) `Methods for the chemical analysis of biomass process streams', in Handbook on Bioethanol, ed. C.E. Wymen, Taylor & Francis, Washington, D.C., 395±415. Ehrman, T. (1996) `Determination of acid-soluble lignin in biomass', National Renewable Energy Laboratory: Golden, CO; http://cobweb.ecn.purdue.edu/~lorre/ 16/research/LAP-004.pdf (accessed March 2009). El Rassi, Z. (1994) `Capillary electrophoresis of carbohydrates', Advances in Chromatography, 34, 177±250. El Rassi, Z. (1997) `Recent developments in capillary electrophoresis of carbohydrate species', Electrophoresis, 18 (12), 2400±2407. Fales, F.W., Russell J.A. and Fain J.N. (1961) `Some applications and limitations of the enzymic, reducing (Somogyi), and Anthrone methods for estimating sugars', Clinical Chemistry, 7 (4), 289±303. Fenske, J.J., Griffin, D.A. and Penner, M.H. (1998) `Comparison of aromatic monomers in lignocellulosic biomass Prehydrolysates', Journal of Industrial Microbiology Biotechnology, 20 (6), 364±368. DOI: 10.1038/sj.jim.2900543. Fournier, E. (2005) `Colorimetric quantification of carbohydrates', in Current Protocols in Food Analytical Chemistry, John Wiley and Sons, 653±660. DOI: 10.1002/ 0471142913.fae0101s00. Fox, A. (1999) `Carbohydrate profiling of bacteria by gas chromatography±mass spectrometry and their trace detection in complex matrices by gas chromatography± tandem mass spectrometry', Journal of Chromatography A, 843 (1±2), 287±300. DOI: 10.1016/S0021-9673(98)00884-X. Fox, K.F., Wunschel, D.S., Fox, A. and Stewart, G.C. (1998) `Complementarity of GC± MS and LC±MS analyses for determination of carbohydrate profiles of vegetative cells and spores of bacilli', Journal of Microbiological Method, 33 (1), 1±11. DOI: 10.1016/S0167-7012(98)00030-X. Fujikawa, T., Shiomi, M., Mukai, S. and Wada, M. (1974) `Phenol sulphuric acid for the quantitative analysis of sugar. Influence of impurities', Nippon Nogeikagaku Kaishi, 48 (9), 483±491. Galbe, M. and Zacchi, G. (2002) `A review of the production of ethanol from softwood', Applied Microbiology Biotechnology, 59 (6), 618±628. DOI: 10.1007/s00253-0021058-9. Ganzera, M. and Stuppner, H. (2005) `Evaporative light scattering detection (ELSD) for the analysis of natural products', Current Pharmaceutical Analysis, 1 (2), 135±144. DOI: 10.2174/1573412054022699. Gey, M. (1987) `Characterization of biotechnological processes and products using highperformance liquid chromatography (HPLC). I. Separations of carbohydrates, organic acids and lipids', Acta Biotechnologica, 8 (2), 197±205. DOI: 10.1002/ abio.370080216. Gholipour, Y., Nonami, H. and Erra-Balsells, R. (2008a) `Application of pressure probe and UV-MALDI-TOF MS for direct analysis of plant underivatized carbohydrates in subpicoliter single-cell cytoplasm extract', Journal of the American Society of

ß Woodhead Publishing Limited, 2010

306

Bioalcohol production

Mass Spectrometry, 19, 1841±1848. Gholipour, Y., Nonami, H. and Erra-Balsells, R. (2008b) `In situ analysis of plant tissue underivatized carbohydrates and on-probe enzymatic degraded starch by matrixassisted laser desorption/ionization time-of-flight mass spectrometry by using carbon nanotubes as matrix', Analytical Biochemistry, 383, 159±167. Gilar, M., Bouvier, E.S.P. and Compton, B.J. (2001) `Advances in sample preparation in electromigration, chromatographic and mass spectrometric separation methods', Journal of Chromatography A, 909, 111±135. Gomez, L., Rubio, E. and Auge, M. (2002) `A new procedure for extraction and measurement of soluble sugars in ligneous plants', Journal of the Science of Food and Agriculture, 82, 360±369. Gomez, L.D., Bristow, J.K., Statham, E.R. and McQueen-Mason, S.J. (2008a) `Analysis of saccharification in Brachypodium distachyon stems under mild conditions of hydrolysis', Biotechnology for Biofuels, 1 (15), pp. 15, http:// www.ncbi.nlm.nih.gov/pmc/articles/PMC2577630/pdf/1754-6834-1-15.pdf (accessed 28 October 2009). DOI: 10.1186/1754-6834-1-15. Gomez, L.D., Steele-King, C.G. and McQueen-Mason, S.J. (2008b) `Sustainable liquid biofuels from biomass: the writing's on the walls', New Phytologist, 178, 473±485. DOI: 10.1111/j.1469-8137.2008.02422.x. Green, F., Clausen, C.A. and Highley, T.L. (1989) `Adaptation of the Nelson-Somogyi reducing-sugar assay to a microassay using microliter plates', Analytical Biochemistry, 182 (2), 197±199. DOI: 10.1016/0003-2697(89)90578-2. Guignard, C., Jouve, L., BogeÂat-Triboulot, M.B., Dreyer, E., Hausman, J.F. and Hoffmann, L. (2005) `Analysis of carbohydrates in plants by high-performance anion-exchange chromatography coupled with electrospray mass spectrometry', Journal of Chromatography A, 1085 (1), 137±142. DOI: 10.1016/ j.chroma.2005.05.068. Guo, G.-L., Chen, W.-H., Men, L.-C. and Hwang, W.-S. (2008) `Characterization of dilute acid pretreatment of silvergrass for ethanol production', Bioresource Technology, 99, 6046±6053. Hameister, D. and Kragl, U. (2006) `Selective recovery of carbohydrates from aqueous solution facilitated by a carrier', Engineering in Life Sciences, 6, 187±192. Hamelinck, C.N., van Hooijdonk, G. and Faaij, A.P.C. (2005) `Ethanol from lignocellulosic biomass: techno-economic performance in short-, middle- and long-term', Biomass and Bioenergy, 28 (4), 384±410. Hames, B.R. (2009) `Biofuels: methods and protocols', in Methods in Molecular Biology, ed. J.R. Mielenz, Humana Press, Clifton, NJ, pp. 145±167. Hames, B.R., Thomas, S.R., Sluiter, D., Roth, C.J. and Templeton, D.W. (2003) `Rapid biomass analysis: new tools for compositional analysis of corn stover feedstocksand process intermediates from ethanol production', Applied Biochemistry and Biotechnology, 105, 5±16. Harley, W.M., Kozar, M.P. and Fox, A. (2002) `Trace analysis of muramic acid in indoor air using an automated derivatization instrument and GC±MS2 or GC±MS', Jornal of Microbiological Method, 51 (1), 95±104. Hashimoto, A. and Kameoka, T. (2000) `Mid-infrared spectroscopic determination of sugar contents in plant-cell culture media using an ATR method', Appied Spectroscopy, 54 (7), 1005±1011. Hashimoto, A., Yamanaka, A., Kanou, M., Nakanishi, K. and Kameoka, T. (2005) `Simple and rapid determination of metabolite content in plant cell culture medium using an FT-IR/ATR method', Bipprocess Biosystem Engineering, 27, 115±123.

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

307

Hashir, M.A., Stecher, G., Bakry, R., Kasemsook, S., Blassnig, B., Feuerstein, I., Abel, G., Popp, M., Bobleter, O. and Bonn, G.K. (2007) `Identification of carbohydrates by matrix-free material-enhanced laser desorption/ionization, mass spectrometry', Rapid Communications in Mass Spectrometry, 21, 2759±2769. Hicks, K.B. (1998) `High-performance liquid chromatography of carbohydrates', Advances in Carbohydrate Chemistry and Biochemistry, 46, 17±72. Honda, S. (1996) `Postcolumn derivatization for chromatographic analysis of carbohydrates', Journal of Chromatography A, 720 (1±2), 183±199. DOI: 10.1016/0021-9673(95)00212-X. Honda, S., Iwase, S., Makino, A. and Fujiwara, S. (1989) `Simultaneous determination of reducing monosaccharides by capillary zone electrophoresis as the borate complexes of N-2-pyridylglycamines', Analytical Biochemistry, 176, 72±77. Horning, C.G., Horning, M.J., Carroll, D.I., Dzidic, I. and Stillwell, R.N. (1973) `New picogram detection system based on a mass spectrometer with an external ionization source at atmospheric pressure', Analytical Chemistry, 45, 936±943. Howard, R.L., Abotsi, E., Jansen van Rensburg, E. L. and Howard, S. (2003) `Lignocellulose biotechnology: issues of bioconversion and enzyme production', African Journal of Biotechnology, 2 (12), 602±619. Huber, C., Grill., E., Oefner, P. and Bobleter, O. (1994) `Capillary electrophoretic determination of the component monosachharides in hemicelluloses', Fresenious Journal of Analytical Chemistry, 348, 825±831. Huggett, A.S.G. and Nixon, D.A. (1956) `Use of glucose oxidase, peroxidase and odianisidine in determination of blood and urinary glucose', The Lancet, 270 (6991), 368±370. DOI: 10.1016/S0140-6736(57)92595-3. Iinuma, Y. and Herrmann, H. (2003) `Method development for the analysis of particle phase substituted methoxy phenols and aromatic acids from biomass burning using capillary electrophoresis/electrospray ionization mass spectrometry (CE/ESI-MS)', Journal of Chromatography A, 1018, 105±115. Ingram, L.O., Aldrich, H.C., Borges, A.C.C., Causey, T.B., Martinez, A., Morales, F., Saleh, A., Underwood, S.A., Yomano, L.P., York, S.W., Zaldivar, J. and Zhou, S. (1999), `Enteric bacterial catalysts for fuel ethanol production', Biotechnology Progress, 15, 855±866. Irick, T.J., West, K., Brownell, H.H., Schwald, W. and Saddler, J.N. (1988), `Comparison of colorimetric and HPLC techniques for quantitating the carbohydrate components of steam-treated wood', Applied Biochemistry and Biotechnology, 17 (1±3), 137± 149. DOI: 10.1007/BF02779152. Jeffries, T.W., Yang, V.W. and Davis, M.W. (1998) `Comparative study of xylanase kinetics using Dinitrosalicylic, Arsenomolybdate, and ion chromatographic assays', Applied Biochemistry and Biotechnology, 70±72 (1), 257±265. DOI: 10.1007/ BF02920142. Jonsson, J.A. and Mathiasson, L. (2001), `Membrane extraction techniques for sample preparation', Advances in Chromatography, 41, 53±91. Kaar, W.E., Gutierrez, C.V. and Kinoshita, C.M. (1998) `Steam explosion of sugarcane bagasse as a pretreatment for conversion of ethanol', Biomass and Bioenergy, 14 (3), 277±287. DOI: 10.1016/S0961-9534(97)10038-1. Kamm, B., Kamm, M., Schmidt, M., Starke, I. and Kleinpeter, E. (2006) `Chemical and biochemical generation of carbohydrates from lignocellulose-feedstock (Lupinus nootkatensis) ± quantification of glucose', Chemosphere, 62 (1), 97±105. DOI: 10.1016/j.chemosphere.2005.03.073. Kano, K., Takagi, K., Inoue, K., Ikeda, T. and Ueda, T. (1996) `Copper electrodes for

ß Woodhead Publishing Limited, 2010

308

Bioalcohol production

stable subpicomole detection of carbohydrates in high-performance liquid chromatography', Journal of Chromatography A, 721 (1), 53±57. DOI: 10.1016/ 0021-9673(95)00757-1. Karlsson, G., Winge, S. and Sandberg, H. (2005) `Separation of monosaccharides by hydrophilic interaction chromatography with evaporative light scattering detection', Journal of Chromatography A, 1092 (2), 246±249. DOI: 10.1016/ j.chroma.2005.08.025. Katona, Z.F., Sass, P. and MolnaÂr-Perl, I. (1999) `Simultaneous determination of sugars, sugar alcohols, acids and amino acids in apricots by gas chromatography±mass spectrometry', Journal of Chromatography A, 847, 91±102. Khandurina, J. (2008) `Analysis of carbohydrates by capillary electroelectrophoresis', in Handbook of Capillary and Microchip Electrophoresis and Associated Microtechniques, 3rd edition, Landers, J.P. (ed.), CRC Press, Boca Raton, FL, pp. 251±294. Khandurina, J. and Guttman, A. (2005) `High resolution capillary electrophoresis of oligosaccharide structural isomers', Chromatographia, 62, S37±S41. Kim, J.H., Shome, B., Liao, T.H. and Pierce, J.G. (1967) `Analysis of neutral sugars by gas-liquid chromatography of alditol acetates: Application to thyrotropic hormone and other glycoproteins', Analytical Biochemistry, 20 (2), 258±274. King, E.J. and Garner, R.J. (1947) `The colorimetric determination of glucose', Journal of Clinical Pathology, 1, 30±33. Klampfl, C.W. (2006) `Recent advances in the application of capillary electrophoresis with mass spectrometric detection', Electrophoresis, 27, 3±34. Klinke, H.B., Thomsen, A.B. and Ahring, B.K. (2001) `Potential inhibitors from wet oxidation of wheat straw and their effect on growth and ethanol production by Thermoanaerobacter mathranii', Applied Microbiololy and Biotechnology, 57 (5± 6), 631±638. DOI: 10.1007/s002530100825. Klinke, H.B., Ahring, B.K., Schmidt, A.S. and Thomsen, A.B. (2002) `Characterization of degradation products from alkaline wet oxidation of wheat straw', Bioresource Technology, 82 (1), 15±26. DOI: 10.1016/S0960-8524(01)00152-3. Klinke, H.B., Thomsen, A.B. and Ahring, B.K. (2004) `Inhibition of ethanol-producing yeast and bacteria by degradation products produced during pre-treatment of biomass', Applied Microbiology and Biotechnology, 66 (1), 10±26. DOI: 10.1007/ s00253-004-1642-2. Knapp, D.R. (1979) Handbook of Analytical Derivatization Reactions, John Wiley and Sons. Kootstra, A.M.J., Mosier, N.S., Scott, E.L., Beeftink, H.H. and Sanders, J.P.M. (2009) `Differential effects of mineral and organic acids on the kinetics of arabinose degradation under lignocellulose pretreatment conditions', Biochemical Engineering Journal, 43, 92±97. Ladisch, M.R., Flickinger, M.C. and Tsao, G.T. (1979) `Fuels and chemicals from biomass energy', The International Journal, 4 (20), 135±164. Laine, C., Tamminen, T., Vikkula, A. and Vuorinen, T. ( 2002) `Methylation analysis as a tool for structural analysis of wood polysaccharides', Holzforschung, 56, 607±614. Landucci, R., Goodman, B. and Wyman, C. (1994) `Methodology for evaluating the economics of biologically producing chemicals and materials from alternative feedstocks', Applied Biochemistry and Biotechnology, 45 (1), 677±696. Lapierre, C., Rolando, C. and Monties, B. (1983) `Characterization of poplar lignins acidolysis products: capillary gas-liquid and liquid-liquid chromatography of monomeric compounds', Holzforschung, 37, 189±198.

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

309

Larsson, S., Palmqvist, E.H., B., Tengborg, C., Stenberg, K., Zacchi, G. and Nilvebrant, N.O. (1999a) `The generation of fermentation inhibitors during dilute acid hydrolysis of softwood', Enzyme Microbial Technology, 24 (3±4), 151±159. DOI: 10.1016/S0141-0229(98)00101-X. Larsson, S., Reimann, A., Nilvebrant, N.O. and Jonsson, L.J. (1999b) `Comparison of different methods for the detoxification of lignocellulose hydrolysates of spruce', Applied Biochemistry and Biotechnology, Twentieth Symposium on Biotechnology for Fuels and Chemicals (77±79), pp. 91±103. Laureano-Perez, L., Teymouri, F., Alizadeh, H. and Dale, B.E. (2005) `Understanding factors that limit enzymatic hydrolysis of biomass: characterization of pretreated corn stover', Applied Biochemistry and Biotechnology, Twenty-Sixth Symposium on Biotechnology for Fuels and Chemicals (121±124), pp. 1081±1099. Law, W.S., Zhao, J.H., Hauser, P.C. and Li, S.F.Y. (2007) `Capillary electrophoresis with capacitively coupled contactless conductivity detection for low molecular weight organic acids in different samples', Journal of Separation Science, 30, 3247±3254. Lee, J.W., Gwak, K.S., Park, J.Y., Park, M.J., Choi, D.H., Kwon, M. and Choi, I.G. (2007) `Biological pretreatment of softwood Pinus densiflora by three White Rot Fungi', Jornal of Microbiology, 45 (6), 485±491. Lee, J.W., Kim, H.Y., Koo, B.W., Choi, D.H., Kwon, M. and Choi, I.G. (2008) `Enzymatic saccharification of biologically pretreated Pinus densiflora using enzymes from Brown Rot Fungi', Journal of Bioscience and Bioenineering, 106 (2), 162±167. DOI: 10.1263/jbb.106.162. Lee, Y.C. (1990) `High-performance anion-exchange chromatography for carbohydrate analysis', Analytical Biochemistry, 189 (2), 151±162. DOI: 10.1016/00032697(90)90099-U. Luo, C., Brink, D.L. and Blanch, H.W. (2002) `Identification of potential fermentation inhibitorsin conversion of hybrid poplar hydrolysate to ethanol', Biomass Bioenergy, 22 (2), 125±138. DOI: 10.1016/S0961-9534(01)00061-7. Lynd, L.R. (1996) `Overview and evaluation of fuel ethanol from cellulosic biomass: technology, Economics, the Environment, and Policy', Annual Review of Energy and the Environment, 21, 403±465. DOI: 10.1146/annurev.energy.21.1.403. Lynd, L.R., Cushman, J.H., Nichols, R.J. and Wyman, C.E. (1991) `Fuel ethanol from cellulosic biomass', Science, 251 (4999), 1318±1323. DOI: 10.1126/science.251. 4999.1318. Marko-Varga, G., Buttler, T., Gorton, L., Olsson, L., Durand, G. and Barcelo, D. (1994) `Qualitative and quantitative carbohydrate analysis of fermentation substrates and broths by liquid chromatographic techniques', Journal of Chromatography A, 665 (2), 317±332. Marks, V. (1996) `An improved glucose-oxidase method for determining blood, C.S.F. and urine glucose levels', Clinica Chimica Acta, 251 (1), 19±24. Martens, D.A. and Frankenberger, W.T. (1991) `Determination of saccharides in biological materials by high-performance anion-exchange chromatography with pulsed amperometric detection', Journal of Chromatography, 546, 297±309. Matsumuto, M., Ueba, K. and Kondo, K. (2005) `Separation of sugar by solvent extraction with phenylboronic acid and trioctylmethylammonium chloride', Separation and Purification Technology, 43, 269±274. Matta, K., Kamal-Eldin, E. and Torronen, A.R. (2003) `High-performance liquid chromatography (HPLC) analysis of phenolic compounds in berries with diode array and electrospray ionization mass spectrometric (MS) detection: Ribes species', Journal of Agricultural and Food Chemistry, 51, 6736±6744.

ß Woodhead Publishing Limited, 2010

310

Bioalcohol production

Matuszewski, B.K., Constanzer, M.L., and Chavez-Eng, C. M. (1998) `Matrix effect in quantitative LC/MS/MS analyses of biological fluids: a method for determination of finasteride in human plasma at picogram per milliliter concentrations', Analytical Chemistry, 70, 882±889. Matuszewski, B.K., Constanzer, M.L. and Chavez-Eng, C.M. (2003) `Strategies for the assessment of matrix effect in quantitative bioanalytical methods based on HPLC± MS/MS', Analytical Chemistry, 75, 3019±3030. Mecozzi, M. (2005) `Estimation of total carbohydrate amount in environmental samples by the phenol-sulphuric acid method assisted by multivariate calibration', Chemometrics and Intelligent Laboratory Systems, 79, 84±90. Medeiro, P.M. and Simoneit, B.R.T. (2007) `Analysis of sugars in environmental samples by gas chromatography±mass spectrometry', Journal of Chromatography A, 1141 (2), 271±278. DOI: 10.1016/j.chroma.2006.12.017. Mielenz, J.R. (2001) `Ethanol production from biomass: technology and commercialization status', Ecology and Industrial Microbiology, 4 (3), 324±329. DOI: 10.1021/ac60147a030g. Miller, G.L. (1959) `Use of dinitrosalicylic acid reagent for determination of reducing sugar', Analytical Chemistry, 31 (3), 426±428. Mishima, D., Tateda, M., Ike, M. and Fujita, M. (2006) `Comparative study on chemical pretreatments to accelerate enzymatic hydrolysis of aquatic macrophyte biomassused in water purification processes', Bioresource Technology, 97 (16), 2166±2172. Mitchell, S., Dobson, P., Davies, H.V. and Viola, R. (1998) ``Potential for artifacts in the measurement of fructose and sucrose in extracts of potato tubers using the microplate reader assay', Potato Research, 41 (4), 383±386. DOI: 10.1007/ BF02358970. MolnaÂr-Perl, I. (2000) `Role of chromatography in the analysis of sugars, carboxylic acids and amino acids in food', Journal of Chromatography A, 891 (1), 1±32. DOI: 10.1016/S0021-9673(00)00598-7. MolnaÂr-Perl, I., Pinter-Szakacs, M., Kovago, A. and Petroczy, J. (1984) `Gas±liquid chromatographic determination of the raffinose family of oligosaccharides and their metabolites present in soy beans', Journal of Chromatography A, 295 (2), 433±443. Morales-Cid, G., Simonet, B.M., CaÂrdenas, S. and ValcaÂrcel, M. (2007) `On-capillary sample cleanup method for the electrophoretic determination of carbohydrates in juice samples', Electrophoresis, 28, 1557±1563. Moreno, M.V.G., MejõÂas, R.C., MarõÂn, R.N. and Barroso, C.G. (2002) `Analysis of sugar acids by capillary electrophoresis with indirect UV detection. Application to samples of must and wine', European Food and Research Technology, 215, 255± 259. Mosier, N., Hendrickson, R., Ho, N., Sedlak, M. and Ladisch, M.R. (2005a) `Optimization of pH controlled liquid hot water pretreatment of corn stover', Bioresource Technology, 96 (18), 1986±1993. DOI: 10.1016/j.biortech.2005.01. 013d a. Mosier, N., Wyman, C., Dale, B., Elander, R., Lee, Y.Y., Holtzapple, M. and Ladisch, M. (2005b) `Features of promising technologies for pretreatment of lignocellulosic biomass', Bioresource Technology, 96 (6), 673±686. DOI: 10.1016/ j.biortech.2004.06.025. Nelson, N. (1994) `A photometric adaptation of the somogyi method for the determination of glucose', Journal of Biological Chemistry, 153 (2), 375±380. Nichols, N.N., Sharma, L.N., Mowery, R.A., Chambliss, C.K., van Walsum, G.P., Dien,

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

311

B.S. and Iten, L.B. (2008) `Fungal metabolism of fermentation inhibitors present in corn stover dilute acid hydrolysate', Enzyme Microbial Technology, 42 (7), 624± 630. DOI: 10.1016/j.enzmictec.2008.02.008. Niemela, K. (1988) `GLC-MS Studies on pine kraft black liquors. I: Identification of monomeric compounds', Holzforschung, 42 (3), 169±173. DOI: 10.1016/S00219673(00)94439-X. Niemela, K. (1998) `Gas±liquid chromatography±mass spectrometry studies on pine kraft black liquors: III. The liberation of carboxylic acids in the initial phase of pulping', Journal of Chromatography A, 446, 247±252. Niemela, K. and Sjostrom, E. (1986) `Simultaneous identification of aromatic and aliphatic low molecular weight compounds from alkaline pulping liquor by capillary gas-liquid chromatography-mass spectrometry', Holzforschung, 40 (6), 361±368. Nogueira, L.C., Silva, F., Ferreira, I. M. P. L. V. O. and Trugo, L.C. (2005) `Separation and quantification of beer carbohydrates by high-performance liquid chromatography with evaporative light scattering detection', Journal of Chromatography A, 1065, 207±210. Noll, G.R., Nagle, N.J., Mitchell, D.J., Baker, J.O., Grohmann, K. and Himmel, M.E. (1990) `Evaluation of alkaline earth and transition metals for the use in the ion moderated partition chromatography of sugars', Journal of Liquid Chromatography, 13 (4), 703±714. Oefner, P.J. and Chiesa, C. (1994) `Capillary electrophoresis of carbohydrates', Glycobiology, 4 (4), 397±412. Oefner, P.J., Lanziner, A.H., Bonn, G. and Bobleter, O. (1992) `Quantitative studies on furfural and organic acid formation during hydrothermal, acidic and alkaline degradation of D-Xylose', Monatshefte fur Chemie, 123 (6±7), 547±556. DOI: 10.1007/BF00816848. Okatch, H. and Torto, N. (2003) `Profiling of carbohydrate polymers in biotechnology using microdialysis sampling, high performance anion exchange chromatography with integrated pulsed electrochemical detection/mass spectrometry', African Journal of Biotechnology, 2 (12), 636±644. Olsson, L. and Hahn-HaÈgerdal, B. (1996) `Fermentation of lignocellulosic hydrolysates for ethanol production', Enzyme Microbial Technology, 18 (5), 312±331. DOI: 10.1016/0141-0229(95)00157-3. Palmqvist, E. and Hahn-HaÈgerdal, B. (2000) `Fermentation of lignocellulosic hydrolysates. II: inhibitors and mechanisms of inhibition', Bioresource Technology, 74 (1), 25±33. DOI: 10.1016/S0960-8524(99)00161-3. Pecina, R., Burtscher, P., Bonn, G. and Bobleter, O. (1986) `GC-MS and HPLC analyses of lignin degradation products in biomass hydrolysates', Fresenius Z Anal. Chem., 325 (5), 461±465. DOI: 10.1007/BF00489135. Persson, P., Andersson, J., Gorton, L., Larsson, S., Nilvebrant, N.O. and Jonsson, L.J. (2002) `Effect of different forms of alkali treatment on specific fermentation inhibitors and on the fermentability of lignocellulose hydrolysates for production of fuel ethanol', Journal of Agricultural and Food Chemistry, 50 (19), 5318±5325. DOI: 10.1021/jf025565o. Raemy, A. and Schweizer, T.F. (1983) `Thermal behaviour of carbohydrates studies by heat flow calorimetry', Journal of Thermal Analysis, 28, 95±108. Ramirez, A. J., Mottaleb, M.A., Brooks, B.W. and Chambliss, C.K. (2007) `Analysis of pharmaceuticals in fish using liquid chromatography-tandem mass spectrometry', Analytical Chemistry, 79 (8), 3155±3163.

ß Woodhead Publishing Limited, 2010

312

Bioalcohol production

Richardson, S., Cohen, A. and Gorton, L. (2001) `High-performance anion-exchange chromatography±electrospray mass spectrometry for investigation of the substituent distribution in hydroxypropylated potato amylopectin starch', Journal of Chromatography A, 917 (1±2), 111±121. DOI: 10.1016/S0021-9673(01)00690-2. Rivasseau, C., Boisson, A.-M., Mongelard, G., Couram, G., Bastien, O. and Bligny, R. (2006) `Rapid analysis of organic acids in plant extracts by capillary electrophoresis with indirect UV detection Directed metabolic analyses during metal stress', Journal of Chromatography A, 1129, 283±290. Rogatsky, E., Jayatillake, H., Goswami, G., Tomuta, V. and Stein, D. (2005) `Sensitive LC MS quantitative analysis of carbohydrates by Cs+ attachment', Journal of the American Society for Mass Spectrometry, 16 (11), 1805±1811. DOI: 10.1016/ j.jasms.2005.07.017. Rumbold, K., Okatch, H., Torto, N., Siika-Aho, M., GuÈbitz, G., Robra, K.H. and Prior, B. (2002) `Monitoring on-line desalted lignocellulosic hydrolysates by microdialysis sampling micro-high performance anion exchange chromatography with integrated pulsed electrochemical detection/mass spectrometry', Biotechnology and Bioengineering, 78 (7), 822±828. DOI: 10.1002/bit.10264. Saha, B.C., Iten, L.B., Cotta, A. and Wu, Y.V. (2005) `Dilute acid pretreatment, enzymatic saccharification and fermentation of wheat straw to ethanol', Process Biochemestry, 40 (12), 3693±3700. DOI: 10.1016/j.procbio.2005.04.006. Saifer, A. and Gerstenfeld, S. (1958) `The photometric microdetermination of blood glucose with glucose oxidase', Journal of Laboratory and Clinical Medicine, 51 (3), 448±460. Sanz, M.L. and Martinez-Castro, I. (2007) `Recent developments in sample preparation for chromatographic analysis of carbohydrates', Journal of Chromatography A, 1153 (1±2), 74±89. Sawardeker, J.S., Sloneker, J.H. and Jeanes, A. (1965) `Quantitative determination of monosaccharides as their alditol acetates by gas liquid chromatography', Analytical Chemistry, 37 (12), 1602±1604. Schiller, M., von der Heydt, H., Marz, F. and Schmidt, P.C. (2002) `Quantification of sugars and organic acids in hygroscopic pharmaceutical herbal dry extracts', Journal of Chromatography A, 968, 101±111. Schwald, W., Chan, M., Breuil, C. and Saddler, J.N. (1988) `Comparison of HPLC and colorimetric methods for measuring cellulolytic activity', Applied Microbiololy and Biotechnology, 28 (4±5), 398±403. DOI: 10.1007/BF00268203. Sevcik, R.S., Mowery, R.A., Becker, C. and Chambliss, C.K. (2010) `Rapid high performance liquid chromatography analysis of sugars in aqueous extracts and dilute-acid hydrolysates of herbaceous biomass', Analytical Chemistry, Submitted. Sharma, L.N., Becker, C. and Chambliss, C.K. (2009) `Analytical characterization of fermentation inhibitors in biomass pretreatment samples using liquid chromatography UV±visible spectroscopy, and tandem mass spectrometry', in Methods in Molecular Biology, ed. J.R. Mielenz, Humana Press, Clifton, NJ, 125± 143. DOI: 10.1016/j.clinbiochem.2004.11.007. Shui, G. and Leong, P. (2002) `Separation and determination of organic acids and phenolic compounds in fruit juices and drinks by high-performance liquid chromatography', Journal of Chromatography A, 977 (1), 89±96. Skomarovsky, A.A., Markov, A.V., Gusakov, A.V., Kondrat'eva, E.G., Okunev, O.N., Bekkarevich, A.O., Matys, V.Y. and Sinitsyn, A.P. (2006) `New cellulases efficiently hydrolyzing lignocellulose pulp', Applied Biochemistry and Microbiology, 42 (6), 592±597. DOI: 10.1134/S0003683806060093.

ß Woodhead Publishing Limited, 2010

Analytical monitoring of pretreatment and hydrolysis processes

313

Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D. and Crocker, D. (2008) `Determination of Structural Carbohydrates and Lignin in Biomass', National Renewable Energy Laboratory: Golden, CO, http://www.nrel.gov/ biomass/pdfs/42618.pdf (accessed March 2009). Smith, R.M. (2003) `Before the injection ± modern methods of sample preparation for separation techniques', Journal of Chromatography A, 1000, 3±27. Soga, T. and Serwe, M. (2000) `Determination of carbohydrates in food samples by capillary electrophoresis with indirect UV detection', Food Chemistry, 69, 339± 344. Somogyi, M. (1945) `Determination of blood sugar', Journal of Biological Chemistry, 160 (1), 69±73. Somogyi, M. (1952) `Notes on sugar determination', Journal of Biological Chemistry, 70 (3), 19±23. Spackman, V.M.T. and Cobb, A.H. (2002) `An enzyme-based method for the rapid determination of sucrose, glucose and fructose in sugar beet roots and the effects of impact damage and postharvest storage in clamps', Journal of the Science of Food and Agriculture, 82 (1), 80±86. DOI: 10.1002/jsfa.1005. Spindler, D., Wyman, C. and Grohmann, K. (1990) `Evaluation of pretreated herbaceous crops for the simultaneous saccharification and fermentation process', Applied Microbiology and Biotechnology, 24±25 (1), 275±286. Starke, I., Holzberger, A., Kamm, B. and Kleinpeter, E. (2000) `Qualitative and quantitative analysis of carbohydrates in green juices (wild mix grass and alfalfa) from a green biorefinery by gas chromatography/mass spectrometry', Fresenius Z Anal Chem, 367 (1), 65±72. Steinberg, P. and Fox, A. (1999), `Automated Derivatization Instrument: Preparation of alditol acetates for analysis of bacterial carbohydrates using gas chromatography/mass spectrometry', Analytical Chemistry, 71 (9), 1914±1917. DOI: 10.1021/ac981155g. Sun, Y. and Cheng, J. (2002) `Hydrolysis of lignocellulosic materials for ethanol production: a review', Bioresource Technology, 83 (1), 1±11. DOI: 10.1016/S09608524(01)00212-7. Suzuki, S. and Honda, S. (1998), `A tabulated review of capillary electrophoresis of carbohydrates', Electrophoresis, 19, 2539±2560. Sweeley, C.C., Bentley, R., Makita, M. and Wells, W.W. (1963) `Gas-liquid chromatography of trimethylsilyl derivatives of sugars and related substances', Journal of the American Chemical Society, 85 (16), 2497±2507. Szczodrak, J. and Fiedurek, J. (1996) `Technology for conversion of lignocellulosic biomass to ethanol', Biomass Bioenergy, 10 (5±6), 367±375. Taherzadeh, M.J. and Karimi, K. (2008) `Pretreatment of lignocellulosic wastes to improve ethanol and biogas production: a review', International Journal of Molecular Sciences, 9, 1621±1651. Takeuchi, Y., Jin, F., Enomoto, H. and Tohji, K. (2007) `Application of acid catalytic Hydrothermal reaction to Conversion of carbohydrate biomass into valuable substances', Water Dynamics: 4th International workshop on water dynamics, 190± 192. Tang, L. and Kebarle, P. (1993) `Dependence of ion intensity in electrospray mass spectrometry on the concentration of the analytes in the electrosprayed solution', Analytical Chemistry, 65, 3654±3668. Tarkosova, J. and Kopikova, J. (2000) `Determination of carbohydrate content in bananas during ripening and storage by near infrared spectroscopy', Journal of Near Infrared Spectroscopy, 8, 21±26.

ß Woodhead Publishing Limited, 2010

314

Bioalcohol production

Taylor, P.J. (2005) `Matrix effects: the Achilles heel of quantitative high-performance liquid chromatography±electrospray±tandem mass spectrometry', Clinical Biochemistry, 38, 328±334. Tolstikov, V.V. and Fiehn, O. (2002) `Analysis of highly polar compounds of plant origin: combination of hydrophilic interaction chromatography and electrospray ion trap mass spectrometry', Analytical Biochemistry, 301, 298±307. Vanderford, B.J. and Snyder, S.A. (2006) `Analysis of pharmaceuticals in water by isotope dilution liquid chromatography/tandem mass spectrometry', Environmental Science and Technology, 40 (23), 7312±7320. Verzele, M., Simoens, G. and Damme, F.V. (1987) `A critical review of some liquid chromatography systems for the separation of sugars', Chromatographia, 23 (4), 292±300. DOI: 10.1007/BF02311783. Viola, R. and Davies, H.V. (1992) `A microplate reader assay for rapid enzymatic quantification of sugars in potato tubers', Potato Research, 35 (1), 55±58. DOI: 10.1007/BF02357723. Wei, Y. and Ding, M. (2000) `Analysis of carbohydrates in drinks by high-performance liquid chromatography with a dynamically modified amino column and evaporative light scattering detection', Journal of Chromatography A, 904 (1), 113±117. DOI: 10.1016/S0021-9673(00)00909-2. WillfoÈr, S., Pranovich, A., Tamminen, T., Puls, J., Laine, C., SuurnaÈkki, A., Saake, B., Uotila, K., Simolin, H., Hemming, J. and Holmbom, B. (2009) `Carbohydrate analysis of plant materials with uronic acid-containing polysaccharides ± a comparison between different hydrolysis and subsequent chromatographic analytical techniques', Industrial Crops and Products, 29 (2±3), 571±580. Wright, P.J. and Wallis, A.F.A. (1996) `Rapid determination of carbohydrates in hardwoods by high performance anion exchange', Holzforschung, 50 (6), 518±524. Wyman, C.E., Dale, B.E., Elander, R.T., Holtzapple, M., Ladisch, M.R. and Lee, Y.Y. (2005) `Coordinated development of leading biomass pretreatment technologies', Bioresource Technology, 96 (18), 1959±1966. DOI: 10.1016/ j.biortech.2005.01.010. Yourchisin, D.M. and van Walsum, G.P. (2004), `Comparison of microbial inhibition and enzymatic hydrolysis rates of liquid and solid fractions produced from pretreatment of biomass with carbonic acid and liquid hot water', Applied Biochemistry and Biotechnology, 115 (1±3), 1073±1086. DOI: 10.1385/ABAB:115:1-3:1073. Yu, Y., Lou, X. and Wu, H. (2008) `Some recent advances in hydrolysis of biomass in hot-compressed water and its comparisons with other hydrolysis methods', Energy and Fuels, 22 (1), 46±60. Zaldivar, J., Nielsen, J. and Olsson, L. (2001) `Fuel ethanol production from lignocellulose: a challenge for metabolic engineering and process integration', Applied Microbiology and Biotechnology, 56 (1±2), 17±34. DOI: 10.1007/ s002530100624. Zook, C.M. and LaCours, W.R. (1995) `Pulsed amperometric detection of carbohydrates in fruit juices following high performance anion exchange chromatography', Current Separations, 14 (2), 48±52.

ß Woodhead Publishing Limited, 2010

12

Online monitoring of fermentation processes in lignocelluloses-to-bioalcohol production A . E L I A S S O N L A N T Z and K . V . G E R N A E Y , Technical University of Denmark, Denmark and C . J . F R A N Z EÂ N and L . O L S S O N , Chalmers University of Technology, Sweden

Abstract: Online monitoring of bioethanol production is challenging due to the complex, viscous sample matrix, also containing solids. Suitable targets for online measurement are identified and potential monitoring techniques in addition to standard physicochemical measurements are reviewed. Advantages and limitations of chromatographic techniques, spectroscopic methods and software sensors are discussed in detail. In conclusion, bioethanol production processes should certainly benefit from online measurement of key process variables, especially if measurements are incorporated in process control loops. However, convincing case studies are lacking, and therefore development of suitable sampling methods and online measurement techniques should first be prioritized. Key words: bioethanol, chromatography, fermentation, fluorescence, nearinfrared, sampling, software sensor, spectroscopy.

12.1

Introduction

Fermentation processes have traditionally been rather poorly monitored, leaving opportunities to improve the processes by increased monitoring and control. For a long time fermentation monitoring has been dominated by online measurement of physical variables (such as pressure, stirrer speed, weight) and some physicochemical variables (such as pH, redox potential), whereas more process specific variables such as substrate and product concentrations are seldom monitored (Sonnleitner, 1999). A challenge with monitoring and control of fermentation processes is that for many important process variables there are very few or no sensors that are suitable for online measurements. For example, for measuring the concentration and activity of the cell mass no well-established and wellaccepted standard method is available, even though a number of options are available on the market (Kiviharju et al., 2008). In later years a renewed interest for online monitoring of fermentation processes has emerged. This trend has largely been motivated by the changed legislation for approval of processes by FDA and corresponding bodies (http://

ß Woodhead Publishing Limited, 2010

316

Bioalcohol production

www.fda.gov/cder/guidance/6419fnl.pdf). The change in legislation has brought about a new view on quality control in pharmaceutical processes. Traditionally, ensuring the quality of the product was based on extensive off-line analysis in the lab, whereas currently it is encouraged to ensure the end product quality on the basis of in-process measurements, or in other words on the basis of online measurement of the process performance. This has opened up for the need and the development of online monitoring of fermentation processes. Even though the interest in pharmaceutical industry is a major driving force, the progress made in that industry is transferring much more broadly into the fermentation industry (Junker and Wang, 2006), where the potential economical benefit of increased process understanding combined with more tightly controlled processes is an important driving force. Nevertheless, there are still several major challenges to be solved in order to introduce online monitoring widely in industrial fermentation processes. Bioethanol production is an example of a fermentation process that belongs to the group of `low-value ± large-volume' processes. For such processes to be economically viable and competitive, the production cost should be kept to a minimum, which means that the product yield and productivity must constantly be optimized. The process streams in bioethanol processes are both chemically and physically complex. It is therefore especially challenging to adapt online measurement techniques to this demanding environment. Typical bioethanol process flowsheets and the different unit operations are covered in detail in other chapters of this book. Here, a short process overview will be given in the perspective of monitoring needs. Today bioethanol is produced from sucrose or starch-based crops whereas the future major production is expected to be from lignocellulosic material (a polymeric material composed of the major polymers cellulose, hemicelluloses and lignin). Many possible process schemes may be depicted, but a simplified and general process scheme applying to both types of feedstock can be found in Fig. 12.1. As a first process step the feedstock (raw material) will be pretreated, such that it can be used as a substrate stream in the subsequent process steps. When starch is the raw material, this process step will lead to a partial degradation of the starch to a maltodextrin rich stream. When using lignocellulosic material as the raw material, the pretreatment will primarily lead to solubilization of the hemicelluloses and partial opening of the cellulose structure as well. Thereafter the ethanol production will proceed either through a separate hydrolysis and fermentation process route (SHF) (A) or through a simultaneous saccharification and fermentation process (SSF) (B). In the SHF process the hydrolysis can take place using either chemicals or enzymes, whereas in the SSF process, enzymes will be used. Enzymes may be produced on-site, or they may be purchased from an enzyme producer. Ethanol is the final product, but the use of side-streams such as carbon dioxide, animal feed, biogas, fertilizers, energy, and steam contribute significantly to the overall process economy.

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

317

12.1 General bioethanol process scheme.

During the pretreatment step, the feedstock is subject to high temperature and/or high pressure, possibly under acidic or alkaline conditions. Under such conditions the sugar polymer of starch and hemicelluloses will be hydrolysed into mono- and oligosaccharides, and in addition a number of degradation products will be formed. These degradation products encompass sugar degradation products (furfurals and acids), lignin degradation products (aromatic acids, alcohols and aldehydes) as well as components released from the plant wall structure (such as acids and extractives). As illustrated in Fig. 12.1, the liquid part of a typical lignocellulosic hydrolysate is brownish and opaque. The complete lignocellulosic hydrolysate culture medium is a slurry of remaining solid material (cellulose and part of the lignin) suspended in the liquid hydrolysate. A gas chromatography (GC) analysis of a lignocellulosic hydrolysate has revealed 40 peaks that could be identified, in addition to a number of unidentified peaks, illustrating the large complexity of its composition (Buchert et al., 1990). As the sample matrix of the process streams are complex not only from a chemical but also a physical point of view, online analysis is a very challenging task. Not only are the streams complex in nature, it may also be expected that the content varies as a function of time. During a fermentation process the substrate undergoes constant changes, as the saccharides are consumed by the microorganisms and additionally some of the other components may be converted, e.g. furfurals are converted to the corresponding alcohols. Furthermore, the feedstock fed to a given process will undergo changes as a consequence of

ß Woodhead Publishing Limited, 2010

318

Bioalcohol production

feedstock transport and storage, harvest time (season) and the growth conditions of the plants. Bioethanol production is a large-scale production process and the conditions prevailing in the production facilities are such that the applied monitoring technologies need to be robust to ensure reliable measurements. It is not that difficult to acquire a lot of online data. However, the main challenge is to ensure that the sensors provide reliable data, and subsequently to make sure that relevant information is extracted from the available data. Analysis of historical data can be very useful, but is in practice not so often done because it is typically very time consuming to retrieve the online data from databases. Moreover, stored data are often useless because of the data compression algorithms that are applied; often data compression does not allow the reconstruction of the dynamics of the process when data are retrieved again. It is also hard to unambiguously find cause and effect chains from historical data, since they only show correlations. Other problems that are encountered in practice are: how to combine data from different sources (e.g. synchronization)?; how to cope with outliers and other faulty measurements?; how to differentiate between detection of a process fault on the one hand and normal variation, e.g. due to variation in feedstock, on the other hand? Several of the measurement techniques that presently receive growing interest are non-invasive spectroscopic techniques. Such techniques rely on the treatment of the collected spectral information with multivariate data analysis methods to extract process relevant information. As these spectra contain a wealth of information, the data treatment is one of the major challenges in establishing reliable online monitoring and control based on spectroscopic techniques.

12.2

Variables of interest to monitor in bioethanol production processes

The core variables to monitor in fermentation processes are the substrate consumption, product formation and the cell mass concentration and its activity. Considering that the cells are the biocatalyst in the fermentation, monitoring of the amount of cells and its activity is of utmost importance. Even though continued efforts have been made, cell mass concentrations and activities are still very challenging to measure, as, for example, demonstrated in a recent review by Kiviharju et al. (2008). Due to the complex sample matrix of lignocellulosic hydrolysates/slurries, classical methods such as optical density (OD) measurements and gravimetric determination can not be used. Even though it is hard to monitor some of the core variables directly, better information extraction from standard fermentation variables such as pH, temperature, agitation, feed rate, base addition rate, nutrient addition rate, etc., may allow prediction of cell mass concentration and other important variables by means of appropriate software sensors (see Section 12.6 for more details).

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

319

Taking the process steps outlined in Fig. 12.1 as a basis we can summarize the variables that ideally should be evaluated and controlled during lignocellulosic bioethanol production in addition to those already discussed. The feedstock should be characterized in terms of content and composition of cellulose, hemicelluloses, extractives and lignin, to enable application of optimal pretreatment conditions. After pretreatment, the presence of inhibitors should be evaluated online as these are fundamental for the performance of both the enzymes as well as the microorganisms in subsequent process steps. In addition, a general insight into the composition of the fermentation substrate feed stream would for example facilitate more optimal media design and would be helpful also in deciding on fermentation conditions for enhanced performance, for example nutrient addition. All the fermentation processes involved, be it SSF, SHF or the enzyme production step, must be monitored and controlled such that a reliable production is ensured. This also includes monitoring nutrient concentrations, and byproduct formation. A lot of information may be extracted indirectly from physical measure-ments (as discussed below in Section 12.6 on software sensors). Some additional variables should be monitored during enzyme production, e.g. measurement of specific activities and concentrations of the produced enzymes. Moreover, side streams and byproducts (e.g., carbon dioxide, biogas, animal feed, fertilizers, lignin) should also be evaluated in terms of quantity, composition and quality, since they make a major contribution to the overall process economy. In the future, online (or at-line) monitoring of an increasing number of these variables will provide real-time information which can directly be used to decide on the most suitable conditions for the next step in the process. In the case of lignocellulose-based bioethanol production, even slight changes in the feedstock, pretreatment and/or hydrolysis conditions may cause large differences in the composition of the substrate stream and specifically increased formation of inhibitory or even toxic compounds may result in the medium. Due to this large variability, applying the same conditions every time will lead to low yields and productivities, or even a complete process breakdown in the case of increased toxicity, unless the toxicity can be predicted and corrective actions taken, based on reliable monitoring techniques.

12.3

Sampling issues and overview of potential monitoring techniques

Before any reliable online measurement can be established, a number of sampling issues have to be solved. In this section, typical sampling issues will be introduced briefly, and problems related to specific online monitoring techniques will be highlighted. A concise overview of potential applications of different online monitoring techniques to a bioethanol production process is provided in Table 12.1.

ß Woodhead Publishing Limited, 2010

Table 12.1 Concise overview of different online measurement methods that can be applied to a bioethanol production process Method

What can be analyzed?

Influence of process conditions

Sampling

Others

ß Woodhead Publishing Limited, 2010

Substrates Biomass

Ethanol

Byproducts

LC

Yes

No

Yes

Yes

Particles and viscosity

Needed, filtering

Well-known method

GC

No

No

Yes

Some

Not influenced if sample taken from gas phase

Needed, from gas phase directly or filtering from liquid phase

Well-known method

Fluorescence spectroscopy

Some

Yes

No

Some

Particles, air bubbles

No

Multivariate analysis needed

NIR spectroscopy

Yes

Yes

Yes

Yes

Particles, air bubbles

No

Multivariate analysis needed

Software sensors

Yes

Yes

Yes

Yes

Quality of available data will influence quality of predicitions

No

Need for detailed process knowledge and/or advanced algorithms

Online monitoring of fermentation processes

321

In order to obtain reliable online measurements, one first of all has to be careful when selecting the point where the online sensor is installed or sample is withdrawn. Usually, the reactor content has to be well-mixed at the location of the measurement point, to ensure that the collected data are representative for the phenomena taking place in the reactor. Moreover, care should be taken that the time constant of the online measurement is significantly lower than the important time constants of the process to be monitored. Biological reactions are characterized by relatively large time constants and therefore this is not a major problem. Time delays introduced by an online measurement that has a long response time (e.g., due to a long sample preparation time) can considerably reduce the performance of a control loop using that online measurement as an input. Last but not least, it is important to ensure that the sampling approach/ applied measurement technique is sufficiently robust to generate reliable data in a harsh process environment without too much need for maintenance. Despite the obvious online monitoring potential of chromatography (see also Section 12.4), the online application of gas and liquid chromatography to fermentation monitoring is still uncommon. The major issue is the sampling, which must be representative for the whole cultivation and robust enough to ensure problem-free operation. Online HPLC relies on sampling by ultra- or microfiltration. Applications in the early 1990s of hollow-fiber and flat ultrafiltration units in connection with online liquid chromatography have been reviewed elsewhere (van de Merbel, 1997; SchuÈgerl, 2001). Van de Merbel (1997) concluded that when membrane characteristics, ultrafiltration devices, and the sample flow rate are optimized, many bioprocesses using complex industrial media, including waste from the pulping industry, can be monitored successfully. In small, ideally mixed bioreactors, an in situ filter located inside the bioreactor is preferred since it does not disturb the mixing characteristics of the fermentation (FranzeÂn, 2003; Schultze, 1995). External cross-flow filtration units have the dual benefits of offering a possibility for periodic cleaning or replacement of the filter, and of cell recycling to the main reactor (Brandberg et al., 2007; FranzeÂn et al., 1994). Cell recycle by filtration is probably not reasonable as a single option in a large-scale process based on lignocellulose hydrolysate, but after primary separation of cells and lignocellulose debris by centrifugation the product stream may be easier to filter, ensuring a particle-free sample for column injection. HPLC has been applied for online monitoring of, e.g., Bacillus subtilis cultures (Filippini et al., 1991), antibiotic production (SchuÈgerl, 1988; SchuÈgerl and Seidel, 1998), and recombinant protein production (SchuÈgerl et al., 1993; Turner et al., 1994). Removal of proteins before analysis has been shown to be imperative for robust operation in industrial media (Scheller et al., 1994a, b). Sample withdrawal from the fermentation broth/substrate in bioethanol production may be complicated due to the fact that the sample matrix generally contains particles and is relatively viscous, hence there is a risk for clogging and/

ß Woodhead Publishing Limited, 2010

322

Bioalcohol production

or fouling of filter units. Since inline spectroscopic methods are based on interaction of light with the fermentation broth (see also Section 12.5), the measurement may for most spectroscopic techniques be performed through a glass window using standard ports in the fermenter vessel, thereby providing an advantage over, e.g., chromatographic methods, when it comes to sampling a thick, heterogeneous broth.

12.4

Chromatographic techniques

Chromatographic techniques offer quantitative multicomponent analysis of most classes of chemical compounds, with high precision, accuracy, specificity and sensitivity. When using spectroscopic detectors it is even possible to obtain positive identification of unknown analytes. Moreover, chromatography is a mature technology. Thus, there exist chromatography-based protocols for the determination of both organic and inorganic components of lignocellulosic material, hydrolysates, and fermentation broths. Nevertheless, chromatography is still under constant development and there is still room for improvement and extension of existing protocols. A thorough description of both the theory and many applications of chromatography can be found in Heftmann (2004a, b). A wide variety of available column geometries, packing materials, eluents, chromatography systems and detectors allows the optimization of separation, detection, and quantification of specific compounds, even in complex mixtures. Ideally, all compounds under scrutiny should be retained on the column and eluted at different retention times, while all uninformative compounds should either not be retained at all on the column, or be strongly adsorbed, but eluted during regeneration of the column. Gas chromatography In the field of fermentation monitoring, gas chromatography has traditionally been used off-line to determine the concentration of volatile products in liquid broth samples (e.g. Weuster-Botz et al., 1993). Online gas chromatography has, e.g., been applied to acetone-butanol fermentations both using head-space sampling (Comberbach et al., 1985) as well as sampling and analysis of the liquid broth (McLaughlin et al., 1985). In the latter case, an ultrafiltration membrane was used to obtain a cell-free liquid sample to be analyzed by an online gas chromatograph. Ethanol and volatile byproducts of yeast metabolism have been determined by online gas chromatography during ethanol production from beet molasses by means of a gas-permeable PTFE membrane probe, inserted directly into the reactor (Groboillot et al., 1989). Furthermore, online gas chromatography has been applied to quantify ethanol and the fermentation byproducts acetoin, acetaldehyde, and iso-propanol during continuous ethanol

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

323

fermentation from hydrolyzed B-starch with immobilized Zymomonas mobilis (Diamantis et al., 2006). To our knowledge, the number of online applications of gas chromatography in bioethanol production from lignocellulose hydrolysates is very low. Liquid chromatography Gas chromatography is limited by the requirement that the analyte must be volatile, either in its natural state or after chemical derivatization. Liquid chromatography, on the other hand, can be used to quantify non-volatile components of fermentation broths, like different sugars, directly from the liquid solution. In addition, many (semi-)volatile fermentation products can also be separated and detected using liquid chromatography. Two HPLC protocols have become more or less standard analytical procedure for fermentation samples. One is the application of Bio-Rad Aminex HPX87H ion exclusion/cation exchange column using 5 mM H2SO4 as eluent. Since the hydrophobic non-functionalized resin also interacts with neutral analytes, the column can be used to separate several sugars, acids and alcohols. The other is the Dionex ion chromatography column CarboPac, which allows for separation of mixtures of different mono- and disaccharides and sugar alcohols. Both of these systems have been used extensively for quantitative off-line analysis of fermentation processes (e.g. Zaldivar et al., 2002, de Jongh et al., 2008). The Aminex HPX87H column, in combination with a refractive index detector, has also been used for online monitoring of lab-scale ethanol fermentation (FranzeÂn et al., 1994; FranzeÂn, 2003). A slightly different column, the BioRad Aminex HPX-87C, with Ca2+ as leaving ion, has been applied in an online monitoring and control system for fermentation processes, including automatic sampling by filtration, dilution, and injection on the column for analysis of glucose and ethanol (Liu et al., 2001). The system was applied to Zymomonas mobilis fermentations, and was used to generate input data for a controller of the feed flow rate that kept the glucose concentration at a preset profile. The Pb2+ column HPX87P from BioRad has been used for online monitoring of carbohydrates in a small-scale fermentation of spent sulfite liquor, using a refractive index detector (Buttler et al., 1993). In that investigation the fermentation broth was filtered using a Waters Durapore filter and tangential flow filter/acquisition module. Sample cleanup was performed online using a precolumn filled with a mixture of anion and cation exchange resins. In parallel to carbohydrate analysis on the HPX87P column, ethanol was also determined by injecting part of the automatically sampled broth on a reversed phase PLRP-S column followed by detection using an ethanol biosensor (Buttler et al., 1993). Buttler et al. (1994) used microdialysis for sampling a fermentation of spent sulfite liquor. Ethanol was determined using a fast chromatographic separation followed by enzymatic

ß Woodhead Publishing Limited, 2010

324

Bioalcohol production

detection (alcohol biosensor). This system allowed determination of ethanol in only 5 minutes. Rumbold et al. (2002) used microdialysis sampling and microhigh performance anion exchange chromatography with integrated pulsed electrochemical detection to monitor the enzymatic hydrolysis of Eucalyptus grandis. Method development is continuing into so-called ultra-high performance liquid chromatography (UHPLC), using ever higher pressures and smaller particle sizes, thus shortening analysis times and improving efficiency (CastroPerez et al., 2004; Xiayan and Legido-Quigley, 2008). These improvements may be very useful in online applications.

12.5

Spectroscopic methods

In addition to being non-invasive, spectroscopic methods are attractive for bioprocess monitoring and control since they have the potential to provide continuous measurements of multiple analytes. The spectroscopic methods that have received most attention with respect to bioprocess monitoring are different forms of infrared (IR) spectroscopy and fluorescence spectroscopy (Marose et al., 1999; StaÈrk et al., 2002; Scarff et al., 2006; Cervera et al., 2009). Raman spectroscopy is used sporadically. Basic principles of infrared and multi-wavelength fluorescence spectroscopy Infrared (IR) spectroscopy is based on the fact that the covalent bonds in organic molecules respond to specific frequencies with rotation or vibration. Below a short description of the principles for IR spectroscopy is given, more information can be found, for example, in Infrared Spectroscopy for Food Quality Analysis and Control (Sun, 2008) and Near-Infrared Technology in the Agricultural and Food Industries (Williams and Norris, 1987). When molecules are irradiated, they can absorb photons that carry energy coinciding with the characteristic vibrations of the molecule. Absorption of this energy results in excitation of the molecule to a higher energy level. Different types of chemical bonds respond to different frequencies, and IR absorption at certain frequencies may thus be related to specific chemical groups. Most organic molecules may be seen in the IR regions, since C±H, N±H, O±H, CN, CO and CC bonds will give IR absorption. IR spectroscopy has traditionally been divided into near-, mid- and far-IR. The far-IR is the region with the lowest energy and is used for rotational spectroscopy. In the mid-IR (MIR) region (~2500±25 000 nm) the fundamental vibrations are seen whereas, in the higher energy near-IR (NIR) region (~800±2500 nm) overtones and combination bands of fundamental absorptions are observed. Rather few molecules manifest overtones, and consequently, the overtone bands are generally much weaker in intensity than the fundamental absorption bands. Combination bands occur when

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

325

two or more vibrations are excited simultaneously. Water has characteristic absorption bands in IR that are very strong and thereby limit the use of this technology for online monitoring of fermentation processes. Due to the weaker absorptions in NIR, this technique is more suited for fermentation applications. The light is selectively absorbed when passing through the sample and the remaining part of the light may be partially transmitted and reflected. The amount of absorbed light cannot be measured directly; instead it has to be calculated from either the transmitted or the reflected light. In a fermentation process the morphology and the size distribution of the cells influence how much light is reflected. Fluorescence spectroscopy was first introduced for measuring culture fluorescence by Duysens and Amesz (1957). This first fluorescence monitoring was performed with single wavelength instruments. Multi-wavelength fluorescence, with multiple excitation and emission wavelength combinations, for inline measurements of cultivations has not been used for much more than about 15 years (Tartakovsky et al., 1996; Marose et al., 1998). These studies showed that an increased amount of useful information can be extracted when monitoring cultivations with multi-wavelength fluorescence compared to monitoring a single excitation/emission wavelength combination. Fluorescence is caused by the absorption of a photon by a molecule producing an electronically excited molecule that later will emit light while returning to its ground state. It is the delocalized electrons in a molecule that enable it to be excited and exhibit fluorescence, e.g. valence electrons in covalent bonds in highly conjugated systems, lone pair electrons or electrons in aromatic rings. First UV-VIS light is absorbed, which causes an electron to be excited from the singlet ground state to a vibrational state of an excited singlet state. Next the electron will fall to the ground state of the lowest excitable state. The electron can then undergo different processes. Fluorescence is the result of the return to a vibrational ground state by emission of a photon. The emitted photon always has a longer wavelength (= lower energy) than the absorbed photon. For more information about fluorescence consult Valeur (2002) and Lakowics (2006). There are several important organic fluorophores, such as the three aromatic amino acids tryptophan, phenylalanine and tyrosine. Tryptophan has a more rigid structure and one extra ring compared to phenylalanine and tyrosine, and consequently most fluorescence from proteins comes from tryptophan. Other important biological compounds with fluorescence are some vitamins, riboflavin, flavin mononucleotide (FMN), flavin adenine dinucleotide (FAD), and the reduced forms of the pyridine nucleotides NADH and NADPH. Data analysis As stated in the introduction, spectroscopic techniques result in multivariate data and require some sort of advanced data treatment for interpretation. Every

ß Woodhead Publishing Limited, 2010

326

Bioalcohol production

spectrum that is collected consists of about 100 to 1000 values. As a consequence, the data matrixes that are obtained from online spectroscopy can become quite large. The developments in computer science and software have largely facilitated the implementation of spectroscopy based process monitoring. The two chemometric methods principal component analysis (PCA) (Wold et al., 1987) and partial least squares (PLS) (Geladi and Kowalski, 1986; Wold et al., 2001) regression are commonly applied together with spectroscopy. PCA decomposes data into a structure part and a noise part and reduces the number of variables, which are linear combinations of the original variables. These new variables are the so-called principal components (PCs) and they are selected so that the first PC covers as much of the variation in the data as possible, the second PC is orthogonal to the first one and covers as much of the remaining variation as possible, and so forth. By applying PCA large multi-dimensional collinear data sets, typically obtained when applying online monitoring to bioprocesses, can often be reduced to a limited number (e.g. 3 or 4) of dimensions without significant loss of information. PLS is a multivariate regression method that provides a mathematical relationship between the online data and off-line data, and can be used for future prediction of Y data from measurements of X. Also for PLS an important attribute is reduction in number of variables. The basis for the algorithm is to maximize the covariance between a model of X and a model of Y. Examples of NIR and fluorescence applications in fermentation processes Raw material composition and quality Different types of agricultural and forest waste material usually serve as feedstock for bioethanol production. These raw materials consist of very complex polymer-like molecules, and their composition may vary with season and also depend on the pretreatment applied before fermentation. Online determination of feedstock composition would to a large extent facilitate control of the process, where it would be obvious to apply such online information in a feedforward controller to improve the performance of, e.g., pre-treatment and fermentation process steps. NIR-based methods have been applied to determine the composition of a number of different plant- and wood-based feedstocks. In one study a total of 121 different samples from eight various feedstock sources were used to build a PLS model (Sanderson et al., 1996). The samples were ground before scanning with NIR reflectance. Good correlations (above 0.9) were seen for extractives, lignin, arabinose, xylose, glucose and nitrogen. Ye et al. (2008) applied FT-NIR to determine variation of chemical composition in corn stover fractions. The developed PLS model yielded root mean square error of prediction (RMSEP %w/w) values of 0.92, 1.03, 0.17, 0.27, 0.21 and 1.12 for glucan, xylan, galactan, arabinan, mannan and lignin, respectively. In both the above referred

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

327

investigations, the original feedstock, before any pretreatment or hydrolysis, was the material to which NIR was applied. It would be interesting to also monitor the feedstock composition after pretreatment in order to better design, operate and control the fermentation step, e.g. determine the optimal feed rate in a fedbatch fermentation process or decide media design. It is likely that NIR could be applied to determine the sugar composition, even though the treated feedstock probably provides a more complex matrix than the untreated, due to formation of degradation products, etc. Furthermore, determination of the content of different phenolic compounds and other degradation products that may be inhibitory to the microorganisms when present in too high concentrations, may enable even further possibilities to achieve a more optimal process by adopting process conditions based on feedstock composition. The combination of IR and fluorescence spectroscopy is expected to allow detection of several of these compounds. Monitoring of substrates and ethanol Carbohydrates and ethanol do not possess fluorescence and hence, the spectroscopy based method of choice for determination of these variables is mainly IR absorption. Several studies have investigated the use of NIR or other IR spectroscopy methods to predict glucose, ethanol and ammonium concentrations in fermentations and reported chemometric models with low prediction errors. However, the majority of these studies have been performed off-line and on clear liquid samples. There are only few online applications reported in the literature (Tamburini et al., 2003; Veale et al., 2007; Petersen et al., 2010). Even less can be found about monitoring of bioethanol processes with `real' substrates, and in these studies samples where usually withdrawn and centrifuged, before applying an off-line measurement (Rathore et al., 2007; Liebmann et al., 2009). To the best of our knowledge, the application of IR spectroscopy in situ in a bioethanol process with `real', viscous and particle-containing medium has thus far not been demonstrated. However, models that have been developed for off-line prediction of substrate and ethanol are promising. FT-NIR was applied on filtered samples from a corn dry grind fermentation (Rathore et al., 2007). Both transflectance and transmission spectra were applied to predict various fermentation quality indicators, e.g. ethanol, total soluble sugars, fermentable sugars, glucose, maltose and maltotriose. Reliable PLS models with low RMSEP were obtained for total soluble sugars and ethanol, whereas the other variables could not be predicted well. The observed RMSEP values were 1.5% v/v (range 0±13.1%) and 1.5% w/v (range 0.7±27.5%) for ethanol and total sugars, respectively. Liebmann et al. (2009) used a NIR transflectance probe to measure ethanol in stillage and glucose in the mash from filtrated wheat and rye mash samples. PLS models with a standard deviation of prediction errors of 2% of the mean ethanol concentration (range 0±56 g/L) and 9.6% of the mean glucose concentration (range 0±55 g/L) were obtained.

ß Woodhead Publishing Limited, 2010

328

Bioalcohol production

Ethanol and substrate concentrations have also been successfully predicted on the basis of multi-wavelength fluorescence spectra (Skibsted et al., 2001; Solle È dman et al., et al., 2003; Hagedorn et al., 2003, 2004; Surribas et al., 2006; O 2009). However, as neither ethanol nor glucose or nitrate are fluorescent, the resulting chemometric models are indirect and based on correlations, most probably to cell mass. This is fine in principle, but one should bear in mind that the correctness of the models relies on the assumption that the correlations between the analytes remain constant. This may not be the case in a fermentation, for example if a change in the feedstock composition leads to a change in the cell mass and product yields or cell morphology. Monitoring of cell mass NIR spectroscopy has recently been applied to in situ process monitoring of cell mass concentration (Arnold et al., 2002; Tamburini et al., 2003; Veale et al., 2007; Petersen et al., 2010). The maximum cell mass concentrations were typically below 20 g/L, and PLS models resulted in predictions with an error of 5±10%. Cell mass concentration was monitored over a wider concentration range, 1±60 g/L, in fed-batch fermentations of Saccharomyces cerevisiae (Ge et al., 1994) by applying short-wavelength NIR. The measurements were performed through the glass walls of the vessel using a bifurcated fiber-optic bundle. It was observed that when the concentration of cell mass increased, the relative amount of backscattered light reaching the detector also increased, causing a decrease in apparent absorbance (Ge et al., 1994). For transmission probes increased scattering will lead to an increase in the overall absorption (Crowley et al., 2005; Finn et al., 2006). The second derivative of the spectra is typically used as input to the chemometric model instead of the raw spectra to avoid inaccurate predictions due to variations in scattering. Hereby baseline shifts are removed, and this pretreatment can furthermore also assist in resolving spectral overlaps and will thus allow to show absorbance peaks more clearly (Vaidyanathan et al., 2001a, b; Finn et al., 2006). Multi-wavelength fluorescence has been applied to determine the cell mass concentration for a number of different microorganisms, e.g. Saccharomyces cerevisiae and Escherichia coli (Tartakovsky et al., 1996; Marose et al., 1998; Haack et al., 2004; Rhee and Kang, 2007). Cell mass concentration estimation errors were around 10% for the constructed PLS models. The fluorescent spectra of S. cerevisiae cell mass could be decomposed into peaks corresponding to those of NAD(P)H, tryptophan and riboflavin (Haack et al., 2004). In addition to monitoring the cell mass concentration, fluorescence has also been employed to determine the metabolic state of cell mass (Lindemann et al., 1998; Johansson and LideÂn, 2006). A drawback with fluorescence is that the medium may have high background fluorescence, hence disturbing the measurement and increasing the detection limit. This is, for example, the case for molasses (unpublished results).

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

329

There are also other methods commonly used for cell mass determination, such as optical density and dielectric spectroscopy. Optical density is typically not applicable in bioethanol processes due to the particle-containing media used. Most dielectric spectroscopy measurements are based on the change in capacitance, which is proportional to the total cell volume. The media composition may also influence the capacitance measurement. The effect can to some extent be circumvented by measurement at different frequencies (Neves et al., 2000). Dielectric spectroscopy is frequently used in the brewing industry and has been proven useful (KronloÈf et al., 1991). A distinct difference between dielectric spectroscopy and the other methods mentioned here is that only viable cell mass results in a signal, whereas the other methods register total cell mass. The cell membrane has to be intact for the cells to possess capacitance. As lignocellulosic hydroltsates may act inhibitory on the cells, a measure on viable cells may prove to be valuable. A more novel method for cell mass determination is photo-acoustic spectroscopy (Lindgren and Hamp, 2006; Schmid, 2006). The methodology is based on absorption of electromagnetic radiation. When sound waves with a certain frequency range are sent through a fermentation broth, the microbial cells will absorb part of the radiation. The cell mass concentration as well as the morphology and viscosity of the broth influence the signal. It is an interesting sensor as the cost for purchasing the equipment is several orders of magnitudes lower than, for example, NIR and multi-wavelength fluorescence sensors. Raman spectroscopy Raman spectroscopy is not so often used for online monitoring of fermentation processes. Picard et al. (2007) applied Raman spectroscopy for in situ monitoring of ethanol in a concentration range of 0±0.4 mol/L during an alcoholic fermentation with Saccharomyces cerevisiae. Mendes et al. (2003) demonstrated very convincingly that Raman spectroscopy could be used to determine ethanol content in ethanol fuel samples (84.9±100% ethanol) and in beverages (0.0±100.0% ethanol). Clearly, this indicates that Raman spectroscopy could be useful for product formation monitoring in bioethanol production. However, it should be added explicitly here that the work of Mendes et al. (2003) was performed on a relatively simple sample matrix. For bioethanol production processes the sample matrix is considerably more complex, and practically this hurdle might create difficulties when trying to establish an online ethanol measurement on the basis of Raman spectroscopy. Challenges in implementation Probes for in situ usage must be able to withstand steam sterilization, may not be too sensitive to fouling and must have long-term stability. IR based spectroscopy

ß Woodhead Publishing Limited, 2010

330

Bioalcohol production

probes fulfill these demands. Online fluorescence measurements typically take place through a glass window on the fermenter, and hence they do not need to be steam-sterilizable. Both types of probes are promising sensors for online applications. Thick, particle containing feedstocks present a challenge. In an off-line setting, filtration or centrifugation is possible in order to remove particles and hence remove scattering effects from the spectra, whereas particle removal is not possible when performing measurements online. However, the multivariate data treatment methods can to some extent cope with scattering. Moreover, if one is interested in monitoring cell mass concentration, filtering is not an option as it also removes the microbial cells. Aeration and stirring also pose a challenge as air bubbles may disturb the spectroscopy measurements. Yet, for bioethanol processes this is most likely not a major problem as aeration and stirring intensities usually are quite low. Another limitation with online NIR applications is that the spectral region above 2100 nm, where the strongest cell mass signal is found, is too noisy due to the optical fibres, and hence this part of the spectrum can not be used (Petersen et al., 2010). However, by optimization of optical path length and number of scans the weaker signals at lower wavelengths can be improved (Arnold et al., 2002), and thus online NIR monitoring of cell mass concentration is indeed feasible. Last but not least, acceptance of multivariate data analysis methods is also an important issue. There might be a general skepticism towards this type of data compared to standard analytical data which are not multivariate. The availability of a set of convincing examples in the form of successful case studies, clear communication between chemometricians and the operators and analysts, and illustrative user-friendly software that elegantly converts a complex set of online spectroscopic data into useful information will all help to bridge that gap.

12.6

Software sensors

The estimation of the concentration of analytes of interest using so-called software sensors is in many cases a fruitful alternative to direct (or analytespecific) measurements using for example chromatographic and spectroscopic methods This is, for example, illustrated in a recent review on methods that allow the online measurement of the cell mass concentration (Kiviharju et al., 2008), where software sensors are included as an alternative approach that can compete with methods such as dielectric spectroscopy, OD, IR spectroscopy and fluorescence for in situ measurement of the cell mass concentration. The central idea behind a software sensor is thus to use (relatively) easily accessible online data for the estimation of other cultivation variables that are otherwise either difficult to measure or only measured at low frequency. The dissolved oxygen tension in the liquid phase, the gas flow rates, the off-gas

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

331

oxygen and carbon dioxide concentrations, as well as feeding and titration rates are the variables that are the most frequently used online data that serve as an input to software sensors (SundstroÈm and Enfors, 2008). The advantage of using software sensors in a process is three-fold: first of all, no extra hardware cost is involved since already available data are exploited; secondly, the estimates might serve as a data input to control algorithms that could not be established without the software sensor, thus resulting in improved system operation; thirdly, the estimates will usually contribute to obtaining an improved process understanding. The software sensors can be divided in three classes: (1) software sensors based on stoichiometry, elementary balances and first-principles models; (2) software sensors based on data-driven modelling methods, i.e. black-box approaches; (3) hybrid software sensors, which are partly based on firstprinciples, but supplemented with black-box approaches for parts of the system that are not sufficiently well understood. This third class of software sensors will not be discussed here. The first class of software sensors exploits the available process knowledge on the stoichiometric relations between the variables in a fermentation process. Application of stoichiometric relations and elementary balances in software sensors is illustrated by Stephanopoulos and San (1984). Relatively simple software sensors can be obtained by expressing the variable to be predicted as a function of known constants/parameters and available online measurements. The respiratory quotient (RQ), the ratio between the CO2 production rate and the O2 consumption rate in aerobic fermentations, is one of the most well-known software sensor examples (SundstroÈm and Enfors, 2008). RQ can be used for detection of the start of oxidoreductive growth at conditions where the degree of reduction of the substrate is significantly different from the degree of reduction of the product. This is the case for ethanol production from glucose by Saccharomyces cerevisiae, where the RQ shows values above one when ethanol is formed (Jobe et al., 2003). The CO2 evolution rate, obtained via gas analysis, can also be used as a measure for metabolic actvity, and has been used to control the feed rate in fed-batch fermentation of inhibitory lignocellulose hydrolysate (Taherzadeh et al., 2000, Nilsson et al., 2001). Titration data, i.e. cumulative addition profiles of acid and base, collected during the fermentation, also form a very useful source of information that can be exploited in a software sensor, for example for the calculation of the cell mass concentration and the specific growth rate of cells (San and Stephanopoulos, 1984; SundstroÈm and Enfors, 2008). Interestingly, SundstroÈm and Enfors (2008) have proposed the use of a software sensor for estimation of a variable representing the oxygen uptake per energy substrate (RO/S). This variable was suggested to represent the extent of physiological stress, and it was demonstrated for fed-batch Escherichia coli fermentations how this variable could be used for online detection of stress induced by sudden pH shocks. Similar software

ß Woodhead Publishing Limited, 2010

332

Bioalcohol production

sensors could undoubtedly also be useful in biofuel production processes for online detection of process faults. The performance of simple software sensors outlined in the previous paragraphs, where the variable to be predicted is expressed as a mathematical function of known constants/parameters and available online measurements, can be poor due to measurement noise and uncertainties about the process (Stephanopoulos and San, 1984). The estimation results can, however, be improved significantly by applying estimation algorithms such as the Kalman filter (for linear systems) and the extended Kalman filter (for nonlinear systems) (Stephanopoulos and San, 1984). In practice, a number of advanced estimation methods are available. For details on receding-horizon observers, asymptotic observers and hybrid observers, the review paper by Bogaerts and van de Wouwer (2003) is a good starting point. Bastin and Dochain (1990) should be consulted for a more extended introduction to estimation algorithms. Note that applying such algorithms will of course result in more complicated software sensors, which unfortunately might limit the use of this type of algorithms in a production environment. The second class of software sensors exploits correlations between the variables in the fermentation process, without seeking any mechanistic explanation for the observed correlations. Methods such as artifical neural networks (ANN) and chemometric modelling techniques, for example partial least squares (PLS), belong here. Online estimation of the cell mass concentration with ANN has been demonstrated by VaneÏk et al. (2004) for fed-batch yeast cultivations, È dman et al. and by AcunÄa et al. (1998) for lactic acid bacteria batch cultures. O (2009) successfully applied PLS to standard bioreactor data obtained from a Saccharomyces cerevisiae batch fermentation for the prediction of cell mass, substrate and ethanol concentration. Montague et al. (2008) recently demonstrated the use of PLS models on industrial process data, where multi-way PLS was used to predict the end point of a batch process, in this case a beer fermentation. In a very recent publication, a moving window principal component analysis (MW-PCA) method was successfully used to identify phase changes in several different industrially relevant batch processes (Maiti et al., 2009). The MW-PCA method was solely based on changes in the statistical properties of online data, such as pH, dissolved oxygen concentration, agitator speed and concentrations of CO2 and O2 in the exhaust gas, and appears to be useful for any case where even slight changes in process properties must be identified. The main message is that black-box methods are certainly useful, on condition that the user is aware of the limitations of the methods. Moreover, some of these methods have the advantage that they can be applied rather easily in practice. Software for building chemometric models is, for example, available from several software vendors or can be downloaded as freeware. Since bioethanol production, especially second-generation processes, is still a developing industry, little is published specifically on the use of software sensors on such

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

333

processes. The bioethanol industry should therefore learn from the experiences with software sensors that are available in other branches of the fermentation industry. However, a few warnings have to be issued also. Industrial bioprocess data are noisy and are heterogeneous with respect to time scale, i.e. not all variables are recorded with the same sampling interval (Charaniya et al., 2008; Montague et al., 2008). They are also heterogeneous with respect to data types (Charaniya et al., 2008): many variables are continuous (e.g., pH, dissolved oxygen concentration), others (e.g., valve settings) are discrete or binary (onoff). Thus, substantial data pre-processing might be necessary, including missing value replacement, outlier removal, interpolation and re-sampling, before one can start testing software sensors. Montague et al. (2008) indicate, for example, that off-line data pre-treatment can take up to 75% of the time in a proof-ofconcept project. In an online application all this needs to be performed automatically, which is considerably more challenging. Results obtained by SundstroÈm and Enfors (2008) were, for example, obtained off-line, following forward and reverse filtering of historical batch data. In an online (production) setting, forward filtering will introduce delays that can for example influence performance of controllers. Reverse data filtering typically requires that a process is finished before the data are filtered, and is thus not really applicable in a production environment.

12.7

Conclusions

The major benefit of using online monitoring techniques, in contrast to off-line lab analysis of the process, is the possibility of coupling online measurements to process control. The major benefits of automated process control are to be expected in dynamic processes, where the quality of the feedstock varies, or where the process operation may suffer from unforeseen disturbances, or in processes which are inherently nonlinear. All these criteria apply to bioethanol production from both starch- and lignocellulosic-based feedstocks, i.e. both forestry and argricultural waste. Conditions during batch or fed-batch operation are inherently dynamic, especially during the SSF process. Variation in the source and type of raw material causes varying quality of the hydrolysate, in terms of sugar contents and inhibitor concentration. The response to some inhibitors is nonlinear. For example, furfural and HMF can be converted to their less inhibitory alcohols by the yeast, as long as the concentration is low enough, leading to improved fermentation performance (Taherzadeh et al., 2000). However, at higher concentrations the fermentation performance deteriorates dramatically. Online monitoring combined with feedback and feedforward control could make the difference between a working process and a failure. In conclusion, the benefit of online monitoring of feedstock quality, hydrolysis, fermentation performance and biological activity should be very high in largescale production of bioethanol. However, there is still a lack of convincing case

ß Woodhead Publishing Limited, 2010

334

Bioalcohol production

studies where the full potential of online monitoring has been exploited on a larger-scale bioethanol process by inserting the online measurement system in a suitable control loop. Nevertheless, for successful application, development must continue to proceed in terms of: · prediction of medium toxicity based on feedstock characterization and analysis of the hydrolysis process, · rapid detection of inhibition of metabolic activity, i.e. a measure of the cell mass activity, · reliable algorithms for large-scale bioreactor control for enzyme and bioethanol production, or in short, research must proceed in terms of robust sampling procedures, analytical methods, and data analysis and integration. Achieving successful implementation of an increasing number of online monitoring methods necessitates that industry is convinced that these methods bring along a benefit and can be implemented in a reliable way. The benefit, that should be a more efficient process operation, can either be in the form of improved yield, or shorter processing times (lower capital investment cost), or more consistent product quality.

12.8

References

AcunÄa G., Latrille E., BeÂal C., Corrieu G. (1998), Static and dynamic neural network models for estimating biomass concentration during thermophilic lactic acid bacteria batch cultures, J. Ferment. Bioeng., 85, 615±622. doi: 10.1016/S0922338X(98)80015-9 Arnold S.A., Gaensakoo R., Harvey L.M., McNeil B. (2002), Use of at-line and in-situ near-infrared spectroscopy to monitor biomass in an industrial fed-batch Escherichia coli process, Biotechnol. Bioeng., 80, 405±413. doi: 10.1002/bit.10383 Bastin G., Dochain D. (1990), On-line Estimation and Adaptive Control of Bioreactors, Elsevier, Amsterdam. Bogaerts P., van de Wouwer A. (2003), Software sensors for bioprocesses, ISA Trans., 42, 547±558. doi: 10.1016/S0019-0578(07)60005-6 Brandberg T., Karimi K., Taherzadeh M.J., FranzeÂn C.J., Gustafsson L. (2007), Continuous fermentation of wheat-supplemented lignocellulose hydrolysate with different types of cell retention, Biotechnol. Bioeng., 98, 80±90. doi: 10.1002/bit.21410 Buchert J., NiemelaÈ K., Puls J., Poutanen K. (1990), Improvement in the fermentability of steamed hemicellulose hydrolysate by ion exclusion, Process Biochem., 25, 176± 180. Buttler T., Johansson K.A.J., Gorton L.G.O., Marko-Varga G.A. (1993), On-line fermentation process monitoring of carbohydrates and ethanol using tangential-flow filtration and column liquid chromatography, Anal. Chem., 65, 2628±2636. doi: 10.1021/ac00067a014 Buttler T., Gorton, L., Jarskog, H., Marko-Varga, G., Hahn-HaÈgerdal, B., Meinander, N., Olsson, L. (1994), Monitoring of ethanol during fermentation of a lignocellulose

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

335

hydrolysate by on-line microdialysis sampling, column liquid chromatography and an alcohol biosensor. Biotechnology and Bioengineering, 44, 322±328. Castro-Perez J., Plumb R., Granger J. (2004), A new paradigm for metabolism studies: UPLC/Q-Tof', Application note. Waters Corporation, Milford, USA. Accessed 7 December 2008. http://www.waters.co.jp/application/product/uplc/pdf/ 720000953en.pdf. Cervera A.E., Petersen N., Eliasson Lantz A., Larsen A., Gernaey K.V. (2009), Application of near-infrared spectroscopy for monitoring and control of cell culture and fermentation, Biotechnol. Prog., 25, 1561±1581. doi: 10.1021/bp.280 Charaniya S., Hu W., Karypis G. (2008), Mining bioprocess data: opportunities and challenges, Trends Biotechnol., 26, 690±699. doi: 10.1016/j.tibtech.2008.09.003 Comberbach D.M., Scharer J.M., Moo-Young M. (1985), Application of the headspace gas chromatographic technique for continuous monitoring of the acetone-butanolethanol fermentation, Enzyme Microb. Technol., 7, 543±548. doi: 10.1016/01410229(85)90096-1 Crowley J., Arnold S.A., Wood N., Harvey L.M., McNeil B. (2005), Monitoring a high cell density recombinant Pichia pastoris fed-batch bioprocess using transmission and reflectance near infrared spectroscopy, Enzyme Microb. Technol., 36, 621±628. doi: 10.1016/j.enzmictec.2003.12.016 de Jongh W.A., Bro C., Ostergaard S., Regenberg B., Olsson L., Nielsen J. (2008), The roles of galactitol, galactose-1-phosphate, and phosphoglucomutase in galactoseinduced toxicity in Saccharomyces cerevisiae, Biotechnol. Bioeng., 101, 317±326. doi: 10.1002/bit.21890 Diamantis V., Melidis P., Aivasidis A. (2006), Continuous determination of volatile products in anaerobic fermenters by on-line capillary gas analysis, Anal. Chim. Acta, 573±574, 189±194. doi: 10.1016/j.aca.2006.05.036 Duysens L.N.M., Amesz J. (1957), Fluorescence spectrometry of reduced phosphopyridine nucleotide in intact cells in the nearultraviolet and visible region, Biochim. Biophys. Acta, 24, 19±26. doi: 10.1016/0006-3002(57)90141-5 Filippini C., Moser J.U., Sonnleitner B., Fiechter A. (1991), On-line capillary gas chromatography with automated liquid sampling, a powerful tool in biotechnology, Anal. Chim. Acta., 255, 91±96. doi:10.1016/0003-2670(91)85091-6 Finn B., Harvey L.M., McNeil B. (2006), Near-infrared spectroscopic monitoring of biomass, glucose, ethanol and protein content in a high cell density baker's yeast fed-batch bioprocess, Yeast, 23, 507±517. doi: 10.1002/yea.1371 FranzeÂn C.J. (2003), Metabolic flux analysis of RQ-controlled microaerobic ethanol production by Saccharomyces cerevisiae, Yeast, 20, 117±132. doi: 10.1002/yea.956 FranzeÂn C.J., LideÂn G., Niklasson C. (1994), An experimental guide to the relevant aeration rates in microaerobic bioprocesses, In: Galindo E., RamõÂrez O.T. (eds), Advances in Bioprocess Engineering. Amsterdam, Kluwer Academic Publishers, pp. 255±265. Ge Z., Cavinato A.G., Callis J.B. (1994), Noninvasive spectroscopy for monitoring cell density in a fermentation process, Anal. Chem., 66, 1354±1362. doi: 10.1021/ ac00080a023 Geladi P., Kowalski B.R. (1986), Partial least-squares regression: A tutorial, Anal. Chim. Acta., 185, 1±17. doi: 10.1016/0003-2670(86)80028-9 Groboillot A., Pons M.N., Engasser J. (1989), Monitoring of volatiles in alcoholic fermentations on molasses via a gas membrane sensor, Appl. Microbiol. Biotechnol., 32, 37±44. doi: 10.1007/BF00164820 Haack M.B., Eliasson A., Olsson L. (2004), On-line cell mass monitoring of

ß Woodhead Publishing Limited, 2010

336

Bioalcohol production

Saccharomyces cerevisiae cultivations by multi-wavelength fluorescence, J. Biotechnol., 114, 199±208. doi: 10.1016/j.jbiotec.2004.05.009 Hagedorn A., Legge R.L., Budman H. (2003), Evaluation of spectrofluorometry as a tool for estimation in fed-batch fermentations, Biotechnol. Bioeng., 83, 104±111. doi: 10.1002/bit.10649 Hagedorn A., Levadoux W., Groleau D., Tartakovsky B. (2004), Evaluation of multiwavelength culture fluorescence for monitoring the aroma compound 4hydroxy-2(or 5)-ethyl-5(or 2)-methyl-3(2H)-furanone (HEMF) production, Biotechnol. Prog., 20, 361±367. doi: 10.1021/bp0300321 Heftmann E. (ed.) (2004a), Chromatography 6th edition: fundamentals and applications of chromatography and related differential migration methods. Part A: Fundamentals and techniques. Amsterdam, Elsevier BV. Heftmann E. (ed.) (2004b), Chromatography 6th edition: fundamentals and applications of chromatography and related differential migration methods. Part B: Applications. Amsterdam, Elsevier BV. Jobe A.M., Herwig C., Surzyn M., Walker B., Marison I., von Stockar U. (2003), Generally applicable fed-batch culture concept based on the detection of metabolic state by on-line balancing, Biotechnol. Bioeng., 82, 627±639. doi: 10.1002/ bit.10610 Johansson L., LideÂn G. (2006), A study of long-term effects on plasmid-containing Escherichia coli in carbon-limited chemostat using 2D-fluorescence spectrofluorimetry', Biotechnol. Prog., 22, 1132±1139. doi: 10.1021/bp060061m Junker B.H., Wang H.Y. (2006), Bioprocess monitoring and computer control: Key roots of the current PAT initiative, Biotechnol. Bioeng., 95, 226±261. doi: 10.1002/ bit.21087 Kiviharju K., Salonen K., Moilanen U., EerikaÈinen T. (2008), Biomass measurement online: the performance of in situ measurements and software sensors, J. Ind. Microbiol. Biotechnol., 35, 657±665. doi: 10.1007/s10295-008-0346-5 KronloÈf J. (1991), Evaluation of a capacitance probe for determination of viable yeast biomass, In: Proceedings of the 23rd European Brewery Convention, pp. 233±239. Lakowicz, J. R. (2006), Principles of Fluorescence Spectroscopy, Springer Science+Business media, New York. Liebmann B., Friedl A., Varmuza K. (2009), Determination of glucose and ethanol in bioethanol production by near infrared spectroscopy and chemometrics, Anal. Chim. Acta, 642, 171±178. doi: 10.1016/j.aca.2008.10.069 Lindemann C., Marose S., Nielsen H.O., Scheper T. (1998), 2-Dimensional fluorescence spectroscopy for on-line bioprocess monitoring, Sens. Actuators B, 51, 273±277. doi: 10.1016/S0925-4005(98)00195-6 Lindgren T., Hamp S. (2006), Biomass monitoring using acoustic spectroscopy, IEEE Sensors J., 6, 1068±1075. doi: 10.1109/JSEN.2006.881434 Liu Y.C., Wang F.S., Lee W.C. (2001), On-line monitoring and controlling system for fermentation processes, Biochem. Eng. J., 7, 17±25. doi: 10.1016/S1369703X(00)00100-5 Maiti S.K., Srivastava R.K., Bhushan M., Wangikar P.P. (2009), Real time phase detection based online monitoring of batch fermentation processes, Process Biochem., 44, 799±811. doi: 10.1016/j.procbio.2009.03.008 Marose S., Lindemann C., Scheper T. (1998), Two-dimensional fluorescence spectroscopy: a new tool for on-line bioprocess monitoring, Biotechnol. Prog., 14, 63±74. doi: 10.1021/bp970124o Marose S., Lindemann C., Ulber R., Scheper T. (1999), Optical sensors systems for

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

337

bioprocess monitoring, Trends Biotechnol., 17, 30±34. doi: 10.1016/S01677799(98)01247-5 McLaughlin J.K., Meyer C.L., Papoutsakis E.T. (1985), Gas chromatography and gateway sensors for on-line state estimation of complex fermentations (butanolacetone fermentation), Biotechnol. Bioeng., 27, 1246±1257. doi: 10.1002/ bit.260270821 Mendes L.S., Oliveira F.C.C., Suarez P.A.Z. and Rubim J.C. (2003), Determination of ethanol in fuel ethanol and beverages by Fourier transform (FT)-near infrared and FT-Raman spectrometries, Anal. Chim. Acta, 493, 219±231. doi: 10.1016/S00032670(03)00870-5 Montague G.A., Martin E.B., O'Malley C.J. (2008), Forecasting for fermentation operational decision making, Biotechnol. Prog., 24, 1033±1041. doi: 10.1021/bp.29 Neves A.A., Pereira D.A., Vieira L.M., Menezes J.C. (2000), Real time monitoring biomass concentration in Streptomyces clavuligerus cultivations with industrial media using a capacitance probe, J. Biotechnol., 84, 45±52. doi: 10.1016/S01681656(00)00325-4 Nilsson A., Taherzadeh M.J., LideÂn G. (2001), Use of dynamic step response for control of fed-batch conversion of lignocellulosic hydrolyzates to ethanol, J. Biotechnol., 89, 41±53. doi: 10.1016/S0168-1656(01)00283-8 È dman P., Lindvald Johansen C., Olsson L., Gernaey K.V., Eliasson Lantz A. (2009), O On-line estimation of biomass, glucose and ethanol in Saccharomyces cerevisiae cultivations using in-situ multiwavelength fluorescence and software sensors, J. Biotechnol., 144, 102±112. doi: 10.1016/j.jbiotec.2009.08.018 È dman P., Cervera Padrell A.E., Stocks S., Eliasson Lantz A., Gernaey K.V. Petersen N., O (2010), In-situ NIR spectroscopy for analyte specific monitoring of glucose and ammonium in Streptomyces coelicolor fermentations', Biotechnol. Prog., 26, 263± 271. doi: 10.1021/bp.288 Picard A., Daniel I., Montagnac G. and Oger P. (2007), In situ monitoring by quantitative Raman spectroscopy of alcoholic fermentation by Saccharomyces cerevisiae under high pressure, Extremophiles, 11, 445±452. doi: 10.1007/s00792-006-0054-x Rathore S.S.S., Paulsen M.R., Sharma V., Singh V. (2007), Use of near-infrared spectroscopy for monitoring fermentation in a corn dry grind ethanol process, Transactions ASABE, 50, 2337±2344. Rhee J.I., Kang T.H. (2007), On-line process monitoring and chemometric modeling with 2D fluorescence spectra obtained in recombinant E coli fermentations, Process Biochem., 42, 1124±1134. doi: 10.1016/j.procbio.2007.05.007 Rumbold K., Okatch H., Torto N., Siika-Aho M., GuÈbitz G., Robra K.H., Prior B. (2002), Monitoring on-line desalted lignocellulosic hydrolysates by microdialysis sampling micro-high performance anion exchange chromatography with integrated pulsed electrochemical detection/mass spectrometry, Biotechnol. Bioeng., 78, 822±828. doi: 10.1002/bit.10264 San K., Stephanopoulos G. (1984), Studies on on-line bioreactor identification. IV. Utilization of pH measurements for product estimation, Biotechnol. Bioeng., 26, 1209±1218. doi: 10.1002/bit.260261006 Sanderson M.A., Agblevor F., Collins M., Johnson D.K. (1996), Compositional analysis of biomass feedstocks by near infrared reflectance spectroscopy, Biomass Bioenergy, 11, 365±370. doi: 10.1016/S0961-9534(96)00039-6 Scarff M., Arnold S.A., Harvey L.M., McNeil B. (2006), Near infrared spectroscopy for bioprocess monitoring and control: current status and future trends, Crit. Rev. Biotechnol., 26, 17±39. doi: 10.1080/07388550500513677

ß Woodhead Publishing Limited, 2010

338

Bioalcohol production

Scheller U., Siedenberg D., SchuÈgerl K. (1994a), Ion chromatography for monitoring biotechnological processes. Part I. System development for analysis of industrial cultivation media of microorganisms, Anal. Chim. Acta, 285, 169±176. doi: 10.1016/0003-2670(94)85020-8 Scheller U., Siedenberg D., HuÈbner U., Siebold M., Kretzmer G., SchuÈgerl K. (1994b), Ion chromatography for monitoring biotechnological processes. Part II. Applications for industrial cultivation media, Anal. Chim. Acta, 286, 179±187. doi: 10.1016/0003-2670(94)80158-4 Schmid T. (2006), Photoacoustic spectroscopy for process analysis, Anal. Bioanal. Chem., 384, 1071±1086. doi: 10.1007/s00216-005-3281-6 SchuÈgerl K. (1988), On-line analysis and control of production of antibiotics, Anal. Chim. Acta, 213, 1±9. doi: 10.1016/S0003-2670(00)81334-3 SchuÈgerl K. (2001), Progress in monitoring, modeling and control of bioprocesses during the last 20 years, J. Biotechnol., 85, 149±173. doi: 10.1016/S0168-1656(00)00361-8 SchuÈgerl K., Seidel G. (1998), Monitoring of the concentration of -lactam antibiotics and their precursors in complex cultivation media by high-performance liquid chromatography, J. Chromatogr. A, 812, 179±189. doi: 10.1016/S00219673(98)00264-7 SchuÈgerl K, Brandes L, Wu X, Bode J, Ree JI, Brandt J, Hitzmann B (1993). Monitoring and control of recombinant protein production, Anal. Chim. Acta, 279, 3±16. doi: 10.1016/0003-2670(93)85061-N. Schulze, U. (1995). Anaerobic physiology of Saccharomyces cerevisiae. PhD thesis, Dept. Biotechnology, Technical University of Denmark, Lyngby, Denmark, pp. 77± 82. ISBN 87-88584-14-3. Skibsted E., Lindemann C., Roca C., Olsson L. (2001), On-line bioprocess monitoring with a multi-wavelength fluorescence sensor using multivariate calibration, J. Biotechnol., 88, 47±57. doi: 10.1016/S0168-1656(01)00257-7 Solle D., Geissler D., Stark E., Scheper T., Hitzmann B. (2003), Chemometric modelling based on 2D-fluorescence spectra without a calibration measurement, Bioinformatics, 19, 173±177. Sonnleitner B. (1999), Instrumentation of small-scale bioreactors, Manual of Industrial Microbiology and Biotechnology, 2nd edition, ed. by Demain et al. American Society for Microbiology, pp. 221±235 StaÈrk E., Hitzmann B., SchuÈgerl K., Scheper T., Fuchs C., Koster D., Markl H. (2002), Insitu-fluorescence-probes: a useful tool for non-invasive bioprocess monitoring, Adv. Biochem. Eng. Biotechnol., 74, 21±38. doi: 10.1007/3-540-45736-4_2 Stephanopoulos G., San K. (1984), Studies on on-line bioreactor identification. I. Theory, Biotechnol. Bioeng., 26, 1176±1188. doi: 10.1002/bit.260261006 Sun D.-W. (Ed) (2008), Infrared Spectroscopy for Food Quality Analysis and Control, Academic Press, London. SundstroÈm H, Enfors S-O (2008), Software sensors for fermentation processes, Bioprocess Biosyst. Eng., 31, 145±152. doi: 10.1007/s00449-007-0157-5 Surribas A., Geissler D., Gierse A., Scheper T., Hitzmann B., Montesinos J.L., Valero F. (2006), State variables monitoring by in situ multi-wavelength fluorescence spectroscopy in heterologous protein production by Pichia pastoris, J. Biotechnol., 124, 412±419. doi: 10.1016/j.jbiotec.2006.01.002 Taherzadeh M.J., Eklund R., Gustafsson L., Niklasson C., LideÂn G. (1997), Characterization and fermentation of dilute-acid hydrolyzates from wood, Ind. Eng. Chem. Res., 36, 4659±4665. doi: 10.1021/ie9700831 Taherzadeh M.J., Niklasson C., LideÂn G. (2000), On-line control of fed-batch

ß Woodhead Publishing Limited, 2010

Online monitoring of fermentation processes

339

fermentation of dilute-acid hydrolyzates' Biotechnol. Bioeng., 69, 330±338. doi: 10.1002/1097-0290(20000805)69:3<330::AID-BIT11>3.0.CO;2-Q Tamburini E., Vaccari G., Tosi S., Trilli A. (2003), Near-infrared spectroscopy: a tool for monitoring submerged fermentation processes using an immersion optical-fiber probe, Appl. Spectrosc., 57, 132±138. doi: 10.1366/000370203321535024 Tartakovsky B., Sheintuch M., Hilmer J.-M., Scheper T. (1996), Application of scanning fluorometry for monitoring of a fermentation process, Biotechnol. Prog., 12, 126± 131. doi: 10.1021/bp950045h Turner C., Gregory M.E., Thornhill N.F. (1994), Closed-loop control of fed-batch cultures of recombinant Escherichia coli using on-line HPLC, Biotechnol. Bioeng., 44, 819±829. doi: 10.1002/bit.260440707 Vaidyanathan S., Harvey L.M., McNeil B. (2001a), Deconvolution of near-infrared spectral information for monitoring mycelial biomass and other key analytes in a submerged fungal bioprocess, Anal. Chim. Acta, 428, 41±59. doi: 10.1016/S00032670(00)01205-8 Vaidyanathan S., Macaloney G., Harvey L.M., McNeil B. (2001b), Assessment of the structure and predictive ability of models developed for monitoring key analytes in a submerged fungal bioprocess using near-infrared spectroscopy, Appl. Spectrosc., 55, 444±453. doi: 10.1366/0003702011951957 Valeur B. (2002), Molecular fluorescence: principles and applications, Wiley-VCH Verlag, Weinheim. van de Merbel N.C. (1997), The use of ultrafiltration and column liquid chromatography for on-line fermentation monitoring, Trends Anal. Chem., 16, 162±173. doi: 10.1016/S0165-9936(97)00008-3 VaneÏk M., HrncÏirÏÂõk P., VovsõÂk J., NaÂhlõÂk J. (2004), On-line estimation of biomass concentration using a neural network and information about metabolic state, Bioprocess Biosyst. Eng., 27, 9±15. doi: 10.1007/s00449-004-0371-3 Veale E.L., Irudayaraj J., Demirci A. (2007) An on-line approach to monitor ethanol fermentation using FTIR spectroscopy, Biotechnol. Prog., 23, 494±500. doi: 10.1021/bp060306v Weuster-Botz D., Aivasidis A., Wandrey C, (1993), Continuous ethanol production by Zymomonas mobilis in a fluidized bed reactor. Part II: Process development for the fermentation of hydrolysed B-starch without sterilisation, Appl. Microbiol. Biotechnol., 39, 685±690. doi: 10.1007/BF00164450 Williams P., Norris, K. (Eds.), (1987), Near-Infrared Technology in the Agricultural and Food Industries, American Association of Cereal Chemists, Inc. Wold S., Esbensen K., Geladi P. (1987), Principal component analysis ± a tutorial, Chem. Int. Lab. Sys., 2, 37±52. doi: 10.1016/0169-7439(87)80084-9 Wold S., SjoÈstroÈm M., Eriksson L. (2001), PLS-regression: a basic tool of chemometrics, Chem. Int. Lab. Sys., 58, 109±130. doi: 10.1016/S0169-7439(01)00155-1 Xiayan L., Legido-Quigley C. (2008), Advances in separation science applied to metabonomics, Electrophoresis, 29, 3724±3736. doi: 10.1002/elps.200700851 Ye X.P., Liu L., Hayes D., Womac A., Hong K., Sokhansanj S. (2008), Fast classification and compositional analysis of cornstover fractions using Fourier transform nearinfrared techniques, Bioresource Technol., 99, 7323±7332. doi: 10.1016/ j.biortech.2007.12.063 Zaldivar J., Borges A., Johansson B., Smits H.P., Villas-BoÃas S.G., Nielsen J., Olsson L. (2002), Fermentation performance and intracellular metabolite patterns in laboratory and industrial xylose-fermenting Saccharomyces cerevisiae, Appl. Microbiol. Biotechnol., 59, 436±442. doi: 10.1007/s00253-002-1056-y

ß Woodhead Publishing Limited, 2010

13

Modelling hydrolysis and fermentation processes in lignocelluloses-to-bioalcohol production T . T S O U T S O S , Technical University of Crete, Greece

Abstract: The lignocellulosic complex is the most abundant biopolymer on the planet. To date, many lignocellulosic materials have been tested for bioethanol production (crop residues, wood, herbaceous biomass, agroindustrial residues and cellulosic wastes). The pretreatment of raw material, and the detoxification of the hydrolysates to be fermented, are critical for the economic and technology viability of the process. This chapter reviews: the main lignocellulose-to-bioethanol technology pathways; the critical issues of the saccharification of lignocellulosics (LGCs); the role of the cellulose and hemicellulose degradation products; the relevant sustainability issues; and the future trends. Key words: lignocellulose, lignocellulosic biomass, bioethanol, biofuel, cellulose saccharification.

13.1

Introduction

Several reviews have been published on the fuel ethanol production from lignocellulosic biomass (Tsoutsos and Koukios, 1990; Lynd, 1996; Chandrakant and Bisaria, 1998); lignocellulosic complex is the most abundant biopolymer on the planet. It is estimated that lignocellulosic materials comprise about 50% of world biomass (10±50 billion t/year). Many lignocellulosic materials have been tested for bioethanol production, and could be grouped as follows: · crop residues (cane bagasse, straws of cereals, barley straw, sweet sorghum bagasse) · wood (hardwood, softwood) · herbaceous biomass (alfalfa hay, switchgrass, reed canary grass) · agroindustrial residues (sawdust, wood chips, olive kernels and pulp) · cellulosic wastes (newsprint, waste office paper, recycled paper sludge). These types of materials comprise primarily cellulose, hemicellulose and lignin. As Table 13.1 indicates carbohydrates account for roughly 50±70% of the dry mass of most biomass species, with lignin accounting for approximately 20% in additional; ash and other minor components make up the balance.

ß Woodhead Publishing Limited, 2010

Table 13.1 Composition (wt%, dry basis) of representative lignocellulosic biomass species (McMillan, 1997) Hexosan ß Woodhead Publishing Limited, 2010

Biomass type

Pentosan

Acetyl

Lignin (Klason)

Total carbohydrate

Glucan

Galactan

Mannan

Xylan

Arabinan

Hardwood species Silver maple Sycamore Black locust Poplar hybrid NE388 Sweetgum

45.9 44.0 49.4 48.6 49.5

0.0 0.0 0.0 0.3 0.3

1.2 0.9 1.0 0.5 0.4

17.1 16.3 16.2 14.6 17.5

0.7 0.6 0.4 0.3 0.4

3.9 3.6 3.8 2.2 2.3

20.8 22.8 21.5 21.8 21.8

64.9 61.8 67.0 64.3 68.1

Herbaceous sp. Switchgrass Weeping lovegrass Flatpea hay

36.6 36.7 28.9

1.2 1.7 1.5

0.0 0.0 0.1

16.1 17.6 7.4

2.2 2.6 2.0

1.1 1.1 1.4

21.9 21.2 24.5

56.1 58.6 39.9

Ag. residues Corn cobs Corn stover

39.4 40.9

1.1 1.0

0.0 0.0

28.4 21.5

3.6 1.8

1.9 1.9

17.5 16.7

72.5 65.2

342

Bioalcohol production

This chapter reviews: · · · · ·

The The The The The

13.2

main lignocellulose-to-bioethanol technology pathways. critical issues of the saccharification of lignocellulosic materials. role of the cellulose and hemicellulose degradation products. relevant sustainability issues. future trends.

Saccharification of lignocellulosic biomass by chemical/enzymatic processes

13.2.1 Pretreatment of lignocellulosic biomass The production of lignocellulosic sugars is the most critical stage in the process of ethanol production, because good quality solutions favour efficient conversion to ethanol. However, the process is not complete, because the ligninhemicellulose complex inhibits the penetration of the hydrolytic means, while the chemical breakdown of cellulose is not easy due to its crystallinity. As a result, for the improvement of reaction productivity, pretreatment of the raw material is essential; the cellulose-lignin matrix should be broken in order to increase the fraction of amorphous cellulose, the most preferable form for acid or (mainly) enzymatic attack, and also the major part of hemicellulose should be hydrolyzed and lignin should be released or even degraded (Sanchez and Cardona, 2008). Furthermore the cellulose hydrolysis is affected by the porosity of lignocellulosic materials; the yield of cellulose hydrolysis is less than 20% of the theoretical when pretreatment is not carried out, whereas the yield after pretreatment often exceeds 90% of theoretical value (Lynd, 1996); therefore removal of lignin and hemicellulose, reduction of crystalline cellulose and increase in porosity of the materials are performed. Additionally, the pretreatment should improve the formation of sugars during the succeeding hydrolysis, and avoid the formation of inhibitors for subsequent hydrolysis and fermentation processes. For the pretreatment of lignocellulosic materials several processes have been proposed and developed (Table 13.2).

13.2.2 Hydrolysis process Typically, after the pretreatment process three hydrolysis alternatives exist: · enzymatic hydrolysis · concentrated acid hydrolysis · dilute acid hydrolysis (Sidiras and Koukios, 2004). Dilute acid hydrolysis is selected for the production of fermentable sugars via softer conditions than those in the case of concentrated acid. This process uses

ß Woodhead Publishing Limited, 2010

Modelling hydrolysis and fermentation processes

343

Table 13.2 Pretreatment methods of lignocellulosic biomass for fuel ethanol production (Sanchez and Cardona, 2008) Methods

Procedure/agents

Physical methods Mechanical Chipping, gridding, contribution milling Pyrolysis

T > 300 ëC

Examples of pretreated materials

References

Wood and forestry wastes (hardwood, straw)

Papatheophanous et al. (1998) Sun and Cheng (2002) Khiyami et al. (2005); Sun and Cheng (2002)

Wood, waste cotton, corn stover

Physical-chemical methods Steam Saturated steam at explosion 160±290 ëC, p ˆ 0:69±4.85 MPa

Eucalyptus, softwood

Liquor hot water

Bagasse, corn stover, olive pulp

Ammonia fiber explosion (AFEX)

Pressured hot water, p > 5 MPa, T ˆ 170±230 ëC, >1 min, solid/liquid < 20% 1±2 kg NH3/kg dry biomass, 90 ëC, 30 min, p ˆ 1:12± 1.36 MPa

CO2 explosion

4 kg CO2/kg fiber, p ˆ 5:62 MPa

Chemical methods Ozonolysis Ozone, room temperature and pressure Dilute-acid 0.5±5% H2SO4, HCl, p ˆ 1 MPa (solid/ hydrolysis liquid < 10%), T > 100 ëC Concentrated acid hydrolysis Alkaline hydrolysis

10±30% H2SO4, 170±190 ëC; solid/ liquid 20±60% Dilute NaOH, 1 day, 60 ëC; Ca(OH)2, 4 h, 120 ëC, potentially H2O2

Ballesteros et al. (2002); Lynd et al. (2002); SÎderstrÎm et al. (2003); Sun and Cheng(2002) Ballesteros et al. (2002); Lynd et al. (2002)

Bagasse, wheat straw, barley straw, rice hulls, corn stover, switchgrass, alfalfa Municipal solid wastes; bagasse; alfalfa; recycled paper

Lynd et al. (2002); Sun and Cheng (2002)

Poplar sawdust, bagasse, wheat straw, cotton straw Poplar wood, bagasse, corn stover, wheat straw, rice hulls, switchgrass Poplar sawdust; bagasse

Sun and Cheng (2002)

Hardwood; bagasse; corn stover; straws with low lignin

ß Woodhead Publishing Limited, 2010

Sun and Cheng (2002)

Lynd et al. (2002); Sun and Cheng (2002) Teixeira et al. (1999) Teixeira et al. (1999); Sun and Cheng (2002); Lynd et al. (2002)

344

Bioalcohol production

Table 13.2 Continued Methods

Procedure/agents

Examples of pretreated materials

References

Oxidative delignification Wet oxidation

Peroxidase, 2% H2SO4, 20 ëC, 8 h

Bagasse

Sun and Cheng (2002)

1.2 MPa oxygen pressure, 195 ëC, 15 min, small amount of Na2CO3, H2SO4 Organic solvents (MeOH, EtOH, acetone, ethylene glycol, triethylene glycol); mixture with 1% H2SO4 or HCl; 185±198 ëC, 30± 60 min, pH ˆ 2:0± 3.4

Corn stover, wheat straw

Bjerre et al. (1996)

Corn stover, wheat straw, poplar wood

Lynd et al. (2002); Sun and Cheng (2002)

Corn stover, wheat straw

Sun and Cheng (2002)

Beech wood

Itoh et al. (2003)

Organosolv

Biological methods Fungai Fungi pretreatment Cellulase and hemicellulase production BioC. subvermispora for organosolv 2±8 weeks; ethanolysis at 140± 200 ëC

dilute acid concentration (up to 3±4%) in temperatures 100±240 ëC. A number of acids may be used, such as HCl, H2SO4, H3PO4 and HNO3. In temperatures between 110 and 140 ëC hemicellulose is hydrolyzed, while crystallic cellulose remains practically unchanged up to 170 ëC and takes place up to 240 ëC. The difference between these two parts dominates the design of a two-stage process. The hydrolysis of, separately, the hemicellulose and the cellulose parts, has already been studied for wood biomass (Stinson, 1983; Tsoutsos and Koukios, 1990) and concerns initial hemicellulose hydrolysis in low temperature (120±150 ëC) and then cellulose hydrolysis at high temperature up to 240 ëC. During prehydrolysis the lignine-hemicellulose complex is broken down, facilitating the hemicellulose hydrolysis and the sugar production, mainly xylose, under relatively soft conditions. However, at increased temperature xylose is broken down and undesirable byproducts are formulated. The next step is the application of higher temperatures (>170 ëC) and potentially increased acid concentrations, so that the cellulosic part may be hydrolyzed. The two stages process has several advantages such as:

ß Woodhead Publishing Limited, 2010

Modelling hydrolysis and fermentation processes

345

· it allows the production of useful byproducts such as xylitol and aravitol · it increases the cellulose breakdown during hydrolysis and consequently the sugar yield · it is more economic than the concentrated acid reaction because it requires cheaper equipment · it avoids important environmental problems related to the use of strong acids · its general design and management are less complicated than enzymatic hydrolysis. However, it is noted that the operating conditions need to be carefully selected in order to avoid high byproduct concentrations with significant inhibitory effect during the sequential fermentation. Also, before the fermentation process, the hydrolysates' pH should be regulated, in order not to suspend the metabolism of fermentation microbial cultures.

13.2.3 Kinetics of the dilute acid hydrolysis Kinetics of the acid cellulose hydrolysis Saeman suggested the cellulose hydrolysis as a pseudo-homogenous system of two simple sequential reactions (Saeman, 1945; Tsoutsos and Koukios, 1991). During the first reaction cellulose is hydrolyzed to glucose, while during the second glucose was degraded, producing byproducts as follows: k1

k2

cellulose ÿ! glucose ÿ! degradation products Saeman modified the Arrhenius type constants as follows: ki ˆ Ai
C ni
e…ÿEi =…R
T††

13:1

where: i ˆ 1 for the production reaction, i ˆ 2 for the degradation reaction, C is acid concentration (%w/w), T is temperature (ëK), E1 and E2 are the activation energy of the production and degradation of glucose, n1 and n2 are exponential acid parameters, and A1 and A2 are pre-exponential parameters (minÿ1). So the glucose concentration (G) versus time is described as in equation [13.2]: dG 13:2 ˆ k1 C ÿ k2 G dt Examples of kinetics for lignocellulose are presented in Table 13.3. The values of the above parameters depend on the proportion of three main biomass components and mainly on the way of the inter-connection of these components. This model includes several simplifications, as the system is, in fact, heterogeneous, while the crystallinity of cellulose and the acid diffusion in this are not taken into account. Having this in mind, more complex models have

ß Woodhead Publishing Limited, 2010

Table 13.3 Kinetics of lignocellulosic hydrolysis ki ˆ Ai
Cni
e…ÿEi =…R
T††

R ˆ 8; 314  10ÿ3 kJ/Mol K) T (ëK) i ˆ 1; 2

xylan ! xylose Lignocellulosic

xylose ! F

ß Woodhead Publishing Limited, 2010

A1 (minÿ1)

n1

E1 (kJ/mol)

A2 (minÿ1)

n2

E2 (kJ/mol)

Corn stover

3:68
1020 …2:08 ÿ 1=C†

0

171.6

1:95
1014 …2:36 ÿ 1=c†

0

133.9

Bhandari et al. (1984)

Wheat straw

2:025
1020

1.55

167.0

1:52
1015

2.00

141.0

Ranganathan et al. (1985)

0.688

113.51

Kwarteng (1983)

Hardwood

13

6:23
10

1.17

116.43

2:33
10

12

gluan ! glucose Lignocellulosic Corn stover Wheat straw Hardwood

Source

glucose ! HMF

A1 (minÿ1)

n1

E1 (kJ/mol)

A2 (minÿ1)

n2

E2 (kJ/mol)

2:71
1019

2.74

189.6

2:01
1014

1.86

137.3

Bhandari et al. (1984)

2:21
10

14

0.68

150.62

Tsoutsos and Koukios (1990)

2:75
10

12

1.17

124.68

Lee et al. (1999)

19

1:68
10 2:85
10

13

0.7 1.2

190.37 133.05

Source

Modelling hydrolysis and fermentation processes

347

been developed, which examine the proportion between crystalline ± amorphous cellulose, as well as the acid diffusion conditions in the raw material, but also the polymerization degree of the cellulose parts. Kinetics of the hemicellulose acid hydrolysis Based on the Saeman's model Chambers developed the equivalent for hemicellulose hydrolysis and this study gave satisfactory results for compatibility with experimental results (Chambers et al., 1979): k1

k2

hemicellulose ÿ! xylose ÿ! degradation products To successfully understand their experimental results, a more complicated model has been suggested, which describes the split of the tree-like hemicellulose in two xylan parts (Veeraraghavan et al., 1982). The first (I) is hydrolyzed quickly to oligomers, while the second (II) follows the same course but much slower. Afterwards the oligomers are hydrolyzed to xylose, which then is hydrolyzed to degradation products. Enzymatic hydrolysis The enzymatic hydrolysis is carried out using microbial cellulolytic enzymes; it has verified enhanced results for the subsequent fermentation because no degradation components of glucose are formed although the process is slower. Most of the commercial cellulases are obtained aerobically from T. reesei, though a small portion is obtained from A. niger. T. reesei releases a blend of cellulases, such as -glucosidases, cellobiohydrolases, endoglucanases, and hemicellulases (Zhang and Lynd, 2004). Cellobiohydrolases causes a gradual decrease in the polymerization degree, endoglucanases cause the break of cellulose in smaller chains, endoglucanases particularly act on amorphous cellulose, and cellobiohydrolases can act on crystalline cellulose (Lynd et al., 2002). Unfortunately, cellobiohydrolases are inhibited by the cellobiose. So, -glucosidase from other source needs to be added in order to complement the action of the cellulases of this fungus. Cellulases should be adsorbed on the surface of substrate particles before hydrolysis of insoluble cellulose takes place. The three-dimensional structure of these particles in combination with their size and shape determines whether glucosidic linkages are accessible to enzymatic attack (Zhang and Lynd, 2004). This makes cellulose hydrolysis slower compared to the enzymatic degradation of other biopolymers. For instance, the hydrolysis rate of starch by amylases is 100 times faster than hydrolysis rate of cellulose by cellulases under industrial processing conditions. A mechanistic model of enzymatic hydrolysis for cellulose by T. reesei cellulase is presented in Fig. 13.1.

ß Woodhead Publishing Limited, 2010

348

Bioalcohol production

13.1 Mechanistic model of enzymatic hydrolysis for cellulose by T. reesei (Zhang and Lynd, 2004).

A classification scheme for quantitative models of enzymatic hydrolysis of cellulose is proposed in Table 13.4: · `Nonmechanistic models' are those based on data correlation without an explicit calculation of adsorbed cellulase concentration. Although these models may be useful for correlating data, they are unlikely to be reliable under conditions different from those for which the correlation was developed, and they have limited utility for testing and developing understanding. · `Semimechanistic models' are featuring defensible adsorption, but are based on concentration as the only variable describing the state of the substrate and/ or are based on a single cellulose hydrolyzing activity. Most of the hydrolysis models proposed to date for design of industrial systems fall into the category of semimechanistic models. · `Functionally based models' are featuring adsorption, substrate state variables in addition to concentration, and multiple enzyme activities. Functionally based models are particularly useful for developing and testing understanding at the level of substrate features and multiple enzyme activities, including identification of rate-limiting factors and strategies to alleviate such factors. Examples of nonmechanistic mathematical models of the enzymatic cellulose hydrolysis are presented in Table 13.5.

ß Woodhead Publishing Limited, 2010

Table 13.4 Classification scheme for models of enzymatic cellulose hydrolysis (Zhang and Lynd, 2004)

ß Woodhead Publishing Limited, 2010

Model category

Defining feature and basis

Utility

Limitations

Non mechanistic

Not based on a defensible adsorption model

Data correlation

· Reliability under conditions different from those used to develop the correlation · Does not enhance understanding

Semimechanistic

Based on a defensible adsorption model

· Data correlation · Reactor design · Identification of essential features

· Understanding at the level of substrate features and multiple enzyme activities

Functionally based

Includes an adsorption model, substrate state variables in addition to concentration, multiple solubilizing activities

· Testing and developing understanding at the level of substrate features and multiple enzyme activities · Identifying rate-limiting factors · Reactor design

· Molecular design · State of model development and data availability currently limit application to design

Structurally based

Based on structural information pertaining to cellulase components

· Molecular design · Testing and developing understating of structure/ function relationship

· Challenging to develop meaningful kinetic models based on structural information

350

Bioalcohol production

Table 13.5 Mathematical model of nonmechanistic enzymatic cellulose hydrolysis (Zhang and Lynd, 2004) Dependent variables Authors

Independent variable

Time

Enzyme

Miyamoto and Nisozawa (1942) Holtzapple et al. (1984) Gharphuray et al. (1983) Sattler et al. (1989) Koullas et al. (1992) Karrer et al. (1925)

Conversion

X

X

Conversion

X

13.3

Substrate

>1 substrate variable

Conversion Conversion Conversion Hydrolysis rate

X X X X

Fermentation by various microorganisms

13.3.1 The inhibition of fermentation During pretreatment and hydrolysis of lignocellulosic biomass, various compounds ± inhibitors of the subsequent fermentation are formed in parallel to fermentable sugars. Inhibitors are generated as a result of the hydrolysis of the extractive components, organic and sugar acids esterified to hemicellulose (acetic, formic, glucuronic, galacturonic), and solubilized phenolic derivatives. In the same way, inhibitors are produced from the degradation products of soluble sugars (furfural, HMF) and lignin (cinnamaldehyde, p-hydroxybenzaldehyde, syringaldehyde), and as a consequence of corrosion (metal ions) (Lynd, 1996; Taherzadeh, 1999; Palmqvist and Hahn-HaÈgerdal, 2000). The role of furfural (F) Furfural (F) is the main degradation product of xylose:

F inhibits the in vitro activity of important glucolytic enzymes resulting in the reduction of rhythm of reproduction of microbial population and ethanol productivity. The relation between F concentration and inhibition is influenced by several factors, mainly the type of used microorganism but also the initial vaccination concentrations of the culture (Olsson and Hahn-HaÈgerdal, 1996).

ß Woodhead Publishing Limited, 2010

Modelling hydrolysis and fermentation processes

351

It is reported that concentrations >1 g/L in cultures of Saccharomycae cerevisae (SC) decrease considerably the rhythm of CO2 receipt (Sanchez and Bautista, 1988) and the cell development in the initial stage of fermentation. With use of high initial quantity of vaccination SC 3.5 g/L important reductions both at the increase rate of microbial population and in the ethanol production rate were observed with addition 3.0 F g/L (Taherzadeh, 1999). The role of 5-hydroxy-methyl-furfural (HMF) 5-hydroxy-methyl-furfural (HMF) is the main product of the glucose degradation:

HMF inhibits the fermentation process likewise F. It has been found that HMF has lower activity; however, it remains in the culture four times more than F. 1 g HMF/ L has no inhibitory effect in the yeast. However, corresponding addition of 2 g/L resulted in the light reduction of proteins quantity and reduction of the growth rate at 23%. In other research on SC culture, concentration 2 g/L simply elongated the stagnation phase of the culture (Sanchez and Bautista, 1988). Additional 4 g/L resulted in reduction 32% CO2 receipt, 40% in ethanol production and 70% in the increase rate of microorganisms. However, it has been found that increase of the initial glucose concentration decreases the effect of HMF (Mielenz, 2001). Detoxification of lignocellulosic hydrolyzates For the above reasons and depending on the type of adopted pretreatment and hydrolysis, detoxification of the solutions that will undergo fermentation is required. Detoxification methods can be physical, chemical or biological. These methods vary in the neutralization degree of the inhibitors. Also, the fermenting microorganisms have different tolerances to the inhibitors. The main characteristics of the current detoxification methods are reviewed in Table 13.6.

13.3.2 Fermentation of biomass hydrolysates A variety of microorganisms, generally either bacteria, yeast, or fungi, ferment carbohydrates to ethanol under oxygen-free conditions according to the reaction: C6H12O6 ! 2C2H5OH + 2CO2 The classic configuration involves a sequential process where the hydrolysis of cellulose and the fermentation are carried out in different units (separate

ß Woodhead Publishing Limited, 2010

352

Bioalcohol production

Table 13.6 Detoxification methods of streams resulting of pretreatment and hydrolysis of lignocellulosic biomass for bioethanol production (Sanchez and Cardona, 2008) Methods

Procedures/agents

Microorganism

References

S. cerevisae

Palmqvist and HahnHÌgerdal (2000)

S. cerevisae; P. stipitis S. cerevisae

Palmqvist and HahnHÌgerdal (2000); Cantarella et al. (2004) Persson et al. (2002)

S. cerevisae

Lee et al. (1999)

Recombinant; E. coli

Weil et al. (2002)

Chemical methods NeutralizaCa(OH)2 or CaO, pH=6 tion

S. cerevisae

Alkaline Ca(OH)2; pH=9±10.5 detoxification

S. cerevisae; E. coli

Combined KOH, pH=20, then pH alkaline adjustment to 6.5 detoxification Ionic Weak base resins exchange

P. stipitis; S. cerevisae

Yu and Zhang (2003); Cantarella et al. (2004) Yu and Zhang (2003); Martin et al. (2002) Palmqvist and HahnHÌgerdal (2000)

Physical methods Evaporation Evaporation, separation of volatile and nonvolatile fractions and dilution of non volatile fraction Extraction Organic solvents, 3:1 org. phase: aqueous phase

Adsorption

Supercritical solvent, 20 MPa, 40 ëC Activated carbon, 0.05±0.20 g/g glucose Amberlite hydrophobic polymeric adsorbent XAD-4, 8% w/v, 1.5 h, 25 ëC

Biological methods Enzymatic Laccase (phenol detoxification oxidase); lignin peroxidise Microbial Trichoderma reesei; detoxification Pseudomonas putida; Streptomyces sentonii

Z. mobilis; S. cerevisae

Wooley et al. (1999); Palmqvist and HahnHÌgerdal (2000)

S. cerevisae

Palmqvist and HahnHÌgerdal (2000); Martin et al. (2002) Khiyami et al. (2005)

ß Woodhead Publishing Limited, 2010

Modelling hydrolysis and fermentation processes

353

hydrolysis and fermentation ± SHF-). The most used microorganism for fermenting lignocellulosic hydrolysates is SC, which ferments selectively the hexoses of the hydrolysates excluding the pentoses. In this case, once hydrolysis is completed, the produced hydrolysate is fermented and converted into ethanol.

13.3.3 Fermentation of pentoses One of the main problems in bioethanol production from lignocellulosic biomass is that SC can ferment only certain mono- and disaccharides like glucose, fructose, maltose and sucrose, but not pentoses obtained during hemicellulose hydrolysis (mainly xylose). Methods to overcome this obstacle are through recombinant DNA technology or the use of pentose fermenting microorganisms like some species of yeasts and bacteria: 3C5H10O5 ! 5C2H5OH + 5CO2 In this case, configurations involving the separate fermentation of pentoses and hexoses have been proposed. Olsson and Hahn-HaÈgerdal (1996) extensively reviewed the microorganisms employed for pentose fermentation noting that most of them are inhibited by substrate, besides product inhibition. Pentose fermenting yeasts require a careful control for maintaining low oxygen levels in the culture medium needed for their oxidative metabolism. Additionally, these yeasts successfully ferment pure xylose, but not the aqueous hemicellulose flows generated during the biomass pretreatment, probably due to the presence of different inhibitors (Chandrakant and Bisaria, 1998). For pentose utilizing microorganisms, the hexoses are unquestionably the easier and faster assimilable substrate to ethanol. If fermentation time is not sufficiently long, pentoses remain in the medium decreasing the utilization rates of the lignocellulosic complex. As a rule, microorganisms prefer glucose over galactose followed by xylose and arabinose. This is explained by the catabolic repression that glucose exerts on the uptake rates of xylose and other pentoses as in the case of C. shehatae. To offset this effect, sequential fermentations are employed in which SC utilizes the hexoses during the first days of cultivation and later xylose-utilizing yeast is added in order to complete the conversion to ethanol (Chandrakant and Bisaria, 1998). Thermophilic and saccharolytic clostridia are an important group of ethanolproducing microorganisms and include species as Clostridium thermohydrosulfuricum, Thermoanaerobacter thermosaccharolyticum and C. thermocellum. These bacteria may transform pentoses and aminoacids into ethanol and can synthesize up to 2 mole EtOH/mole hexose. Having saccharolytic properties, these microorganisms have the ability to grow on a wide variety of non-treated wastes. C. thermocellum can even directly convert lignocellulosic materials into ethanol (McMillan, 1997). The main drawback is their very low ethanol

ß Woodhead Publishing Limited, 2010

354

Bioalcohol production

tolerance. Consequently the maximal reached concentrations of ethanol are less than 30 g/L. Xylose-fermenting termophilic bacteria are prospective organisms to be cocultured with cellulose hydrolyzing bacteria such C. thermocellum in order to directly convert pretreated lignocellulosic biomass into ethanol, process named consolidated bioprocessing (Sanchez and Cardona, 2008).

13.4

Simultaneous saccharification and fermentation (SSF)

In simultaneous saccharification and fermentation (SSF), hydrolysis and fermentation are performed in a single unit. SSF demonstrates better performance than the SHF process such as higher ethanol yields and less energetic consumption. In this case, cellulases and microorganisms are added to the same process unit allowing the glucose formed during the enzymatic hydrolysis to be immediately consumed by the microbial cells converting it into ethanol. Thus, the inhibition effect caused by the sugars over the cellulases is neutralized. However, the need to use more dilute media to reach suitable rheological properties means the final product concentration is low. In addition, this process operates at non-optimal conditions for hydrolysis and requires a higher enzyme dose, which influences substrate conversion positively, but process costs negatively. Considering that enzymes are one of the main reasons for high production costs, it is necessary to find methods to reduce the cellulases doses to be utilized. Alkasrawi et al. (2003) showed that the addition of the non-ionic surfactant Tween-20 to the steam exploded wood during a batch SSF using SC has some effects: 8% increase in ethanol yield, 50% reduction in cellulases dose, increase of enzyme activity at the end of the process, and decrease in the time required for reaching the highest ethanol concentration. It is assumed that the surfactant avoids or reduces the non-useful adsorption of cellulases to the lignin. Hari Krishna et al. (1998) evaluated the optimal conditions of the SSF of sugar cane leaves, as they did for the SHF. They defined a temperature of 40 ëC and pH = 5.1 as the best conditions for 3-d cultivation, achieving 31 g/L of ethanol from an initial substrate load as high as 15%. Nevertheless, the enzyme dose was quite high (100 FPU/g cellulose). Softwood is more difficult to degrade by SSF than hardwood. Stenberg et al. (2000) used the resulting slurry of the steam pretreatment of SO2-impregnated spruce in SSF tests using yeasts and determined that the best initial load of substrate was 5% (w/w) reaching an 82% yield based on the cellulose and soluble hexoses present at the start of the process. The productivity was doubled related to SHF. The cellulases load was in the range of 5±32 FPU/g cellulose. Varga et al. (2004) proposed a non-isothermal regime for batch SSF process in the case of wet oxidized corn stover: in the first step, small amounts of

ß Woodhead Publishing Limited, 2010

Modelling hydrolysis and fermentation processes

355

cellulases were added at 50 ëC to obtain better mixing conditions. In the second step, more cellulases were added along with the yeast SC at 30 ëC. In this way, the final solid concentration in the hydrolysate could be increased up to 17% dry matter concentration achieving 78% ethanol yield. In general, increased cultivation temperature accelerates metabolic processes and lowers the refrigeration requirements. Kadar et al. (2004) compared the performance of thermotolerant K. marxianus and SC during batch SSF of waste cardboard and paper sludge not finding great differences between both microorganisms at 40 ëC, although cellulose conversions (55±60%) and ethanol yields (0.30±0.34 g/g cellulose) were relatively low. With the aim of increasing ethanol yields from olive pulp, Ballesteros et al. (2002) employed LHW pretreatment reaching a yield of 80% of theoretical and recovering potentially valuable phenolic compounds.

13.5

Environmental issues

Presumed environmental benefits are important drivers for greater use of biofuels ± particularly the benefits of reduced GHG emissions. Yet, no fuel system is free of environmental concerns (Hoekman, 2009). Biomass exploitation provides large challenges due to raw material abundance, particularly in the technological scheme of fuel bioethanol, which is environmentally friendly and directly exploitable for substitution of petrochemicals, which today are used for the 97% of transport needs (Mielenz, 2001). Besides bioethanol production is greenhouse gas neutral, if only renewable raw materials will be used. Some of the major concerns in biofuel production deal with water implications. Currently, biofuels are a marginal additional stress on water supplies at the regional to local scale. However, significant acceleration of biofuel production could cause much greater water quantity problems depending on where the crops are grown. Concerns about both water quantity and quality are less severe for biofuels from lignocellulosic biomass; also in application rates of nitrogen and pesticides (potentially minimizing nutrient pollution of waterways) and less erosion ± though some of these impacts could be mitigated by improved agricultural practices. There has been considerable global debate about the life-cycle GHG effects (and energy balance) of ethanol fuel (Tsoutsos et al., 2007). Numerous studies have reported widely differing results, largely because of different assumptions regarding agricultural practices, fossil energy sources used in producing ethanol, allocation of GHG emissions (and energy inputs) to co-products that are manufactured along with the ethanol and the extent and impacts of land use changes. For larger GHG reductions, cellulosic ethanol is required; Hammerschlag (2006) determined small energy and GHG benefits for most corn ethanol processes, and much larger benefits from cellulosic processes. In parallel Farrell

ß Woodhead Publishing Limited, 2010

356

Bioalcohol production

et al. (2006) also used their model to determine GHG emissions and energy inputs for three ethanol cases compared to gasoline. The `cellulosic ethanol' scenario reduces GHGs by over 85% compared to the gasoline baseline. Very recently, several reports have suggested that biofuels may have significant adverse GHG impacts. This stems largely from concerns that changes in land use patterns can increase GHG emissions, and these increases are not treated properly in life cycle assessment (LCA) models. Although many LCA studies have considered ± at least partially ± the impacts of direct land use changes (for example, growing more corn at the expense of soy beans or pasture land), they have not generally accounted for indirect land use changes (for example, clearing rainforests to grow crops to satisfy food/feed requirements that have been disrupted by USA renewable fuel policies). Other important considerations are the specific cropping practices being utilized to grow the biofuel feedstocks, and the impacts of these practices upon GHG emissions ± particularly N2O. All these issues regarding biofuels' lifecycle GHG impacts remain controversial, and areas for on-going study.

13.6

Successful examples

Four main regions have a head start in cellulosic ethanol ± USA, Europe, China and Brazil ± but other regions and countries are starting to focus their efforts on second generation ethanol. USA Corn stover and wheat straw are the two types of feedstock considered to hold the major potential for ethanol production. However, woody substrates (sawdust, wood trimmings, soft wood and hard wood), dedicated energy crops and urban waste or rice straw may have significant potential as well. Table 13.7 presents the existing commercial plants in USA (Hoekman, 2009). Europe In Europe efforts are primarily set on producing ethanol from wheat straw, but other potential feedstocks are citrus peels and urban waste (Ebio, 2008). China There is a major effort from the NDRC (National Development and Reform Commission) to help Chinese production through subsidies. A substantial number of companies are working in the cellulosic ethanol field, focusing on microorganisms, pretreatment technology and ethanol production.

ß Woodhead Publishing Limited, 2010

Modelling hydrolysis and fermentation processes

357

Table 13.7 Commercial plants in USA (Hoekman, 2009) Lead company

Feedstocks

Process details

Project location

Abengoa Bioenergy

700 t/d ag. Residues (corn stover, wheat straw, switchgrass)

Enzymatic Kansas, hydrolysis, USA saccharification, fermentation

Ethanol (15 mg/y) heat, power

Blue Fire Ethanol

700 t/d landfill green waste and wood waste

Concentrated acid hydrolysis, fermentation

Ethanol (19 mg/y), heat, power

California, USA

Expected products

Broin 840 t/d ag. (now Poet) Residues (corn fiber, cobs and stalks)

Hydrolysis, Iowa, USA saccharification, fermentation

Ethanol (30 mg/y)

Iogen 700 t/d ag. Biorefinery Residues (wheat and barley straw, corn stover, swithcgrass, etc.)

Hydrolysis, Idaho, USA saccharification, fermentation

Ethanol (18 mg/y)

t/d: dry tons per day; mg/y: million gallon ts/year

Brazil Today, Brazil is the country in the world with the lowest production costs for first generation ethanol. In Brazil ethanol is produced from sugar cane. Bagasse, the fibrous residue remaining after sugar extraction from sugar cane, is a potential biomass resource for new bioethanol processes. In Brazil many universities and research centers are focused on developing a technology for treatment of bagasse to be able to convert it into ethanol.

13.7

Future trends

Cellulosic ethanol is not expected to be realized in short-term. Relevant companies operate with a 5-year timeline. Still today, all projects focusing on ethanol production from cellulosic feedstocks experience very basic problems with feedstock collection, processing and high treatment costs. Enzyme costs are the biggest cost contributor for the production of second generation ethanol. Enzyme cost has been viewed as the major barrier to biomass conversion ± this is no longer the case even though improvements still need to be achieved to make it an economically viable process. Grassi (1999) points out that the development of bioenergy production technologies would represent the creation of 200 000 direct and indirect jobs and the reduction of 255 CO2 Mts/year of in 2010. Nevertheless, the main limitation

ß Woodhead Publishing Limited, 2010

358

Bioalcohol production

in the use of this resource is the conversion process of biomass into ethanol as analyzed in Section 13.3. Once these technological limitations are overcome, lignocellulosic biomass will be the main feedstock for ethanol production. Certainly, a detailed economic and environmental evaluation of the different feedstocks is required in order to make decisions on the most appropriate raw materials for fuel ethanol production in each case. Several studies for developing large-scale production schemes have been carried out; however, the main determinant is the higher degree of complexity inherent in the processing of this feedstock, due to the nature and composition of lignocellulosic materials. Two of the main polymers of the biomass should be broken down into fermentable sugars in order to be converted into ethanol or other valuable products; however, this degradation process is complicated, energy-consuming and under development. The massive utilization of fuel ethanol requires that its production technology be economically and environmentally sustainable. For current technologies at commercial level, the main share in the cost structure corresponds to the feedstocks (above 60%) followed by processing expenditures. The lignocellulosic biomass represents the most prospective feedstock for ethanol production. The availability and low cost of a wide range of lignocellulosic materials offer many possibilities for the development of bioindustries that could contribute to the reduction of greenhouse gas emissions worldwide. The current research trends for improving fuel ethanol production are linked to the nature of used raw materials, and process engineering issues. The main trends in the conversion of different feedstock into ethanol are summarized in Table 13.8. For the three main types of feedstocks (Table 13.1), the development of effective, continuous fermentation technologies with near to 100% yields and elevated volumetric productivities is one of the main research subjects in the ethanol industry. To this end, many newly proposed technologies for reducing the product inhibition effect on the cell growth rate should be scaled-up to industrial level.

13.8

Sources of further information and advice

Nowadays there is a global interest in cellulosic bioethanol. Science and technology (S&T) knowledge is under development, amongst the important electronic information sources are suggested: · EU project `New Improvements for Ligno-cellulosic Ethanol' www.nile-bioethanol.org · European Bioethanol Fuel Association (Ebio) www.ebio.org · European Biomass Industry Association www.eubia.org

ß Woodhead Publishing Limited, 2010

Modelling hydrolysis and fermentation processes

359

Table 13.8 Research trends and priorities for improving fuel ethanol from different feedstocks (Sanchez and Cardona, 2008) Issue

Critical issues

Feedstock

Evaluation of the use of dedicated energy crops Generic modification of herbaceous plants for changing their carbohydrate content Economic utilization of different and alternative wastes like municipal solid wastes (MSW)

Pretreatment

Reduction of the milling power Optimization of steam explosion and dilute-acid pretreatment Development of LHW, AFEX and alkaline hydrolysis Reduced formation of inhibitors Recycling of concentrated acids

Hydrolysis

Increase in specific activity, thermal stability and cellulosespecific binding of cellulases (e.g., by protein engineering) Reduction of costs of cellulases production (10-fold reduction) Cellulases production by solid-state fermentation Recycling of cellulases Improvement of acid hydrolysis of municipal solid wastes (MSW)

Fermentation

Increase in conversion of glucose and pentoses to ethanol Recombinant strains with increased stability and efficiency for assimilating hexoses and pentoses, and for working at higher temperatures Development of strains more tolerant to the inhibitors Increase of ethanol tolerance in pentose-fermenting microorganisms

· European Environment Agency www.eea.europa.eu · National Renewable Energy Laboratory (USA) www.nrel.gov

13.9

Acknowledgements

I thank Professor Dr Vassilis Gekas for valuable discussions and recommendations.

13.10 References Alkasrawi M, Eriksson T, Borjesson J, Wingren A, Galbe M, Tjerneld F, Zacchi G (2003), `The effect of Tween-20 on simultaneous saccharification and fermentation of softwood to ethanol', Enzyme Microb Technol, 33, 71±78. Ballesteros I, Oliva JM, Negro MJ, Manzanares P, Ballesteros M (2002), `Ethanol production from olive oil extraction residue pretreated with hot water', Appl Biochem Biotechnol, 717±732.

ß Woodhead Publishing Limited, 2010

360

Bioalcohol production

Bhandari N, Macdonald DG, Bakhshi NN (1984), `Kinetic studies of corn stover saccharification using sulphuric acid', Biotechnol Bioeng, 26, 320±327. Bjerre AB, Olesen AB, Fernqvist T, Ploger A, Schmidt AS (1996), `Pretreatment of wheat straw using combined wet oxidation and alkaline hydrolysis resulting in convertible cellulose and hemicellulose', Biotechnol Bioeng, 49, 568± 577. Cantarella M, Cantarella L, Gallifuoco A, Spera A, Alfani F (2004), `Comparison of different detoxification methods for steam-exploded poplar wood as a substrate for the bioproduction of ethanol in SHF and SSF', Process Biochem, 39(11), 1533± 1542. Chambers RP, Lee YY, McCaskey TA (1979), `Liquid fuel and chemical production from cellulosic biomass ± Hemicellulose recovery and pentose utilization in a biomass processing complex', 3rd Annual Biomass System Conference Proceedings, DE-AS051, Golden, CO, 255±264. Chandrakant P, Bisaria VS (1998), `Simultaneous bioconversion of cellulose and hemicellulose to ethanol', Crit Rev Biotechnol, 18 (4), 295±331. Ebio, www.ebio.org, accessed December 2008. Farrell AE, Plevin RJ, Turner BT, Jones AD, O'Hare M, Kammen DM (2006), `Ethanol can contribute to energy and environmental goals', Science, 311, 506±508. Gharpuray MM, Lee YH, Fan LT (1983), `Structural modification of lignocellulosic by pretreatments to enhance enzymatic hydrolysis', Biotechnol Bioeng, 25, 157±172. Grassi G (1999), `Modern bioenergy in the European Union', Renewable Energy, 16, 985±990. Hammerschlag R (2006), `Ethanol's energy return on investment: a survey of the literature 1990 ± present', Environ Sci Technol, 40, 1744±1750. Hari Krishna S, Prasanthi K, Chowdary G, Ayyanna C (1998), `Simultaneous saccharification and fermentation of pretreated sugar cane leaves to ethanol', Process Biochem, 33 (8), 825±830. Hoekman SK (2009), `Biofuels in the US ± challenges and opportunities', Renewable Energy, 34, 14±22. Holtzapple MT, Caram HS, Humphrey AE (1984), `Determining the inhibition constants in the HCH-1 model of cellulose hydrolysis', Biotechnol Bioeng, 26, 735±757. Itoh H, Wada M, Honda Y, Kuwahara M, Watanabe T (2003), `Bioorganosolve pretreatments for simultaneous saccharification and fermentation of beech wood by ethanolysis and white rot fungi', J Biotechnol, 103(3), 273±280. Kadar Zs, Szengyel Zs, Reczey K (2004), `Simultaneous saccharification and fermentation (SSF) of industrial wastes for the production of ethanol', Ind Crops Products, 20, 103±110. Karrer P, Schubert P, Whrli W (1925), `Polysaccharide XXXIII: uber enzymatischen abbau von kunstseide und nativer cellulose', Helv Chim Acta, 8, 797±810. Khiyami M, Pometto AL, Brown RC (2005), `Detoxification of corn stover and corn starch pyrolysis liquors by ligninolytic enzymes of Phanerochaete chrysosporium', J Agric Food Chem, 53(8), 2969±77. Koullas DP, Christakopoulos P, Kekos D, Macris BJ, Koukios EG (1992), `Correlating the effect of pretreatment on the enzymatic hydrolysis of straw'. Biotechnol Bioeng, 38, 113±16. Kwarteng IK (1983), `Kinetics of acid hydrolysis of hardwood in a continuous plug flow reactor'. PhD thesis. Dartmouth Coll., Hanover, NH. Lee WG, Lee JS, Shin CS, Park SC, Chang HN, Chang YK (1999), `Ethanol production using concentrated oak wood hydrolysates and methods to detoxify', Appl Bioch Biot, 13, 547±559.

ß Woodhead Publishing Limited, 2010

Modelling hydrolysis and fermentation processes

361

Lynd LR (1996), `Overview and evaluation of fuel ethanol from cellulosic biomass: technology, economics, the environment, and policy', Ann Rev Energy Environ, 21, 403±465. Lynd LR, Weimer PJ, van Zyl WH, Pretorious IS (2002), `Microbial cellulose utilization: fundamentals and biotechnology', Microbiol Molecular Biol Rev, 66 (3), 506±577. MartõÂn C, Galbe M, Wahlbom C F, Hahn-HaÈgerdal B, JoÈnsson L J (2002), `Ethanol production from enzymatic hydrolysates of sugarcane bagasse using recombinant xylose-utilising Saccharomyces cerevisiae', Enzyme Microb Technol, 31( 3), 274±282. McMillan JD (1997), `Bioethanol production: status and prospects', Renewable Energy, 10, 295±302. Mielenz JR (2001), `Ethanol production from biomass: technology and commercialization status', Curr Opin Microbiol, 4 (3), 324±329. Miyamoto S, Nisizawa K (1942), `Model for cellulose hydrolysis by cellulase', J Vet Soc Arm Jpn, 396, 778±784. Olsson L, Hahn-HaÈgerdal B (1996), `Fermentation of lignocellulosic hydrolyzates for ethanol production', Enzyme Microb Technol, 18, 312±331. Palmqvist E, Hahn-HaÈgerdal B (2000), `Fermentation of lignocellulosic hydrolyzates. II: inhibitors and mechanisms of inhibition', Biores Technol, 74, 25±33. Papatheophanous MG, Billa E, Koullas DP, Monties B, Koukios EG (1998), `Optimizing multisteps mechanical-chemical fractionation of wheat straw components', Ind Crops Products, 7(2±3), 249±256. Ê , Holmgren J, SoÈderman U (2002), `Detecting and measuring individual trees Persson A using an airborne laser scanner', Photogram Eng Remote Sensing, 68(9), 925±932. Ranganathan DG, MacDonald DG, Bakhshi NN (1985), `Kinetic studies of wheat straw hydrolysis using sulphuric acid', Can J Chem Eng, 63, 840±844. Saeman JF (1945), `Kinetics of wood saccharification: hydrolysis of cellulose and decomposition of sugars in dilute acid at high temperature', Ind Eng Chem, 37, 43± 52. Sanchez B, Bautista J (1988), `Effects of furfural and 5-hydroxymethylfurfural on the fermentation of Saccharomyces cerevisiae and biomass production from Candida guilliermondii', Enzyme Microb Technol, 10, 315±318. Sanchez OJ, Cardona CA (2008), `Trends in biotechnological production of fuel ethanol from di?erent feedstocks', Biores Technol, 99, 5270±5295. Sattler W, Esterbauer H, Glatter O, Steiner W (1989), `The effect of enzyme concentration on the rate of the hydrolysis rate of cellulose', Biotechnol Bioeng, 33, 1221±1234. Sidiras D, Koukios E (2004), `Simulation of acid-catalysed organosolv fractionation of wheat straw', Biores Technol, 94, 91±98. SoÈderstroÈm J, Pilcher L, Galbe M, Zacchi G (2003), `Two-step steam pretreatment of softwood by dilute H2SO4 impregnation for ethanol production', Biomass Bioenergy, 24(6), 475±486. Stenberg K, Bollok M, Reczey K, Galbe M, Zacchi G (2000), `Effect of substrate and cellulase concentration on simultaneous saccharification and fermentation of steampretreated softwood for ethanol production', Biotechnol Bioeng, 68 (2), 204±210. Stinson JM (1983), `Energy from biomass', Technical Insights, Inc., New York. Sun Y, Cheng J (2002), `Hydrolysis of lignocellulosic materials for ethanol production: a review', Biores Technol, 83, 1±11. Taherzadeh MJ (1999), `Ethanol from Lignocellulose: Physiological Effects of Inhibitors and Fermentation Strategies', PhD Thesis, Chalmers, University of Technology Goteborg, Sweden.

ß Woodhead Publishing Limited, 2010

362

Bioalcohol production

Teixeira LC, Linden JC, Schroeder HA (1999), `Optimizing peracetic acid pretreatment conditions for improved simultaneous saccharification and co-fermentation (SSCF) of sugar cane bagasse to ethanol fuel', Renewable Energy, 16(1±4), 1070±1073. Tsoutsos TD, Koukios EG (1990), `Modeling and comparison of major types of dilute acid hydrolysis reactors for cellulose saccharification', Cellulose Chem Technol, 24 (6), 713±725. Tsoutsos TD, Koukios EG (1991), `Selection of a reaction system for cellulose saccharification: a global approach', Chem Biochem Eng Quart, 3, 151±156. Tsoutsos T, Bethanis D, Gekas V (2007), `Simulation of fermentable sugar production from lignocellulosics to fuel bioethanol', 15th European Biomass Conference and Exhibition, 7±11 May 2007, Berlin, e-proceedings. Varga E, Klinkle HB, Reczey K, Thomsen AB (2004), `High solid simultaneous saccharification and fermentation of wet oxidized corn stover to ethanol', Biotechnol Bioeng, 88 (5), 567±574. Veeraraghavan S, Chambers RP, Myles MY, Lee A (1982), `Kinetic model and reactor development in hemicellulose hydrolysis', AIChE National Meeting, Orlando. Weil JR, Dien B, Bothast R, Hendrickson R, Mosier NS, Ladisch MR (2002), `Removal of fermentation inhibitors formed during pretreatment of biomass by polymeric adsorbents', Ind Eng Chem Res, 41, 6132±6138. Wooley R, Ruth M, Glassner D, Sheehan J (1999), `Process design and costing of bioethanol technology: A tool for determining the status and direction of research and development', Biotechnol Bioeng, 15, 794±803. Yu Z, Zhang H (2003), `Pretreatments of cellulose pyrolysate for ethanol production by Saccharomyces cerevisiae, Pichia sp. YZ-1 and Zymomonas mobilis', Biomass Bioenergy, 24(3), 257±262. Zhang YHP, Lynd LR (2004), `Toward an aggregated understanding of enzymatic hydrolysis of cellulose: noncomplexed cellulose systems', Biotechnol Bioeng, 8 (7), 797±824.

ß Woodhead Publishing Limited, 2010

14

Environmental life cycle assessment of lignocellulose-to-bioalcohol production Y . Z H A N G , J . M C K E C H N I E and H . L . M A C L E A N , University of Toronto and S . S P A T A R I , Drexel University, USA

Abstract: Environmental life cycle assessment methods have become integral tools for developing and implementing renewable and low carbon fuel policies in several jurisdictions. We discuss the current state of knowledge on environmental implications, focused on greenhouse gas emissions, of lignocellulosic bioalcohol fuels as understood through a life cycle framework and identify areas for continued research to develop robust evaluation methods to support informed decision making and investment in emerging biofuel technologies. Understanding direct and market-induced effects and uncertainties across the life cycle of biofuels and, more broadly, their use in vehicles is critical for establishing sustainable biofuel markets. Key words: life cycle assessment, greenhouse gas emissions, lignocellulosic ethanol, bioethanol, biofuels.

14.1

Introduction

In spite of the many benefits offered by personal transportation vehicles, the current sector, in part due to its almost entire dependence on fossil fuels for propulsion, is unsustainable. This dependence results in the sector being responsible for a significant proportion of air pollutant and greenhouse gas (GHG) emissions as well as other negative impacts on environmental sustainability. Worldwide, transportation (all modes) accounts for approximately 1/5 of GHG emissions but this value is close to 1/3 in most industrialized countries (Lutsey and Sperling, 2008). As one component of a move in the direction of sustainability, the sector must undergo a decarbonizing transition. This transition is currently evidenced through policy efforts under development in many jurisdictions to implement low carbon and renewable fuel requirements. Such low carbon fuel policies, as they have come to be known, measure the carbon intensity of transportation fuels on a life cycle basis and are designed to encourage market innovation rather than having government select winners (Sperling and Yeh, 2009). In the future, these policies are planned to move more broadly toward sustainable fuel policies, those that include environmental,

ß Woodhead Publishing Limited, 2010

366

Bioalcohol production

social, and economic measures of sustainability (Spatari et al., 2008). Bioalcohols derived from lignocellulosic feedstocks utilizing biochemical processes have been reported to have much lower life cycle fossil and petroleum energy use and GHG emissions than conventional petroleum-derived gasoline and diesel (e.g., Sheehan et al., 2003; Larson, 2006; CONCAWE, 2007). Initiatives such as low carbon fuel standards, perhaps the most prominent of these being the California Low Carbon Fuel Standard (LCFS) and the United Kingdom's Renewable Transportation Fuel Obligation Programme (RTFO), incorporate life cycle-based standards into climate change regulations for transportation fuels (State of California Office of the Governor, 2007; UK Department of Transport, 2007). These standards require reductions in the life cycle carbon intensity of transportation fuels, based on estimates of CO2 equivalent (eq.) per energy unit of fuel on a life cycle basis. These regulations are the first to be based on systematic life cycle assessment (LCA) and will require that the life cycle GHG emissions (and potentially other sustainability metrics) associated with the production of a large set of transportation fuels be quantified. These regulations are `raising the bar' with respect to the quality and transparency required of LCAs of fuels, particularly biofuels as these options have been given considerable focus during the development of the standards due to their potential to assist in meeting the GHG emissions reduction requirements. LCA is an analytical tool which can be employed to examine the implications on the environment of the full set of activities (and their supply chains) associated with the production, use and end-of-life of a product, process or project. The International Organization for Standardization (ISO) publishes guidelines for the completion of LCAs (ISO, 2006). While ISO provides guidance, an analyst must make many judgments during the completion of an LCA. This chapter reviews key issues in evaluating the environmental performance of bioalcohols through the use of LCA. The chapter reviews the biofuels' LCA literature, presents LCAs of bioalcohols produced through biochemical conversion of lignocellulosic feedstocks, compares the environmental performance of these bioalcohols to other road transportation fuels, including other biofuels and conventional fossil fuels, and then compares the bioalcohols with alternative options for lignocellulosic biomass utilization, specifically electricity production. Insights are provided on aspects of the lignocellulosic bioalcohol life cycles that can enhance their environmental performance and finally, likely future trends in the field are discussed. Owing to the complexity of LCAs of biofuels and the local nature of biomass production, definitive results for biofuel pathways cannot be presented with certainty, rather the chapter outlines key issues that must be considered in developing and interpreting LCAs of biofuels with a focus on lignocellulose-derived bioalcohols.

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

14.2

367

Life cycle assessment (LCA) of biofuels

Effective decision making regarding biomass utilization, as one of many inputs, requires a life cycle approach. When life cycle methods are applied to fuels and vehicles as in the majority of biofuels studies, this is generally referred to as a well-to-wheel (WTW) analysis. This terminology was initially adopted because the first stage of the life cycle (resource extraction) for conventional fuels is the oil `well'. Although feedstocks for biofuels do not come from a well, the WTW terminology is commonly applied in the field and is utilized in this chapter. The fuel cycle stages from resource recovery through processing and the delivery of the fuel to the vehicle are referred to as the well-to-tank (WTT) component, the use (combustion) of the fuel in the vehicle is referred to as the tank-to-wheel (TTW) component and the entire fuel cycle is termed the WTW. In WTW analyses, it is customary to omit the accounting of energy and emissions associated with materials required for infrastructure (e.g., fuel production facilities), vehicle manufacture, non-fuel inputs over the vehicle life (e.g., maintenance), and end-of-life processes when retiring the vehicle: these would typically be included in a complete LCA. While a few life cycle-based studies and models of biofuels have included (or have the option to include, in the case of the models) activities beyond those included in WTW analysis (e.g., Hill et al., 2006; NRCan, 2008; ANL, 2009), by far the majority of studies of biofuels have been limited to WTW activities and this approach is therefore the focus of this chapter.

14.2.1 Activities included in WTW studies of biofuel production and use The following are the primary activities (each activity consists of several subactivities) involved in WTW studies of biofuel production and the use of the biofuel in a light-duty vehicle (e.g., automobile, light truck). 1. Feedstock production/collection. Biomass feedstock production and/or collection includes the activities and the inputs, outputs (discharges) and impacts associated with production and harvest of a crop, collection of a residue, or the cellulosic fraction of municipal solid waste. 2. Feedstock transport. Biomass transport includes the inputs, outputs and impacts associated with the use of a transport mode (e.g., truck, rail) to move the feedstock from the harvest/collection location to the conversion plant gate. 3. Biofuel production. Feedstock conversion to ethanol includes the inputs, outputs and impacts associated with processing the delivered feedstock into fuel grade biofuel. Any required fuel blending (e.g., denaturant) is usually included in this activity. 4. Biofuel distribution. Distribution includes the inputs, outputs and impacts from the distribution of the fuel from the fuel conversion facility to intermediate bulk storage and consumer refueling stations.

ß Woodhead Publishing Limited, 2010

368

Bioalcohol production

5. Biofuels use. Use considers the inputs, outputs and impacts due to use of the fuel in a light-duty vehicle, i.e., due to driving the vehicle.

14.2.2 Review of WTW studies of biofuels There is a several decade history of WTW studies of transportation fuels, beginning with analyses of conventional gasoline and diesel fuels, followed by studies of alternative fuels (for a review of studies up to 2001, see MacLean and Lave, 2003). Over the last decade numerous studies of biofuels have been completed, the majority considering starch, sugar and oil seed-derived (often referred to as `1st generation') fuels, such as ethanol derived from corn, and biodiesel from soybean or rapeseed in North American or European settings, respectively (see Larson, 2006 for a review). There have been fewer studies of biofuels produced from lignocellulosic biomass (often referred to as `2nd generation'), in large part because these feedstock/fuel `pathways' are not yet at commercial scale and reliable data are therefore limited in availability. This situation may change in the near future as the first commercial plants are expected to begin operation within the next few years (US DOE, 2009). Outside of North America and the European Union, a few studies have been completed in other developed countries such as Australia and Japan (e.g., Beer et al., 2002; Hayashi et al., 2008). Recently, a limited set of studies has been published of biofuels produced in developing and transition economies such as Thailand (e.g., Nguyen et al., 2007a; 2007b). Feedstock availability varies by location; feedstocks include dedicated herbaceous (e.g., switchgrass, miscanthus) and woody (e.g., hybrid poplar, willow) energy crops, agricultural residues (e.g. corn stover, wheat straw), forest and mill residues, and the cellulosic portion of municipal solid waste. Most of the lignocellulosic studies have modeled ethanol produced through biochemical conversion processes while fewer studies have been completed of other alcohols (e.g., methanol) or of lignocellulosic fuels produced through thermochemical conversion processes (Fleming et al., 2006). Key issues in WTW studies of biofuels In a review of LCAs of liquid biofuels for the transport sector (both 1st and 2nd generation), Larson (2006) discusses key observations in line with those presented in other studies (Quirin et al., 2004; Fleming et al., 2006; Tripp, 2007). The studies show wide ranges in net energy balance and GHG impacts among different biofuels and even for the same feedstock/biofuel pathway. Key issues, which in part explain these differences, include the wide range of plausible values for many of the input parameters in the feedstock and fuel conversion technologies (some variation in results is expected due to different geographic regions and inherent differences in the pathways) as well as

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

369

assumptions made in individual studies. Specific to GHG results, Larson (2006) notes four main contributors to variation in results among studies: 1) the climateactive species (direct species, CO2, CH4, N2O, and in some cases the indirect species, CO, NOx, and NMOC) included in the calculation of CO2 equivalence (eq.); 2) assumptions around N2O emissions (N2O is related to feedstock production and fertilizer application); 3) the allocation methods used for coproduct credits; and 4) soil carbon impact of feedstock production. Noteworthy for evaluating different lignocellulosic ethanol technologies is the co-product allocation procedure employed for the conversion facility because electricity is assumed as a co-product for most lignocellulose-based biochemical ethanol facilities (other co-products are also likely for future biorefineries). While many studies calculate an electricity credit based on the displacement/system expansion method by subtracting the energy and related emissions displaced through conventional routes (e.g., the electricity grid) from the energy and emissions associated with ethanol production, others allocate the life cycle inventory burdens by the products' mass or energy contents (Spatari et al., 2005), and some studies do not employ co-product allocation at all, employing the most conservative assumption of allocating all energy and emissions to the biofuel (Woods and Bauen, 2003). Based on our comparisons of biofuels' LCAs, we would add to Larson's list the importance of the boundary definition (what activities are included in a study) as this can vary among analyses. For example, many studies have not included the process chemicals and enzymes needed for biochemical conversion (MacLean and Spatari, 2009). The European Commission Joint Research Centre (JRC), Oil Companies' European Association for Environment, Health and Safety in Refining and Distribution (CONCAWE), and European Council for Automotive Research & Development (EUCAR) (CONCAWE, 2007) also compared biofuels' WTW results, focusing primarily on commercial ethanol production pathways through use of the feedstocks sugar beet, sugar cane, and wheat; however, they also investigated the lignocellulosic feedstocks wheat straw, wood, and wood waste. The study focuses on near-term technologies (ca. 2010), whereas Larson (2006) reviews some longer-term studies. Lignocellulosic biofuel WTW studies have focused on energy consumption, GHG emissions and to a lesser extent, criteria air pollutant emissions (e.g., MacLean and Lave, 2003; Spatari et al., 2005; Hill et al., 2006; NRCan, 2008; ANL, 2009). Far fewer studies have considered other environmental metrics such as human health impacts, land and water use, and impacts on biodiversity although some recent studies have made progress on some of these metrics (e.g., Sheehan et al., 2003; Hill et al., 2009). The quality of published LCA studies has varied, although on average it has improved over the past decade with advances in theory, science, available data and attention to more transparent and complete documentation.

ß Woodhead Publishing Limited, 2010

370

Bioalcohol production

Results of WTW studies of lignocellulosic ethanol While results have varied, in comparison to crude oil-derived gasoline and corn ethanol, the studies generally agree that lignocellulosic ethanol can reduce fossil fuel use, petroleum use, and GHG emissions across the life cycle and can potentially reduce emissions of criteria pollutants such as CO, PM, and NOx, (Sheehan et al., 2003; Spatari et al., 2005; Wu et al., 2006; NRCan, 2008; ANL, 2009; Hill et al., 2009). For example, CONCAWE (2007) found reductions in WTW fossil energy use (67% to 93%) and GHG emissions (64% to 92%) for lignocellulosic ethanol relative to conventional gasoline measured on a per kilometre basis. The work of Sheehan et al. (2003) likely remains the most detailed lignocellulosic ethanol WTW study, particularly as it pertains to the examination of the feedstock, corn stover, (an agricultural residue) and the biochemical conversion technology developed by the US National Renewable Energy Laboratory (NREL). Most WTW studies have modeled either NREL technology (e.g., Sheehan et al., 2003; Spatari et al., 2005; ANL, 2009) or Iogen technology (Levelton, 1999; CONCAWE, 2007). A larger set of potential technologies, consisting of different pretreatment methods, and hydrolysis and fermentation orientations, are under development for converting lignocellulose to ethanol and studies are underway to evaluate these (e.g., Spatari, 2007). Wu et al. (2005) studied advanced lignocellulosic conversion technologies for the mid-term (ca. 2030) along with different scenarios for biorefineries that produce transportation fuels (ethanol, Fischer-Tropsch diesel, and dimethyl ether), electricity, and animal feed, all using switchgrass as a feedstock. The ethanol conversion technology they examined uses one pretreatment method in combination with advanced hydrolysis-fermentation technology. Wu et al. (2005) concluded that petroleum use and GHG emissions were reduced by 68± 93% and 82±87%, respectively, and criteria pollutants were also reduced or unchanged (in the case of CO) through the use of the biofuels examined. The authors noted, however, that substantial petroleum and GHG savings are due in large part to co-producing electricity through advanced integrated gasification combined cycle technologies. Kim and Dale (2003) also considered life cycle aspects of products and fuels derived from agricultural sources in a conceptual future biorefinery, which makes use of a variety of agricultural biomass feedstocks to produce fuels, chemicals, proteins, and plastic materials. The authors argue that the biorefinery holds the potential to reduce the US dependence on foreign petroleum sources, reduce the consumption of fossil fuels in North America and finally, reduce the resulting GHG and other emissions from the consumption of fossil fuels. In spite of the attractiveness of the majority of the WTW studies' results for lignocellulosic biofuels, due to the lack of commercial production and that the feedstocks are not currently produced for this purpose, the specific results and

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

371

assumptions made vary considerably. In addition, recently, a small number of studies have shown less attractive performance of biofuels (both 1st generation and lignocellulosic) where the feedstock production has been reported to result in land use change (LUC) and release ecologically-stored carbon to the atmosphere (Searchinger et al., 2008, Fargione et al., 2008). Ecological carbon storage is significant within the global carbon cycle. While the combustion of fossil fuels is responsible for releasing approximately 270 Pg C to the atmosphere from the beginning of industrialization to the end of the 20th century, LUC, largely due to the conversion of forest and native grassland to intensive agriculture, is responsible for emitting 136 Pg C during the same period (Lal, 2003). Although attention to LUC implications of biofuel production has intensified as of late, some life cycle analysts have long been aware of the potential consequences of LUC (e.g. Deluchi et al., 1987). Deluchi (1991) first incorporated LUC-associated GHG emissions into a biofuel life cycle study. Two classifications of LUC exist. Direct LUC refers to a change in the use of a parcel of land, for example, by switching crops on agricultural land or introducing previously unmanaged land to a productive use. Cultivating energy crops for biofuel production can either increase or decrease the quantity of carbon stored within agricultural ecosystems, depending on the selected energy crop, ecosystem characteristics, agricultural management practices, and the crop that is displaced. Indirect LUC occurs when a given market commodity, for example corn used for producing a biofuel, increases its demand for land, thereby inducing an increase in market price for the agricultural commodity (corn) as well as for other commodities (e.g., soy), thereby inducing additional land to be put into cultivation to meet the demand for all agricultural commodities in the global economy. The conversion of forest and native land into agricultural use releases carbon to the atmosphere from carbon stores, and this release has been termed indirect land use change (iLUC). Given the global nature of the economy, iLUC could occur in entirely different regions of the world from where the initial crop substitution took place, thus establishing accurate relationships between local decisions to produce biofuel feedstock and iLUC impacts is challenging. Searchinger et al. (2008) employed partial equilibrium economic modeling and historical land clearing patterns to estimate the iLUC of biofuel production in the United States, finding LUC emissions to outweigh fuel substitution benefits of both corn and switchgrass-derived ethanol over a period of 50 years. There has been significant debate on this issue (e.g. Wang and Haq, 2008) and researchers have begun rebutting the work of Searchinger et al. (e.g., Kim et al., 2009). Research on iLUC change is ongoing (e.g., Birur et al., 2008) and the issue remains to be resolved. Moreover, it is important to examine market-mediated effects (including iLUC), when any new fuel or technology is introduced. Examining those market-mediated effects requires using economic models such as partial equilibrium or computable general equilibrium models as well as conventional WTW/LCA models. Research is underway which is

ß Woodhead Publishing Limited, 2010

372

Bioalcohol production

examining how the economic and life cycle models could be integrated, or at least as a starting point, how the results of the economic models could be integrated into WTW models (for more details see Section 14.7). The significant uncertainties in both types of models as well as the complexity of the issues make this a very challenging research area which will take considerable time to develop, verify and validate models. Life cycle-based studies of alternative applications of lignocellulosic biomass Comparing LCAs of biofuels with those of alternative uses of biomass, e.g. cofiring with coal in electricity generating stations and dedicated biomass electricity generation, is important for determining the relative environmental merit of the different pathways. These types of analyses are critical for informing policy decisions regarding efficient and effective uses of biomass resources. There have been far fewer studies that have compared these disparate options. The following studies include comparisons of biobased liquid fuels with stationary uses of biomass; Elsayed et al. (2003), Greene (2004), Larson (2006), CONCAWE (2007), and Zhang et al. (2007). One key issue to note in the comparisons and the GHG mitigation potential is the carbon content of the baseline fuel that is being displaced, e.g., coal vs. natural gas vs. gasoline. The higher the carbon content of the fuel being displaced, all other things equal, the larger the GHG mitigation benefit with a biobased resource. Larson (2006 p.121) reports, `it is difficult to make unequivocal statements regarding the relative GHG-mitigating merits of different biomass applications without specific case comparisons.' As discussed earlier, the complexity of biofuel systems and their site-specific nature makes it risky to generalize their environmental performance.

14.3

Life cycle assessment (LCA) of biochemical lignocellulosic alcohol production

Ethanol has received by far the most attention of any alcohol to be produced through biochemical conversion and therefore, we limit our discussion in this section to this fuel. Feedstocks vary by region and in their properties (e.g., cellulose, hemicellulose, and lignin contents, energy density). Because of structural and compositional differences between feedstocks, status of conversion technology development, economics, and preferences of the developer, a set of distinct pretreatment, enzymatic hydrolysis and fermentation methods is currently under development. There is uncertainty as to which feedstocks and technologies will be used in the first set of commercial facilities that convert lignocellulosic feedstock to fuels and no consensus on preferred feedstockconversion technology pathways in the literature (Spatari, 2007). In this section we present results for lignocellulosic ethanol biochemical conversion pathways from the US Department of Energy Argonne National

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

373

Laboratory's Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) model (version 1.8b) (ANL, 2009). This model has been widely utilized and cited in the literature (see, for example, Kim et al., 2009 and Searchinger et al., 2008) and is therefore utilized to highlight some key issues in WTW modeling of lignocellulosic ethanol. GREET models the NREL conversion technology which consists of dilute acid pretreatment followed by enzymatic hydrolysis. The feedstocks for the lignocellulosic biochemical pathways included in GREET and discussed in this chapter are: · `farmed tree', which represents a dedicated woody energy crop (hybrid poplar), · `herbaceous biomass', which represents a herbaceous energy crop (switchgrass), and, · corn stover, an agricultural residue. For details on the specific pathways see ANL (2009). The model was run for the year 2010 and for US national average cases, with default values. GREET analyzes energy use (total, fossil, renewable), GHG emissions, and criteria air pollutant emissions. For our analysis, we limit the discussion to GHG emissions as energy analysis is considered in another chapter of this volume. In this section we examine ethanol well-to-ethanol plant exit gate (WTG) results (biomass production and fuel processing stages), allowing for a more detailed comparison of the fuel production. In the next section, we include as well the distribution of the ethanol and its use in the vehicle. Results are presented on the basis of one MJ of ethanol produced, which is referred to as the functional unit in LCA terminology. The quantitative results in this chapter do not include iLUC or emissions associated with the production of chemical or enzyme inputs required for ethanol conversion as these components are not currently included in GREET. Both of these components may be fairly significant contributors (Searchinger et al. 2008; MacLean and Spatari, 2009, respectively) and require more attention in biofuel LCA studies. The WTG GHG emissions associated with the production of ethanol through the farmed tree pathway are shown in Fig. 14.1. The results are broken down by life cycle activity. The WTG activities are disaggregated into feedstock production (emissions occurring during cultivation, harvest, and transport of biomass to a conversion facility), fertilizer production and application, direct LUC, ethanol conversion (emissions occurring during biomass pretreatment, fermentation, and purification stages) and co-product credit (calculated as the GHG emissions associated with the production of a commodity/service that is displaced by the co-product of ethanol conversion). The WTG emissions are shown both including and excluding the co-product credit. The co-product of the ethanol production is electricity. GREET utilizes the displacement method discussed above to determine this credit and assumes that the electricity produced by the facility offsets that which would otherwise be produced by the US national

ß Woodhead Publishing Limited, 2010

374

Bioalcohol production

14.1 Well-to-gate GHG emissions associated with the production of lignocellulosic ethanol from farmed trees through biochemical conversion. Adapted from GREET 1.8b (ANL 2009). Notes: N2O (N fert.) refers to nitrogen fertilizer induced N2O emissions as the result of microbial nitrification and denitrification in the soil. Land use change refers to direct land use change. WTG ˆ well-to-ethanol plant exit gate. Without credit is the value without including co-product credits. With credit is the value including co-product credits.

average grid mix. Applicable to all pathways modeled here, differing treatment of co-products can significantly impact WTW results (see Hill et al., 2006). Owing to the carbon-intensive nature of the US mix (51% coal), a considerable co-product credit results and the WTG results are more attractive with the credit (ÿ15.7 g CO2 eq./MJ compared to ÿ5.5 g CO2 eq./MJ of ethanol produced). The negative values signify sequestration associated with ethanol production as opposed to the release of GHGs. The feedstock production (excluding the N fertilizer component) is responsible for about 6 g CO2 eq., mostly resulting from the combustion of the diesel fuel in harvest machinery and transportation vehicles. Fertilizer requirements for trees are low and therefore do not result in significant emissions. The direct LUC component represents the estimated change in soil carbon that would occur due to cultivation of the woody energy crops. The assumption of a significant increase in soil carbon (112 500 g CO2/dry ton of farmed tree or 15.5 g CO2 eq/ MJ of ethanol) is a major factor in this pathway having the best WTG GHG emissions performance of the three lignocellulosic ethanol pathways modeled in GREET. The assumption in GREET is that the land utilized for tree farming is converted from land currently idle or used as pastureland. There are differing views on whether this carbon credit is appropriate for farmed tree soil carbon sequestration. In modeling ethanol produced from farmed trees, the California

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

375

Air Resources Board (CARB, 2009, p. 17) did not consider this benefit, and indicated that this issue would be reviewed in the future. Ethanol conversion results in a relatively modest amount of GHG emissions, in large part because the process modeled assumes that much of the process energy is provided by the lignin portion of the biomass feedstock. The WTG GHG emissions associated with the herbaceous biomass ethanol pathway are presented in Fig. 14.2. Compared to the farmed tree pathway, the feedstock production emissions are slightly higher, but more so, the emissions associated with N fertilizer application are much larger. These emissions result due to the larger nitrogen fertilizer requirements for the herbaceous biomass. The impact of fertilizer usage is emphasized due to the global warming potential of nitrous oxide (298 compared to 25 for methane and 1 for CO2), which is the primary emission that results from nitrogen fertilizer application. The land use change component for this pathway is less than half that of the farmed tree pathway due to lesser accumulation of carbon in the soil and the root structure by herbaceous crops. The co-product credit for electricity production is half that of the farmed tree pathway due primarily to the lower lignin content of herbaceous biomass. The combination of higher fertilizer usage, less carbon accumulation and lower co-product credit results in net GHG emissions for the herbaceous ethanol pathway of 9 g CO2 eq./MJ (with credit) and 14 g CO2 eq./ MJ of ethanol (without credit).

14.2 Well-to-gate GHG emissions associated with the production of lignocellulosic ethanol from herbaceous biomass through biochemical conversion. Adapted from GREET 1.8b (ANL 2009). Notes: See Fig. 14.1.

ß Woodhead Publishing Limited, 2010

376

Bioalcohol production

14.3 Well-to-gate GHG emissions associated with the production of lignocellulosic ethanol from corn stover through biochemical conversion. Adapted from GREET 1.8b (ANL 2009). Notes: See Fig. 14.1.

The final biochemical ethanol pathway is the conversion of corn stover (Fig. 14.3). The `feedstock production' activity consists primarily of the collection of the stover, which is assumed to be collected in a second pass of the field, following collection of the grain (equipment is under development for one pass harvesting where the corn and stover would be collected together during a single pass of the field). Most stover is not currently removed from the field but left to provide nutrients and to lessen erosion of the soil. There are differing views on the amount of stover that can be removed from a field without negative consequences. In general terms, the allowable removal percentage varies depending on soil type, tillage, removal practices, etc. In most LCA studies, the nutrient value (or a portion of it) in the stover removed has been assumed to need to be replaced through the application of fertilizer. The emissions associated with the fertilizer production and application are included in GREET. Partially offsetting these emissions are the N2O emissions that would have resulted if the stover had been left on the field. There is no LUC component associated with the stover as it is assumed in GREET that the removal of stover does not result in any change in soil carbon. The co-product credit for electricity production is identical in this pathway to that of the herbaceous biomass pathway due to the similar lignin contents of the feedstocks. The stover WTG results (5 g CO2 eq./MJ (with credit) and 10 g CO2 eq./MJ (without credit)) fall in between those of the farmed tree and herbaceous ethanol pathways but are much closer to the latter.

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

377

From the above discussion, many of the issues reported in the literature are those that impact the GHG emissions results of the lignocellulosic ethanol pathways discussed above. Of particular note are: the assumptions around fertilizer application and associated N2O emissions; co-product credits; and the impact of LUC on carbon storage. There remain gaps in the literature with respect to these issues and others (e.g., process chemical and enzyme inputs) and further research will improve our understanding of these aspects and benefit future LCA studies.

14.4

Comparison of lignocellulosic alcohol biofuel life cycle assessments (LCAs) with those of other fuels

LCA/WTW results are most informative when they are utilized in a comparative/ relative sense. However, it remains difficult to do inter-study comparisons due to different study boundaries, geographic locations, and software computational algorithms combined with the complexity, and sometimes lack of transparency, of the studies. For these reasons and for consistency with the previous section, we present results of the GREET model and compare the lignocellulosic ethanol pathways discussed above with other feedstock/fuel pathways to examine the relative GHG performance. The comparison fuels are fossil fuels (conventional petroleum gasoline (26 ppm sulfur, no oxygenate) and diesel (200 ppm sulfur) and compressed natural gas (CNG)), as well as 1st generation biofuels (corn ethanol, corn butanol, Brazilian sugarcane ethanol, soybean biodiesel, and soybean renewable diesel) and ethanol derived from forest residues (2nd generation pathway) through thermochemical conversion (gasification). The soybean renewable diesel is produced through the UOP process which employs petroleum refining technology to convert soy oil to a useable fuel rather than the conventional trans-esterification route for producing biodiesel (Huo et al., 2009). With the exception of the Brazilian sugarcane ethanol, all other fuels are modeled based on US data. The fuels are assumed to be utilized in dedicated vehicles which are able to utilize neat fuels. All vehicle data are taken from GREET. With the exception of the conventional gasoline, diesel and CNG vehicles, none of the vehicles (e.g., a vehicle able to run on 100% ethanol, E100) is sold in North America. Comparing the WTW performance of the lignocellulosic ethanol biochemical pathways, the farmed tree ethanol, due to its superior WTG performance as discussed in the prior section has the best WTW performance (see Table 14.1 and Fig. 14.4). Due to the carbon sequestration associated with direct LUC and the co-product credits, the WTW emissions are negative. The herbaceous biomass ethanol and corn stover ethanol pathways result in positive, but low GHG emissions. In comparing the performance of these fuels with the set of conventional and alternative fuels, the following observations are made. We do

ß Woodhead Publishing Limited, 2010

378

Bioalcohol production

Table 14.1 Well-to-wheel GHG emissions associated with the lignocellulosic ethanol and alternative fuel pathways. Adapted from GREET 1.8b (ANL 2009) Greenhouse gas emissions (g CO2 eq./km driven)

WTT

TTW

WTW

Conventional gasoline Farm. tree ethanol Herb. ethanol Stover ethanol Forest residue ethanol Corn ethanol Corn butanol Br. sugarcane ethanol Conventional diesel Soybean biodiesel Soybean renewable diesel CNG

58 ÿ248 ÿ173 ÿ185 ÿ142 ÿ17 39 ÿ158 42 ÿ196 ÿ108 58

241 220 220 220 220 220 229 220 205 209 201 199

300 ÿ28 47 35 78 203 268 62 247 13 93 257

Notes: WTT ˆ well-to-tank,TTW ˆ tank-to-wheel, WTW ˆ well-to-wheel, gasoline vehicle fuel economy is 10.1L gasoline equivalent (Lge)/100 km (23.2 mpgge), ethanol vehicles are E100 dedicated vehicles with 9.5 Lge/100 km (24.8 mpgge), butanol vehicle is B100 dedicated vehicle with 10.1Lge/100 km (23.2 mpgge), diesel vehicles have 8.5 Lge/100 km (27.8 mpgge) and are able to use 100% conventional, biodiesel or renewable diesel. CNG vehicle has 10.7 Lge/100 km (22 mpgge). Farm. tree ethanol ˆ ethanol produced from farmed trees, Herb. ethanol ˆ ethanol produced from herbaceous lignocellulosic biomass, Forest res. ethanol ˆ ethanol produced from forest residues, Br. ˆ Brazil, Soybean renewable diesel ˆ diesel produced from soybean oil using the UOP process, CNG ˆ compressed natural gas. All co-product credits based on displacement method with the exception of forest residue-derived ethanol based on thermochemical process which assumes energy basis for co-product allocation. The conventional gasoline and diesel pathways have been modified so that they do not include a Canadian oil sands component, a minor adjustment. Reaccumulation of CO2 in biomass regrowth is credited to the WTT stage, resulting in negative emissions for that stage for many of the biofuel pathways. Adapted from GREET1.8b (ANL 2009).

not elaborate on the performance of all of the comparison fuels as that is beyond the scope of the chapter and has been discussed elsewhere (e.g., MacLean and Lave, 2003). In line with prior studies, the majority (77±83%) of WTW GHG emissions from the conventional gasoline, diesel and CNG internal combustion engine (ICE) vehicles result from the operation of the vehicle (the TTW portion of the life cycle which consists of the combustion of the fuel in the vehicle). The WTW emissions for these pathways range from 247 to 300 g CO2 eq./km driven. While the TTW emissions for the biofuels are similar in magnitude to those of the fossil fuels, due to the assumption that the CO2 resulting from the combustion of the biofuel is offset by carbon incorporated during regrowth of the biomass feedstock (e.g., switchgrass) within the time period considered, the resulting WTW emissions are small, sometimes negative. The corn-derived fuels (ethanol and butanol) have the poorest GHG emissions performance of all of the biofuels as a result of the fossil fuel inputs

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

379

14.4 Well-to-wheel GHG emissions associated with the lignocellulosic ethanol and alternative fuel pathways. Adapted from GREET 1.8b (ANL 2009). Note: Pathway abbreviations as in Table 14.1.

which are utilized throughout the life cycle of their production (fertilizer inputs, diesel utilization during harvest and transportation, and fossil fuel used in the conversion process). The Brazilian sugarcane ethanol process has comparatively low emissions due in large part to the high yields of sugarcane, the less intensive farming practices in Brazil, and the use of bagasse (sugarcane residue) for process energy during conversion. The soybean-derived diesel pathways also have fairly low GHG emissions due primarily to the low nitrogen fertilizer requirements resulting from the nitrogen fixing nature of soybeans and the relatively low fossil energy inputs to the conversion process. As has been common practice to date in WTW studies, GREET does not report estimates of land area needed to support fuel production. The evaluation of GHG (and other) impacts, however, on a land area basis offers important insights (Larson, 2006) and therefore we utilize data within GREET to represent the fuel options on this basis. The amount of fuel/energy that could be produced from a hectare of land each year and the associated GHG emissions avoided based on land area are shown in Figs 14.5 and 14.6, respectively (only land utilized for feedstock production is included, not that required for the conversion facility or other infrastructure). While the issues are also of importance for conventional fuels, we limit our discussion here to a comparison of the biofuels which are generally suggested to entail greater land area requirements than fossil fuels. The avoided emissions are calculated relative to a reference pathway, petroleum gasoline or diesel, the latter for the biobased diesel fuel options.

ß Woodhead Publishing Limited, 2010

380

Bioalcohol production

14.5 Energy yield per agricultural land area for biofuels. Adapted from GREET 1.8b (ANL 2009).

While the farmed tree ethanol has the lowest WTW GHG emissions (per km driven ± see Fig. 14.4) of the biochemical lignocellulosic pathways, on a land area basis the herbaceous biomass ethanol has the highest energy yield (115 Gigajoules per ha per year (GJ/ha-y)) and slightly better GHG avoided performance (10.6 Mg CO2 eq./ha-y vs. 9.6 Mg CO2 eq./ha-y). Of all of the biofuels examined, the Brazilian sugarcane ethanol has the highest energy yield and the highest GHG avoided on a land area basis, slightly greater than the herbaceous biomass ethanol. The soybean diesel pathways have the lowest energy yield and poor GHG emissions avoided performance on a land area basis, in spite of their

14.6 GHG emissions avoided with biofuels compared to gasoline or diesel per agricultural land area required for biofuel feedstock production. Adapted from GREET 1.8b (ANL 2009). Note: Avoided emissions associated with soybean biodiesel and renewable diesel are calculated based on petroleum diesel reference. All other emissions avoided are based on petroleum gasoline reference.

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

381

relatively good GHG emissions performance on a WTW basis (per km driven). This results primarily from the soybean's relatively low yields (approximately 2.5 Mg/ha-y in the US) and oil seed content which is among the lowest of all oil seed crops (approximately 18%). A comparison of oil seed yields and content is presented in Tripp (2007). The above examines pathways utilizing default values estimated in GREET for the year 2010 and results should be interpreted with this in mind. Different agricultural feedstock yields and conversion yields will impact these results. In addition, there are expected to be tradeoffs in some cases between agricultural and conversion inputs (e.g., more intensive farming practices may result in higher yields but higher emissions as well) that should be examined.

14.5

Comparison of life cycle studies of lignocellulosic bioalcohols with those of alternative biomass utilization

There are few studies which examine alternative uses for biomass, although this is critical from a research perspective and for informing industry decision making and public policies. We developed life cycle models for lignocellulosic ethanol production and also the production and use of wood pellets and agricultural residues for use in electricity generation in Ontario, Canada (Spatari et al. 2005; Zhang et al., 2007; 2010). The wood pellets and agricultural residues were assumed to be co-fired with coal in existing coal generating stations. In addition, we examined the use of 100% wood pellets in the coal-fired boilers in the generating stations. The analyses are case studies and represent conditions in the province; however, in comparisons with analyses of other locations, the trends in results are similar (Zhang et al., 2010). The pathways studied include: · Ethanol produced from corn stover through biochemical conversion and utilized as E85 (85% ethanol, 15% gasoline blend by volume) in an ICE automobile (9.0 liters gasoline equivalent (Lge)/100 km, 26.1 miles per gallon gasoline equivalent (mpgge). · Ethanol produced from switchgrass through biochemical conversion and utilized as E85 in an ICE automobile. · Electricity produced through co-firing wood pellets with coal in existing coal generating stations. · Electricity produced through 100% wood pellet utilization in existing coal generating stations. · Electricity produced through co-firing agricultural residues with coal in existing coal generating stations. The amount of GHG emissions that could be avoided by utilizing one tonne of biomass in each of the above applications is shown in Fig. 14.7. Utilizing

ß Woodhead Publishing Limited, 2010

382

Bioalcohol production

14.7 GHG emissions avoided with alternative uses of biomass: Transportation fuels (ethanol) and electricity. Adapted from: Spatari et al. (2005), Zhang et al. (2007, 2010). Note: Avoided emissions associated with E85 are calculated based on petroleum gasoline reference. Avoided emissions for electricity pathways are based on coal electricity generation reference.

biomass to displace coal for electricity generation provides greater GHG emissions reductions than displacing petroleum with biomass-derived fuels on the basis of one oven dry tonne (ODT) of biomass utilized. The higher carbon content of the fuel being displaced (coal) in the case of the electricity generation as well as the lower processing requirement (and associated lower GHG emissions) for the agricultural residue and wood pellet pathways lead to their more promising results. The GHG emissions avoided are almost three times those of the transportation pathways (1340 to 1575 vs. 450 to 530 kg CO2 eq./ ODT of biomass). The 10% wood pellet co-firing pathway results in slightly more GHG emissions avoided than the 100% wood pellet option per unit of pellets due to the relatively higher efficiency of the co-firing in the coal generating stations. The 10% co-firing of agricultural residues has slighter higher GHG avoided than the 10% co-firing with wood pellets because of the lower GHG emissions associated with the life cycle of the agricultural residues, and that in the wood pellet production a portion of the biomass harvested is used as process fuel for drying during pelletization and is not used directly for electricity generation. While the advantages of the electricity generation pathways are shown in the analysis, there are arguments in support of the biofuels pathways as well. The liquid biofuels may provide greater overall benefits than electricity because there are alternative means of generating low carbon electricity but no alternative relatively moderate cost low carbon transportation fuels that are com-

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

383

patible with existing infrastructure and vehicles. Despite lesser GHG emission reductions attainable by utilizing biomass in producing transportation fuels, these pathways may be a more effective use of the resource in attaining economy-wide reductions in GHGs. The economic benefit of producing liquid fuels may also be greater. Displacing petroleum-based transportation fuels would reduce the dependence of many countries on imported petroleum products or, alternatively, increase export possibilities. To support research and development as well as policy decision making, further life cycle-based studies of alternative biomass uses, not just limited to fuels and electricity but also including other biobased products, should be conducted.

14.6

Routes for environmental improvement

There are steps that can be taken throughout the life cycle of biofuel production and use to enhance environmental performance. Below we highlight some key issues along the life cycle.

14.6.1 Feedstock production improvements Indirect LUC impacts can be avoided by ensuring that biofuel feedstocks are not grown on land utilized for food and livestock feed production. Sustainable biomass certification could be employed to partially address this issue. Direct LUC is less of an issue for lignocellulosic ethanol pathways since the production of these feedstocks typically either increases carbon storage in the agroecosystem (farmed tree biomass and perennial herbaceous biomass) or residues are utilized from traditional agricultural production (corn stover) with allowance for returning a portion of the residue to sustain soil organic carbon. The overall GHG performance of producing agricultural and energy crops could be improved by adopting more sustainable agricultural processes. No-till agriculture has been proposed as an effective method to increase carbon storage in soils (for example, Smith et al., 1998), although some uncertainty exists as to whether this does in fact provide benefits when spatial variability and carbon storage in deeper soil layers are considered (VandenBygaart et al., 2002; Baker et al., 2007) and no-till cannot be universally employed. Despite uncertainties regarding soil carbon impact of no-till, adopting this practice would reduce onfarm fuel use. Winter cover crops, planted following the harvest of an annual crop, offer an additional practice to increase soil carbon in agriculture (Kim et al., 2009). Moreover, some researchers have found that other biofuel conversion technologies, such as biomass pyrolysis with biochar production for return to soil rather than combustion to electricity, offer better opportunities for restoring soil organic carbon and greater avoided carbon emissions than if used to displace fossil energy (e.g., electricity) (Gaunt and Lehmann, 2008). Biomass productivity and the impact of cultivation on soil carbon are site-specific. Improve-

ß Woodhead Publishing Limited, 2010

384

Bioalcohol production

ments in the environmental performance of lignocellulosic fuels could be achieved by matching species selection with local site characteristics to optimize productivity and soil carbon accumulation. Development of high yielding, low input feedstocks with attractive properties for conversion to biofuels is expected to continue to be investigated. Beyond ecosystem carbon storage, environmental performance of biofuels could be enhanced by introducing more sustainable agricultural practices. Integrating nutrient application with global positioning system (GPS) measurements and soil sampling can help ensure optimal quantities are applied, reducing the excess typical of current agricultural practice (Smil, 2001). Proper management of the timing, rate, application method, and source of nitrogen fertilizer can help minimize nitrogen loss, leading to high crop yield, high profitability and low emissions to air and water (Scharf et al., 2006). Diesel-fueled farm equipment and transportation vehicles could be designed/modified to utilize low carbon, low impact fuels and to be more fuel efficient. More efficient crop production/ collection would require less energy use and lower the negative impact on ecosystems (e.g., less soil erosion).

14.6.2 Lignocellulosic biomass conversion to bioalcohols Developing specialized processes for different feedstock properties will improve the overall efficiency and yield of lignocellulosic alcohol conversion. Novel conversion methods, potentially combining biochemical and thermochemical processes, offer a range of possibilities for significant improvements in biomass conversion in biorefineries likely to be constructed over the next decades. The production of chemicals and enzymes utilized in conversion of biomass to biofuels is often not included in life cycle studies. Technological advances to integrate chemical, enzyme, and N-fertilizer (or nutrient) production within a biorefinery could offer opportunities for reducing operating costs as well as lowering emissions during process conversion. As well, continued research to reduce chemical and enzyme loadings can improve the environmental performance of biofuel systems and improve their economic viability. Enzyme production in particular offers the potential for improvement with the current rapid development of biological sciences (Schoemaker et al., 2003).

14.6.3 Public policy issues Policies aimed solely at supporting the production of liquid transportation fuels from biomass exclude other potential uses of biomass to achieve GHG reductions or to meet other decision maker goals. Such policies also risk ignoring the multitude of impacts that biofuel production might have on a surrounding ecosystem due to their narrow scope, including but not limited to carbon storage, water availability and pollution, and biodiversity. Policies must consider issues

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

385

beyond the liquid biofuel system to encourage optimal decisions regarding land and resource utilization. Issues such as the viability of alternate emission mitigation methods in sectors that could make use of biomass resources, the economic and social implications of these utilization options, and the ecological impacts of various biomass utilization strategies, as well as broader sustainability implications, must be included in the decision-making process.

14.7

Future trends

In this section we discuss sources of life cycle information for biofuels, recent life cycle-related initiatives and future research directions to support improved life cycle environmental assessment with respect to lignocellulosic bioalcohols. The most comprehensive life cycle-based models specific to transportation fuels and their use in vehicles are: · United States: GREET developed by the US Department of Energy's Argonne National Laboratory. Available at: http://www.transportation.anl.gov/ modeling_simulation/GREET/index.html and LEM (Life Cycle Emissions Model) developed by Mark Delucchi of University of California at Davis (http://www.its.ucdavis.edu/people/faculty/delucchi/index.php). · Canada: GHGenius maintained by Natural Resources Canada. Available at: http://www.ghgenius.ca · European Union: Joint Research Centre (JRC) model and associated research in collaboration with EUCAR and CONCAWE (see ies.jrc.ec.europa.en/WTW). In addition to the above groups, many others worldwide are moving forward in developing and applying life cycle methods to biofuel pathways. From a regulatory perspective, as noted earlier, low carbon fuel initiatives such as the UK's Renewable Fuel Transportation Obligation (http://www.dft.gov.uk/pgr/ roads/environment/rtfo), California's Low Carbon Fuel Standard (http:// www.energy.ca.gov/low_carbon_fuel_standard/index.html), a proposed US national low carbon fuel standard and those in other jurisdictions (e.g., various US states and Canadian provinces) will require that life cycle emissions performance be determined for a large set of transportation fuels. It should also be noted that the United States Department of Energy's call for proposals for `Demonstration of Integrated Biorefinery Operations' requires a life cycle study of GHG emissions associated with the proposed process be included with each application (http://www1.eere.energy.gov/biomass). The Energy Independence and Security Act of 2007 (EISA) establishes new renewable fuel categories and eligibility requirements. EISA sets mandatory life cycle GHG reduction thresholds for renewable fuel categories. The regulatory purpose of the life cycle GHG emissions analysis is to determine whether renewable fuels meet the GHG thresholds for the different categories of renewable fuel.

ß Woodhead Publishing Limited, 2010

386

Bioalcohol production

From a research perspective, there are many topics which are undergoing further study to improve life cycle environmental assessments of biofuels. These include: · The majority of LCAs have focused on GHG emissions, energy use and air pollutant emissions. Future studies will consider a broader set of metrics including land and water use, discharges to air, land and water, as well as improving impact assessment, moving in the direction of quantifying impacts on human and ecosystem health. · Impacts of iLUC. The Global Trade Analysis Project (GTAP, https:// www.gtap.agecon.purdue.edu/default.asp), based in the Department of Agricultural Economics at Purdue University, is developing a global computable general equilibrium model that could be used to estimate the market forces resulting from the diversion of agricultural land and products to biofuel production. We are aware as well that there are other initiatives on iLUC taking place in the US and Europe (e.g., http://www.europarl. europa.eu/sides/getDoc.do?pubRef=-//EP//TEXT+TA+P6-TA-20080613+0+DOC+XML+V0//EN). Models such as GTAP or the Forestry and Agriculture Sector Optimization Model with Greenhouse Gases (FASOMGHG) developed at Texas A&M University are becoming more critical inputs into evaluations of the economic and environmental impacts of new biofuel policies. They are expected to be incorporated into life cycle/ WTW models estimating GHG emissions implications of iLUC for low carbon and renewable fuels policies under development worldwide (e.g., Gallagher, 2008); most earlier government initiatives only acknowledged the importance of direct LUC-related GHG emissions (e.g., Cramer et al. 2007). As noted earlier in the paper, there are many challenges in this integration and the incorporation of market-mediated effects into LCA studies and biofuels regulation more generally. Resolution of many sources of uncertainty in current iLUC studies will not be possible before policy decisions regarding the continuation or expansion of biofuel programs are made. It will be critical that uncertainties in the models are explicitly treated and transparently presented, and that decisions with implications on stakeholders be made with a balanced perspective. · Uncertainty in quantifying soil carbon accumulation/depletion further impairs the ability to accurately assess LUC impacts. Data collection to allow for more reliable means of assessing the carbon storage potential of lands under different use scenarios and production techniques is being undertaken at US Geological Survey Soil Carbon Research (http://carbon.wr.usgs.gov/) and the Integrated Sink Enhancement Assessment (INSEA) project for the European Commission Joint Research Centre (http://eusoils.jrc.ec.europa.eu/projects/insea/index.htm), while research into better understanding soil carbon dynamics under land management alternatives is ongoing (e.g., Climate Science Department at

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

387

Lawrence Berkeley National Laboratory (http://esd.lbl.gov/CSD). Kutsch et al. (2009) provide an overview of the state of soil carbon science. · Integrating more detailed chemical process modeling (such as with ASPEN Plus) into LCA studies of biofuels (and other fuels). · With the many uncertainties that propagate throughout life cycle models, use of stochastic analysis methods is an important component that is expected to become more commonly and better integrated into LCA/WTW models in the future (see Spatari, 2007). We cannot know what will occur in the future; however, moving forward to ensure that biobased product initiatives and public policies make net positive contributions to society will be key. Life cycle-based methods are continually improving and expanding in their scope. They are but one input into decisions and their results should be interpreted with a broad knowledge of the scope and limitations of the study.

14.8

References

ANL (2009), The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET) Model. Version 1.8b. Argonne National Laboratory, Argonne, IL. Available from: (accessed March 4, 2009). Baker JM, Oschner TE, Venterea RT and Griffis TJ (2007), `Tillage and soil carbon sequestration ± What do we really know?' Agriculture, Ecosystems, and Environment, 118, 1±5, doi:10.1016/j.agee.2006.05.014. Beer T, Grant T, Williams D and Watson H (2002), `Full fuel-cycle greenhouse gas emissions analysis of alternative fuels for Australian heavy vehicles', Atmospheric Environment, 36, 753±763, doi:10.1016/S1352-2310(01)00514-3. Birur DK, Hertel TW and Tyner WE (2008), Impact of biofuel production on world agricultural markets: a computable general equilibrium analysis. Department of Agricultural Economics. Purdue University. GTAP working paper No. 53. California Air Resources Board (2009), Detailed California-modified GREET pathway for cellulosic ethanol from farmed trees by fermentation. Released February 27, 2009. Version 2.1. Available at http://www.arb.ca.gov/fuels/lcfs/ 022709lcfs_trees.pdf. CONCAWE (Oil Companies' European Association for Environment, Health and Safety in Refining and Distribution). Joint Research Centre of the EU Commission, and European Council for Automotive R&D (2007), Well-to-Wheels Analysis of Future Automotive Fuels and Powertrains in the European Context. Version 2c, January. Available at: http://ies.jrc.ec.europa.eu/WTW. Cramer J, Wissema E, de Bruijne M, Lammers E, Dijk D, Jager H, van Bennekom S, Breunesse W, Horster R, van Leenders C, Wonink S, Wolters W, Kip H, Stam H, Faaij A and Kwant E (2007), Testing framework for sustainable biomass. Final report from the project group Sustainable production of biomass. Dutch stakeholder report to the Dutch Government. Available at: http://www.lowcvp.org.uk/assets/ reports/070427-Cramer-FinalReport_EN.pdf. Deluchi MA (1991), Emissions of greenhouse gases from the use of transportation fuels

ß Woodhead Publishing Limited, 2010

388

Bioalcohol production

and electricity, ANL/ESD/TM-22, Volume 1, Center for Transportation Research, Argonne National Laboratory, Argonne, IL. Available at: http:// www.its.ucdavis.edu/people/faculty/delucchi. Deluchi MA, Johnston RA and Sperling D (1987), Transportation fuels and the greenhouse effect, UERG-180, University Energy Research Group, University of California, Berkeley, CA. Elsayed MA, Matthews R, and Mortimer ND (2003), Carbon and Energy Balances for a Range of Biofuel Options. Resources Research Unit, Sheffield Hallam University, UK, March. Fargione J, Hill J, Tilman D, Polasky S and Hawthorne P (2008), `Land clearing and the biofuel carbon debt', Science, 319, 1235±1238, doi: 10.1126/science.1152747. Fleming JR, Habibi S and MacLean HL (2006), `Investigating the sustainability of lignocellulose-derived fuels for light-duty vehicles'. Transportation Research Part D: Transport and Environment, 11(2), 146±159, doi:10.1016/j.trd.2006.01.001. Gallagher E (2008), The Gallagher Review of the indirect effects of biofuels production. Renewable Fuels Agency (RFA), UK, Available at: http://www.dft.gov.uk/rfa/_db/ _documents/Report_of_the_Gallagher_review.pdf. Gaunt JL and Lehmann J (2008), `Energy balance and emissions associated with biochar sequestration and pyrolysis bioenergy production'. Environ. Sci. Technol., 42, 4152±4158, doi: 10.1021/es071361i. Greene H (2004), Growing energy: How biofuels can help end America's oil dependence. Natural Resources Defense Council, New York, December. Hayashi T, Masuda K and Yamamoto M (2008), Measuring economic and CO2 impacts of bio-ethanol production in Hokkaido, Japan. Proceedings of the International Conference on Sustainable Agriculture for Food, Energy, and Industry 2008, July 2±6, Hokkaido, Japan. Hill J, Nelson E, Tilman D, Polasky S and Tiffany D (2006), `Environmental, economic, and energetic costs and benefits of biodiesel and ethanol biofuels', PNAS, 103(30), 11206±11210, doi: 10.1073/pnas.0604600103. Hill J, Polasky S, Nelson E, Tilman D, Huo H, Ludwig L, Neumann J, Zheng H and Bonta D (2009), `Climate change and health costs of air emissions from biofuels and gasoline'. PNAS, 106(6), 2077±2082, doi: 10.1073/pnas.0812835106. Huo H, Wang M, Bloyd C and Putsche V (2009), `Life cycle assessment of energy use and greenhouse gas emissions of soybean-derived biodiesel and renewable fuels', Environ. Sci. Technol., 43(3), 750±756, doi: 10.1021/es8011436. ISO (2006) ISO 14041: Environmental Management: Life cycle assessment ± Goal and scope defintion and inventory analysis. International Organization for Standardization, Geneva. Kim H, Kim S and Dale BE (2009), `Biofuels land use change and greenhouse gas emissions: Some unexplored variables', Environ. Sci. Technol., 43(3), 961±967, doi: 10.1021/es802681k. Kim S and Dale BE (2003), `Cumulative energy and global warming impact from the production of biomass for biobased products', J. Ind. Ecol., 7(3±4), 147±162. Kim S and Dale BE (2005), `Life cycle assessment of various cropping systems utilized for producing biofuels: bioethanol and biodiesel', Biomass and Bioenergy, 29(6), 426±439, doi: 10.1016/j.biombioe.2005.06.004. Kutsch W, Bahn M and Heinemeyer A (2009), Soil Carbon Dynamics. Cambridge, UK, Cambridge University Press. Lal R (2003), `Global potential of soil carbon sequestration to mitigate the greenhouse effect'. Critical Reviews in Plant Sciences, 22(2), 151±184, doi: 10.1080/

ß Woodhead Publishing Limited, 2010

Environmental life cycle assessment

389

713610854 Larson ED (2006), `A review of life-cycle analysis studies on liquid biofuel systems for the transport sector', Energy for Sustainable Development, X(2), 109±126. Levelton (1999), Assessment of Net Emissions of Greenhouse Gases from EthanolBlended Gasolines in Canada: Lignocellulosic Feedstocks, Levelton Engineering Ltd. and S&T Consultants. Prepared for Agriculture and Agrifood Canada, Ottawa, ON, File 499-0893. Lutsey N and Sperling D (2008), Transportation and greenhouse gas mitigation. University of California, Davis, Institute for Transportation Studies, Davis, CA. 2008 UCD ITS RP-08-06. MacLean HL and Lave LB (2003), `Evaluating automobile fuel/propulsion technologies', Progress in Energy and Combustion Science, 29(1), 1±69, doi:10.1016/S03601285(02)00032-1. MacLean HL and Spatari S (2009), `Contribution of enzymes and process chemicals to the life cycle of ethanol', Environmental Research Letters, 4, 1±10, doi: 10.1088/ 1748-9326/4/1/014001. NRCan (2008), GHGenius ± a model for lifecycle assessment of transportation fuels. Version 3.12b. Natural Resources Canada, Ottawa, ON, Canada. Available at: http://www.ghgenius.ca/forum/index.php?action=vthread&forum=7&topic=208> (accessed June 12, 2008). Nguyen TLT, Gheewala SH and Garivait S (2007a), `Energy balance and GHGabatement cost of cassava utilization for fuel ethanol in Thailand', Energy Policy, 35(9), 4585±4596, doi: 10.1016/j.enpol.2007.03.012. Nguyen TLT, Gheewala SH and Garivait S (2007b). `Full chain energy analysis of fuel ethanol from cassava in Thailand', Environ. Sci. Technol., 41(11), 4135±4142, doi: 10.1021/es062064. Quirin M, Gartner SO, Pehnt M and Reinhardt GA (2004), CO2 mitigation through biofuels in the transport sector: status and perspectives. Main report, Institute for Energy and Environmental Research (IFEU), Heidelberg, Germany. Scharf P, Lory J and Grundler J (2006), Integrated pest management. Best management practices for nitrogen fertilizer in Missouri. IPM 1027. Plant Protection Programs, College of Agriculture, Food and Natural Resources, University of MissouriColumbia, USA. Schoemaker HE, Mink D and Wubbolts MG (2003). `Dispelling the myths: biocatalysis in industrial synthesis', Science, 299, 1694±1697, doi: 10.1126/science.1079237. Searchinger T, Heimlich R, Houghton RA, Dong F, Elobeid A, Fabiosa J, Tokgoz S, Hayes D and Yu T-H (2008), `Use of U.S. croplands for biofuels increases greenhouse gases through emissions from land-use change', Science, 319, 1238± 1240, doi: 10.1126/science.1151861. Sheehan J, Aden A, Paustian K, Kendrick K, Brenner J, Walsh M and Nelson R (2003), `Energy and environmental aspects of using corn stover for fuel ethanol', Journal of Industrial Ecology, 7(3±4), 117±146, doi: 10.1162/108819803323059433. Smil V (2001), Feeding the World: A Challenge for the Twenty-First Century, Cambridge, MA, MIT Press. Smith P, Powlson DS, Glendining MJ and Smith JU (1998), `Preliminary estimates of the potential for carbon mitigation in European soils through no-till farming', Global Change Biology, 4, 679±685. Spatari S, Zhang Y and MacLean HL (2005), `Life cycle assessment of switchgrass and corn stover ± derived ethanol fueled automobiles', Environ. Sci. Technol., 39, 9750±9758, doi:10.1021/es048293+.

ß Woodhead Publishing Limited, 2010

390

Bioalcohol production

Spatari S (2007), Biomass to ethanol pathways: Evaluation of lignocellulosic ethanol production technologies. PhD, thesis, Department of Civil Engineering, University of Toronto, Toronto, Canada. Spatari S, O'Hare M, Fingerman K and Kammen D (2008), Sustainability and the low carbon fuel standard, Report to the California Air Resources Board, October. Sperling D and Yeh S (2009), `Low carbon fuel standard', Issues Sci. Technol., Winter, 57±66. State of California Office of the Governor (2007), Executive Order S-1-07, The Low Carbon Fuel Standard, January 18. Tripp B (2007), Evaluating the life-cycle of biodiesel in North America, MASc. thesis, Department of Civil Engineering, University of Toronto, Toronto, Canada. UK Department of Transport (2007), About the Renewable Transport Fuel Obligation Programme (RTFO). http://www.dft.gov.uk/roads/RTFO. US DOE (2009), Biomass multi-year program plan, Office of Biomass Program, Energy Efficiency and Renewable Energy, US Department of Energy. Available at: http:// www1.eere.energy.gov/biomass. VandenBygaart AJ, Yang XM, Kay BD and Aspinall JD (2002), `Variability in carbon sequestration potential in no-till soil landscapes of southern Ontario'. Soil and Tillage Research, 65, 231±241, doi:10.1016/S0167-1987(02)00003-X. Wang MQ and Haq Z (2008), `Ethanol's effect on greenhouse gas emissions', E-Letter, Science, Aug 12, 2008, http://www.sciencemag.org/cgi/eletters/319/5867/ 1238#10977. Woods J and Bauen A (2003) Technology Status Review and Carbon Abatement Potential of Renewable TransportFuels (RTF) in the UK. DTI; B/U2/00785/REP URN 03/982, http://www.berr.gov.uk/files/file15003.pdf. Wu M, Wu Y and Wang M (2005), Mobility chains analysis of technologies for passenger cars and light-duty vehicles fuelled with biofuels: Application of the GREET model to the Role of Biomass in America's Energy Future (RBAEF) Project. Center for Transportation Research, Argonne National Laboratory, Argonne, IL, May. Wu M, Wang M and Huo H (2006), Fuel-cycle assessment of selected bioethanol production pathways in the United States. Center for Transportation Research, Argonne National Laboratory, Argonne, IL, Nov 7, ANL/ESD/06-7. Zhang Y, Habibi S and MacLean HL (2007), `Environmental and economic evaluation of bioenergy in Ontario, Canada', J. of the Air & Waste Management Association, 57, 919±933, doi:10.3155/1047-3289.57.8.919. Zhang Y, McKechnie J, Cormier D, Lyng R, Mabee W and MacLean HL (2010), `Comparison of life cycle greenhouse gas and air pollutant emissions associated with electricity production from coal, natural gas and wood pellets in Ontario, Canada'. Environ. Sci. Technol., 44(1), 538±544.

ß Woodhead Publishing Limited, 2010

15

Chemical production from lignocellulosic biomass: thermochemical, sugar and carboxylate platforms A . D . S M I T H , M . L A N D O L L , M . F A L L S and M . T . H O L T Z A P P L E , Texas A&M University, USA

Abstract: Lignocellulose is abundant, inexpensive, and renewable. Further, it can replace fossil fuels as our primary source for chemicals and energy. To transition from fossil fuels to lignocellulose-based chemicals and fuels, an economically viable commercial biorefinery must be developed. This chapter reviews three major chemical platforms for a biorefinery. The thermochemical platform thermally degrades biomass into gaseous, liquid, and solid components. The sugar platform utilizes enzymes and/or chemicals to create sugars. The carboxylate platform ferments biomass into carboxylic acids/salts. The intermediate products of these three platforms serve as a `biocrude' for the production of chemicals and fuels from lignocellulose. Key words: thermochemical platform, sugar platform, carboxylate platform, gasification, pyrolysis, biorefinery, lignocellulose, chemical production, biomass, chemicals.

15.1

Introduction

Continued dependence on petroleum presents many challenges. Demand grows faster than supply, which increases fuel costs. Petroleum use increases the atmospheric carbon dioxide concentration, thus contributing to global warming. The United States imports 72% of its oil supply, which affects national security and contributes to the trade deficit (EIA, 2007). Producing chemicals and fuels from renewable sources addresses all of these issues. Lignocellulose is the primary component of plant material, and thus is the most abundant renewable material on earth. It is an attractive feedstock for conversion into chemicals and fuels because it is abundant and inexpensive. `The key to exploiting the chemical value of lignocellulosics is to depolymerize the lignocellulosic matrix in order to obtain smaller molecules that can be utilized, or further converted to platform chemicals and biofuels' (Hayes, 2009). This chapter reviews three platforms for producing chemicals and fuels from lignocelluloses: thermochemical, sugar, and carboxylate. The purpose of this chapter is to present an overview of each platform, and to review process basics, primary and secondary products, advantages and disadvantages, research and development, and future trends.

ß Woodhead Publishing Limited, 2010

392

15.2

Bioalcohol production

Lignocellulose feedstocks

15.2.1 What is lignocellulose? Lignocellulosic biomass is composed of cellulose (40±50%), lignin (15±20%), and hemicellulose (25±35%). Cellulose is a linear polymer of glucose (sixcarbon sugar) linked with -1,4 bonds, as compared to the -1,4 bonds of starch. This difference makes cellulose ridged and more difficult to digest. Hemicellulose is a randomly acetylated branch polymer of five-carbon sugars (mostly xylose) that are covalently bonded to lignin. Lignin is a polymer of phenyl propane and serves as the `binder' that holds the structural cellulose bundles together. Biochemical platforms must disrupt and/or remove lignin because it reduces the digestion of the cellulose and hemicellulose (Sierra et al., 2008).

15.2.2 Is there enough? In 2005, the world used about 400 quadrillion Btu (quads) of fossil fuel energy. Annually, about 2700 quads of biomass are produced by photosynthesis (Chen et al., 2003; EIA, 2007), so there are sufficient quantities of biomass if it can be collected. Realistically, biomass should be reserved for high-value applications, such as liquid transportation fuels and chemicals. Other renewables (e.g., wind, solar) should be used for stationary power. On an energy basis, lignocellulosic biomass is less expensive than petroleum ($5±20 USD per barrel of oil equivalent verses $50±140 USD) (Huber et al., 2006); Thus, it is realistic that biomass can supply a significant fraction of global energy needs.

15.2.3 Waste and crops Lignocellulose feedstocks can be categorized as waste materials or energy crops. Wastes materials (e.g., municipal solid waste, agricultural residues, and sewage sludge) are ideal feedstocks because they are abundant and low cost. Using wastes help reduce landfill volumes, add value to negative-value products, and improve biorefinery economics. Disadvantages of using municipal solid waste include inconsistent composition, cost of separation, and transportation logistics. Three major logistical issues to consider when choosing a feedstock are harvesting, transportation, and storage. Single-pass harvesting techniques should be employed to reduce handling and cost. Because biomass has a low energy density (17±19 000 btu/kg) and medium to high moisture content (0.5±0.95 g moisture/g total) biorefineries must be near (<100 km) the biomass source (farmland, landfill, sewage treatment plant, etc.). Biomass storage increases the biomass handling and typically results in a 5±10% loss of dry matter (Atchison and Hettenhaus, 2003; Sierra et al., 2008). Schemes that minimize storage, adapt to seasonal fluctuations in biomass supply, and minimize handling/ transportation should be used.

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

393

The logistics used to transport biomass from the field significantly impact the economics of the biorefinery. Thus, an economical means of transportation must be realized before agricultural waste can be considered as a viable feedstock (Rooney, 2007). Ultimately, the cost per delivered dry tonne dictates if a feedstock is economically attractive. If wastes are unavailable or uneconomical, energy crops can be grown. The ideal crop feedstock has the following characteristics: · · · · · ·

has a high annual yield (>30 dry tonne per hectare) uses little or no irrigation does not require fertilizers grows on marginal land is drought tolerant is a non-food crop (Rooney, 2007).

Potential energy crops include sorghum, energy cane, switchgrass, and water hyacinths.

15.3

Thermochemical platform

Thermochemical processing is an established technology that uses heat and/or catalysts to convert lignocellulose into gases, liquids, and solids. The amount and composition of each depend on specific process conditions (temperature, heating rate, equivalence ratio, etc.) and the composition of the feedstock (moisture, ash, element ratios) (Reed, 1981). Lignocellulose is an attractive feedstock for thermochemical conversion because of its abundance, high volatile content (~70±90%), high oxygen content, and low sulfur content. Disadvantages include its high moisture content and its dispersed nature, which increase the drying costs and transportation costs, respectively.

15.3.1 Feedstock considerations When choosing a feedstock for thermochemical processing, it is important to consider the elemental composition, moisture content, heating value, ash content, ratio of volatile carbon to fixed carbon, ash-softening temperature, and thermal conductivity. Elemental (ultimate) analysis helps to predict the product spectrum, identify potential pollutants or hazardous byproducts, and compare feedstocks. Moisture in feedstock increases drying costs and decreases the higher (gross) heating value. Moisture content must be less than 50% to be economically viable (Phillips et al., 2007). Because ash is inert and contributes to slagging/fouling, lower ash contents are preferred. An NREL report shows that increasing the ash content from 1% to 18% increases ethanol selling price 60% and decreases yield 32% (Phillips et al., 2007; Reed, 1981). Low ash-

ß Woodhead Publishing Limited, 2010

394

Bioalcohol production

softening temperatures can foul equipment. Thermal conductivity of the biomass influences the heating rate and affects its behavior during gasification (Reed, 1981). The logistics and economics of gaseous feedstocks must be considered. For thermochemical processes operating with an equivalence ratio > 0, an oxygen source must be supplied. (Note: `equivalence ratio' is defined as the amount of oxygen supplied compared to the stoichiometric amount needed for complete combustion.) Air is inexpensive and ideal if liquids and solids are the target products. If gaseous products are desired, air will dilute the gas with nitrogen, requiring further purification. Use of pure oxygen eliminates the need for gas purification, but limits the facility location and increases cost. Gaseous products are dominated by CO and H2, but there is often more CO than is desired for downstream processing. In this case, the water-gas shift reaction can be employed: CO + H2O $ H2 + CO

15.1

It requires steam addition to produce H2; the resulting CO2 is removed. When more reduced products are desired, the hydrogen content must be enriched (Reed, 1981). Biomass geometry is critical to the equipment employed. Fluidized-bed reactors require small particles, whereas fixed-bed gasifiers need large chunks or pellets (Reed, 1981). Mechanical processing of biomass typically requires less than three per cent of the energy content of the biomass (Reed, 1981). However, particle reduction for an entrained flow gasifier (which requires the smallest particles) may consume a third of the total energy demand of the biorefinery (Phillips et al., 2007).

15.3.2 Process basics The equivalence ratio (ER) is the most important process variable because it has the greatest influence on primary products produced (Reed, 1981). Pyrolysis (ER ~ 0), gasification (0 < ER < 1), and combustion (ER  1) define regions of the ER continuum. Low ERs (<0.15) yield high fractions of solid and liquid components. Increasing the ER increases the decomposition of solid and liquid components into gaseous components. Because combustion (ER  1) is only useful for heat and power generation, it will not be discussed. In addition to ER, heating rate, residence time, pressure, and temperature can be adjusted to optimize the yield of the target product. Figure 15.1 illustrates the gasification and pyrolysis pathways with the general influence of the process variables indicated. Figure 15.2 shows how biomass chemically changes during different thermochemical processes and the general products that are favored. There are many variations of pyrolysis and gasification; however, the general principles discussed can be applied throughout.

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

395

15.1 Gasification and pyrolysis pathways (Huber et al., 2006).

15.2 Chemical changes during thermochemical processing of biomass (Reed, 1981).

ß Woodhead Publishing Limited, 2010

396

Bioalcohol production

Pyrolysis Pyrolysis is the thermal degradation of a material when an oxidizing agent is absent or present in very limited quantities. It results in gases, liquids, and char (Bridgwater and Grassi, 1991; LePori and Soltes, 1985). The proportions of the primary products depend on process variable (e.g., heating rate, residence time, particle size, temperature, pressure) (Bridgwater and Grassi, 1991). Fast (or flash) pyrolysis uses rapid heating rates (100±10 000 ëC/min) and short residence times at high temperatures (450±900 ëC) and near-atmospheric pressures (~100 kPa). Because of the fast heating rate, fine particles must be used to ensure rapid heat transfer. These conditions favor gaseous and liquid products, which are quenched to prevent further degradation (Bridgwater and Grassi, 1991; Goyal et al., 2008). Slow pyrolysis uses heating rates from 5 to 7 ëC/min at temperatures ranging from 400 ëC to 600 ëC at sub-atmospheric pressures (1±100 kPa). These conditions produce higher yields of solids (char) with less liquids and gases. Because the heat transfer is slower, larger particles are used (Bridgwater and Grassi, 1991; Chen et al., 2003; Hayes, 2009; Goyal et al., 2008). The mechanisms of biomass pyrolysis are complex and not completely understood (LePori and Soltes, 1985). `The most accepted theory is that primary vapors are produced first, the characteristics of which are most influenced by heating rate. These primary vapors then further degrade to secondary tars and gases if held at a high temperature for long enough for secondary reactions to occur' (Bridgwater and Grassi, 1991). Gasification Gasification is the partial oxidation of biomass to produce gaseous products with a usable heating value (LePori and Soltes, 1985; Higman and Burgt, 2003). Typical gasification conditions have an equivalence ratio of 0.2±0.4, a temperature of 1000±1500 ëC, and a pressure of 100±2000 kPa (Reed, 1981; Bridgwater and Grassi, 1991; LePori and Soltes, 1985). Biomass gasification is similar to coal gasification, except lower operating temperatures are used because biomass is more reactive and has high metal contents (Na, K, Ca, etc.), which increase slagging/fouling problems (Huber et al., 2006). Tar removal is the greatest technological challenge hindering the development of commercial biomass gasifiers (Bridgwater, 2001; Huber et al., 2006). Fouling and slagging due to ash must also be addressed. Air gasification produces a low-heating-value gas (150±200 Btu/scf; diluted with nitrogen), which is uneconomical to distribute but useful on-site for heat or power generation. Oxygen gasification produces a higher heating value gas (300 Btu/scf), which is useful for conversion to methanol, gasoline, ammonia, methane, or hydrogen (Reed, 1981). Typically, 10±30% of the energy content of the solid fuel is lost during gasification (Reed, 1981).

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

397

Gasifiers are classified according to the method by which the fuel moves through the gasifier: moving-bed, fluidized-bed, rotary-kiln, and multiple-hearth types (LePori and Soltes, 1985). Further, moving-bed gasifiers may be classified by the direction of air moving through the gasifier: downdraft, updraft, or crossdraft (LePori and Soltes, 1985).

15.3.3 Primary products Each thermochemical process is designed to optimize the production of a target product. The same basic primary products are produced in both pyrolysis and gasification processes; the relative amounts depend on the operating parameters and the feedstock. In this section, the basic primary products have been grouped to simplify our discussion. Figure 15.3 (Reed, 1981) shows the equilibrium composition for adiabatic air/biomass reaction. Although, equilibrium is rarely achieved in a thermochemical reaction, this figure helps illustrate how ER influences the primary product composition.

15.3 Equilibrium composition for adiabatic air/biomass reaction (Reed, 1981).

ß Woodhead Publishing Limited, 2010

398

Bioalcohol production

Char Char (charcoal) is the most common pyrolysis product, which is primarily used as a fuel for heating or cooking because of its high energy content (~30 GJ/t) (LePori and Soltes, 1985). Alternatively, char can be used as a soil amendment which protects important soil fungi and increases land productivity. Additional benefits include sequestering carbon in a recalcitrant form, reducing N2O emissions from the soil, and trapping nitrates and organics from drainage waters (Hayes, 2009). Industrially, activated carbon (activated charcoal) is commonly used in filters to remove chlorine and volatile organic compounds. Pyrolysis tars and oils Pyrolysis liquids (tars and oils) are complex chemicals resulting from the uncontrolled degradation of the biomass (LePori and Soltes, 1985). Bio-oil can contain over 100 compounds including carboxylic acids, esters, alcohols, ketones, aldehydes, phenols, alkenes, alkanes, amines, furans, guaiacols, syringols, and miscellaneous oxygenates (Goyal et al., 2008). Therefore, pyrolysis liquid is a biocrude that requires refining, separation, and upgrading into more useful products. Some direct applications of pyrolysis tars and oils include liquid smoke, binders, and combustion fuel for power generation (Goyal et al., 2008). Pyrolysis oils must be upgraded for use as motor fuel because of their high oxygen (30±40% w/w) and water contents. Additionally, these oils are less stable and less miscible than conventional fuels (Goyal et al., 2008; Bridgwater and Grassi, 1991). Some problematic issues ± especially in storage ± include phase separation, polymerization, and corrosion (LePori and Soltes, 1985; Goyal et al., 2008). Synthesis gas (syngas) and producer gas Syngas is made in oxygen-fed gasifiers whereas producer gas is made in air-fed gasifiers. Producer gas is diluted with nitrogen and has little use other than a low-heating-value fuel (Huber et al., 2006). Like pyrolysis liquid, synthesis gas is a mixture of decomposition products (e.g., CO, H2, CO2, CH4, and light hydrocarbons) that can be separated and sold, or used to synthesize secondary chemicals. CO and H2 are the building blocks for synthesizing a variety of hydrocarbons and organic compounds. As mentioned earlier, the water-gas shift reaction can be used to enrich the hydrogen content (Huber et al., 2006; Reed, 1981). Water Water is a byproduct of most thermochemical processes and the dissolved hydrocarbon content must addressed by an appropriate disposal method.

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

399

15.3.4 Chemistry and secondary chemicals Before discussing chemistry and secondary chemicals, it must be understood that the quality and composition of the primary products are specific to the feedstock and the process conditions. Further, bio-oil is very complex, potentially containing over 100 compounds; therefore, the following discussion is very general. Hydrotreatment (adding hydrogen or a hydrogen-donor solvent) is a wellestablished technology that upgrades bio-oil to produce hydrocarbons by removing oxygen, sulfur, and nitrogen. Hydrotreating catalysts typically contain nickel, molybdenum, and cobalt. Zeolite catalysts (aluminum and silicon) can be used to directly convert bio-oil into high-quality hydrocarbons and aromatics (Bridgwater and Grassi, 1991). Separation of each individual component is not practical because of the complexity of the bio-oil and the low concentration of each chemical species; however, fractions with similar functionalities and chemical properties can be separated. Figure 15.4 shows the array of chemicals that can be made from syngas and some of the catalysts involved. Conversion of syngas to hydrocarbons using Fisher-Tropsch catalysts was practiced by the Germans in World War II, and is currently practiced in South Africa. Alcohols, aldehydes, and ethers can be used as solvents and reagents. Methanol is a platform chemical that can be converted into olefins, gasoline, methyl tert-butyl ether (MTBE), acetic acid, hydrogen, and formaldehyde (Hayes, 2009; Huber et al., 2006). Hydrogen has wide

15.4 Potential chemicals from syngas and some of the required catalysts (Hayes, 2009).

ß Woodhead Publishing Limited, 2010

400

Bioalcohol production

application in upgrading liquid products and chemical synthesis. Ammonia and urea are major components of fertilizers. Olefins such as ethylene can be used to make polymers and other chemicals. Traditional transportation fuels (gasoline and diesel) may be synthesized to displace petroleum-derived fuel.

15.3.5 Current production of chemicals Gasification technology has been used on a large scale for over 150 years (Reed, 1981). Many gasification technologies have been demonstrated on a large scale. Table 15.1 lists companies that are working to develop full-scale biorefineries. Most industrial gasifiers use coal because of its low cost. Many companies such as American Electric Power, Duke Energy, Southern Company, and General Electric use the Integrated Gasification Combined Cycle (IGCC), which is commonly described as `clean coal technology' to produce power and chemicals. If low-cost biomass is available, these plants can use it to replace or supplement coal (Gasification Technology Council, 2008).

15.3.6 Research and development Table 15.1 lists companies that will likely have small commercial-scale biomass facilities within five years. The following topics are key areas of research: · Production of phenolic resins and tar removal (Effendi et al., 2008; Huber et al., 2006). · Investigation of mechanisms that govern thermochemical decomposition reactions to optimize the production of specific chemicals (Hayes, 2009). · Catalysts that increase the selectivity of target chemical and reduce poisoning and deactivation of catalysts (Hayes, 2009). · Biocatalysts that produce chemicals from synthesis gas. · Characterization of the performance of different lignocellulosic feedstocks. · Investigation of different heating sources such as solar and microwaves (Huang et al., 2008).

15.4

Sugar platform

The sugar platform converts lignocelluloses to sugars, which are converted into fuels and chemicals. Common routes rely heavily on enzymatic saccharification and fermentation. Ethanol is the most researched product, but is not necessarily the most economically attractive product (Gray et al., 2006).

15.4.1 Feedstock considerations Traditional feedstocks for the sugar platform are sugarcane and corn. Sugarcane is heavily favored in Brazil because of its tropical climate. Sugars can be

ß Woodhead Publishing Limited, 2010

Table 15.1 Companies that are likely to have commercial-scale (>10 ML/y) biorefineries in the next five years (Hayes, 2009)

ß Woodhead Publishing Limited, 2010

Platform

Company

Process

Main outputs

Current facility

Full scale

Output

Thermochemical

Alico

Biocatalytic gasification

Ethanol

Pilot Plant (2003)

2009

53 ML

Choren

Gasification and FT synthesis

FT-diesel

2011

200,000 tonnes

BlueFire Ethanol Concentrated acid hydrolysis Range Fuels Gasification and mixed alcohols Coskata Plasma gasification and biocatalytic synthesis

Ethanol Ethanol, methanol Ethanol

Pilot Plant. Demo plant (15,000 tonnes) expected 2008 2 pilot plants, largest = 81 kL Pilot Plant

2010 2008

72 ML 38 ML

Lab-scale. 150 kL (late 2008)

2011

370 ML

Iogen

Ethanol

Demo plant (2004): 3 ML

2011

90 ML

Ethanol

Pilot scale

2009

45 ML

Ethanol Ethanol

Lab-scale 3 pilot plants. Largest 1.3 ML (2007) Demo plant ± 5 ML Pilot Plant

2011 2011

200 ML 110 ML

2010 2011

57 ML 115 ML

Pilot Plant (2001) Demo Plant (2008)

2011

5 ML

Demo plant (poplar wood) 5.7 ML (2009)

2010

6 ML

Sugar

Royal Nedalco Verenium

Steam explosion and enzymatic hydrolysis Solvent treatment and enzymatic hydrolysis Enzymatic hydrolysis Enzymatic hydrolysis

Abengoa Poet

Enzymatic hydrolysis Enzymatic hydrolysis

Ethanol Ethanol

Terrabon

MixAlco process

ZeaChem

Zeachem technology

Carboxylate salts/Synthetic gasoline Chemicals and ethanol

Colusa

Carboxylate

402

Bioalcohol production

converted directly into ethanol and other products via fermentation. Because of its temperate climate, the United States favors corn. To utilize corn, amylase enzymes are employed to hydrolyze starch into sugar. In 2008, the annual US ethanol capacity of ~46 billion liters (~12 billion gallons) was much less than the gasoline usage of 530 billion liters (140 billion gallons) (Gray et al., 2006; Rooney et al., 2007; RFA, 2008). Additionally, using food crops to produce chemicals and fuels creates an unnecessary ethical dilemma. The sugar platform can accept any lignocellulosic material, provided it does not inhibit or degrade the hydrolytic enzymes. From a process economics viewpoint, the ideal feedstock has low lignin content for better digestibility, low ash content for better yield, and a low cost per delivered tonne. Unreactive biomass increases the fermentor volume, which increases capital costs.

15.4.2 Process basics In the early twentieth century, mineral acids (e.g., sulfuric acid) were used to hydrolyze both starch and cellulose. Because the acids are not very selective and produce degradation products (e.g., hydroxymethyl furfural), newer approaches emphasize enzymes. Amylase enzymes saccharify corn starch into glucose, which is fermented to ethanol by Saccharomyces yeasts (Gray et al., 2006). The enzymatic hydrolysis of cellulose is performed by cellulase enzymes; unfortunately, lignocellulosic biomass is more complex and recalcitrant. Cellulase activity is 100 times less than amylase activity, prompting the need for more efficient cellulase systems and pretreatment methods (Bok et al., 1994). The goal of lignocellulose pretreatment is to increase enzymatic digestibility of the biomass. This is achieved in several ways: remove lignin, remove acetyl groups from hemicellulose, increase material porosity, and reduce cellulose crystallinity (Sun, 2002). There are three main pretreatment methods: · Physical pretreatment methods use grinding, milling, or chipping to reduce size and lower crystallinity. · Physico-chemical pretreatments use steam (most popular), liquid ammonia, or carbon dioxide to degrade the hemicellulose and lignin via explosive decompression. · Chemical pretreatments use solvents, oxidants, acids, or bases. Solvents (e.g., ethanol, butanol) increase the hydrophobic nature of the liquid to increase the solubility of lignin fragments. Oxidants (e.g., ozone, hydrogen peroxide) remove lignin. Acid pretreatment (e.g., sulfuric acid) hydrolyze hemicellulose to xylose and other sugars, which opens pore structure. Alkaline pretreatments (e.g., NaOH, NH3, Ca(OH)2) remove lignin and acetyl groups from hemicellulose; both inhibit enzymatic accessibility (Mosier et al., 2005). Chemical treatments can be used in combination (e.g., NH3 + solvent, NaOH + H2O2, Ca(OH)2 + O2) to improve pretreatment effectiveness.

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

403

Once biomass has been pretreated, cellulase enzymes degrade the cellulose into glucose. Enzymes account for a significant portion of the total ethanol production cost (Walker and Wilson, 1991). There are three types of cellulase enzymes, each with a different role. Endogluconase (1,4-D-glucanohydrolase) is responsible for hydrolyzing low-crystallinity regions to create free-chain ends. Exogluconase (1,4- -D-glucan cellobiohydrolase) works from the free chainends to hydrolyze cellulose into cellobiose. The third enzyme ( -glucosidase) cleaves the -1,4 linkage of cellobiose to give two molecules of glucose (Coughlan and Ljungdahl, 1988). Many cellulases have cross reactivity and also hydrolyze the -1,4 links in hemicellulose. Other enzymes ± such as glucuronidase, acetylesterase, xlanase, and glucomannase ± specialize in hydrolyzing hemicellulose polymers (Duff and Murray, 1996). The fermentation of lignocellulose-derived sugars is a new challenge. Sugars derived from starch and sucrose are hexoses, which easily ferment by the yeast S. cerevisiae (Aiello-Mazzarri et al., 2006). In contrast, enzymatic hydrolysis of lignocellulosic biomass produces a mixture of hexoses (six-carbon sugars) and pentoses (five-carbon sugars), primarily glucose and xylose. There are three approaches being explored: (1) use two microorganisms, one that ferments hexoses (e.g., S. cerevisiae) and one that ferments pentoses (e.g., Pichia stipitis), (2) engineer pentose metabolic pathways into yeast (e.g., S. cerevisiae) (Jeffries and Jin, 2004), and (3) modify microorganisms that metabolically convert both glucose and xylose, and genetically engineer them to be more efficient (e.g., Escherichia coli, Klebsiella oxytoca, and Zymomonas mobilis) (Dien et al., 2003).

15.4.3 Primary products Many products can be obtained via fermentation of sugars. The microorganism used dictates the products (e.g., ethanol, succinic acid, lactic acid, and acetic acid (Kamm and Kamm, 2007). Ethanol receives the most interest because of its use as fuel; however, it is not an economically feasible primary product on its own (Gray et al., 2006). Complementing ethanol production with another product makes the sugar platform a feasible choice for a biorefinery system. Figure 15.5 illustrates secondary products that can be made from primary products derived from hexoses.

15.4.4 Chemistry and secondary chemicals Succinic acid For many anaerobic microorganisms, succinic acid is a common metabolite. In contrast to ethanol fermentation, succinic acid fermentation consumes CO2 rather than producing it. A biorefinery could combine succinate and ethanol

ß Woodhead Publishing Limited, 2010

404

Bioalcohol production

15.5 Secondary products from hexoses.

fermentation processes to reduce carbon waste. Succinic acid can be produced by many microorganisms such as Actinobacillus succinogenes, a facultative anaerobe producing the highest yields of succinate and concentrations up to 110 g/L (Guettler and Jain, 1996). It can ferment a wide range of substrates including glucose, D-xylose, cellobiose, galactose, fructose, mannose, lactose, and L-arabinose (Guettler et al., 1999). According to MBI International, succinic acid can be produced via fermentation for $2.20/kg (production of 5000 tonnes/year) or as low as $0.55/kg (production greater than 75 000 tonnes/ year). In the United States, succinic acid sells for $5.90 to $8.80/kg (Zeikus et al., 1999). Succinic acid can be chemically converted to high-value secondary chemicals, including adipic acid and the solvents N-methylpryrrolidone, 1,4butanediol, -butyrolactone, and tetrahydrofuran. Adipic acid is used to manufacture lubricants, foams, and food products and is a precursor to nylon 6,6 (Zeikus et al., 1999). Lactic acid Lactic acid can be obtained through hexose fermentation. The most promising lactic-acid producing microorganisms come from the genera Lactobacillus, Bacillus, and Rhizopus (Gruber et al., 2006). Currently, only a few manufacturers produce lactic acid by fermentation. Cargill Dow has the largest plant with a capacity of 182 million kg (400 million lbs) per year. To obtain commercialgrade lactic acid, there are four main processing steps. Initially, the sugars are fermented to lactic acid, which is buffered and becomes lactate salts. The salts are acidified back to acids. The fermentation broth is purified to remove cells, nutrients, and residual sugars. Finally, it is concentrated for sale. Commercial-grade lactic acid can be used to produce propylene glycol, propylene oxide, and epoxides. Propylene glycol is used as a moisturizer and solvent. Propylene oxide can be further processed to polyesters, polycarbonates, and polyurethanes (Kamm and Kamm, 2004), which are used in fabrics, highimpact and temperature-resistant materials, and adhesives. Poly(lactic) acid is a high-potential polymer derived from lactic acid. With improvements in technology and cost, poly(lactic) acid could replace petrochemical-based

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

405

polymers, like polystyrene and polypropylene. Typical applications of these polymers include foams and packaging materials (Gruber et al., 2006). Levulinic acid Levulinic acid, another platform chemical, can be obtained through the biofine process. Rather than relying on fermentation, the biofine process uses dilute sulfuric acid as a catalyst to convert the five- and six-carbon sugars respectively to furfural and hydroxymethyl fufural (HMF) (Brownlee, 1927; Watanabe et al., 2005). Longer residence times convert HMF to levulinic acid. Having both ketone and carboxylic acid functionality allows for chemistry to produce a wide variety of secondary products such as solvents, fuel additives, or solid resins (Huber et al., 2006). Levulinic acid can be converted into diphenolic acid, a component in lubricants, adhesives, and paints. In most commercial applications, bisphenol A is used because it is inexpensive; however, the biofine process can make diphenolic acid more cost effective. Succinic acid may also be produced from levulinic acid. Methyltetrahydrofuran, a fuel additive, is the most promising secondary chemical derived from levulinic acid. The most common by-product of the conversion of cellulose to levulinic acid is formic acid, a highvalue commodity chemical (Hayes et al., 2006).

15.4.5 Research and development A primary research goal of the sugar platform is to develop more capable and efficient cellulase enzymes. Rational design and directed evolution are the most common methods employed to improve cellulase properties. Rational design relies on a detailed understanding of the protein structure and the structure± function relationship. Through site-directed mutagenesis, targeted amino-acid sequences are modified to achieve the desired effect. Because enzymatic cellulose hydrolysis is very complex and current knowledge is limited, design remains a trial-and-error process (Zhang et al., 2006). Directed evolution relies solely on random mutagenesis and screening. Because of its importance, research is being performed to improve the screening and selection strategies (Zhang et al., 2006). Table 15.1 lists companies that will likely have small commercial-scale facilities within five years.

15.5

Carboxylate platform

The carboxylate platform includes all processes that produce carboxylate salts or carboxylic acids from anaerobic fermentation of biomass. These acids/salts can be sold as a primary product, used as chemical intermediates, or fermented with different microorganisms into secondary products. Carboxylic acids are an attractive target for a biorefinery because their carboxyl group enables down-

ß Woodhead Publishing Limited, 2010

406

Bioalcohol production

stream upgrading to a variety of useful compounds, including alcohols, aldehydes, ketones, esters, and olefins (Eggeman and Elander, 2005).

15.5.1 Feedstock considerations Theoretically and in practice, any biodegradable material can be a feedstock for the carboxylate platform. Like the sugar platform, the ideal feedstock has low lignin content, low ash content, low cost per delivered dry tonne, and favorable reactivity. Logistics and availability of feedstocks are critical if a twocomponent feedstock is used; it is rare for both the carbohydrate and the nutrient sources to be co-located and available in the ideal proportion.

15.5.2 Fermentation Lignocellulose hydrolysis produces both five- and six-carbon sugars. The sugar platform ferments these sugars using genetically modified organisms (GMOs) to metabolize both five- and six-carbon sugars. GMOs are expensive to engineer and increase operating costs because sterile conditions must be maintained. Further, GMOs may revert back to the wild-type organism. Alternatively, a defined mixed-culture fermentation can be employed, but they are challenging in practice (Eggeman and Verser, 2006). Rather than fermenting sugars to alcohols, they can be fermented to carboxylic acids. Carboxylic acid fermentations are thermodynamically favored over alcohol fermentations, as shown in Equations 15.2 and 15.3. C6H12O6 ! 2C2H5OH + 2CO2


G = ÿ203.2 kJ/mol

15.2

C6H12O6 ! 3CH3COOH


G = ÿ258.6 kJ/mol

15.3

Mixed-culture acid-producing fermentations are self-regulating because competition forces only the most fit microorganisms to survive (Angenent and Wrenn, 2008). Many known species of acidogens can metabolize both five- and six-carbon sugars to produce carboxylic acids. A mixed-culture of acidogens is ideal for lignocellulose digestion because the microorganisms produce all the enzymes necessary to utilize biomass components (e.g., cellulose, hemicellulose, starch, free sugars, pectin, protein, and fats); thus enabling them to cope with feedstock variation (Angenent et al., 2004). The mixed-acid fermentation is robust and not easily contaminated because the products are thermodynamically favored. An exception is that methanogens can transform carboxylic acids to methane and carbon dioxide. Fortunately, methanogens are fairly fragile and susceptible to inhibitors (e.g., methane analogs like iodoform, ammonium ion, oxygen, ionophores like monensin). The advantages of mixedacid fermentations include the use of a wide variety of feedstocks (e.g., lignocellulose, sewage sludge, and manure), non-sterile operating conditions, use of

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

407

inexpensive materials of construction, and energy savings associated with concentrating nonvolatile organic salts, which are easier to concentrate than a mixture of volatile solvents and water. Mixed-culture fermentations produce a mixture of carboxylic acids ranging from acetic (two carbons) to heptanoic (seven carbons). Single carboxylic acids (e.g., acetic acid) can be obtained using single- or co-culture fermentations. In mixed-culture fermentations, individual acid concentrations can be controlled by adding inhibitors, and using bioaugmentation (plasmid transfer), and phage control (elimination of specific species with viruses) (Angenent and Wrenn, 2008).

15.5.3 The MixAlco process An example of mixed-acid fermentation is the MixAlco process shown in Fig. 15.6. First, the biomass is pretreated with lime and oxygen (or air). The biomass is then fermented into carboxylic acids. Typical acid concentrations and yields are 20±50 g/L and 40±80%, respectively, depending on substrate. Mixed-acid concentrations as high as 100 g/L have been observed with food scraps (Coleman, 2007). Volatile solid loading rates (VSLR) range from 2 to 12 g of volatile solid/ (Lliquid
day). The liquid retention times range from 25 to 70 days. To maintain pH between 5 and 7, the fermentation is typically buffered using calcium carbonate or ammonium bicarbonate. The resulting broth is clarified (descummed) to remove cellular material and suspended solids. The clarified liquid (3±5% w/w salts) is concentrated (dewatered) to ~20% w/w salts (Holtzapple et

15.6 MixAlco process flow diagram.

ß Woodhead Publishing Limited, 2010

408

Bioalcohol production

al., 1999), and then dried to produce crystallized carboxylate salts. Because a mixed-culture is used, the MixAlco process produces a spectrum of carboxylates (two- to seven-carbon salts). Key advantages of the MixAlco process include non-sterile fermentation operating conditions, no enzyme costs, low capital costs, inexpensive high-yield feedstocks, and high-value primary products.

15.5.4 ZeaChem technology ZeaChem Inc. is a startup company in Lakewood, CO. Their process is shown in Fig. 15.7. Currently, ZeaChem focuses on using dried distillers' grain residues and poplar wood chips as feedstocks; however, in theory, their process can use any biomass as a feedstock. Pretreated biomass is enzymatically hydrolyzed to release sugars from cellulose and hemicellulose. The sugar solution of approximately ~5% (w/w) is fed to two-stage fermentation, each stage with a different culture. The first stage ferments the sugars into lactic acid. The optimal pH and temperature are ~6 and 35±45 ëC, respectively. Lactic acid production is completely inhibited when its concentration reaches 10±12%. A 90% conversion is possible with 18±24 hour residence times. The second stage ferments the lactic acid into acetic acid under thermophilic conditions (50±70 ëC) at pH ~7. The acetic acid yield is 85% or better. The fermentation broth from both stages is centrifuged and/or filtered so that only acids proceed to the next step. The second-stage broth contains approximately 5% (w/w) acetic acid. The remaining residues can be sold as single-cell protein or fed back into the fermentors (Verser and Eggeman, 2008).

15.7 Zeachem process (Verser and Eggeman 2008).

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

409

15.5.5 Chemistry and secondary chemicals A myriad of chemicals can be produced from carboxylate salts and carboxylic acids. Figure 15.8 provides an overview of some major pathways to secondary chemicals. Dry calcium carboxylates can undergo thermal conversion to yield ketones. If calcium formate is present, then aldehydes will be formed. Carboxylic acids can undergo vapor-phase decomposition over a thorium or cesium oxide catalyst to produce ketones and aldehydes, respectively. Ketones and aldehydes can be hydrogenated to produce secondary and primary alcohols using high-pressure hydrogen (1.38 MPa, 200 psi) and catalysts (e.g., Raney nickel, reduced CuO/ ZnO, or copper chromite). Using the Williamson ether synthesis process, ethers can be produced from alcohols reacted at 50±100 ëC in the presence of a strong acid catalyst. High yields are achieved using primary alcohols whereas secondary alcohols give poor yields. Reacting alcohols with carboxylic acids in the presence of a strong acid catalyst yields esters, which can then be hydrogenated to produce two alcohols. Using a zeolite catalyst, alcohols can be dehydrated to alkenes, which are subsequently oligomerized into hydrocarbons (e.g., paraffins, cycloalkanes, and aromatics). Amides can be produced by reaction with amines (e.g., ammonia) at high temperatures (200 ëC) and pressures (350±700 kPa; 50±100 psi). To achieve 90% conversion, several catalysts may be used including boric acid, aluminum oxide,

15.8 Flow diagram showing secondary chemicals that can be derived from carboxylic acids and carboxylate salts.

ß Woodhead Publishing Limited, 2010

410

Bioalcohol production

or tin and zinc alkoxides. Acid halides can be produced from carboxylic acids and thionyl chloride in an imidazole-catalyzed reaction. Dehydration of a carboxylic acid yields the corresponding anhydride. The dehydration reaction occurs at high temperatures (750 ëC) and can be promoted by ethyl phosphate catalyst. Carboxylic acids can be fermented to polyhydroxyalkanoates (PHA), which are polyesters used to make biodegradable plastics (Ojumu et al., 2004). Acetic, propanoic, butyric, lactic, and higher-molecular-weight organic acids (i.e., nonanoic acid and octanoic acid) can be converted to PHA by prokaryotic microorganisms (Du and Yu, 2002a). Physical and mechanical properties of PHA polymers can be tuned by regulating the substrate composition, which affects the copolymer composition (Du and Yu, 2002b). Waste solids (lignin, undigested residues, etc.) from pretreatment and fermentation can be burned to generate steam/electricity or gasified to produce synthesis gas (syngas). As described earlier, syngas can be converted to fuels and chemicals, including hydrogen needed for downstream processing in the carboxylate platform. Alternatively, lignin can be dehydroxygenated to yield fuels and other valuable chemicals, such as phenols, cyclohexane, benzene, naphthalene, and phenanthrene (Huber et al., 2006). Other chemical applications of lignin include polypropylene and polyurethane filler (Alexy et al., 2000) and extenders, and supplements in wood adhesives (Kadam et al., 2008).

15.5.6 Research and development Research is being conducted to improve pretreatment methods, evaluate different feedstocks, improve fermentation performance, and find microorganisms that produce higher acid concentrations. Downstream processing is being tested on a pilot scale. Currently, no company utilizes the carboxylate platform at an industrial scale. Table 15.1 lists companies that plan to have small commercial-scale facilities within five years. Terrabon L.L.C. has successfully operated a pilot fermentation and is building a demonstration plant (~5 tonnes sorghum per day) to test a small commercial-scale mixed-acid fermentation. Byogy Renewables Inc. plans to integrate the MixAlco process with a thermochemical process to produce chemicals and fuels. ZeaChem Inc. has successfully operated a pilot plant at its Menlo Park, CA laboratory and is developing a small-scale biorefinery using poplar wood (ZeaChem, 2008).

15.6

Conclusions

Analogous to petroleum refineries, biorefineries process and refine heterogeneous feedstocks into useful chemicals and fuels. Because it is abundant and inexpensive, lignocellulose is the ideal biomass feedstock. Successful commer-

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

411

cial biorefineries will allow nations to become energy independent and produce chemicals that provide a high standard of living. To transition from fossil fuels to lignocellulosic biomass, economically viable biorefineries must be developed. At present, no lignocellulose-based biorefineries exist. Compared to coal, the thermochemical platform is currently too expensive. Additionally, fouling from tar and ash present difficult technical challenges (Wright and Brown, 2007; Yang and Shang-Tian, 2007). The sugar platform requires costly enzymes, but can produce high-value specialty chemicals that offset production costs (Walker and Wilson, 1991; Zeikus et al., 1999). The carboxylate platform shows economic promise and is being vetted at the pilot and demonstration scales. Researchers and investors continue to pursue all three platforms with many companies developing small demonstration facilities within five years or less (see Table 15.1).

15.7

Sources of further information and advice

www.gasification.org http://whitecoalenergy.com www.biomassmagazine.com http://alternativefuels.about.com/od/researchdevelopment/a/gasification.htm www.ge-energy.com/prod_serv/products/gasification/en/overview.htm www.terrabon.com www.zeachem.com www.byogy.com www.rangefuels.com http://en.wikipedia.org/wiki/Bioconversion_of_biomass_to_mixed_alcohol_ fuels

15.8

References

Aiello-Mazzarri, C., Agbogbo, F. K. and Holtzapple, M. T. (2006) Conversion of municipal solid waste to carboxylic acids using a mixed culture of mesophilic microorganisms. Bioresource Technology, 97, 47±56. Alexy, P., KosõÂkovaÂ, B. and PodstraÂnska, G. (2000) The effect of blending lignin with polyethylene and polypropylene on physical properties. Polymer, 41, 4901±4908. Angenent, L. T. and Wrenn, B. A. (2008) Optimizing mixed-culture bioprocessing to convert wastes into bioenergy. In Wall, J. D., Harwood, C. S. and Demain, A. (Eds.) Bioenergy. Washington, DC, ASM Press, pp. 179±194. Angenent, L. T., Karim, K., Al-Dahhan, M. H., Wrenn, B. A. and DomõÂguez-Espinosa, R. (2004) Production of bioenergy and biochemicals from industrial and agricultural wastewater. Trends in Biotechnology, 22, 477±485. Atchison, J. E. and Hettenhaus, J. R. (2003) Innovative Methods for Corn Stover Collecting, Handling, Storing and Transporting. Golden, CO, National Renewable Energy Laboratory. Bok, J. D., Goers, S. K. and Eveleigh, D. E. (1994) Cellulase and xylanase systems of

ß Woodhead Publishing Limited, 2010

412

Bioalcohol production

Thermotoga neapolitana. In Himmel, M. E., Baker, J. O. and Overend, R. P. (Eds.) Enzymatic Conversion of Biomass for Fuels Production. ACS Symposium Series 566, Washington, DC, American Chemical Society. Bridgwater, A. V. (Ed.) (2001) Progress in Thermochemical Biomass Conversion, Oxford, Blackwell Science. Bridgwater, A. V. and Grassi, G. (Eds.) (1991) Biomass Pyrolysis Liquids Upgrading and Utilisation, New York, Elvevier Applied Science. Brownlee, H. J. (1927) Furfural manufacture from oat hulls: I ± a study of the liquid-solid ratio. Ind. Eng. Chem., 19, 422±424. Chen, G., Andries, J., Luo, Z. and Spliethoff, H. (2003) Biomass pyrolysis/gasification for product gas production: the overall investigation of parametric effects. Energy Conversion and Management, 44, 1875±1884. Coleman, S. (2007) Continuous fermentation of food scraps with constant pH control to produce carboxylic acids. Chemical Engineering. College Station, TX, Texas A&M University. Coughlan, M. P. and Ljungdahl, L. G. (1988) Comparative biochemistry of fungal and bacterial cellulytic enzyme system. In Aubert, J. P., Beguin, P. and Millet, J. (Eds.) Biochemistry and Genetics of Cellulose Degradation. New York, Academic Press. Dien, B. S., Cotta, M. A. and Jeffries, T. W. (2003) Bacteria engineered for fuel ethanol production: current status. Applied Microbiology and Biotechnology, 63, 258±266. Du, G. and Yu, J. (2002a) Green technology for conversion of food scraps to biodegradable thermoplastic polyhydroxyalkanoates. Environ. Sci. Technol., 36, 5511±5516. Du, G. and Yu, J. (2002b) Metabolic analysis on fatty acid utilization by Pseudomonas oleovorans: mcl-poly(3-hydroxyalkanoates) synthesis versus [beta]-oxidation. Process Biochemistry, 38, 325±332. Duff, S. J. B. and Murray, W. D. (1996) Bioconversion of forest products industry waste cellulosics to fuel ethanol: a review. Bioresource Technology, 55, 1±33. Effendi, A., Gerhauser, H. and Bridgwater, A. V. (2008) Production of renewable phenolic resins by thermochemical conversion of biomass: a review. Renewable and Sustainable Energy Reviews, 12, 2092±2116. Eggeman, T. and Elander, R. T. (2005) Process and economic analysis of pretreatment technologies. Bioresource Technology, 96, 2019±2025. Eggeman, T. and Verser, D. (2006) The importance of utility systems in today's biorefineries and a vision for tomorrow. Applied Biochemistry and Biotechnology, 130, 361±381. EIA (2007) Annual Energy Review 2007. Washington, DC, Energy Information Agency. Gasification Technology Council (2008). Arlington, VA, Gasification Technology Council. Goyal, H. B., Seal, D. and Saxena, R. C. (2008) Bio-fuels from thermochemical conversion of renewable resources: a review. Renewable and Sustainable Energy Reviews, 12, 504±517. Gray, K. A., Zhao, L. and Emptage, M. (2006) Bioethanol. Current Opinion in Chemical Biology, 10, 141±146. Gruber, P., Henton, D. E. and Starr, J. (2006) Polylactic acid from renewable resources. In Kamm, B., Gruber, P. R. and Kamm, M. (Eds.) Biorefineries ± Industrial Processes and Products. Status Quo and Future Directions. Weinheim, Wiley-VCH. Guettler, M. V. and Jain, M. K. (1996) Method for making succinic acid, bacterial variants for use in the process, and methods for obtaining variants. US patent 5,573,931.

ß Woodhead Publishing Limited, 2010

Chemical production from lignocellulosic biomass

413

Guettler, M. V., Rumbler, D. and Jain, M. K. (1999) Acinobacillus succinogenes sp. nov., a novel succinic-acid-producing strain from the bovine rumen. International Journal of Systematic Bacteriology, 49, 207±216. Hayes, D. J. (2009) An examination of biorefining processes, catalysts and challenges. Catalysis Today, 145, 138±151. Hayes, D. J., Fitzpatrick, S., Hayes, M. H. B. and Ross, J. R. H. (2006) The biofine process ± production of levulinic acid, furfural, and formic acid from lignocellulosic feedstocks. In Kamm, B., Gruber, P. R. and Kamm, M. (Eds.) Biorefineries ± Industrial Processes and Products: Status Quo and Future Directions. Weinheim, John Wiley & Sons, Inc. Higman, C. and Burgt, M. V. D. (2003) Gasification, New York, Elsevier. Holtzapple, M. T. et al. (1999) Biomass conversion to mixed alcohol fuels using the MixAlco process. Applied Biochemistry and Biotechnology, 79, 609±631. Huang, Y. F., Kuan, W. H., Lo, S. L. and Lin, C. F. (2008) Total recovery of resources and energy from rice straw using microwave-induced pyrolysis. Bioresource Technology, 99, 8252±8258. Huber, G. W., Iborra, S. and Corma, A. (2006) Synthesis of transportation fuels from biomass: chemistry, catalysts, and engineering. Chem. Rev., 106, 4044±4098. Jeffries, T. W. and Jin, Y. S. (2004) Metabolic engineering for improved fermentation of pentoses by yeasts. Applied Microbiology and Biotechnology, 63, 495±509. Kadam, K. L., Chin, C. Y. and Brown, L. W. (2008) Flexible biorefinery for producing fermentation sugars, lignin and pulp from corn stover. J Ind Microbiol Biotechnol, 35, 331±341. Kamm, B. and Kamm, M. (2004) Principles of biorefineries. Applied Microbiology and Biotechnology, 64, 137±145. Kamm, B. and Kamm, M. (2007) Biorefineries ± multi product processes. White Biotechnology. Berlin, Springer-Verlag Berlin. Lepori, W. A. and Soltes, E. J. (1985) Thermochemical conversion for energy and fuel. In Hiler, E. A. and Stout, B. A. (Eds.) Biomass Energy: A Monograph. College Station, TX, Texas A&M Univeristy Press. Mosier, N., Wyman, C., Dale, B., Elander, R., Lee, Y. Y., Holtzapple, M. and Ladisch, M. (2005) Features of promising technologies for pretreatment of lignocellulosic biomass. Bioresource Technology, 96, 673±686. Ojumu, T. V., Yu, J. and Soloman, B. O. (2004) Production of Polyhydroxyalkanoates, a bacterial biodegradable polymer. African Journal of Biotechnology, 3(1), 18±24. Phillips, S., Aden, A., Jechura, J., Dayton, D. and Eggeman, T. (2007) Thermochemical Ethanol via Indirect Gasification and Mixed Alcohol Synthesis of Lignocellulosic Biomass. Golden, CO, NREL. Reed, T. B. (Ed.) (1981) Biomass Gasification: Principles and Technology, Park Ridge, NJ, Noyes Data Corporation. RFA (2008) US Fuel Ethanol Industry biorefineries and Production Capacity. Washington, DC, Renewable Fuels Association. Rooney, W. L. (2007) Sorghum Breeding and Genetics, ed. College Station, TX. Rooney, W. L., Blumenthal, J., Bean, B. and Mullet, J. E. (2007) Designing sorghum as a dedicated bioenergy feedstock. Biofuels Bioproducts & Biorefining, 1, 147±157. Sierra, R., Smith, A., Granda, C. and Holtzapple, M. T. (2008) Producing fuels and chemicals from lignocellulosic biomass. Chemical Engineering Progress. American Institute of Chemical Engineers. Sun, Y. (2002) Enzymatic hydrolysis of rye straw and bermudagrass for ethanol production. PhD dissertation, Raleigh, N.C.: North Carolina.

ß Woodhead Publishing Limited, 2010

414

Bioalcohol production

Verser, D. and Eggeman, T. J. (2008) Process for Producing Ethanol. United States, ZeaChem, Inc. Walker, L. P. and Wilson, D. B. (1991) Enzymatic Hydrolysis of cellulose: an overview. Bioresource Teghnology, 36, 3±14. Watanabe, M., Aizawa, Y., Iida, T., Levy, C., Aida, T. M. and Inomata, H. (2005) Glucose reactions within the heating period and the effect of heating rate on the reactions in hot compressed water. Carbohydrate Research, 340, 1931±1939. Wright, M. M. and Brown, R. C. (2007) Comparative economics of biorefineries based on the biochemical and thermochemical platforms. Biofuels, Bioproducts and Biorefining, 1, 49±56. Yang, S.-T. and Shang-Tian, Y. (2007) Bioprocessing ± from biotechnology to biorefinery. Bioprocessing for Value-Added Products from Renewable Resources. Amsterdam, Elsevier. ZeaChem (2008). Lakewood, CO, ZeaChem. Zeikus, J. G., Jain, M. K. and Elankovan, P. (1999) Biotechnology of succinic acid production and markets for derived industrial products. Applied Microbiology and Biotechnology, 51, 545±552. Zhang, Y. H. P., Himmel, M. E. and Mielenz, J. R. (2006) Outlook for cellulase improvement: Screening and selection strategies. Biotechnology Advances, 24, 452±481.

ß Woodhead Publishing Limited, 2010

16

Production of longer-chain alcohols from lignocellulosic biomass: butanol, isopropanol and 2,3-butanediol  PEZ CONTRERAS, W. KUIT, A. M. LO M. A. J. SIEMERINK, S. W. M. KENGEN, J . S P R I N G E R and P . A . M . C L A A S S E N , Wageningen University and Research Centre, The Netherlands

Abstract: The four carbon alcohols butanol and 2,3-butanediol are the longest chain alcohols found as natural major end-products of microbial fermentation. They represent important bulk chemicals widely used in industry as solvents for lacquers, paints and similar, or as intermediates in chemical synthesis reactions. In this chapter the biological methods for the production of butanol, isopropanol and 2,3-butanediol, including advances in strain development and in the fermentation processes, are described in detail to provide the reader with a state-of-the-art view on these fields. In addition, the biological production of other important commercial alcohols (2-butanol, 1,2- and 1,3-propanediol) with the potential to be biologically produced is briefly reviewed. Key words: butanol, biobutanol, acetone, ethanol, isopropanol, 2,3butanediol, ABE fermentation, white biotechnology.

16.1

Introduction

The four carbon alcohols butanol and 2,3-butanediol are the longest chain alcohols found as natural major end-products of microbial fermentation. They represent important bulk chemicals widely used in industry as solvents for lacquers, paints and similar, or as intermediates in chemical synthesis reactions (see Section 16.2). With the exception of a relatively small proportion of butanol produced by fermentation in China and Brazil (Ni and Sun, 2009), these alcohols are currently petrochemically synthesized. The annual production of butanol has been estimated to be approximately 2.8 million tons in 2006, with continuous increase in demand and capacity in the coming years (Yuan, 2007). The good properties of 2,3-butanediol and specially of butanol as fuel additive (or as precursors thereof) for the replacement of fossil-derived transport fuels, together with the versatility of the solvent-producing organisms to ferment most sugars present in nature, have renewed the attention on the fermentative production of

ß Woodhead Publishing Limited, 2010

416

Bioalcohol production

these components from renewable biomass resources in the recent years (DuÈrre, 2007, 2008, Celinska and Grajek, 2009). The butanol fermentation process is generally known as the acetone, butanol and ethanol (ABE) fermentation. Some of the butanol-producing microorganisms produce isopropanol as co-product of the fermentation instead of acetone (the IBE fermentation process). Since isopropanol is considered an interesting product and preferred over acetone for some applications (i.e. as fuel extender) (Rogers et al., 2006) it is included in this chapter. The ABE fermentation has a long industrial history and has been in operation almost uninterrupted from the 1910s, when the first plants were built in the United Kingdom (Jones and Woods, 1986). Presently, the process is being re-introduced in China (Ni and Sun, 2009) and other countries. The ABE process was the first industrial process for the production of butanol, and started to be replaced by the emerging petrochemical industry from the 1960s due to economical considerations. However, large-scale ABE plants continued to be in operation in Russia and China until the mid 1990s (Zverlov et al., 2006, Chiao and Sun, 2007). The fermentation of sugars to 2,3-butanediol was described for the first time in 1906, and since then it has been widely studied (Syu, 2001, Rogers et al., 2006, Celinska and Grajek, 2009). This fermentation process has not been operated industrially, although there are reports of successful pilot-scale tests during the 1930s±1940s for the production of 2,3-butanediol as precursor for 1,3-butadiene (Celinska and Grajek, 2009). As it occured with the ABE fermentation, the developments in the petrochemical industry resulted in a more cost-effective process for the production of 2,3-butanediol, causing loss of interest in the fermentative process. In the present chapter the biological methods for the production of butanol, isopropanol and 2,3-butanediol, including recent advances in strain development and in the fermentation processes, are described in detail to provide the reader with a state-of-the-art view on these fields. In addition, the biological production of other important commercial alcohols (1,2- and 1,3-propanediol) or an alcohol with the potential to be biologically produced (2-butanol) is briefly reviewed.

16.2

Characteristics and uses of butanol, acetone, isopropanol and 2,3-butanediol

Some physical and chemical characteristics of the alcohols and by-products mentioned in this chapter are summarized in Table 16.1. Butanol (n-butyl alcohol, 1-butanol, CAS number [71-36-3]) represents a widely used bulk chemical with a broad range of industrial applications. Currently, butanol is mainly petrochemically synthesized by reduction of butyraldehyde or from ethylene oxide (O'Neil et al., 2001, Yuan, 2007). During the first part of last century, butanol was produced at an industrial scale by fermentation of sugars or starch, via the ABE fermentation process (Jones and

ß Woodhead Publishing Limited, 2010

Table 16.1 Physical and chemical properties of alcohols and other products studied in this chapter Property

Ethanol

ß Woodhead Publishing Limited, 2010

1-Butanol

Acetone

74.12 ÿ89.8

58.08 ÿ94.7

117.7

56.1

Isopropanol

1,2-Propanediol

1,3-Propanediol

2,3-Butanediol

Chemical structure Molecular weight Melting point at 101.3 kPa (ëC) Boiling point at 101.3 kPa (ëC) Specific gravity* Vapor pressure at 25 ëC (kPa)

46.07 ÿ114.1 78.29 0.78920 7.87

0.80920 0.86

* Temperature indicated as superscript. Data from Lide (2008).

0.78025 30.8

60.1 ÿ87.9

76.1 ÿ60

76.1 ÿ27.7

82.3

187.6

214.4

0.78125 6.02

1.03620 0.02

1.05320 0.007

90.12 7.6 182.5 1.00320 0.02

418

Bioalcohol production

Woods, 1986). From the 1960s, this process was slowly replaced by the petrochemical process for butanol production due to economical considerations. Butanol is used primarily as a chemical intermediate in the production of butyl acrylate, used in the manufacturing of, e.g., polymeric coatings, adhesives, elastomers and plastics; and butyl methacrylate, used in resins, dental products, as an oil additive and in the leather and paper industry. Other products derived from butanol are butyl glycol, used mainly as an industrial solvent, butyl acetate, used in paints and as flavouring agent in the food industry, butylamine, e.g. used in the production of thiocarbazides and butylbenzenesulfonamide, used as plasticizer in nylon (DuÈrre, 2008). In response to the increasing global demand on transportation fuels derived from renewable resources instead of fossil fuels, interest in the fermentative production of butanol has emerged in the recent years. Presently, ethanol produced from biomass is the most well-known alcohol for blending with gasoline, even though butanol has a number of advantages over ethanol. These, derived from the physical properties of butanol (see Tables 16.1 and 16.2), include: (i) higher energy density, (ii) lower vapour pressure, (iii) less corrosive and (iv) lower solubility in water (preventing phase separation of the mix alcohol/ gasoline) and better miscibility with the fuel. Butanol and its blends can be transported using existing pipelines (this is not the case for ethanol) and used in current engines without modifications (Ladisch, 1991, Anonymous, 2006, DuÈrre, 2007, 2008). The possibility to use pure butanol as a motor fuel has been shown by David Ramey, from ButylFuel, LLC, during two long trips in the US in 2005 and 2007 (www.butanol.com, accessed by 13 January 2009). In 2006, the multinationals British Petroleum and Dupont started a joint venture to further develop and commercialize the fermentative production of butanol. The start of the construction of an ABE demonstration plant together with a full scale bioethanol plant (capacity 420 ML/y from wheat) at BP's industrial complex in Table 16.2 Properties of alcohols as fuels compared to gasoline Compound

Butanol Ethanol Methanol Isopropanol Gasoline 2,3-Butanediol 2-Butanol

Heat of Heat of Energy combustion vaporization density (
cHë) (kJ/mol) (
vapHmë) (kJ/mol) (MJ/kg) 2670 1368 726 2006 4817* 2461 2661

51 42 37 45 36 67 42

36.0 29.7 22.7 33.4 43.5 27.3 35.9

RON MON

113 130 133 121 95

94 96 99 96 85

Data from NIST (2009), Houben (1995), Riley (1995) Ladisch (1991) and Euse¨bio (2003); * data for heptane. Abbreviations: RON, Research octane number, and MON, Motor octane number, are octane rating numbers related to the performance of a liquid in a combustion engine.

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

419

Hull (UK) was announced in 2007, and it is expected to be operational in 2009. In China the current ABE production is based on corn, where 11 large scale butanol fermentation facilities were in operation by mid 2008 (Ni and Sun, 2009). In Brazil, one large-scale ABE plant, using sugar cane as substrate, is in operation and will be described later in this chapter. Acetone (2-propanone, CAS number [67-64-1]) can be derived from fossil oil by synthesis from cumene (a by-product in phenol manufacture) by the hydroperoxide process or by catalytic dehydrogenation of isopropanol or produced from renewable resources using the ABE fermentation process. It represents a widely used solvent, for example in printing inks, oils, plastics, adhesives, varnishes and resins and in the pharmaceutical industry. Acetone is also a precursor of other products such as acrylic plastics and explosives (O'Neil et al., 2001). Isopropanol (2-propanol, CAS number [67-63-0]) is currently manufactured from propylene, a by-product of petroleum refining, either by an indirect or a catalytic hydration process. Alternatively, some of the bacterial strains performing the ABE fermentation reduce the acetone directly to isopropanol, resulting in the IBE fermentation process (Jones and Woods, 1986). Isopropanol is an inexpensive alcohol with a large number of applications. Besides being an important chemical intermediate in many processes, it is used in antifreeze solutions (dry gas) for fuel tanks, as a solvent in products such as oils, ink and cosmetics, as antiseptic and disinfectant agent replacing ethanol, and in the pharmaceutical industry (O'Neil et al., 2001). Isopropanol also has good properties as a blending component for gasoline (Rogers et al., 2006). A mix of 99.5% (v/v) of isopropanol-butanol-ethanol (3:6:1 volume ratio) and 0.5% (v/v) water has been tested as motor fuel replacing gasoline satisfactorily (Groot et al., 1986). Diisopropyl ether (DIPE), which can be derived from isopropanol by a reduction reaction, has been identified as a possible alternative to methyl tertiary butyl ether (MTBE) as oxygenated blending additive to gasoline (Huang et al., 1990). In this way, the applications of isopropanol are extended. DIPE can also be manufactured from acetone feedstock, in a route where acetone is first hydrogenated to isopropanol. Direct conversion of acetone to isopropanol and DIPE using metallic catalysts has been reported (single step catalytic production of diisopropyl ether (DIPE) from acetone feedstock over nickel based catalysts (Chidambarama and Viswanathan, 2007). 2,3-Butanediol (2,3-butylene glycol, CAS number [513-85-9]) can be produced by a number of microorganisms in the genera Bacillus, Enterobacter, Klebsiella, Pseudomonas, and Serratia. Three different stereomers exist, viz. D(-)- (2R,3R)-2,3-butanediol, L(+)-(2S,3S)-2,3-butanediol and meso-(2R,3S)-2,3butanediol, which can all be produced biologically. The 2,3-butanediol fermentation process at pilot scale using strains of Klebsiella oxytoca and Bacillus polymyxa has been described during the years of World War II for the production of 2,3-butanediol as precursor of 1,3-butadiene used for synthetic rubber.

ß Woodhead Publishing Limited, 2010

420

Bioalcohol production

However, the process was not commercially implemented due to competition with the petrochemical process for the production of 2,3-butanediol (Celinska and Grajek, 2009, Syu, 2001, Voloch et al., 1985). Nowadays, 2,3-butanediol is still manufactured by the petrochemical industry. The commercial product is usually either the meso- or the D(-)-form. The meso-form is derived from trans2,3-epoxybutane and the D/L mixture is derived from cis-2,3-epoxybutane (O'Neil et al., 2001). 2,3-Butanediol can be used as component in anti-freeze agents, printing inks, perfumes, fumigants, moistening and softening agents, explosives and plasticizers, or carriers for pharmaceuticals (Xiu and Zeng, 2008). Furthermore, dehydration of 2,3-butanediol results in methylethyl ketone, also usable as fuel additive, or as organic solvent for resins and lacquers. The 2,3-butanediol analogues, acetoin and diacetyl, obtained after dehydrogenation, can be used as flavouring agents in dairy products, margarines or in cosmetics (Garg and Jain, 1995).

16.3

Production of butanol and isopropanol by clostridia

Strains of anaerobic bacteria belonging to the genus Clostridium ferment carbohydrates (both C5 or C6 mono-saccharides, or polymers such as starch) into a mix of alcohols and ketones through a shared metabolic pathway (Fig. 16.1). In this fermentation, known as `solvent fermentation', butanol is the main product, with acetone or isopropanol as major co-products. Therefore, the fermentative production of these important chemicals is usually studied together.

16.3.1 Microorganisms and metabolism Louis Pasteur was the first to describe the production of butanol by microbial fermentation (Pasteur, 1862). A number of studies on butanol production by anaerobic bacteria followed, and in 1905, Schardinger reported the production of acetone by fermentation of glucose by a Bacillus species (Beijerinck, 1893, Schardinger, 1905). The production of isopropanol, together with butanol, by fermentation was reported in 1906 (Pringsheim, 1906a, 1906b). The combined formation of acetone and butanol by one single organism was reported by Fernbach in 1911, during research on finding a biological route for production of butanol, which could be used as a precursor for synthetic rubber, at industrial scale (DuÈrre and Bahl, 1996). The isolation by Chaim Weizmann of bacterial strains able to convert starchy materials into butanol and acetone in 1912 formed the basis for the ABE industrial fermentation process that started in the United Kingdom and expanded worldwide during the first half of the 20th century. The history of the development, expansion and decline of the ABE industrial process

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

421

16.1 Biochemical pathways in C. acetobutylicum and C. beijerinckii, modified from Jones and Woods (1986). Enzymes catalizing the different reactions are indicated by abbreviation of their name: Ldh, lactate dehydrogenase; Als, acetolactate synthase; Ald, acetolactate decarboxylase; HydA, hydrogenase; Pta, phosphate acetyltransferase (phosphotransacetylase); AckA, acetate kinase; ThlA, thiolase; Hbd, 3-hydroxybutyryl-CoA dehydrogenase; Crt, crotonase; Bcd, butyryl-CoA dehydrogenase; Ptb, phosphate butyryltransferase (phosphotransbutyrylase); Buk1, Buk2, butyrate kinase 1 and 2, respectively; AdhE, AdhE2, aldehyde-alcohol dehydrogenase 1 and 2, respectively; BdhA, BdhB, butanol dehydrogenase A and B, respectively; CtfAB, acetoacetyl-CoA:acetate/butyrate:CoA transferase subunits A and B; Adc, acetoacetate decarboxylase; Adh, primary and secondary alcohol dehydrogenase (in C. beijerinckii NRRL B593) (Hiu et al., 1987). Other abbreviations: CoA, coenzyme A; Pi, phosphate; Fd, ferredoxin; Etf, electron transfer flavoprotein.

ß Woodhead Publishing Limited, 2010

422

Bioalcohol production

is a prime example of the entanglement of industrial development and political implications, with Weizmann becoming the first president of the state of Israel as an example. Detailed reviews covering this exciting history from different viewpoints have been published (Rose, 1961, Jones and Woods, 1986, DuÈrre and Bahl, 1996, DuÈrre, 1998, Walton and Martin, 1979, Jones, 2001, Rogers et al., 2006), and therefore we refer the reader to these and it will not be addressed in this chapter. A large number of clostridial strains able to produce neutral solvents (acetone, butanol, ethanol, isopropanol) from different carbohydrate substrates have been isolated and used in patent applications over the years (Jones and Keis, 1995). Initially, solvent-producing strains were classified mainly as C. acetobutylicum or C. beijerinckii, but important physiological and genetic differences observed between strains belonging to the same group made it necessary to make a clear classification of the existing strains. Detailed DNA similarity studies and 16S rDNA sequence comparisons between strains belonging to different culture collections showed that the existing strains can be classified into four distinct groups (Johnson et al., 1997, Johnson and Chen, 1995, Jones and Keis, 1995, Keis et al., 2001) that are all members of cluster I of the clostridia: I, C. acetobutylicum (type strain ATCC 824); II, C. beijerinckii (type strain NCIMB 9362); III, C. saccharoacetobutylicum (type strain NCP 262), and IV, C. saccharoperbutylacetonicum (type strain N1-4). All known solventogenic clostridia are mesophilic and contain DNA with low GC-content. The biochemical pathways used for the conversion of carbohydrates into hydrogen, carbon dioxide, volatile fatty acids, ethanol, butanol, acetone or isopropanol by solvent-producing clostridial species have been extensively studied and characterized (Jones and Woods, 1986). Hexose sugars are metabolized via the Embden-Meyerhof pathway (Fig. 16.1). One mol of hexose is converted to 2 moles of pyruvate, with the net production of 2 moles of ATP and 2 moles NADH. The utilization of pentoses takes place via the pentose phosphate pathway (Warburg-Dickens pathway), yielding 5 moles of ATP and 5 moles of NADH and 2 moles of fructose-6-phosphate and 1 mole of glyceraldehyde-3phosphate (which both enter the glycolytic pathway) per 3 moles of pentoses (Jones and Woods, 1986). The pyruvate resulting from the glycolysis is cleaved by pyruvate ferredoxin oxidoreductase in the presence of coenzyme A to yield CO2, acetyl-CoA and reduced ferredoxin. Acetyl-CoA is the central intermediate in the branched fermentation pathways leading to both acid and solvent production (Fig. 16.1). Despite numerous physiological studies, it is still not completely understood how solvent production is regulated at the molecular level. During the exponential growth phase mainly acids are produced (acidogenic phase) by most strains, including strains of C. acetobutylicum. By reaching the early stationary phase the production of solvents starts, a phenomenon known as `metabolic switch'. Initiation of solvent formation requires low pH, threshold concentrations of

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

423

acetate and butyrate, and a suitable growth-limiting factor such as phosphate or sulfate (Andreesen et al., 1989, Roos et al., 1985). Solvent formation appears to be associated with the availability of ATP and NAD(P)H (Meyer and Papoutsakis, 1989) and can be controlled, in continuous culture, by varying the glucose concentration (Bahl et al., 1986). Increasing the reducing power in the cell by inhibiting hydrogenase can also enhance solvent formation. The reported role of a DNA-binding protein, Spo0A, on the expression of genes that are jointly involved in solvent production and sporulation in C. beijerinckii (Ravagnani et al., 2000), suggests that these two phenomena may be connected. Most of the genes that encode enzymes involved in primary metabolism have been characterized, and there is extensive knowledge on the regulation of the expression of genes involved in acid and solvent production and their function in the metabolic pathways (Girbal and Soucaille, 1998, Woods, 1995). The beststudied strain, at the genetic level is C. acetobutylicum ATCC 824. In this strain the genes involved in solvent production are located on a megaplasmid of 210 kb (pSOL1). The loss of this megaplasmid results in asporogenous strains unable to make solvents (Cornillot et al., 1997a). The whole genome of C. acetobutylicum ATCC 824 has been sequenced and is publicly available (http:// www.ncbi.nlm.nih.gov) (NoÈlling et al., 2001). The presence of a similar megaplasmid in other strains has not been reported. The genome of another butanol producer, strain C. beijerinckii NCIMB 8052 has been sequenced (http:// genome.jgi-psf.org/finished_microbes/clobe/clobe.home.html), although a scientific publication describing it has not been published yet. The solvent-producing clostridial strains gradually lose their ability to produce solvents when they are kept at vegetative stage for long periods of time (i.e., during repeated transfers of growing cultures or long periods of continuous cultivation), a phenomenon known as `degeneration' (Jones and Woods, 1986). In addition to their lack of solvent production, degenerated strains show different morphological and physiological characteristics compared to the parent strain; larger and translucent colonies with irregular shapes (Woolley and Morris, 1990), a longer or thinner cell shape and a characteristic infrared spectrum (Schuster et al., 2001). The degeneration of the strains is a result of genetic changes in the cells, which are not fully characterized yet. The loss of solvent production seems to be linked to the loss of the ability to sporulate, since several sporogenous degenerated mutants have been isolated during prolonged continuous cultures (Stephens et al., 1985, Woolley and Morris, 1990). However, solvent-producing asporogenous mutants are also known. The role of the regulator Spo0A in the initiation of solventogenesis and sporulation has been elucidated (Ravagnani et al., 2000), but whether it plays a role in the regulation of degeneration is still not known. As mentioned earlier, the loss of the pSOL1 plasmid in C. acetobutylicum ATCC 824 results in degenerated strains (Cornillot et al., 1997b), but in other solvent-producing strains, a different mechanism for degeneration is expected, since they do not contain such

ß Woodhead Publishing Limited, 2010

424

Bioalcohol production

a plasmid and the genes involved in solvent production are still present in the degenerated mutant (Chen and Blaschek, 1999, Kosaka et al., 2007). Recently, a cell density-dependent regulatory mechanism (quorum sensing) has been proposed to be involved in the degeneration of strain C. saccharoperbutylacetonicum N1-4, since solvent production in a degenerated mutant could be restored after the addition to the cultivation medium of concentrated broth extract from a wild-type fermentation (Kosaka et al., 2007).

16.3.2 Process technology The ABE process was the first industrial fermentation to be run using pure cultures under aseptic conditions, and represented the second largest fermentation (after the ethanol fermentation) from the late 1910s to the 1960s of the 20th century. During this period, plants were built and operated all over the world (Rogers et al., 2006, Jones and Woods, 1986), with butanol being the main product of interest. The years of the two World Wars were an exception since acetone from the ABE fermentation was needed for cordite manufacturing and became the most valuable product. Details on the operation conditions of many of these ABE plants and processes are available from a number of publications and patent applications written during those times and in later years (Beesch, 1952, 1953, Prescott and Dunn, 1959a, Rose, 1961; Gibbs, 1983, Zverlov et al., 2006, Rogers et al., 2006; Chiao and Sun, 2007). Data on the operation of the plant run at Germiston (South Africa) by National Chemical Products (now NCPalcohols) from 1937 to 1982 (including fermentation procedures and reports on phage infections) have been summarized by several authors (Spivey, 1978, Jones and Woods, 1986, Jones, 2001). Although the ABE fermentation was the major process, several industrial scale plants were run using strains producing butanol and isopropanol (the IBE process) as main products (Walton and Martin, 1979, Prescott and Dunn, 1959b). Most solvent-producing clostridial strains known to date are able to grow on a variety of carbohydrates (including hexoses such as glucose and arabinose, pentoses such as xylose and polymers such as starch or xylan) (Mitchell, 1998), with yields of total solvents (ABE and/or IBE) produced varying between 0.3 and 0.4 grams of total solvents per gram of sugar equivalents consumed. The relative concentration of end products in the fermentation broth is characteristic for the clostridial strain used and the growth conditions. The ratio 6 : 3 : 1 (B : A : E) is the ratio found for the Weizmann strain and considered typical for strains belonging to the species C. acetobutylicum (Jones and Woods, 1986). In Table 16.3, data on production of solvents at commercial scale from starch or sugary substrates are summarized. The original microorganism isolated by Weizmann (classified as C. acetobutylicum) had the capacity to ferment starches (e.g., from cornmeal mash or from potatoes) directly to butanol and acetone as major products, with minor

ß Woodhead Publishing Limited, 2010

Table 16.3 Solvent production during large-scale fermentation of molasses or starchy substrates %, weight product/weight substrate

ß Woodhead Publishing Limited, 2010

Culture

Substrate

Yield (%)

Butanol

Acetone

Ethanol

C. acetobutylicum C. acetobutylicum

Corn starch Wheat flour. Data from plant in Grosny (Russia) Wheat flour/sugar beet molasses mix (38.8/61.2%). Data from plant in Grosny (Russia) Wheat flour/sugar beet molasses/pentosane hydrolysate from lignocellulose mix (28.6/66.4/5.0). Plant in Dokshukino (Russia) Molasses

36.2 38.2

22.5 21.8

11 12.1

2.7 4.3

Ni and Sun (2009) Zverlov et al. (2006)

34.8

20.8

10.5

3.5

Zverlov et al. (2006)

35.8

20.3

9.9

5.6

Zverlov et al. (2006)

29.9

18.0

9.0

3.0

38.0 30±33

22.0 15.9± 21.4

10.6 0.3±7.9

5.3

Spivey (1978) Keis et al. (2001) Beesch (1953) Prescott and Dunn (1959b)

C. acetobutylicum C. acetobutylicum

C. acetobutylicum (now C. saccharobutylicum) C. acetobutylicum C. toanum

Corn starch Cane juice (cane sugar mixed with blackstrap molasses)

Data shown correspond to yields in percentage of weight of product by weight of substrate in the medium.

Isopropanol

5.7±14.5

References

426

Bioalcohol production

quantities of acetic and butyric acids, carbon dioxide and hydrogen as byproducts. The strains used in starch-based processes were able to degrade the starch polymer, making the addition of enzymes for hydrolysis unnecessary (Beesch, 1953). After some years, starchy substrates were replaced by molasses as the main raw material for the fermentation, and new solvent-producing strains with good fermentative properties on molasses were isolated and patented (Beesch, 1952, Walton and Martin, 1979). However, usage of both types of substrates (in some cases mixed, see Table 16.3) continued, depending on raw material availability and local economical conditions. Corn starch has been the most widely used starchy raw material. Besides, sweet potato, cassava and wheat starch (Zverlov et al., 2006) have been reported as substrates for the industrial process. Procedures for the pre-treatment of the corn substrates for the fermentation have been described in detail (Beesch, 1953, Walton and Martin, 1979). In the past, the corn kernel was milled and, if economically interesting, the corn oil was extracted from the germ and treated separately. The degerminated meal was mixed with water (and eventually stillage) at concentrations of approx. 8.5% in weight (based on the original corn dry weight), to ensure complete utilization of the sugars. Because of the toxicity of the products (mainly butanol) to the bacteria, concentrations of butanol higher than 13 g/L (corresponding to a sugar consumption of approximately 60 g/L) were not achieved, and excess substrate remained unfermented. In the case of molasses, two types were the most commonly used; blackstrap molasses and high test molasses. Blackstrap molasses refers to the concentrated syrup remaining after the crystallization of sucrose from sugar cane juice containing an average of 52% total sugars, consisting of approximately 30% sucrose and 22% invert sugar (glucose and fructose). High-test molasses (also called invert molasses) is the concentrate of the sugar cane juice and contains mainly glucose and fructose (originated from the `inversion' of sucrose done to prevent crystallization) (Walton and Martin, 1979). High-test molasses was preferred over the blackstrap molasses, since it contains less solids and salts and its composition is more consistent from one batch to another. In contrast to starchbased media, molasses media needed supplementation with buffering agents and nutrients to give good solvent yields. Normally, ammonia (added at concentrations of 1.2±1.3% in weight of sugar content of the raw material) was used as buffering agent and as nitrogen source. Phosphate, in the form of superphosphate, was added in concentrations varying between 0.05 and 0.2% in weight of sugar content (Walton and Martin, 1979). Also, molasses media were supplemented with sources of complex nitrogen and other nutrients, such as yeast water, corn steep liquor or stillage for improved solvent yields (Beesch, 1952). For optimal fermentation yields and sugar utilization, the total sugar concentration in the molasses media was adjusted to 5.5±7.5% (sugar weight per volume), depending on the fermentation conditions, similarly as described for starchy media. In Table 16.3, examples of solvent production on molasses media are shown.

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

427

The traditional solvent fermentation is a batch process operated under sterile conditions to prevent contaminations (Jones and Woods, 1986), although continuous fermentation at large scale has been reported recently to be conducted in China (Chiao and Sun, 2007). The growth conditions for the solvent-producing bacteria (temperature between 30 and 37 ëC, starting pH of the medium between 6 and 7) are also suitable for other anaerobic organisms. Therefore, the fermentation equipment (tanks, pipes, etc.) needs to be sterilized (normally by using steam) and kept sterile at all times during operation. The fermentation media are sterilized by cooking using conditions specific for the substrate. Corn starch was usually cooked for approximately 60 minutes at 121±127 ëC in order to gelatinize the starch and sterilize the mash, while molasses media were treated using milder conditions (Beesch, 1952, 1953). The most contamination problems were caused by lactic acid bacteria, resulting in fermentations with low or no solvent production (Beesch, 1953). After sterilization, filling and inoculation of the bioreactor, the anaerobic fermentation of the sugars took between 50 and 60 hours to be completed. The fermented liquor was then subjected to a multi-stage distillation process to separate and purify the solvents (Walton and Martin, 1979). Part of the still residue (containing residual nutrients) was in many cases reused for the preparation of the fermentation mash, a practice known as `slopback', with positive effect on the fermentation (higher yields, less water consumption) (Rogers et al., 2006, Walton and Martin, 1979). Besides solvents, other commercially interesting products are formed during the ABE fermentation, which contribute positively to the economics of the industrial process. Fermentation gases are hydrogen and carbon dioxide, which can be separated and sold for different applications. Pure hydrogen could be used as a fuel in fuel cell applications or as reducing agent for chemical reactions. Carbon dioxide could be used, among other possibilities, for dry ice production (Spivey, 1978), or oil recovery operations. Moreover, the production of ammonia and methanol from industrial ABE-fermentation off-gases has been reported (Linden et al., 1986). The fermentation slops contain bacterial cells that are rich in riboflavin (vitamin B2), cyanocobalamin (vitamin B12) and proteins. Therefore, the slops were normally concentrated, dried and sold as cattle feed nutrient, given the right market conditions (Spivey, 1978, Linden et al., 1986). A recurrent problem in the industrial ABE process was the infection of the cultures by bacteriophages. Many different cases of infections caused by bacteriophages have been documented in many plants over the years (Rogers et al., 2006). Such an infection has a dramatic influence on the cultures, causing `sluggish' fermentations with extended fermentation times and reduced solvent yields (Jones et al., 2000). Based on archived company reports and records from the NCP owned ABE plant at Germiston (South Africa), Jones and co-workers (Jones et al., 2000) have summarized the data on infected fermentations, or where infections were suspected, by bacteriophages occurring in several years between 1943 and 1980. In this study, the phage CA1, that caused the infection

ß Woodhead Publishing Limited, 2010

428

Bioalcohol production

16.2 Commercial production plant of ABE from sugar cane feedstock operated in Brazil (Photograph by courtesy of David Jones).

in 1980, has been characterized. The host range of the bacteriophages infecting solvent producing clostridia has been shown to be very limited, with specific bacteriophages infecting only certain related species (Keis et al., 1995). Protocols for developing phage-immune strains have been reported (Walton and Martin, 1979, Hongo and Murata, 1965). However, immunized strains may become lysogenic (releasing infecting phages), which makes it necessary to search for new immune strains. It is very difficult to prevent the occurrence of this kind of infections, and therefore it is important to develop strategies to minimize the impact on the whole process. A conventional commercial ABE fermentation plant has been built recently in the state of Rio de Janeiro in Brazil (see photo in Fig. 16.2). This new plant, commissioned in 2006, is an addition to an existing sugar mill and ethanol fermentation plant. It makes use of the strains and know-how of a pre-existing ABE facility that operated in Brazil from the 1940s to early 1990s. In addition, it has a dedicated routine laboratory that is responsible for culture maintenance and for propagation and monitoring the fermentation process. The plant is operated using conventional computer-controlled technology. A flow-sheet of this process is shown in Fig. 16.3. The fermentation substrate used is clarified sugar cane juice with some supplementary nutrients being added. The substrate is sterilized by continuous heat sterilization prior to entering the fermentors. Culture biomass build-up involves standard laboratory propagation followed by two pre-fermentor stages before inoculation into the production-scale fermentor. The full-scale fermentation part consists of eight fermentors of 350 m3 volume each linked to a conventional solvent distillation facility. The energy requirements for operating the new plant have been met by pre-existing capacity available at the sugar mill. The fermentation times, yields, and solvent concentrations achieved in this process are comparable to those reported for the most advanced and successful commercial ABE batch fermentations that operated world wide during the last century. The market for the new plant is the

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

429

16.3 Schematic flow sheet from a sugarcane-based ABE plant located in Brazil.

solvent and chemical market in Brazil and at this stage possible application in the biofuels market is not being considered (David Jones, personal communication).

16.4

Advances in the production of butanol and isopropanol

16.4.1 Improvement of the butanol and isopropanol producing strains Metabolic engineering of solvent-producing clostridia is still a developing field, which has been hampered for a long time by genetic inaccessibility of the organisms. Techniques like random mutagenesis (Bowring and Morris, 1985) and transformation have been developed for solvent-producing strains (Young et al., 1989, Young, 1993), making the engineering of recombinant strains with altered product formation possible. Integration-vectors have been developed and used to specifically disrupt genes in C. acetobutylicum. In the strain C. acetobutylicum ATCC 824 inactivation of the aad gene (coding for an aldehyde/alcohol dehydrogenase AdhE; see Fig. 16.1) eliminates acetone formation and reduces butanol production by 85% (Green and Bennett, 1996). On the other hand, inactivation of the buk gene (coding for butyrate kinase 1; see Fig. 16.1) reduces butyrate production but increases butanol production by 15% (Green et al., 1996). Complementation of the aad mutant with a functional aad gene restored butanol production but not acetone production and complementation of the buk mutant with a functional buk gene restored the production of butyrate in acidogenic cultures (Green and Bennett, 1998). In

ß Woodhead Publishing Limited, 2010

430

Bioalcohol production

addition, a mutant strain with an inactivated solR gene, located on the megaplasmid, produced higher amounts of solvents compared to the wild type (Harris et al., 2001). Antisense RNA techniques have also been used successfully for the study of regulation of product formation (Desai et al., 1999). Recently, Sillers et al. (2009) reported the construction of a mutant strain with a down-regulated acetone pathway using antisense RNA, combined with over-expression of the main aldehyde alcohol dehydrogenase enzyme. This resulted in production of solvents up to 30 g/L, with ethanol being the major product for the first time (Table 16.4). Recently, two different methodologies for stable integration of foreign DNA into the clostridial genome have been developed. These techniques allow stable and selective gene-disruption and/or the insertion of an additional sequence into the genome (Heap et al., 2007, 2009, Soucaille et al., 2008, Shao et al., 2007), representing an important addition to the genetic tools already available. The first methodology is based on the use of the mobile group II intron Ll.LtrB from Lactoccocus lactis, firstly reported by Karberg et al. (2001). The insertion site of this group of introns is determined to a large extent by the sequence of a specific region of the intron. An algorithm was developed to detect potential insertion sites in a given sequence. Based on this, the sequence of the intron can be adapted to be targeted towards a desired insertion site. This system can be applied to create targeted insertions in a wide range of bacteria, both Gram-positive and Gram-negative, and has been commercialized under the name TargeTron (Sigma-Adrich (Zhong et al., 2003)). This system has been adjusted to be used in C. acetobutylicum (Heap et al., 2007, Shao et al., 2007) and further optimized to generate antibiotic marker-free mutants (Heap et al., 2009). The second method is based on homologous recombination using a negative selection marker reported for B. subtilis (Fabret et al., 2002) and adjusted for C. acetobutylicum (Soucaille et al., 2008). Using this technique, a mutant has been created in which a native hydrogenase gene has been replaced by an algal one with no antibiotic resistance gene left in the genome (von Abendroth et al., 2008). A chemically-induced mutant of the strain C. beijerinckii NCIMB 8052, strain BA101, that produces significantly increased amounts of butanol (Parekh and Blaschek, 1999, Qureshi and Blaschek, 1999, 2000a, 2001a), constitutes one of the most interesting strains developed so far (see Table 16.4). This strain consistently produced double amounts of butanol (250 mM vs 121 mM final butanol concentrations for BA101 and wild-type, respectively) and increased butanol tolerance when grown in batch cultures on glucose compared to the wild-type strain (Qureshi and Blaschek, 2001b). An economic assessment of butanol production from corn using this mutant strain resulted in a price for butanol below the price of petrochemically produced butanol at that time (Qureshi and Blaschek, 2000b). Genetic tools have also been applied to create strains with an increased butanol tolerance, since the toxicity of butanol is one of the factors limiting the

ß Woodhead Publishing Limited, 2010

Table 16.4 Overview of mutants of C. acetobutylicum (strain ATCC 824) and C. beijerinckii (strain NCIMB 8052) and their fermentation performance reported since the year 2000 Genes inactivated or repressed

ß Woodhead Publishing Limited, 2010

C. acetobutylicum ATCC 824 Wild type buk1 orf5 (solR) orf5 (solR) ctfB (asRNA) ctfB (asRNA) ctfB (asRNA)

Genes (over)expresseda

Notes

41 14 13 12 0 2 2 34 nr

85 66 97 141 26 61 30 148 nr

158 226 197 238 130 178 130 231 140

16 98 29 47 190 300 240 21 nr

23 9 17 180 248

73 75 26 75 87

± ± ± ± ±

± ± ± 92 150

6 7 3 22 44

Sillers et al. (2008) Sillers et al. (2008) Sillers et al. (2008) Sillers et al. (2008) Sillers et al. (2008)

200 L scale

nr

nr

103

171

±

200 L scale

nr

nr

95

240

22

Qureshi and Blaschek (2001b) Qureshi and Blaschek (2001b)

adhE adhE Pptb-adhE Pptb-adhE, thlA PthlA-groESL CAC1869

C. acetobutylicum M5 (ATCC 824 mutant deficient in pSOL1) M5 reference strain M5 ackA Static flask experiment M5 buk1 Static flask experiment M5 ackA Pptb-adhE Static flask experiment M5 Pptb-adhE

BA101 (chemically induced mutant) a b

References

60 51 68 85 124 85 110 80 nr

adhE

C. beijerinckii NCIMB 8052 Wild type

End product concentrations (mM)b Acetate Butyrate Acetone Butanol Ethanol

Genes are overexpressed using their native promoter site unless otherwise noted. nr, not reported; ±, not detected.

Harris et al. (2000) Harris et al. (2000) Harris (2001) Harris (2001) Tummala et al. (2003a,b) Sillers et al. (2009) Sillers et al. (2009) Tomas (2003a, b) Borden and Papoutsakis (2007)

432

Bioalcohol production

productivity of the ABE fermentation. Due to its hydrophobic chain and polar group, butanol distorts the lipidic cell membranes, causing severe cell damage. The mechanisms involved in butanol tolerance are very complex, since the membrane composition of an organism is controlled by many different, highly regulated routes. Many efforts have been made in the past to obtain strains of Clostridium with increased butanol tolerance. Both chemical mutagenesis (Hermann et al., 1985, Westhuizen et al., 1982, Allcock et al., 1981) and adaptation strategies (Lin and Blaschek, 1983, Baer et al., 1987, Soucaille et al., 1987) have been used to isolate mutants with increased butanol resistance. In some cases this resulted in slight to moderate increases in butanol production levels, while in some cases solvent production was lost (Baer et al., 1987). In C. saccharobutylicum NCP 262, the increased butanol resistance was an unexpected trait of an autolysis-deficient mutant (Westhuizen et al., 1982). In contrast, Soucaille and co-workers reported a butanol resistant mutant of C. acetobutylicum ATCC 824 which displayed increased autobacteriocin activity (Soucaille et al., 1987). Using a genetic approach, the over-expression of two heat-shock related proteins (GroES and GroEL) (Tomas et al., 2003a, b) and, in a different study, an endogenous putative transcriptional regulator (CAC1869) (Borden and Papoutsakis, 2007) in C. acetobutylicum ATCC 824, mutants with increased butanol tolerance have been constructed successfully. In the first study, the mutant strain showed an increased butanol production of 32% (17 g/L vs 13 g/L final butanol concentration) compared to the wild-type strain, while in the second case the levels of solvent production remained unchanged.

16.4.2 Production of butanol and isopropanol by species other than clostridia Recently, a number of studies and patent applications have been published concerning the production of butanol by other microorganisms to overcome some of the problems inherent to the butanol fermentation by clostridia, such as the formation of co-products (acetone, isopropanol, ethanol), degeneration of the strains and sensitivity to phage infections. Both Atsumi and co-workers (Atsumi and Liao, 2008b, Atsumi et al., 2008) and Inui and co-workers (Inui et al., 2008) describe the cloning of part of the metabolic route of C. acetobutylicum into E. coli resulting in recombinant strains producing butanol. Atsumi et al. introduced in E. coli the C. acetobutylicum genes adhE2, crt, bcd, etfA, etfB and hbd plus the E. coli gene atoB (encoding an acetyl-CoA acetyltransferase) and combined it with deletions in adhE, ldhA, frdBC, fnr and pta obtaining a strain that under semi-aerobic conditions produced 5 mM butanol when grown on mineral medium with glucose as sole carbon source. Growth on rich medium with 2% (v/v) glycerol increased the butanol production to 7 mM. Inui et al. introduced the C. acetobutylicum genes adhE, crt, bcd, etfA, etfB, hbd and thlA in E. coli

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

433

and found formation of 16 mM butanol with highly concentrated cells under anaerobic conditions in medium with glucose as carbon source (Inui et al., 2008). Patent applications in which the introduction of the C. acetobutylicum butanol pathway in alternative micro-organisms is claimed have been filed by Dupont de Nemours (WO 2007/041269) and by DSM (2008/05991). Dupont claims bacteria, yeasts and fungi to produce butanol while the DSM patent is restricted to yeasts and fungi. The examples shown in the mentioned patent applications are E. coli with a production of 0.9 mM and S. cerevisiae with a production of 0.02 mM butanol in the Dupont patent and a genetically modified S. cerevisiae with a production of 0.3 mM butanol in the DSM patent. Alternative routes for the production of alcohols (butanol, propanol, isobutanol and branched chain alcohols) have been described by Atsumi and co-workers (Atsumi and Liao, 2008b) and Shen and Liao (Shen and Liao, 2008). These alcohols are produced via keto-acid intermediates found in the native amino acid pathways of E. coli. Atsumi and co-workers reported E. coli strains able to produce 0.2 mM butanol on medium with glucose and 3.9 mM butanol when 2-ketovalerate was added to the medium. Shen and Liao describe a systematic improvement of the butanol and propanol synthesis by means of metabolic engineering of the amino acid biosynthesis and competing pathways resulting in production titers of 16.6 mM propanol and 13.5 mM butanol by a single recombinant strain. In addition, Atsumi and Liao (Atsumi and Liao, 2008a) have shown the possibility to bypass threonine biosynthesis in the 2-keto acid-based approach by introducing a citramalate pathway in E. coli. The key element in this approach is the cimA gene (encoding a citramalate synthase enzyme) from the thermophilic archeon Methanococcus jannaschii. The cimA gene is subjected to directed evolution to enhance the specific activity over a wide temperature range and to be insensitive to feed back inhibition, and then transformed into E. coli. The recombinant strains expressing this gene produced 1-propanol up to 37.8 mM plus 1-butanol up to 5.3 mM. The same research group reported the extension of the branched-chain amino acid pathways to produce non-natural longer chain keto acids and alcohols by engineering the chain elongation activity of 2-isopropylmalate synthase and altering the substrate specificity of downstream enzymes through rational protein design. When introduced into E. coli, this constructed biosynthetic pathway produces various long-chain alcohols with carbon number ranging from 5 to 8. Although the levels of new products obtained were very low (in the range of M), this study demonstrated that it is possible to engineer strains with a broader product range than the known natural products (Zhang et al., 2008). It is difficult to predict the possible toxic effect of butanol or longer chain alcohols on the new strains in view of its generic nature. Clostridia can tolerate concentrations up to 2% (v/v) of butanol in the medium, although the inhibitory effect of butanol can be already observed at lower concentrations. Fischer and coworkers (2008) present an overview of the butanol tolerance of seven mesophilic,

ß Woodhead Publishing Limited, 2010

434

Bioalcohol production

facultative anaerobic bacteria showing a great distribution in tolerance amongst the different micro-organisms. The authors also emphasize the fact that identifying the ideal strain for biofuel production is a balance between selecting the right host as a starting point and choosing which properties to change. In general the experimental tools for pathway manipulation and the metabolic models required for evolutionary strain optimization are well developed but similar toolsets for the definition of strategies for productivity and titer improvement are not available, and regulatory pathways are usually not completely understood. Metabolic engineering has also been applied to create pathways for isopropanol production in E. coli. Previously it has been shown that introduction of four genes from C. acetobutylicum (ctfA, ctfB, adc and thlA) into E. coli generated a strain capable of producing 93 mM acetone (Bermejo et al., 1998). Introduction of a secondary alcohol dehydrogenase gene (adh from C. beijerinckii) in combination with the aforementioned genes led to isopropanol synthesis. After optimization, the combination of C. acetobutylicum thlA, E. coli atoAD (acetoacetyl-CoA transferase), C. acetobutylicum adc and C. beijerinckii adh achieved the highest titre of 82 mM isopropanol with a production rate of 0.4 g/L.h (Hanai et al., 2007). Almost the same approach was followed by Jojima and co-workers (2008). These authors employed four C. acetobutylicum genes (thl, ctfA, ctfB and adc) and one C. beijerinckii gene (adh) for construction of a synthetic isopropanol pathway in E. coli. Expression of these genes under control of a tac promoter resulted in the production of 227 mM isopropanol under aerobic conditions. The engineered E. coli surpass the best reported strain of C. beijerinckii which produces isopropanol at end-concentrations of approximately 67 mM (Groot and Luyben, 1986) and also C. isopropylicum which produces isopropanol at approximately 77 mM (Matsumura et al., 1992). A major advance of the engineered E. coli strains is the lack of a butanol pathway which is a competing pathway for isopropanol production in Clostridium.

16.4.3 Advances in process technology Alternative fermentation substrates The costs of the substrates are a key factor in the economics of the ABE fermentation as in most fermentations for production of bulk chemicals. The commercial fermentations in the past and the current ones are based on starchy (corn, wheat, etc) or sugary (molasses) substrates. At present their market price is too high and the potential competition with the food market makes these substrates unsuitable as viable feedstocks to be used in the long-term for the large-scale production of butanol. The ability of saccharolytic clostridia to utilize a wide range of carbohydrate substrates, including mono- and disaccharides as glucose, xylose or cellobiose and polymers such as xylan or

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

435

Table 16.5 Substrates used for the ABE fermentation as alternative to starch-based or sugary substrates Substrates

Main carbon components

References

Non-cellulosic Apple pomace Jerusalem artichokes Low-grade potatoes Whey Soy molasses Low grade glycerol

fructose, glucose, sucrose polyfructans glucose lactose dextrose, sucrose, fructose glycerol

Voget et al. (1985) Marchal et al. (1986) Nimcevic et al. (1998) Maddox et al. (1993) Qureshi et al. (2001a) Andrade and Vasconcelos (2003)

Lignocellulosic Wood hydrolysate

glucose, mannose

Maddox and Anne (1983), Yu et al. (1984) Huang et al. (1986) Sombrutai et al. (1996) Lo¨pez-Contreras et al. (2000) Qureshi et al. (2007) Ezeji and Blaschek (2008)

Peat Palm oil effluent Domestic organic waste Wheat straw Dried distillers grains and solubles (DDGS) Corn fiber xylan

glucose, xylose oil, glucose, xylose glucose, xylose glucose, xylose glucose, xylose, arabinose, galactose glucose, xylose, arabinose

Qureshi et al. (2006)

starch (Mitchell, 1998, DuÈrre and Bahl, 1996), stimulated a search for alternative cheaper substrates since the early years of the ABE process. Although some strains produce cellulolytic enzymes (Zappe et al., 1988, LoÂpez-Contreras et al., 2004), none of the known solventogenic clostridia is able to degrade cellulose, and therefore, addition of external cellulases is needed to degrade cellulosecontaining substrates for the fermentation. A selection of cellulosic and noncellulosic substrates that have been tested over the years as possible alternatives with different solvent-producing strains is listed in Table 16.5. The tubers from the Jerusalem artichoke plant (Helianthus tuberosus, also called sunchoke) have been studied as a potential substrate for the ABE fermentation (Marchal et al., 1985, 1986) and also for 2,3-butanediol production (see Table 16.7). In the tubers, the storage polysaccharide is inulin, a polymer typically consisting of approximately 35 fructose units with a terminal glucose unit, instead of starch as in the case of potatoes. Marchal and co-workers used Jerusalem artichokes juice, supplemented only with buffering agents, performing simultaneous saccharification and fermentation (SSF) of the inuline polymers by externally added enzymes and C. acetobutylicum. Solvent yields of 33.9% (w/w) with respect to sugars consumed and final solvent concentrations of 23±24 g/L were achieved (Marchal et al., 1985). Another interesting non-cellulosic substrate is whey permeate. This waste stream from the dairy industry contains mostly lactose and approximately 1%

ß Woodhead Publishing Limited, 2010

436

Bioalcohol production

(w/w) of protein. Fermentation of whey by clostridia species resulted in altered end product ratios towards butanol (i.e. 10:1, ratio B:A) compared to those found on starch or molasses (2:1, ratio B:A) (Maddox et al., 1993). Economical evaluations dating from the 1980s for the production of butanol and acetone (Linden et al., 1986) and butanol and isopropanol (Schoutens and Groot, 1985) showed that these processes were not, or only marginally viable, and that the transportation and/or concentration costs of whey are critical parameters negatively affecting the process. Low grade potatoes have been tested as substrate for fermentation by C. beijerinckii with good yields and productivities (Nimcevic et al., 1998). This fermentation has been upscaled in a pilot facility in Austria during the late 1990s by Gapes and co-workers (Nimcevic and Gapes, 2000, Gapes, 2000), operating in a continuous mode using partially liquefied substrates. In this facility, different techniques for (on-line) product removal (including gas-stripping, vacuum distillation and reverse osmosis) were tested resulting in promising results for a full-scale process. Lignocellulosic substrates, defined as those derived from plant material and in particular agricultural wastes, composed mainly of lignin and carbohydrate polymers (cellulose and hemicellulose), are considered the substrates with the greatest potential for the ABE fermentation due to their wide availability, sugar composition and low price. Many studies have shown that lignocellulosic materials from a variety of biomass sources hydrolysed with acid, alkali and/or enzymes are potential feedstocks for ABE fermentation (Gapes et al., 1983, Jones and Woods, 1986, Claassen et al., 1998, Ezeji and Blaschek, 2008, Qureshi et al., 2007). Hemicellulosic hydrolysates from lignocellulosic waste materials (hemp waste, corn cobs and sunflower shells) prepared by diluted acid treatment were used successfully in Russia for the partial replacement of grains as substrates in industrial ABE plants (see Table 16.3), constituting the only report on the use of these kind of substrates at large scale (Zverlov et al., 2006). In most cases, the lignocellulosic material has to be pretreated in order to make it more accessible to chemical or enzymatic hydrolysis. The most common pretreatment is steam-explosion (Focher et al., 1991, Gapes et al., 1983), but extrusion has also been used (LoÂpez-Contreras et al., 2000, Claassen et al., 1998). The use of wheat straw for the production of acetone, butanol and ethanol has been described recently (Qureshi et al., 2007). In this report, wheat straw was ground to fine particles, re-suspended in a solution of 1% (v/v) sulfuric acid and autoclaved. The pH of the resulting slurry was adjusted to 5.0 using a concentrated NaOH solution, and commercial cellulases, xylanases and -glucosidases were added to hydrolyse the sugar polymers to soluble sugars. After hydrolysis, the hydrolysates were filtered and sterilized to be used for fermentation. The concentration of total sugars varied between 50 and 60 g/L, being glucose, xylose and mannose the most abundant (in this order). Using C. beijerinckii

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

437

NCP260 in a batch reactor with wheat straw hydrolysate (WSH) as substrate and supplemented with nutrients, yields of 0.4 g ABE/g sugar consumed, and end concentrations of acetone and butanol of 11.9 g/L and 12 g/L, respectively, were obtained, very similar to the results on synthetic medium with glucose as substrate. Interestingly, the WSH was fermentable without detoxification steps, in contrast to other lignocellulosic hydrolysates, where inhibiting concentrations of toxic components (furfurals, phenolic compounds or organic acids) were present and needed to be detoxified prior to the fermentation (Claassen et al., 2000, Ezeji and Blaschek, 2008, Maddox and Anne, 1983, Walton and Martin, 1979). The same bacterial strain was used for the fermentation of another potentially interesting substrate, corn fiber xylan (CFX), which is a by-product from the wet-milling of corn (Qureshi et al., 2006). In an integrated batch process with `in-situ' substrate degradation and ABE removal, a total of 24.7 g ABE/L were produced with yields of 0.44 g ABE/g sugar fermented. For combining efficient simultaneous fermentation and saccharification (SSF), CFX-based medium needed supplementation with xylose, nutrients and xylanase enzymes. Other interesting lignocellulose-containing material tested as substrate for the ABE fermentation are dried distillers grains and solubles (DDGS). DDGS, derived as a by-product from a corn-based ethanol facility in Iowa (USA) were pretreated using three different methods (diluted acid, hot water and ammonium fiber expansion) and enzymatically hydrolysed. After detoxification by overliming in the case of the acid-pretreated material, the hydrolysates were fermented to acetone, butanol, and ethanol by five different strains. Best results were obtained with the hot-water pretreated DDGS, with end concentrations of total solvents between 10 and 13 g/L for all strains tested. Good yields of solvents, above 0.30 g ABE/g sugar consumed, were obtained for all sugars present in the hydrolysates, with total sugar consumption, showing the versatility of solvent-producing strains (Ezeji and Blaschek, 2008). To enable direct utilization of lignocellulosic biomass as substrate, application of simultaneous saccharification and fermentation systems has been studied. On the one hand, the use of co-cultures of solventogenic clostridia with true cellulolytic organisms has the advantage of eliminating the pre-hydrolysis step. The direct conversion of cellulose to solvents by co-cultures of C. acetobutylicum with C. cellulolyticum or C. thermocellum has been shown. In both cases, the production of solvents was very poor, possibly due to the fact that the concentrations of sugars and butyric acid in the medium were too low to induce the solventogenic phase. To increase the level of butyric acid in the medium, co-cultures of C. acetobutylium and C. beijerinckii with butyric acid producing strains of C. butyricum and C. pasteurianum have been tested, but the amount of solvents produced was not higher than in the mono-cultures (Jones and Woods, 1986). On the other hand, supplementing the culture medium with cellulases is another alternative to pre-hydrolysis of the substrate. Strains of C.

ß Woodhead Publishing Limited, 2010

438

Bioalcohol production

acetobutylicum and C. beijerinckii are able to grow on and produce solvents from crystalline cellulose (LoÂpez-Contreras, 2003) or from wheat straw (Soni et al., 1982) when a cellulase preparation from the fungus Trichoderma reesei is present in the medium. These results may represent an economic improvement with respect to separate hydrolysis and fermentation of these substrates. Advanced fermentation and down stream processing of products In the traditional ABE fermentation, the volatile products are recovered from the fermentation broth by distillation followed by fractionation. The operation costs of the distillation and fractionation units needed in a full-scale facility are relatively high. This is the most energy-demanding step of the whole process, due to the low concentrations of the solvents in the broth (approximately 2% (v/v)) and the mixed products to be separated. Therefore, since the 1970s, there has been a search for alternative, more energy efficient recovery techniques, summarized by a number of authors (Ennis et al., 1986, Groot et al., 1989, 1992, Santangelo et al., 1998, DuÈrre, 1998). Recently, Oudshoorn and co-workers have made an evaluation of the most promising separation techniques for 1-butanol from aqueus solutions (membrane-based systems, such as reverse osmosis, perstraction and pervaporation, as well as liquid/liquid extraction, adsorption and gas-stripping) to be applied to the ABE process (Oudshoorn et al., 2009). Currently, there is an interest in the use of `in-situ' product recovery methods (for removal of solvents during the fermentation) because these offer two main advantages compared to conventional down-stream techniques: (i) these methods could be suitable for continuous and selective toxic product recovery (especially butanol). Production is limited by product inhibition, which limits the extent of substrate utilization. In-situ removal offers the potential of using concentrated feed solutions, resulting in an important reduction of process streams (including waste water stream), and (ii) reduction of the costs of product recovery if the separation technique is competitive with distillation. The methods where the most developments have been reported towards an integrated fermentation process in the recent years are pervaporation and gas-stripping, and are described below. In Table 16.6, fermentation parameters of different integrated processes are shown. Pervaporation Pervaporation is a process in which a liquid (or gas) stream containing two or more miscible components is placed in contact with one side of a non-porous polymeric membrane or a molecularly porous inorganic membrane (such as a zeolite membrane) while a vacuum or gas purge is applied to the other side. The components in the liquid stream sorb into/onto the membrane, then permeate through the membrane and evaporate into the vapour phase. The resulting

ß Woodhead Publishing Limited, 2010

Table 16.6 Performance of integrated fermentation processes for production of solvents Type of process

Recovery technique

Batch ß Woodhead Publishing Limited, 2010

Substrate

Yield g ABE/ g sugar

Total solvent Total solvents productivity produced g/L.h g/L in broth

Glucose

0.42

0.34

24.2

60 g/L of glucose consumed

Qureshi and Blaschek (1999)

0.69

51.5

121 g/L of glucose consumed

Qureshi and Blaschek (1999)

119

342 g/L of glucose consumed

Qureshi et al. (2001b)

Batch

Pervaporation

Glucose

0.42

Fed-batch

Pervaporation

Glucose

0.35

Fed-batch

Pervaporation

Glucose

0.43

0.98

Continuous with immobilized cells

Pervaporation (polypropylene membrane)

Whey permeate medium

0.39

3.5

Batch

Gas-stripping

Wheat straw hydrolysate

0.37

Fed-batch

Gas-stripping

Glucose

0.39

Continuous

Gas-stripping

Glucose

0.40

* Solvents are isopropanol, butanol and ethanol.

Comments

165.1

References

Groot et al. (1984)* Lactose conc. of 47 g/L in feed. D ˆ 0:59 hÿ1

Friedl et al. (1991)

47.6

128 g/L total sugars consumed

Qureshi et al. (2007)

1.16

40±172

500 g/L of glucose consumed

Ezeji et al. (2004a)

0.91

460

1163 g/L of glucose consumed

Ezeji et al. (2004b)

440

Bioalcohol production

vapour is then condensed. Physicochemical interactions between the diffusing species and the membrane material determine the ratio of the diffusing species (Vane, 2005, Ennis et al., 1986). This technique has been successfully applied to the separation of ABE from aqueous systems using different conditions. Several membrane types have been studied (silicone, PDMS, polypropylene, or liquid membranes (oleyl alcohol and propylene), poly(tetraflouroethylene)). Best results in terms of selectivity have been obtained using silicone-based membranes (Qureshi and Blaschek, 1999, Huang and Meagher, 2001). Integration of pervaporation with continuous fermentation resulted in higher productivities compared to batch processes; Silicone/silicalite and polypropylene membranes have been used (Groot et al., 1991, 1992, Friedl et al., 1991, Huang and Meagher, 2001). However, some practical problems, such as fouling of the pumps were encountered. Batch fermentations on media with high substrate concentration with in-situ pervaporation using a silicone membrane produced 51 g/L of solvents, but utilized only 80% of the substrate due to nutrient limitation (Qureshi and Blaschek, 1999) (Table 16.6). Despite the relatively good results obtained in laboratory experiments, the application of pervaporation in full-scale fermentation processes is still hampered by the high costs of the membranes and modules needed and by operational problems such as stability or fouling. Gas-stripping The removal of volatile fermentation products from fermentation broth by gasstripping is a relatively simple process that, similar to pervaporation, does not harm the culture and can be operated in a continuous mode (Qureshi and Blaschek, 2001c). In this process, gas (anoxic nitrogen or fermentation gas) is sparged through the medium during fermentation, capturing volatile products (butanol, isopropanol, acetone and ethanol). The gas containing the products is led to a cooling unit where products condense resulting in concentrated solvent solutions. In Table 16.6, some examples of the application of gas-stripping in ABE fermentation processes are given. In all cases, gas-stripping has a very positive effect on the fermentation, since the toxic volatile products are removed selectively, eliminating the toxicity effects of solvents and resulting in enhanced sugar utilization of concentrated sugar substrates, higher biomass concentration in the fermentor and higher productivities. Before this technique can be applied at large scale, important aspects need to be solved, including: (i) excessive foaming of the broth, due to the combination of proteins and other components with the gas; (ii) need for large volumes of fermentors to allow gas-stripping; (iii) high costs of the equipment needed (i.e. compressors) for an industrial facility; and (iv) high energy costs for the separation of the volatiles from the gas-stream.

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

16.5

441

Methods for biological production of 2,3-butanediol

16.5.1 Microorganisms and metabolic pathways Several 2,3-butanediol-producing bacterial species have been described, belonging to the genera Bacillus, Enterobacter, Klebsiella, Pseudomonas, and Serratia. In literature, the biological production of all three different stereomers of 2,3-butanediol has been reported (Syu, 2001, Voloch et al., 1985). The synthesis route towards 2,3-butanediol proceeds via the central intermediate acetolactate, involved in the branched-chain amino acid (isoleucine, leucine and valine) biosynthesis pathway. Acetolactate is formed by coupling two pyruvate molecules with the concomitant formation of one carbon dioxide molecule, catalyzed by acetolactate synthase. Subsequently, acetolactate is decarboxylated by acetolactate decarboxylase to yield acetoin (Wardwell et al., 2001, Xiao and Xu, 2007). Alternatively, acetoin can also be produced from pyruvate and acetaldehyde by pyruvate decarboxylase (Crout et al., 1986). Reduction of acetoin by acetoin reductase results in 2,3-butanediol production (Fig. 16.4).

16.4 Different metabolic pathways for the production of stereoisomers of 2,3butanediol from pyruvic acid. Scheme adapted from Taylor and Juni (1960) and Ui et al. (1986).

ß Woodhead Publishing Limited, 2010

442

16.6

Bioalcohol production

Advances in the production of 2,3-butanediol

16.6.1 Improvement of 2,3-butanediol producing strains Currently, two bacterial species, Klebsiella pneumoniae and Bacillus polymyxa, have demonstrated their potential for 2,3-butanediol production on a commercial scale. K. pneumoniae is known to produce mainly meso-2,3butanediol, while B. polymyxa produces mainly D(-)-2,3-butanediol (Garg and Jain, 1995, Qin, 2006). K. pneumoniae has several advantages over B. polymyxa as 2,3-butanediol producer. It grows rapidly, metabolizes a wider range of simple sugars and produces higher amounts of 2,3-butanediol. Furthermore, all of the major sugars present in hemicellulose and cellulose hydrolysates are well suited as substrates for this bacterial species (Garg and Jain, 1995). The highest 2,3-butanediol levels reported so far were obtained with K pneumoniae and reached values as high as 150 g 2,3-butanediol/L and a productivity of 4.21 g/Lh (see Table 16.7). These fed-batch fermentation experiments were conducted in a 5-L bioreactor with 3-L initial medium volume and initial glucose concentration of 70 g/L. An optimal residual glucose concentration of 20±30 g/L was maintained during the fermentation process (Ma et al., 2009, Syu, 2001).

16.6.2 Process technology: fermentation parameters and product separation Several parameters have been reported to have a tremendous effect on the production of 2,3-butanediol. Culturing factors such as pH, temperature, aeration, agitation, and inoculum size are important aspects influencing the 2,3butanediol production process, as well as factors like initial substrate concentration, product concentration, medium supplements and the water activity of the medium. In general, the pH optimum for 2,3-butanediol production lies normally in the acidic range (approximately pH 6). The temperature optimum for growth and 2,3-butanediol production is reported to be between 30 and 35 ëC. Furthermore, aeration of the system results in higher 2,3-butanediol efficiencies (Garg and Jain, 1995). The oxygen transfer rate is noted as the most important operating factor for 2,3-butanediol production. Lowering the oxygen supply rate increases the 2,3-butanediol yield, although it also decreases the cell density. Therefore, an optimal balance is needed to create optimal conditions for 2,3-butanediol production (Syu, 2001). Agitation is also beneficial, since it permits, besides improved oxygen-transfer rates, a better availability of fresh substrates and prevents high local concentrations of products. Increasing the size of inoculum of K. pneumoniae and B. polymyxa did not improve 2,3-butanediol production (Yu and Saddler, 1982, Laube et al., 1984). In contrast, the acclimatization of pre-cultures to new environmental conditions improved the production tremen-

ß Woodhead Publishing Limited, 2010

Table 16.7 Comparison of improved fermentation systems reported for the production of 2,3-butanediol Organism

ß Woodhead Publishing Limited, 2010

Fermentation type

Substrate

Fermentation time (h)

Concentration g 2,3-BD/L

Yield Productivity g 2,3-BD/ (g/L/h) g substrate

Recovery technique

Comments

Klebsiella pneumoniae CICC 10011 Klebsiella pneumoniae CICC 10011

Fed-batch

40

84.0

0.28*

Fed-batch

Jerusalem artichoke powder Glucose

60

63.8

Klebsiella pneumoniae G31 Klebsiella pneumoniae SDM

Fed-batch

Glycerol

300

Fed-batch

Glucose

Klebsiella oxytoca ME-UD-3 Klebsiella pneumoniae PTCC 1290 Klebsiella pneumoniae CICC 10011 Klebsiella oxyroca ATCC 8724

Batch

Reference

2.10

±

SSF

Sun et al. (2009b)

0.44*

1.06*

Aqueous two-phase extraction

Based on 2-propanol/ (NH4)2SO4 system

Sun et al. (2009a)

49.2

0.36*

0.16*

±

38

150

0.44*

4.21

±

Glucose

56

95.5

0.48

1.71

±

Batch

Glucose

35

23.01

NR

0.66

Liquid-liquid extraction

Fed-batch

Glucose

50

92.4

0.43*

1.84*

±

Fed-batch

Glucose

NR

NR

0.47

1.40

Vacuum membrane distillation

Petrov and Petrova (2009) Residual glucose concentration 20±30 g/l Two-stage agitation

Ma et al. (2009)

Extraction with oleyl alcohol

Anvari and Khayati (2009)

Ji et al. (2009)

Qin (2006) Final conc. 2,3-BD is 430 g/L

Qureshi (1994)

* Calculated from reported data; NR = not reported; SSF, simultaneous saccharification and fermentation; 2,3-BD, 2,3-butanediol.

444

Bioalcohol production

dously specially when concentrated wood hydrolysates and agricultural residues have been used as substrates (Grover, 1990, Yu and Saddler, 1983). Carbohydrate concentrations used in most studies, using particularly wood hydrolysates or various types of molasses as substrates, vary between 5 and 10% (w/v). More concentrated substrates (with higher sugar concentrations) contained inhibiting concentrations of toxic substances like furfurals and phenolic compounds. Significant inhibition of 2,3-butanediol formation and sugar utilization was observed when substrate levels of only xylose, glucose, arabinose, galactose, mannose and cellobiose exceeded 50 g/L. It has been reported that 2,3-butanediol is less toxic than ethanol, acetone or butanol to several microbial species (Yu and Saddler, 1982). As an example, growth of K. oxytoca is sustained at 2,3-butanediol concentrations up to 105 g/L, and 2,3-butanediol production continues up to concentrations of this product up to 130 g/L. Therefore it was concluded that 2,3-butanediol at high concentrations may inhibit bacterial growth, but has little effect on its biosynthesis (Fond et al., 1985). Another important factor is the composition of the culture medium. It must contain all necessary components for optimal growth and 2,3-butanediol production. A cheap and adequate nitrogen source in 2,3-butanediol production is urea, instead of yeast-extract. Urea has been added to substrates like hydrolysed wheat mashes and wood hydrolysates. Generally, K. pneumoniae has low nutritional requirements and produces already satisfactory amounts of 2,3-butanediol in media containing inorganic salts and sugar. Addition of extra peptone/beef-, wheat-, malt- or yeast extract are proven to enhance the production of 2,3butanediol. Next to these extracts, extra phosphate and the trace metals Fe2+, and Mn2+ were found to significantly improve 2,3-butanediol yields. The water activity of the medium is another important factor, related to the osmotic pressure experienced by the microorganism. Increasing the concentration of certain solutes, such as starch or sugars, in the medium results in a decrease of the water activity. K. pneumoniae is known to possess a relatively weak osmotolerance and therefore, water activity may have a great influence in its industrial process environment. This effect may partially explain why the butanediol process is more difficult with natural, complex sources of carbohydrates (e.g., starch-containg substrates), than with simple sugars as substrate (Garg and Jain, 1995). An important aspect in the 2,3-butanediol fermentation is the recovery from the fermentation broth. Major difficulties in this recovery are due to the fact that 2,3-butanediol has a high boiling point and high affinity for water, but also the presence of dissolved and solid components of fermentation mashes interferes with the recovery (Garg and Jain, 1995). In microbial production systems, the costs of the separation of 2,3-butanediol from the fermentation broth have been estimated to account for more than 50% of the total expenses. Most studies regarding downstream processing of 2,3butanediol are focused on steam stripping, reverse osmosis, pervaporation, and

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

445

solvent extraction (Xiu and Zeng, 2008). In Table 16.7, a summary of recent results described for 2,3-butanediol processes at laboratory scale using innovative separation techniques is shown. Pervaporation, explained in Section 16.4.2, has also been applied for 2,3butanediol separation with promising results (Syu, 2001). Solvent extraction may also become an effective method for 2,3-butanediol recovery. Many solvents, like ethyl acetate, diethyl ether, and n-butanol are suitable for this extraction system. Countercurrent steam stripping and reversed osmosis are also regarded as feasible techniques (Syu, 2001). However, these methods are still difficult and deficient, and more research is needed to improve their yield, purity and reduce their energy consumption (Xiu and Zeng, 2008).

16.7

Other long chain alcohols that can be produced biologically

16.7.1 1,2-propanediol and 1,3-propanediol The three-carbon diols 1,3-propanediol and 1,2-propanediol, which are of significant commercial interest, can both be biologically produced. Since C3 alcohols are beyond the scope of this chapter they will be described only briefly. A recent review covering research on these compounds can be found in the book The Prokaryotes (Rogers et al., 2006). 1,2-Propanediol can be produced naturally by E. coli and several other microorganisms from 6-deoxyhexoses like L-rhamnose and L-fucose (Bennett and San, 2001). Production of 1,2-propanediol at 70 ëC was recently reported for the extreme thermophile Caldicellulosiruptor saccharolyticus (Werken et al., 2008). The metabolic pathway followed for the production of 1,2-propanediol involves the formation of L-lactaldehyde and dihydroxyacetone phosphate, and the subsequent reduction of L-lactaldehyde to (S)-1,2-propanediol (Sawada and Takagi, 1964). The production via this pathway is currently considered not commercially feasible due to the low availability of the deoxysugars needed as substrate. Production from common sugars like glucose and xylose has been described for Clostridium sphenoides (Tran-Din and Gottschalk, 1985) and Thermoanaerobium thermosaccharolyticum (Altaras and Cameron, 2001). For these bacteria it was found that the R- or the S-enantiomer is produced, and that a different pathway is used, involving methylglyoxal and either D-lactaldehyde or acetol. Various articles and patents have been issued that describe an improved production of 1,2-propanediol by recombinant E. coli, Klebsiella or yeast (Altaras and Cameron, 1999). The current industrial production of 1,2propanediol occurs chemically by hydration of propylene oxide, in a process that requires and produces several toxic compounds. Moreover, the chemical process yields a racemic mixture. Despite these disadvantages, the chemical process is still the predominant method of production for this chemical.

ß Woodhead Publishing Limited, 2010

446

Bioalcohol production

1,3-Propanediol is produced by a limited group of bacteria, including enterobacteria, lactobacilli, and clostridia (SchuÈtz and Radler, 1984, Homann et al., 1990, Biebl et al., 1999). In all cases glycerol is the starting substrate, and 1,3-propanediol is produced involving a vitamin B12-dependent dehydration step. The only exception reported so far is the production by Clostridium butyricum, that uses a B12-independent pathway (Raynaud et al., 2003). This pathway has been characterized and could be engineered into C. acetobutylicum, yielding an almost 2-fold improved production in titer and productivity of 1,3propanediol compared to the natural producer C. butyricum (GonzaÂlez-Pajuelo et al., 2005, 2006). In recent years, research on the microbial production of 1,3propanediol has considerably expanded as this diol can be used for the production of a bio-based polyester. Biobased polyester production is feasible when the production of 1,3-propanediol can become more cost-effective, which means that a cheaper substrate is used, such as glucose instead of glycerol. Many efforts are being made to combine the pathway from glucose to glycerol with the pathway from glycerol to 1,3-propanediol using metabolic engineering (Biebl et al., 1999).

16.7.2 2-Butanol Another long-chain alcohol with the potential to be produced biologically is 2butanol. However, whereas butanol is a common metabolic end product of various bacteria species, 2-butanol is not. This makes the achievement of high productivities a real challenge. However, based on the octanol-water partition coefficient (log Kow), 2-butanol is expected to be less toxic to the producing microorganism compared to butanol (0.61 and 0.88, respectively), and thus higher levels might be reached during fermentative production of 2-butanol (Schultz et al., 2002, Weber and de Bont, 1996). Currently, 2-butanol is produced on a large scale from petroleum and used mainly for the production of methyl-ethyl ketone, which is a solvent for cleaning agents and paint removers. Volatile esters of 2-butanol are used in perfumes and artificial flavours. It is a chiral compound, and thus it exists as two enantiomers. As 2-butanol is an isomer of butanol it is expected to have similar fuel properties. Because 2-butanol is an important industrial chemical attempts have been undertaken to construct a biosynthetic pathway for its production using genetic engineering. Donaldson and co-workers (2007) describe a recombinant microbial production host that expresses a 2-butanol biosynthetic pathway. The pathways involve conversion of pyruvate via -acetolactate, acetoin, 3-amino-2butanol, 3-amino-2-butanol phosphate, and 2-butanone to 2-butanol. Alternatively, 2-butanol can be produced via acetoin, meso-2,3-butanediol and 2-butanone, as was described for the lactobacilli (Speranza et al., 1997). From the patent literature, enzymatic in vitro 2-butanol production is also known.

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

447

Reduction of ketones to 2-butanol by secondary alcohol dehydrogenases derived from Rhodococcus rubur (Stampfer et al., 2003) or from Candida parapsilosis (Kojima et al., 1997) has been reported.

16.8

Future trends

The large-scale butanol fermentation plants that are currently in operation, or planned for the near future, use sugary substrates (sugar cane), grains (wheat, barley) or corn as substrates and distillation for the separation of butanol. If biologically produced butanol (biobutanol) needs to compete with other biofuels or petrochemicals, improvements in the existing process are needed with respect to substrate use (replacement by lignocellulosic substrates, no competition with food industry), product yields and titers (increase of butanol tolerance) and separation technologies (distillation is a high energy demanding process). As shown in Section 16.4, important advances in different aspects of the fermentation process have been achieved, including innovative set-ups for integrated fermentation and product removal, which are expected to contribute to the development of a more efficient process. It should be noted that, in the particular case of the butanol fermentation, improvements with respect to strain performance (resistant to higher butanol concentrations, higher product yields or with no production of co-products), will have a dramatic effect on the downstream costs of the products. Therefore, much effort is currently devoted to the metabolic engineering of the solvent-producing species and to the development of tailor-made strains of other species (see Sections 16.4.1 and 16.4.2). Data and expertise on the large-scale operation of improved processes using lignocellulosic substrates (Zverlov et al., 2006) and continuous fermentations (Ni and Sun, 2009) are available. These and the existence of current operating ABE production plants would allow for the rapid application of new developments (e.g., improved strains, integrated fermentations) into an economical ABE or IBE process within a reasonable period of time. The current interest in biobutanol is reflected by the establishment in the last years of a number of start-up companies focused on research and development of the butanol fermentation process. Examples in Europe are Green Biologics (www.greenbiologics.com) established in the United Kingdom, and the Swiss/ German Butalco (www.butalco.com), the latter focusing on butanol production from yeasts. In the USA, Cobalt Biofuels (www.cobaltbiofuels.com), Tetravitae Bioscience (www.tetravitae.com), ButylFuel, LLC (www.butanol.com) and Baer (baer-enterprises.com) are to be mentioned. The partnership of the multinationals BP and Dupont to develop and commercialize biobutanol has been mentioned earlier in this chapter. The 2,3-butanediol fermentation process is further away from commercial implementation than the process for butanol. One of the main causes for this is the low price of the petrochemically produced 2,3-butanediol and its inter-

ß Woodhead Publishing Limited, 2010

448

Bioalcohol production

mediates (Rogers et al., 2006). Although many advances are being made in this process, as described in Section 16.6, the most economically feasible applications for biologically produced 2,3-butanediol are more likely to be linked to its use in high value products (in pharmacy, personal care or chemistry) rather than as a bulk chemical or a biofuel (Rogers et al., 2006). The future of all biobased industrial processes to replace conventional petrochemical production processes for bulk chemicals relies for a major part on the stable availability of cheap substrates, and efficient technologies to treat them (Wilke and Vorlop, 2004). Besides technological factors, the market situation (e.g., the price of fossil oils, volume and demand of the products) and the political climate (e.g., governmental policies to stimulate environmentally friendly initiatives, implementation of taxes on polluting systems) are crucial factors for the implementation of a biobased industry. In a recent study, different scenarios for the future market potentials for biobased chemicals (including butanol) have been described (Dornburg et al., 2008). Here, Dornburg and coworkers used a variety of data on several current chemical processes with the potential to be replaced by white biotechnology (using fermentation or enzymatic processes) to make a prediction on the economical and energetical benefits that could be obtained in Europe in 2050. Assumptions were made concerning the following parameters; fossil fuel prices (varying from US$30/ barrel to US$83/barrel), technological developments (from current state of the art to implemented new technologies), costs of feedstocks for the biological processes (from ¨400/tonne sugar to ¨70/tonne sugar), growth of the chemical market (no growth and up to 3% tonnage increase per year) and the support to the biobased chemical industry in the form of subsidies (no subsidy or subsidies from 1 to 5% of product value). Using these parameters, three scenarios, low, medium and high scenario, corresponding to bad, medium and good conditions for biobased chemicals, respectively, were defined. Data on the different parameters for all chemicals studied, under the conditions of the three mentioned scenarios were modelled to give a prediction on the market potential and the economical and energetical benefits of the biobased chemicals during the years 2000±2050. It is shown that, under the high scenario, a total energy benefit of 18±32% and an economical saving of ¨74.8 billion compared to the traditional chemical processes could be achieved by 2050 in Europe. These predictions are encouraging and show the great potential of white biotechnology (including the butanol fermentation process) to contribute to a sustainable biobased economy.

16.9

Sources of further information and advice

Internet resources on solvent-producing clostridia · www.clostridia.net Website hosted by the group of Prof. Nigel Minton at the University of Nottingham (UK). Information on pathogenic and solvent-

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

449

producing clostridia, including references to European projects and related activities can be found here. · www.biobutanol.nl Website of the project `EOS-LT: Biobutanol' by Dr. Ana M. LoÂpez Contreras from the Biconversion group of the Agrotechnology and Food Sciences Group from the Wageningen UR (The Netherlands). The project `EOS: Biobutanol', financed by the Dutch Ministry of Agriculture, Nature and Fisheries through SenterNovem, has a goal to develop a biobutanol process based on wheat straw and grass using an integrated fermentation system. Recent books and book chapters related to this subject · Handbook on Clostridia (DuÈrre, 2005). This book presents information on both clostridia of biotechnological importance, with a focus on the solventproducing clostridia, and clostridia of medical importance. The fields of methods, physiology, medical significance, regulation, ecosystems, genomics, and current and potential applications of the most important clostridial species are described and discussed in the different sections of this book. · The Prokaryotes. Volume 1: Symbiotic Associations, Biotechnology, Applied Microbiology (Dworkin et al., 2006). In volume 1 of the third edition of the series The Prokaryotes, a recognized reference in the field of microbiology, section 3 deals with biotechnology and applied microbiology. Within this section, in subsection 3.1 Rogers and co-workers (2006) review the status and advances of the existing microbial processes for the production of organic acids and solvent production, including interesting historical data. · Clostridia ± Molecular Biology in the Post-genomic Era (BruÈggemann and Gottschalk, 2009). The focus of this book is primarily on advances in the molecular biology of pathogenic clostridia. However, Chapter 10 (entitled `Development of genetic knock-out systems for clostridia') gives a detailed overview of the recently developed genetic tools for gene inactivation in, among others, solventogenic clostridia, and Chapter 12 (entitled `Metabolic networks in Clostridium acetobutylicum: Interaction of sporulation, solventogenesis and toxin formation') describes the latest insights in the metabolic networks involved in solvent formation and the role of the regulator Spo0A in the metabolism of C. acetobutylicum.

16.10 Acknowledgements Work in the research group of authors Ana M. LoÂpez-Contreras, Jan Springer and Pieternel A. M. Claassen was supported by the Dutch Ministry of Economic Affairs through the program EOS-LT (Subsidy for Research on Energy, www.senternovem.nl/eos). Support by the Bio-Based Sustainable Industrial Chemistry (B-BASIC, www.b-basic.nl) program of NWO-ACTS in The Nether-

ß Woodhead Publishing Limited, 2010

450

Bioalcohol production

lands is gratefully acknowledged by all authors. Ms. Hetty van der Wal and Prof. David Jones are gratefully acknowledged for their contribution in Fig. 16.3.

16.11 References Allcock, E. R., Reid, S. J., Jones, D. T. and Woods, D. R. (1981) `Autolytic activity of an autolysin-deficient mutant of Clostridium acetobutylicum', Appl Environ Microbiol, 42, 929±935. Altaras, N. E. and Cameron, D. C. (1999) `Metabolic engineering of a 1,2-propanediol pathway in Escherichia coli', Appl. Environ. Microbiol., 65, 1180±1185. Altaras, N. E. and Cameron, D. C. (2001) `Conversion of sugars to 1,2-propanediol by Thermoanaerobacterium thermosaccharolyticum HG-8', Biotechnol Prog, 17, 52±56. Andrade, J. C. and Vasconcelos, I. (2003) `Continuous cultures of Clostridium acetobutylicum: culture stability and low-grade glycerol utilization', Biotechnol Lett, 25, 121±125. Andreesen, J. R., Bahl, H. and Gottschalk, G. (1989) `Introduction to the physiology and biochemistry of the genus Clostridium'. In Biotechnology Handbooks, Vol. 3 (Eds, Minton, N. P. and Clark, D. J.), Plenum Press, New York. Anonymous (2006) `BP-DuPont biofuels fact sheet', Vol. 2009 BP-Dupont. Anvari, M. and Khayati, G. (2009) `In situ recovery of 2,3-butanediol from fermentation by liquid-liquid extraction', J Ind Microbiol Biotechnol, 36 (2), 313±317. Atsumi, S., Cann, A. F., Connor, M. R., Shen, C. R., Smith, K. M., Brynildsen, M. P., Chou, K. J. Y., Hanai, T. and Liao, J. C. (2008) `Metabolic engineering of Escherichia coli for 1-butanol production', Metabol Eng, 10, 305±311. Atsumi, S. and Liao, J. C. (2008a) `Directed evolution of Methanococcus jannaschii citramalate synthase for biosynthesis of 1-propanol and 1-butanol by Escherichia coli', Appl Environ Microbiol, 74, 7802±7808. Atsumi, S. and Liao, J. C. (2008b) `Metabolic engineering for advanced biofuels production from Escherichia coli', Curr Opin Biotechnol, 19 (5), 414±419. Baer, S. H., Blaschek, H. P. and Smith, T. L. (1987) `Effect of butanol challenge and temperature on lipid composition and membrane fluidity of butanol-tolerant Clostridium acetobutylicum', Appl Environ Microbiol, 53, 2854±2861. Bahl, H., Gottwald, M., Kuhn, A., Rale, V., Andersch, W. and Gottschalk, G. (1986) `Nutritional factors affecting the ratio of solvents produced by Clostridium acetobutylicum', Appl Environ Microbiol, 52 (1), 169±172. Beesch, S. C. (1952) `Acetone-butanol fermentation of sugars', Ind Eng Chem, 44, 1677± 1682. Beesch, S. C. (1953) `Acetone-butanol fermentation of starches' Appl Microbiol, 1, 85± 95. È ber die ButylalkoholgaÈrung und das butyl Ferment', Beijerinck, M. W. (1893) `U Verhandel. Akad. Wetenschappen Amsterdam, Afdeel. Natuurkunde, Sectie II, Deel I, 3±51. Bennett, G. N. and San, K. Y. (2001) `Microbial formation, biotechnological production and applications of 1,2-propanediol', Appl Microbiol Biotechnol, 55, 1±9. Bermejo, L. L., Welker, N. E. and Papoutsakis, E. T. (1998) `Expression of Clostridium acetobutylicum ATCC 824 genes in Escherichia coli for acetone production and acetate detoxification', Appl Environ Microbiol, 64, 1079±85. Biebl, H., Menzel, K., Zeng, A.-P., Deckwer, W.-D. (1999) `Microbial production of 1,3propanediol', Appl Microbiol Biotechnol, 52, 289±297.

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

451

Borden, J. R. and Papoutsakis, E. T. (2007) `Dynamics of genomic-library enrichment and identification of solvent tolerance genes for Clostridium acetobutylicum', Appl Environ Microbiol, 73 (9), 3061±3068. Bowring, S. N. and Morris, J. G. (1985) `Mutagenesis of Clostridium acetobutylicum', J Appl Bacteriol, 58, 577±84. BruÈggemann, H. and Gottschalk, G. (Eds.) (2009) Clostridia ± Molecular Biology in the Post-genomic Era, Caister Academic Press, Norfolk, UK. Celinska, E. and Grajek, W. (2009) `Biotechnological production of 2,3-butanediol ± current state and prospects', Biotechnol Adv, 27 (6), 715±725. Chen, C. K. and Blaschek, H. P. (1999) `Effect of acetate on molecular and physiological aspects of Clostridium beijerinckii NCIMB 8052 solvent production and strain degeneration' Appl Environ Microbiol, 65, 499±505. Chiao, J.-S. and Sun, Z.-H. (2007) `History of the acetone-butanol-ethanol fermentation industry in China: development of continuous production technology', J Mol Microbiol Biotechnology, 13, 12±14. Chidambarama, V. and Viswanathan, B. (2007) `Single step catalytic production of disopropyl ether (DIPE) from acetone feedstock over nickel based catalysts', Applied Catalysis, 71 (1±2), 32±43. Claassen, P. A. M., Budde, M. A. W., Buitelaar, R. M. and Tan, G. B. N. (1998) `Production of Acetone Butanol and Ethanol (ABE) from agricultural residues or Domestic Organic Waste (DOW) and long-term fermentation on glucose'. In 10th European Conference and Technology Exhibition `Biomass for Energy and Industry' (Eds, Kopetz, H., Weber, T., Palz, W., Chartier, P. and Ferrero, G. L.), CARMEN, Rimpar, Germany, WuÈrzburg, Germany, pp. 138±141. Claassen, P. A. M., Budde, M. A. W. and LoÂpez-Contreras, A. M. (2000) `Acetone, butanol and ethanol production from domestic organic waste by solventogenic Clostridia', J Mol Microbiol Biotechnol, 2, 39±44. Cornillot, E., Croux, C. and Soucaille, P. (1997a) `Physical and genetic map of the Clostridium acetobutylicum ATCC 824 chromosome', J Bacteriol, 179, 7426±7434. Cornillot, E., Nair, R. V., Papoutsakis, E. T. and Soucaille, P. (1997b) `The genes for butanol and acetone formation in Clostridium acetobutylicum ATCC 824 reside on a large plasmid whose loss leads to degeneration of the strain', J Bacteriol, 179, 5442±5447. Crout, D. H. G., Littlechild, J. and Morrey, S. M. (1986) `Acetoin metabolism ± stereochemistry of the acetoin produced by the pyruvate decarboxylase of wheatgerm and by the alpha-acetolactate decarboxylase of Klebsiella aerogenes', J Chem Soc ± Perkin Trans, 1, 105±108. Desai, R. P., Harris, L. M., Welker, N. E. and Papoutsakis, E. T. (1999) `Metabolic flux analysis elucidates the importance of the acid-formation pathways in regulating solvent production by Clostridium acetobutylicum', Metabol Eng, 1, 206±13. Donaldson, G. K., Huang, L. L., Nagarajan, L. A. M.-H. V., Nakamura, C. E. and Suh, W. (2007) `Fermentative production of four carbon alcohols', Patent No. WO/2007/ 041269, filed by E.I. Du Pont De Nemours and Company. Dornburg, V., Hermann, B. G. and Patel, M. K. (2008) `Scenario projections for future market potentials of biobased bulk chemicals', Environ Sci Technol, 42, 2261± 2267. DuÈrre, P. (1998) `New insights and novel developments in clostridial acetone/butanol/ isopropanol fermentation', Appl Microbiol Biotechnol, 49, 639±648. DuÈrre, P. (Ed.) (2005) Handbook on Clostridia, Taylor & Francis Group, Boca Raton, FL. DuÈrre, P. (2007) `Biobutanol: An attractive biofuel', Biotechnol J, 2, 1525±1534.

ß Woodhead Publishing Limited, 2010

452

Bioalcohol production

DuÈrre, P. (2008) `Fermentative butanol production. Bulk chemical and biofuel', Ann NY Acad Sci, 1125, 353±362. DuÈrre, P. and Bahl, H. (1996) `Microbial production of acetone/butanol/isopropanol'. In Products of Primary Metabolism, VCH Publisher, Weinheim, Germany, pp. 229±268. Dworkin, M., Falkow, S., Rosenberg., E., Schleifer, K.-H. and Stackebrandt, E. (Eds.) (2006) The Prokaryotes. Volume 1: Symbiotic Associations, Biotechnology, Applied Microbiology, Springer Verlag, Berlin. Ennis, B. M., Gutierrez, N. A. and Maddox, I. S. (1986) `The acetone-butanol-ethanol fermentaiton: a current assessment', Process Biochem, 21, 131±147. EuseÂbio, M. E., Lopes Jesus, A. J., Cruz, M. S. C., LeitaÄo, M. L. and Redinha, J. S. (2003) `Enthalpy of vaporisation of butanediol isomers', J Chem Thermodynamics, 35 (1), 123±129. Ezeji, T. and Blaschek, H. P. (2008) `Fermentation of dried distillers' grains and solubles (DDGS) hydrolysates to solvents and value-added products by solventogenic clostridia', Biores Technol, 99, 5232±5242. Ezeji, T. C., Qureshi, N. and Blaschek, H. P. (2004a) `Acetone butanol ethanol (ABE) production from concentrated substrate: reduction in substrate inhibition by fedbatch technique and product inhibition by gas stripping', Appl Microbiol Biotechnol, 63, 653±658. Ezeji, T. C., Qureshi, N. and Blaschek, H. P. (2004b) `Butanol fermentation research: upstream and downstream manipulations', The Chemical Record, 4, 305±314. Fabret, C., Ehrlich, S. D. and Noirot, P. (2002) `A new mutation delivery system for genome-scale approaches in Bacillus subtilis', Mol Microbiol, 46, 25±36. Fischer, C. R., Klein-Marcuschamer, D. and Stephanopoulos, G. (2008) `Selection and optimization of microbial hosts for biofuels production', Metab Eng, 10, 295±304. Focher, B., Marzetti, A., Beltrame, P. L. and Cartini, P. (1991) `Structural features of cellulose and cellulose derivatives, and their effects on enzymatic hydrolysis'. In Biosynthesis and Biodegradation of Cellulose (Eds, Haigler, C. H. and Weimer, P. J.), Marcel Dekker, New York, pp. 293±310. Fond, O., Jansen, N.B., and Tsao G.T. (1985) `A model of acetic acid and 2,3-butanediol inhibition of the growth and metabolism of Klebsiella oxytoca', Biotechnol Lett, 7, 727±732. Friedl, A., Qureshi, N. and Maddox, I. S. (1991) `Continuous acetone-butanol-ethanol (ABE) fermentation using immobilized cells of Clostridium acetobutylicum in a packed bed reactor and integration with product removal by pervaporation', Biotechnol Bioeng, 38, 518±527. Gapes, J. R. (2000) `The economics of acetone-butanol fermentation: theoretical and market considerations', J Mol Microbiol Biotechnol, 2, 27±32. Gapes, J. R., Larsen, V. F. and Maddox, I. S. (1983) `A note on procedures for inoculum development for the production of solvents by a strain of Clostridium butylicum', J Appl Bacteriol, 55, 363±365. Garg, S. K. and Jain, A. (1995) `Fermentative production of 2,3-butanediol: a review', Biores Technolol, 51, 103±109. Gibbs, D. F. (1983) `The rise and fall (.... and rise?) of acetone/butanol fermentations', Trends Biotechnol, 1, 12±15. Girbal, L. and Soucaille, P. (1998) `Regulation of solvent production in Clostridium acetobutylicum', Trends Biotechnol, 16, 11±16. GonzaÂlez-Pajuelo, M., Meynial-Salles, I., Mendes, F., Andrade, J. C., Vasconcelos, I., Soucaille, P. (2005) `Metabolic engineering of Clostridium acetobutylicum for the industrial production of 1,3-propanediol from glycerol', Metab Eng, 7, 329±336.

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

453

GonzaÂlez-Pajuelo, M., Meynial-Salles, I., Mendes, F., Soucaille, P., Vasconcelos, I. (2006) `Microbial conversion of glycerol to 1,3-propanediol: physiological comparison of a natural producer, Clostridium butyricum VPI 3266, and an engineered strain, Clostridium acetobutylicum DG1(pSPD5)', Appl Environ Microbiol, 72, 96±101. Green, E. M. and Bennett, G. N. (1996) `Inactivation of an aldehyde/alcohol dehydrogenase gene from C. acetobutylicum ATCC 824', Appl Biochem Biotechnol, 57±58, 213±221. Green, E. M. and Bennett, G. N. (1998) `Genetic manipulation of acid and solvent formation in Clostridium acetobutylicum ATCC 824', Biotechnol Bioeng, 58, 215± 221. Green, E. M., Boynton, Z. L., Harris, L. M., Rudolph, F. B., Papoutsakis, E. T. and Bennett, G. N. (1996) `Genetic manipulation of acid formation pathways by gene inactivation in Clostridium acetobutylicum ATCC 824', Microbiology, 142, 2079±86. Groot, W. J. and Luyben, K. (1986) `In situ recovery by adsorption in the butanol/ isopropanol batch fermentation', Appl Microbiol Biotechnol, 25, 29±31. Groot, W. J., Oever, C. E. v. d. and Kossen, N. W. F. (1984) `Pervaporation for simultaneous product recovery in the butanol/isopropanol batch fermentation', Biotechnol Lett, 6, 709±714. Groot, W. J., Schoutens, G. H. and Luyben, K. C. A. M. (1986) `Fermentatieve produktie van een isopropanolbutanol mengsel', I2-processtechnologie, 4, 27±31. Groot, W. J., Lans, R. G. J. M. v. d. and Luyben, K. C. A. M. (1989) `Batch and continuous butanol fermentations with free cells: integration with product recovery by gas-stripping', Appl Microbiol Biotechnol, 32, 305±308. Groot, W. J., Reyer, M. C. H. d., Faille, T. B. d. l., Lans, R. G. J. M. v. d. and Luyben, K. C. A. M. (1991) `Integration of pervaporation and continuous butanol fermentation with immobilized cells. I: Experimental results', Chem Eng J, 46, B1±B10. Groot, W. J., Lans, R. G. J. M. v. d. and Luyben, K. C. A. M. (1992) `Technologies for butanol recovery integrated with fermentations', Process Biochem, 27, 61±75. Grover, B. P., Garg S.K., and Verma J. (1990) `Production of 2,3-butanediol from wood hydrolysate by Klebsiella pneumoniae', World J Microbiol Biotechnol, 6, 328±332. Hanai, T., Atsumi, S. and Liao, J. C. (2007) `Engineered synthetic pathway for isopropanol production in Escherichia coli', Appl Environ Microbiol, 73, 7814±7818. Harris, L. M., Desai, R. P., Welker, N. E. and Papoutsakis, E. T. (2000) `Characterization of recombinant strains of the Clostridium acetobutylicum butyrate kinase inactivation mutant: Need for new phenomenological models for solventogenesis and butanol inhibition?', Biotechnol Bioeng, 67 (1), 1±11. Harris, L. M., Blank, L., Desai, R. P., Welker, N. E. and Papoutsakis, E. T. (2001) `Fermentation characterization and flux analysis of recombinant strains of Clostridium acetobutylicum with an inactivated solR gene', J Ind Microbiol Biotechnol, 27 (5), 322±328. Heap, J. T., Pennington, O. J., Cartman, S. T., Carter, G. P. and Minton, N. P. (2007) `The ClosTron: a universal gene knock-out system for the genus Clostridium', J Microbiol Meth, 70, 452±464. Heap, J. T., Cartman, S. T., Pennington, O. J., Cooksley, C. M., Scott, J. C., Blount, B., Burns, D. A. and Minton, N. P. (2009) `Development of genetic knock-out systems for Clostridia'. In Clostridia Molecular Biology in the Post-genomic Era (Eds, BruÈggemann, H. and Gottschalk, G.), Caister Academic Press, Norfolk, UK, pp. 179±198. Hermann, M., Fayolle, F., Marchal, R., Podvin, L., Sebald, M. and Vandecasteele, J. P.

ß Woodhead Publishing Limited, 2010

454

Bioalcohol production

(1985) `Isolation and characterization of butanol-resistant mutants of Clostridium acetobutylicum', Appl Environ Microbiol, 50, 1238±43. Hiu, S. F., Zhu, C. X., Yan, R. T. and Chen, J. S. (1987) `Butanol-ethanol dehydrogenase and butanol-ethanol-isopropanol dehydrogenase: Different alcohol dehydrogenases in two strains of Clostridium beijerinckii (Clostridium butylicum)', Appl Environ Microbiol, 53 (4), 697±703. Homann, T., Tag, C., Biebl, H., Deckwer, W.-D., Schink, B. (1990) `Fermentation of glycerol to 1,3-propanediol by Klebsiella and Citrobacter strains', Appl Microbiol Biotechnol, 33, 121±126. Hongo, M. and Murata, A. (1965) Agricol Biol Chem, 29, 1140±1145. Houben, M. C. M. (1995) `Oxygenated blending components for gasoline-alcohols and ethers'. In Motor Gasoline: Critical Reports on Applied Chemistry (Eds, Marshall, E. L. and Owen, K.), Royal Society of Chemistry, Cambridge, 45±71. Huang, J. C. and Meagher, M. M. (2001) `Pervaporative recovery of n-butanol from aqueous solutions and ABE fermentation broth using thin-film silicalite-filled silicone composite membranes', J Membr Sci, 192, 231±242. Huang, L., Forsberg, C. W. and Gibbins, L. N. (1986) `Influence of external pH and fermentation products on Clostridium acetobutylicum intracellular pH and cellular distribution of fermentation products', Applied Environ Microbiol, 1230±1234. Huang, T. J., Sorensen, C. M. and Varghese, P. (1990) `Process for the production of ethers'. US Patent No. 4906787, Mobil Oil Corporation. Inui, M., Suda, M., Kimura, S., Yasuda, K., Suzuki, H., Toda, H., Yamamoto, S., Okino, S., Suzuki, N. and Yukawa, H. (2008) `Expression of Clostridium acetobutylicum butanol synthetic genes in Escherichia coli', Appl Microbiol Biotechnol, 77, 1305±1316. Ji, X. J., Huang, H., Du, J., Zhu, J. G., Ren, L. J., Hu, N. and Li, S. (2009) `Enhanced 2,3butanediol production by Klebsiella oxytoca using a two-stage agitation speed control strategy', Bioresour Technol, 100 (13), 3410±3414. Johnson, J. L. and Chen, J.-S. (1995) `Taxonomic relationships among strains of Clostridium acetobutylicum and other phenotypically similar organisms', FEMS Microbiol Rev, 17, 233±240. Johnson, J. L., Toth, J., Santiwatanakul, S. and Chen, J. S. (1997) `Cultures of "Clostridium acetobutylicum" from various collections comprise Clostridium acetobutylicum, Clostridium beijerinckii, and two other distinct types based on DNA-DNA reassociation', Int J Syst Bacteriol, 47, 420±424. Jojima, T., Inui, M. and Yukawa, H. (2008) `Production of isopropanol by metabolically engineered Escherichia coli', J Appl Microbiol Biotechnol, 77, 1219±1224. Jones, D. T. (2001) `Applied acetone-butanol fermentation'. In Clostridia. Biotechnology and medical application (Eds, Bahl, H. and DuÈrre, P.), Wiley-VCH Verlag, Weinheim, Germany, pp. 125±168. Jones, D. T. and Keis, S. (1995) `Origins and relationship of industrial solvent-producing clostridial strains', FEMS Microbiol Rev, 17, 233±240. Jones, D. T. and Woods, D. R. (1986) `Acetone-butanol fermentation revisited', Microbiol Rev, 50, 484±524. Jones, D. T., Shirley, M., Wu, X. and Keis, S. (2000) `Bacteriophage infections in the industrial AB fermentation', J Mol Microbiol Biotechnol, 2, 21±26. Karberg, M., Guo, H. T., Zhong, J., Coon, R., Perutka, J. and Lambowitz, A. M. (2001) `Group II introns as controllable gene targeting vectors for genetic manipulation of bacteria', Nature Biotechnol, 19, 1162±1167. Keis, S., Bennett, C. F., Ward, V. K. and Jones, D. T. (1995) `Taxonomy and phylogeny of industrial solvent-producing clostridia', Int J Syst Bacteriol, 45, 693±705.

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

455

Keis, S., Shaheen, R. and Jones, D. (2001) `Emended descriptions of Clostridium acetobutylicum and Clostridium beijerinckii, and descriptions of Clostridium saccharoperbutylacetonicum sp. nov. and Clostridium saccharobutylicum sp. nov', Int J System Evolution Microbiol, 51, 2095±2103. Kojima, T., Yamamoto, H., Kawada, N. and Matsuyama, A. (1997) `Stereospecific alcohol dehydrogenase isolated from Candida parapsilosis, amino acid and DNA sequences therefor, and method of preparation thereof', US 6255092 B1. Kosaka, T., Nakayama, S., Nakaya, K., Yoshino, S. and Furukawa, K. (2007) `Characterization of the sol operon in butanol-hyperproducing Clostridium saccharoperbutylacetonicum strain N1-4 and its degeneration mechanism', Biosci Biotechnol Biochem, 71, 58±68. Ladisch, M. R. (1991) `Fermentation-derived butanol and scenarios for its uses in energyrelated applications', Enzyme Microbiol Technol, 13, 280±283. Laube, V. M., Groleau, D. and Martin, S. M. (1984) `2,3-Butanediol production from xylose and other hemicellulosic components by Bacillus polymyxa', Biotechnol Lett, 6, 257±262. Lide, D. R. (2008) The CRC Handbook of Chemistry and Physics, 89th edition, CRC Press, Boca Raton, FL. Lin, Y.-L. and Blaschek, H. P. (1983) `Butanol production by a butanol-tolerant strain of Clostridium acetobutylicum in extruded corn broth', Appl Environ Microbiol, 45, 966±973. Linden, J. C., Moreira, A. R. and Lenz, T. G. (1986) `Acetone and butanol'. In Comprehensive Biotechnology. The Principles of Biotechnology: Engineering Consideration (Eds, Cooney, C. L. and Humphrey, A. E.), Pergamon Press, Oxford, pp. 915±931. LoÂpez-Contreras, A. M. (2003) `Utilisation of lignocellulosic substrates by solventproducing clostridia'. In Laboratory of Microbiology/Agrotechnology and Food Innovations BV, Wageningen University and Research Centre, Wageningen, pp. 142. LoÂpez-Contreras, A. M., Claassen, P. A. M., Mooibroek, A. and de Vos, W. M. (2000) `Utilisation of saccharides in extruded domestic organic waste by Clostridium acetobutylicum ATCC 824 to produce acetone, butanol and ethanol', Appl Microbiol Biotechnol, 54, 162±167. LoÂpez-Contreras, A. M., Gabor, K., Martens, A. A., Renckens, B. A. M., Claassen, P. A. M., van der Oost, J. and de Vos, W. M. (2004) `Substrate-induced production and secretion of cellulases by Clostridium acetobutylicum', Appl Environ Microbiol, 70, 5238±5243. Ma, C., Wang, A., Qin, J., Li, L., Ai, X., Jiang, T., Tang, H. and Xu, P. (2009) `Enhanced 2,3-butanediol production by Klebsiella pneumoniae SDM', Appl Microbiol Biotechnol, 82, 49±57. Maddox, I. S. and Anne, E. M. (1983) `Production of n-butanol by fermentation of wood hydrolysate', Biotechnol Lett, 5, 175±178. Maddox, I. S., Qureshi, N. and Gutierrez, N. A. (1993) `Utilization of whey by Clostridia and process technology'. In The Clostridia and Biotechnology (Ed, Woods, D. R.) Butterworth-Heinemann, Stoneham, pp. 343±370. Marchal, R., Blanchet, D. and Vandecasteele, J. P. (1985) `Industrial optimization of acetone-butanol fermentation: a study of the utilization of Jerusalem artichokes', Appl Microbiol Biotechnol, 23, 92±98. Marchal, R., Ropars, M. and Vandecasteele, J. P. (1986) `Conversion into acetone and butanol of lignocellulosic substrates pretreated by steam explosion', Biotechnol Lett, 8, 365±370.

ß Woodhead Publishing Limited, 2010

456

Bioalcohol production

Matsumura, M., Takehara, S., Kataoka, H. (1992) `Continuous butanol/isopropanol fermentation in down-flow column reactor coupled with pervaporation using supported liquid membrane', Biotechnol Bioeng, 39 (2), 148±156. Meyer, C. L. and Papoutsakis, E. T. (1989) `Increased levels of ATP and NADH are associated with increased solvent production in continuous cultures of Clostridium acetobutylicum', Appl Microbiol Biotechnol, 30, 450±459. Mitchell, W. J. (1998) `Physiology of carbohydrate to solvent conversion by Clostridia', Adv Microb Physiol, 39, 31±130. Ni, Y. and Sun, Z. (2009) `Recent progress on industrial fermentative production of acetone-butanol-ethanol by Clostridium acetobutylicum in China', Appl Microbiol Biotechnol, 83, 415±423. Nimcevic, D. and Gapes, J. R. (2000) `The acetone-butanol fermentation in pilot plant and pre-industrial scale', J Mol Microbiol Biotechnol, 2, 15±20. Nimcevic, D., Schuster, M. and Gapes, J. R. (1998) `Solvent production by Clostridium beijerinckii NRRL B592 growing on different potato media', Appl Microbiol Biotechnol, 50, 426±428. NIST (2009) Webbook Chemistry. Retrieved 5 June 2009, from http://webbook.nist.gov/ chemistry/. NoÈlling, J., Breton, G., Omelchenko, M. V., Marakova, K. S., Zeng, Q., Gibson, R., Lee, H. M., Dubios, J., Qiu, D., Hitti, J., GTC Sequencing Center Production, F. a. B. T., Wolf, Y. I., Tatusov, R. L., Sabathe, F., Doucette-Stamm, L., Soucaille, P., Daly, M. J., Bennett, G. N., Koonin, E. V. and Smith, D. R. (2001) `Genome sequence and comparative analysis of the solvent-producing bacterium Clostridium acetobutylicum', J Bacteriol, 183, 4823±4838. O'Neil, M. J., Smith, A. and Heckelman, P. E. (Eds.) (2001) The Merck Index, Merck & Co., Whitehouse Station. Oudshoorn, A., Wielen, L. A. M. v. d. and Straathof, A. J. J. (2009) `Assessment of options for selective 1-butanol recovery from aqueous solutions', Ind Eng Chem Res, 28, 7325±7336. Parekh, M. and Blaschek, H. P. (1999) `Butanol production by hypersolvent-producing mutant Clostridium beijerinckii BA101 in corn steep water medium containing maltodextrin [or glucose, maltose or xylose]', Biotechnol Lett, 21, 45±48. Pasteur, L. (1862) `Extrait des proces-verbaux', Soc Chim Paris, 52. Petrov, K. and Petrova, P. (2009) `High production of 2,3-butanediol from glycerol by Klebsiella pneumoniae G31', Appl Microbiol Biotechnol, 84 (4), 659±665. Prescott, S. C. and Dunn, C. G. (1959a) `The acetone-butanol fermentation'. In Industrial Microbiology (Eds, Prescott, S. C. and Dunn, C. G.), McGraw-Hill, New York, pp. 250±284. Prescott, S. C. and Dunn, C. G. (1959b) `The butanol-isopropanol fermentation'. In Industrial Microbiology (Eds, Prescott, S. C. and Dunn, C. G.), McGraw-Hill, New York, pp. 285±294. Pringsheim, H. (1906a) `Ueber den Ursprung des FuseloÈls und eine alkohole bildende Bakterienfrom', Zentralbl II Bakteriol Parasitenkd Infektionskr Abt 15, 300±321. Pringsheim, H. (1906b) `Ueber ein stickstoff assimilierendes Clostridium', Zentralbl. Bakteriol. Parasitenkd. Infektionskr II Abt 15, 795±800. Qin, J. (2006) `Production of 2,3-butanediol by Klebsiella pneumoniae using glucose and ammonium phosphate', Chin J Chem Eng, 14, 132±136. Qureshi, N. (1994) `Recovery of 2,3-butanediol by vacuum membrane distillation', Sep Sci Technol, 29, 1733±1748. Qureshi, N. and Blaschek, H. P. (1999) `Production of acetone butanol ethanol (ABE) by

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

457

a hyper-producing mutant strain of Clostridium beijerinckii BA101 and recovery by pervaporation', Biotechnol Prog, 15, 594±602. Qureshi, N. and Blaschek, H. P. (2000a) `Butanol production using Clostridium beijferinckii BA101 hyper-butanol producing mutant strain and recovery by pervaporation', Appl Biochem Biotechnol, 84±86, 225±235. Qureshi, N. and Blaschek, H. P. (2000b) `Economics of butanol fermentation using hyper-butanol producing Clostridium beijerinckii BA101', Food Bioprod Process, 78, 139±144. Qureshi, N. and Blaschek, H. P. (2001a) `ABE production from corn: a recent economic evaluation', J Ind Microbiol Biotechnol, 27, 292±297. Qureshi, N. and Blaschek, H. P. (2001b) `Recent advances in ABE fermentation: hyperbutanol producing Clostridium beijerinckii BA101', J Ind Microbiol Biotechnol, 27 (5), 287±291. Qureshi, N. and Blaschek, H. P. (2001c) `Recovery of butanol from fermentation broth by gas stripping', Renewable Energy, 22, 557±564. Qureshi, N., Lolas, A. and Blaschek, H. P. (2001a) `Soy molasses as fermentation substrate for production of butanol using Clostridium beijerinckii BA101', J Ind Microbiol Biotechnol, 290±295. Qureshi, N., Meagher, M. M., Huang, J. and Hutkins, R. W. (2001b) `Acetone butanol ethanol (ABE) recovery by pervaporation using silicalite±silicone composite membrane from fed-batch reactor of Clostridium acetobutylicum', J Membr Sci, 187, 93±102. Qureshi, N., Li, X. L., Hughes, S., Saha, B. C. and Cotta, M. A. (2006) `Butanol production from corn fiber xylan using Clostridium acetobutylicum', Biotechnol Prog, 22, 673±680. Qureshi, N., Saha, B. and Cotta, M. (2007) `Butanol production from wheat straw hydrolysate using Clostridium beijerinckii', Bioprocess Biosystems Eng, 30, 419±427. Ravagnani, A., Jennert, K. C., Steiner, E., Grunberg, R., Jefferies, J. R., Wilkinson, S. R., Young, D. I., Tidswell, E. C., Brown, D. P., Youngman, P., Morris, J. G. and Young, M. (2000) `Spo0A directly controls the switch from acid to solvent production in solvent-forming clostridia', Mol Microbiol, 37, 1172±1185. Raynaud, C., Sarcabal, P., Meynial-Salles, I., Croux, C., Soucaille, P. (2003) `Molecular characterization of the 1,3-propanediol operon of Clostridium butyricum encoding a novel coenzyme B12 independent glycerol dehydratase and a 1,3 propanediol dehydrogenase', Proc Natl Acad Sci USA, 100, 5010±5015. Riley, R. Q. (1995) `Alternative fuels for spark-ignition engines'. In Motor Gasoline: Critical Reports on Applied Chemistry (Eds, Marshall, E. L. and Owen, K.), Royal Society of Chemistry, Cambridge, 239±274. Rogers, P., Chen, J.-S. and Zidwick, M. J. (2006) `Organic acid and solvent production. Part III: Butanol, acetone and iso-propanol; 1,3 and 1,2 propanediol production; and 2,3 butanediol production'. In The Prokaryotes: Symbiotic Associations, Biotechnology, Applied Microbiology, Vol. 1 (Eds, Dworkin, M., Falkow, S., Rosenberg, E., Schleifer, K.-H. and Stackebrandt, E.), Springer, Berlin, pp. 672±755. Roos, J. W., McLaughlin, J. K. and Papoutsakis, E. T. (1985) `The effect of pHon nitrogen supply, cell lysis, and solvent production in fermentations of C. acetobutylicum', Biotechnol Bioeng, 27, 681±694. Rose, A. H. (1961) `Acetone-butanol fermentation'. In Industrial Microbiology (Ed, Rose, A. H.), Butterworth, London, pp. 160±166. Santangelo, J. D., Kuhn, A., Treuner Lange, A. and DuÈrre, P. (1998) `Sporulation and time course expression of sigma-factor homologous genes in Clostridium

ß Woodhead Publishing Limited, 2010

458

Bioalcohol production

acetobutylicum', FEMS Microbiol Lett, 161, 157±164. Sawada, H. and Takagi, Y. (1964) `The metabolism of L-rhamnose in Escherichia coli. 3. L-Rhamulose-phosphate aldolase', Biochim Biophys Acta, 92, 26±32. Schardinger, F. (1905) `Bacillus macerans, ein aceton bildender Rottebacillus', Zentr. Bakt Parasitenk, 14, 772±781. Schoutens, G. H. and Groot, W. J. (1985) `Economic feasibility of the production of isopropanol-butanol-ethanol fuels from whey permeate', Process Biochem, 20, 117±121. Schultz, T. W., Cronin, M.T.D., Netzeva, T.I., Aptula. A.O. (2002) `Structure±toxicity relationships for aliphatic chemicals evaluated with Tetrahymena pyriformis', Chem. Res. Toxicol, 15, 1602±1609. Schuster, K. C., Goodacre, R., Gapes, J. R. and Young, M. (2001) `Degeneration of solventogenic Clostridium strains monitored by Fourier transform infrared spectroscopy of bacterial cells', J Ind Microbiol Biotechnol, 27, 314±21. SchuÈtz, H. and Radler, F. (1984) `Anaerobic reduction of glycerol to propanediol-1,3 by Lactobacillus brevis and Lactobacillus buchneri', Syst Appl Microbiol, 5, 169±178. Shao, L., Hu, S., Yang, Y., Gu, Y., Chen, J., Yang, Y., Jiang, W. and Yang, S. (2007) `Targeted gene disruption by use of a group II intron (targetron) vector in Clostridium acetobutylicum', Cell Res, 17, 963±965. Shen, C. R. and Liao, J. C. (2008) `Metabolic engineering of Escherichia coli for 1butanol and 1-propanol production via the keto-acid pathways', Metab Eng, 10, 312±320. Sillers, R., Chow, A., Tracy, B. and Papoutsakis, E. T. (2008) `Metabolic engineering of the non-sporulating, non-solventogenic Clostridium acetobutylicum strain M5 to produce butanol without acetone demonstrate the robustness of the acid-formation pathways and the importance of the electron balance', Metab Eng, 10 (6), 321±332. Sillers, R., Al-Hinai, M. A. and Papoutsakis, E. T. (2009) `Aldehyde-alcohol dehydrogenase and/or thiolase overexpression coupled with CoA transferase downregulation lead to higher alcohol titers and selectivity in Clostridium acetobutylicum fermentations', Biotechnol Bioeng, 102 (1), 38±49. Sombrutai, W., Takagi, M. and Yoshida, T. (1996) `Acetone-butanol fermentation by Clostridium aurantibutyricum ATCC 17777 from a model medium using palm oil mill effluent', J Ferment Bioeng, 81, 543±547. Soni, B. K., Das, K. and Ghose, T. K. (1982) `Bioconversion of agro-wastes into acetone butanol', Biotechnol Lett, 4, 19±22. Soucaille, P., Joliff, G., Izard, A. and Goma, G. (1987) `Butanol tolerance and autobacteriocin production by Clostridium acetobutylicum', Curr Microbiol, 14, 295±299. Soucaille, P., Figge, R. and Croux, C. (2008) `Process for chromosomal integration and DNA sequence replacement in clostridia', Application No. 11/737,025 at US Patent Office, 27. Speranza, G., Corti, S., Fontana, G., Manitto, P., Galli, A., Scarpellini, M. and Chialva, F. (1997) `Conversion of meso-2,3-butanediol into 2-butanol by lactobacilli. Stereochemical and enzymatic aspects', J Agric Food Chem, 45, 3476±3480. Spivey, M. J. (1978) `The acetone/butanol/ethanol fermentation', Process Biochem, 13, 2±4, 25. Stampfer, W., Kosjek, B., Kroutil, W. and Faber, K. (2003) `Alcohol dehydrogenases with high solvent and temperature stability', US2004157305 (WO03078615). Stephens, G. M., Holt, R. A., GottschalK, J. C. and Morris, J. G. (1985) `Studies on the stability of solvent production by Clostridium acetobutylicum in continuous

ß Woodhead Publishing Limited, 2010

Production of longer-chain alcohols from lignocellulosic biomass

459

culture', J Appl Bacteriol, 59, 597±605. Sun, L. H., Jiang, B. and Xiu, Z. L. (2009a) `Aqueous two-phase extraction of 2,3butanediol from fermentation broths by isopropanol/ammonium sulfate system', Biotechnol Lett, 31, 371±376. Sun, L. H., Wang, X. D., Dai, J. Y. and Xiu, Z. L. (2009b) `Microbial production of 2,3butanediol from Jerusalem artichoke tubers by Klebsiella pneumoniae', Appl Microbiol Biotechnol, 82, 847±852. Syu, M. J. (2001) `Biological production of 2,3-butanediol', Appl Microbiol Biotechnol, 55, 10±18. Taylor, M. B. and Juni, E. (1960) `Stereoisomeric specificities of 2,3-butanediol dehydrogenases', Biochimica et Biophysica Acta, 39 (3), 448±457. Tomas, C. A., Welker, N. E. and Papoutsakis, E. T. (2003a) `Overexpression of groESL in Clostridium acetobutylicum results in increased solvent production and tolerance, prolonged metabolism, and changes in the cell's transcriptional program', Appl Environ Microbiol, 69 (8), 4951±4965. Tomas, C. A., Alsaker, K. V., Bonarius, H. P. J., Hendriksen, W. T., Yang, H., Beamish, J. A., Paredes, C. J. and Papoutsakis, E. T. (2003b) `DNA array-based transcriptional analysis of asporogenous, nonsolventogenic Clostridium acetobutylicum strains SKO1 and M5', J Bacteriol, 185, 4539±4547. Tran-Din, K. and Gottschalk, G. (1985) `Formation of D(-)-1,2-propanediol and D(-)lactate from glucose by Clostridium sphenoides under phosphate limitation', Arch Microbiol, 142, 87±92. Tummala, S. B., Welker, N. E. and Papoutsakis, E. T. (2003a) `Design of antisense RNA constructs for downregulation of the acetone formation pathway of Clostridium acetobutylicum'. J Bacteriol, 185, 1923±1934. Tummala, S. B., Junne, S. G. and Papoutsakis, E. T. (2003b) `Antisense RNA downregulation of coenzyme A transferase combined with alcohol-aldehyde dehydrogenase overexpression leads to predominantly alcohologenic Clostridium acetobutylicum fermentations', J Bacteriol, 185 (12), 3644±3653. Ui, S., Masuda, T., Masuda, H. and Muraki, H. (1986) `Mechanism for the formation of 2,3butanediol stereoisomers in Bacillus polymyxa', J Ferment Technol, 64(6), 481±486. Vane, L. M. (2005) `A review of pervaporation for product recovery from biomass fermentation processes', J Chem Technol Biotechnol, 80, 603±629. Voget, C. E., Mignone, C. F. and Ertola, R. J. (1985) `Butanol production from apple pomace', Biotechnol Lett, 7, 43±46. Voloch, M., Jansen N.B., Ladisch M.R. Tsao, G.T. Narayan, R. and Rodwell, V.W. (1985) `2,3-Butanediol'. In Comprehensive Biotechnology: The Principles, Applications and Regulations of Biotechnology in Industry, Agriculture and Medicine. Pergamon Press, Oxford. von Abendroth, G., Stripp, S., Silakov, A., Croux, C., Soucaille, P., Girbal, L. and Happe, T. (2008) `Optimized over-expression of [FeFe] hydrogenases with high specific activity in Clostridium acetobutylicum', Int J Hydrogen Energy, 33, 6076±6081. Walton, M. T. and Martin, J. L. (1979) `Production of Butanol-Acetone by Fermentation'. In Microbial Technology, Vol. 1 (Eds, Peppler, H. J. and Perlman, D.), Academic Press, New York, pp. 187±209. Wardwell, S. A., Yang, Y. T., Chang, H. Y., San, K. Y., Rudolph, F. B. and Bennett, G. N. (2001) `Expression of the Klebsiella pneumoniae CG21 acetoin reductase gene in Clostridium acetobutylicum ATCC 824', J Ind Microbiol Biotechnol, 27, 220±227. Weber, F. J. and de Bont, J. A. M. (1996) `Adaptation mechanisms of microorganisms to the toxic effects of organic solvents on membranes', Biochimica et Biophysica Acta,

ß Woodhead Publishing Limited, 2010

460

Bioalcohol production

1286, 225±245. Werken, H. J. G. v. d., Verhaart, M. R. A., VanFossen, A. L., Willquist, K., Lewis, D. L., Nichols, J. D., Goorissen, H. P., Mongodin, E. F., Nelson, K. E., Niel, E. W. J. v., Stams, A. J. M., Ward, D. E., Vos, W. M. d., Oost, J. v. d., Kelly, R. M. and Kengen, S. W. M. (2008) `Hydrogenomics of the extremely thermophilic bacterium Caldicellulosiruptor saccharolyticus', Appl Environ Microbiol, 74, 6720±6729. Westhuizen, A. v. d., Jones, D. T. and Woods, D. R. (1982) `Autolytic activity and butanol tolerance of Clostridium acetobutylicum', Appl Environ Microbiol, 44, 1277±1281. Wilke, T. and Vorlop, K.-D. (2004) `Industrial bioconversion as an alternative to conventional chemistry', Appl Microbiol Biotechnol, 66, 131±142. Woods, D. R. (1995) `The genetic engineering of microbial solvent production', Trends Biotechnol, 13, 259±264. Woolley, R. C. and Morris, J. G. (1990) `Stability of solvent production by Clostridium acetobutylicum in continuous culture: strain differences', J Appl Bacteriol, 69, 718± 728. Xiao, Z. and Xu, P. (2007) `Acetoin metabolism in bacteria', Crit Rev Microbiol, 33, 127±140. Xiu, Z. L. and Zeng, A. P. (2008) `Present state and perspective of downstream processing of biologically produced 1,3-propanediol and 2,3-butanediol', Appl Microbiol Biotechnol, 78, 917±926. Young, M. (1993) `Development and exploitation of conjugative gene transfer in clostridia'. In The Clostridia and Biotechnology (Ed, Woods, D. R.), ButterworthHeinemann, Stoneham. Young, M., Minton, N. P. and Walter, L. S. (1989) `Recent advances in the genetics of the clostridia', FEMS Microbiol Rev, 63, 301±326. Yu, E. K. and Saddler, J. N. (1982) `Enhanced production of 2,3-butanediol by Klebsiella pneumoniae grown on high sugar concentrations in the presence of acetic acid', Appl Environ Microbiol, 44, 777±784. Yu, E. K. and Saddler, J. N. (1983) `Fed-batch approach to production of 2,3-butanediol by Klebsiella pneumoniae grown on high substrate concentrations', Appl Environ Microbiol, 46, 630±635. Yu, E. K. C., Deschatelets, L. and Saddler, J. N. (1984) `The bioconversion of wood hydrolyzate to butanol and butanediol', Biotechnol Lett, 6, 327±332. Yuan, L. (2007) `Supply-demand analysis of n-butanol and octanol both home and abroad as well as their technical development trends', Petroleum and Petrochemical Today, 15, 33±44 (In Chinese). Zappe, H., Jones, W. A., Jones, D. T. and Woods, D. R. (1988) `Structure of an endo-b1,4-glucanase gene from Clostridium acetobutylicum P262 showing homology with endoglucanase genes from Bacillus spp', Appl Environ Microbiol, 54, 1289±1292. Zhang, K., Sawaya, M. R., Eisenbergb, D. S. and Liao, J. C. (2008) `Expanding metabolism for biosynthesis of nonnatural alcohols', Proc Natl Acad Sci, 105, 20653±20658. Zhong, J., Karberg, M. and Lambowitz, A. M. (2003) `Targeted and random bacterial gene disruption using a group II intron (targetron) vector containing a retrotransposition-activated selectable marker', Nucleic Acids Res, 31, 1656±1664. Zverlov, V. V., Berezina, O., Velikodvorskaya, G. A. and Schwarz, W. H. (2006) `Bacterial acetone and butanol production by industrial fermentation in the Soviet Union: use of hydrolyzed agricultural waste for biorefinery', Appl Microbiol Biotechnol, 71, 587±597.

ß Woodhead Publishing Limited, 2010

Index

-amylase, 260 -arabinofuranosidase, 31 -glucuronidase, 31 ABE fermentation see acetone, butanol and ethanol fermentation acetaldehyde, 322 acetate, 153 acetic acid, 296, 399, 410 acetoacetyl-CoA transferase, 434 Acetobacter xylinum, 218 acetoin, 322, 420 acetolactate, 441 acetone, 419 acetone, butanol and ethanol fermentation, 416, 419, 438, 447 commercial production plant, 428 schematic flow sheet from sugarcane-based plant, 429 substrates used as alternative to starch-based or sugary substrates, 435 acetyl xylan esterases, 31 acid hydrolysis, 143±56 concentrated acid hydrolysis, 144±5 dilute acid hydrolysis, 143±4 ethanol production plants, 149±52 future trends, 155±6 pre-treatment process and apparatus, 145±9 unit operations pertinent to ethanol industry, 152±5 acid pre-treatment, 35±7, 402 acidogens, 406 Actinobacillus succinogenes, 404 activated charcoal, 398 advanced hydrolysis-fermentation technology, 370 aeration, 330 AFEX see ammonia fibre expansion agitation, 442 agroecosystem, 383 air gasification, 396 alcohols, 406

aldehydes, 299, 406 alkali pulping, 29 alkaline pre-treatment, 38±9, 402 alkaline wet oxidation, 39 aluminium, 399 aluminium oxide, 409 Amazonian stinkbird Hoatzin see Opisthocomus hoazin American Electric Power, 400 amino-modified silica gel, 294 ammonia, 400, 409, 426 ammonia fibre expansion, 25±6, 29±30, 32, 38, 44, 47, 89, 99, 128, 211, 212 ammonia peroxide pre-treatment, 38 ammonia recycle percolation, 38, 89 ammonium bicarbonate, 407 Anthrone method, 291 antisense RNA techniques, 430 aquasolv treatment see liquid hot water pretreatment Arabidopsis thaliana, 227, 229 arabinose, 226 arabitol, 231 Arkenol Inc., 149±52, 155 Arkenol model, 150, 151 aromatic acid, 299 ASPEN Plus, 252, 386 model, 136 aspen wood, 93 Asplund process, 5 autohydrolysis, 11, 144 Avicel, 35, 98 Avicel FMC PH105, 126 azeotropic distillation, 249±50 -branched alcohols, 258 -glucosidases, 30, 161, 166, 179, 187, 238, 347, 403 -xylosidase, 31 Bacillus, 404, 419 Bacillus polymyxa, 419, 442

ß Woodhead Publishing Limited, 2010

462

Index

Bacillus subtilis, 321 bacterial cellulose, 98 bacterial microcrystalline cellulose, 98, 126 bacteriophages, 427 Baer, 447 baggase, 80 baker's yeast, 224±39 ball milling, 33 Bancroft co-ordinates, 258 Bankia setacea, 189 barley straw, 164 batch reactor, 145±6 beet molasses, 322 benzene, 410 beta-glucosidase, 124 bio-based adsorbents, 259±60 bio-oil, 398, 399 Bio-Rad Aminex HPX-87C, 323 Bio-Rad Aminex HPX-87H ion exclusion/ cation exchange column, 323 Bio-Rad Pb2+ HPX-87H column, 323 bioalcohol, 246 bioaugmentation, 407 biochar production, 383 bioethanol, 246 process, 327 different online measurement methods, 320 general process scheme, 317 biofuels activities included in WTW studies, 367±8 life cycle assessment, 367±72 routes for environmental improvement, 383±5 feedstock production improvements, 383±4 lignocellulosic biomass conversion to bioalcohols, 384 public policy issues, 384±5 vs other fuels life cycle assessments, 377±81 energy yield per agricultural land area for biofuels, 380 GHG emissions avoided with biofuels vs gasoline or diesel, 380 WTW studies review, 368±72 key issues, 368±9 lignocellulosic biomass alternative applications, 372 lignocellulosic ethanol results, 370±2 biomass, 73, 143 biodegradation, 160 geometry, 394 hydrolysates, 351±3 hydrolysis, 159 pyrolysis, 383 simplified conceptual model of structure, 99

see also lignocellulosic biomass Bioscience, 447 bisphenol A, 405 black-box methods, 332 blackstrap molasses, 426 BlueFire Inc., 155 BMCC see bacterial microcrystalline cellulose borate, 300 boric acid, 409 Brazilian sugarcane ethanol, 377, 380 process, 379 British Petroleum, 418 bubble cap tray column, 251 2,3-butanediol advances in production, 442±5 2,3-butanediol producing strains improvement, 442 process technology, 442±5 characteristics and uses, 419±20 different metabolic pathways, 441 improved fermentation systems comparison, 443 production from lignocellulosic biomass, 415±48 production methods, 441 butanol, 379, 416 advances in process technology, 434±40 advanced fermentation and down stream products processing, 438 alternative fermentation substrates, 434±8 gas-stripping, 440 pervaporation, 438±40 advances in production, 429±34 butanol-producing strains improvement, 429±32 production by species other than clostridia, 432±4 characteristics and uses, 416, 418±19 production by clostridia, 420±9 microorganisms and metabolism, 420±4 process technology, 424±9 production from lignocellulosic biomass, 415±48 2-butanol, 446±7 butanol fermentation process, 416, 448 ButylFuel LLC, 418, 447 butyric acid, 410 Byogy Renewables Inc., 410 caboxymethylcellulase, 218 Cadoxen, 129 caffeic acid, 287 CAFI Project, 41, 48 calcium carbonate, 407 calcium carboxylates, 409 calcium formate, 409

ß Woodhead Publishing Limited, 2010

Index Caldicellulosiruptor saccharolyticus, 445 California Air Resources Board, 375 California's Low Carbon Fuel Standard, 385 Candida intermedia, 227 Candida parapsilosis, 447 Candida shehatae, 213, 353 capillary electrophoresis, 299, 300 carbohydrate-binding modules, 161, 167±8 carbohydrates alternative analysis techniques, 300±1 analysis, 288±96 colorimetric analysis, 289±91 chemical assay, 290±1 enzyme assay, 289±90 gas chromatography, 291±3 carbohydrate derivatisation, 292±3 detection technique utilised in gas chromatography, 291±2 GC analysis detected D-glucose isomers, 293 HPLC, 293±6 alternative separation techniques, 294±5 ion-exchange techniques, 294 selective detection techniques, 295±6 universal techniques for analysis, 295 sugars in biomass hydrolysates of lignocellulosic materials, 289 carbonic acid, 37 carboxylate platform chemistry and secondary chemicals, 409±10 feedstock considerations, 406 fermentation, 406±7 lignocellulosic biomass chemical production, 405±10 major pathways to secondary chemicals, 409 MixAlco process, 407±8 flow diagram, 407 research and development, 410 ZeaChem technology, 408 carboxylate salts, 405 carboxylic acid, 405, 409, 410 carboxymethylcellulose, 182 Cargill Dow, 404 catalytic hydratation process, 419 CBP see consolidated bioprocessing cellobiohydrolase I, 81 cellobiohydrolases, 30, 31, 124, 161, 167, 179, 347 cellobiose, 179, 184 cellobiose dehydrogenase, 185 Celluclast 1.5 L, 166 cellulases, 75, 80, 161, 183, 402, 403 adsorption capacity, 76 components and activity on cellulose substrates, 82

463

classes, 30 development, 187±91 fungal expression systems, 190±1 genomics and metagenomics approaches, 189 identifying novel cellulases, 188±9 protein engineering, 189±90 synthetic cellulase mixtures, 191 development to improve enzymatic hydrolysis of lignocellulosic biomass future trends, 193±4 issues, 192±3 recent developments, 191±2 lignocellulosic biomass enzymatic hydrolysis improvement, 178±94 structure and function, 179±87 biomass saccharification time-dependence, 186 cellulolysis kinetics, 185±7 cellulolysis mechanisms, 182±3 cellulose degradation, 184±5 cellulose degradation enzymatic mechanisms, 185 major secreted proteins of T. reesei, 180 non-cellulase secreted factors, 184 structure, 181±3 types of fungal cellulases, 179±80 cellulolysis kinetics, 185±7 mechanisms, 182±3 cellulose, 3, 6, 10, 26±7, 124±6, 129, 162, 283, 319, 392 breakdown mechanism, 161 crystallinity, 15 maximum cellulose adsorption capacity, 80 maximum cellulose adsorption capacity and equilibrium constants, 81 degradation, 184±5 degree of polymerisation, 15 key substrate features controlling hydrolysis crystallinity, 75±84 degree of polymerisation, 84±8 hemicellulose and degree of hemicellulose acetylation, 88±91 lignin, 91±7 relative polymer degradation, 13 cellulose-binding module, 126, 181 cellulose conversion, 152 cellulose I, 38 cellulose III, 38 cellulose solvent- and organic solvent-based lignocellulose fractionation, 125, 129±34, 135 concentrated phosphoric acid and acetone recycling flowchart, 146

ß Woodhead Publishing Limited, 2010

464

Index

flowchart, 130 vs dilute acid pre-treatment on biomass composition, 131 cellulose wastes, 210 cellulosic biomass, 210 cellulosic ethanol, 355, 356, 357 cellulosomes, 183 cesium oxide catalyst, 409 CFX see corn fibre xylan char, 398 chemical pre-treatment, 35±40 acidic-based, 35±7 alkali-based, 38±9 oxidative, 39±40 chemometric models, 282, 332 chitosan, 262 Chloella kessleri, 227 citramalate synthase enzyme, 433 clean coal technology, 400 Clostridium, 420, 432 Clostridium acetobutylicum, 422, 423, 424, 430, 433, 446 mutans overview and their fermentation performance, 431 Clostridium beijerinckii, 422, 423, 430, 436 mutans overview and their fermentation performance, 431 Clostridium butyricum, 446 Clostridium isoprpylicum, 434 Clostridium saccharobutylicum, 432 Clostridium saccharoperbutylacetonicum, 422, 424 Clostridium sphenoides, 445 Clostridium thermocellum, 217, 353, 354 Clostridium thermohydrosulfuricum, 257, 353 Clostridium thermosaccharolyticum, 217 CNG see compressed natural gas CNG internal combustion engine, 378 co-current reactors, 8, 145 co-fermentation, 205±19 challenges in lignocelluloses-derived sugars, 224±39 combining recombinant pathways, 230±2 enzyme cross-affinity, 231±2 xylose and arabinose pathways, 231 transcriptional regulation in mixed substrate fermentations, 232±8 engineering regulatory response of S. cerevisiae, 235±7 glucose signalling, 232±5 physiological response to altered glucose signalling, 233 regulatory response to novel substrates, 237±9 transporter preferences, 225±30 Gxf1 transporter expression, 229

heterologous transporters expression, 227, 229±30 heterologous xylose transporters, 228 hexose sugars uptake, 226 is transport controlling pentose sugars uptake rate, 230 kinetic parameters for Hxt transporters for glucose, 226 pentose sugars transport improvement, 227 pentose sugars uptake, 226±7 coal, 382 Cobalt Biofuels, 447 cohesin, 183 combined severity factor, 36 comminution, 4 composite membranes, 265±6 compressed natural gas, 377 CONCAWE see Health and Safety in Refining and Distribution concentrated acid hydrolysis, 144±5 conductometric detection, 285 consolidated bioprocessing, 59, 193, 206, 208±9, 219, 354 feedstocks and pre-treatment, 210±13 microbial strains, 217±18 scheme, 209 corn, 400 corn fibre xylan, 437 corn starch, 426 corn stover, 41±7, 42, 43, 86, 87, 92, 93, 135, 211, 376, 383 thermochemical pre-treatments effect, 45±6 corrosion, 398 Corynebacterium glutamicum, 153 COSLIF see cellulose solvent- and organic solvent-based lignocellulose fractionation cotton stalks, 92 counter current reactors, 8 countercurrent steam stripping, 445 Crabtree metabolism, 236 crop residues, 210 crystallinity accessibility, 75±83 controlling cellulose hydrolysis, 75±84 effect on maximum cellulose adsorption capacity, 80 effect on maximum cellulose adsorption capacity and equilibrium constants, 81 effectiveness, 83±4 vs cellulose viscosity degree of polymerisation, 87 cyclohexane, 410 D-arabinose, 293

ß Woodhead Publishing Limited, 2010

Index D-arabitol, 293 D-xylose, 293 degeneration, 423 degree of polymerisation accessibility, 85±7 controlling cellulose hydrolysis, 84±8 effect of pre-treatment severity for corn stover solids, 86 effectiveness, 87±8 Delta-T Company, 259 descummed, 407 diacetyl, 420 diazomethane, 298 dielectric spectroscopy, 329 diesel, 378, 400 diesel-fuelled farm, 384 diisopropyl ether, 419 dilute acid, 211, 212 hydrolysis, 153±4 pre-treatment, 36, 131 dimethyl ether, 370 dimethyl sulfoxide, 33 Dionex application note 363, 147 Dionex ASE 350 Accelerated Solvent Extractor, 147, 148 Dionex ion chromatography column CarboPac, 323 DIPE see diisopropyl ether diphenolic acid, 405 direct LUC, 371 direct microbial conversion, 208 see also consolidated bioprocessing DMSO see dimethyl sulfoxide dockerin, 183 Duke Energy, 400 Dupont, 418 ECOENG 418, 255 electrochemical detection, 300 electrospray ionisation, 296 Embden-Meyerhof pathway, 422 endo-1,4-glucanases, 167 endo- -glucanase, 166 endoglucanase I, 81 endoglucanase II, 83 endoglucanases, 30, 31, 124, 161, 179, 181, 347 endogluconases, 403 endoxylanases, 31 Enterobacter, 419 entrainers, 249 environmental life cycle assessment biochemical production, 372±7 biofuels LCA, 367±72 future trends, 385±7 lignocellulose-to-bioalcohol production, 365±87

465

lignocellulosic alcohol biofuel vs other fuels, 377±81 lignocellulosic bioalcohols vs alternative biomass utilisation, 381±3 routes for environmental improvement, 383±5 enzymatic cellulose hydrolysis, 124±6 enzymatic hydrolysis, 205±6, 347±8 classification scheme for models, 349 functionally based models, 348 nonmechanistic models, 348 semimechanistic model, 348 factors affecting efficiency, 164±8 enzyme-substrate interactions, 167±8 nature of the enzymes, 166±7 substrate structure, 165±6 summary, 165 fermentation related parameter ranges, 54±5 improvement through cellulases development, 178±94 cellulase structure and function, 179±87 cellulases development, 187±91 future trends, 193±4 issues, 192±3 recent developments, 191±2 improvements, 168±70 cellulase performance, 168±9 pre-treated substrate characteristics, 169±70 lignocellulosic biomass, 159±71 cellulose breakdown mechanism, 161 future trends, 170±1 mechanism, 160±2 relative saccharification efficiencies, 163±4 substrate concentration on pre-treated barley straw, 164 time course at different enzyme concentrations, 163 mechanistic model for cellulose by T. reesei, 348 nonmechanistic mathematical model, 350 enzyme cocktails, 160 epoxides, 404 equivalence ratio, 394 Erwinia chrysanthemi, 218 Escherichia coli, 153, 190, 208, 215, 218, 227, 296, 328, 331, 403, 434, 445 Escherichia coli K011, 47, 215 Escherichia coli LY01, 215 ESI see electrospray ionisation esterases, 32 esters, 406 ethanol, 5, 12, 123, 288, 316, 322, 370, 379, 400

ß Woodhead Publishing Limited, 2010

466

Index

dehydration azeotropic distillation flowsheet, 249 different agents for extractive distillation, 256 extractive distillation with gasoline, 251 hydrophilic membranes, 262±6 ethanol fuel, 355 ethyl acetate, 288 ethyl phosphate catalyst, 410 ethylene, 400 ethylene glycol, 250, 253 Eucalyptus grandis, 324 EUCAR see European Council for Automotive Research & Development European Commission Joint Research Centre, 369, 386 European Council for Automotive Research & Development, 369, 385 exo- -glucanase, 166 exoglucanases, 30, 161 exogluconases, 403 extended Kalman filter, 332 extractive distillation, 250±5 conventional liquid solvent and dissolved salt mixture, 253 conventional solvent, 250±1 dissolved salt, 251±3 schematics, 252 ethanol dehydration typical process, 250 gasoline for ethanol dehydration, 255 hyper-branched polymers, 253 ionic liquid, 254±5 separating agents comparison, 255 extractive fermentation, 255±8 extractives, 319 Fabospora, 214 fast pyrolysis, 396 feedstock, 152 characteristics, 8±9 production, 376 steam explosion pre-treatment conditions and results, 7 Fenton chemistry, 185 fermentation, 56, 205±19, 406±7 fermentation gases, 427 fermentation processes biomass hydrolysates, 351±3 chromatographic techniques, 322±4 gas chromatography, 322±3 liquid chromatography, 323±4 general bioethanol process scheme, 317 inhibition, 350±1 lignocellulosic hydrolysates detoxification, 351 role of furfural, 350±1

role of HMF, 351 modelling in lignocellulose-to-bioalcohol production, 340±59 NIR and fluorescence applications, 326±30 cell mass monitoring, 328±9 implementation challenges, 329±30 Raman spectroscopy, 329 raw material composition and quality, 326±7 substrates and ethanol monitoring, 327±8 online monitoring in lignocellulose-tobioalcohol production, 315±34 pentoses, 353±4 potential monitoring techniques, 319±22 software sensors, 330±3 advantages, 331 data-driven modelling methods, 332±3 hybrid software sensors, 331 stoichiometry, elementary balances and first-principles models, 330±2 spectroscopic methods basic principles, 324±5 data analysis, 325±6 variables of interest to monitor in bioethanol production processes, 318±19 various microorganisms, 350±4 feruloyl esterase, 31 fibrous cellulose powder, 126 FID see flame ionisation detection filtering, 330 filtration, 151, 321 1st generation fuels, 368, 377 Fischer-Tropsch catalysts, 399 Fischer-Tropsch diesel, 370 fixed-bed gasifiers, 394 flame ionisation detection, 285, 291 flash pyrolysis see fast pyrolysis flow-through acid pre-treatment, 144 flow-through reactors, 8, 149 fluidised-bed gasifiers, 397 fluidised-bed reactors, 394 fluorescence spectroscopy, 285, 324, 325 fluoriphoric, 300 forest biorefinery, 247 Forestry and Agriculture Sector Optimisation Model, 386 formaldehyde, 399 fossil fuels, 377 Fourier transform-infrared spectroscopy, 300 fuel ethanol, 343±4 Fungal Genomics Program, 189 furans, 299 furfurals, 15, 57, 127, 153, 296, 297, 317, 350±1, 405 galactose, 226

ß Woodhead Publishing Limited, 2010

Index galactose repressors, 236 gas chromatography, 317, 321 gas-stripping, 440 gasification, 377, 396±7 gasifiers, 397 gasohol, 251 gasoline, 251, 378, 400 General Electric, 400 genetic engineering, 446 genetically modified organisms, 406 genomics, 189 GFP see green fluorescent protein GHG emissions see greenhouse gas emissions GHGenius, 385 Global Trade Analysis Project, 386 glucose, 226, 288, 331, 345, 445 repression, 232, 234±5 signalling, 232±5 glucose repression, 234±5 HXT transporters regulation, 235 Ras2/cAMP/PKA dependent regulation, 235 schematic illustration, 234 glycerol, 238 glycosidic bonds, 10 glycosylamine bond, 295 green fluorescent protein, 126 greenhouse gas emissions, 365 Greenhouse Gases, Regulated Emissions, and Energy Use Transportation model, 376, 377, 379, 385 (version 1.8b), 373 GroEL, 432 GroES, 432 GTAP see Global Trade Analysis Project guaiacyl, 297 H-factor, 11 hardwood, 210, 354 Health and Safety in Refining and Distribution, 369, 370, 385 Helianthus tuberosus, 435 helium, 291 hemicellulases, 347 hemicellulose, 9, 10, 15, 27, 126±7, 128, 162, 283, 316, 319, 392 acid hydrolysis, 347 controlling cellulose hydrolysis, 88±91 accessibility, 88±90 effectiveness, 90±1 degrading enzymes classes, 31 degree of acetylation, 88±91 relative polymer degradation, 13 hemp, 134 herbaceous biomass, 210 hexose, 15, 226, 284

467

high performance anion exchange chromatography, 294 high performance liquid chromatography, 293±6, 298±9 high-test molasses, 426 HMF see 5-hydroxymethyl furfural HPAEC see high performance anion exchange chromatography HPLC see high performance liquid chromatography hydrazine, 33 hydrogen, 399 hydrolysation, 152 hydrolysis, 193 analytical monitoring in lignocellulose-tobioalcohol production, 281±301 alternative analysis techniques, 300±1 biomass hydrolysates preparation, 287±8 bioprocessing analytical perspective, 281±2 carbohydrates analysis, 288±96 detection strategies, 285±7 future trends, 301 lignocellulosic degradation products analysis, 296±9 resulting target analytes, 283±5 lignocellulosic hydrolysis kinetics, 346 modelling in lignocellulose-to-bioalcohol production, 340±59 hydroperoxide process, 419 hydrophilic inorganic membranes, 263±4 hydrophilic mixed matrix, 265±6 hydrophilic polymeric membranes, 262±3 hydrothermal pre-treatment feedstock characteristics, 8±9 hydrothermal reactions, 9±10 lignocellulosic biomass, 3±18 future trends, 17±18 liquid hot water and steam, 5±16 physical comminution, 4 steam explosion vs liquid hot water, 17 method of action, 9±10 pre-treated biomass physical and chemical characteristics, 14±16 liquid hot water, 16 steam explosion, 14±15 process history and description, 5±8 liquid hot water, 7±8 steam explosion, 5±7 severity, 10±14 effect on soluble lignin, pentoses and hexoses concentrations, 13 hemicellulose, lignin and cellulose relative amount, 13 liquid hot water, 13±14 steam explosion, 11±13

ß Woodhead Publishing Limited, 2010

468

Index

hydrothermal treatment see liquid hot water pre-treatment hydrothermolysis, 8 hydrotreating, 399 4-hydroxybenzaldehyde, 153 hydroxylamine, 298 5-hydroxymethyl furfural, 15, 57, 153, 297, 351, 402, 405 Hypocrea jecorina, 179, 227 IBE fermentation see isopropanol, butanol and ethanol fermentation IBUS see Integrated Biomass Utilisation System ICE see CNG internal combustion engine ICE automobile, 381 IGCC see Integrated Gasification Combined Cycle ILCB see integrated lignocellulosic biorefinery indirect LUC, 371, 386 Infrared Spectroscopy for Food Quality Analysis and Control, 324 inorganic adsorbents, 258±9 INSEA see Integrated Sink Enhancement Assessment Integrated Biomass Utilisation System, 8 Integrated Gasification Combined Cycle, 400 integrated hydrolysis, 205±19 integrated lignocellulosic biorefinery, 247±8 Integrated Microbial Genome, 189 Integrated Sink Enhancement Assessment, 386 International Organisation for Standardisation, 366 inulin, 435 IOGEN batch method, 6 IOGEN Corporation, 6 IOGEN technology, 370 ion-exchange chromatography, 294 ionic liquids, 34±5, 254±5 as non-volatile entrainer, 254 Irpex lecteus, 80 ISO see International Organisation for Standardisation iso-octyl alcohol, 258 isoamyl acetate, 258 isopropanol, 322 advances in process technology, 434±40 advanced fermentation and down stream products processing, 438 alternative fermentation substrates, 434±8 gas-stripping, 440 pervaporation, 438±40 advances in production, 429±40 isopropanol-producing strains improvement, 429±32

production by species other than clostridia, 432±4 characteristics and uses, 419 production by clostridia, 420±9 microorganisms and metabolism, 420±4 production from lignocellulosic biomass, 415±48 isopropanol, butanol and ethanol fermentation, 416, 419, 447 2-isopropylmalate synthase, 433 Izumi Inc., 155 Jerusalem artichoke plant, 435 JGC Corp., 155 Joint Research Centre model, 385 Kalman filter, 332 ketones, 406 2-ketovalerate, 433 Klason lignin, 15 Klebsiella, 419, 445 Klebsiella oxytoca, 218, 403, 419, 444 Klebsiella pneumoniae, 442, 444 Kluyveromyces, 214 Kluyveromyces fragilis, 214 Kluyveromyces marxianus, 355 Kluyveromyces marxianus Y01070, 214 laccases, 40 lactic acid, 404±5, 410 Lactobacillus, 404 Lactococcus lactis, 430 land use change, 371 Langmuir adsorption, 126 Langmuir isotherm equation, 74, 76, 93 Langmuir parameters enzyme/protein adsorption on lignin, 94±5 lignocellulosic substrates for various enzymes and proteins, 77±9 LCC linkages see lignin-carbohydrate complex linkages LCFS see Low Carbon Fuel Standard LEM see Life Cycle Emissions Model levulinic acid, 297, 405 life cycle assessment, 56, 58, 366 biochemical lignocellulosic alcohol production, 372±7 biofuels, 367±72 lignocellulosic alcohol biofuel vs other fuels, 377±81 lignocellulosic bioalcohol biofuel vs alternative biomass utilisation, 381±3 models, 356 Life Cycle Emissions Model, 385 ligand-exchange chromatography, 294

ß Woodhead Publishing Limited, 2010

Index lignin, 3, 10, 15, 27, 127, 156, 162, 212, 283, 316, 317, 319, 392 controlling hydrolysis, 91±7 accessibility, 91±6 effectiveness, 96±7 Langmuir parameters for enzyme/protein adsorption, 94±5 guaiacyl alcohol residue representative degradation products, 298 primary residues, 297 relative polymer degradation, 13 lignin-carbohydrate complex linkages, 28, 30, 59, 97, 99 lignocelluloses, 410±11 bioethanol production scheme, 284 biological processing, 206±9 challenges in derived sugars cofermentation, 224±39 combining recombinant pathways, 230±2 future outlook, 238±9 transcriptional regulation in mixed substrate fermentation, 232±8 transporter preferences, 225±30 degradation products analysis, 296±9 alternative techniques, 300±1 gas chromatography, 297±8 HPLC, 298±9 feedstocks, 392±3 definition, 392 quantity, 392 waste and crops, 392±9 lignocelluloses-to-bioalcohol production environmental life cycle assessment, 365±87 biochemical production, 372±7 biofuels LCA, 367±72 future trends, 385±7 lignocellulosic alcohol biofuel vs other fuels LCA, 377±81 lignocellulosic bioalcohols vs alternative biomass utilisation, 381±3 routes for environmental improvement, 383±5 extractive distillation, 250±5 conventional liquid solvent and dissolved salt mixture, 253 conventional solvent, 250±1 dissolved salt, 251±3 ethanol dehydration typical process, 250 gasoline for ethanol dehydration, 255 hyper-branched polymers, 253 ionic liquid, 254±5 schematics of dissolved salt, 252 separating agents comparison, 255 fermentation processes online monitoring, 315±34

469

chromatographic techniques, 322±4 potential monitoring techniques, 319±22 software sensors, 330±3 spectroscopic methods, 324±30 variables of interest to monitor, 318±19 hydrolysis and fermentation processes modelling, 340±59 environmental issues, 355±6 fermentation by various microorganisms, 350±4 future trends, 357±8 saccharification by chemical/enzymatic processes, 342±8 SSF, 354±5 successful examples, 356±7 membrane separation, 261±8 complete membrane pervaporationbioreactor hybrid process, 268 hydrophilic pervaporation membranes, 262±6 hydrophobic pervaporation membranes, 266 membrane pervaporation-bioreactor hybrid, 267±8 vacuum membrane distillation, 268±9 pervaporation performance hydrophilic inorganic membranes, 265 hydrophilic polymeric membranes, 264 hydrophobic membranes, 267 pre-treatment and hydrolysis processes analytical monitoring, 281± 301 alternative analysis techniques, 300±1 biomass hydrolysates preparation, 287±8 bioprocessing analytical perspective, 281±2 carbohydrates analysis, 288±96 current state-of-the-art online vs off-line monitoring techniques, 283 detection strategies, 285±7 future trends, 301 lignocellulosic degradation products analysis, 296±9 resulting target analytes, 283±5 reverse-phase separation selected retention window aliphatic acids, aldehydes, and phenolic compounds in standard solution, 286 degradation products from cornstover hydrolysate, 286 separation and purification processes, 246±69 azeotropic distillation, 249±50 basic lignocellulosic biomass-to-ethanol biorefinery, 246±7 continuous fermentation with in situ extraction, 257

ß Woodhead Publishing Limited, 2010

470

Index

ethanol dehydration AD system flowsheet, 249 extractive distillation agents, 256 extractive fermentation, 255±8 integrated forest biorefinery process block diagram, 248 integrated lignocellulosic biorefinery, 247±8 lignocellulosic biorefineries, 248±9 process block diagram, 247 separation by adsorption, 258±61 advantages and disadvantages, 261 water liquid-phase adsorption, 260±1 water vapor-phase adsorption, 258±60 lignocellulosic alcohol biochemical production LCA, 372±7 comparison with alternative biomass utilisation, 381±3 GHG emissions avoided with alternative uses of biomass, 382 vs other fuels LCAs, 377±81 lignocellulosic biomass, 224, 246±7 acid pre-treatment process and apparatus, 145±52 batch reactor, 145±6 co-current pre-treatment method, 147 co-current reactor, 145 COSLIF with concentrated phosphoric acid and acetone recycling flowchart, 146 Dionex ASE 350 accelerated solvent extractor, 148 flow-through pre-treatment method, 149 flow-through reactor, 149 percolation reactor system laboratory set-up, 148 percolation reactors, 146±7 pre-treatment using a solvent extractor, 147 cellulases development to improve enzymatic hydrolysis, 178±94 cellulase structure and function, 179±87 cellulases development, 187±91 future trends, 193±4 issues, 192±3 recent developments, 191±2 cellulose solvent-based lignocellulose pre-treatment, 128±35 cellulose solvent-only, 129 corn stover cell wall structures and cellulose fibres, 135 COSLIF, 129±34 COSLIF effects vs dilute acid on biomass composition, 131 COSLIF flowchart, 130

hydrolysis curves for different feedstocks, 132 industrial hemp hurds, 134 mass balance for switchgrass via COSLIF and enzymatic hydrolysis, 133 supramolecular structures, 134±5 chemical production, 391±411 carboxylate platform, 405±10 feedstocks, 392±3 sugar platform, 400±5 thermochemical platform, 393±400 commercial plants in USA, 357 concentrated acid hydrolysis, 144±5 dilute acid hydrolysis, 143±4 flow-through acid pre-treatment, 144 dilute acid hydrolysis kinetics, 345±8 acid cellulose hydrolysis, 345±7 enzymatic hydrolysis, 347±8 hemicellulose acid hydrolysis, 347 dilute and concentrated acid hydrolysis, 143±56 future trends, 155±6 environmental issues, 355±6 enzymatic hydrolysis, 159±71 factors affecting hydrolysis efficiency, 164±8 future trends, 170±1 improvements, 168±70 mechanism, 160±2 relative saccharification efficiencies, 163±4 ethanol production plants using acid hydrolysis Arkenol Inc., 149±52 conversion of cellulose/hemicellulose to mixed sugars, 150 future trends, 357±8 hydrolysis process, 342±5 hydrothermal pre-treatment, 3±18 future trends, 17±18 liquid hot water and steam, 5±16 physical comminution, 4 integrated hydrolysis, fermentation and co-fermentation, 205±19 consolidated bioprocessing scheme, 209 feedstocks and pre-treatment for SSF/ CBP, 210±13 future trends, 218±19 lignocellulose biological processing, 206±9 microbial strains for SSF/CBP, 213±18 pre-treatment methods on chemical/ physical structure, 211 process configurations in enzyme-based ethanol production, 207 xylose utilisation metabolic pathway, 216

ß Woodhead Publishing Limited, 2010

Index key features and impact on hydrolysis, 73±101 adsorption capacity for cellulose components, 82 biomass structure conceptual model, 99 cellullase accessibility to cellulose and cellulase effectiveness with impact ranking, 98 cellulose crystallinity on maximum cellulose adsorption capacity, 80, 81 crystallinity, 75±84 crystallinity vs cellulose viscosity degree of polymerisation, 87 degree of polymerisation, 84±8 equilibrium constants for exo and endocellulase, 81 hemicellulose and degree of hemicellulose acetylation, 88±91 Langmuir parameters for lignocellulosic substrates, 77±9, 94±5 lignin, 91±7 pre-treatment effectiveness schematic decision tree, 100 pre-treatment severity on degree of polymerisation, 86 lignocellulosic hydrolysis kinetics, 346 longer-chain alcohols production, 415±48 pathways corn stover, 373 farmed tree, 373 herbaceous biomass, 373 phase classification, 59 pre-treatment, 342 fuel ethanol production methods, 343±4 hydrolysis streams detoxification methods, 352 representative species composition, 341 research trends and priorities for fuel ethanol improvement, 359 saccharification by chemical/enzymatic processes, 342±8 solvent fractionation of solid and liquid components, 122±36 cellulose and enzymatic cellulose hydrolysis, 124±6 future trends, 135±6 hemicellulose, 126±7 lignin, 127 successful examples, 356±7 Brazil, 357 China, 356 Europe, 356 USA, 356 thermochemical pre-treatment, 24±59 comparing effectiveness on corn stover and poplar, 41±7

471

ideal characteristics, 47±58 types, 32±41 why pre-treatment is necessary, 25±32 unit operations pertinent to the ethanol industry, 152±5 feedstock preparation, 152 minimising microbe inhibiting compounds, 153±4 sugar/acid separation technologies, 155±6 lignocellulosic biorefinery, 50 lignocellulosic hydrolysates, 351 lignocellulosic substrates, 316, 436 Lignol Innovation Corporation, 34 lime pre-treatment, 44 lime solubilisation, 212 liquid chromatography, 321 liquid hot water, 7±8, 13±14, 16, 211, 212 vs steam explosion, 17 liquid hot water pre-treatment, 8, 37 longer-chain alcohols advances in 2,3-butanediol production, 442±5 advances in butanol and isopropanol production, 429±40 integrated fermentation processes for solvent production, 439 process technology advances, 434±40 producing strains improvement, 429±32 production by species other than clostridia, 432±4 biological production methods, 441 characteristics and uses, 416±20 future trends, 447±8 other long chain alcohols that can be produced biologically, 445±7 2-butanol, 446±7 1,2-propanediol and 1,3-propanediol, 445±6 physical and chemical properties, 417 production by clostridia, 420±9 biochemical pathways in C. acetobutylicum and C. beijerinckii, 421 microorganisms and metabolism, 420±4 process technology, 424±9 solvent production during molasses or starchy substrates large-scale fermentation, 425 production from lignocellulosic biomass, 415±48 vs gasoline properties as fuels, 418 Low Carbon Fuel Standard, 366 low wavelength ultraviolet detection, 285 LUC see land use change lysogenic, 428

ß Woodhead Publishing Limited, 2010

472

Index

MALDI see matrix assisted laser desorption ionisation maleic anhydride, 262 mannose, 226 Masonite process, 5, 6 mass separating agent, 249 mass spectrometry, 285, 292, 299 matrix assisted laser desorption ionisation, 296 matrix effects, 299 membrane pervaporation, 261 membrane separation, 261±8 metabolic engineering, 434 metabolic switch, 422 metagenomics, 189 Methanococcus jannaschii, 433 methanogens, 406 methanol, 399 methyl-ethyl ketone, 446 methyl tert-butyl ether, 288, 399, 419 4-methyl-umbelliferyl- -D-cellobioside, 190 4-methyl-umbelliferyl- -D-pyranoside, 190 methylene chloride, 288 methyltetrahydrofuran, 405 Microbial Genome Program, 189 Microbiome informatics, 189 milling, 4 mineral acids, 33, 36 MixAlco process, 407±8, 410 flow diagram, 407 Mn-dependent peroxidases, 40 mobibdate, 300 molasses, 328, 426 molecular sieve dehydration process, 259 molecular sieve dehydration unit, 259 monophenol oxidase, 40 mordenite, 263 moving-bed gasifiers, 397 moving window principal component analysis, 332 MSW see municipal solid wastes MTBE see methyl tert-butyl ether multi-ply homogeneous membranes, 265 multi-wavelength fluorescence, 325, 328 multiple-hearth gasifiers, 397 municipal solid wastes, 210 mushroom spent straw, 41 MW-PCA see moving window principal component analysis N-ethyl-pyridinium chloride, 35 naphthalene, 410 National Chemical Products, 424 National Development and Reform Commission, 356 National Renewable Energy Laboratory, 160, 215, 370

technology, 370, 373 native cellulolytic strategy, 209, 217 Natural Resources Canada, 385 NDRC see National Development and Reform Commission Near-Infrared Technology in the Agricultural and Food Industries, 324 negative selection marker, 430 nitrogen, 291, 355 NREL see National Renewable Energy Laboratory octanol-water partition coefficient, 446 octylsulphate, 255 ODT see oven dry tonne Oil Companies' European Association for Environment, 369 oils, 398 olefins, 399, 400, 406 oleyl alcohol, 257, 258 oligosaccharides, 126 online monitoring chromatographic techniques, 322±4 gas chromatography, 322±3 liquid chromatography, 323±4 fermentation processes in lignocelluloses-tobioalcohol production, 315±34 potential monitoring techniques, 319±22 software sensors, 330±3 advantages, 331 data-driven modelling methods, 332±3 hybrid software sensors, 331 stoichiometry, elementary balances and first-principles models, 330±2 spectroscopic techniques, 324±30 applications in fermentation processes, 326±30 basic principles, 324±5 data analysis, 325±6 variables of interest to monitor in bioethanol production processes, 318±19 Opisthocomus hoazin, 189 optical density, 318, 329 organosolv process, 33±4 oven dry tonne, 382 oxidants, 402 oxidative pre-treatment, 39±40 oxygen gasification, 396 oxygen transfer rate, 442 ozone, 39 ozonolysis, 39 P-factor, 11 Pachysolen tannophilus, 238 pacific shipworm see Bankia setacea PAD see pulsed amperometric detection

ß Woodhead Publishing Limited, 2010

Index para-hydroxyphenol, 297 Parr high pressure reactor, 145 partial least squares, 326, 332 models, 327, 328 PCA see principal component analysis pentose, 226±7, 284, 353±4 pentose phosphate pathway, 422 Peoria Process, 155 percolation reactors, 146±7 laboratory set-up, 148 pervaporation, 261, 264, 438±40, 445 pesticides, 355 petrochemical process, 418 petroleum gasoline, 377 petroleum refining, 419 phage CA1, 427 phage control, 407 Phanerochaete chrysosporium, 41, 184, 185, 189 phase separation, 398 phenanthrene, 410 phenolic acid esterases, 31 phenols, 410 phenylalanine, 325 phosphate, 426 phosphoric acid, 145 photo-acoustic spectroscopy, 329 Pichia stipitis, 213, 229, 403 Plant Genome Program, 189 plasmid transfer, 407 PLRP-S reversed phase column, 323 PLS see partial least squares poly(1-trimethylsilyl-1-propyne), 266 poly(acrylic acid), 253 polycarbonates, 404 poly(dimethyl siloxane), 266 polyelectrolytes, 262 polyesters, 404 poly(ethylene glycol), 253 polyglycerol, 253 polyhydroxyalkanoates, 410 polyimides, 262, 263 poly(lactic) acid, 404±5 polymerisation, 398 polymers cellulose, 316 polypropylene, 405, 410 polystyrene, 405 polystyrenesulfonate-alumina, 265 polysulfone, 262, 263 polyurethane filler, 410 polyurethanes, 404 poly(vinyl alcohol), 262 poplar, 41±7, 93 effect of leading thermochemical pre-treatments, 45±6 poplar wood, 410

473

Postia placenta, 185, 189 potassium acetate, 251 pre-treatment processes, 9 analytical monitoring in lignocellulose-tobioalcohol production, 281± 301 biomass hydrolysates preparation, 287±8 bioprocessing analytical perspective, 281±2 carbohydrates analysis, 288±96 carbohydrates and degradation products alternative analysis techniques, 300±1 detection strategies, 285±7 future trends, 301 lignocellulosic degradation products analysis, 296±9 resulting target analytes, 283±5 principal component analysis, 326 producer gas, 398 1,2-propanediol, 445±6 1,3-propanediol, 445±6 propanoic acid, 410 propylene glycol, 404 propylene oxide, 404 protein engineering, 189±90 protein kinase A, 235 pseudolignin, 15 Pseudomonas, 419 pSOL1 plasmid, 423 PTFE membrane probe, 322 pulsed amperometric detection, 285, 295 pyrolysis, 396 tars and oils, 398 quorum sensing, 424 RAC see regenerated amorphous cellulose Raman spectroscopy, 324, 329 random mutagenesis, 429 recombinant cellulolytic strategy, 209 refractive index, 295 refractive index detection, 285 refractometric detection, 300 regenerated amorphous cellulose, 126, 129 Renewable Transportation Fuel Obligation Programme, 366, 385 respiratory quotient, 331 reverse phase liquid chromatography, 294, 299 reversed osmosis, 445 Rhizopus, 404 Rhodococcus rubur, 447 RMSEP see root mean square error of prediction root mean square error of prediction, 326, 327 rot fungus, 171 rotary-kiln gasifiers, 397

ß Woodhead Publishing Limited, 2010

474

Index

rotational spectroscopy, 324 RPLC see reverse phase liquid chromatography RTFO Programme see Renewable Transportation Fuel Obligation Programme saccharification, 163±4 lignocellulosic biomass by chemical/ enzymatic processes, 342±8 saccharification and fermentation process, 316, 319, 333 Saccharomyces, 402 Saccharomyces cerevisiae, 12, 37, 193, 208, 213±17, 225, 296, 331±2, 351 engineering regulatory response, 235±7 expression of Gxf1 transporter, 229 glucose signalling scheme, 234 heterologous xylose transporters, 228 physiological response to altered glucose signalling, 233 Saccharomyces cerevisiae LNH-ST 424A, 47 Saccharomyces cerevisiae TMB 3400, 217 Saeman's model, 347 2nd generation fuels, 368, 377 separate hydrolysis and fermentation process, 169, 206±7, 316, 319, 351, 353, 354 Serratia, 419 severity factor, 11 SHF see separate hydrolysis and fermentation process Sigmacell, 35 silicon, 399 simultaneous saccharification and cofermentation, 206, 208 microbial strains, 214±17 simultaneous saccharification and extractive fermentation, 257 simultaneous saccharification and fermentation, 169, 189, 193, 206, 354±5, 435 feedstocks and pre-treatment, 210±13 future trends, 218±19 microbial strains, 213±14 single-pass harvesting techniques, 392 siropulping, 6 slopback, 427 sodium acetate, 251 sodium alginate, 262 sodium carboxymethyl cellulose, 263 softwood, 210, 354 solid phase extraction, 288 solid state fermentation, 40±1 solvent, 249, 402 solvent fermentation, 420, 427 solvent fractionation, 33±5 ionic liquid-based, 34±5 organosolv process, 33±4

phosphoric acid, 34 solid and liquid components from pretreated lignocellulosic biomass, 122±36 cellulose and enzymatic cellulose hydrolysis, 124±6 cellulose solvent-based lignocellulose pretreatment, 128±35 future trends, 135±6 hemicellulose, 126±7 lignin, 127 solventogenesis, 423 solventogenic clostridia, 435 soybean diesel pathways, 381 SPE see solid phase extraction spectroscopy, 282 Spezyme, 166 Spo0A, 423 sporulation, 423 SSEF see simultaneous saccharification and extractive fermentation SSF see simultaneous saccharification and fermentation SSFC see simultaneous saccharification and co-fermentation STAKE continuous process, 6 STAKE Technology, 6 stannate, 300 starchy biomass adsorbents, 259±60 steam explosion, 5±7, 11±13, 14±15, 16, 37, 211, 212 pre-treatment conditions and results for feedstocks, 7 vs liquid hot water, 17 stirring, 330 succinic acid, 403±4 sucrose, 316 sugar platform chemistry and secondary chemicals, 403±5 lactic acid, 404±5 levulinic acid, 405 secondary products from hexoses, 404 succinic acid, 403±4 feedstock considerations, 400±2 lignocellulosic biomass chemical production, 400±5 pre-treatment methods chemical, 402 physical, 402 physicochemical, 402 primary products, 403 process basics, 402±3 research and development, 405 sugarcane, 400 sulphite, 154 sulphur dioxide, 12 sulphuric acid, 12, 143, 402

ß Woodhead Publishing Limited, 2010

Index sunchoke, 435 supercellulase, 171 supercritical ammonia treatment, 38 superoxide dismutase, 40 superphosphate, 426 switchgrass, 133, 134 synchronisation, 318 syngas, 398, 410 synthesis gas see syngas syringaldehyde, 153 syringic acid, 287 syringyl, 297 tank-to-wheel component, 367 tar removal, 396 TargeTron, 430 tars, 398 TCD see thermal conductivity detection termites, 40 Terrabon L.L.C., 410 Tetravitae, 447 TFA see trifluoroacetyl The Prokaryotes, 445 thermal conductivity detection, 285 Thermoanaerobacter thermosaccharolyticum, 353 Thermoanaerobacterium thermosaccharolyticum, 217±18 Thermoanaerobium thermosaccharolyticum, 445 thermochemical platform chemistry and secondary chemicals, 399±400 companies that are likely to have commercial-scale biomass facilities, 401 current chemicals production, 400 feedstock considerations, 393±4 lignocellulosic biomass chemical production, 393±400 potential chemicals from syngas and some required catalysts, 399 primary products, 397±8 adiabatic air/biomass reaction equilibrium composition, 397 char, 398 pyrolysis tars and oils, 398 syngas and producer gas, 398 water, 398 process basics, 394±7 chemical changes during biomass processing, 395 gasification, 396±7 gasification and pyrolysis pathways, 395 pyrolysis, 396 research and development, 400

475

thermochemical pre-treatment comparing effectiveness on corn stover and poplar, 41±7 effect at the end of pre-treatment and enzymatic hydrolysis, 45±6 optimum pre-treatment conditions, 42 prominent physicochemical effects, 43 detailed mass balance for generic unit operations, 50 ideal characteristics, 47±58 detailed mass balance, 50 economic and environmental feasibility, 58 enzymatic hydrolysis-fermentation related parameter ranges, 54±5 factors affecting viability, 47±56 interdependence with other operations and process-product economic viability, 51 mass and energy balances, 56±7 material and energy flow balance, 49 pretreatment-related parameters, 52±3 product yield, reaction rate, product concentration, 57 lignocellulosic biomass, 24±59 types, 32±41 biological, 40±1 chemical, 35±40 physical, 33 solvent fractionation-based, 33±5 why necessary for lignocellulosics, 25±32 effect on cell wall composition and ultrastructure, 28±30 glycosyl hydrolases necessary for saccharification, 30±2 native plant cell wall recalcitrance, 26±7 untreated grass cell-wall structural components, 28 thionyl chloride, 410 thorium oxide catalyst, 409 threonine biosynthesis, 433 tin alkoxide, 410 TMS see trimethylsilyl Trametes versicolor, 154 Trichoderma enzymes, 31, 32 Trichoderma harzanium, 153 Trichoderma harzianum, 93 Trichoderma reesei, 81, 83, 84, 154, 160, 166, 167, 171, 179, 181, 184, 185, 187, 189±90, 217, 347, 438 major secreted proteins, 180 time-dependence of biomass saccharification by cellulases, 186 trifluoroacetyl, 292 trimethylsilyl, 292 Tripp, 381 tryptophan, 325

ß Woodhead Publishing Limited, 2010

476

Index

TTW component see tank-to-wheel component Tween-20, 354 tyrosisne, 325 UAS see upstream activating sequences ultrafiltration, 170, 322 ultrahigh performance liquid chromatography, 324 uncatalysed solvolysis, 8 UOP process, 377 upstream activating sequences, 235 urea, 400, 444 US Department of Energy Argonne National Laboratory, 372±3, 385 Joint Genome Institute, 189 US Geological Survey Soil Carbon Research, 386 vacuum membrane distillation, 268±9 vanillin, 153 vapour-liquid equilibrium, 253, 255 volatile solid loading rates, 407 Warburg-Dickens pathway, 422 waste materials, 392 waste solids, 410 water, 398 water-gas shift reaction, 394, 398 water vapor-phase adsorption, 258±60 bio-based adsorbents, 259±60 inorganic adsorbents, 258±9 Waters Durapore filter, 323 Weizmann strain, 424 well-to-ethanol plant exit gate, 373 GHG emissions associated with lignocellulosic ethanol production cornstover through biochemical conversion, 376 farmed trees through biochemical conversion, 374 herbaceous through biochemical conversion, 375 well-to-tank component, 367

well-to-wheel analysis, 367 GHG emissions associated with lignocellulosic ethanol and alternative fuel pathways, 378, 379 wheat straw, 92, 163 wheat straw hydrolysate, 437 white biotechnology, 448 white-rot fungi, 40 Williamson ether synthesis process, 409 wood hydrolysates, 444 WSH see wheat straw hydrolysate WTG see well-to-ethanol plant exit gate WTT component see well-to-tank component WTW analysis see well-to-wheel analysis xylan, 98 xylanases, 32, 47 xylitol dehydrogenase, 216, 230 xylose, 12, 126±7, 226, 445 fermentation Crabtree-negative characteristics, 238 effect of glucose concentration, 238 heterologous transporters expressed in Saccharomyces cerevisiae, 228 metabolic pathway for utilisation, 216 recovery, 12 xylose isomerase, 216, 230 xylose reductase, 216, 230, 231 xylulokinase, 216, 237 YLR042C, 237 YPL230w, 237 ZeaChem Inc, 410 ZeaChem technology, 408 zeolite A, 260, 263 zeolite catalysts, 399, 409 zeolite membrane, 438 zeolite X, 263 zeolite Y, 263 zeolites molecular sieves, 261 zinc alkoxide, 410 Zymomonas mobilis, 208, 215, 218, 323, 403

ß Woodhead Publishing Limited, 2010