Separation of Functional Molecules in Food by Membrane Technology 9780128150566, 2052062092, 2292302322, 3053053063, 0128150564

Citation preview

Separation of Functional Molecules in Food by Membrane Technology

This page intentionally left blank

Separation of Functional Molecules in Food by Membrane Technology Edited by

Charis M. Galanakis Department of Research & Innovation, Galanakis Laboratories, Chania, Greece Food Waste Recovery Group, ISEKI Food Association, Vienna, Austria

Academic Press is an imprint of Elsevier 125 London Wall, London EC2Y 5AS, United Kingdom 525 B Street, Suite 1650, San Diego, CA 92101, United States 50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom Copyright r 2019 Elsevier Inc. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions. This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein). Notices Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary. Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility. To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein. 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 ISBN: 978-0-12-815056-6 For Information on all Academic Press publications visit our website at https://www.elsevier.com/books-and-journals

Publisher: Andre Gerhard Wolff Acquisition Editor: Nina Rosa Bandeira Editorial Project Manager: Katerina Zaliva Production Project Manager: Omer Mukthar Cover Designer: Matthew Limbert Typeset by MPS Limited, Chennai, India

Contents List of Contributors Preface

1.

Introduction in Membrane Technologies

xi xv 1

Maria Norberta de Pinho and Miguel Minhalma

2.

1 3 3 3 6 8 8 8 9 10 12 12

1.1 Introduction 1.2 Materials and Structures of Membranes 1.2.1 Materials 1.2.2 Structures 1.3 Classification of Membrane Processes 1.4 Membrane Modules 1.4.1 Plate-and-Frame Modules 1.4.2 Tubular Modules 1.4.3 Hollow Fiber Modules 1.4.4 Spiral Wound Modules 1.5 Pressure-Driven Membrane Processes 1.5.1 Membrane Characterization 1.5.2 Mass Transfer at the Fluid Phase Circulating Tangentially to the Membrane: Concentration Polarization 1.5.3 Operation Modes 1.5.4 Case Study Applications 1.6 Electrodialysis 1.6.1 Characterization of Ion-Exchange Membranes 1.6.2 Process Operation 1.7 Pervaporation 1.8 Conclusion References

15 17 19 21 22 23 24 27 27

Introduction in Functional Components for Membrane Separations

31

˘ s and Charis M. Galanakis Sonia A. Socaci, Anca C. Farca¸ 2.1 Introduction 2.2 Challenges in Functional Food Development

31 32

v

vi

Contents

2.3 Recovery of Bioactive Compounds From Conventional and Nonconventional Sources 2.3.1 Proteins and Active Peptides 2.3.2 Polyphenols 2.3.3 Polysaccharides 2.3.4 Lipids 2.3.5 Bioactive Compounds of Animal Origin 2.4 Separation and Recovery of Macro- and Micromolecules Using Membrane Technologies 2.5 Conclusions References Further Reading

3.

Membrane Filtration of Biosurfactants

33 34 42 45 48 49 52 68 68 77 79

Paula Jauregi and Konstantina Kourmentza

4.

79 79 84 87 90 91 92

3.1 Introduction 3.1.1 Biosurfactants 3.1.2 Production of Biosurfactants 3.2 Downstream Processing of Biosurfactants 3.2.1 The Ultrafiltration Process and Equipment 3.2.2 Assessment of Separation Performance 3.3 Ultrafiltration of Lipopeptide Biosurfactants 3.3.1 Surfactin Separation by the Two-Step Ultrafiltration Method 3.3.2 Hybrid Recovery Processes Using Ultrafiltration 3.3.3 Separation of Lipopeptide Mixtures 3.3.4 Recovery of Lipopeptides From Complex Culture Medium 3.4 Ultrafiltration of Rhamnolipid Biosurfactants 3.5 Ultrafiltration of Mannosylerythritol Lipids 3.6 Further Considerations 3.6.1 Membrane Choice 3.6.2 Membrane Cleaning 3.7 Conclusions and Final Outlook References Further Reading

95 97 98 100 100 101 102 103 112

Membrane Technology for the Purification of Enzymatically Produced Oligosaccharides

113

92 94 95

Andre´s Co´rdova, Carolina Astudillo and Andre´s Illanes 4.1 Introduction to Oligosaccharide Prebiotics 4.2 Synthesis of Oligosaccharides in Membrane Bioreactors 4.3 Strategies and Mechanisms Involved in the Nanofiltration of Oligosaccharides 4.3.1 Retention Due to Concentration Polarization 4.3.2 Effect of Temperature

113 118 122 122 124

Contents

5.

vii

4.3.3 Effect of Transmembrane Pressure 4.3.4 Effect of Solute Concentration and Feed Composition 4.4 Current Status of the Purification of Specific Oligosaccharides by Membrane Technology 4.4.1 Purification of Xylooligosaccharides 4.4.2 Purification of Isomaltooligosaccharides 4.4.3 Purification of Galactooligosaccharide and Other Lactose-Derived Prebiotics 4.4.4 Purification of Fructooligosaccharides 4.4.5 Purification of Other Oligosaccharides 4.5 Challenges and Perspectives 4.6 Conclusions References Further Reading

128 131

Pectin Removal and Clarification of Juices

155

133 133 137 139 143 144 145 146 147 153

Sankha Karmakar and Sirshendu De 5.1 Introduction 5.2 Various Processing or Preservation Techniques Used for Fruit and Vegetable Juice 5.2.1 Depectinization of Fruit Juices 5.3 Other Treatment Methods for Fruit Juice Clarification/Depectinization 5.3.1 Membrane 5.3.2 Membrane Classification 5.3.3 Membrane Modules 5.4 Membrane Based Separation Processes 5.5 Depectinization and Membrane Based Clarification of Some Typical Fruit and Vegetable Juices 5.5.1 Enzymatic Depectinization of Juices 5.5.2 Membrane Based Clarification Process for Some Typical Fruit and Vegetable Juice 5.6 Conclusion References

6.

Recovery of Phenolic-Based Compounds From Agro-Food Wastewaters Through Pressure-Driven Membrane Technologies

155 157 163 163 164 166 166 168 168 168 172 182 184

195

Roberto Castro-Mun˜oz, Carmela Conidi and Alfredo Cassano Abbreviations 6.1 Introduction 6.2 Membrane-Based Technologies as Emerging Tools for Food Wastewaters Valorization 6.3 Recovery of Phenolic Compounds From Agro-Food Wastewaters 6.3.1 Olive Mill Wastewaters

195 196 197 199 199

viii

7.

Contents

6.3.2 Artichoke Wastewaters 6.3.3 Wastewaters From Winemaking Industry 6.3.4 Other Food Wastewaters and By-Products 6.4 Current Uses of Phenolic-Based Compounds Extracted From Food Wastewaters 6.5 Economic Overview of Membrane-Based Technologies in Phenolic Recovery 6.6 Conclusions and Future Trends References

205 206 209

Lignin Separation and Fractionation by Ultrafiltration

229

214 220 222 222

Javier Ferna´ndez-Rodrı´guez, Xabier Erdocia, Fabio Herna´ndez-Ramos, Marı´a Gonza´lez Alriols and Jalel Labidi

8.

7.1 Introduction 7.2 Lignin Chemistry 7.3 Lignin Isolation 7.3.1 Sulfite Process 7.3.2 Kraft Process 7.3.3 Alkaline Treatments 7.3.4 Organosolv Process 7.4 Lignin Application in the Food Industry 7.4.1 Lignin as Part of Dietary Fiber 7.4.2 Lignin as Feed Additive 7.4.3 Antimicrobial Effects 7.4.4 Prebiotic Effects of Lignin and Weight Gain 7.4.5 Packaging and Films 7.5 Separation of Lignin by Ultrafiltration 7.5.1 Kraft Lignin Purification 7.5.2 Lignosulfonates Purification 7.5.3 Sulfur-Free Lignin Fractionation: Alkaline and Organosolv Lignin 7.5.4 Hydrolysis Lignin Recovery 7.5.5 Membrane Flux Decay in Lignin Separation Process 7.5.6 Ultrafiltration Purification as Source for Food Industry 7.6 Conclusions References Further Reading

229 230 232 232 233 234 235 236 237 238 238 239 240 241 245 247

Membrane Separations in the Dairy Industry

267

248 249 250 253 257 258 265

George Q. Chen, Thomas S.H. Leong, Sandra E. Kentish, Muthupandian Ashokkumar and Gregory J.O. Martin 8.1 Introduction 8.2 Milk and Whey 8.2.1 Milk Components and Composition

267 270 270

Contents

8.2.2 Milk Proteins 8.2.3 Casein Micelles 8.3 Milk Processing 8.3.1 On Farm Concentration 8.3.2 Control of Microbial Growth 8.3.3 Milk Protein Fractionation 8.3.4 Milk Fat Fractionation 8.4 Cheese Processing 8.4.1 Cheese Milk Standardization 8.5 Whey Processing 8.5.1 Whey Protein Concentration 8.5.2 Whey Protein Fractionation 8.5.3 Whey Concentration 8.5.4 Whey Demineralization 8.5.5 Cheese Brine Purification 8.6 Waste Treatment 8.7 Fouling 8.8 Application of Sonication 8.8.1 Physical and Chemical Effects of Ultrasound 8.8.2 Ultrasonic Reduction of Fouling Buildup 8.8.3 Mechanisms of Ultrasonic Membrane Cleaning in Dairy Applications 8.8.4 Application of Ultrasound to Membrane Cleaning 8.9 Conclusions and Future Trends References

9.

Current and Future Applications of Nanofiltration in Food Processing

ix 272 272 273 273 273 275 278 278 279 280 280 281 282 283 284 285 287 289 289 290 291 292 295 295

305

Alfredo Cassano, Carmela Conidi and Roberto Castro-Mun˜oz Abbreviations 9.1 Introduction 9.2 General Aspects of NF Membranes 9.3 Fruit Juice Processing 9.4 Wine and Must Processing 9.5 Dairy Industry 9.5.1 Concentration and Demineralization of Whey 9.5.2 Concentration and Demineralization of UF Whey Permeate 9.5.3 Recovery of Lactic Acid 9.6 Sugar Industry 9.7 Recovery of Functional Compounds From Food Processing Byproducts 9.8 Conclusions and Future Trends References Further Reading

305 306 307 310 314 319 319 320 322 325 329 342 342 348

x

Contents

10. Electrodialysis-Based Separation Technologies in the Food Industry

349

Yaoming Wang, Chenxiao Jiang, Laurent Bazinet and Tongwen Xu 10.1 General Introduction of Electrodialysis and Electrodialysis With Bipolar Membrane 10.2 Electrodialysis in Food Processing 10.2.1 Whey 10.2.2 Fruit Juice 10.2.3 Wine 10.2.4 Sauce 10.2.5 Sugars 10.2.6 Amino Acids 10.3 Electrodialysis With Bipolar Membrane Applications in Food Processing 10.3.1 Applications to Fruit Juices 10.3.2 Applications to Proteins 10.3.3 Application to Phospholipids 10.4 Electrodialysis With Filtration Membrane Applications in Food Processing 10.4.1 Applications to Protein Hydrolysates 10.4.2 Separation of Protein 10.4.3 Separation of Chitosan Oligomers 10.4.4 Concentration of Cranberry Juice in Phenolic Compounds 10.5 Conclusion Acknowledgments References Further Reading

11. Osmotic Driven Membrane Processes for Separation of Special Food Compounds

350 351 351 353 354 355 356 357 359 359 361 364 366 367 370 371 372 373 374 374 381

383

´ Claus He´lix-Nielsen, Katalin Belafi-Bako, Irena Petrinic, Guofei Sun, Ye Wee Siew, Simon Alvisse, Nguyen Xuan Tung, Andras Boor and Nandor Nemestothy 11.1 Introduction 11.2 Forward Osmosis for Up-Concentrating Food Compounds 11.3 Membrane Osmotic Distillation 11.3.1 Preservation of Valuable Compounds 11.3.2 Sensory Evaluation 11.4 Draw Solutions 11.5 Conclusions Acknowledgments References Further Reading Index

383 386 389 392 394 396 398 398 398 401 403

List of Contributors Marı´a Gonza´lez Alriols Biorefinery Processes Research Group (BioRP), Chemical and Environmental Engineering Department, University of the Basque Country UPV/EHU, Plaza Europa, Donostia-San Sebastia´n, Spain Simon Alvisse Aquaporin Asia Pte. Ltd, Singapore, Singapore Muthupandian Ashokkumar ARC Dairy Innovation Hub, The University of Melbourne, Parkville, VIC, Australia; School of Chemistry, The University of Melbourne, Parkville, VIC, Australia Carolina Astudillo Escuela de Alimentos, Facultad de Agronomı´a y de los Alimentos, Pontificia Universidad Cato´lica de Valparaı´so, Valparaı´so, Chile Laurent Bazinet Institute of Nutrition and Functional Foods (INAF), Dairy Research Center (STELA), Department of Food Sciences, Laboratory of Food Processing and ElectroMembrane Processes (LTAPEM), Laval University, Quebec City, QC, Canada Katalin Belafi-Bako University of Pannonia, Veszprem, Hungary Andras Boor University of Pannonia, Veszprem, Hungary Alfredo Cassano Institute on Membrane Technology, ITM-CNR, Rende, Italy Roberto Castro-Mun˜oz Institute on Membrane Technology, ITM-CNR, Rende, Italy; University of Chemistry and Technology, Prague, Prague, Czech Republic; Nanoscience Institute of Aragon (INA), University of Zaragoza, Zaragoza, Spain George Q. Chen ARC Dairy Innovation Hub, The University of Melbourne, Parkville, VIC, Australia; Department of Chemical Engineering, The University of Melbourne, Parkville, VIC, Australia Carmela Conidi Institute on Membrane Technology, ITM-CNR, Rende, Italy Andre´s Co´rdova Escuela de Alimentos, Facultad de Agronomı´a y de los Alimentos, Pontificia Universidad Cato´lica de Valparaı´so, Valparaı´so, Chile Sirshendu De Department of Chemical Engineering, Indian Institute of Technology, Kharagpur, West Bengal, India Maria Norberta de Pinho Chemical Engineering Department, Instituto Superior Te´cnico, Universidade de Lisboa, Portugal; Center of Physics and Engineering of Advanced Materials, CeFEMA, Instituto Superior Te´cnico, Universidade de Lisboa, Portugal

xi

xii

List of Contributors

Xabier Erdocia Biorefinery Processes Research Group (BioRP), Chemical and Environmental Engineering Department, University of the Basque Country UPV/EHU, Plaza Europa, Donostia-San Sebastia´n, Spain Javier Ferna´ndez-Rodrı´guez Biorefinery Processes Research Group (BioRP), Chemical and Environmental Engineering Department, University of the Basque Country UPV/EHU, Plaza Europa, Donostia-San Sebastia´n, Spain Anca C. F˘arca¸s Faculty of Food Science and Technology, University of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, Cluj-Napoca, Romania Charis M. Galanakis Department of Research & Innovation, Galanakis Laboratories, Chania, Greece; Food Waste Recovery Group, ISEKI Food Association, Vienna, Austria Claus He´lix-Nielsen Technical University of Denmark, Lyngby, Denmark Fabio Herna´ndez-Ramos Biorefinery Processes Research Group (BioRP), Chemical and Environmental Engineering Department, University of the Basque Country UPV/EHU, Plaza Europa, Donostia-San Sebastia´n, Spain Andre´s Illanes Escuela de Ingenierı´a Bioquı´mica, Facultad de Ingenierı´a, Pontificia Universidad Cato´lica de Valparaı´so, Valparaı´so, Chile Paula Jauregi Department of Food and Nutritional Science, Harry Nursten building, The University of Reading, Whiteknights, Reading, United Kingdom Chenxiao Jiang CAS Key Laboratory of Soft Matter Chemistry, Collaborative Innovation Centre of Chemistry for Energy Materials, School of Chemistry and Materials Science, University of Science and Technology of China, Hefei, P.R. China Sankha Karmakar Department of Chemical Engineering, Indian Institute of Technology, Kharagpur, West Bengal, India Sandra E. Kentish ARC Dairy Innovation Hub, The University of Melbourne, Parkville, VIC, Australia; Department of Chemical Engineering, The University of Melbourne, Parkville, VIC, Australia Konstantina Kourmentza Department of Food and Nutritional Science, Harry Nursten building, The University of Reading, Whiteknights, Reading, United Kingdom Jalel Labidi Biorefinery Processes Research Group (BioRP), Chemical and Environmental Engineering Department, University of the Basque Country UPV/EHU, Plaza Europa, Donostia-San Sebastia´n, Spain Thomas S.H. Leong ARC Dairy Innovation Hub, The University of Melbourne, Parkville, VIC, Australia; Department of Chemical Engineering, The University of Melbourne, Parkville, VIC, Australia; School of Chemistry, The University of Melbourne, Parkville, VIC, Australia Gregory J.O. Martin ARC Dairy Innovation Hub, The University of Melbourne, Parkville, VIC, Australia; Department of Chemical Engineering, The University of Melbourne, Parkville, VIC, Australia

List of Contributors

xiii

Miguel Minhalma Center of Physics and Engineering of Advanced Materials, CeFEMA, Instituto Superior Te´cnico, Universidade de Lisboa, Portugal; Chemical Engineering Department, Instituto Superior de Engenharia de Lisboa, Instituto Polite´cnico de Lisboa, Portugal Nandor Nemestothy University of Pannonia, Veszprem, Hungary Irena Petrini´c University of Maribor, Maribor, Slovenia Ye Wee Siew Aquaporin Asia Pte. Ltd, Singapore, Singapore Sonia A. Socaci Faculty of Food Science and Technology, University of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, Cluj-Napoca, Romania; Institute of Life Sciences, University of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, Cluj-Napoca, Romania Guofei Sun Aquaporin Asia Pte. Ltd, Singapore, Singapore Nguyen Xuan Tung Aquaporin Asia Pte. Ltd, Singapore, Singapore Yaoming Wang CAS Key Laboratory of Soft Matter Chemistry, Collaborative Innovation Centre of Chemistry for Energy Materials, School of Chemistry and Materials Science, University of Science and Technology of China, Hefei, P.R. China Tongwen Xu CAS Key Laboratory of Soft Matter Chemistry, Collaborative Innovation Centre of Chemistry for Energy Materials, School of Chemistry and Materials Science, University of Science and Technology of China, Hefei, P.R. China

This page intentionally left blank

Preface Nowadays, functional compounds and the so-called nutraceuticals are used as additives in foods due to their ability to provide advanced technological and health properties to the final product. For instance, macromolecules like proteins have been utilized as fat replacement in meat and milk products, flavor enhancers in confectionery, as well as food and beverage stabilizers. Soluble dietary fiber is not only known for its ability to lower blood lipid level, but also for its advanced gelling properties that can replace fat in foods, stabilize emulsions, and improve the shelf life of the product. Smaller molecules such as natural antioxidants (polyphenols, carotenoids, tocopherols, ascorbic acid, etc.) have been connected to both nutritional (reduction of oxidative stress, prevention of cancer, arteriosclerosis, aging processes) and functional (preservative of vegetable oils and emulsions) properties in foods. On the other hand, undervalued bioresources and natural products are considered as target substrates for the separation and/or recovery of the previously mentioned functional compounds. The later process follows typically the principles of analytical chemistry in five stages (the so-called “5-Stages Universal Recovery Processing”): (1) macroscopic pretreatment, (2) separation of macro- and micromolecules, (3) extraction, (4) isolation
purification, and finally (5) product formation or encapsulation. Membrane technologies are among the key physicochemical and nondestructive techniques applied in the second, third, and fourth steps of the above downstream processing. As a general rule, membrane separations have been employed to concentrate macromolecules in the retentate and subsequently release smaller molecules in the permeate stream. This is the theoretic approach based on the sieving mechanism and molecular weight of the solutes. However, in practice the asymmetric manufacture of membrane pores does not always reflect molecular weight as the one and only criterion for the separation of macro- and micromolecules. In terms of food separation, the simultaneous recovery of macro- and micromolecules in one stream is a problem leading to additional purification stages, yield loss, and finally increased cost. The urgent need for sustainability within the food industries as well as the increasing stress for innovative, optimized, and tailor-made products has pushed food manufacturers to optimize membrane applications. Subsequently, there is a need for a new guide covering the latest developments in this particular direction.

xv

xvi

Preface

Over the last 3 years, the Food Waste Recovery Group (www.foodwasterecovery.group) of ISEKI Food Association has organized a series of activities (webinars, workshops, e-courses, etc.) and published books describing food waste recovery processing, highlighting sustainable solutions for the management of specific food processing byproducts (e.g., from the olive, grape, coffee, and cereal industries), describing tools for the implementation of innovations in the food industry, exploring the effect of emerging and nonthermal technologies on functional foods development, as well as dealing with target functional compounds like polyphenols. Following these advances, this book covers all the important of aspects of functional macroand micromolecule separation in the food industry using membrane technologies. The ultimate goal is to support the scientific community, professionals, and enterprises that aspire to develop new applications, as well as to improve efficiency and sustainability of food systems. The book consists of 11 chapters. Chapter 1 introduces membrane technologies (e.g., membrane materials and structures, modules and processes), while Chapter 2 introduces food components, and particularly discusses their properties, exploring recovery technologies and highlighting challenges. Chapter 3 deals with the recovery and purification of biosurfactants (specifically lipopeptides) using membranes, highlighting challenges in spite of cost reduction and future developments. Biosurfactants have a range of biological (e.g., antimicrobial) activities, whereas their biodegradability makes them very attractive molecules for a range of applications. In a similar approach, Chapter 4 reports the main technological advances and perspectives regarding the use of membrane technology as a strategy for manufacturing and purifying enzymatically produced oligosaccharides. Chapter 5 deals with the removal of pectin from fruit juices for clarification purposes. The processed juice after complete removal of pectin and aseptic packaging can have long shelf life along with its original taste and nutritional qualities. Chapter 6 provides a critical overview of the most relevant applications of pressure-driven membrane operations for the recovery of phenolic-based compounds from agro-food wastewaters in the light of recent studies carried out on both laboratory and pilot scale. Chapter 7 discusses the recovery and fractionation of lignin, the valorization of which could be the basis of the production of a broad variety of chemicals, such as vanillin, that can be employed in the food industry. Chapter 8 denotes the various applications of reverse osmosis, nanofiltration, ultrafiltration, and microfiltration in the processing of milk, and dairy waste streams, including on-farm concentration of milk, removal of microorganisms, and fractionation of protein and fat. Chapter 9 focuses only on nanofiltration applications in some selected areas of food processing, such as fruit juice clarification, and the dairy, sugar, and wine industries. Insights are given in light of the recent developments on pilot and large scale investigations. On the other hand, Chapter 10 discusses similar applications of electrodialysis and finally Chapter 11 presents two promising osmotic

Preface

xvii

membrane processes, where forward osmosis and membrane osmotic distillation were applied for up-concentration of a coconut milk solution and juices obtained from various colorful wild fruits, respectively. Conclusively, the book fills the gap existing in the current literature by analyzing in detail separation of macro- and micromolecules in the food industry using membrane technologies, aiming at generating new food products and addressing trends in the market. It addresses researchers, specialists, chemical engineers, processors, and manufacturers working in the food industry, as well as new product developers in the food and agricultural industry. It could also be purchased by university libraries and institutes all around the world to be used as a textbook and/or ancillary reading in graduate and postgraduate level multidisciplinary courses dealing with food engineering, food science and technology, and chemical and environmental engineering. Herein, I would like to thank all the authors of the book for their collaboration and work in bringing together different topics regarding membrane separation in one comprehensive text. Their acceptance of my invitation, their dedication to editorial guidelines, the book’s concept and timeline were an honor for me and are highly appreciated. I consider myself fortunate to have had the opportunity to collaborate with so many experts from Australia, Canada, Chile, China, Czech Republic, Denmark, Hungary, Greece, India, Italy, Singapore, Slovenia, Spain, Portugal, Romania, and the United Kingdom. I would also like to thank acquisition editor Nina Bandeira and book manager Katerina Zaliva for their assistance during editing and all of the Elsevier team during the production process. I would also like to express my gratitude to the support of Food Waste Recovery Group, who indicated the relevant experts and provided us with information regarding food separation with membrane technologies. Ultimately, a message for the reader: such collaborative projects always contain minor errors and gaps. Thereby, instructive comments and even criticism are and always will be welcome. So, please do not hesitate to contact me to discuss further insights of the book. Charis M. Galanakis1,2 1

Food Waste Recovery Group, ISEKI Food Association, Vienna, Austria, 2Research & Innovation Department, Galanakis Laboratories, Chania, Greece

This page intentionally left blank

Chapter 1

Introduction in Membrane Technologies Maria Norberta de Pinho1,2 and Miguel Minhalma2,3 1

Chemical Engineering Department, Instituto Superior Te´cnico, Universidade de Lisboa, Portugal, 2Center of Physics and Engineering of Advanced Materials, CeFEMA, Instituto Superior Te´cnico, Universidade de Lisboa, Portugal, 3Chemical Engineering Department, Instituto Superior de Engenharia de Lisboa, Instituto Polite´cnico de Lisboa, Portugal

Chapter Outline 1.1 Introduction 1.2 Materials and Structures of Membranes 1.2.1 Materials 1.2.2 Structures 1.3 Classification of Membrane Processes 1.4 Membrane Modules 1.4.1 Plate-and-Frame Modules 1.4.2 Tubular Modules 1.4.3 Hollow Fiber Modules 1.4.4 Spiral Wound Modules 1.5 Pressure-Driven Membrane Processes

1 3 3 3 6 8 8 8 9 10 12

1.5.1 Membrane Characterization 12 1.5.2 Mass Transfer at the Fluid Phase Circulating Tangentially to the Membrane: Concentration Polarization 15 1.5.3 Operation Modes 17 1.5.4 Case Study Applications 19 1.6 Electrodialysis 21 1.6.1 Characterization of IonExchange Membranes 22 1.6.2 Process Operation 23 1.7 Pervaporation 24 1.8 Conclusion 27 References 27

1.1 INTRODUCTION Membrane processes occupy an important place among separation techniques nowadays. In fact, in the last few decades, the conventional separations or chemical engineering unit operations like distillation, solvent extraction, etc. have been in many situations substituted or complemented by membrane separation processes. This is of particular relevance in the food industry, which constitutes the second membrane market after that of water, including wastewater and desalination. In the early 1960s, the development by Loeb and Sourirajan of cellulose acetate asymmetric membranes for sea water Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00001-2 © 2019 Elsevier Inc. All rights reserved.

1

2

Separation of Functional Molecules in Food by Membrane Technology

desalination made it possible to envisage their use in reverse osmosis (RO) at a large scale and as an alternative to the energy-intensive thermal processes. The cellulose acetate asymmetric membranes display the singular feature of combining high transmembrane fluxes with high rejection coefficients to NaCl, whereas they are synthesized by the phase-inversion technique. The versatility of this technique opened the way for the production of membranes from many other polymers and with a wide range of structures. The extension beyond polymers to inorganic and hybrid materials together with the recourse of advanced preparation techniques gave rise to the production of a wider range of membrane structures as described in Section 1.2. The main role of synthetic membranes as agents of selective transport or chemical species separation is the result of the synergy between a membrane structure and a given driving force. For example, the RO, nanofiltration (NF), or ultrafiltration (UF) processes utilize asymmetric microporous membrane structures and hydrostatic pressure differences as the driving force and electrodialysis, which utilizes ion-exchange membranes, requires an electrical potential difference as the driving force for the separation of charged species. These combinations lead to a number of membrane separation processes that are classified in Section 1.3 upon membrane structure, driving force, separation mechanism, and range of application. The industrial scale-up of these processes is very much linked to the development in the 1970s of compact membrane modules, such as the spiral wound or the hollow fiber, that can accommodate hundreds of square meters of membrane surface area per cubic meter of volume. The performance of the pressure-driven processes of microfiltration (MF), UF, NF, and RO depends not only on the membrane characteristics but is also highly dependent on the mode and conditions of operation. The hydraulic permeability is the parameter that assesses the membrane capability to water permeation. It is the water permeation rate per unit of membrane surface area and unit of transmembrane pressure and on the average decreases an order of magnitude from UF, NF to RO. The selective permeation to solutes is assessed through rejection coefficients that evaluate the fraction of the solute in the feed that does not permeate through the membrane. Their values depend on the membrane characteristics and on the operating conditions, namely the feed cross flow velocity and the transmembrane pressure. In fact, a membrane that is a selective barrier is always associated to the phenomenon of concentration polarization that occurs as a consequence of the accumulation of solute material rejected by the membrane, and leads to the development of a concentration profile from a higher value at the membrane/feed interface to a lower value in the feed bulk. Concentration polarization acts as an additional mass transfer resistance that leads to the decrease of the permeate fluxes and therefore is essential to be taken into account in the design of membrane systems. In addition, it is a precursor of membrane fouling, as the material accumulated at the membrane/feed

Introduction in Membrane Technologies Chapter | 1

3

interface may be adsorbed and modify the membrane surface porous structure. In the long term, this leads to reversible (removed by membrane cleaning) or even to irreversible fouling, meaning the decrease of membrane lifetime. The electrically driven process electrodialysis has been traditionally used for the desalination of brackish water and is now the best available technique for many applications in the food industry, for example, in the desalination of cheese whey and tartaric stabilization of wine. In general, electrodialysis represents a strong asset in the removal of salts and acids in food processing with preservation of the product’s organoleptic properties. In contrast with the two previously mentioned processes, pressure-driven and electrically driven processes, pervaporation still has a very small worldwide market, for economic reasons. This concentration-driven process is a technical alternative solution to distillation for the situations of azeotropic mixtures and of mixtures with components having very close boiling points. A strong asset of pervaporation in the food industry is the processing of temperature-sensitive mixtures, as it can be run at room temperature. The increase of the feed temperature within a range compatible with the thermolabile products leads to the enhancement of the productivity, therefore making it economically feasible. In fact, the economic limitation of pervaporation is mainly due to the low permeate fluxes of the commercial pervaporation membranes in the present market.

1.2 MATERIALS AND STRUCTURES OF MEMBRANES 1.2.1 Materials The development of Loeb and Sourirajan cellulose acetate asymmetric membranes (Loeb and Sourirajan, 1963) led to the application of membranes in large-scale processes and created the need of membrane production from a wide variety of synthetic polymers, inorganic materials, and hybrid polymer/ inorganic materials. Table 1.1 presents the most common materials used in membrane synthesis.

1.2.2 Structures In parallel with the developments in materials science, the advancement in membrane preparation techniques leads to the production of membranes with a myriad of physical structures that can be classified in four major groups: 1. 2. 3. 4.

Microporous Homogeneous Homogeneous ion exchange Asymmetric: integral and composite

4

Separation of Functional Molecules in Food by Membrane Technology

TABLE 1.1 Most Common Materials Used in the Manufacturing of Synthetic Membranes (Strathmann, 1986) Material Type

Material

Inorganic

Glass Ceramic Metallic Zeolitic

Natural based polymers

Cellulose diacetate and triacetate (CA) Cellulose acetate propionate (CAP) Cellulose acetate butyrate (CAB) Cellulose acetate methacrylate (CAM)

Synthetic polymers

Polyamide (PA) Polyether sulfone (PES) Polyimide (PI) Polyacrylonitrile (PAN) Polysulfone (PS) Polyvinylchloride (PVC) Polyvinylidene fluoride (PVDF) Polypropylene (PP)

1.2.2.1 Microporous Membranes With the manufacturing process, microporous membranes may acquire a variety of structures as shown in Table 1.2. They are applied in MF, which is an operation widely used in the food industry for beverage clarification and in general for the removal of suspended solids and colloidal matter. 1.2.2.2 Homogeneous Membranes One type of homogeneous membrane is constituted by dense films that allow the separation of molecules with similar dimensions due to the fact that they present different solubilities and diffusivities in the membrane matrix. The most important applications for these membranes are gas permeation and pervaporation. Another type of homogeneous membranes is the ionexchange membrane, which is mainly used in electrodialysis. Table 1.3 presents the most common materials and manufacturing processes for these membranes.

Introduction in Membrane Technologies Chapter | 1

5

TABLE 1.2 Most Common Microporous Membranes (Strathmann, 1986) Material

Pore Dimension (μm)

Manufacturing Process

Ceramic, metal, polymer powder

1
20

Pressing and sintering

Homogeneous polymer sheets

0.5
10

Stretching

Homogeneous polymer sheets

0.02
10

Track-etching

Polymer solution

0.01
5

Phase inversion

TABLE 1.3 Most Common Homogeneous Membranes Material

Manufacturing Processes

Polymer

Extrusion of films and casting of polymer solutions

Polymer ion-exchange resin

Pressing and casting of solutions

0.1-1 μm 100-200 μm

FIGURE 1.1 Scheme and cross-sectional diagram of an asymmetric membrane.

1.2.2.3 Asymmetric Membranes Asymmetric membranes have a central place in membrane separation technology due to the fact that they combine high permeation fluxes and high selectivity. In fact, that was a milestone in promoting from the 1960s onward the industrial scale-up of the pressure-driven membrane processes. The feature responsible for the high permeation/selectivity is the asymmetric crosssectional structure of a very thin dense layer whose thickness can vary from 0.1 to 1 μm and a subjacent porous layer with thicknesses varying from 100 to 200 μm. This is schematically illustrated in Fig. 1.1. The thin dense layer is responsible for the membrane selectivity, often called the active layer, and the porous layer gives essentially mechanical strength to the membrane. In contrast with the symmetric membranes, which act as traditional filters and retain particles in their pores with subsequent fouling and decline of permeation fluxes, the asymmetric membranes act as surface filters retaining the rejected material at the membrane surface and not inside the pores and therefore avoiding pore blocking/fouling, as shown in Fig. 1.2.

6

Separation of Functional Molecules in Food by Membrane Technology

200 μm

FIGURE 1.2 Scheme and cross-sectional diagram of a symmetric membrane.

1.2.2.3.1 Integral Asymmetric Membranes These membranes are prepared by the phase inversion method (Loeb and Sourirajan, 1963; Strathmann, 1986; Kesting, 1985), in which a starting polymeric casting solution, containing a polymer and a solvent system, is cast into a film and after some evaporation time the film is quenched in a nonsolvent, which for most cases is water. The versatility of this process both in terms of casting solution composition and in terms of the casting parameters (evaporation time, quenching media type, and temperature) allows the tailoring of membranes with different structures and permeation characteristics.

1.2.2.3.2

Composite Asymmetric Membranes

These membranes are the result of a deposition of a dense selective layer onto a porous substrate of a different polymeric material. Commercial composite asymmetric membranes may be prepared by several methods, such as (Kesting, 1985; Cadotte et al., 1981): 1. Preparation of a dense thin layer, which is then mechanically laminated on a porous support layer. 2. Preparation of a dense film directly over the microporous substrate layer by dip coating. 3. Plasma polymerization at the surface of the porous membrane. 4. Polycondensation or interfacial polymerization over the surface of a porous support layer.

1.3 CLASSIFICATION OF MEMBRANE PROCESSES The selective permeation of membranes with different structures and/or physicochemical properties can be attributed to different mechanisms associated to sieving, steric hindrances, membrane/solvent/solute(s) interactions, solution/diffusion characteristics, and electrical migration. The external action of pressure, electrical, and concentration driving forces together with different transport mechanisms leads to the membrane separation processes described in Table 1.4.

TABLE 1.4 Membrane Separation Processes: Operating Principles and Applications Separation Process

Membrane Type

Driving Force

Separation Mechanism

Range of Application

Microfiltration (MF)

Microporous 0.1
10 μm

Pressure 0.1
1 bar

Sieving

Sterilization

Asymmetric

Pressure 0.5
8 bar

Ultrafiltration (UF)

Clarification Sieving

Separation of macromolecular solutes

Membrane/solvent/solute interactions Nanofiltration (NF)

Asymmetric “skin type”

Pressure 10
40 bar

Reverse osmosis (RO)

Asymmetric “skin type”

Pressure 20
100 bar

Solution/diffusion

Separation of salts and microsolutes

Dialysis

Microporous 0.1
10 μm

Concentration gradient

Diffusion

Separation of salts and microsolutes from macromolecular solutions

Electrodialysis (ED)

Ion exchange

Electrical potential gradient

Electrical migration

Desalting of ionic solutions

Gas permeation (GP)

Homogeneous

Pressure and concentration gradient

Solution/diffusion

Separation of gas mixtures

Concentration gradient

Solution/diffusion

Concentration and separation of small organic solutes

Symmetric and asymmetric Pervaporation (PV)

Homogeneous Symmetric and asymmetric

Sieving Membrane/solvent/solute interactions Solution/diffusion

Separation of salts and small organic solutes

8

Separation of Functional Molecules in Food by Membrane Technology

1.4 MEMBRANE MODULES Although the development of asymmetric membranes, which are highly selective and have high permeation fluxes, has significantly contributed to the industrial development of membrane processes, this had to be accompanied by the development of a membrane support system, where membranes are arranged in different configurations characterized by high membrane surface area per unit of volume and with the capability to process fluids at the right hydrodynamic conditions and operating transmembrane pressures. Besides those characteristics, this support system should also be easily cleaned and maintained. In membrane separation processes the system where the membranes are inserted is called the membrane module. Membrane modules have the functions of supporting the membranes, separating physically the feed solution from the permeate solution, promoting a good mass transfer in the different compartments, and allowing the establishment of the process driving force. Additional constraints related to technical and economical parameters have to be taken into consideration. The modules should accommodate membranes with a very high packing degree and the pumping system for solutions distribution should present a reasonable cost. There are several modules capable of satisfying the above requisites and they should be carefully chosen considering the physicochemical characteristics of the feed to be processed. Industrial membrane modules may be classified into four types: plate-and-frame, tubular, hollow fibers, and spiral wound.

1.4.1 Plate-and-Frame Modules In the plate-and-frame modules the feed flows in a narrow channel with a rectangular cross section where the channel walls may be one or two membranes. Due to the very small channel thickness, which restrains the feed flow rate, the normal flow conditions in industrial modules are limited to laminar flow. Although laminar flow is in place, the mass transfer is kept at high values because the channel length is only of a few centimeters and therefore the growth of the concentration boundary layer is controlled. The easy cleaning and the low internal volume make these modules adequate in the cases where periodical disinfections are necessary. In the plate-and-frame modules represented in Fig. 1.3 the membrane support plates have a membrane sheet in each side and they are stacked alternately with spacers. The feed stream, which flows tangentially to the membrane surface (active layer), follows the path described in Fig. 1.3. The pressurized fluid that permeates through the membrane is collected in the permeate channels at atmospheric pressure.

1.4.2 Tubular Modules In tubular modules, represented in Fig. 1.4, tubes of a porous material with adequate chemical, thermal, and mechanical resistances are coated with a

Introduction in Membrane Technologies Chapter | 1

9

FIGURE 1.3 Representation of a plate-and-frame module.

FIGURE 1.4 Representation of a tubular module. Figure courtesy of Koch Membrane system, Inc. (www.kochmembrane.com) and used by the publisher with permission.

film that constitutes the membrane. The feed solution circulates in the inner part of the tubes in contact with the active layer, while the permeate passes through the tubes. There are monochannel or multichannel configurations for these membranes and they are mounted in outer shells that may accommodate several membrane units.

1.4.3 Hollow Fiber Modules The hollow fibers module, shown in Fig. 1.5, is composed by several polymeric capillary fibers that are introduced inside an outer shell. The feed solution may flow on the inside of the fibers and the permeate is collected in the

10

Separation of Functional Molecules in Food by Membrane Technology

FIGURE 1.5 Representation of a hollow fiber module. Figure courtesy of Koch Membrane system, Inc. (www.kochmembrane.com) and used by the publisher with permission.

FIGURE 1.6 Representation of a spiral wound module. Figure courtesy of Koch Membrane system, Inc. (www.kochmembrane.com) and used by the publisher with permission.

space between the fibers and the outer shell. In this situation the membrane active layer is inside the fibers. One other situation is when the feed solution flows on the outside of the fibers, the active layer is on the outer fiber face, and the permeate is collected inside the fibers. The choice between these two situations depends mainly on the application purpose and on the feed solution characteristics.

1.4.4 Spiral Wound Modules The configuration of the spiral wound modules is presented in Fig. 1.6. These modules have several membrane sheets folded and connected to a perforated central tube. The feed solution flows tangentially to the membrane

Introduction in Membrane Technologies Chapter | 1

11

surface (active layer side) and the permeate gets through the membranes and is collected in the central tube. The comparison of the membrane modules in terms of properties and applications is shown in Table 1.5. The 4th column is related to the control degree of the concentration polarization phenomenon, which is explained in the Section 1.4.2. The dimensions of tubular membranes are compatible with high feed flow rates and turbulent flows. Table 1.6 provides the most adequate modules for each of the membrane processes. TABLE 1.5 Membrane Modules: Properties and Applications Module Type

Compaction (m2/m3)

Price

Concentration Polarization Control

Application

Plate-andframe

400
600

High

Acceptable

MF, UF, NF, RO, PV, and GP

Tubular

20
30

Very high

Very good

MF, UF (tolerance of feed with high solids content)

Hollow fiber

600
1200

Very low

Poor

MF, UF, NF, RO, PV, and GP

Spiral wound

800
1000

Low

Acceptable

UF, NF, RO, PV, and GP

TABLE 1.6 Membrane Processes and Modules Separation Process

Plate-andFrame Module

Tubular Module

Hollow Fiber Module

Spiral Wound Module

Microfiltration

1

11

1

2

Ultrafiltration

11

11

1

1

Reverse osmosis

1

1

11

11

Pervaporation

11

2

1

1

Gas permeation

2

2

11

11

Electrodialysis

11

2

2

2

11, Very adequate; 1 , adequate; 2 , not adequate.

12

Separation of Functional Molecules in Food by Membrane Technology

TABLE 1.7 Modules Comparison in Reverse Osmosis Module

Packing Density

Fouling Control

Cleaning Performance

Plate-and-frame

Moderate

Fair

Good

Tubular

Low

Very good

Very good

Hollow fiber

Very high

Poor

Poor

Spiral wound

High

Fair/poor

Fair/poor

The selection of the membrane modules has to take into account the physicochemical feed characteristics and the possible pretreatments required. Given the industrial importance of RO, the Table 1.7 compares the different RO modules’ characteristics.

1.5 PRESSURE-DRIVEN MEMBRANE PROCESSES The pressure-driven processes, MF, UF, NF, and RO, and their main characteristics and applications are summarized in Table 1.4 and in Fig. 1.7. Fig. 1.8 represents schematically the operation mode in any of the membrane processes. The feed stream flows tangentially to the membrane surface and generates two other streams: the concentrate stream enriched in the solutes that are rejected by the membrane and the permeate stream that passes through the membrane. The efficiency of the UF, NF, and RO processes depends greatly on the concentration polarization phenomenon that occurs when a concentration profile is developed at the membrane surface, due to the rejected solutes accumulation. The membrane characteristics and the operating conditions are important factors in the development of this phenomenon and its quantification is crucial for the membrane processes’ design. From this perspective, membrane characterization is presented below.

1.5.1 Membrane Characterization The hydraulic permeability, LP, is the characterization parameter that quantifies the membrane capacity to permeate pure water. This parameter can be determined by representing the permeate fluxes to pure water, JP, in function of the transmembrane pressure applied. The slope of the straight line obtained is the hydraulic permeability, LP, as depicted in Fig. 1.9. The variation of the permeate volume per unit of time, membrane surface area, and transmembrane pressure is the hydraulic permeability, LP, according to Eq. (1.1): JP;pure water 5 LP ΔP

ð1:1Þ

Introduction in Membrane Technologies Chapter | 1

13

Particles and colloids

Microfiltration (MF) Proteins, polysaccharides, polyphenols, and other macromolecules

Ultrafiltration (UF) Glucose, fructose, amino acids and small organic solutes, and bivalent salts

Nanofiltration (NF) Salts and small organic solutes

Reverse Osmosis (RO)

FIGURE 1.7 Applications of pressure-driven membrane processes.

Feed Q0, CA0

Concentrate QC, CAC

Permeate QP, CAP

FIGURE 1.8 Schematic representation of membrane process. QA0, QAP, QAC—Flow rate of feed, permeate and concentrate streams, respectively; CA0, CAP, CAC—concentration of solute A in the feed, permeate and concentrate, respectively.

For an aqueous solution of a given solute A, that is totally or partially rejected by the membrane, the permeation flux, JP, is given by the ratio of the permeate flow rate, QP, divided by the membrane surface area. For the same transmembrane pressure, the JP value is generally lower than the one for pure water (JP, pure water), and is given by Eq. (1.2). JP 5 LP ðΔP 2 ΔπÞ

ð1:2Þ

where Δπ 5 π0
πP, is the average osmotic pressure difference between the feed side (π0) and the permeate side (πP). The membrane separation performance for a given solute A is given by the apparent rejection coefficient, fA, defined as: fA 5

CA0 2 CAP CA0

ð1:3Þ

where CA0 and CAP are the solute A concentration in the feed and permeate streams, respectively. The accumulation of solute A at the membrane surface leads to an increase of its concentration, CAM, and therefore a new intrinsic rejection coefficient, f0 A, is defined as: f 0A 5

CAM 2 CAP CAM

ð1:4Þ

14

Separation of Functional Molecules in Food by Membrane Technology 120.0

J P (kg/h/m2)

100.0 80.0 60.0 40.0 Lp

20.0 0.0 0

0.5

1

1.5

2

2.5

3

ΔP (bar)

FIGURE 1.9 Variation of pure water permeate flux, JP, with the transmembrane pressure applied, ΔP.

100% 90% 80%

fA (%)

70% 60% 50% 40% 30% 20% 10% 0% 0

10

20

30

40

50

Neutral solutes, A, molecular weight (kDa)

FIGURE 1.10 Example of representation of the profile of rejection coefficients to model solutes, A.

One other important characterization parameter is the molecular weight cutoff (MWCO), which is defined as the molecular weight (MW) of a solute A that is rejected by the membrane in the range from 0.9 to 0.99, depending on the author/manufacturer criteria. The rejection coefficient profile, fA versus MW, obtained for a serious of solutes of increasing MW is shown in Fig. 1.10. The upper part of this figure is nearly a plateau, making it very difficult to have an accurate intersection point of the curve, fA versus MW, with the horizontal line set at the fA value selected by the author between 0.9 and 0.99. In order to overcome this problem, the higher range of fA values is linearized through the representation of log(fA/(1
fA)) versus MW. This results in a well-defined intersection point for MWCO determination. As an

Introduction in Membrane Technologies Chapter | 1

15

example, this is shown in Fig. 1.11, for setting the value of fA equal to 0.91 and log(fA/(1
fA)) 5 1.

1.5.2 Mass Transfer at the Fluid Phase Circulating Tangentially to the Membrane: Concentration Polarization

log (fA/(1-fA))

The tangential circulation of a solute A solution at a given operating pressure results in a convective flow perpendicular to the membrane surface. Depending on the membrane selective characteristics the solvent permeates preferentially through the membrane and the solute A is totally or partially rejected by the membrane. The rejected solute accumulates near the membrane surface and in steady state, the convective flux toward the membrane is balanced by the solute flux through the membrane and the diffusive flux from the membrane surface to the bulk solution. The development of a concentration profile in the liquid boundary layer, as schematically shown in Fig. 1.12, is designated by concentration polarization. 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

y = 0.0552x + 0.1826 R2 = 0.9606

MWCO = 14.8 kDa 0

10

20

30

40

50

Neutral solutes molecular weight (kDa)

FIGURE 1.11 Example of graphic representation for membrane MWCO determination.

x Feed

CA0

δ

Δx

CA CAm

vp.CA

–DAB.

Membrane Permeate, vp vp.CAp

FIGURE 1.12 Concentration polarization, concentration profile, CA, in steady state.

∂CA ∂x

16

Separation of Functional Molecules in Food by Membrane Technology

The quantification of the concentration polarization phenomenon is carried out through a differential mass balance to solute A in a volume with differential thickness, Δx, in the fluid boundary layer adjacent to the membrane. The result is the differential Eq. (1.5): vp CA 2 DAW

dCA 5 vp CAp dx

ð1:5Þ

where vp is the permeate flux (m/s) and DAw is the diffusivity of solute A in the solvent. For a boundary layer of thickness δ the boundary conditions are: x 5 0; CA 5 CAM

ð1:6Þ

x 5 δ; CA 5 CA0

ð1:7Þ

The integration of Eq. (1.5) with the boundary conditions (Eqs. 1.6 and 1.7) results in: vpδ CAM 2 CAp 5 eDAW CA0 2 CAp

ð1:8Þ

Using the film theory (Bird et al., 2002), a mass transfer coefficient k, is introduced as k 5 DAw/δ, and the concentration polarization is quantified by Eq. (1.9): vp CAM 2 CAp 5ek CA0 2 CAp

ð1:9Þ

The mass transfer coefficient is obtained by empirical correlations for Sherwood number, Sh 5 Sh(Re, Sc). These correlations are obtained in the literature for different geometries and flow regimes. For example, for a circular geometry the most common correlations are: Sh 5

kDe 5 0:04Re 0:75 Sc 0:33 DAW

Turbulent regimen

ð1:10Þ

and Sh 5


 kDe 5 Re Sc De =L DAW

Laminar regimen

ð1:11Þ

where Sh, Re, and Sc are Sherwood, Reynolds, and Schmidt numbers, with De and L being the diameter and length of a cylindrical channel, respectively. As previously mentioned, two different rejection coefficients can be defined as membrane characterization parameters, an apparent rejection coefficient, fA, and an intrinsic rejection coefficient, f0 A, based on the solute concentration in the bulk feed solution, CA0, and in the fluid/membrane interface, CAM, respectively. These two different coefficients are the result

Introduction in Membrane Technologies Chapter | 1

17

of the occurrence of the concentration polarization phenomenon, and they can be related to each other using Eqs. (1.3), (1.4), and (1.9). Eqs. (1.12) and (1.13) express these interrelationships.
 fA exp vp =k 0 
 fA 5 ð1:12Þ 1 2 fA 1 2 exp vp =k and


 exp vp =k CAM 
 5 0
0 CA0 fA 1 1 2 fA exp vp =k

ð1:13Þ

Simplifying the concentration polarization module, described in Eq. (1.9), for the particular case of total rejection of solute A, CAP 5 0, one gets Eq. (1.13). The CAM/CA0 ratio varies with the operating parameters and membrane characteristics (f0 A). The operating parameters of transmembrane pressure and feed circulation velocity (or Re) are directly determining vp and k, respectively. The formation of concentration polarization acts as an additional resistance to mass transfer and leads to permeate flux decline (Macedo et al., 2011). The apparent rejection coefficient, fA, is not only dependent on the membrane characteristics but is also strongly dependent on the process operating conditions, namely feed flow velocity and transmembrane pressure.

1.5.3 Operation Modes The pressure-driven membrane processes, MF, UF, NF, and RO, can be operated in batch or in continuous mode.

1.5.3.1 Batch Operation The batch operation can be run in total recirculation of the feed/concentrate stream and continuous withdraw of the permeate stream, Fig. 1.13A. A recirculation loop may be introduced, in order to better control the hydrodynamic conditions inside the membrane module, Fig. 1.13B (Kulkarni et al., 1992). (A)

(B)

Feed tank

Feed Pump Permeate

Feed tank

Feed pump Recirculation pump Permeate

FIGURE 1.13 Batch operation (B) with and (A) without recirculation loop.

18

Separation of Functional Molecules in Food by Membrane Technology

1.5.3.2 Continuous Operation The continuous operation of MF/UF/NF/RO is used for large-scale production systems. In these systems the solution to be processed is fed into a set of modules, arranged in stages, with a large membrane area, and after a single passage on each module two streams is obtained, a permeate stream and a concentrate stream. Due to the large surface area in each module, which can reach 30 m2/module in the case of the spiral wound modules, the longitudinal average circulation velocity decreases and the solute concentration increases are very significant, and these variations have to be taken into account in membrane process design. In the case of spiral wound modules that have small feed channels that accommodate low feed flow rates, characteristic of laminar flow and low Reynolds numbers, the introduction of spacers in the feed channels promotes mixing, which leads to the minimization of the concentration polarization phenomenon. Furthermore, the hydrodynamic conditions needed to achieve a good mass transfer can be reached though the recirculation of the concentrate stream. On the other hand, for UF with tubular modules, the module’s dimensions allow the operation with higher feed circulation flow rates, characteristic of turbulent regimen, and therefore a good feed mixing can be achieved. Additionally, feed recirculation loops can be introduced in each stage to maintain adequate hydrodynamic and mass transfer conditions. This continuous operation mode with recirculation loops in each stage is shown in Fig. 1.14 (Kulkarni et al., 1992). The increase of solute concentration along the module leads to the axial decrease of the permeate flux. For the same total membrane area the increase of the number of stages in series leads to the increase of the productivity in the first stages due to smoother axial concentration profiles, and therefore to an increase of the overall operation efficiency. However, the technoeconomic analysis shows that for a number of stages above four the increase of the overall efficiency does not compensate the additional costs introduced by the recirculation loops. To assure the hydrodynamic conditions recommended by the manufacturers the number of modules in parallel in each stage should decrease in successive stages. This is commonly designated by the Christmas tree configuration.

Feed tank

Stage 2

Stage 1 Q1 Feed pump

Recirculation pump Permeate 1

Q2

Recirculation pump

Q3 Permeate 2

FIGURE 1.14 Continuous operation with recirculation loops in each stage.

Introduction in Membrane Technologies Chapter | 1

19

1.5.4 Case Study Applications Several applications of pressure-driven processes in the dairy industry are schematically presented in Figs. 1.15
1.17. In the first example, UF is used to concentrate the sheep milk proteins. This enrichment of milk in proteins is the basis for the yield increase in cheese manufacturing (Catarino et al., 2013).

1.5.4.1 Technoeconomical Analysis In batch mode, the processing of 1000 L/day of sheep milk is performed with an operation daily time of 4 hours, using UF membranes characterized by total rejection to the milk proteins and by permeate fluxes of 50 L/m2/h at the working transmembrane pressure. For a volumetric concentration factor (VCF) of 4 (from 1000 to 250 L) a membrane surface area of 3.75 m2 is required. A preliminary technoeconomic analysis is carried out in Table 1.8. A very common industrial use of UF is the fractionation of cheese whey into a concentrate stream enriched in cheese whey proteins and a permeate enriched in lactose and salts. Sheep milk

Vi = 1000 l Vf = 250 l Feed pump

Permeate with lactose and salts Jp = 50 L/m2/h

FIGURE 1.15 Batch sheep milk concentration by ultrafiltration.

Cheese whey Spray dryer

Cheese whey powder Feed pump Cheese whey concentrated for higher yield production of whey cheeses Permeate rich in lactose and salts

FIGURE 1.16 Concentration of cheese whey by ultrafiltration for subproducts recovery.

20

Separation of Functional Molecules in Food by Membrane Technology Sheep cheese whey

Feed pump

Cheese whey concentrated for further use in food products

UF

Permeate rich in lactose

NF

RO

Concentrate

Water for reuse

FIGURE 1.17 Fractionation of sheep cheese whey by an integrated sequence of UF/NF/RO for water and subproducts recovery.

TABLE 1.8 Technoeconomic Analysis of UF Sheep Milk Concentration Capacity

1000 L/day

Daily operation time

4h

Operation days per year

330

Average permeate flux

50 l/m2/h

Concentration factor

4

Membrane surface area

3.75 m2

Investment (2000 h/m2)

7500 m2

Lifetime

10 years

Annual maintenance cost (5% of investment)

380 h

Capital costs

2.3 h/m3

Energy costs

0.5 h/m3

Maintenance costs

1.1 h/m3

Total costs

3.9 h/m3

As seen in Fig. 1.16 the concentrate can be further dried for cheese whey powder production or used for the higher yield production of whey cheeses (Macedo et al., 2015). An integrated process of UF/NF is presented in Fig. 1.17 as an example of processing sheep cheese whey for byproduct recovery and effluent minimization. The UF yields a concentrate that is protein enriched and a

Introduction in Membrane Technologies Chapter | 1

21

permeate that is further processed by NF for the production of a concentrate rich in lactose and a permeate that can be desalinated by RO to produce water to be used in the factory (Minhalma et al., 2007; Magueijo et al., 2006). The very abundant literature on the use of MF, UF, NF, and RO in the must and wine industries goes from the concentration and rectification of grape must (Santos et al., 2008; Catarino et al., 2008; de Pinho et al., 2015), wine clarification (Gonc¸alves et al., 2001; de Pinho, 2010), to recovery of polysaccharides and polyphenols from wine and winery wastes (Resende et al., 2013; Giacobbo et al., 2013a,b, 2015, 2017). The wide range of MWCO of UF and NF membranes is a strong asset in the fractionation of saccharide mixtures (Catarino et al., 2008; Minhalma et al., 2006).

1.6 ELECTRODIALYSIS Electrodialysis is a separation process that uses ion-exchange membranes under the action of an electric field (driving force) for the separation of ionic species from aqueous solutions or other neutral solutes. The ion-exchange membranes are one of the key elements of this unit operation and they are selectively permeable to ions. Depending on the membrane characteristics, cation exchanger or anion exchanger, a preferential cation or anion permeation through the membrane will take place, respectively. The basic functioning principle of an electrodialyzer is depicted in Fig. 1.18. The feed solution containing ions flows through several compartments that have a cation-exchange membrane on one side and an anionexchange membrane on the other. Under the action of an electric field that is established between the anode and the cathode, the cations migrate into the cathode, permeating selectively through the cation-exchange membrane, while the anions migrate toward the anode permeating selectively through the anion-exchange membrane. In this way, there is the transport of the salt from the feed compartment, designated also by the diluate compartment, to the concentrate under the action of the electric field, which is the driving force. The salt is retained in the adjacent compartments, also designated by concentrate compartment because the anion-exchange membranes retain the cations that are migrating towards the cathode under the electric field action while the cation-exchange membranes block the anion transport that is moving toward the anode. The ions that reach the electrodes suffer redox reactions and there is the formation of chemical compounds, which are removed by the electrodes’ cleaning solution. An electrodialysis cell has a diluate compartment and a concentrate compartment, that together with a pair of ion-exchange membranes, one anionic and other cationic, constitute the basic unit of this separation process. An

22

Separation of Functional Molecules in Food by Membrane Technology Electrode washing solution

Feed Concentrate + + + + + + + + +

Anode

+ + + + +

-

+

+

+

+

+ + n pairs of + cation and anion + transfer + membranes + + + + + Electric current + + -

+ + + + + + + + + + + +

-

Cathode

Concentrate Diluate

FIGURE 1.18 Schematic diagram of an electrodialysis stack operation.

industrial electrodialyzer has a few hundred of these cells, and Fig. 1.18 represents schematically the membrane stack part of an electrodialysis unit.

1.6.1 Characterization of Ion-Exchange Membranes The ion-exchange commercial membranes are films of ion-exchange resins with thicknesses between 0.1 and 1.5 mm, with exchange capacities ranging from 1 to 3.5 eq-mol/kg of dry resin equilibrated with sodium and water content that varies between 30% and 50%. For conventional applications, the membranes are generally produced from reticulated polystyrene that is further sulfonated to produce (2SO2 3 ) groups or quaternized to produce (2NR1 3 ) groups in the polymeric matrix, yielding cation-exchange and anion-exchange membranes, respectively. Due to the low mechanical resistance of these membranes it is necessary to add a fabric support layer (Strathmann, 1992). Membrane characteristics like electric resistance (Ω cm2), burst strength (MPa), and thickness (mm) are provided by the ion-exchange membrane manufacturers.

Introduction in Membrane Technologies Chapter | 1

23

Electrodes washing wastewater

Power

Diluate

+

Feed

-

Microfilter "Make-up" Concentrate purge Feed tank

Concentrate tank

FIGURE 1.19 Flow diagram of an electrodialysis plant.

1.6.2 Process Operation Fig. 1.18 shows the principle of operation of an electrodialysis stack with “n” membrane pairs. At a large scale an electrodialysis unit can comprise several stacks and each one may have from 20 to 500 pairs of membranes per electrode pair, with surface areas ranging from 0.02 to 2 m2. The diluate and concentrate compartments are delimited by the membranes, which have in between spacers that prevent the membranes from contacting each other and at the same time promote mass transfer. The spacers have usually a thickness ranging from 0.5 to 2 mm depending on the application and on the manufacturer. An electrodialysis plant in addition to the membrane stacks comprises power supply, pumps, tanks, process control devices, feed solution pretreatment unit, etc. A typical flow diagram is presented in Fig. 1.19. After MF the feed solution is circulated into the electrodialysis stacks through the diluate channels. The deionized solution and the concentrated brine streams leaving the last stack are collected in storage tanks when the desired deionization is reached. In case of salt precipitation risk the pH in the concentrate tank may be adjusted. In order to prevent the formation of free chlorine the electrodes are often rinsed with a separate chloride-free solution. Besides the typical use of electrodialysis for the production of potable water from brackish water, the food, pharmaceutical, and the chemical industries also use this technology. In the food industry, desalination of cheese whey and the wine tartaric stabilization by electrodialysis are wellestablished processes with economic and product quality advantages. In fact, for wine tartaric stabilization the electrodialysis presents two strong assets over the traditional cold tartaric stabilization method: 1. Preservation of the wine organoleptic properties 2. Avoidance of the diatomaceous earth filtration (Gonc¸alves et al., 2001; 2003; Narciso et al., 2005; Soares et al., 2009).

24

Separation of Functional Molecules in Food by Membrane Technology

Other applications with different levels of technology maturity are found in the dairy, juice, and sugar industries. Hybrid processes of electrodialysis with other membrane processes or conventional unit operations are the subject of intense research and development (de Pinho et al., 2015).

1.7 PERVAPORATION In pervaporation the feed as a liquid mixture, at atmospheric pressure, circulates tangentially to the membrane and due to a concentration driving force the permeate in the vapor form is enriched in the preferential permeating component(s). This operation differs from the other membrane separation processes because there is a change of physical state between the liquid feed and the gaseous permeate stream (Aptel, 1986; Rautenbach and Albrecht, 1989). The driving force that is responsible for the mass transfer across the membrane is a gradient of chemical potentials between the liquid phase and the gaseous phase. This can be achieved by lowering the activity of the permeating component(s) in the permeate side, by maintaining the partial pressure lower than the saturation vapor pressure. This can be carried out by applying vacuum or using an inert carrier gas in the permeate side, as presented in Figs. 1.20 and 1.21, respectively. In the vacuum method, the permeate is kept at low pressure through the use of a vacuum pump, and the permeate condensation is carried out under vacuum. The inert carrier gas method is not generally used at the industrial scale.

Permeate Vacuum pump FIGURE 1.20 Schematic representation of pervaporation by vacuum.

Permeate Carrier gas FIGURE 1.21 Schematic representation of pervaporation with inert carrier gas.

Introduction in Membrane Technologies Chapter | 1

25

The heat required for the liquid/vapor phase change, from the feed to the permeate side, is supplied by the sensible heat from the liquid mixture, and therefore there is a decrease of the feed temperature along the membrane. In industrial applications the pervaporation units incorporate heat exchangers, in order to compensate the liquid mixture temperature decrease. The pervaporation performance depends greatly on the membrane/solute (s)/solvent interactions. The mass transfer is usually described by the solution
diffusion model, which assumes a three-step mechanism: (de Pinho et al., 1990). 1. Preferential sorption at the membrane/liquid feed interface 2. Diffusion through the membrane, described by Fick’s first law 3. Desorption on the membrane/permeate interface There are two main pervaporation types: the most common one uses membranes that permeate preferentially water, and the other one uses membranes that permeate preferentially organics (de Pinho et al., 1990; Cipriano et al., 1991; Neto and de Pinho, 2000). In fact, the first commercial application of pervaporation reports to the preferential water permeation in water/ ethanol systems as part of the downstream processing in the bioethanol production. The two main parameters used to access the pervaporation separation performance are the selectivity and the permeate flux. The membrane selectivity results from the preferential affinity of the membrane material for a given component of the liquid mixture, that can be quantified by two parameters, α or β, defined as: α5

Yi =Yj Y i ð 1 2 Xi Þ 5 Xi =Xj Xi ð1 2 Yi Þ

ð1:14Þ

Yi Xi

ð1:15Þ

β5

where Xi and Yi correspond to the feed and permeate mass fractions, respectively, for a component i (the species that permeates preferentially), and Xj and Yj correspond to the mass fractions of the other component in the feed and permeate, respectively. For a membrane that permeates preferentially water, this is very well illustrated in Fig. 1.22, for the water (i)/isopropanol (j) system. As it can be seen, the selectivity depends greatly on the water fraction in the feed. For very low feed water content the permeate is rich in isopropanol, but for increasing feed water content the permeate gets enriched in water with low amounts of isopropanol. In systems that have an azeotrope, as shown by the vapor/liquid equilibrium of the water/isopropanol system also represented in Fig. 1.22, pervaporation constitutes a viable technical alternative to distillation.

26

Separation of Functional Molecules in Food by Membrane Technology

Mole fraction of water in permeate

1 0.8 0.6 0.4 ELV 0.2 0 0

0.2 0.4 0.6 0.8 Mole fraction of water in feed

1

FIGURE 1.22 Comparison of the water/isopropanol separation by distillation and pervaporation with a membrane that permeates preferentially water.

Permeate flux (kg/h/m2)

1 Memb. A

0.8 0.6

Memb. B -70ºC

0.4 0.2

Memb. B -35ºC

0 0

0.2

0.4 0.6 0.8 Mole fraction of water in feed

1

FIGURE 1.23 Variation of permeate flux with feed water content and temperature for a water/ isopropanol system.

On the other hand, permeate flux quantifies the mass transport through the membrane and it is dependent on the composition and temperature of the feed, as depicted in Fig. 1.23. As the feed temperature and the preferential permeating solute concentration increases, the permeate flux increases. The application of pervaporation at industrial scale has not been as successful as in the cases of the pressure-driven and electrically driven membrane processes. This is mainly due to the higher energy requirements associated to the occurring phase change. However, pervaporation is still an important solution for processes where distillation is not technoeconomically viable, like the systems that have azeotropes or mixtures whose components have very close boiling points. One other important application is the processing of temperature-sensitive mixtures.

Introduction in Membrane Technologies Chapter | 1

27

The pervaporation application can be arranged into three main categories: 1. Dehydration of organic solvents with membranes permeating preferentially water. 2. Purification of aqueous streams contaminated with traces of organic compounds. Although membranes permeating preferentially organics exhibit very low permeation fluxes, this separation can be technoeconomically feasible because of the small organic quantities that have to be removed. 3. Organic
organic separation.

1.8 CONCLUSION The design of membrane processes requires not only deep knowledge of the feed solution to be processed and of the products to be obtained, but also careful selection of membranes, membrane modules, and pretreatments. This will allow production enhancement through the optimization of the operating conditions that lead to the minimization of the concentration polarization and fouling phenomena. The decision of the mode of operation (batch or continuous) depends mainly on the feed volume to be processed. Two main features of membrane processes should be emphasized in the development of new applications: (1) the modular character of the membrane technology, which allows to envisage its application in small, medium, and large scale installations; and (2) the versatility in the synthesis of hybrid processes involving membrane processes with other operation units either upstream as pretreatments, or downstream, as polishing stages for the recovery of valuable compounds from multicomponent mixtures.

REFERENCES Aptel, P., 1986. In: Bungay, P.M., Lonsdale, H.K., de Pinho, M.N. (Eds.), Synthetic Membranes: Science, Engineering and Applications. Elsevier, p. 403. Bird, R.B., Stewart, W.E., Lightfoot, E.N., 2002. Transport Phenomena, second ed. John Wiley & Sons, New York. Cadotte, J.E., King, R.S., Marjele, R.J., Peterson, R.J., 1981. Interfacial synthesis in the preparation of reverse osmosis membranes. J. Macromol. Sci. Chem. A 15, 727. Catarino, I., Minhalma, M., Beal, L.L., Mateus, M., de Pinho, M.N., 2008. Assessment of saccharide fractionation by ultrafiltration and nanofiltration. J. Membr. Sci. 312, 34
40. Catarino, I., Martins, A.P.L., Duarte, E., Prudeˆncio, E.S., de Pinho, M.N., 2013. Rennet coagulation of sheep milk processed by ultrafiltration at low concentration factors. J. Food Eng. 114, 249
254. Cipriano, M.M., Geraldes, V., de Pinho, M.N., 1991. Membrane separation processes for the clean production of xanthates. J. Membr. Sci. 62, 103. de Pinho, M.N., 2010. In: Peinemann, K.-V., Pereira, S., Giorno, L. (Eds.), Membrane Processes in Must and Wine Industries in Membranes for food Applications, Volume 3. Wiley-VCH. de Pinho, M.N., Rautenbach, R., Herion, C., 1990. Mass transfer in radiation-grafted pervaporation membranes. J. Membr. Sci. 54, 131.

28

Separation of Functional Molecules in Food by Membrane Technology

de Pinho, M.N., Geraldes, V., Catarino, I., Method for Simultaneous Concentration and Rectification of Grape Must Using Nanofiltration and Electrodialysis, US 8,945,645 B2, Feb. 3 2015. Giacobbo, A., Oliveira, M., Duarte, E., Mira, H.M.C., Bernardes, A.M., de Pinho, M.N., 2013. Ultrafiltration based process for the recovery of polysaccharides and polyphenols from winery effluents. Sep. Sci. Technol. 48, 445
454. Giacobbo, A., Bernardes, A.M., de Pinho, M.N., 2013. Nanofiltration for the recovery of low molecular weight polysaccharides and polyphenols from winery effluents. Sep. Sci. Technol. 48, 2524
2530. Giacobbo, A., Prado, J.M., Meneguzzi, A., Bernardes, A.M., de Pinho, M.N., 2015. Microfiltration for the recovery of polyphenols from winery effluents. Sep. Purif. Technol. 143, 12
18. Giacobbo, A., Meneguzzi, A., Bernardes, A.M., de Pinho, M.N., 2017. Pressure-driven membrane processes for the recovery of antioxidant compounds froth winery effluents. J. Cleaner Prod. 155, 172
178. Gonc¸alves, F., Fernandes, C., de Pinho, M.N., 2001. White wine clarification by micro/ultrafiltration: effect of removed colloids in tartaric stability. Sep. Purif. Technol. 22
23, 423
429. Gonc¸alves, F., Fernandes, C., Cameira dos Santos, P., de Pinho, M.N., 2003. Wine tartaric stabilization by electrodialysis and its assessment by the saturation temperature. J. Food Eng. 59, 229
235. Kesting, R.E., 1985. Synthetic Polymeric Membranes—A Structural Perspective, second ed. John Wiley & Sons, New York. Kulkarni, S.S., Funk, E.W., Li, N.N., Winston Ho, W.S., Sirkar, K.K. (Eds.), 1992. Membrane Handbook. Van Nostrand Reinhold, New York. Loeb, S., Sourirajan, S., 1963. Adv. Chem. Ser 38, 117. Macedo, A., Duarte, E., Pinho, M., 2011. The role of concentration polarization in ultrafiltration of ovine cheese whey. J. Membr. Sci. 381, 34
40. Macedo, A., Duarte, E., Fragoso, R., 2015. Assessment of the performance of three ultrafiltration membranes for fractionation of ovine second cheese whey. Int. Dairy J. 48, 31
37. Magueijo, V., Minhalma, M., Queiroz, D., Geraldes, V., Macedo, A., de Pinho, M.N., 2006. Reduction of wastewaters and valorisation of by-products from “Serpa” cheese manufacture using Nanofiltration. Water Sci. Technol. 52 (10
11), 393
399. Minhalma, M., Beal, L.L., Catarino, I., Mateus, M., de Pinho, M.N., 2006. Optimization of saccharide fractionation using nanofiltration/ultrafiltration. Desalination 199 (1
3), 337
339. Minhalma, M., Magueijo, V., Queiroz, D., de Pinho, M.N., 2007. Optimization of “Serpa” cheese whey nanofiltration for effluent minimization and byproducts recovery. J. Environ. Manage. 82 (2), 200
206. Narciso, D., Fernandes, C., Cameira dos Santos, P., de Pinho, M.N., 2005. New method for the estimation of the potassium hydrogen tartrate saturation temperature of Port wines. Am. J. Enol. Viticult. 56 (3), 255
266. Neto, J.M., de Pinho, M.N., 2000. Mass transfer modelling for solvent dehydration by pervaporation. Sep. Purif. Technol. 18, 151. Rautenbach, R., Albrecht, R., 1989. Membrane Processes. John Wiley & Sons, UK. Resende, A.M., Catarino, S., Geraldes, V., de Pinho, M.N., 2013. Separation and purification by ultrafiltration of white wine high molecular weight polysaccharides. Ind. Eng. Chem. Res. 52, 8875
8879. Santos, F.R., Catarino, I., Geraldes, V., de Pinho, M.N., 2008. Concentration and rectification of grape must by nanofiltration. Am. J. Enol. Viticult. 59, 446
450.

Introduction in Membrane Technologies Chapter | 1

29

Soares, P.A.M.H., Geraldes, V., Fernandes, C., Cameira dos Santos, P., de Pinho, M.N., 2009. Wine tartaric stabilization by electrodialysis: prediction of required deionization degree. Am. J. Enol. Viticult. 60, 183
188. Strathmann, H., 1986. In: Bungay, P.M., Lonsdale, H.K., de Pinho, M.N. (Eds.), Synthetic Membranes: Science, Engineering and Applications. Elsevier, p. 1. Strathmann, H., 1992. Electrodialys and related processes. In: Membranes: Proceedings of the CEE-Brazil Workshop on Membrane Separation Processes Rio de Janeiro, May 3
8.

This page intentionally left blank

Chapter 2

Introduction in Functional Components for Membrane Separations ˘ s1 and Charis M. Galanakis3,4 Sonia A. Socaci1,2, Anca C. Farca¸ 1

Faculty of Food Science and Technology, University of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, Cluj-Napoca, Romania, 2Institute of Life Sciences, University of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, Cluj-Napoca, Romania, 3 Department of Research & Innovation, Galanakis Laboratories, Chania, Greece, 4 Food Waste Recovery Group, ISEKI Food Association, Vienna, Austria

Chapter Outline 2.1 Introduction 2.2 Challenges in Functional Food Development 2.3 Recovery of Bioactive Compounds From Conventional and Nonconventional Sources 2.3.1 Proteins and Active Peptides 2.3.2 Polyphenols 2.3.3 Polysaccharides

31 32

33 34 42 45

2.3.4 Lipids 2.3.5 Bioactive Compounds of Animal Origin 2.4 Separation and Recovery of Macro- and Micromolecules Using Membrane Technologies 2.5 Conclusions References Further Reading

48 49

52 68 68 77

2.1 INTRODUCTION Functional ingredients and foods are constantly gaining importance in the everyday choices of consumers, due to the increasing interest of consumers in promoting and improving their health through diet. This new trend, based on the principle that it is more effective to prevent a disease than to cure it, has motivated the scientific world in increasing efforts toward research focused on identifying, isolating, and purifying bioactive compounds with health-promoting properties that can be used as functional ingredients to fortify foods and beverages (McClement, 2012). Another point of interest is finding new sources of bioactive compounds, optimizing the extraction and Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00002-4 © 2019 Elsevier Inc. All rights reserved.

31

32

Separation of Functional Molecules in Food by Membrane Technology

purification methods, as well as maintaining the bioactivity of these compounds throughout the product shelf life. There is no consistent definition of the term “functional food” throughout the world’s countries, the term being constantly revised by regulatory bodies. In Europe, for example, the EU government does not have a formal legislative definition for “functional foods” (Martirosyan and Singh, 2015). A commonly accepted definition by several organizations is that “functional foods” are “foods or ingredients of foods that provide an additional physiological benefit, beyond their basic nutrition” (Day et al., 2009). Recently, the Functional Food Center, USA, defined the term as “natural or processed foods that contains known or unknown biologically-active compounds; the foods, in defined, effective, and nontoxic amounts, provide a clinically proven and documented health benefit for the prevention, management, or treatment of chronic disease” and it is currently trying to standardized this definition (Martirosyan and Singh, 2015). According to the above definitions, the simplest examples of functional foods are fruits and vegetables due to their content in phytochemicals such as polyphenols, carotenoids, dietary fiber, vitamins, fatty acids, minerals, proteins, etc. These bioactive compounds have known beneficial health effects proven by various in vitro or in vivo studies (e.g., antioxidant activity protecting the cells from reactive oxygen species damage and thus lowering the risk of developing different diseases associated with oxidative stress, antimicrobial properties, antiinflammatory, antitumor activity, antihypertensive, antidiabetic) (Ganesan and Xu, 2017; Hasler, 2002; Williamson, 2009). Examples of functional foods may also include foods in which a component was added (e.g., enriched in omega-3 fatty acids) or removed/reduced (e.g., low-fat products, gluten-free) or a food in which one or several components have been modified, replaced, or enhanced to improve its health properties (e.g., yogurt with probiotic bacteria) (Stein and Rodriguez-Cerezo, 2008).

2.2 CHALLENGES IN FUNCTIONAL FOOD DEVELOPMENT Functional foods exist at the interface between food and drugs, and therefore offer great potential for health improvement and prevention of diseases when ingested as part of a balanced diet (Hilliam, 2000; Otles and Cagindi, 2012). The functional properties of the bioactive compounds are proven by extensive scientific research, but to be used in human nutrition, their safety, efficacy, and bioavailability must also be ensured by a regulatory framework to provide clear information to consumer and not misleading claims. In this sense, the benefits versus the risk of long-term intake need a thorough assessment and further investigation. Several aspects have to be considered and characterized: 1. the safety of the intake of these bioactive compounds (with or without nutrient value) related to the long-term consumption of functional foods

Introduction in Functional Components Chapter | 2

33

2. the interactions between the bioactive compounds used to fortified the food product and other ingredients 3. the impact of the technological processing of food on the functionality of bioactive compounds The requirements of functional food safety indicated by FAO/WHO impose on manufacturers the obligation to conduct placebo-controlled clinical studies and to evaluate their results in four phases: safety, efficiency, effectiveness, and surveillance (Otles and Cagindi, 2012). Even though the functional food market is growing every year, there is still the need for developing new functional food. The process of developing new functional food is an expensive one, involving technological, legal, and commercial challenges (McClements et al., 2009). Regarding the technological challenges, one has to take into consideration the compatibility between the bioactive compound and food matrix (e.g., solubility, stability) and the processing flow, the development of adequate delivery systems for bioactive compounds. Moreover, the substantial investment of a company in the research required for the functional food to meet the efficacy and safety criteria is not always returned within the company. In other words, after the health claim is adequately documented, other competing companies may use the claim and make profit. Incentives, such as a period of exclusivity or tax incentives, would encourage food companies to pursue functional food development by ensuring a profitable return on successful products (IFT, 2018; Khan et al., 2013). Another opportunity is to find renewable and lowcost sources of high-value compounds. The generation of food waste is inevitable, especially during the preconsumption stage (Ravindran and Jaiswal, 2016). The implementation of strict legislation for human health and environmental safety and the emergence of novel techniques for the recovery of commercially important biomolecules has caused enormous interest in food supply chain waste valorization. Technologies for the recovery of high added value compounds are pivotal to the utilization of food waste for commercial applications. In this conjuncture, the exploitation of food byproducts for the recovery and reuse of valuable bioactive compounds is one of the most sustainable approaches.

2.3 RECOVERY OF BIOACTIVE COMPOUNDS FROM CONVENTIONAL AND NONCONVENTIONAL SOURCES Fruit and vegetable byproducts such as peel, bark, seeds, leaves, etc., often contain more bioactive compounds and with higher antioxidant activities than those found in the edible portion (Can-Cauich et al., 2017). Thereby, more and more research is focused on exploiting these unconventional sources for the recovery of valuable molecules. Use of these byproducts as sources of bioactive compounds usually requires preliminary processing

34

Separation of Functional Molecules in Food by Membrane Technology

steps before the extraction, like reducing the water content or drying in order to minimize the microbiological and biochemical reactions, to increase the concentration in bioactive compounds, to ease the handling, and to extend the storage period. There is no universal method for the extraction of bioactive compounds, but several criteria must be fulfilled for an extraction process to be considered efficient and economically sustainable: selectivity toward the analyte, high extraction yields, possibility of solvent recovery or using “green solvents,” use of low-cost reagents, low energy consumption, reduced extraction time, maintaining the functionality of the recovered molecules, and possibility to be scaled up (Azabou et al., 2016; Tunchaiyaphum et al., 2013; Socaci et al., 2017b). Thereby, the classical solid
liquid and liquid
liquid extraction methods are being optimized, to overcome the limitations related to the high amount of used solvents, toxicity of the solvents, waste management, etc. Among the currently most employed modern extraction techniques we can mention are microwave-assisted extraction, ultrasound-assisted extraction, pressurized liquid extraction (e.g., pressurized hot water extraction), enzymeassisted extraction, supercritical CO2-based extraction, membrane filtration processes, and other emerging techniques (Angiolillo et al., 2015; Banerjee et al., 2017; Goula et al., 2017; Heng et al., 2017, pp. 25
27; Loginov et al., 2013; Socaci et al., 2017a). Choosing the appropriate extraction method is crucial in the recovery of the bioactive molecules, because it significantly influences the yield and the composition of the extract. Several examples of extraction methods used for the recovery of bioactive compounds from different matrices are presented in Table 2.1.

2.3.1 Proteins and Active Peptides Proteins and peptides include a group of macromolecules that consists of amino acids of polymeric chains of amino acids. They are known to possess a variety of nutritional, functional, and biological properties. Nutritionally, the proteins are a source of energy and amino acids, which are essential for growth and maintenance. Functionally, the proteins contribute to the physicochemical and sensory properties of various protein-rich foods. For example, when incorporated in food products, protein may exert emulsifying properties, film forming properties, flavor binding capacity, viscosity increase by binding the water, and gelation properties (Aydemir et al., 2014; Sharma et al., 2010; Socaci et al., 2017a; Yu et al., 2016). Proteins are used in the food industry as fat replacement in meat and milk products, flavor enhancers in confectionery, and food and beverage stabilizers (Patsioura et al., 2011). Furthermore, many dietary proteins possess specific biological properties that make these components potential ingredients of functional or healthpromoting foods. The majority of the functional properties of proteins are attributed to physiologically active peptides encrypted in protein molecules

TABLE 2.1 Examples of Bioactive Compounds From Plant and Animal-Derived Products or Waste and the Employed Extraction Techniques Compound Class

Waste Origin

Byproduct Source

Extraction Techniques

References

Proteins and bioactive peptides

Cereals

Brewer’s spent grain

Ultrasonic-assisted extraction

Tang et al. (2010)

Sequential extraction of proteins and arabinoxylans

Vieira et al. (2014)

Enzymatic assisted extraction

Niemi et al. (2013)

Ultrafiltration

Tang et al. (2009)

Defatted rice bran

Alkali extraction and isoelectric precipitation

Han et al. (2015)

Rice byproducts

Enzymatic hydrolysis and membrane filtration technique

Ferry et al. (2017)

Soybean

Subcritical water hydrolysis

Pinkowska and Oliveros (2014)

Rapeseed meal

Ultrasound-assisted aqueous extraction

Yu et al. (2016)

Subcritical water hydrolysis

Pinkowska et al. (2013)

Sunflower meal

Alkaline solubilization and acid precipitation

Salgado et al. (2012)

Hazelnut meal

Solvent extraction (water, acetone)

Aydemir et al. (2014)

Flaxseed hull

Coagulation and ultrafiltration

Loginov et al. (2013)

Oil crops

(Continued )

TABLE 2.1 (Continued) Compound Class

Waste Origin

Byproduct Source

Extraction Techniques

References

Canola meal

Alkaline solubilization and acid precipitation (isoelectric precipitation)

Manamperi et al. (2011) Karaca et al. (2011)

Fruits and vegetable

Animal byproducts

Electroactivated solutions (noninvasive extraction method)

Gerzhova et al. (2015)

Salt precipitation

Karaca et al. (2011)

Palm kernel cake

Enzymatic hydrolysis

Ng et al. (2013)

Defatted cherry seeds

Enzymatic hydrolysis and membrane ultrafiltration

Garcia et al. (2015)

Apricot kernel cake

Alkaline solubilization and acid precipitation

Sharma et al. (2010)

Spirulina

Enzymatic hydrolysis and membrane filtration

Ma et al. (2007)

Fish and chicken

Isoelectric solubilization/precipitation

Shi et al. (2017)

Chicken feathers

Enzymatic hydrolysis and membrane ultrafiltration

Fontoura et al. (2014)

Whey

Ion exchange chromatography/cationexchange selective adsorption process

El-Sayed and Chase (2010a,b)

Membrane filtration techniques

Ndiaye et al. (2010)

Egg yolk

Ultrafiltration, gel filtration and reversedphase HPLC

Pokora et al. (2014)

Shellfish

Enzymatic hydrolysis and micro-, ultra-, and nanofiltration, ion exchange chromatography

Beaulieu et al. (2013)

Polysaccharides

Cereals

Fruits and vegetables

Brewer’s spent grain

Enzymatic hydrolysis

Niemi et al. (2012a,b)

Sequential extraction of proteins and arabinoxylans

Vieira et al. (2014)

Acid hydrolysis

Mussatto and Roberto (2005)

Defatted flaxseed Rice bran Sesame husk

Enzymatic hydrolysis

Nandi and Ghosh (2015)

Citrus peel and apple pomace

Subcritical water extraction

Wang et al. (2014)

Orange peel

Ultrahigh pressure/microwave

Guo et al. (2012)

Microwave extraction

Maran et al. (2013)

Microwave

Seixas et al. (2014)

Passion fruit Pomelo

Quoc et al. (2015)

Pumpkin kernel cake and oilpumpkin biomass

Sequential extraction

Koˇsta´lova´ et al. (2013)

Berry, bilberry, and black currant press cake from juice production

Aqueous extraction, sequential buffer extraction and ultrafiltration

Mu¨ller-Maatsch et al. (2016) Hilz et al. (2005)

Mandarin peel

Crossflow microfiltration

Cho et al. (2003)

Olive mill wastewaters

Ultrafiltration

Galanakis et al. (2010) (Continued )

TABLE 2.1 (Continued) Compound Class

Waste Origin

Byproduct Source

Extraction Techniques

References

Lipids, fatty acids

Cereals

Brewer’s spent grain

Soxhlet extraction

Niemi et al. (2012b)

Rice bran

Solid
liquid extraction supercritical fluid to the extraction

Oliveira et al. (2012) Perretti et al. (2003)

Fruit and vegetables

Animal byproducts

Grape seeds

Pressurized carbon dioxide extraction with compressed carbon dioxide as solvent and ethanol as cosolvent

Dalmolin et al. (2010)

Supercritical fluid extraction

Prado et al. (2012)

Microalgae

Cell disruption method

Oviyaasri et al. (2017)

Byproducts from the palm oil extraction

Separation techniques through membranes

Tan et al. (2007)

Fish and chicken

Isoelectric solubilization/precipitation

Shi et al. (2017)

Fish

Supercritical fluid extraction

Sarker et al. (2012) Ferdosh et al. (2015)

Polyphenols

Cereals

Brewer’s spent grain

Alkaline hydrolysis

Mussatto et al. (2007)

Rice bran biomass (byproduct)

Subcritical water extraction

Pourali et al. (2010)

Roasted wheat germ (byproduct)

Supercritical fluids CO2

Gelmez et al. (2009)

Oil crops

Fruits and vegetables

Rapeseed

Ultrasound-assisted aqueous extraction

Yu et al. (2016)

Byproducts from the palm oil extraction

Separation techniques through membranes

Tan et al. (2007)

Olive byproducts

Continuous countercurrent liquid
liquid extraction

Allouche et al. (2004)

Chemical (acid) hydrolysis

Bouallagui et al. (2011)

Sunflower meal

Mild-acidic protein extraction with adsorptive removal of phenolic compounds

Weisz et al. (2013)

Flaxseed hull

Coagulation and ultrafiltration

Loginov et al. (2013)

Tomato pomace and skin

Enzymatic assisted extraction/solvent extraction

Azabou et al. (2016)

Potato peels and tubers

Pressurized liquid extractor

Luthria (2012)

Solvent extraction (stirring)

Ieri et al. (2011)

Ultrasound extraction

Singhai et al. (2011)

Purple sweet potatoes

Ultrasound-assisted extraction, centrifugation and ultrafiltration

Zhu et al. (2016)

Orange peels

Nanofiltration

Conidi et al. (2012)

Bergamot juice

Ultrafiltration, nanofiltration

Conidi et al. (2011)

Artichoke wastewater

Membrane filtration (ultrafiltration, nanofiltration) and polymeric resins

Conidi et al. (2015) (Continued )

TABLE 2.1 (Continued) Compound Class

Waste Origin

Byproduct Source

Extraction Techniques

References

Onion skin (byproduct)

Subcritical water

Ko et al. (2011)

Forest fruit pomaces

Supercritical fluid extraction

Laroze et al. (2010)

Peanut skins byproduct

Microwave

Ballard et al. (2010)

Apple pomace

Ultrasound extraction

Pingret et al. (2012)

Apple and peach pomaces

Subcritical water

Adil et al. (2007)

Grape seeds

Supercritical fluid extraction

Agostini et al. (2012) Yilmaz et al. (2010)

Carotenoids

Essential oils

Fruits and vegetables

Fruit and vegetables

Concentration and ultrafiltration

Nawaz et al. (2006)

Enzymatic assisted extraction

Azabou et al. (2016)

Supercritical fluids CO2

Yi et al. (2009)

Citrus peel

Ultrasound extraction

Sun et al. (2011)

Sea buckthorn seeds

Supercritical carbon dioxide fluid extraction

Kagliwal et al. (2011)

Citrus peel

Solvent extraction, distillation, hydrodistillation

Thongnuanchan and Benjakul (2014)

Tomato pomace and skin

Introduction in Functional Components Chapter | 2

41

(Roupas et al., 2007). Depending on the source, the molecular weight (MW) of proteins ranges from decades to hundreds of kDa. Proteins may also have different charges because amino acids are amphoteric and the charge of their molecule depends on its isoelectric point. For example, oat contains mainly globulin (20
35 kDa and pI 5 5.5), albumin (14
17 kDa, 4.0 , pI , 7.0), and prolamin (17
34 kDa, 5.0 , pI , 9.0) (Klose and Arendt, 2012). These characteristics are used for their selective fractionation during membrane processing. Peptides can show enhanced bioactivities and different functional properties to those reported for the parent protein. Structurally, peptides are short-chain peptide segments of protein molecules, including 2
20 amino acid residues. Their biological activities (e.g., antioxidant, opioid or mineralbinding, immunomodulatory, antimicrobial, hypocholesterolemic, antihypertensive, antithrombotic, antidiabetic) are tailored by their molecular weights and amino acid sequences (Atef and Ojagh, 2017; Kadam et al., 2015; Lemes et al., 2016). Thus, the antimicrobial peptides usually have a molecular weight below 10 kD, and contain up to 50 amino acids, about half being hydrophobic (Atef and Ojagh, 2017). Besides, antioxidant activity has been proved to increase when small peptides are obtained (Moure et al., 2006). Based on these properties, peptides and amino acids recovered by means of subcritical water hydrolysis can be of interest for nutraceutical products, the food industry (e.g., food additives), pet food, or managing patients unable to digest proteins (Marcet et al., 2016). Most bioactive peptides are derived from expensive protein matrices, which in most cases make their application unfeasible. Hence the proteinrich waste generated by agroindustries has become an attractive alternative for protein recovery and reuse (Lemes et al., 2016). But, in order for a byproduct to be considered as a source of protein, it has to fulfill two essential criteria: a high protein content and a high-quality protein (well-balanced essential amino acid composition and bioavailability) (WHO, 2007). The advances made in the isolation and purification technologies led to an increased acceptance of recovered soy proteins from byproducts as functional ingredients in foods. This growing acceptance is also attributed to the nutritional and potential health benefits of the recovered proteins, such as the prevention of hypercholesterolemia, atherosclerosis, and cancer (Yadav and Joyner, 2014). The main byproducts with a relatively high content of protein are the defatted meals resulting from oil production, including olive, palm, sunflower, rapeseed, but also soybean meal, rice bran, cereal spent grain, and microalgae, these biomasses being also available in large quantities and at a low cost. Rice byproducts have been suggested as a cheap, renewable, and abundant source of antioxidant compounds and bioactive peptides. In a recent study, a protein byproduct from the rice starch industry was hydrolyzed with

42

Separation of Functional Molecules in Food by Membrane Technology

five commercial proteolytic enzymes, avoiding the use of solvents or chemicals. The liquid supernatants of the scaled-up digestions were subfractionated by crossflow membrane filtration to isolate peptide samples having different molecular weight ranges. Isolated peptide fractions were demonstrated to possess antioxidant, antihypertensive, antityrosinase and/or antiinflammatory activities, while all were shown to be neither cytotoxic nor irritant (Ferry et al., 2017). A highly attractive waste for the production of bioactive peptides is the one resulting from olive oil manufacturing, considering that from 100 kg of fresh olives around 2 kg of flour with approximately 22% protein is generated in the process (Rodriguez et al., 2008). In another study peptide extracts were obtained by the digestion of the cherry seed protein isolate with alcalase and thermolysin. The results showed that fractions obtained by ultrafiltration possess relative high antioxidant and antihypertensive properties (Garcia et al., 2015). Application of ultrafiltration on recovery of protein from brewer’s spent grain waste water was studied by Tang et al. (2009). More than 92% of the protein was retained by membranes with both molecular weight cutoff (MWCO) of 5 and 30 kDa, the protein contents in the final product being 20.09% and 15.98%, respectively, compared with that of 4.86% concentrated by rotary evaporation (Tang et al., 2009).

2.3.2 Polyphenols Polyphenols are plant secondary metabolites, containing more than one phenol unit or building block per molecule. They are mostly found in fruits, vegetables, cereals, tea, coffee, cacao, etc., but also in the derived foods and in the byproducts resulting after the processing of the raw materials. Beside the fact that polyphenols influence the sensorial attributes of plants and foods (e.g., aroma, taste, astringency, and color), they are also among the most active natural antioxidants (Landete, 2012). There are over 8000 polyphenolic compounds identified in nature (Ganesan and Xu, 2017; Landete, 2012). Polyphenols present in plants are usually conjugated to sugars and organic acids and can be divided in two big groups: flavonoid and nonflavonoid phenolics (Landete, 2012). But, depending on the number of phenolic groups and structural elements, they can be further classified in four classes: G

Flavonoids—including flavones, flavonols, flavanones, flavanols, anthocyanidins, and isoflavonoids, constitute the most important class, with over 5000 compounds being currently described (Lima et al., 2014). They are found in fruits and vegetables, green tea, red wine, etc. Flavonoids include phenolic alcohols (i.e., flavan-3-ols), flavonols, flavones, anthocyanins, and secoiridoids. Phenolic alcohols such as tyrosol

Introduction in Functional Components Chapter | 2

G

G

G

43

and aldehydes such as isovanillic acid are present in grapes and olives, having a low MW (MW 5 110
228) similar to nonflavonoids. Flavonols, like procyanidin B2, quercetin, and kaempferol, are larger molecules (MW 5 286
579), while flavones (e.g., apigenin) have similar characteristics to flavonols. Anthocyanidins (e.g., malvidin, cyanidin-3-glucoside, rutin) are mainly comprised of malvidin 3-glucoside and respective pyruvic acid derivatives as well as pigments of anthocyanins linked either to a catechin unit or to a procyanidin dimer or to a 4-vinylphenol group. The MW of anthocyanidins and their structural characteristics varies importantly depending on their degree of polymerization, for example, starting from simple dimeric acetaldehyde malvidin 3-glucoside structures and reaching heavier fractions (Galanakis, 2015a). Stilbens—belong to a relatively small group of nonflavonoid class of phenolic compounds. The main representative compound is trans-resveratrol. Major dietary sources include grapes, wine, soy, peanuts (Ozcan et al., 2014). Lignans—also a nonflavonoid group of phenolic compounds, these are found in seeds, cereals, legumes, and algae. Phenolic acids—including hydroxybenzoic acids (e.g., gallic acid), hydroxycinnamic acids (e.g., ferulic acid, chlorogenic acid), and their derivatives are the most common nonflavonoid naturally occurring phenolics. They are widespread in the plant kingdom, with appreciable amounts being found in fruits, tea, coffee, vegetables, etc. Other nonflavonoid phenolics include o-diphenols like hydroxytyrosol, gallic, and protocatechuic acids. All these compounds possess low MW (148
194) due to the appearance of only one aromatic ring. Indeed, o-diphenols (i.e., gallic and caffeic acid) are generally smaller and have more polar molecules than the rest hydroxycinnamic acid derivatives (Galanakis et al., 2013).

The polyphenols can also be divided in extractable and nonextractable compounds. Extractable polyphenols are low- and mediummolecular mass phenolics, they can be extracted from the plant matrix using different solvents (water, methanol, aqueous acetone, etc.), and are potentially bioavailable in the small intestine. On the other hand, the nonextractable polyphenols are macromolecular compounds (polymeric polyphenols) or single phenols bound to dietary fiber or protein, which remain insoluble in the usual aqueous-organic solvent and an additional treatment should be applied for their release from the matrix (e.g., acidic hydrolysis, thiolysis). Nonextractable polyphenols are bioavailable only in the large intestine and, as the soluble ones, have remarkable antioxidant activity (Kristl et al., 2011; Landete, 2012; Matthews et al., 1997; Perez-Jimenez and Calixto Saura, 2015). The great scientific interest in polyphenols is mainly due to their functionality as free radical scavengers, but other biological activities (e.g.,

44

Separation of Functional Molecules in Food by Membrane Technology

antiinflammatory, vasodilating, etc.) may be also assigned to these compounds. Depending on the quantity ingested and on their bioavailability in the human body, polyphenols could have desired health effects but also adverse ones (Landete, 2012). Thus, an adequate daily intake of polyphenols, or food containing polyphenols, may improve the protection of the human body against several chronic diseases such as cancers, cardiovascular diseases, cerebrovascular diseases, diabetes, aging and neurodegenerative diseases, and diabetes mellitus (Ganesan and Xu, 2017; Vladimir-Kneˇzevi´c et al., 2012; Landete, 2012; Lima et al., 2014; Ozcan et al., 2014). There is no accurate data available in the literature regarding polyphenol dietary intake. Different attempts were made in establishing a total daily polyphenol intake, but this estimation is difficult not only due to the large extent of dietary habits and preferences, but also due to the structural diversity of polyphenols, their uneven distribution in plant tissues, and their fractioning during processing of raw materials. Scalbert and Williamson (2000), estimated that the daily intake of polyphenols could reach up to 1 g/day. This value is higher than for the rest of the ingested phytochemicals (B10 times higher than the intake of vitamin C and 100 times higher than the intakes of vitamin E and carotenoids) (Scalbert et al., 2005). Another aspect that should be considered is that in most studies, when evaluating the polyphenol dietary intake, the contribution of the nonextractable fraction was ignored. A recent study conducted by Perez-Jimenez and Calixto Saura (2015), focused on determined macromolecular polyphenol (including hydrolyzable and nonextractable proanthocyanidins) content from different fruits and vegetables consumed in four European countries (France, Germany, Spain, the Netherlands) in order to estimate the macromolecular polyphenol intake from food and vegetables. The results showed that the daily intake of this type of polyphenols, in all four European countries taken into study, was about 200 mg. Therefore, it is fair to say that the macromolecular polyphenols are a major part of the total polyphenol content and total antioxidant capacity of fruits and vegetables. Depending on the class of polyphenolic compounds, the extraction process can be achieved by different techniques but all involve the dissolution of compounds at the cellular level in the plant matrix followed by their diffusion in the extraction solvent. Nowadays there is an emphasis on developing or optimizing “eco-friendly” extraction processes in order to minimize the negative impact on the environment and also to obtain functional compounds that do not pose safety issues when used in foods. Membrane processing is one method that reduces the use of toxic organic solvents and concentrates the final product. The main advantages of membrane filtration, compared to the classical extraction procedures, are high purity (clarification, purification, and concentration), reduced energy consumption, higher separation efficiency, milder extraction conditions, and possibility to scale up (Nawaz et al., 2006). Furthermore, membrane filtration is suitable for the recovery of

Introduction in Functional Components Chapter | 2

45

bioactive molecules, including polyphenols, from agrifood byproducts (Loginov et al., 2013; Nawaz et al., 2006). A membrane processing method that can be used with success in the separation of polyphenols is ultrafiltration because it can reject compounds with molecular weight higher than 1000 (Nawaz et al., 2006). Many studies were performed in recent years and promising results were obtained for the recovery of polyphenols from different plant materials (e.g., sweet potatoes, bergamot juice) or food waste (e.g., grape seeds, flaxseed hull, artichoke wastewater, citrus byproducts) (Conidi et al., 2011, 2012, 2015; Loginov et al., 2013; Zhu et al., 2016).

2.3.3 Polysaccharides Polysaccharides, which are classified as carbohydrates, are long-chain molecules made up of simple sugar molecules connected by glycosidic linkages. In nature, there is a wide diversity of polysaccharide molecular structures, with different functional properties. The polysaccharides are commonly categorized as starche polysaccharides (SPs) and nonstarch polysaccharides (NSPs); the latter can be further divided in soluble and insoluble fiber (Goh et al., 2014). Demonstration of the beneficial implications of fiber-rich foods on human health has led to their integration in the category of functional ingredients. Soluble dietary fibers reduce the intestinal absorption of blood cholesterol whereas insoluble dietary fibers are associated to water absorption and intestinal regulation apart from the well-known prebiotic and health benefits (Zhu et al., 2015). Consequently, many studies are still ongoing in order to develop nutraceutical and fiber-rich products (Han et al., 2017). NSPs play a crucial role in the food systems in which they are present because they have the ability to control the mobility of free water or to form a three-dimensional network structure in foods. In addition, NSPs’ fraction is capable to create links with other components that exist in the food matrix thus influencing the structure, physical functionality, sensory attributes, and nutritional value of the foods (Goh et al., 2014; Venzon et al., 2015). It is a known fact that the rate at which a product with high starch content is digested can affect the blood glucose level, thus creating major implications for human health. The introduction of NSPs into starch-rich foods can have a positive influence on the reduction of blood glucose levels (Chiu et al., 2011). According to their composition, the pectic polysaccharides recovered from fruit and vegetable byproducts can be used both as food additives due to their physical properties and also as functional foods for their bioavailability and bioactivity in the human body (Mu¨ller-Maatsch et al., 2016). In terms of chemical structure, pectin consists of D-galacturonic acid polymeric units and there are two dominant types in bioresources: homo- and rhamnogalacturonans. The first polymer consists of a backbone of α-1,4-linked

46

Separation of Functional Molecules in Food by Membrane Technology

galacturonic acid residues. Homogalacturonan molecules are surrounded by numerous hydroxyl groups that provide the ability to form hydrogen bonds, whereas these molecules are negatively charged due to the demethylation of carboxylic groups. The rhamnogalacturonan backbone contains fewer carboxylic groups compared to the homogalacturonan structure since it is composed of repeated α-L-rhamnose-(1-4)-α-D-galacturonic acid units (Galanakis, 2015a). Pectin molecules are also classified into two groups, high methoxyl pectin and low methoxyl pectin, depending on the degree of esterification (Quoc et al., 2015). The latest affects importantly their technological applications and especially their gelling properties. β-Glucans (soluble dietary fiber with advanced gelling properties, too) are linear homopolysaccharides composed of continuant (1,4)-linked β-D-glucose segments. The latest are separated by single (1,3) linkages. β-Glucan is water soluble due to the presence of the β-(1,3)-linked β-glycosyl residue, which prevents alignment of glucose segments and increases the corresponding solubility (Lazaridou and Biliaderis, 2007). Currently, apple pomace, citrus peels, and to some extent sugar beet pulp are used as sources for commercial pectin production (Di Donato et al., 2014; Poli et al., 2011). These pectic compounds are not only present in high amounts in citrus peels but have also important functionality due to the presence of associated bioactive compounds. Thus, the press cake, beside cell walls polysaccharides, contains B70% of polyphenols originally present in the raw materials (Hilz et al., 2005). Many studies have highlighted that defatted oil bearing materials (such as sesame husk, rice bran, and flaxseed) give interesting health benefits. Compositional analysis reveals that sesame husk, rice bran, and flaxseed consist of almost 68%, 27%, and 39% dietary fiber respectively and have been reported to have positive health effects (e.g., laxative, cholesterol-lowering agent). An added value of the polysaccharides extracted from sesame husk, rice bran, flax seed, or even oilpumpkin biomass is the high antioxidant activity, which makes them suitable functional ingredients to be considered for the development of new dietary supplements and functional foods (Koˇsta´lova´ et al., 2013; Nandi and Ghosh, 2015; Nosa´lova´, et al., 2011). Moreover, the pectic polysaccharide fractions extracted from pea pods were rich in uronic acid (70%
95%); highly methylated (30%) and acetylated (10%); and rich in arabinose, xylose, and galactose, demonstrating a high purity and strong methylation degree (Mu¨ller-Maatsch et al., 2016). The heterogeneity of the pectin structure depends on the plant origin, the part of the plant where it is located (peels, pulp, seed, etc.), and how it is extracted. Nevertheless, the data obtained in the study conducted by Mu¨llerMaatsch et al. (2016), on 26 vegetable wastes, showed that the structure of the pectin extracted from waste is similar to that from the raw matrices, although the methylation and acetylation degrees are lower due to the processing and/or enzymatic actions.

Introduction in Functional Components Chapter | 2

47

There are several classical ways to extract pectin from different materials, all of them using traditional heating. The physicochemical process involves hydrolysis and extraction of pectin macromolecules from plant tissue, purification of the liquid extract, and isolation of the extracted pectin from the liquid, steps that are influenced by various factors, mainly temperature, pH, and time. In order to obtain better yields and desired functionality, the isolation, purification, and characterization techniques need to be progressively improved. Compared with the conventional methods, microwave extraction proves to be more efficient in terms of extraction yields and time and solvent consumption (Seixas et al., 2014; Quoc et al., 2015). Also, subcritical water extraction method, used for citrus and apple pectin extraction, has proved to be effective, generating a good yield (22% and 17%, respectively). In addition, the extracted pectin showed a high antioxidative and antitumor activity (Wang et al., 2014). In another study ultrahigh-pressure extraction, as an emerging novel technology, was applied to extract pectin from orange peel. The obtained results demonstrated that ultrahigh-pressure was an efficient, timesaving, and ecofriendly method, the extraction yield and stability being significantly higher than those of traditional heating and microwave extraction (Guo et al., 2012). The ultrafiltration technology can be employed to purify pectin extracted from fruit byproducts and β-glucan recovered from cereal byproducts (Patsioura et al., 2011; Galanakis et al., 2013). Like in the case of proteins, the MW of both pectin and β-glucan moieties also ranges from decades to hundreds of kDa. The large size of these molecules restricts their permeation through small pores, while the surrounding hydroxyl groups form hydrogen bonds with the water and hydrophilic membranes (Galanakis, 2015a). Qiu et al. (2009) used five types of ultrafiltration membranes with different molecular weight cutoffs to separate apple pectin with different molecular weights. The results showed that galacturonic acid contents and esterification degrees increase with an increase in molecular weight as well as the monosaccharide composition (Qiu et al., 2009). The efficiency of membrane filtration processes was highlighted also in the case of pectin isolated from an extract prepared from mature citrus peel: the use of a crossflow microfiltration (MW) step contributed to a 75% saving in the solvent consumption, and although the pectin recovery yield decreased from 10.5% to 9.9 %, the galacturonic acid content increased from 68.0% to 72.2% (Li and Chase, 2010). In addition, by ultrafiltration, the pigments and impurities can be removed due to the absorptive function. The soluble pigments in the sunflower head are strongly associated with the pectin extract and brought undesired color to the final products. In this regard, by purifying the extracted pectin by the ultrafiltration process, approximately 50% of pigments and 75% of salts have been removed (Kang et al., 2015).

48

Separation of Functional Molecules in Food by Membrane Technology

Brewer’s spent grain, the main waste from the beer production process, is another valuable source of carbohydrates, their level being up to 50% of the byproduct weight. The main carbohydrates in brewer’s spent grain are cellulose (B17% dw) and hemicelluloses, mainly arabinoxylan (25
28% dw) (Niemi et al., 2012a; Vieira et al., 2014). The easiest way to exploit brewer’s spent grain as a functional ingredient involves drying and converting it into flour. Thus, by adding brewer’s spent grain flour to various food products (e.g., bakery products, extruded products, and meat products) an improvement in their fiber, protein, mineral content, and water holding capacity can be achieved (Farcas et al, 2017; Nagy et al., 2017; Stojceska et al., 2008). Being a material rich in hemicelluloses, brewer’s spent grain can be subjected to various hydrolysis processes in order to release the monosaccharides (xylose and arabinose), which can be further fermented in order to generate valuable products (e.g., xylitol, a sweetener used in food industry) (Mussatto and Roberto, 2005). Also, arabinoxylans, which are considered a dietary fiber with many potential applications as functional ingredients, can be extracted from brewer’s spent grain under strong alkaline conditions and also by using an innovative fully integrated process that sequentially extracts the proteins and arabinoxylans (Vieira et al., 2014).

2.3.4 Lipids Most fresh fruits and vegetables exhibit a low content of fat; in contrast, dry vegetables and many fruit seeds and kernels contain large amounts of (often discarded) high-value oils that are potentially useful for incorporation in functional foods, cosmetics, and pharmaceutical products (Femenia, 2007). Polyunsaturated fatty acids (PUFAs) and phospholipids as well as the lipid-soluble minor components such as carotenoids, tocopherols, and phytosterols have become the focus of lipid-based nutraceutical products. The main source of PUFA is fish oils whereas the main source of phospholipids, carotenoids, tocopherols, and phytosterols is vegetable oils and other plant materials (Akin et al., 2012) According to the World Health Organization (WHO, 2010), saturated fatty acid consumption, mainly myristic and palmitic acids, is directly related to the risk of cardiovascular diseases. Therefore, diets should provide an adequate supply of polyunsaturated fatty acids and oleic acid. Rice bran, a low-value coproduct obtained from rice processing, could represent a potential source of healthy products due to the unique antioxidant and nutraceutical complex present in its composition (Oliveira et al., 2012). Rice bran contains almost 12%
18.5% oil, with a well-balanced fatty acid profile in terms of unsaturated-to-saturated fatty acids ratio and the ratio of diunsaturated (linoleic acid) to saturated (palmitic acid) and monounsaturated (oleic acid) fatty acids. Rice bran contains a range of fats, of which 47% are monounsaturated, 33% polyunsaturated, and 20% saturated. Also, it

Introduction in Functional Components Chapter | 2

49

comprises around 4.3% highly unsaponifiable components like tocotrienols (a form of vitamin E), gamma oryzanol, and betasitosterol (Marcet et al., 2016). Tocols have proven effectiveness in preventing cardiovascular disease and some forms of cancer, whereas oryzanol, a mixture of triterpene alcohols and phytosterols esterified with ferulic acid, has shown hypocholesterolemic activity being effective in decreasing early atherosclerosis (Pal and Pratap, 2017). By using ethanol as extraction solvent, it was possible to obtain values from 123 to 271 mg of tocols/kg of fresh rice bran and 1527 to 4164 mg of oryzanol/kg of fresh rice bran, indicating that it is feasible to obtain enriched oil when the proper solvent is used (Oliveira et al., 2012). Many marine microalgae strains have oil contents of between 10% and 50% (w/w) and produce a high percentage of total lipids (up to 30%
70% of dry weight). They contain a considerable amount of high-quality oils, partly consisting of omega-3 and omega-6 fatty acids, which can be used as bioactive ingredients in the development of functional foods. Long-chain polyunsaturated fatty acids (LC-PUFAs) such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are considered essential omega-3 fatty acids in human nutrition, microalgae being considered a valuable vegetal source for the extraction of these compounds. Microalgae, and supplements derived from it, are excellent alternative sources of EPA, DHA, and other fatty acids, since fish often contain toxins due to pollution (Oviyaasri et al., 2017). Research underway in the palm oil industry has revealed a range of bioactives that can be used as functional components for food products that enhance health. Palm pressed fibers contain extractable phytochemicals such as sterols, vitamin E, carotenoids, phospholipids, squalene, and phenolics that are earmarked for applications in functional foods, nutraceuticals, pharmaceuticals, and the cosmetics industry. Although oil is the primary product from the palm, various byproducts are generated during the extraction (Tan et al., 2007).

2.3.5 Bioactive Compounds of Animal Origin Meat and meat products represent an important segment of the human diet because they provide essential nutrients that cannot be easily obtained from vegetables and their derived products. Over the last decade, the amount of byproducts generated from the animal products industry has continued to witness tremendous growth. These byproducts, which are usually perceived as waste, can in fact be further reprocessed for extraction of valuable bioactive compounds (Alao et al., 2017). The major sources of animal waste are represented by the meat, fish, and poultry industries, this material including slaughterhouse derivatives that cannot be sold, such as organs and other visceral mass, feathers, heads,

50

Separation of Functional Molecules in Food by Membrane Technology

bones, fat and meat trimmings, blood and other fluids, wastes from seafood, skins, and wastes from dairy processing such as curd, whey, and milk sludge from the separation process (Bordenave et al., 2002; Durham and Hourigan, 2007). Meat byproducts are rich in lipids, carbohydrates, and proteins and can be subjected to different processes in order to recover nutrients (Shi et al., 2017). In addition, bioactive peptides can be created from meat proteins using different types of hydrolysis and purification procedures. These bioactive compounds are known to have antimicrobial, antioxidative, antithrombotic, antihypertensive, anticarcinogenic, satiety regulating, and immunomodulatory activities and may affect the cardiovascular, immune, nervous, and digestive systems. Peptides may also be effective in the treatment of mental health diseases, cancer, diabetes, and obesity (Helkar et al., 2016; Lafarga and Teagase, 2014; Marcet et al., 2016). Animal blood is a byproduct from meat processing and contains a high level of protein and iron that can be reclaimed and reused. Also, the plasma protein
HEC complex was reported to contain a large amount of essential amino acids and electrophoretic separation of plasma protein resulted in complex albumin forming the major fraction (Wan et al., 2002). The recovery of protein from red cells of bovine blood by ultrafiltration indicated that the obtained fraction had a good amino acid profile for use as an ingredient in formulated foods for human consumption (Roupas et al., 2007). Collagen is another component that can have nutraceutical and biomedical applications since microporous collagen films are used for delivery of anticancer drugs and collagen matrices are used as gene delivery agents that promote bone and cartilage formation. Also, gelatin is found to be a potent antioxidant and antihypertensive and to enhance dietary calcium absorption. Today there is an increasing demand for functional molecules, which promotes fish raw materials and utilization of the byproducts as raw materials for food, nutraceutical, pharmaceutical, and biotechnological applications. The fish industry is one of the most problematic food industries, since around a 40%
50% of total weight of the animal is considered to be waste (Kaspar and Reichert, 2013). The popularity of fish oil dietary supplements has been steadily growing due to their high content of omega-3 polyunsaturated fatty acids (ω-3 PUFAs) such as eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) (Rizliya and Mendis, 2014; Shi et al., 2017). The modern diet is insufficient in omega fatty acids, hence the intake of oily fish two to three times a week is recommended. Incorporation of fish oils into normal food ingredients can be considered as an alternative in order to increase the intake of polyunsaturated fatty acids, especially EPA and DHA. Nutritionally, EPA and DHA are the most important components of the omega-3 fatty acids family, as they have significant human health benefits. There is already plenty of scientific evidence proving that these compounds have a positive impact on human health by reducing the risk of

Introduction in Functional Components Chapter | 2

51

cardiovascular disease, cancer, diabetes, and depression, as well as improving the immune system, and ensuring proper neural development (Oviyaasri et al., 2017). The food products fortified with ω-3 PUFAs provide a way to achieve desired biochemical effects of these nutrients without the ingestion of dietary supplements, medications, or a major change in dietary habits. Also, the isoelectric solubilization/precipitation (ISP) allows efficient recovery of fish protein isolate that could be used in functional foods. Based on proximate composition, high protein recovery yield and high reduction of lipid in the recovered protein indicate that ISP processing effectively recovers nutrients from various representative species (Tahergorabi et al., 2012). Carnivorous fishes’ stomachs have a high pepsin content that can be separated by ultrafiltration, concentration, and spray-drying. Fish skin and bones are also potential sources of collagen and gelatin. Collagen is obtained by acid treatment of the byproducts whereas gelatin is derived from fish skin by enzymatic hydrolysis (Gildberg, 2004). The byproducts generated by the fish industry can also be exploited as sources of bioactives. The fish frame resulting after the filleting of fishes is a complete source of protein because it contains appreciable levels of muscle protein with essential amino acids. Moreover, the fish protein isolate can be used for the isolation of bioactive peptides, which have manifold applications in functional foods and pharmaceutical products due to their antihypertensive or angiotensin-converting enzyme (ACE) inhibition, antiproliferative, anticoagulant, immunomodulatory, and chelating effects (Atef and Ojagh, 2017). Another source of animal-derived food waste is the dairy industry. Whey, the principal byproduct from dairy processing, contains 95% of the original water, most of the lactose, 20% of the milk protein, and traces of fat (Russ and Meyer-Pittroff, 2004). The problem with whey utilization is that a very large volume of whey is produced worldwide each year and it contains only dilute concentrations of these valuable proteins and other biochemicals (Ravindran and Jaiswal, 2016). Recently, methods based on membrane technologies have been developed for the treatment of wastewater and aqueous food systems (Castro-Munoz et al., 2015). An example is the ultrafiltration of milk for cheese manufacture, which increases the recovery of whey proteins in the cheese and reduces whey volumes (Durham and Hourigan, 2007). It was found that column chromatography and membrane separation remain the most commonly used techniques for whey protein fractionation. Better results, particularly in terms of purity levels, were obtained with the coupling approach involving a combination of more than one technique (El-Sayed and Chase, 2011). The whey fractionation process and ultrafiltration of permeate impose greater processing demands, but result in a wide range of products with unique functionalities and high value. Fractionated whey products include whey protein concentrate; whey protein isolate; α and β fractions; bioactive

52

Separation of Functional Molecules in Food by Membrane Technology

proteins such as immunoglobulins, lactoferrin, lactoperoxidase, and glycomacropeptide; bioactive peptides; and minerals. These products have an excellent potential for use in the development of new functional foods and nutraceuticals (Durham and Hourigan, 2007) Also, whey proteins have demonstrated different physiological functions due to their numerous bioactive peptides that exhibit various properties such as antioxidative, antihypertensive, antimicrobial, immunoregulatory, angiotensinconverting enzyme inhibition, and mineral carrying capacity. These bioactive compounds can be produced by enzymatic hydrolysis of the protein coupled with various membrane separation procedures that can be used to fractionate the peptides based on size (Korhonan and Pihlanto, 2003). In order to recover all the bioactive fractions dairy processors and researchers continue to develop separation technologies based on size, density, charge, or hydrophobicity of the different components. In this respect different techniques are often combined in order to develop elaborate processes of sequential fractionation and purification and increase the recovery yield of high-value byproducts (Durham and Hourigan, 2007; El-Sayed and Chase, 2010a,b; Ndiaye et al., 2010). Thereby, the “bulk” whey proteins such as immunoglobulin (150
1000 kDa and 5.5 , pI , 8.3) and ovine serum albumin (66 kDa and pI 5 5.0 6 0.1) are positive in the acidic (pH 5 4.8 6 0.1) whey solutions, whereas for the smaller α-lactalbumin (14 kDa and pI 5 4.5 6 0.3) and β-lactoglobulin (18 kDa and pI 5 5.3 6 0.1), charge is expected to be weakly negative and positive, respectively (Galanakis, 2015a). For instance, immunoglobulin and ovine serum albumin have been quantitatively recaptured with a 300-kDa tubular ceramic membrane and pH 5 4 (Alme´cija et al., 2007). The selective recovery of a particular protein (i.e., α-lactalbumin) can be conducted by combining several effects, including reduced pore size and enhanced electrostatic repulsion between the charged membrane and protein species (Cowan and Ritchie, 2007).

2.4 SEPARATION AND RECOVERY OF MACRO- AND MICROMOLECULES USING MEMBRANE TECHNOLOGIES The most common methodology for the recovery of functional compounds from the aforementioned conventional and nonconventional bioresources is the so-called 5-Stages Universal Recovery Process, which progresses as follows (Galanakis, 2012, 2015a): 1. 2. 3. 4. 5.

macroscopic pretreatment, separation of macro- and micromolecules extraction isolation-purification product formation or encapsulation

Introduction in Functional Components Chapter | 2

53

MF, ultrafiltration (UF), nanofiltration (NF), and electrodialysis (ED) are physicochemical and most importantly nondestructive technologies that could be used in the four first steps this methodology. Indeed, a usual practice is to concentrate bigger molecules (macromolecules) in the retentate and release smaller (micromolecules) in the permeate stream, respectively (Galanakis, 2015b). This approach looks simple at a theoretical level as the main fractionation mechanism of membranes is the “sieving” one that separates the molecules according to their MW. Nevertheless, in practice, this is not always the case since membrane pores are asymmetric, while the MWCO reflects the mean size of all membrane pores. To this line, the sieving effect is not always working strictly at the molecular level and indeed attenuates when the hydrophobic nature of the membrane surface and the solubility of the solutes are incorporated (Pinelo et al., 2009; Galanakis, 2015b). Subsequently, MWCO (although being an important factor) does not reflect an absolute barrier for the separation of macro- and micromolecules. Another issue to deal with is that functional molecules of natural sources do not move freely in solutions, but exist trapped within clusters of bioresource matrices. A typical example is phenols that bind noncovalently to proteins (Rawel et al., 2005) and dietary fibers (Bravo et al., 1994) of bioresources. In other words, smaller molecules like antioxidants could be trapped and recaptured in the retentate due to the structural characteristics of the macromolecules existing in the feed, and macromolecules could pass to the permeate due to the local structure of membrane pores in some points. If macroscopic pretreatment and extraction procedures are conducted prior the application of UF (e.g., via vacuum or thermal concentration, acid or enzyme extractions, etc.), the bioresource clusters could break, but also macromolecules may decompose leading to the formation of oligopolymer structures. This process affects the sieving mechanism and separation performance of UF membranes. On the other hand, if macro- and micromolecule complexes (i.e., whey proteins with phenols) have not been separated prior to the UF procedure, smaller molecules are covered in the concentrate stream (Galanakis, 2015a). The simultaneous recapture of macro- and micromolecules in one stream is a yield problem that leads to additional and costly purification stages. However, this fact may be desirable depending on the product that the food technologist is willing to develop (Galanakis, 2015b). In any case, researchers and professionals are nowadays preferring more and more membrane applications in order to separate all kind of functional compounds that exist in foods, for example, proteins, pectin, β-glucan, polyphenols, anthocyanins, tannins, flavonoids, carbohydrates and sugars in fruit juices, solutions, agricultural wastewaters, and beverages (Dı´az-Reinoso et al., 2009; Garcı´a-Martı´n et al., 2010; Galanakis et al., 2010; Kuhn et al., 2010; Cassano et al., 2013; He et al., 2013; Li et al., 2013).

54

Separation of Functional Molecules in Food by Membrane Technology

Galanakis (2015a) revised five research studies (Galanakis et al., 2010, 2013, 2014, 2015; Patsioura et al., 2011) dealing with membrane separation mechanisms of macro- and micromolecules under similar processing conditions. These studies were conducted using a broad range of MWCO for the membranes, starting with UF and 100 kDa to the border of NF and 1 kDa. Experiments were conducted using two crossflow UF systems (DSS Labstak M20 and M10), seven membrane materials (GR40PP-100 kDa, GR51PP50 kDa, GR60PP-25 kDa, GR70PP-20 kDa, GR81PP-10 kDa, GR95PP10 kDa, and ETNA01PP-1 kDa) of the same manufacturer (Alfa Laval Nakskov), and different feeds extracted from food wastes and beverages. According to this review, separations were discussed in three categories according to the MW of the solutes (100
50, 25
10, and 2
1 kDa), providing insights about the separation mechanism of functional macro- and micromolecules using membrane technology. Thereby, by applying wide pore membranes of 100
50 kDa, the sieving mechanism dominates the separation of macro- and micromolecules making it rather distinct (Table 2.2). Experimenting with narrower membrane pores of 25
10 kDa (Table 2.3) and 2
1 kDa (Table 2.4), the separation based on molecular cutoff becomes more difficult and progresses in terms of components solubility, membrane hydrophobicity, and polarity resistance. The latest phenomenon leads to concentration polarization and fouling caused by the aggregation of proteins and polysaccharides on the membrane surface, or gel formation. However, separation opportunities for particular applications exist for the whole range (1
100 kDa) of the tested MWCO. The application of membranes such as polysulfone, which has a more asymmetric cutoff profile, can provide a degree of flexibility to the assayed separation. At this case, macromolecules can step in “gaps” and pass through membranes pores, whereas small polar molecules can stick in the polar membrane parts and get adsorbed on them. Polysulfone membranes in the range of 20
25 kDa were shown to be very efficient for the separation of: 1. polymeric from monomeric anthocyanins 2. pectin from phenolic compounds and cations 3. removal of “heavier” phenolic classes (autoxidated or compounds linked to macromolecules) without affecting the overall antioxidant properties of the permeate The utilization of a less hydrophobic and narrower membrane such as polyethersulfone of 10 or 2 kDa was shown to be reversely efficient. In particular, this membrane was able to absorb rapidly polar micromolecules such as hydroxycinnamic acid derivatives and at the same time to release oligosaccharide pectin fragments in the permeate. On the other hand, the more hydrophobic composite fluoropolymer was able to separate efficiently phenolic classes like hydroxycinnamic acids from flavonols (Galanakis, 2015a).

TABLE 2.2 Concentrations of Macro- and MicroMolecules Originating From Different Sources and Corresponding Retention Percentages Obtained Using Ultrafiltration Membranes (100
50 kDa) Substrate

Feed (Target Compounds)

Membrane Barrier MWCOa (kDa)

Material

Compounds Macromolecules

References Micromolecules

Group

C (mg/L)

R (%)

Group

Standard

Solution (β-glucan)

100

PSb

β-glucan

200
2000

92
95

n.d.c

Oat mill waste

Extract (β-glucan)

100

PSb

β-glucan

302
442

53
67

Saccharidesd

2317
5344

4
11

Proteins

190
376

35
40

Total phenols

16
31

3
9

Monov. ionse

1057
1699

2
3

Total sugars

384

,1

Total phenols

280

,1

o-diphenols

57

,1

Hyd.-cin. acidsf

19

,1

Flavonols

18

10

Monov. ionse

122

,1

Antirad. effic.g

1.7g

,1

Total phenols

68

13

Monov. ionse

1304

2

Olive millwastewater

AIRh from olive mill wastewater

Beverage (phenol)

Water soluble extract (pectin)

100

100

PSb

PSb

n.d.c

Pectin

87

79

C (mg/L)

R (%) Patsioura et al. (2011) Patsioura et al. (2011)

Galanakis et al. (2010)

Galanakis et al. (2010)

(Continued )

TABLE 2.2 (Continued) Substrate

Winery sludge

Feed (Target Compounds)

Extract (phenols and anthocyanins)

Membrane Barrier a

MWCO (kDa) 100

Compounds

Material

PS

b

References

Macromolecules

Micromolecules

Group

C (mg/L)

R (%)

Pectin

6443

12

Polanthoc.i

172

59

Group

C (mg/L)

R (%) Galanakis et al. (2013)

Total sugars

3910

61

Red. sugarsj

412

50

Nonred. sugarsk

3498

62

Total phenols

1965

64

o-diphenols

560

52

Hyd.-cin. acidsf

297

57

265

43

Flavonols l

Total anth.

249

61

Monom. anth.m

76

59

Diluted extract (phenols and anthocyanins)

100

PSb

Pectin

1670

6

Total sugars

1065

74

Polanthoc.h

172

77

Red. sugarsj

129

58

Nonred. sugarsk

935

76

Total phenols

476

69

o-diphenols

560

80

Hyd.-cin. acidsf

297

99

265

68

Flavonols l

Anari cheese

Halloumi cheese

Whey (protein and sugars)

Whey (protein and sugars)

100

100

PSb

PSb

Proteins

Proteins

3723

1087

76

69

Total anth.

249

61

Monom. anth.m

76

77

Total sugars

49,703

9

Red. sugarsj

469

86

Nonred. sugarsk

48,581

7

Total phenols

112

78

Total sugars

12,729

2

Red. sugarsj

182

50

Nonred. sugarsk

12,533

2

Total phenols

43

55

Galanakis et al. (2013)

Galanakis et al. (2014)

Galanakis et al. (2014)

(Continued )

TABLE 2.2 (Continued) Substrate

Anari cheese

Halloumi cheese

a

Feed (Target Compounds)

Whey (protein and sugars)

Whey (protein and sugars)

Membrane Barrier a

MWCO (kDa) 50

50

Compounds

Material

PS

b

PSb

Macromolecules

References Micromolecules

Group

C (mg/L)

R (%)

Group

C (mg/L)

R (%)

Proteins

3723

73

Total sugars

49703

8

Red. sugarsj

469

82

Nonred. sugarsk

48,581

7

Total phenols

112

74

Total sugars

12,729

18

Red. sugarsj

182

33

Nonred. sugarsk

12,533

18

Total phenols

43

66

Proteins

1087

68

Galanakis et al. (2014)

Galanakis et al. (2014)

MWCO, molecular weight cutoff. PS, polysulfone. n.d., not determined. d Saccharides include oligo-, di-, and monosaccharides. e Monov. ions, monovalent ions. f Hyd.-cin. acids, hydroxycinnamic acid derivatives. g Antirad. effic., antiradical efficacy expressed in mg DPPH/g. h AIR, alcohol insoluble residue. i Pol-anthoc., polymeric anthocyanins. j Red. sugars, reducing sugars. k Nonred. sugars, nonreducing sugars. l Total anth., total anthocyanins. m Monom. anth., monomeric anthocyanins. Source: Adapted from Galanakis, C.M., Kotanidis, A., Dianellou, M., & Gekas, V. (2015). Phenolic content and antioxidant capacity of Cypriot Wines. Czech J. Food Sci. 33(2), 126
136. b c

TABLE 2.3 Concentrations of Macro- and Micromolecules Originated From Different Sources and Corresponding Retention Percentages Obtained Using Ultrafiltration Membranes (25
10 kDa) Substrate

Olive mill wastewater

Feed (Target Compounds)

Beverage (phenol)

Membrane Barrier MWCOa (kDa) 25

Compounds

Material

Macromolecules Group

PS

b

C (mg/L)

References Micromolecules

R (%)

c

n.d.

Group

C (mg/L)

R (%)

Total sugars

384

18

Total phenols

280

10

o-Diphenols

57

6

Hyd.-cin. acidsd

19

32

18

37

Flavonols Monov. ions

e

122 f

g

AIR from olive mill wastewater

Water soluble extract (pectin)

25

PS

b

Pectin

87

98

f

Galanakis et al. (2010)

26

Antirad. effic.

1.7

8

Total phenols

68

40

Monov. ionse

1304

10

Galanakis et al. (2010)

(Continued )

TABLE 2.3 (Continued) Substrate

Winery sludge

Feed (Target Compounds)

Extract (phenols and anthocyanins)

Membrane Barrier MWCO (kDa) 20

a

Compounds

Material

PS

b

References

Macromolecules

Micromolecules

Group

C (mg/L)

R (%)

Group

C (mg/L)

R (%)

Pectin

6443

64

Total sugars

1065

87

Polanthoc.h

172

92

Red. sugarsi

129

78

Nonred. sugarsj

935

88

Total phenols

476

85

o-diphenols

560

87

Hyd.-cin. acidsd

297

81

265

65

249

89

Monom. anth.

76

86

Flavonols k

Total anth.

l

Diluted extract (phenols and anthocyanins)

20

PS

b

Pectin

1670

52

Total sugars

1065

78

Polanthoc.h

172

94

Red. sugarsi

129

78

Nonred. sugarsj

935

78

Total phenols

476

77

o-diphenols

560

85

Galanakis et al. (2013)

Galanakis et al. (2013)

Hyd.-cin. acidsd

297

99

265

75

249

62

Monom. anth.

76

17

Total sugars

49,703

21

Red. sugarsi

469

74

Nonred. sugarsj

48,581

22

Total phenols

112

78

Total sugars

12729

34

Red. sugarsi

182

54

Nonred. sugarsj

12,533

35

Flavonols k

Total anth.

l

Anari cheese

Anari cheese

Olive mill wastewater

Whey (protein and sugars)

Whey (protein and sugars)

Beverage (phenol)

20

20

10

PS

PS

b

b

PESf

Proteins

Proteins

n.d.c

3723

1087

84

76

Total phenols

43

59

Total sugars

384

32

Total phenols

280

21

o-Diphenols

57

32

Hyd.-cin. acidsd

19

44

18

56

Flavonols e

Monov. ions

f

Antirad. effic.

122

23

f

24

1.7

Galanakis et al. (2014)

Galanakis et al. (2014)

Galanakis et al. (2010)

(Continued )

TABLE 2.3 (Continued) Substrate

g

AIR from olive mill wastewater a

Feed (Target Compounds)

Water soluble extract (pectin)

Membrane Barrier MWCO (kDa) 10

a

Compounds

Material

f

PES

Macromolecules

References Micromolecules

Group

C (mg/L)

R (%)

Group

C (mg/L)

R (%)

Pectin

87

98

Total phenols

68

71

Monov. ionse

1304

49

Galanakis et al. (2010)

MWCO, molecular weight cutoff. PS, polysulfone. n.d., not determined. d Hyd.-cin. acids, hydroxycinnamic acid derivatives. e Monov. ions, monovalent ions. f Antirad. effic., antiradical efficacy expressed in mg DPPH/g. g AIR, alcohol insoluble residue. h Pol-anthoc., polymeric anthocyanins. i Red. sugars, reducing sugars. j Nonred. sugars, nonreducing sugars. k Total anth., total anthocyanins. l Monom. anth., monomeric anthocyanins. Source: Adapted from Galanakis, C.M. (2015a). Chapter 3: Development of a universal recovery strategy. In: Galanakis, C.M. (Ed.), Food Waste Recovery: Processing Technologies and Techniques. Elsevier Inc.: Waltham; Galanakis, C.M. (2015b). Separation of functional macromolecules and micromolecules: from ultrafiltration to the border of nanofiltration. Trends Food Sci. Technol. 42, 44
63. b c

TABLE 2.4 Concentrations of Macro- and Micromolecules Originated From Different Sources and Corresponding Retention Percentages Obtained Using Membranes in the Edge of Ultrafiltration With Nanofiltration (2
1 kDa) Substrate

Olive mill wastewater

Feed (Target Compounds)

Beverage (phenol)

Membrane Barrier a

MWCO (kDa) 2

Compounds

Material

Macromolecules Group

b

PES

References

C (mg/L)

Micromolecules R (%)

c

n.d.

Group

C (mg/L)

R (%)

Total sugars

384

38

Total phenols

280

25

57

48

19

53

Flavonols

18

62

Monov. ionse

122

27

Total phenols

68

81

1304

55

o-Diphenols Hyd.-cin. acids

f

b

PES

Pectin

87

99

d

Galanakis et al. (2010)

AIR from olive mill wastewater

Water soluble extract (pectin)

2

Winery sludge

Extract

2

PESb

n.p.g

n.p.g

Galanakis et al. (2013)

Diluted extract

2

PESb

n.p.g

n.p.g

Galanakis et al. (2013)

e

Monov. ions

Galanakis et al. (2010)

(Continued )

TABLE 2.4 (Continued) Substrate

Anari cheese

Feed (Target Compounds)

Ultrafiltered whey (protein and sugars)

Membrane Barrier MWCOa (kDa) 2

Compounds

Material

b

PES

Macromolecules

References Micromolecules

Group

C (mg/L)

R (%)

Group

C (mg/L)

R (%)

Proteins

652

35

Total sugars

38,402

13

135

11

Nonred. sugars

37,805

13

Total phenols

27

9

Total sugars

8212

5

81

11

Nonred. sugars

7668

5

Total phenols

16

31

Red. sugarsh i

Halloumi cheese

Ultrafiltered whey (protein and sugars)

2

b

Proteins

PES

275

47

Red. sugarsh i

Winery sludge

Extract (phenols and anthocyanins)

1

j

ETNA

Pectin

6443

47

Total sugars

1065

76

Polanthoc.k

172

56

Red. sugarsh

129

66

Nonred. sugarsi

935

77

Total phenols

476

74

560

67

Hyd.-cin. acids

297

77

Flavonols

265

45

Total anth.l

249

57

76

55

o-diphenols d

Monom. anth.

m

Galanakis et al. (2014)

Galanakis et al. (2014)

Galanakis et al. (2013)

Diluted extract (phenols and anthocyanins)

1

ETNAj

Pectin Polanthoc.k

1670 172

39

Total sugars

45

h

1065

64

Red. sugars

129

57

Nonred. sugarsi

935

65

Total phenols

476

56

o-diphenols

560

64

297

80

265

53

249

44

Monom. anth.

76

41

Total sugars

38,402

24

135

62

Nonred. sugars

37,805

23

Total phenols

27

36

Total sugars

8212

23

Red. sugarsh

81

21

Nonred. sugarsi

7668

23

Total phenols

16

23

Hyd.-cin. acids

d

Flavonols l

Total anth.

m

Anari cheese

Ultrafiltered whey (protein and sugars)

1

j

ETNA

Proteins

652

24

Red. sugarsh i

Ultrafiltered whey (protein and sugars)

1

j

ETNA

Proteins

275

42

Galanakis et al. (2013)

Galanakis et al. (2014)

Galanakis et al. (2014)

(Continued )

TABLE 2.4 (Continued) Substrate

Dry red wine

Feed (Target Compounds)

Sixfold diluted (phenolics and anthocyanins)

Membrane Barrier a

MWCO (kDa) 1

Compounds

Material

Macromolecules Group

j

ETNA

References

c

n.d.

C (mg/L)

Micromolecules R (%)

Group

C (mg/L)

Total phenols

1930
4413

69
90

Hyd.-cin. acidsd

121
444

42
53

Flavonols

118
369

9
40

75
559

21
71

Total anth.l n

n

28
63

Antirad. effic.

o

Sweet red wine

Sixfold diluted (phenolics and anthocyanins)

1

ETNA

c

n.d.

466
806

66
85

730

65

87

67

Flavonols

81

25

Total anth.l

19

26

Antirad. effic.n

6n

31

FRAP activityo

98o

60

Total phenols Hyd.-cin. acids

d

Galanakis et al. (2015)

11
57 o

FRAP activity j

R (%)

Galanakis et al. (2015)

Dry white wine

Sixfold diluted (phenolics and anthocyanins)

1

ETNAj

n.d.c

Total phenols

224

23

19

63

30

49

15

28

Antirad. effic.

4

n

29

FRAP activityo

95o

57

Hyd.-cin. acids Flavonols Total anth.l n

a

MWCO, molecular weight cutoff. PES, polyethersulfone. n.d., not determined. d Hyd.-cin. acids, hydroxycinnamic acid derivatives. e Monov. ions, monovalent ions. f AIR, alcohol insoluble residue. g n.p., no permeate. h Red. sugars, reducing sugars. i Nonred. sugars, nonreducing sugars. j ETNA, composite fluoropolymer. k Pol-anthoc., polymeric anthocyanins. l Total anth., total anthocyanins. m Monom. anth., monomeric anthocyanins. n Antirad. effic., antiradical efficacy expressed in mg DPPH/g. o FRAP activity expressed in μg TROLOX/mL. b c

d

Galanakis et al. (2015)

68

Separation of Functional Molecules in Food by Membrane Technology

2.5 CONCLUSIONS Taking into consideration all of the above, the future trend in the development of novel functional foods and nutraceutical products has already begun by accepting the concept of waste recovery and must continue through applying the most suitable extraction and purification technologies. The effective recovery of biologically-active components from different types of raw materials and byproducts is a constant challenge. In the recovery process it is necessary not only to ensure a good extraction yield at low cost and energy consumption, but also to ensure the purification of the bioactives by separating them from the complex matrix while maintaining their functional properties. Thus, success in developing new functional ingredients and innovative products is closely related to the efficiency of the separation/purification step of bioactive compounds. Membrane processing is one of the most promising technologies for the recovery of these valuable compounds from agroindustrial waste streams. Membrane systems, compared to the conventional methods, have several advantages, including versatility (e.g., depending on the membrane characteristics and on the used solvent, different compounds can be separated); mild operating conditions (e.g., minimal thermal damage); scalability at industrial level; environmental friendliness (e.g., low energy consumption); and economic feasibility (e.g., low cost).

REFERENCES Adil, I., Cetin, H., Yener, M., Bayindirli, A., 2007. Subcritical (carbon dioxide 1 ethanol) extraction of polyphenols from apple and peach pomaces, and determination of the antioxidant activities of the extracts. J. Supercrit. Fluids 43 (1), 55
63. Agostini, F., Bertussi, R.A., Agostini, G., dos Santos, A.C., Rossato, M., Vanderlinde, R., 2012. Supercritical extraction from vinification residues: fatty acids, α-tocopherol, and phenolic compounds in the oil seeds from different varieties of grape. Sci. World J. 790
486. Akin, O., Temelli, F., Ko¨seo˘glu, S., 2012. Membrane applications in functional foods and nutraceuticals. Crit. Rev. Food. Sci. Nutr. 52 (4), 347
371. Alao, B.O., Falowo, A.B., Chulayo, A., Muchenje, V., 2017. The potential of animal byproducts in food systems: production, prospects and challenges. Sustainability 9 (7), 1089. Allouche, N., Fki, I., Sayadi, S., 2004. Toward a high yield recovery of antioxidants and purified hydroxytyrosol from olive mill wastewaters. J. Agric. Food. Chem. 52 (2), 267
273. Alme´cija, M.C., Iba´n˜ez, R., Guadix, A., Guadix, E.M., 2007. Effect of pH on the fractionation of whey proteins with a ceramic ultrafiltration membrane. J. Memb. Sci. 288, 28
35. Angiolillo, L., Del Nobile, M.A., Conte, A., 2015. The extraction of bioactive compounds from food residues using microwaves. Curr. Opin. Food Sci. 5, 93
98. Atef, M., Ojagh, S.M., 2017. Health benefits and food applications of bioactive compounds from fish byproducts: a review. J. Funct. Foods 35, 673
681. Aydemir, L.Y., Go¨kbulut, A.A., Baran, Y., Yemenicioglu, A., 2014. Bioactive, functional and edible film-forming properties of isolated hazelnut (Corylus avellana L.) meal proteins. Food Hydrocolloids 36, 130
142.

Introduction in Functional Components Chapter | 2

69

Azabou, S., Abid, Y., Sebii, H., Felfoul, I., Gargouri, A., Attia, H., 2016. Potential of the solidstate fermentation of tomato by products by Fusarium solani pisi for enzymatic extraction of lycopene. LWT- Food Sci. Technol. 68, 280
287. Ballard, T.S., Mallikarjunan, P., Zhou, K., O’Keefe, S., 2010. Microwave-assisted extraction of phenolic antioxidant compounds from peanut skins. Food. Chem. 120 (4), 1185
1892. Banerjee, J., Singh, R., Vijayaraghavan, R., MacFarlane, D., Patti, A.F., Arora, A., 2017. Bioactives from fruit processing wastes: green approaches to valuable chemicals. Food. Chem. 225, 10
22. Beaulieu, L., Thibodeau, J., Bonnet, C., Bryl, P., Carbonneau, M.E., 2013. Detection of antibacterial activity in an enzymatic hydrolysate fraction obtained from processing of Atlantic rock crab (Cancer irroratus) byproducts. Pharm. Nutr. 1, 149
157. Bordenave, S., Fruitier, I., Ballandier, I., Sannier, F., Gildberg, A., Batista, I., et al., 2002. HPLC preparation of fish waste hydrolysate fractions. Effect on Guinea pig ileum and ACE activity. Prep. Biochem. Biotechnol. 32, 65
77. Bouallagui, Z., Han, J., Isoda, H., Sayadi, S., 2011. Hydroxytyrosol rich extract from olive leaves modulates cell cycle progression in MCF-7 human breast cancer cells. Food Chem. Toxicol. 49, 179
184. Bravo, L., Abia, R., Saura-Calixto, F., 1994. Polyphenols as dietary fiber associated compounds. Comparative study on in vivo and in vitro properties. J. Agric. Food Chem. 42, 1481
1487. Can-Cauich, C.A., Sauri-Duch, A., Betancur-Ancona, D., Chel-Guerrero, L., Gonza´lez-Aguilar, G.A., Cuevas-Glory, L.F., et al., 2017. Tropical fruit peel powders as functional ingredients: evaluation of their bioactive compounds and antioxidant activity. J. Funct. Foods 37, 501
506. Cassano, A., Conidi, C., Giorno, L., Drioli, E., 2013. Fractionation of olive mill wastewaters by membrane separation techniques. J. Hazard. Mater. 248
249, 185
193. ´ lvareza, C., Ya´n˜ez-Ferna´ndeza, J., 2015. Recovery of bioactive Castro-Munoz, R., Orozco-A compounds from food processing wastewaters by ultra and nanofiltration: a review. Adv. Biores. 6 (3), 152
158. Chiu, C.J., Liu, S., Willett, W.C., Wolever, T.M., Brand-Miller, J.C., Barclay, A.W., et al., 2011. Informing food choices and health outcomes by use of the dietary glycemic index. Nutr. Rev. 69 (4), 231
242. Cho, C.W., Lee, D.Y., Kim, C.W., 2003. Concentration and purification of soluble pectin from mandarin peels using crossflow microfiltration system. Carbohyd. Polym. 54, 21
22. Conidi, C., Cassano, A., Drioli, E., 2011. A membrane-based study for the recovery of polyphenols from bergamot juice. J. Memb. Sci. 375, 182
190. Conidi, C., Cassano, A., Drioli, E., 2012. Recovery of phenolic compounds from orange press liquor by nanofiltration. Food Bioprod. Process. 90, 867
874. Conidi, C., Lopez, R.A.D., Garcia-Castello, E.M., Cassano, A., 2015. Purification of artichoke polyphenols by using membrane filtration and polymeric resins. Sep. Purif. Technol. 144, 153
161. Cowan, S., Ritchie, S., 2007. Modified polyethersulfone (PES) ultrafiltration membranes for enhanced filtration of whey proteins. Sep. Sci. Technol. 42, 2405
2418. Dalmolin, I., Mazutti, M.A., Batista, E.A.C., Meireles, M.A.A., Oliveira, J.V., 2010. Chemical characterization and phase behavior of grape seed oil in compressed carbon dioxide and ethanol as co-solvent. J. Chem. Therm. 42 (6), 797
801. Day, L., Seymour, R.B., Pitts, K.F., Konczak, I., Leif, L., 2009. Incorporation of functional ingredients into foods. Trends Food Sci. Technol. 20, 388
395.

70

Separation of Functional Molecules in Food by Membrane Technology

Di Donato, P., Poli, A., Taurisano, V., Nicolaus, B., 2014. Polysaccharides: applications in biology and biotechnology/Polysaccharides from bioagro-waste new biomolecules-life. Polysaccharides 1
29. Dı´az-Reinoso, B., Moure, A., Domı´nguez, H., Parajo´, J.C., 2009. Ultra- and nanofiltration of aqueous extracts from distilled fermented grape pomace. J. Food Eng. 91, 587
593. Durham, R.J., Hourigan, J.A., 2007. Waste management and co-product recovery in dairy processing. In: Waldron, K. (Ed.), Handbook of Waste Management and Co-product Recovery in Food Processing. Woodhead Publishing Limited, Cambridge, England, pp. 332
328. El-Sayed, M.H., Chase, H., 2011. Trends in whey protein fractionation. Biotechnol. Lett. 33, 1501
1511. El-Sayed, M.H., Chase, H.A., 2010a. Simulation of the breakthrough curves for the adsorption of alpha-lactalbumin and beta-lactoglobulin to SP sepharose FF cationexchanger. Biochem. Eng. J. 49, 221
228. El-Sayed, M.H., Chase, H.A., 2010b. Purification of the two major proteins from whey concentrate using a cation-exchange selective adsorption process. Biotechnol. Prog. 26, 192
199. Farcas, A.C., Socaci, S.A., Mudura, E., Dulf, F.V., Vodnar, D.C., Tofana, M., et al., 2017. Exploitation of brewing industry wastes to produce functional ingredients. In: Kanauchi, M. (Ed.), Brewing Technology. InTech, Rijeka, Croatia, pp. 137
156. Femenia, A., 2007. High-value co-products from plant foods: cosmetics and pharmaceuticals. In: Waldron, K. (Ed.), Handbook of Waste Management and Co-Product Recovery in Food Processing, 2007. Woodhead Publishing Limited, Cambridge, England, pp. 470
501. Ferdosh, S., Sarker, Z.I., Norulaini, N., Oliveira, A., Yunus, K., Chowdury, A.J., et al., 2015. Quality of tuna fish oils extracted from processing the byproducts of three species of neritic tuna using supercritical carbon dioxide. J. Food Proc. Preservation 39, 432
441. Ferry, M., Graen-Heedfeld, J., Bretz, K., Guillon, F., Michelini, E., Calabretta, M.M., et al., 2017. Peptide fractions obtained from rice byproducts by means of an environment-friendly process show in vitro health-related bioactivities. PLoS One 12 (1), e0170954. Fontoura, R., Daroit, D.J., Correa, A.P.F., Meira, S.M.M., Mosquera, M., Brandelli, A., 2014. Production of feather hydrolysates with antioxidant, angiotensin-I converting enzyme and dipeptidyl peptidase-IV-inhibitory activities. N. Biotechnol. 31, 506
513. Galanakis, C.M., 2012. Recovery of high added-value components from food wastes: conventional, emerging technologies and commercialized applications. Trends Food Sci. Technol. 26 (2), 68
87. Galanakis, C.M., 2015a. Chapter 3: development of a universal recovery strategy. In: Galanakis, C.M. (Ed.), Food Waste Recovery: Processing Technologies and Techniques. Elsevier Inc, Waltham. Galanakis, C.M., 2015b. Separation of functional macromolecules and micromolecules: from ultrafiltration to the border of nanofiltration. Trends Food Sci. Technol. 42, 44
63. Galanakis, C.M., Tornberg, E., Gekas, V., 2010. Clarification of high-added value products from olive mill wastewater. J. Food Eng. 99, 190
197. Galanakis, C.M., Markouli, E., Gekas, V., 2013. Fractionation and recovery of different phenolic classes from winery sludge via membrane filtration. Sep. Purif. Technol. 107, 245
251. Galanakis, C.M., Chasiotis, S., Botsaris, G., Gekas, V., 2014. Separation and recovery of proteins and sugars from Halloumi cheese whey. Food Res. Int. 65, 477
483. Galanakis, C.M., Kotanidis, A., Dianellou, M., Gekas, V., 2015. Phenolic content and antioxidant capacity of Cypriot Wines. Czech J. Food Sci. 33 (2), 126
136. Ganesan, K., Xu, B., 2017. A critical review on polyphenols and health benefits of black soybeans. Nutrients 9, 455.

Introduction in Functional Components Chapter | 2

71

Garcia, M.C., Endermann, J., Gonzalez-Garcia, E., Marina, M.L., 2015. HPLC-Q-TOF-MS identification of antioxidant and antihypertensive peptides recovered from cherry (Prunus cerasus L.) subproducts. J. Agric. Food Chem. 63, 1514
1520. Garcı´a-Martı´n, N., Perez-Magarin˜o, S., Ortega-Heras, M., Gonza´lez-Huerta, C., Mihnea, M., Gonza´lez-Sanjose´, M.L., et al., 2010. Sugar reduction in musts with nanofiltration membranes to obtain low alcohol-content wines. Sep. Purif. Technol. 76, 158
170. Gelmez, N., Kincal, N.S., Yener, M.E., 2009. Optimization of supercritical carbon dioxide extraction of antioxidants from roasted wheat germbased on yield, total phenolic and tocopherol contents, and antioxidant activities of the extracts. J. Supercritical Fluids 48, 217
224. Gerzhova, A., Mondor, M., Benali, M., Aider, M., 2015. Study of the functional properties of canola protein concentrates and isolates extracted by electro-activated solutions as noninvasive extraction method. Food Biosci. 12, 128
138. Gildberg, A., 2004. Enzymes and bioactive peptides from fish waste related to fish silage, fish feed and fish sauce production. J. Aquatic Food Product Technol. 13, 3
11. Goh, K.K.T., Kumar, R., Wong, S.S., 2014. Functionality of non-starch polysaccharides (NSPs). In: Noomhorm, A., Ahmad, I., Anal, A.K. (Eds.), Functional Foods and Dietary Supplements: Processing Effects and Health Benefits. John Wiley & Sons, Ltd, Chichester, UK, pp. 187
225. Goula, A.M., Ververi, M., Adamopoulou, A., Kaderides, K., 2017. Green ultrasound-assisted extraction of carotenoids from pomegranate wastes using vegetable oils. Ultrason. Sonochem. 34, 821
830. Guo, X., Han, D., Xi, H., Rao, L., Liao, X., Hu, X., et al., 2012. Extraction of pectin from navel orange peel assisted by ultra-high pressure, microwave or traditional heating: a comparison. Carbohyd. Polym. 88, 441
448. Han, S., Chee, K., Cho, S., 2015. Nutritional quality of rice bran protein in comparison to animal and vegetable protein. Food. Chem. 172, 766
769. Han, S., Ma, S., Li, L., Wang, X., Zheng, X., 2017. Application and development prospects of dietary fibers in flour products. Hindawi J. Chem. Available from: https://doi.org/10.1155/ 2017/2163218. Hasler, C.M., 2002. Functional foods: benefits, concerns and challenges-a position paper from the American Council on science and health. J. Nutr. 132 (12), 3772
3781. He, R., Girgih, A.T., Malomo, S.A., Ju, X., Aluko, R.E., 2013. Antioxidant activities of enzymatic rapeseed protein hydrolysates and the membrane ultrafiltration fractions. J. Funct. Foods 5, 219
227. Helkar, P.H., Sahoo, A.K., Patil, N.J., 2016. Review: food industry byproducts used as a functional food ingredient. Int. J. Waste Resour. 6, 248
253. Heng, M.Y., Katayama, S., Mitani, T., Ong, E.S., Nakamura, S., 2017. Solventless extraction methods for immature fruits: evaluation of their antioxidant and cytoprotective activities. Food. Chem. 221, 1388
1393. Hilliam, M., 2000. Functional Food
How big is the market. World Food Ingred. 12, 50
52. Hilz, H., Bakx, E.J., Schols, H.A., Voragen, A.G.J., 2005. Cell wall polysaccharides in black currants and bilberries—characterisation in berries, juice, and press cake. Carbohyd. Polym. 59, 477
488. IFT, http://www.ift.org/B/media/Knowledge%20Center/Science%20Reports/Expert%20Reports/ Functional%20Foods/Functionalfoods_expertreport_full.pdf, accessed 20.01.2018 Ieri, F., Innocenti, M., Andrenelli, L., Vecchio, V., Mulinacci, N., 2011. Rapid HPLC/DAD/MS method to determine phenolic acids, glycoalkaloids and anthocyanins in pigmented potatoes (Solanum tuberosum L.) and correlations with variety and geographical origin. Food. Chem. 125, 750
759.

72

Separation of Functional Molecules in Food by Membrane Technology

´ lvarez, C., O’Donnell, C.P., 2015. Ultrasound applications for the Kadam, S.U., Tiwari, B.K., A extraction, identification and delivery of food proteins and bioactive peptides. Trends Food Sci. Technol. 46, 60
67. Kagliwal, L.D., Patil, S.C., Pol, A.S., Singhal, R.S., Patravale, V.B., 2011. Separation of bioactives from sea buckthorn seeds by supercritical carbon dioxide extraction methodology through solubility parameter approach. Sep. Purif. Technol. 80, 533
540. Kang, J., Hu, X., Yang, R., Chen, Y., Yang, H., 2015. Characterization of natural low-methoxyl pectin from sunflower head extracted by sodium citrate and purified by ultrafiltration. Food. Chem. 180, 98
105. Karaca, A.,C., Low, N., Nickerson, M., 2011. Emulsifying properties of canola and flaxseed protein isolates produced by isoelectric precipitation and salt extraction. Food Res. Int. 44, 2991
2998. Kaspar, A.A., Reichert, J.M., 2013. Future directions for peptide therapeutics development. Drug Discov. Today 18, 807
817. Khan, R.S., Grigor, J., Wingera, R., Win, A., 2013. Functional food product development
opportunities and challenges for food manufacturers. Trends Food Sci. Technol. 30, 27
37. Klose, C., Arendt, E.K., 2012. Proteins in oats; their synthesis and changes during germination: a review. Crit. Rev. Food. Sci. Nutr. 52, 629
639. Ko, M.J., Cheigh, C.I., Cho, S.W., Chung, M.S., 2011. Subcritical water extraction of flavonol quercetin from onion skin. J. Food Eng. 102 (4), 327
333. Korhonan, H., Pihlanto, A., 2003. Bioactive peptides: new challenges and opportunities for the dairy industry. Aust. J. Dairy Technol. 58 (2), 129
133. Koˇsta´lova´, Z., Hroma´dkova´, Z., Ebringerova´, A., 2013. Structural diversity of pectins isolated from the Styrian oil-pumpkin (Cucurbita pepo var. styriaca) fruit. Carbohyd. Polym. 93 (1), 163
171. Kristl, J., Slekovec, M., Tojnko, S., Unuk, T., 2011. Extractable antioxidants and non-extractable phenolics in the total antioxidant activity of selected plum cultivars (Prunus domestica L.): evolution during on-tree ripening. Food. Chem. 125, 29
34. Kuhn, R.C., Filho, F.M., Silvia, V., Palacio, L., Herna´ndez, A., Pra´danos, P., 2010. Mass transfer and transport during purification of fructooligosaccharides by nanofiltration. J. Memb. Sci. 365, 356
365. Lafarga, T., Teagase, M.H., 2014. Bioactive peptides from meat muscle and byproducts: generation, functionality and application as functional ingredients. Meat Sci. 98, 227
239. Landete, J.M., 2012. Updated knowledge about polyphenols: functions, bioavailability, metabolism, and health. Crit. Rev. Food. Sci. Nutr. 52 (10), 936
948. Laroze, L.E., Dı´az-Reinoso, B., Moure, A., Zu´n˜iga, M.E., Domı´nguez, H., 2010. Extraction of antioxidants from several berries pressing wastes using conventional and supercritical solvents. Europ. Food Res. Technol. 231, 669
677. Lazaridou, A., Biliaderis, C.G., 2007. Molecular aspects of cereal β-glucan functionality: physical properties, technological applications and physiological effects. J. Cereal Sci. 46, 101
118. Lemes, A.C., Sala, L., da Costa, O.J., Cavalcante Braga, A.R., Egea, M.B., Fernandes, K.F., 2016. A review of the latest advances in encrypted bioactive peptides from protein-rich waste. Int. J. Mol. Sci. 17 (6), 950
974. Li, J., Chase, H.A., 2010. Applications of membrane techniques for purification of natural products. Biotechnol. Lett. 32 (5), 601
608. Li, Q., Bi, Q.-Y., Lin, H.-H., Bian, L.-X., Wang, X.-L., 2013. J. Memb. Sci. 427, 155
167.

Introduction in Functional Components Chapter | 2

73

Lima, G.P.P., Vianello, F., Correˆa, C.R., da Silva Campos, R.A., Borguini, M.G., 2014. Polyphenols in fruits and vegetables and its effect on human health. Food Nutr. Sci. 5, 1065
1082. Loginov, M., Boussetta, N., Lebovka, N., Vorobiev, E., 2013. Separation of polyphenols and proteins from flaxseed hull extracts by coagulation and ultrafiltration. J. Memb. Sci. 442, 177
186. Luthria, D.L., 2012. Optimization of extraction of phenolic acids from a vegetable waste product using a pressurized liquid extractor. J. Funct. Foods 4, 842
850. Ma, C., Ni, X., Chi, Z., Ma, L., Gao, L., 2007. Purification and characterization of an alkaline protease from the marine yeast Aureobasidium pullulans for bioactive peptide production from different sources. Marine Biotechnol. (NY) 9 (3), 343
351. Manamperi, W.A.R., Wiesenborn, D.P., Chang, S.K.C., Pryor, S.W., 2011. Effects of protein separation conditions on the functional and thermal properties of canola protein isolates. J. Food. Sci. 76 (3), 266
273. Maran, J.P., Sivakumar, V., Thirugnanasambandham, K., Sridhar, R., 2013. Optimization of microwave assisted extraction of pectin from orange peel. Carbohyd. Polym. 9, 703
709. ´ lvarez, C., Paredes, B., Dı´az, M., 2016. The use of sub-critical water hydrolysis for Marcet, I., A the recovery of peptides and free amino acids from food processing wastes. Review of sources and main parameters. Waste Manage. 49, 364
371. Martirosyan, D.M., Singh, J., 2015. A new definition of functional food by FFC: what makes a new definition unique? Funct. Foods Health Dis. 5 (6), 209
223. Matthews, S., Mila, I., Scalbert, A., Donnelly, D.M.X., 1997. Extractable and non-extractable proanthocyanidins in barks?. Phytochemistry 45 (2), 405
410. McClement, D.J., 2012. Edible delivery systems for nutraceuticals: designing functional foods for improved health. Ther. Deliv. 3 (7), 801
803. McClements, D.J., Decker, E.A., Park, Y., Weiss, J., 2009. Structural design principles for delivery of bioactive components in nutraceuticals and functional foods. Crit. Rev. Food. Sci. Nutr. 49 (6), 577
606. Moure, A., Domı´nguez, H., Parajo´, J.C., 2006. Antioxidant properties of ultrafiltration-recovered soy protein fractions from industrial effluents and their hydrolysates. Process Biochem. 41, 447
456. Mu¨ller-Maatsch, J., Bencivenni, M., Caligiani, A., Tedeschi, T., Bruggeman, G., Bosch, M., et al., 2016. Pectin content and composition from different food waste streams. Food. Chem. 201, 37
45. Mussatto, S.I., Roberto, I.C., 2005. Acid hydrolysis and fermentation of brewer’s spent grain to produce xylitol. J. Sci. Food. Agric. 85 (14), 2453
2460. Mussatto, S.I., Dragone, G., Roberto, I.C., 2007. Ferulic and p-coumaric acids extraction by alkaline hydrolysis of brewer’s spent grain. Ind. Crops Products 25, 231
237. Nagy, M., Semeniuc, C.A., Socaci, S.A., Pop, C.A., Rotar, A.M., S˘al˘agean, C.D., et al., 2017. Utilization of brewer’s spent grain and mushrooms in fortification of smoked sausages. Food Sci. Technol. Camp. 37 (2), 315
320. Nandi, I., Ghosh, M., 2015. Studies on functional and antioxidant property of dietary fibre extracted from defatted sesame husk, rice bran and flaxseed. Bioactive Carbohyd. Dietary Fibre 5, 129
136. Nawaz, H., Shi, J., Mittal, G.S., Kakuda, Y., 2006. Extraction of polyphenols from grape seeds and concentration by ultrafiltration. Sep. Purif. Technol. 48, 176
181.

74

Separation of Functional Molecules in Food by Membrane Technology

Ndiaye, N., Pouliot, Y., Saucier, L., Beaulieu, L., Bazinet, L., 2010. Electroseparation of bovine lactoferrin from model and whey solutions. Sep. Purif. Technol. 74, 93
99. Ng, K.L., Ayob, M.K., Said, M., Osman, M.D.A., Ismail, A., 2013. Optimization of enzymatic hydrolysis of palm kernel cake protein (PKCP) for producing hydrolysates with antiradical capacity. Ind. Crops Products 43, 725
731. Niemi, P., Faulds, C.B., Sibakov, J., Holopainen, U., Poutanen, K., Buchert, J., 2012a. Effect of a milling pre-treatment on the enzymatic hydrolysis of carbohydrates in brewer’s spent grain. Bioresour. Technol. 116, 155
160. Niemi, P., Tamminen, T., Smeds, A., Viljanen, K., Ohra-Aho, T., Holopainen-Mantila, U., et al., 2012b. Characterization of lipids and lignans in brewer’s spent grain and its enzymatically extracted fraction. J. Agric. Food. Chem. 60, 9910
9917. Niemi, P., Martins, D., Buchert, J., Faulds, C.B., 2013. Pre-hydrolysis with carbohydrases facilitates the release of protein from brewer’s spent grain. Bioresour. Technol. 136, 529
534. Nosa´lova´, G., Prisenznakova´, L., Kosta´lova´, Z., Ebringerova´, A., Hroma´dkova´, Z., 2011. Suppressive effect of pectic polysaccharides from Cucurbita pepo L. var. Styriaca on citric acid-induced cough reflex in guinea pigs. Fitoterapia 82, 357
364. Oliveira, R., Oliveira, V., Aracava, K.K., da Rodrigues, C.E., 2012. Effects of the extraction conditions in the yield and composition of rice bran oil extracted with ethanol
A response surface approach. Food Bioprod. Process. 90, 22
31. Otles, S., Cagindi, O., 2012. Safety considerations of nutraceuticals and functional foods. In: McElhatton, A., Sobral, P.J.A. (Eds.), Novel Technologies in Food Science. Springer, New York, US, pp. 121
136. Oviyaasri, M., Manjuladevi, M., Anitha, R., Narmatha, M., Kavitha, J., 2017. Extraction and isolation of polyunsaturated fatty acids from microalgae Tetraselmis. Sp. Int. J. Innovative Res. Sci. Eng. Technol. 6 (3), 4810-481. Ozcan, T., Akpinar-Bayizit, A., Yilmaz-Ersan, L., Delikanli, B., 2014. Phenolics in Human Health. Int. J. Chem. Eng. Applicat. 5, 393
396. Pal, Y.P., Pratap, A.P., 2017. Rice Bran Oil: a versatile source for edible and industrial applications. J. Oleo. Sci. 66 (6), 551
556. Patsioura, A., Galanakis, C.M., Gekas, V., 2011. Ultrafiltration optimization for the recovery of β-glucan from oat mill waste. J. Memb. Sci. 373, 53
63. Perez-Jimenez, J., Calixto Saura, F., 2015. Macromolecular antioxidants or non-extractable polyphenols in fruit and vegetables: intake in four European countries. Food Res. Int. 74, 315
323. Perretti, G., Miniati, E., Montanari, L., Fantozzi, P., 2003. Improving the value of rice byproducts by SFE. J. Supercrit. Fluids 26, 63
71. Pinelo, M., Jonsson, G., Meyer, A.S., 2009. Membrane technology for purification of enzymatically produced oligosaccharides: molecular and operational features affecting performance. Sep. Purif. Technol. 70, 1
11. Pingret, D., Fabiano-Tixier, A.S., Bourvellec, C.L., Renard, C.M.G.C., Chemat, F., 2012. Lab and pilotscale ultrasound-assisted water extraction of polyphenols from apple pomace. J. Food Eng. 111, 73
81. Pinkowska, H., Oliveros, E., 2014. Application of the Doehlert matrix for the determination of the optimal conditions of hydrolysis of soybean protein in subcritical water. Ind. Eng. Chem. Res. 53, 1320
1326. Pinkowska, H., Wolak, P., Oliveros, E., 2013. Application of Doehlert matrix for determination of the optimal conditions of hydrothermolysis of rapeseed meal in subcritical water. Fuel 106, 258
264.

Introduction in Functional Components Chapter | 2

75

Pokora, M., Zambrowicz, A., Dabrowska, A., Eckert, E., Setner, B., Szoltysik, M., et al., 2014. An attractive way of egg white protein byproduct use for producing of novel antihypertensive peptides. Food. Chem. 151, 500
505. Poli, A., Anzelmo, G., Fiorentino, G., Nicolaus, B., Tommonaro, G., Di Donato, P., 2011. Polysaccharides from wastes of vegetable industrial processing: new opportunities for their eco-friendly re-use. In: Elnashar, M. (Ed.), Biotechnology of Biopolymers. Rijeka, Croatia InTech. pp. 32
56. Pourali, O., Asghari, F., Yoshida, H., 2010. Production of phenolic compounds from rice bran biomass under subcritical water conditions. Chem. Eng. J. 160 (1), 259
266. Prado, M.J., Dalmolin, I., Carareto, N.D., Basso, R.C., Meirelles, A.J., Oliveira, J.V., et al., 2012. Supercritical fluid extraction of grape seed: process scale-up, extract chemical composition and economic evaluation. J. Food Eng. 109, 249
257. Qiu, N.X., Tian, Y.X., Qiao, S.T., Deng, H., 2009. Apple pectin behavior separated by ultrafiltration. Agric. Sci. China 8 (10), 1193
1202. Quoc, L.P.T., Huyen, V.T.N., Hue, L.T.N., Hue, N.T.H., Thuan, N.H.D., Tam, N.T.T., et al., 2015. Extraction of pectin from pomelo (Citrus maxima) peels with the assistance of microwave and tartaric acid. Int. Food Res. J. 22 (4), 1637
1641. Ravindran, R., Jaiswal, A.K., 2016. Exploitation of food industry waste for high-value products. Trends Biotechnol. 34, 58
69. Rawel, H.M., Meidtner, K., Kroll, J., 2005. Binding of selected phenolic compounds to proteins. J. Agric. Food. Chem. 53, 4228
4235. Rizliya, V., Mendis, E., 2014. Biological, physical, and chemical properties of fish oil and industrial applications. In: Kim, S.K. (Ed.), Seafood Processing Byproducts: Trends and Applications. Springer Science Business Media, New York. Rodriguez, G., Lama, A., Rodriguez, R., Jimenez, A., Guillen, R., Fernandez-Bolanos, J., 2008. Olive stone an attractive source of bioactive and valuable compounds. Bioresour. Technol. 99, 5261
5269. Roupas, P., De Silva, K., Smithers, J.W., Ferguson, A., 2007. Waste management and coproduct recovery in red and white meat processing. In: Waldron, K. (Ed.), Handbook of Waste Management and Co-product Recovery in Food Processing. Woodhead Publishing Limited, Cambridge, England, pp. 305
328. Russ, W., Meyer-Pittroff, R., 2004. Utilizing waste products from the food production and processing industries. Crit. Rev. Food. Sci. Nutr. 44, 57
62. Salgado, P.R., Molina Ortiz, S.E., Petruccelli, S., Mauri, A.N., 2012. Functional food ingredients based on sunflower protein concentrates naturally enriched with antioxidant phenolic compounds. J. Am. Oil Chem. Soc. 89, 825
836. Sarker, M.Z.I., Selamat, J., Habib, A.S.Md. A., Ferdosh, S., Akanda, M.J.H., et al., 2012. Optimization of supercritical CO2 extraction of fish oil from viscera of African Catfish (Clarias gariepinus). Int. J. Mol. Sci. 13, 11312
11322. Scalbert, A., Williamson, G., 2000. Dietary intake and bioavailability of polyphenols. chocolate: modern science investigates an ancient medicine. J. Nutr. 130 (8S Suppl), 2073S
2085S. Scalbert, A., Johnson, I.T., Saltmarsh, M., 2005. Polyphenols: antioxidants and beyond. Am. J. Chem. Nutr. 81, 215S
217S. Seixas, F.L., Fukuda, D.L., Turbiani, F.R.B., Garcia, P.S., Petkowicz, C.L.O., Jagadevan, S., et al., 2014. Extraction of pectin from passion fruit peel (Passiflora edulis f. flavicarpa) by microwave-induced heating. J. Food Hydrocolloid 38, 186
192. Sharma, P.C., Tilakratne, B.M.K.S., Gupta, A., 2010. Utilization of wild apricot kernel press cake for extraction of protein isolate. J. Food Sci. Technol. 47 (6), 682
685.

76

Separation of Functional Molecules in Food by Membrane Technology

Shi, L., Beamer, S.K., Yin, T., Matak, K.E., Yang, H., Jaczynski, J., 2017. Mass balance for isoelectric solubilization/precipitation of carp, chicken, menhaden, and krill. LWT - Food Sci. Technol. 81, 26
34. Singhai, P.K., Sarma, B.K., Srivastava, J.S., 2011. Phenolic acid content in potato peel determines natural infection of common scab caused by Streptomyces spp. World J. Microbiol. Biotechnol. 27, 1559
1567. Socaci, S.A., Farcas, A.C., Vodnar, D.C., Tofana, M., 2017a. Food wastes as valuable sources of bioactive molecules. In: Shiomi, N. (Ed.), Superfood and FunctionalFood—The Development of Superfoods and Their Roles as Medicine. InTech, Rijeka, Croatia, pp. 75
93. Socaci, S.A., Rugin˘a, D.O., Diaconeasa, Z.M., Pop, O.L., F˘arca¸s, A.C., P˘aucean, A., et al., 2017b. Antioxidant compounds recovered from food wastes. In: Chavarri Hueda, M. (Ed.), Functional Food—Improve Health throzugh Adequate Food. InTech, Rijeka, Croatia, pp. 3
21. Stein, A.J., Rodriguez-Cerezo, E., 2008. Functional Food in the European Union. JRC Sci. Tech. Reports. Available from: http://ipts.jrc.ec.europa.eu/publications/pub.cfm?id 5 1719. Stojceska, V., Ainsworth, P., Plunkett, A., Ibanoglu, S., 2008. The recycling of brewer’s processing by-product into ready-to-eat snacks using extrusion technology. J. Cereal Sci. 47 (3), 469
479. Sun, Y., Liu, D., Chen, J., Ye, X., Yu, D., 2011. Effects of different factors of ultrasound treatment on the extraction yield of the all-trans-β-carotene from citrus peels. Ultrason. Sonochem. 18, 243
249. Tahergorabi, R., Beamer, S.K., Matak, K.E., Jaczynski, J., 2012. Functional food products made from fish protein isolate recovered with isoelectric solubilization/precipitation. LWT - Food Sci. Technol. 48, 89
95. Tan, Y.A., Sambanthamurthi, R., Sundram, K., Wahid, M.B., 2007. Valorisation of palm byproducts as functional components. Europ. J. Lipid Sci. Technol. 109, 380
393. Tang, D.S., Yin, G.M., He, Y.Z., Hu, S.Q., Li, B., Li, L., et al., 2009. Recovery of protein from brewer’s spent grain by ultrafiltration. Biochem. Eng. J. 48, 1
5. Tang, D.S., Tian, Y.J., He, Y.Z., Li, L., Hu, S.Q., Li, B., 2010. Optimisation of ultrasonicassisted protein extraction from brewer’s spent grain. Czech J. Food Sci. 28, 9
17. Thongnuanchan, P., Benjakul, S., 2014. Essential oils: extraction, bioactivities, and their uses for food preservation. J. Food. Sci. 79 (7), 1231
1249. Tunchaiyaphum, S., Eshtiaghi, M.N., Yoswathana, N., 2013. Extraction of bioactive compounds from mango peels using green technology. Int. J. Chem. Eng. Applicat. 4 (4), 194
198. Venzon, S.S., Canteri, M.H.G., Granato, D., Demczuk, B.J., Maciel, G.M., Stafussa, A.P., et al., 2015. Physicochemical properties of modified citrus pectins extracted from orange pomace. J. Food Sci. Technol. 52, 4102. Vieira, E., Rocha, M.A.M., Coelho, E., Pinho, O., Saraiva, J.A., Ferreira, I., et al., 2014. Valuation of brewer’s spent grain using a fully recyclable integrated process for extraction of proteins and arabinoxylans. Ind. Crops Products 52, 136
143. ˇ Vladimir-Kneˇzevi´c, S., Blaˇzekovi´c, B., Bival Stefan, M., Babac, M., 2012. Plant polyphenols as antioxidants influencing the human health. In: Venketeshwer, R. (Ed.), Phytochemicals as Nutraceuticals—Global Approaches to Their Role in Nutrition and Health. InTech, Rijeka, Croatia, pp. 155
180. WHO, 2007. Protein and Amino Acid Requirements in Human Nutrition: Report of the Joint WHO/FAO/UNU Expert Consultation. World Health Organization, Geneva.

Introduction in Functional Components Chapter | 2

77

Wan, Y., Ghost, R., Cui, Z., 2002. High resolution plasma protein fractionation using ultrafiltration. Desalination 144, 301
306. Wang, X., Chen, Q., Lu¨, X., 2014. Pectin extracted from apple pomace and citrus peel by subcritical water. Food Hydrocolloids 38, 129
137. Weisz, G.M., Carle, R., Kammerer, D.R., 2013. Sustainable sunflower processing—II. Recovery of phenolic compounds as a by-product of sunflower protein extraction. Innovat. Food Sci. Em. Technol. 17, 169
179. Williamson, C., 2009. Functional foods: what are the benefits. Br. J. Community. Nurs. 14 (6), 230
236. World Health Organization, 2010. Population nutrient intake goals for preventing diet-related chronic diseases, Geneva, witzerland. Available at: http://www.who.int Yadav, B.K., Joyner, J.J., 2014. Processing effects on functional components in cereals and grains. In: Noomhorm, A., Ahmad, I., Anal, A.H. (Eds.), Functional Foods and Dietary Supplements: Processing Effects and Health Benefits. John Wiley & Sons, Ltd, Chichester, UK, pp. 63
89. Yi, C., Shi, J., Xue, S., Jiang, Y., Li, D., 2009. Effects of supercritical fluid extraction parameters on lycopene yield and antioxidant activity. Food. Chem. 113 (4), 1088
1894. Yilmaz, E.E., Vural, H., Ozvural, B.E., 2010. Extraction and identification of proanthocyanidins from grape seed (Vitis vinifera) by using supercritical carbon dioxide. J. Supercrit. Fluids 55 (3), 924
928. Yu, X., Gouyo, T., Grimi, N., Bals, O., Vorobiev, E., 2016. Ultrasound enhanced aqueous extraction from rapeseed green biomass for polyphenol and protein valorization. Comptes Rendus Chimie 19, 766
777. Zhu, F., Du, B., Zheng, L., Li, J., 2015. Advance on the bioactivity and potential applications of dietary fibre from grape pomace. Food. Chem. 186, 207
212. Zhu, Z., Jiang, T., He, J., Barba, F., Cravotto, G., Koubaa, M., 2016. Ultrasound-assisted extraction, centrifugation and ultrafiltration: multistage process for polyphenol recovery from purple sweet potatoes. Molecules 21 (11), 1584.

FURTHER READING Ghoreishi, S., Shahrestani, R., 2009. Subcritical water extraction of mannitol from olive leaves. J. Food Eng. 93 (4), 474
481.

This page intentionally left blank

Chapter 3

Membrane Filtration of Biosurfactants Paula Jauregi and Konstantina Kourmentza Department of Food and Nutritional Science, Harry Nursten building, The University of Reading, Whiteknights, Reading, United Kingdom

Chapter Outline 3.1 Introduction 3.1.1 Biosurfactants 3.1.2 Production of Biosurfactants 3.2 Downstream Processing of Biosurfactants 3.2.1 The Ultrafiltration Process and Equipment 3.2.2 Assessment of Separation Performance 3.3 Ultrafiltration of Lipopeptide Biosurfactants 3.3.1 Surfactin Separation by the Two-Step Ultrafiltration Method 3.3.2 Hybrid Recovery Processes Using Ultrafiltration

79 79 84 87 90 91 92

92

3.3.3 Separation of Lipopeptide Mixtures 3.3.4 Recovery of Lipopeptides From Complex Culture Medium 3.4 Ultrafiltration of Rhamnolipid Biosurfactants 3.5 Ultrafiltration of Mannosylerythritol Lipids 3.6 Further Considerations 3.6.1 Membrane Choice 3.6.2 Membrane Cleaning 3.7 Conclusions and Final Outlook References Further Reading

95

95 97 98 100 100 101 102 103 112

94

3.1 INTRODUCTION 3.1.1 Biosurfactants Biosurfactants are surface active agents of biological origin, produced by bacteria, yeast, and fungi. They are secondary metabolites excreted by microorganisms, or are attached to their outer cellular membrane. Those molecules are of amphiphilic nature and consist of hydrophilic “heads” and hydrophobic (lipophilic) “tails.” Therefore, they can be found between fluid phases of

Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00003-6 © 2019 Elsevier Inc. All rights reserved.

79

80

Separation of Functional Molecules in Food by Membrane Technology

polar and nonpolar media, air
solid, and liquid
solid surfaces as they have the ability to reduce the surface tension (ST) of water and the interfacial tension (IFT) of oil/water interfaces. The effectiveness of a biosurfactant is defined by the concentration where the minimum ST is achieved (critical micelle concentration or CMC), above which no further reduction in the ST occurs. Moreover, they are characterized by the ability to form water
hydrocarbon emulsions and are known to enhance the apparent water solubility of hydrophobic compounds. As biosurfactants are produced by a wide variety of different microorganisms, their chemical structures, and thus physicochemical properties and functions, are quite diverse (Ron and Rosenberg, 2001). The type of biosurfactant occurring is characteristic to the producing microorganism’s physiology and ecology, therefore they are likely to possess several natural roles (Freire et al., 2010; Ron and Rosenberg, 2010). Among them is adhesion of cells to interfaces as a physiological mechanism for the growth of bacteria on hydrophobic substrates, emulsification to promote growth on water immiscible substrates, desorption of substrates from surfaces and/or increase of their apparent water solubility, and defense mechanism for the survival of the producing microbe (Cameotra et al., 2010). Depending on their mass, which ranges between 500 and 1000 kDa, they can be categorized either as low or high molecular mass biosurfactants. Low molecular mass biosurfactants may be composed of amino acids, fatty acids, sugars, as well as functional groups such as carboxylic acids. Glycolipids and lipopeptides are included in this group, with the former consisting of different sugars glycosidically linked to C8-C16 β-hydroxy fatty acids, while the latter consist of cyclopeptides (7
10 peptides) that are also linked to fatty acids of various lengths (Kourmentza et al., 2017; Uzoigwe et al., 2015). On the other hand, high molecular mass biosurfactants comprise of proteins, lipoproteins, polysaccharides, lipopolysaccharides or a combination of them. They are also known as polymeric biosurfactants, with the most studied representatives being the polysaccharide
protein complex alasan and the lipopolysaccharide emulsan. Both compounds have been isolated by Acinetobacter species and are characterized by a molecular mass of 1000 kDa (Navon-Venezia et al., 1995; Rosenberg et al., 1979). In general, low molecular mass biosurfactants are more efficient in reducing the ST of water and IFT against various hydrocarbons, while high molecular mass biosurfactants are able to form stronger and more stable emulsions between two immiscible liquids but are less effective in ST and IFT reduction. For this reason, polymeric biosurfactants are also known as bioemulsifiers (Satpute et al., 2010). In Table 3.1 the major biosurfactant types and the microorganisms involved in their production are summarized. Apart from their surface activity and emulsifying properties they also show high stability in a wide range of temperature, pH, and salinity (Nguyen and Sabatini, 2011; Sobrinho et al., 2013). They are characterized by higher

TABLE 3.1 Biosurfactant Types and Producing Strains With Emphasis on Glycolipids and Lipopeptides Class

Microorganism

References

Rhamnolipids

Pseudomonas aeruginosa

Christova et al. (2011)

Pseudomonas alcaligenes

Oliveira et al. (2009)

Pseudomonas chrororaphis

Gunther et al. (2005)

Pseudomonas putida

Wittgens et al. (2011)

Burkholderia thailandensis

Dubeau et al. (2009), Funston et al. (2016), Kourmentza et al. (2018)

Burkholderia glumae

Costa et al. (2011)

Burkholderia plantarii

Ho¨rmann et al. (2010)

Burkholderia pseudomallei

Ha¨ussler et al. (1998)

Burkholderia seminalis

Arau´jo et al. (2017)

Pantoea stewartii

Rooney et al. (2009)

Enterobacter asburiae Enterobacter hormaechei Acinetobacter cacloaceticus Sophorolipids

Candida bogoriensis

Cutler and Light (1979)

Candida bombicola

Deshpande and Daniels (1995), Solaiman et al. (2004)

Candida apicola

Hommel et al. (1994) (Continued )

TABLE 3.1 (Continued) Class

Mannosylerythritol lipids (MEL)

Microorganism

References

Candida batistae

Konishi et al. (2008)

Candida floricola

Konishi et al. (2016)

Candida tropicalis

Chandran and Das (2012)

Wickerhamiella domerqiae

Chen et al. (2006)

Candida (Pseudozyma) antarctica

Kitamoto et al. (1990)

Candida sp. SY16

Kim et al. (2006)

Pseudozyma aphidis

Rau et al. (2005a,b)

Pseudozyma parantarctica

Tomotake Morita et al. (2009a)

Pseudozyma hubeiensis

Konishi et al. (2011)

Pseudozyma siamensis

Morita et al. (2008)

Pseudozyma churashimaensis

Morita et al. (2013)

Pseudozyma rugulosa

Morita et al. (2006)

Pseudozyma shanxiensis

Fukuoka et al. (2007a)

Pseudozyma tsukubaensis

Andrade et al. (2017), Yamamoto et al. (2013)

Ustilago maydis

Spoeckner et al. (1999)

Ustilago scitaminea

Morita et al. (2009)

Sporisorium sp. aff. Sorgji SAM20

Alimadadi et al. (2018)

Ceriporia lacerate CHZJU

Niu et al. (2017)

Lipopeptides

Bacillus amyloliquefaciens

Ndlovu et al. (2017)

Bacillus methylotrophicus

Jemil et al. (2017)

Bacillus subtilis

Jajor et al. (2016)

Bacillus licheniformis

Favaro et al. (2016)

Bacillus pumilus

Brack et al. (2015)

Bacillus safensis

Domingos et al. (2015)

Bacillus thuringiensis

Deepak and Jayapradha (2015)

Bacillus mojavensis

Ayed et al. (2014), Ma and Hu (2014)

Pseudomonas fluorescens

Nielsen et al. (2003)

Pseudomonas syringae

Burch et al. (2014)

Pseudomonas nitroreducens

de Sousa and Bhosle (2012)

Pseudomonas putida

Li et al. (2013)

Pseudonomas entomophila

Vallet-Gely et al. (2010)

Enterobacter sp. strain N18

You et al. (2015)

Brevibacterium aureum

Kiran et al. (2010)

Nocardiopsis alba

Gandhimathi et al. (2009)

Paenibacillus polymyxa

Grady et al. (2016), Vater et al. (2017)

Serratia marcescens

Matsuyama et al. (1992)

84

Separation of Functional Molecules in Food by Membrane Technology

biodegradability and lower toxicity compared to their synthetic counterparts; have antimicrobial, antifungal, and antiyeast activities; and most importantly they can be produced from renewable resources (Gudin˜a et al., 2015; Kosaric and Vardar-Sukan, 2015; Kourmentza et al., 2017). Due to their versatility and unique properties they can be used as wetting, spreading, foaming, thickening, dispersing, stabilizing, antibacterial and therapeutic agents, solubilizers, detergents, emulsifiers/de-emulsifiers, biofungicides, and antiadhesives (Awada et al., 2011; Leighton, 2013; Sadasivan, 2015; Schelges and Tretyakova, 2016; Suzuki et al., 2012). Therefore, they can find different applications: G G G G

G

G G

in the petroleum industry for enhanced oil recovery and deemulsification in environmental cleanup for bioremediation of water and soil in the mining industry for heavy metals remediation in the food industry for emulsification/deemulsification and as functional ingredients, for the formulation of pharmaceuticals, therapeutics, cosmetics, nanoparticles, and household cleaning products in agriculture for the improvement of soil quality and as biocontrol agents in water treatment in industrial biotechnology as to enhance biofuel and bioenergy production (AGAE Technologies LLC, 2018a; Sachdev and Cameotra, 2013; Santos et al., 2016; Silva et al., 2014; Soliance France, 2018; Toyobo Co. Ltd., 2018)

3.1.2 Production of Biosurfactants Several companies worldwide have initiated the production and commercialization of biosurfactants, which are used by top-leading chemical industries in order to formulate various products. In Table 3.2 companies currently involved in biosurfactant production, and commercialization of biosurfactant-based products, are summarized. Since their possibilities and applications are quite diverse, biosurfactants’ market is expected to grow from approximately $4.2 billion in 2017 up to $5.5 billion in 2022, characterized by a compound annual growth rate of 5.6% (Markets and Markets, 2017). According to statistics released in 2012, Europe was in the lead, holding around 55% of the overall market, while it is projected to continue its dominance in the future (Transparency Market Research, 2014). Key factors promoting their wide commercialization include the increasing awareness among consumers and increasing demand for green solutions by various end-use industries for biodegradable and less toxic alternatives along with stringent regulations regarding the usage of biobased biosurfactants (European Commission, 2012a,b, 2011a,b). However, the high production cost is still restricting the rapid growth of biosurfactants’ market. For

Membrane Filtration of Biosurfactants Chapter | 3

85

TABLE 3.2 Biosurfactant Producing Companies Currently Active Worldwide Biosurfactant

Company

Brand Name

Main Products

Glycolipids

Auravia Latvia (LT)

Aura-Pure

Glycolipid cocktail

Rhamnolipids

Jeneil Biotech. Inc (USA) AGAE

90
95% pure monoand dirhamnolipids

AGAE Technologies LLC (USA)

MayLu¯ cleaning, lotion, and soap products Mono-rhamnolipids C10C10 and C12-C12

Glycosurf LLC (USA) Logos Technologies LLC (USA)

NatSurFact

Aqueous solutions of the sodium salt rhamnolipid form

EcoPlus Wojtek Czech (GE)

Biotenside

Liquid or paste dirhamnolipids

Evonik Industries AG (GE)

Rewoferm

Aqueous solution of sophorolipids (lactone and acid form)

Ecover (BE)

SL18

High lactonic solubilized sophorolipid

Soliance (FR)

Sopholiance S

Wheatoleo (FR)

Sophoclean

Saraya Co. Ltd. (Japan)

Soforo

Sophorolipid-based laundry and dishwashing products

Sopolin

Sopholine Acne, Sopholine Ato, Sopholine functional soaps

Urumqi BioTechnology Co. Ltd. (China)

Sophorolipids

SyntheZyme LLC (USA) MG Intobio Co. Ltd. (South Korea)

(Continued )

86

Separation of Functional Molecules in Food by Membrane Technology

TABLE 3.2 (Continued) Biosurfactant

Company

Brand Name

Allied Carbon Solutions Co. Ltd. (Japan)

ACS-Sophor

Mannosylerytiytol lipids (MEL)

Toyobo Co. Ltd. (Japan)

CERAMELA

Lipopeptides

Kaneka Corporation (Japan)

KANEKA Surfactin

Main Products

Cosmetic ingredients

Boruta-Zachem SA (PL) Lipofabrik (FR)

Surfactin, Fengysin, Iturin, Mycosubtilin, Lichenysin, Plipastatin

example, rhamnolipids of 90% purity cost around $1250/kg whereas high purity synthetic surfactants, such as 99% sodium dodecyl sulfate (SDS), cost between $10 and $20/kg (AGAE Technologies LLC, 2018b; Haihang Industry Co. Ltd, 2018). Therefore, there is a necessity for the development of new strategies and improvement of biotechnological processes that will boost production and the wider application of biosurfactants. Up to date, several approaches are being investigated in order to decrease their overall cost and compete with the synthetic surfactants. Those include: 1. the development and use of mutants and recombinant strains in order to overproduce biosurfactants and obtain higher productivity yields 2. the utilization of inexpensive and renewable raw material in order to form their fermentation medium 3. the selection and optimization of low-cost and efficient downstream processes so as to maximize their recovery and purification (Dhanarajan and Sen, 2015; Kosaric and Vardar-Sukan, 2015; Willenbacher et al., 2015) Mutagenesis has been successfully applied in several studies and the employment of hyperproducers was reported to increase the production of different types of biosurfactants by 2
25 times (Mukherjee et al., 2006; Santos et al., 2016). This strategy has only been studied at laboratory scale and process scale-up may reveal huge potential in terms of yield enhancement, and thus enhance biosurfactant production to a high degree (Randhawa, 2015). Another approach is the utilization of low-cost substrates and the optimization of fermentation conditions, which may lead to a reduction of the overall production cost by 10%
30% (Makkar et al., 2011). Several agroindustrial

Membrane Filtration of Biosurfactants Chapter | 3

87

and food-processing industry byproducts, such as glycerol, used oils, soapstocks, vegetable and fruit processing wastes, etc., have been investigated as potential feedstocks for biosurfactant production within the last decade (Kourmentza et al., 2017; Makkar et al., 2011). Substitution of synthetic substrates with renewable ones is also promoting sustainability via the concept of the circular economy, with wastes being transformed to high value
added products instead of being disposed in landfills (Dhanarajan and Sen, 2015). However, biosurfactant production cost can still be high due to low product yields, long fermentation periods, and most importantly their complex extraction and purification process. Downstream processing of biosurfactants can account for up to 60%
80% of their overall cost and includes their collection, separation, and purification (Najmi et al., 2018). Conventional methods require large volumes of organic solvents, such as ethyl acetate, diethyl ether, or chloroform:methanol. The use of such solvents apart from increasing the production cost will have a negative environmental impact. In addition, low biosurfactant productivity and utilization of media based on waste byproducts increase the complexity of the downstream process. Hence, cost-effective and efficient methods for the separation and purification of biosurfactants from the fermentation broth are crucial for their economical production.

3.2 DOWNSTREAM PROCESSING OF BIOSURFACTANTS Separation of biosurfactants is performed by taking advantage of their unique physicochemical properties, which distinguish them from other molecules present in the fermentation broth. Biosurfactants are amphiphilic compounds and by regulating the pH and ionic strength they tend to accumulate onto water/air, water/solvent, or water/solid interfaces, where one phase is more hydrophobic than the other. Another characteristic is that above their CMC they form supramolecular structures, such as micelles or vesicles, having nominal molecular diameters up to two to three orders of magnitude larger than the single unassociated molecules (Lin and Jiang, 1997). Different methods for downstream processing of various types of biosurfactants have been employed and unit operations, described so far in the literature include, washing, filtration, foam fractionation, precipitation, solvent extraction, crystallization, membrane filtration, absorption, ion-exchange, or thin layer chromatography (Smyth et al., 2010a,b). In order to select the most appropriate strategy for the recovery and purification of biosurfactants certain factors should be taken into account, depending on characteristics of the biosurfactant produced and process requirements. Product yield plays an important role on the process applied. For example, lipopeptides and trehalose lipids are obtained in lower concentrations (0.6
11 g/L) compared to rhamnolipids, sophorolipids, and MEL, which may reach concentrations up to 100 g/L (Kourmentza et al., 2017;

88

Separation of Functional Molecules in Food by Membrane Technology

Mutalik et al., 2008). Moreover, the use of hydrophobic substrates (which are known to induce biosurfactant production), the use of antifoaming agents (due to extensive foaming), and also the presence of impurities such as cellular debris and proteins pose an additional challenge regarding their recovery and purification. Last but not least, depending on the final application of the product, higher purities and possibly certain biosurfactant congeners may be required (Weber and Zeiner, 2015). In Table 3.3 the advantages and drawbacks of each methodology applied for the downstream processing of biosurfactants are summarized. The most applied method used for the recovery of biosurfactants includes acid precipitation of the culture supernatant to pH 2.0 followed by solvent extraction (Smyth et al., 2010b). For their further purification, various chromatographic methods may be employed such as adsorption, ion-exchange, or thin layer chromatography as well as crystallization (Baker and Chen, 2010; Hu and Ju, 2001). In general, this method is time consuming as the extraction step needs to be repeated three times. Moreover, an equal volume of solvent to supernatant (ratio 1:1) needs to be used. As mentioned above, this results in the increase of biosurfactants’ cost and disposal of toxic solvents is an additional issue. An alternative method that can process large volumes of culture supernatants rapidly, continuously, and at low cost is ultrafiltration (UF). UF is used to concentrate, recover, and purify different types of biomolecules according to their size and molecular mass (Najmi et al., 2018). It is performed under relatively low temperatures and pressures and does not require phase changes or chemical additives. Therefore, products with high quality may be recovered, as the extent of their possible denaturation, degradation, and deactivation during the process is minimized (Charcosset, 2006). However, it is not a very selective separation method and it is mainly for recovery and concentration of products. In the case of biosurfactants, UF can lead to a more selective separation by taking advantage of a unique characteristic of surfactant molecules, which is the formation of micelles, at concentrations at and above their critical micelle concentration (CMC). These supramolecular structures have nominal molecular diameters up to two to three orders of magnitude larger than the single molecules. In this way micellar aggregates are retained by relatively high molecular weight cutoff (MWCO) membranes while impurities such as salts, peptides, small proteins, and free amino acids are washed out due to their lower molecular mass (Mulligan and Gibbs, 1990). Mulligan and Gibbs (1990) successfully employed this principle for the recovery and purification of surfactin and rhamnolipids from complex fermentation broths in one UF step. Recovery and purification of surfactin from broth samples in one step of UF was further studied by Sen and Swaminathan (2005) by using a stirred cell device. In this study, they managed to recover surfactin with 70% purity based on its CMC. Another reported method by Lin and Jiang (1997) for the recovery and purification of

Membrane Filtration of Biosurfactants Chapter | 3

89

TABLE 3.3 Summary of Biosurfactant Downstream Processing Methods Downstream Process

Advantages

Disadvantages

References

Acid precipitation (addition of HCl until pH 2)

Low cost

Low purity

Conjunction with solvent extraction or ultrafiltration enhances selectivity

Shah et al. (2016), Smyth et al. (2010b)

Ammonium sulfate precipitation (40% w/v)

Higher recovery yield compared to acid precipitation

Coprecipitation with other small molecules (e.g., riboflavin in Bacillus subtilis) Efficient only for crude biosurfactant recovery

Shah et al. (2016)

Zinc sulfate precipitation (40% w/v) Solvent extraction

Adsorption onto activated carbon

Alajlani et al. (2016), Shah et al. (2016)

Higher recovery yield compared to precipitation Solvents may be reused Best partitioning coefficient of biosurfactant against water achieved using chloroform/methanol

Partial purification

In situ recovery

Concomitant adsorption of other molecules present in spent media has not been considered

Dubey et al. (2005), Liu et al. (2007)

Limited use due to compound specificity

Chen et al. (2008c)

Carbon can be reused for up to three consecutive cycles without decreasing adsorption efficiency

High solvent volumes Increases process cost

Kuyukina et al. (2001), Shah et al. (2016)

High environmental impact Usually prior acid precipitation is required to enhance efficiency

Reduces solvent volume Minimizes product degradation Avoidance of end product inhibition Adsorption onto ion-exchange resins

In situ recovery Biosurfactants recovered in aqueous buffer systems due to their lipophilic ability

(Continued )

90

Separation of Functional Molecules in Food by Membrane Technology

TABLE 3.3 (Continued) Downstream Process

Advantages

Disadvantages

References

Crystallization

High purity

Yang et al. (2012)

Allows separation of lactonic and acidic forms in sophorolipids

Specific to sophorolipids due to high productivity that promotes crystallization of lactonic sophorolipids and recovery of crystals

High purity

High cost

Allows separation of structural groups

Time-consuming

Smyth et al. (2010a)

Low cost

Nonaseptic systems may hamper continuous biosurfactant production

Energy-saving Low cost

Column chromatography

Foam fractionation

High effectiveness Prevention of antifoam chemicals during fermentation Possible continuous removal and in situ recovery

Membrane ultrafiltration

Low cost High purity Continuous recovery Simple Easily scalable

Not applicable in industrial scale Baker and Chen (2010), Dı´az De Rienzo et al. (2016)

Persistent foam difficult to collapse may be formed depending on foam stability of the biosurfactant produced Membrane clogging and formation of biosurfactant layers on the membrane are the major drawbacks in process scale-up

Andrade et al. (2016), Chen et al. (2008a), Juang et al. (2008)

surfactin from fermentation broths consisted of two-step UF (see section 3 for further details on this). In that study, they reported a recovery of 95% although there was no report on the purity of the final product.

3.2.1 The Ultrafiltration Process and Equipment Over the last two decades, new membranes and modules, as well as new operating systems, have been developed in order to meet the requirements of the biotechnology industry (van Reis and Zydney, 2007). UF is a membrane

Membrane Filtration of Biosurfactants Chapter | 3

91

process that lies between nanofiltration and microfiltration, in which the pore sizes of the membranes used are in the range of 1 nm (nanofiltration) to 0.05 μm (microfiltration), and are made mostly of polymeric materials such as polysulfone (PS)/polyethersulfone (PES), polyvinylidene fluoride, polyacrilonitrile, cellulosics, aliphatic, and polyimide by phase inversion process (Mulder, 1997). UF membranes are characterized in terms of their ability to retain proteins of a particular molecular weight (MW) and thus the term molecular weight cutoff (MWCO) is used to define the size of protein that would be (almost completely) retained by a particular UF membrane. Even though it is known that molecular weight alone does not determine the size of a protein, and many manufacturers use dextrans rather than proteins to characterize UF membranes, the term nominal molecular weight cutoff (NMWCO) is still being used. Thus, UF covers particles and molecules that range from approximately 1000 in molecular weight to about 500,000 Da (Cheryan, 1998). The two standard modes of UF are dead-end and crossflow configurations. In the crossflow mode, the fluid to be filtered flows parallel to the membrane surface and permeates through the membrane due to a pressure difference.

3.2.2 Assessment of Separation Performance The assessment of the separation of biosurfactants by membrane filtration has been mainly carried out based on the following parameters: Rejection coefficient ðRÞ 5

Cf 2 Cp Cf

where Cf is the concentration of the molecule (e.g., biosurfactant) in the feed; Cp is the concentration of the molecule in the permeate. The rejection coefficient can be also presented as a percentage (R 3 100) Recovery % 5

Ms 3 100 Mf

where Ms is the mass of biosurfactant in the permeate or retentate and Mf is the mass of the biosurfactant in the feed. The purity of the biosurfactant in relation to protein as main contaminant has been defined as: Pp 5

Ms Mp 1 Ms

where Ms is the mass of biosurfactant and Mp the mass of protein in the recovered fraction (permeated); both determined based on measurements of the concentration of biosurfactant and total protein respectively. Also, purity can be determined based on dry weight as a fraction of biosurfactant mass over total mass.

92

Separation of Functional Molecules in Food by Membrane Technology

3.3 ULTRAFILTRATION OF LIPOPEPTIDE BIOSURFACTANTS Lipopeptide biosurfactants are cyclic compounds, with their hydrophilic part consisting of 7 (surfactins and iturins) to 10 (fengycins) amino acids linked to a lipid hydrophobic moiety. They are produced by both Gram-positive bacterial species, such as Bacillus and Streptomyces as well as Gramnegative, that is, Pseudomonas and Serratia species (Cochrane and Vederas, 2014; Dubern, 2006; Ine`s and Dhouha, 2015). They are widely known for their antimicrobial and antifungal activities and also their strong surface activities (Marchant and Banat, 2012; Mejri et al., 2017; Mihalache et al., 2017). Lipopeptides produced by Bacillus strains are the biosurfactants that have been most studied in relation to their recovery by UF processes.

3.3.1 Surfactin Separation by the Two-Step Ultrafiltration Method Surfactin produced by fermentation of Bacillus subtilis has been recovered and purified from the fermentation broth applying the two-step UF process described by Isa et al. (2007) and illustrated in Fig. 3.1. In the first step, surfactin micelles were retained successfully with membranes of MWCO 10 kDa. As shown by dynamic light scattering (DLS) measurements, the retentate contained particles with a mean diameter of approximately 9 nm, comparable with those found originally in the fermentation broths of 8 nm (Fig. 3.2). Given that the same particle sizes were measured in micellar solutions of a standard surfactin aqueous solution these particles were identified as surfactin micelles. After the addition of 50% (v/v) methanol solution to this retentate, particles with a mean diameter of 9 nm corresponding to surfactin micelles were no longer present in the solution, indicating the rupture of such structures. Moreover, the presence of larger particles was detected, with a mean diameter of 100 nm. Such large particles were identified as protein aggregates; methanol induced aggregation of protein can be assumed to be based on the same principle of ethanol induced aggregation. So, it was concluded that the effective size-based separation of surfactin from proteins

FIGURE 3.1 Two-step ultrafiltration process for surfactin purification where UF1 is the first step of ultrafiltration and UF2 is the second step of ultrafiltration.

Membrane Filtration of Biosurfactants Chapter | 3

93

FIGURE 3.2 Size distribution measured by dynamic light scattering of (A) surfactin standard, (B) fermentation broth, (C) retentate of first step of ultrafiltration, and (D) retentate after addition of methanol (50%).

TABLE 3.4 Summary of Recovery and Purification of Surfactin Obtained by the Two-Step Ultrafiltration Process at Different Conditions of Initial Concentration (Co), Membrane Type and MWCO (in Brackets) References

Co (mg/L)

Membrane

Re (UF1) (%)

Solvent

Re (UF2) (%)

Purity (UF2)

Chen et al. (2007)

400

PES (100)

93

EtOH 33%

87

85a

Chen et al. (2008a)

4020

CE (100)

97

EtOH 50%

80

74a

Isa et al. (2007)

583

RC (10)

98

MeOH 50%

96

93b

Isa et al. (2008)

596

PES (10)

83

MeOH 50%

94

96b

Lin and Jiang (1997)

250

RC (30)

98

MeOH 50%

95

98a



Combined with ammonium sulfate 23%. Purity determination in relation to total dry weight. Purity determination in relation to total protein.

a

b

occurred due to not only the disruption of micelles into free surfactin molecules, but also the formation of large protein aggregates. Therefore, the successful purification of surfactin by UF relied on (1) formation of micelles at and above the CMC of the biosurfactant, (2) retention of successful disruption of micelles by the alcoholic solution and the subsequent recovery of surfactin in the permeate, and (3) alcohol induced aggregation of protein and its subsequent retention. A summary of surfactin purification at different process conditions is shown in Table 3.4.

94

Separation of Functional Molecules in Food by Membrane Technology

The principle and the success of the separation was established using centrifugal devices and a stirred cell filtration device; no differences in results were found between these two devices. However, the scalability was tested using a crossflow filtration unit, which allows continuous operation, higher flow rates, and reduced fouling. High recoveries and purities were obtained with both 10 kDa MWCO regenerated cellulose (RC) membranes and PES (Isa et al., 2008) and overall the performance of the separation was similar to that obtained with the dead end filtration devices (see Table 3.4). The other operating parameter to consider in the scalability of the process is the MWCO of the membrane, as higher MWCO would enable operating at higher flow rates. So higher MWCOs were tested up to 300 kDa. High retention and recovery of surfactin was obtained with MWCO up to 100 kDa; however there was poor rejection by the 300 kDa membrane when tested on a binary mixture of surfactin and mycosubtilin (Jauregi et al., 2013). Further considerations of the scalability and/or applicability of the process led to testing different alcoholic solutions; particularly, the replacement of methanol by ethanol as the latter is less toxic. A summary of the recovery and purity obtained for surfactin with different hydroalcoholic solutions is given in Table 3.4. The replacement of methanol by ethanol led to a reduction in recovery and purity. A further increase in ethanol to 75% led to the same results as with methanol (Jauregi et al., 2013).

3.3.2 Hybrid Recovery Processes Using Ultrafiltration Hybrid processes may be more complicated, compared to two-step UF, but may also come with benefits due to reduced membrane fouling. In their study, Chen et al. (2008b) examined the recovery of surfactin applying ammonium sulfate salting-out, UF, nanofiltration (NF), and also their combination. Several factors were investigated during recovery such as the initial concentration of surfactin (0.21
3.62 g/L); the concentration of ammonium sulfate, which reached up to 46% w/v; solvent ethanol percentage (up to 44% v/v) for micelle-dissociation; as well as membranes of different molecular weight cutoff ranging between 1 and 300 kDa. According to their results, when the surfactin was solubilized first in 33% v/v of ethanol and then NaCl (23% w/v) was added (following this order was important), surfactin micelles were efficiently dissociated and NaCl salting-out effect caused protein precipitation. This resulted in a three-phase system: (1) an ethanol-rich upper layer where most of the surfactin was found, (2) a white precipitate (protein) middle layer, and (3) a water-rich lower layer. This resulted in 93% recovery and 68% purity for surfactin. Compared to the UF-alone process using a PES membrane with 100 kDa MWCO lower surfactin yield (68%) but higher purity (83%) was obtained with the UF. Combination of

Membrane Filtration of Biosurfactants Chapter | 3

95

salting-out and UF showed advantages regarding recovery yield and purity compared to the UF and salting-out alone processes. When surfactin concentration was at 2054 mg/L this hybrid process resulted in a recovery yield of 81% and a purity of 79%, characterized by moderate flux. On the other hand, when salting out was combined with NF results were similar. However, this process was characterized by extremely low flux. Moreover, performing UF first and then salting-out resulted in comparable performance to UF alone. Overall they concluded that the salting-out step led to high recovery but low purity (,69%) whereas UF led to high purity but lower yield (,68%). The combination of the two led to improved performance and particularly when salting-out was applied first followed by UF, which resulted in 81% yield and 78% purity.

3.3.3 Separation of Lipopeptide Mixtures Jauregi et al. (2013) investigated the application of the two-step UF process and showed that the size distributions of mycosubtilin and surfactin single solutions were different from that of the mixture with a monomodal size distribution and an equivalent average diameter of 8 nm, similar to that obtained for single solutions of mycosubtilin suggesting the formation of mixed micelles. These results show that the selective separation of the lipopeptides will be difficult by this membrane separation process. It may be possible to explore strategies that lead to changes in the micellar size of one of the lipopeptides. For example, in our group we found that addition of calcium had an effect on the micelle size of fengycin (unpublished data). However, the formation of mixed micelles may still compete with the counter ions’ effect on micellar size.

3.3.4 Recovery of Lipopeptides From Complex Culture Medium Biosurfactants can be produced using industrial wastes and byproducts as culture medium and this is a way to reduce process cost. For example cassava wastewater has been used successfully in the production of surfactin from B. subtilis (Barros et al., 2008). Cassava wastewater is the main residue of the cassava starch industry. It contains 74 g/L of total solids, and its composition is as follows: proteins 1%; lipids 0.2%; fermentable carbohydrates including glucose, fructose, and saccharose 35%; starch 30%; fibers 1%; nitrogen 0.22%; phosphorus 0.03%; calcium 0.4%; sodium 0.002%; niacin 0.0006%; among others. However, there is a lack of knowledge about technical feasibility of the downstream process that uses industrial wastes as culture medium and the purification of the products obtained from those are rarely reported.

96

Separation of Functional Molecules in Food by Membrane Technology

Andrade et al. (2017) showed for the first time the purification of surfactin using cassava wastewater as fermentation medium and explored different separation strategies including the two-step UF process using PES membranes of 100 kDa MWCO for the first step and 50 kDa for the second step (except in one of the processes where 100 kDa membrane was used in this step too). In all the cases the UF process was applied to the biosurfactant recovered in the foam overflow so foam separation was applied as a pretreatment (Fig. 3.1). This was considered necessary to remove other contaminants in the complex fermentation medium that could result in the fouling of the membrane, and it also resulted in the concentration of the biosurfactant and protein in the foam. Andrade et al. (2016) found that the main differences between the three strategies was in the separation of protein in the second step, that is, in the purity. The worst separation of proteins from surfactin was obtained when using the 100 kDa membrane for the UF2 as both surfactin and protein were rejected similarly leading to low purity (44% purity and little improvement from that in the feed, 35%). The purification improved when a 50 kDa membrane was used in UF2 with 59% purity (from 41% in the feed). However, the best results were obtained when an additional precipitation/extraction step was added before the UF with 80% purity (from 53% in the feed). This compared well with the separation of surfactin produced in synthetic medium using a PES 100 kDa in both steps in which the purity was 94% (Jauregi et al., 2013) and 96% using a PES 10 kDa (Isa et al., 2008). Better results obtained with the synthetic culture medium than with the cassava wastewater may be due to lower protein content in the culture (feed) of the former (75 mg/L whilst 195 mg/L in cassava medium), which facilitated the purification; also when using cassava wastewater proteins were more effectively rejected with a lower MWCO membrane, which suggests different types of proteins present in these two mediums. Thus, it can be concluded that cassava wastewater is a low-cost culture medium. However, the high content of proteins in cassava wastewater make the purification more difficult, and additional purification methods may be required in order to remove the protein more efficiently. Alternatively, protein may be removed from the cassava wastewater as a pretreatment, for instance by precipitation. On the other hand, the protein is a valuable nitrogen source, which has a significant effect on the production of surfactin from B. subtilis (preferably organic nitrogen). It has been reported that the lower the nitrogen source the lower the surfactin production (Davis et al., 1999). This brings about some interesting issues concerning the use of cassava wastewater as a fermentation medium for the production of biosurfactants, since on one hand the use of industrial waste as culture medium could reduce the cost of production, but on the other hand this can make the separation and purification of the products more complicated, which will result in an increase in process cost.

Membrane Filtration of Biosurfactants Chapter | 3

97

3.4 ULTRAFILTRATION OF RHAMNOLIPID BIOSURFACTANTS Rhamnolipids are classified as low molecular mass glycolipid biosurfactants and comprise of rhamnose molecules, glycosidically linked to long-chain hydroxyl fatty acids consisting of 8 to 16 carbon atoms. According to the number of rhamnose molecules present they are categorized as mono- and dirhamnolipids, consisting of one or two rhamnose molecules respectively. Their molecular mass ranges between 302 and 989 Da, while about 60 congeners and homologues have been identified to be produced mainly by Pseudomonas, Acinetobacter, Enterobacter, Pantoea, and Burkholderia species (AbdelMawgoud et al., 2010). They are produced in mixtures of rhamnolipid congeners, and can reduce the ST to 25
30 mN/m and the IFT to 1
4 mN/m while their CMCs range between 20 and 250 mg/L (Kourmentza et al., 2018). Mulligan and Gibbs were the first ones to use UF membranes for the separation of biosurfactants, studying a surfactin and rhamnolipid model (Mulligan and Gibbs, 1990). A variety of molecular weight cutoff (MWCO) membranes were investigated for retaining a rhamnolipid with a molecular mass of 802 Da and according to their findings YM 10 membrane (a centrifugal device with a 10 kDa MWCO membrane) was the most appropriate one for rhamnolipid recovery. In another study, Park and colleagues examined various methods for the recovery and purification of rhamnolipids present in the culture broth and produced by Pseudomonas aeruginosa YPJ-80 using glucose and glycerol (Park et al., 1998). In particular they examined the effect of the pH in the culture supernatant during UF using a magnetically stirred UF cell with YM10 membrane, characterized by a MWCO of 10 kDa. According to their results it was shown that the recovery of rhamnolipids, determined as rhamnose equivalents, was higher at lower pH values. For pH 3 recovery yield reached up to 91%, whereas for pH 7 and 8 recovery of rhamnolipids was 87.1% and 75.7% respectively. It has to be mentioned that UF showed the highest recovery compared to other methods used in the same study. For example, precipitation with HCl and 2% ammonium sulfate resulted in recoveries of 63.9%
78.8%. Extraction with a mixture of chloroform:methanol (2:1 v/v) of the acidified supernatant (pH 5 2) yielded 89% of rhamnolipids. In addition, adsorption on Amberlite XAD-4 ion exchange resin yielded only 29.6% of rhamnolipids. Later on, a two-stage method based on UF was suggested for the recovery and purification of rhamnolipids from the culture broth (Witek-Krowiak et al., 2011). This approach was proposed having previous examples of successful implementation of the ultrafiltration-based recovery of lipopeptides, (see above). The aim of the first stage of the UF process was to concentrate the sterile culture supernatant and retain the rhamnolipid micelles. Apart from the rhamnolipids, other compounds such as proteins were also present

98

Separation of Functional Molecules in Food by Membrane Technology

in the retentate. For the successful separation of rhamnolipids from proteins a second UF stage followed where methanol was added to the retentate in order to disrupt rhamnolipid micelles and recover the biosurfactant in the permeate while proteins were retained in the retentate. The authors studied the effect of membrane type on the effectiveness of rhamnolipid recovery and the effect of pressure on the permeate flux during the first UF step. They also investigated how the concentration of methanol added in the second stage affected the process yield, by determining the rhamnolipid concentration in the feed, retentate, and permeate. According to their results, the highest retention efficacy was achieved with membranes of MWCO below 10 kDa; they selected a 5 kDa U005 Microdyn Nadir membrane for further experimentation because of its high retention rate and high yields. Furthermore they observed a significant flux reduction due to accumulation of both proteins and micelles at the surface of the membrane. Moreover, additional effects such as membrane fouling may occur that have a negative impact on permeate flux. Pressure increase led to a slight increase in permeate flux and further experiments were performed at 140 kPa. Finally, it was shown that methanol concentrations higher than 50% allowed the recovery of more than 90% of rhamnolipids and results were similar to those previously reported for the separation of surfactin. In addition, the dilution of the permeate obtained from the first UF step significantly affected the rhamnolipid recovery yield. Optimum dilution was found to be at a ratio of 1:10 v/v, whereas a fivefold dilution resulted in reduced yield. Overall, the rhamnolipid recovery using a two-stage UF reached up to 84.3%. Optimization of the two-step UF process was also performed. Micelle size at different ethanol concentrations ranging between 0 and 30% and pH between 5 and 9 was measured by DLS. It was shown that when the pH and ethanol concentration increased the mean diameter of micelles decreased. During the first UF step when pH increased from 5 to 8 membrane rejection of rhamnolipids decreased from 98% to 90%. In the second UF step rhamnolipids in the permeate reached 87% at pH 9 and with the addition of 30% v/v ethanol. Thus, it was concluded that recovery was more efficient when UF-1 was performed at pH 5 and UF-2 performed at pH 9 with 30% ethanol.

3.5 ULTRAFILTRATION OF MANNOSYLERYTHRITOL LIPIDS Mannosylerythritol lipids (MEL) are glycolipid biosurfactants mainly produced by yeast strains that belong to Pseudozyma (recently known as Candida) and Ustilago species (Arutchelvi and Doble, 2011). They consist of the hydrophilic group 4-O-ß-D-mannopyranosyl-meso-erythritol, and their hydrophobic moiety may comprise of a fatty acid and/or an acetyl group. They can be further classified according to their degree of acetylation at C-4 and C-6 positions (Kitamoto et al., 1990). MEL-A is the diacetylated structure, MEL-B and MEL-C are the monoacetylated compounds at positions C-6 and

Membrane Filtration of Biosurfactants Chapter | 3

99

C-4 respectively, while MEL-D is completely deacetylated. MEL reduce the ST to 25
34 mN/m, the IFT to about 1
2 mN/m, and their CMC is found between 2.6 3 1024 and 6.4 3 1026 M (Fukuoka et al., 2007b; Tomotake Morita et al., 2009b). Depending on their type MEL are characterized by different selfassembling properties that include liposomes, self-assembled monolayer, sponge phase, bicontinuous cubic phase, and liquid lyotropic crystals (Imura et al., 2007). At concentration higher than 1 mM MEL-A forms a sponge phase, while MEL-B and MEL-C form giant unilamellar vesicles with a diameter higher than 10 nm (Imura et al., 2004; Kitamoto et al., 2009). Compared to other types of biosurfactants, MEL are produced in high yields and they have attracted interest due to their antitumor activities, ability to induce cell differentiation and apoptosis, potential to be used as a vesicle in gene delivery, and inhibition of ice agglomeration, and in cosmetics, due to their ceramide-like skin care property (Arutchelvi and Doble, 2011). In general, the most widely used methodology employed for the purification of MEL is liquid
liquid extraction followed by silica column and HPLC chromatography using silica gel columns (Rau et al., 2005a,b). A recently published study by de Andrade and colleagues investigated the application of UF for the purification of MEL for the first time (Andrade et al., 2017). Production of MEL was performed by cultivation of Pseudozyma tsukubaensis on cassava wastewater. A similar separation strategy to the one developed for surfactin (Fig. 3.1) was followed. First, MEL was recovered in the foam produced during fermentation. The main impurities were proteins that were either present in cassava wastewater or produced by the strain itself. UF was applied after foam fractionation using a PES 100 kDa membrane in a centrifugal UF device. The separation was very efficient with 80% of MEL biosurfactants being retained while 95% of the protein permeated. So high purity was achieved in one UF step (purity 5 0.86). The size distribution measurements of the feed (foam) by DLS showed particles with a monomodal distribution and an equivalent average diameter of 1120 nm at 610.74 mg/L of MEL and d 5 1754 nm at 502.71 mg/L of MEL. This remarkable selective separation obtained in just one step can be explained by the much larger micellar aggregates formed by MEL as compared to those formed by surfactin and mycosubtilin. This process was further scaled up using a 500 mL (250 mL of working volume) and a crossflow UF unit. The initial feed volume was reduced to 25 mL while flux decreased from 90 to 55 L/m2/h during the first 25 minutes and to 45 L m22 h21 during the last 20 minutes due to the presence of proteins. MEL concentration in the retentate increased from 295 to 860 mg/L, indicating a three-fold increase in purity, while the concentration of proteins decreased. Therefore, in agreement with the small-scale UF process, the self-aggregation properties of MEL enabled its purification by membrane filtration in one step. The final product resulted in a highly concentrated

100

Separation of Functional Molecules in Food by Membrane Technology

solution of MEL (860 mg L21) at high purity (from 0.18 to 0.58), that is, at both scales a three-fold increase in purity was achieved.

3.6 FURTHER CONSIDERATIONS 3.6.1 Membrane Choice The most widely used membranes in industry are PS, PES, and RC membranes, which have an asymmetric (skinned) structure, with the thin skin providing the desired selectivity while the more porous structure provides the necessary mechanical support; these membranes have high thermal stability and chemical resistance that allows the use of harsh cleaning chemicals (van Reis and Zydney, 2007). PS and PES membranes are advantageous in terms of having wide temperature limits, wide pH tolerance, fairly good chlorine resistance, easy fabrication for variety of configurations and modules, wide range of pore sizes from 1000 MWCO to 0.2 μ in commercial-size modules, and good chemical resistance to aliphatic hydrocarbons, fully hydrogenated hydrocarbons, alcohols, and acids. In comparison, RC membranes are very hydrophilic and have exceptional nonspecific protein-binding properties. In addition, RC also has good resistance to some common solvents such as 70% butanol and 70% ethanol, and it can tolerate temperatures up to 75  C (Mulder, 1997). In another study, by Isa et al. (2007), regarding the purification of surfactin by the two-step UF process, PES membranes were found to be more affected by concentration polarization (i.e., a decrease in flux due to accumulation of the separated molecules on the surface of the membrane) than RC membranes. This was ascribed to hydrophobic interactions between PES and the aggregated proteins in the second step of UF. In addition, a significant decrease in flux was observed between PES membranes after cleaning and new membranes, which indicated some level of fouling after the second step of surfactin separation. RC membranes, however, showed no significant differences between the cleaned membrane and new indicating no significant fouling with these membranes; this is probably due to the more hydrophilic nature of this type of membrane as compared to PES. However, despite the reduction in flux higher selectivity of surfactin separation was obtained with PES membranes than with RC. This was ascribed to differences in interaction between surfactin molecules and these two membranes. Isa et al (2008) found that surfactin in methanolic solutions were strongly negatively charged (zeta potential 5
27 mV) and according to Salgin et al. (2006) PES had a negative charge of
48.33 mV (in 0.01 M KCl and RC
5.6 mV according to Urba´nski et al. (2002)). Therefore, it was concluded that less surfactin was retained with PES than with RC membrane because electrostatic repulsive interactions between surfactin molecules and the more negatively charged PES membrane increased the permeability of surfactin. A similar effect was observed for SDS; higher

Membrane Filtration of Biosurfactants Chapter | 3

101

permeability was obtained with PES and PS membranes than with RC (Majewska-Nowak et al., 2005). Based on these findings it was concluded that PES membrane was the most suitable membrane for high recovery and selectivity of surfactin separation especially in the second step of UF.

3.6.2 Membrane Cleaning Although membrane filtration can offer many advantages compared to other types of separation technology, the membrane fouling phenomenon, which refers to the formation of deposits on the membrane surface and/or inside the pores of the membrane, is an important limitation of this technology. In the food and dairy industries particularly, membrane fouling by organic molecule adsorption such as proteins can be problematic as it leads to the blockage of the membrane and thereby reduces the throughput or flux. Fouling is governed by a number of factors related to both the foulant and the membrane such as charge and structure of foulant, composition, hydrophobicity, and surface charge of the membrane. Membrane fouling results in loss of productivity due to reduced equipment efficiency, increased material cost of cleaning, and contamination problems due to the growth of microorganisms. Cleaning of membranes can be carried out using cleaning agents such as acids, alkalis, surfactants, disinfectants, and combined cleaning materials. Selection of appropriate chemical cleaning agents is essential as incompatible combinations of cleaning agents and membrane material could lead to irreversible flux loss and reduction in membrane life. Isa et al. (2008) evaluated the permeability of RC and PES membranes based on the flux rates of distilled water at various steps during filtration and after cleaning. The cleaning consisted of rinsing with distilled water for 30 minutes followed by 1.5 hours in 0.1 M NaOH solution for RC membranes and for PES rinsing with water for 30 minutes followed by mixed solution of 0.5 M NaOH with 0.1% SDS for 2 or 3 hours. Although both membranes were affected by similar amount of protein fouling (about 35% protein nonaccounted for, which was assumed to be fouling the membrane), a decrease of 44% and 75% in the flux rates of RC and PES membranes respectively was recorded after the second step of UF. This was ascribed to the pore dilation and constriction experienced by RC and PES membranes respectively in the presence of organic solvents such as methanol as observed by Lencki and Williams (1995) and Isa et al. (2007). The cleaning of the RC membrane after the two-step UF process led to 99% flux recovery compared to the PES membrane, which saw 89% and 98% flux recovery after 2 and 3 hours of cleaning, respectively. Membrane fouling was also addressed in the recovery of rhamnolipids by the two-step UF process using polysulfone (PSU) and PSU-g-PEG (PSU surface modified by poly-ethylene glycol) hollow membranes in a crossflow operation mode by Long et al. (2012). By using PSU-g-PEG membrane

102

Separation of Functional Molecules in Food by Membrane Technology

90.2% of the rhamnolipids were rejected in the first UF step while using PUS membrane rejection was at 88.4%. During the second UF step, when ethanol was used at a final concentration between 40 and 50% membranes swelling was observed and in some cases, this resulted in membrane failure. For this reason, a lower ethanol concentration (20%) was chosen. The flux in the first UF step, with PSU-g-PEG membrane remaining constant for the first 20 min followed by a slight decrease of 11% whilst the flux through PSU seemed to decrease continuously, reaching 58% after 60 min, which was the end of the operation. For both membranes, similar trends were recorded during the second UF step. Membrane fouling was then examined at the end of the two-step UF process. It was shown that for PSU membrane water rinsing retained one-third of the initial flux and further flushing with 0.01 M NaOH did not enhance water flux. On the other hand, PSU-g-PEG membrane retained half of the initial water flux just with water rinsing, while flux was restored at 90% after flushing with 0.01 M NaOH and completely recovered after overnight immersion in 0.01 M NaOH.

3.7 CONCLUSIONS AND FINAL OUTLOOK Ultrafiltration-based separations can be applied for the successful recovery and purification of lipopeptides from clarified fermentation cultures. The two-step UF has proven to be particularly successful for the recovery and purification of lipopeptides from Bacillus subtilis such as surfactin and mycosubtilin. Other biosurfactants such as MEL were successfully recovered following the same UF process; however, MEL was purified from proteins in the medium only in one step. On the other hand, the UF process was unable to separate individual biosurfactants (e.g., surfactin and mycosubtilin) from biosurfactant mixtures. This is probably due to the formation of mixed micelles. The UF process can be scaled up with a crossflow filtration unit and using PES membranes up to 100 kDa MWCO; 300 kDa MWCO membranes proved to be inefficient in the separation of biosurfactant micelles. In the second step of UF the following alcoholic solutions proved to be effective in the disruption of the micelles: 50% methanol and 75% ethanol. PES membranes were more affected than RC membranes by the alcoholic solutions, which led to reduced flux; however, the higher electrostatic charge in PES than RC membranes led to higher permeability of surfactin. Using industrial waste as culture medium (e.g., cassava) could reduce the cost of production; however, as it was found in the separation of surfactin and MEL, this resulted in less pure products than when synthetic medium was used. Thus, in order to obtain products of higher purity, additional steps prior to UF would be necessary, which would increase process cost. In conclusion, using more complex medium from industrial waste may be only viable if the requirement for the purity of the product is not very high and

Membrane Filtration of Biosurfactants Chapter | 3

103

hence, no additional steps are required. Overall it can be concluded that the ultrafiltration-based process is an efficient and scalable process for the recovery and purification of biosurfactants.

REFERENCES Abdel-Mawgoud, A.M., Le´pine, F., De´ziel, E., 2010. Rhamnolipids: diversity of structures, microbial origins and roles. Appl. Microbiol. Biotechnol. 86, 1323
1336. Available from: https://doi.org/10.1007/s00253-010-2498-2. AGAE Technologies LLC, 2018a. Applications of rhamnolipid products [WWW Document]. https://www.agaetech.com/applications/ (accessed 01.17.18). AGAE Technologies LLC, 2018b. Agae Technologies’ Rhamnolipid Product Catalog. Alajlani, M., Shiekh, A., Hasnain, S., Brantner, A., 2016. Purification of bioactive lipopeptides produced by bacillus subtilis strain BIA. Chromatographia 79, 1527
1532. Available from: https://doi.org/10.1007/s10337-016-3164-3. Alimadadi, N., Soudi, M.R., Talebpour, Z., 2018. Efficient production of tri-acetylated monoacylated mannosylerythritol lipids by Sporisorium sp. aff. sorghi SAM20. J. Appl. Microbiol. 124, 457
468. Available from: https://doi.org/10.1111/jam.13642. Andrade, C.J., Barros, F.F.C., de Andrade, L.M., Rocco, S.A., Luis Sforc¸a, M., Pastore, G.M., et al., 2016. Ultrafiltration based purification strategies for surfactin produced by Bacillus subtilis LB5A using cassava wastewater as substrate. J. Chem. Technol. Biotechnol. 91, 3018
3027. Available from: https://doi.org/10.1002/jctb.4928. Andrade, C.J., de, Andrade, L.M., de, Rocco, S.A., Sforc¸a, M.L., Pastore, G.M., Jauregi, P., 2017. A novel approach for the production and purification of mannosylerythritol lipids (MEL) by Pseudozyma tsukubaensis using cassava wastewater as substrate. Sep. Purif. Technol. 180, 157
167. Available from: https://doi.org/10.1016/j.seppur.2017.02.045. Arau´jo, F.D., da, S., Arau´jo, W.L., Eberlin, M.N., 2017. Potential of Burkholderia seminalis TC3.4.2R3 as biocontrol agent against Fusarium oxysporum evaluated by mass spectrometry imaging. J. Am. Soc. Mass Spectrom. 28, 901
907. Available from: https://doi.org/10.1007/ s13361-017-1610-6. Arutchelvi, J., Doble, M., 2011. Mannosylerythritol lipids: microbial production and their applications. In: Steinbu¨chel, A. (Ed.), Biosurfactants From Genes to Applications. Springer-Verlag Berlin Heidelberg, pp. 145
178. Available from: https://doi.org/10.1007/978-3-642-14490-5. Awada, S.M., Awada, M.M., Spendlove, R.S., 2011. Method of controlling pests with biosurfactant penetrants as carriers for active agents. US20110319341 A1. Ayed, H.B., Hmidet, N., Be´chet, M., Chollet, M., Chataigne´, G., Lecle`re, V., et al., 2014. Identification and biochemical characteristics of lipopeptides from Bacillus mojavensis A21. Process Biochem. 49, 1699
1707. Available from: https://doi.org/10.1016/j.procbio.2014.07.001. Baker, S.C., Chen, C.Y., 2010. Enrichment and purification of lipopeptide biosurfactants. In: Sen, R. (Ed.), Biosurfactants. Springer Science 1 Business Media, LLC, pp. 281
288. Barros, F.F.C., Ponezi, A.N., Pastore, G.M., 2008. Production of biosurfactant by Bacillus subtilis LB5a on a pilot scale using cassava wastewater as substrate. J. Ind. Microbiol. Biotechnol. 35, 1071
1078. Available from: https://doi.org/10.1007/ s10295-008-0385-y. Brack, C., Mikolasch, A., Schlueter, R., Otto, A., Becher, D., Wegner, U., et al., 2015. Antibacterial metabolites and bacteriolytic enzymes produced by Bacillus pumilus during

104

Separation of Functional Molecules in Food by Membrane Technology

bacteriolysis of Arthrobacter citreus. Mar. Biotechnol. 17, 290
304. Available from: https://doi.org/10.1007/s10126-015-9614-3. Burch, A.Y., Zeisler, V., Yokota, K., Schreiber, L., Lindow, S.E., 2014. The hygroscopic biosurfactant syringafactin produced by Pseudomonas syringae enhances fitness on leaf surfaces during fluctuating humidity. Environ. Microbiol. 16, 2086
2098. Available from: https:// doi.org/10.1111/1462-2920.12437. Cameotra, S.S., Makkar, R.S., Kaur, J., Metha, S., 2010. Synthesis of biosurfactants and their advantages to microorganisms and mankind. In: Sen, R. (Ed.), Advances in Experimental Medicine and Biology, Volume 672, Biosurfactants. Springer Science 1 Business Media, LLC, New York, USA, pp. 261
280. Chandran, P., Das, N., 2012. Role of sophorolipid biosurfactant in degradation of diesel oil by Candida tropicalis. Bioremediat. J. 16, 19
30. Available from: https://doi.org/10.1080/ 10889868.2011.628351. Charcosset, C., 2006. Membrane processes in biotechnology: an overview. Biotechnol. Adv. 24, 482
492. Available from: https://doi.org/10.1016/j.biotechadv.2006.03.002. Chen, H.L., Chen, Y.S., Juang, R.S., 2007. Separation of surfactin from fermentation broths by acid precipitation and two-stage dead-end ultrafiltration processes. J. Memb. Sci. 299, 114
121. Available from: https://doi.org/10.1016/j.memsci.2007.04.031. Chen, H.L., Chen, Y.S., Juang, R.S., 2008a. Flux decline and membrane cleaning in cross-flow ultrafiltration of treated fermentation broths for surfactin recovery. Sep. Purif. Technol. 62, 47
55. Available from: https://doi.org/10.1016/j.seppur.2007.12.015. Chen, H.L., Chen, Y.S., Juang, R.S., 2008b. Recovery of surfactin from fermentation broths by a hybrid salting-out and membrane filtration process. Sep. Purif. Technol. 59, 244
252. Available from: https://doi.org/10.1016/j.seppur.2007.06.010. Chen, H.L., Lee, Y.S., Wei, Y.H., Juang, R.S., 2008c. Purification of surfactin in pretreated fermentation broths by adsorptive removal of impurities. Biochem. Eng. J. 40, 452
459. Available from: https://doi.org/10.1016/j.bej.2008.01.020. Chen, J., Song, X., Zhang, H., Qu, Y., 2006. Production, structure elucidation and anticancer properties of sophorolipid from Wickerhamiella domercqiae. Enzyme Microb. Technol. 39, 501
506. Available from: https://doi.org/10.1016/j.enzmictec.2005.12.022. Cheryan, M., 1998. Ultrafiltration and Microfiltration Handbook. CRC Press, Boca Raton. Christova, N., Tuleva, B., Cohen, R., Ivanova, G., 2011. Chemical Characterization and Physical and Biological Activities of Rhamnolipids Produced by Pseudomonas aeruginosa BN10. Cochrane, S.A., Vederas, J.C., 2014. Lipopeptides from Bacillus and Paenibacillus spp.: A Gold Mine of Antibiotic Candidates 4
31. https://doi.org/10.1002/med Costa, S.G.V. a O., De´ziel, E., Le´pine, F., 2011. Characterization of rhamnolipid production by Burkholderia glumae. Lett. Appl. Microbiol. 53, 620
627. Available from: https://doi.org/ 10.1111/j.1472-765X.2011.03154.x. Cutler, A.J., Light, R.J., 1979. Regulation of hydroxydocosanoic acid sophoroside production in Candida bogoriensis by the levels of glucose and yeast extract in the growth medium. J. Biol. Chem. 254, 1944
1950. Davis, D.A., Lynch, H.C., Varley, J., 1999. The production of Surfactin in batch culture by Bacillus subtilis ATCC 21332 is strongly influenced by the conditions of nitrogen metabolism. Enzyme Microb. Technol. 25, 322
329. Available from: https://doi.org/10.1016/ S0141-0229(99)00048-4. Deepak, R., Jayapradha, R., 2015. Lipopeptide biosurfactant from Bacillus thuringiensispak 2310: a potential antagonist against Fusarium oxysporum. J. Mycol. Med. 25, e15
e24. Available from: https://doi.org/https://doi.org/10.1016/j.mycmed.2014.10.011.

Membrane Filtration of Biosurfactants Chapter | 3

105

de Sousa, T., Bhosle, S., 2012. Isolation and characterization of a lipopeptide bioemulsifier produced by Pseudomonas nitroreducens TSB. MJ10 isolated from a mangrove ecosystem. Bioresour. Technol. 123, 256
262. Available from: https://doi.org/10.1016/j.biortech.2012.07.056. Deshpande, M., Daniels, L., 1995. Evaluation of sophorolipid biosurfactant production by Candida bombicola using animal fat. Bioresour. Technol. 54, 143
150. Available from: https://doi.org/10.1016/0960-8524(95)00116-6. Dhanarajan, G., Sen, R., 2015. Cost analysis of biosurfactant production from a scientist’s perspective. In: Korasic, N., Vardar-Sukan, F. (Eds.), Biosurfactants Production and Utilization—Processes, Technologies, and Economics. CRC Press, Taylor & Francis Group, pp. 153
164. Dı´az De Rienzo, M.A., Kamalanathan, I.D., Martin, P.J., 2016. Comparative study of the production of rhamnolipid biosurfactants by B. thailandensis E264 and P. aeruginosa ATCC 9027 using foam fractionation. Process Biochem. 51, 820
827. Available from: https://doi.org/ 10.1016/j.procbio.2016.04.007. Domingos, D.F., de Faria, A.F., de Souza Galaverna, R., Eberlin, M.N., Greenfield, P., Zucchi, T.D., et al., 2015. Genomic and chemical insights into biosurfactant production by the mangrove-derived strain Bacillus safensis CCMA-560. Appl. Microbiol. Biotechnol. 99, 3155
3167. Available from: https://doi.org/10.1007/s00253-015-6377-8. Dubeau, D., De´ziel, E., Woods, D.E., Le´pine, F., 2009. Burkholderia thailandensis harbors two identical rhl gene clusters responsible for the biosynthesis of rhamnolipids. BMC Microbiol. 9, 263. Available from: https://doi.org/10.1186/1471-2180-9-263. Dubern, J.-F.B., 2006. Regulation of the Biosynthesis of Novel Cyclic Lipopeptides From Pseudomonas Putida Strain PCL1445. Dubey, K.V., Juwarkar, A.A., Singh, S.K., 2005. Adsorption—desorption process using woodbased activated carbon for recovery of biosurfactant from fermented distillery wastewater. Biotechnol. Prog. 21, 860
867. Available from: https://doi.org/10.1021/bp040012e. European Commission, 2012a. Innovating for Sustainable Growth: A Bioeconomy for Europe [WWW Document]. http://ec.europa.eu/research/bioeconomy/pdf/official-strategy_en.pdf. European Commission, 2012b. A Stronger European Industry for Growth and Economic Recovery. European Commission, 2011a. Mandate Addressed to CEN, CENELEC and ETSI for the Development of European Standards and Technical Specifications and/or Technical Reports for Bio-Surfactants and Bio-Solvents in Relation to Bio-Based Product Aspects. European Commission, 2011b. Mandate Addressed to CEN, CENELEC and ETSI for the Development of Horizontal European Standards and Other Standardisation Deliverables for Bio-Based Products. Favaro, G., Bogialli, S., Di Gangi, I.M., Nigris, S., Baldan, E., Squartini, A., et al., 2016. Characterization of lipopeptides produced by Bacillus licheniformis using liquid chromatography with accurate tandem mass spectrometry. Rapid Commun. Mass Spectrom. 30, 2237
2252. Available from: https://doi.org/10.1002/rcm.7705. Freire, D.M.G., de Araujo, L.V., de Araujo Kronemberger, F., Nitschke, M., 2010. Biosurfactants as emerging additives in food processing. In: Passos, M.L., Ribeiro, C.P. (Eds.), Innovation in Food Engineering: New Techniques and Products. CRC Press, Taylor & Francis Group, pp. 685
706. Available from: https://doi.org/10.1201/9781420086072. Fukuoka, T., Morita, T., Konishi, M., Imura, T., Kitamoto, D., 2007a. Characterization of new types of mannosylerythritol lipids as biosurfactants produced from soybean oil by a basidiomycetous yeast, Pseudozyma shanxiensis. J. Oleo Sci. 56, 435
442. Available from: https:// doi.org/10.5650/jos.56.435.

106

Separation of Functional Molecules in Food by Membrane Technology

Fukuoka, T., Morita, T., Konishi, M., Imura, T., Sakai, H., Kitamoto, D., 2007b. Structural characterization and surface-active properties of a new glycolipid biosurfactant, mono-acylated mannosylerythritol lipid, produced from glucose by Pseudozyma antarctica. Appl. Microbiol. Biotechnol. 76, 801
810. Available from: https://doi.org/10.1007/s00253-0071051-4. Funston, S.J., Tsaousi, K., Rudden, M., Smyth, T.J., Stevenson, P.S., Marchant, R., et al., 2016. Characterising rhamnolipid production in Burkholderia thailandensis E264, a nonpathogenic producer. Appl. Microbiol. Biotechnol. 100, 7945
7956. Available from: https:// doi.org/10.1007/s00253-016-7564-y. Gandhimathi, R., Seghal Kiran, G., Hema, T.A., Selvin, J., Rajeetha Raviji, T., Shanmughapriya, S., 2009. Production and characterization of lipopeptide biosurfactant by a spongeassociated marine actinomycetes Nocardiopsis alba MSA10. Bioprocess Biosyst. Eng. 32, 825
835. Available from: https://doi.org/10.1007/s00449-009-0309-x. Grady, E.N., MacDonald, J., Liu, L., Richman, A., Yuan, Z.C., 2016. Current knowledge and perspectives of Paenibacillus: a review. Microb. Cell Fact. 15, 1
18. Available from: https://doi.org/10.1186/s12934-016-0603-7. Gudin˜a, E.J., Rodrigues, A.I., Alves, E., Domingues, M.R., Teixeira, J.A., Rodrigues, L.R., 2015. Bioconversion of agro-industrial by-products in rhamnolipids toward applications in enhanced oil recovery and bioremediation. Bioresour. Technol. 177, 87
93. Available from: https://doi.org/10.1016/j.biortech.2014.11.069. Gunther, N.W., Nun˜ez, A., Fett, W., Solaiman, D.K.Y., 2005. Production of rhamnolipids by Pseudomonas chlororaphis, a nonpathogenic bacterium. Appl. Environ. Microbiol. 71, 2288
2293. Available from: https://doi.org/10.1128/AEM.71.5.2288-2293.2005. Haihang Industry Co. Ltd, 2018. Factory supply Sodium dodecyl sulfate CAS No.151-21-3 1
19. Ha¨ussler, S., Nimtz, M., Domke, T., Wray, V., Steinmetz, I., 1998. Purification and characterization of a cytotoxic exolipid of Burkholderia pseudomallei. Infect. Immun. 66, 1588
1593. Hommel, R.K., Weber, L., Weiss, A., Himmelreich, U., Rilke, O., Kleber, H.-P., 1994. Production of sophorose lipid by Candida (Torulopsis) apicola grown on glucose. J. Biotechnol. 33, 147
155. Available from: https://doi.org/https://doi.org/10.1016/0168-1656 (94)90107-4. Ho¨rmann, B., Mu¨ller, M.M., Syldatk, C., Hausmann, R., 2010. Rhamnolipid production by Burkholderia plantarii DSM 9509T. Eur. J. Lipid Sci. Technol. 112, 674
680. Available from: https://doi.org/10.1002/ejlt.201000030. Hu, Y., Ju, L.K., 2001. Purification of lactonic sophorolipids by crystallization. J. Biotechnol. 87, 263
272. Available from: https://doi.org/10.1016/S0168-1656(01)00248-6. Imura, T., Yanagishita, H., Kitamoto, D., 2004. Coacervate formation from natural glycolipid: one acetyl group on the headgroup triggers coacervate-to-vesicle transition. J. Am. Chem. Soc. 126, 10804
10805. Available from: https://doi.org/10.1021/ja0400281. Imura, T., Hikosaka, Y., Worakitkanchanakul, W., Sakai, H., Abe, M., Konishi, M., et al., 2007. Aqueous-phase behavior of natural glycolipid biosurfactant mannosylerythritol lipid A: sponge, cubic, and lamellar phases. Langmuir 23, 1659
1663. Available from: https://doi. org/10.1021/la0620814. Ine`s, M., Dhouha, G., 2015. Lipopeptide surfactants: production, recovery and pore forming capacity. Peptides 71, 100
112. Available from: https://doi.org/10.1016/j.peptides.2015.07.006. Isa, M.H.M., Coraglia, D.E., Frazier, R.A., Jauregi, P., 2007. Recovery and purification of surfactin from fermentation broth by a two-step ultrafiltration process. J. Memb. Sci. 296, 51
57. Available from: https://doi.org/10.1016/j.memsci.2007.03.023.

Membrane Filtration of Biosurfactants Chapter | 3

107

Isa, M.H.M., Frazier, R.A., Jauregi, P., 2008. A further study of the recovery and purification of surfactin from fermentation broth by membrane filtration. Sep. Purif. Technol. 64, 176
182. Available from: https://doi.org/10.1016/j.seppur.2008.09.008. Jajor, P., Piłakowska-Pietras, D., Krasowska, A., Łukaszewicz, M., 2016. Surfactin analogues produced by Bacillus subtilis strains grown on rapeseed cake. J. Mol. Struct. 1126, 141
146. Available from: https://doi.org/https://doi.org/10.1016/j.molstruc.2016.02.014. Jauregi, P., Coutte, F., Catiau, L., Lecouturier, D., Jacques, P., 2013. Micelle size characterization of lipopeptides produced by B. subtilis and their recovery by the two-step ultrafiltration process. Sep. Purif. Technol. 104, 175
182. Available from: https://doi.org/10.1016/j. seppur.2012.11.017. Jemil, N., Ayed, H. Ben, Manresa, A., Nasri, M., Hmidet, N., 2017. Antioxidant properties, antimicrobial and anti-adhesive activities of DCS1 lipopeptides from Bacillus methylotrophicus 1
11. https://doi.org/10.1186/s12866-017-1050-2 Juang, R.S., Chen, H.L., Chen, Y.S., 2008. Membrane fouling and resistance analysis in deadend ultrafiltration of Bacillus subtilis fermentation broths. Sep. Purif. Technol. 63, 531
538. Available from: https://doi.org/10.1016/j.seppur.2008.06.011. Kim, H.S., Jeon, J.W., Kim, B.H., Ahn, C.Y., Oh, H.M., Yoon, B.D., 2006. Extracellular production of a glycolipid biosurfactant, mannosylerythritol lipid, by Candida sp. SY16 using fedbatch fermentation. Appl. Microbiol. Biotechnol. 70, 391
396. Available from: https://doi. org/10.1007/s00253-005-0092-9. Kiran, G.S., Thomas, T.A., Selvin, J., Sabarathnam, B., Lipton, A.P., 2010. Optimization and characterization of a new lipopeptide biosurfactant produced by marine Brevibacterium aureum MSA13 in solid state culture. Bioresour. Technol. 101, 2389
2396. Available from: https://doi.org/https://doi.org/10.1016/j.biortech.2009.11.023. Kitamoto, D., Haneishi, K., Nakahara, T., Tabuchi, T., 1990. Production of mannosylerythritol lipids by Candida antarctica from vegetable oils. Agric. Biol. Chem. 54, 37
40. Available from: https://doi.org/10.1080/00021369.1990.10869919. Kitamoto, D., Morita, T., Fukuoka, T., Konishi, M.A., Imura, T., 2009. Self-assembling properties of glycolipid biosurfactants and their potential applications. Curr. Opin. Colloid Interface Sci. 14, 315
328. Available from: https://doi.org/10.1016/j.cocis.2009.05.009. Konishi, M., Fukuoka, T., Morita, T., Imura, T., Kitamoto, D., 2008. Production of new types of sophorolipids by Candida batistae. J. Oleo Sci. 57, 359
369. Available from: https://doi. org/10.5650/jos.57.359. Konishi, M., Nagahama, T., Fukuoka, T., Morita, T., Imura, T., Kitamoto, D., et al., 2011. Yeast extract stimulates production of glycolipid biosurfactants, mannosylerythritol lipids, by Pseudozyma hubeiensis SY62. J. Biosci. Bioeng. 111, 702
705. Available from: https://doi. org/10.1016/j.jbiosc.2011.02.004. Konishi, M., Fujita, M., Ishibane, Y., Shimizu, Y., Tsukiyama, Y., Ishida, M., 2016. Isolation of yeast candidates for efficient sophorolipids production: their production potentials associate to their lineage. Biosci. Biotechnol. Biochem. 80, 2058
2064. Available from: https://doi. org/10.1080/09168451.2016.1191332. Kosaric, N., Vardar-Sukan, F., 2015. Biosurfactants production and utilization—processes, technologies, and economics. In: Kosaric, N., Vardar-Sukan, F. (Eds.), Surfactant Science Series Volume 159. CRC Press, Taylor & Francis Group, Boca Raton, FL, p. 389. Kourmentza, C., Freitas, F., Alves, V., Reis, M.A.M., 2017. Microbial conversion of waste and surplus materials into high-value added products: the case of biosurfactants. In: Kalia, V.C., Kumar, P. (Eds.), Microbial Applications Vol.1: Bioremediation and Bioenergy. Springer International Publishing, pp. 29
77. Available from: https://doi.org/10.1007/978-3-319-52666-9_2.

108

Separation of Functional Molecules in Food by Membrane Technology

Kourmentza, C., Costa, J., Azevedo, Z., Servin, C., Grandfils, C., De Freitas, V., et al., 2018. Burkholderia thailandensis as a microbial cell factory for the bioconversion of used cooking oil to polyhydroxyalkanoates and rhamnolipids. Bioresour. Technol. 247, 829
837. Available from: https://doi.org/10.1016/j.biortech.2017.09.138. Kuyukina, M.S., Ivshina, I.B., Philp, J.C., Christofi, N., Dunbar, S.A., Ritchkova, M.I., 2001. Recovery of Rhodococcus biosurfactants using methyl tertiary-butyl ether extraction. J. Microbiol. Methods 46, 149
156. Available from: https://doi.org/10.1016/S0167-7012(01)00259-7. Leighton, A., 2013. Method for Treating Obesity. US20130310330 A1. Lencki, R.W., Williams, S., 1995. Effect of nonaqueous solvents on the flux behavior of ultrafiltration membranes. J. Memb. Sci. 101, 43
51. Available from: https://doi.org/10.1016/ 0376-7388(94)00282-4. Li, W., Rokni-Zadeh, H., De Vleeschouwer, M., Ghequire, M.G.K., Sinnaeve, D., Xie, G.L., et al., 2013. The antimicrobial compound xantholysin defines a new group of pseudomonas cyclic lipopeptides. PLoS One 8. Available from: https://doi.org/10.1371/journal. pone.0062946. Lin, S., Jiang, H., 1997. Recovery and purification of the lipopeptide biosurfactant of Bacillus subtilis by ultrafiltration. Biotechnol. Tech. 11, 413
416. Available from: https://doi.org/ 10.1023/A:1018468723132. Liu, T., Montastruc, L., Gancel, F., Zhao, L., Nikov, I., 2007. Integrated process for production of surfactin: Part 1: adsorption rate of pure surfactin onto activated carbon. Biochem. Eng. J. 35, 333
340. Available from: https://doi.org/https://doi.org/10.1016/j.bej.2007.01.025. Long, X., Meng, Q., Sha, R., Huang, Q., Zhang, G., 2012. Two-step ultrafiltration of rhamnolipids using PSU-g-PEG membrane. J. Memb. Sci. 409
410, 105
112. Available from: https://doi.org/10.1016/j.memsci.2012.03.039. Ma, Z., Hu, J., 2014. Production and characterization of iturinic lipopeptides as antifungal agents and biosurfactants produced by a marine Pinctada martensii-derived Bacillus mojavensis B0621A. Appl. Biochem. Biotechnol. 173, 705
715. Available from: https://doi.org/ 10.1007/s12010-014-0879-1. Majewska-Nowak, K., Kowalska, I., Kabsch-Korbutowicz, M., 2005. Ultrafiltration of SDS solutions using polymeric membranes. Desalination 184, 415
422. Available from: https://doi. org/10.1016/j.desal.2005.02.057. Makkar, R.S., Cameotra, S.S., Banat, I.M., 2011. Advances in utilization of renewable substrates for biosurfactant production. AMB Express 1, 5. Available from: https://doi.org/10.1186/ 2191-0855-1-5. Marchant, R., Banat, I.M., 2012. Microbial biosurfactants: challenges and opportunities for future exploitation. Trends Biotechnol. 30, 558
565. Available from: https://doi.org/ 10.1016/j.tibtech.2012.07.003. Markets and Markets, 2017. Biosurfactants Market by Type (Glycolipids (Sophorolipids, Rhamnolipids), Lipopeptides, Phospholipids, Polymeric Biosurfactants), Application (Detergents, Personal Care, Agricultural Chemicals, Food Processing), and Region - Global Forecast to 2022. Matsuyama, T., Kaneda, K., Nakagawa, Y., Isa, K., Hara-Hotta, H., Yano, I., 1992. A novel extra cellular cyclic lipopeptide which promotes flagellum-dependent and -independent spreading growth of Serratia marcescens. J Bacteriol 174, 1769
1776. Mejri, S., Siah, A., Coutte, F., Magnin-Robert, M., Randoux, B., Tisserant, B., et al., 2017. Biocontrol of the wheat pathogen Zymoseptoria tritici using cyclic lipopeptides from Bacillus subtilis. Environ. Sci. Pollut. Res. 1
12. Available from: https://doi.org/10.1007/ s11356-017-9241-9.

Membrane Filtration of Biosurfactants Chapter | 3

109

Mihalache, G., Balaes, T., Gostin, I., Stefan, M., Coutte, F., Krier, F., 2017. Lipopeptides produced by Bacillus subtilis as new biocontrol products against fusariosis in ornamental plants. Environ. Sci. Pollut. Res. 1
10. Available from: https://doi.org/10.1007/s11356-017-9162-7. Morita, T., Konishi, M., Fukuoka, T., Imura, T., Kitamoto, D., 2008. Production of glycolipid biosurfactants, mannosylerythritol lipids, by Pseudozyma siamensis CBS 9960 and their interfacial properties. J. Biosci. Bioeng. 105, 493
502. Available from: https://doi.org/ 10.1263/jbb.105.493. Morita, T., Fukuoka, T., Konishi, M., Imura, T., Yamamoto, S., Kitagawa, M., et al., 2009a. Production of a novel glycolipid biosurfactant, mannosylmannitol lipid, by Pseudozyma parantarctica and its interfacial properties. Appl. Microbiol. Biotechnol. 83, 1017
1025. Available from: https://doi.org/10.1007/s00253-009-1945-4. Morita, T., Ishibashi, Y., Fukuoka, T., Imura, T., Sakai, H., Abe, M., et al., 2009. Production of glycolipid biosurfactants, mannosylerythritol lipids, by a smut fungus, Ustilago scitaminea NBRC 32730. Biosci Biotechnol Biochem 73, 788
792. Available from: https://doi.org/ 10.1271/bbb.80901. Morita, T., Ishibashi, Y., Fukuoka, T., Imura, T., Sakai, H., Abe, M., et al., 2009b. Production of glycolipid biosurfactants, mannosylerythritol lipids, using sucrose by fungal and yeast strains and their interfacial properties. Biosci. Biotechnol. Biochem. 73, 2352
2355. Available from: https://doi.org/10.1271/bbb.90439. Morita, T., Kawamura, D., Morita, N., Fukuoka, T., Imura, T., Sakai, H., et al., 2013. Characterization of mannosylerythritol lipids containing hexadecatetraenoic acid produced from cuttlefish oil by Pseudozyma churashimaensis OK96. J. Oleo Sci. 62, 319
327. Available from: https://doi.org/10.5650/jos.62.319. Morita, T., Konishi, M., Fukuoka, T., Imura, T., Kitamoto, D., 2006. Discovery of Pseudozyma rugulosa NBRC 10877 as a novel producer of the glycolipid biosurfactants, mannosylerythritol lipids, based on rDNA sequence. Appl. Microbiol. Biotechnol. 73, 305
313. Available from: https://doi.org/10.1007/s00253-006-0466-7. Mukherjee, S., Das, P., Sen, R., 2006. Towards commercial production of microbial surfactants. Trends Biotechnol. 24, 509
515. Available from: https://doi.org/10.1016/j. tibtech.2006.09.005. Mulder, M., 1997. Basic Principles of Membrane Technology By Marcel Mulder (University of Twente, The Netherlands). Kluwer Academic: Dordrecht. 1996. 564 pp. $255.00. ISBN 07823-4247-X. J. Am. Chem. Soc. 119, 8582. https://doi.org/10.1021/ja975504k Mulligan, C.N., Gibbs, B.F., 1990. Recovery of biosurfactants by ultrafiltration. J. Chem. Technol. Biotechnol. 47, 23
29. Available from: https://doi.org/10.1002/jctb.280470104. Mutalik, S.R., Vaidya, B.K., Joshi, R.M., Desai, K.M., Nene, S.N., 2008. Use of response surface optimization for the production of biosurfactant from Rhodococcus spp. MTCC 2574. Bioresour. Technol. 99, 7875
7880. Available from: https://doi.org/10.1016/j.biortech.2008.02.027. Najmi, Z., Ebrahimipour, G., Franzetti, A., Banat, I.M., 2018. In situ downstream strategies for cost-effective bio/surfactant recovery. Biotechnol. Appl. Biochem. Available from: https:// doi.org/10.1002/bab.1641. Navon-Venezia, S., Zosim, Z., Gottlieb, A., Legmann, R., Carmeli, S., Ron, E.Z., et al., 1995. Alasan, a new bioemulsifier from Acinetobacter radioresistens. Appl. Environ. Microbiol. 61, 3240
3244. Ndlovu, T., Rautenbach, M., Khan, S., Khan, W., 2017. Variants of lipopeptides and glycolipids produced by Bacillus amyloliquefaciens and Pseudomonas aeruginosa cultured in different carbon substrates. AMB Express 7. Available from: https://doi.org/10.1186/s13568-0170367-4.

110

Separation of Functional Molecules in Food by Membrane Technology

Nguyen, T.T., Sabatini, D.A., 2011. Characterization and emulsification properties of rhamnolipid and sophorolipid biosurfactants and their applications. Int. J. Mol. Sci. 12, 1232
1244. Available from: https://doi.org/10.3390/ijms12021232. Nielsen, T.H., Sørensen, J., Nielsen, T.H., Sørensen, J., 2003. Production of Cyclic Lipopeptides by Pseudomonas fluorescens Strains in Bulk Soil and in the Sugar Beet Rhizosphere Production of Cyclic Lipopeptides by Pseudomonas fluorescens Strains in Bulk Soil and in the Sugar Beet Rhizosphere 69, 861
868. https://doi.org/10.1128/AEM.69.2.861 Niu, Y., Fan, L., Gu, D., Wu, J., Chen, Q., 2017. Characterization, enhancement and modelling of mannosylerythritol lipid production by fungal endophyte Ceriporia lacerate CHZJU. Food Chem. 228, 610
617. Available from: https://doi.org/10.1016/j.foodchem.2017.02.042. Oliveira, F.J.S., Vazquez, L., de Campos, N.P., de Franc¸a, F.P., 2009. Production of rhamnolipids by a Pseudomonas alcaligenes strain. Process Biochem. 44, 383
389. Available from: https://doi.org/10.1016/j.procbio.2008.11.014. Park, O.-J., Lee, Y.-E., Cho, J.-H., Shin, H.-J., Yoon, B.-D., Yang, J.-W., 1998. Purification and structural characterization of glycolipid biosurfactants from Pseudomonas aeruginosa YPJ80. Biotechnol. Bioprocess Eng. 3, 61
66. Available from: https://doi.org/10.1007/ BF02932503. Randhawa, K., 2015. Biosurfactants produced by genetically manipulated microorganisms: challenges and opportunities. In: Kosaric, N., Vardar-Sukan, F. (Eds.), BIOSURFACTANTS Production and Utilization—Processes, Technologies, and Economics. CRC Press, Taylor & Francis Group, pp. 49
72. Rau, U., Nguyen, L.A., Roeper, H., Koch, H., Lang, S., 2005a. Fed-batch bioreactor production of mannosylerythritol lipids secreted by Pseudozyma aphidis. Appl. Microbiol. Biotechnol. 68, 607
613. Available from: https://doi.org/10.1007/s00253-005-1906-5. Rau, U., La, A.N., Roeper, H., Koch, H., Lang, S., 2005b. Downstream processing of mannosylerythritol lipids produced by Pseudozyma aphidis. Eur. J. Lipid Sci. Technol. 107, 373
380. Available from: https://doi.org/10.1002/ejlt.200401122. Ron, E., Rosenberg, E., 2010. Role of biosurfactants. In: Timmis, K.N. (Ed.), Handbook of Hydrocarbon and Lipid Microbiology. Springer-Verlag Berlin Heidelberg, pp. 2515
2520. Available from: https://doi.org/10.1007/978-3-540-77587-4. Ron, E.Z., Rosenberg, E., 2001. Natural roles of biosurfactants. Environ. Microbiol. 3, 229
236. Available from: https://doi.org/10.1046/j.1462-2920.2001.00190.x. Rooney, A.P., Price, N.P.J., Ray, K.J., Kuo, T.M., 2009. Isolation and characterization of rhamnolipid-producing bacterial strains from a biodiesel facility. FEMS Microbiol. Lett. 295, 82
87. Available from: https://doi.org/10.1111/j.1574-6968.2009.01581.x. Rosenberg, E., Zuckerberg, A., Rubinovitz, C., Gutnick, D.L., 1979. Emulsifier of arthrobacter RAG-1: isolation and emulsifying properties. Appl. Environ. Microbiol. 37, 402
408. Sachdev, D.P., Cameotra, S.S., 2013. Biosurfactants in agriculture. Appl. Microbiol. Biotechnol. 97, 1005
1016. Available from: https://doi.org/10.1007/s00253-012-4641-8. Sadasivan, L., 2015. Aqueous coatings and paints incorporating one or more antimicrobial biosurfactants and methods for using same. US20130296461 A1. ¨ zdamar, T.H., 2006. Adsorption of bovine serum albumin on polyether Salgin, S., Takac¸, S., O sulfone ultrafiltration membranes: determination of interfacial interaction energy and effective diffusion coefficient. J. Memb. Sci. 278, 251
260. Available from: https://doi.org/ 10.1016/j.memsci.2005.11.008. Santos, D.K.F., Rufino, R.D., Luna, J.M., Santos, V.A., Sarubbo, L.A., 2016. Biosurfactants: multifunctional biomolecules of the 21st century. Int. J. Mol. Sci. 17, 1
31. Available from: https://doi.org/10.3390/ijms17030401.

Membrane Filtration of Biosurfactants Chapter | 3

111

Satpute, S.K., Banpurkar, A.G., Dhakephalkar, P.K., Banat, I.M., Chopade, B.A., 2010. Methods for investigating biosurfactants and bioemulsifiers: a review. Crit. Rev. Biotechnol. 30, 127
144. Available from: https://doi.org/10.3109/07388550903427280. Schelges, H., Tretyakova, M., 2016. Exfoliant with biosurfactants. 20170071846. Sen, R., Swaminathan, T., 2005. Characterization of concentration and purification parameters and operating conditions for the small-scale recovery of surfactin. Process Biochem. 40, 2953
2958. Available from: https://doi.org/10.1016/j.procbio.2005.01.014. Shah, M.U.H., Sivapragasam, M., Moniruzzaman, M., Yusup, S.B., 2016. A comparison of recovery methods of rhamnolipids produced by pseudomonas aeruginosa. Procedia Eng. 148, 494
500. Available from: https://doi.org/10.1016/j.proeng.2016.06.538. Silva, R., de, C.F.S., Almeida, D.G., Rufino, R.D., Luna, J.M., Santos, V.A., et al., 2014. Applications of biosurfactants in the petroleum industry and the remediation of oil spills. Int. J. Mol. Sci. 15, 12523
12542. Available from: https://doi.org/10.3390/ ijms150712523. Smyth, T.J.P., Perfume, A., McClean, S., Marchant, R., Banat, I.M., 2010a. Isolation and analysis of lipopeptides and high molecular weight biosurfactants. In: Timmis, K.N. (Ed.), Handbook of Hydrocarbon and Lipid Microbiology. Springer-Verlag Berlin Heidelberg, pp. 3687
3704. Available from: https://doi.org/10.1007/978-3-540-77587-4. Smyth, T.J.P., Perfumo, A., Marchant, R., Banat, I.M., 2010b. Isolation and analysis of low molecular weight microbial glycolipids. In: Timmis, K.N. (Ed.), Handbook of Hydrocarbon and Lipid Microbiology. Springer-Verlag Berlin Heidelberg, pp. 3705
3723. Available from: https://doi.org/10.1007/978-3-540-77587-4_291. Sobrinho, H.B., Luna, J.M., Rufino, R.D., Porto, A.L.F., Sarubbo, L.A., 2013. Biosurfactants: classification, properties and environmental applications. Recent Dev. Biotechnol. 11, 1
29. Available from: https://doi.org/10.1080/10643389.2014.955631. Solaiman, D.K.Y., Ashby, R.D., Nun˜ez, A., Foglia, T.A., 2004. Production of sophorolipids by Candida bombicola grown on soy molasses as substrate. Biotechnol. Lett. 26, 1241
1245. Available from: https://doi.org/10.1023/B:BILE.0000036605.80577.30. Soliance France, 2018. Sopholiance S: anti-bacterial & Sebore´gulateur [WWW Document]. http://www.soliance.com/dtproduit.php?id 5 42 (accessed 1.22.18). Spoeckner, S., Wray, V., Nimtz, M., Lang, S., 1999. Glycolipids of the smut fungus Ustilago maydis from cultivation on renewable resources. Appl. Microbiol. Biotechnol. 51, 33
39. Available from: https://doi.org/10.1007/s002530051359. Suzuki, M., Kitagawa, M., Yamamoto, S., Sogabe, A., Kitamoto, D., Morita, T., et al., 2012. Activator including biosurfactant as active ingredient, mannosyl erythritol lipid, and production method thereof. US20120070396 A1. Toyobo Co. Ltd., 2018. Ceramela: fermented olive-aided natural moisturizer superior to ceramide [WWW Document]. http://www.toyobo-global.com/seihin/cosme/ceramela.htm Transparency Market Research, 2014. Microbial Biosurfactants Market (Rhamnolipids, Sophorolipids, Mannosylerythritol Lipids (MEL) and Other) for Household Detergents, Industrial & Institutional Cleaners, Personal Care, Oilfield Chemicals, Agricultural Chemicals, Food Processing, Textile and, Transparency Market Research. Urba´nski, R., Go´ralska, E., Bart, H.J., Szymanowski, J., 2002. Ultrafiltration of surfactant solutions. J. Colloid Interface Sci. 253, 419
426. Available from: https://doi.org/10.1006/ jcis.2002.8539. Uzoigwe, C., Burgess, J.G., Ennis, C.J., Rahman, P.K.S.M., 2015. Bioemulsifiers are not biosurfactants and require different screening approaches. Front. Microbiol. 6, 1
6. Available from: https://doi.org/10.3389/fmicb.2015.00245.

112

Separation of Functional Molecules in Food by Membrane Technology

Vallet-Gely, I., Novikov, A., Augusto, L., Liehl, P., Bolbach, G., Pe´chy-Tarr, M., et al., 2010. Association of hemolytic activity of pseudomonas entomophila, a versatile soil bacterium, with cyclic lipopeptide production. Appl. Environ. Microbiol. 76, 910
921. Available from: https://doi.org/10.1128/AEM.02112-09. van Reis, R., Zydney, A., 2007. Bioprocess membrane technology. J. Memb. Sci. 297, 16
50. Available from: https://doi.org/10.1016/j.memsci.2007.02.045. Vater, J., Herfort, S., Doellinger, J., Weydmann, M., Dietel, K., Faetke, S., et al., 2017. Fusaricidins from Paenibacillus polymyxa M-1, a family of lipohexapeptides of unusual complexity—a mass spectrometric study. J. Mass Spectrom. 52, 7
15. Available from: https://doi.org/10.1002/jms.3891. Weber, A., Zeiner, T., 2015. Purification of biosurfactants. In: Kosaric, N., Vardar-Sukan, F. (Eds.), Biosurfactants Production and Utilization—Processes, Technologies, and Economics. CRC Press, Taylor & Francis Group, pp. 129
152. Willenbacher, J., Yeremchuk, W., Mohr, T., Syldatk, C., Hausmann, R., 2015. Enhancement of Surfactin yield by improving the medium composition and fermentation process. AMB Express 5, 145. Available from: https://doi.org/10.1186/s13568-015-0145-0. Witek-Krowiak, A., Witek, J., Gruszczy´nska, A., Szafran, R.G., Ko´zlecki, T., Modelski, S., 2011. Ultrafiltrative separation of rhamnolipid from culture medium. World J. Microbiol. Biotechnol. 27, 1961
1964. Available from: https://doi.org/10.1007/s11274-011-0655-0. Wittgens, A., Tiso, T., Arndt, T.T., Wenk, P., Hemmerich, J., Mu¨ller, C., et al., 2011. Growth independent rhamnolipid production from glucose using the non-pathogenic Pseudomonas putida KT2440. Microb. Cell Fact. 10, 80. Available from: https://doi.org/10.1186/14752859-10-80. Yamamoto, S., Fukuoka, T., Imura, T., Morita, T., Yanagidani, S., Kitamoto, D., et al., 2013. Production of a novel mannosylerythritol lipid containing a hydroxy fatty acid from castor oil by Pseudozyma tsukubaensis. J. Oleo Sci. 62, 381
389. Available from: https://doi.org/ 10.5650/jos.62.381. Yang, X., Zhu, L., Xue, C., Chen, Y., Qu, L., Lu, W., 2012. Recovery of purified lactonic sophorolipids by spontaneous crystallization during the fermentation of sugarcane molasses with Candida albicans O-13-1. Enzyme Microb. Technol. 51, 348
353. Available from: https://doi.org/10.1016/j.enzmictec.2012.08.002. You, J., Yang, S.-Z., Mu, B.-Z., 2015. Structural characterization of lipopeptides from Enterobacter sp. strain N18 reveals production of surfactin homologues. Eur. J. Lipid Sci. Technol. 117, 890
898. Available from: https://doi.org/10.1002/ejlt.201400386.

FURTHER READING Sivapathasekaran, C., Mukherjee, S., Sen, R., Bhattacharya, B., Samanta, R., 2011. Single step concomitant concentration, purification and characterization of two families of lipopeptides of marine origin. Bioprocess Biosyst. Eng. 34, 339
346. Available from: https://doi.org/ 10.1007/s00449-010-0476-9.

Chapter 4

Membrane Technology for the Purification of Enzymatically Produced Oligosaccharides Andre´s Co´rdova1, Carolina Astudillo1 and Andre´s Illanes2 1

Escuela de Alimentos, Facultad de Agronomı´a y de los Alimentos, Pontificia Universidad Cato´lica de Valparaı´so, Valparaı´so, Chile, 2Escuela de Ingenierı´a Bioquı´mica, Facultad de Ingenierı´a, Pontificia Universidad Cato´lica de Valparaı´so, Valparaı´so, Chile

Chapter Outline 4.1 Introduction to Oligosaccharide Prebiotics 113 4.2 Synthesis of Oligosaccharides in Membrane Bioreactors 118 4.3 Strategies and Mechanisms Involved in the Nanofiltration of Oligosaccharides 122 4.3.1 Retention Due to Concentration Polarization 122 4.3.2 Effect of Temperature 124 4.3.3 Effect of Transmembrane Pressure 128 4.3.4 Effect of Solute Concentration and Feed Composition 131 4.4 Current Status of the Purification of Specific Oligosaccharides by Membrane Technology 133

4.4.1 Purification of Xylooligosaccharides 4.4.2 Purification of Isomaltooligosaccharides 4.4.3 Purification of Galactooligosaccharide and Other Lactose-Derived Prebiotics 4.4.4 Purification of Fructooligosaccharides 4.4.5 Purification of Other Oligosaccharides 4.5 Challenges and Perspectives 4.6 Conclusions References Further Reading

133 137

139 143 144 145 146 147 153

4.1 INTRODUCTION TO OLIGOSACCHARIDE PREBIOTICS Due to the alarming numbers of diseases associated with bad nutrition, people’s awareness about eating habits has been increased over the last decade. This situation has opened new business opportunities to the food industry for the Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00004-8 © 2019 Elsevier Inc. All rights reserved.

113

114

Separation of Functional Molecules in Food by Membrane Technology

development of functional foods. Nowadays, many products are launched to a market of increasing demand and consumer acceptance, mainly for functional foods and prebiotics. Prebiotics are nondigestible polysaccharides and oligosaccharides that stimulate the growth of health-promoting bacteria in the gastrointestinal tract (Gibson et al., 2004). They also exert antagonism to Salmonella sp. and Escherichia coli, limiting their proliferation (Patel and Goyal, 2011). Oligosaccharide prebiotics have being implicated in starter culture formulation, gut health maintenance, colitis prevention, cancer inhibition, immunostimulation, cholesterol removal, reduction of cardiovascular disease, prevention of obesity and constipation, as well as other uses in animal feed, for example, poultry, pigs, cattle, and companion animals (Rastall, 2010). Prebiotic oligosaccharides are short-chain carbohydrates that are not degraded by the enzymes of the human digestive tract. Therefore, these molecules will arrive intact to the lower gut to be selectively fermented by the probiotic bacteria. The essential end products of carbohydrate metabolism are shortchain fatty acids, which are used by the host organism as energy source (AlSheraji et al., 2013). Besides their prebiotic effect, oligosaccharides may also impart functional properties making them attractive for use in different food formulations. They are water soluble and mildly sweet, being useful as bulking agents. They also allow increasing viscosity, conferring body and improving mouthfeel to the food matrices, and provide them high moisture-retaining capacity, thus preventing excessive drying (Crittenden and Playne, 1996). In nature, oligosaccharides can be found in different sources such as chicory, onion, garlic, asparagus, artichoke, and different plants, as well as in some processed dairy products. Thus, they can be extracted from natural sources or else produced by chemical or biotechnological processes. In this regard, enzymes (specifically, carbohydrases) are the most used biocatalysts for the large-scale production of oligosaccharides. To such purpose, carbohydrases may act by catalyzing transglycosylation reactions. During transglycosylation, the enzyme transfers one glycosyl unit from an oligosaccharide, usually a disaccharide (the donor) to a hydroxyl containing moiety, usually another oligosaccharide (the acceptor), thus elongating the oligosaccharide chain. If the acceptor is water, the result is the breakdown of the glycosidic bond, leading to hydrolysis. For enzyme extraction of oligosaccharides from plant materials, hydrolysis is the main mechanism, which implies that the carbohydrases must degrade the lignocellulosic material for the oligosaccharides to be extracted. A deeper insight on how hydrolytic and transgylosylation activities of glycosyl hydrolases can be modulated has been recently reviewed (Abdul-Manas et al., 2018). According to Roberfroid (2008), the main prebiotic oligosaccharides can be categorized as shown in Table 4.1. Inuline and soybean oligosaccharides are exclusively produced by direct extraction from natural sources. Therefore, they will not be analyzed in this chapter, which is focused on enzyme-produced oligosaccharides.

TABLE 4.1 Main Categories of Prebiotic Oligosaccharides Name

Main Sources/Substrates

Production Pathway

References

Xylooligosaccharides (XOS)

Bamboo shoots, fruits, vegetables, crop residues, wheat bran, corn, xylose

Hydrothermal treatments or steam explosion at high temperatures

Moure et al. (2006)

Acid and/or enzymatic treatments of lignocellulosic material

Va´zquez et al. (2000)

Transxylosylation with β-xylosidase, usign xylobiose in presence of D-mannose

Kizawa et al. (1991)

Sequential or combined enzyme hydrolysis conducted by amylases, pullanases, and glucosidases

Kuriki et al. (1993)

Transglycosylation of maltose using novel α-glucosidases

Swati et al. (2015)

Asparagus, sugar beet, chicory, garlic, artichoke

Enzymatic hydrolysis

Won Yun (1996)

Sucrose

Transglycosylation of sucrose with fructofuranosidases

Sangeetha et al. (2005)

Lactosucrosea

Lactose and sucrose

Transglycosylation with β-fructofuronasidases

Silve´rio et al. (2015)

Galactooligosaccharides (GOS)

Milk whey permeate, lactose

Transgalactosylation with β-galactosidases

Torres et al. (2010); Vera et al. (2016)

Isomaltooligosaccharides (IMO)

Fructooligosaccharides (FOS)

Starch

(Continued )

TABLE 4.1 (Continued) Name b

Lactulose

Gentiooligosaccharidesc

Main Sources/Substrates

Production Pathway

References

Lactose and fructose

Alkaline isomerization of lactose by molecular rearrangement

Panesar and Kumari (2011)

Enzymatic synthesis from lactose and fructose with β-galactosidases

Guerrero et al. (2011)

Enzymatic synthesis from lactose with cellobiose 2-epimerase

Kim and Oh (2012)

Block condensation

Takiura et al. (1972)

Acid or enzyme hydrolysis of starch transglycosylation reactions mediated by β-glucosidases

Crittenden and Playne (1996)

Gentiobiose starch

Fujimoto et al. (2009)

Inuline

Jerusalem artichokes, garlic, chicory root, wheat

Extraction by using hot water diffusion processes

Boeckner et al. (2001)

Soybean oligosaccharides (SOS)

Soybean seed/soybean whey

Direct extraction

Al Loman and Ju (2016)

a

Lactosucrose is not yet included as a prebiotic, but is considered a potential candidate. Current industrial production of lactulose is conducted by chemical isomerization. Not yet tested in humans.

b c

Membrane Technology for the Purification Chapter | 4

117

When oligosaccharides are extracted from plant materials, several kinds of molecules, interactions, and rheological behaviors are present, reflecting the complexity of the cell wall structures from where they were extracted. This occurs whether extraction is performed by chemical or enzymatic hydrolysis, or both. Nevertheless, the variability of compounds produced does not occur only in oligosaccharides manufactured from plant tissues. During the production of oligosaccharides by enzymatic transglycosylation, hydrolysis will occur simultaneously to a greater or lesser extent depending on the enzyme origin and the reaction conditions. Transglycosylation is a kinetically controlled reaction so that the synthesized oligosaccharide, and also the substrate, will be hydrolyzed, leading to a mixture of oligosaccharides and sugars. In addition, substrates conversions are usually low (below 50%), so that significant amounts of residual substrates will also be present, and oligosaccharides with different degrees of polymerization (DP) and glycosidic linkages will inevitably be produced (Meyer et al., 2015). Since the main application of the produced oligosaccharides is for human consumption, a high degree of product purity is required, and hence, product purification is of paramount importance in the production process (Pinelo et al., 2009). Product purification directly relates to the quality of the oligosaccharide in terms of its functional properties and application in different food matrices. For instance, monosaccharides, which are the usual hydrolytic products contaminating the oligosaccharides, are cariogenic and contribute an undesirable caloric content, their glucose content making the product inadequate for diabetic people. From an economic point of view, purification will be a critical processing step in defining the possible market niches for the product. Conventional strategies such as solvent extraction, solvent precipitation (Moure et al., 2006), adsorption on activated charcoal (Lee et al., 2004), and chromatographic process with ion-exchange resins (Herna´ndez et al., 2009) are frequently reported to be used for the purification of oligosaccharides. However, concerns in terms of food safety, environment protection, and feasibility for large-scale production have prompted the searching of alternative operations. In this context, the purification of prebiotic oligosaccharides by membrane technology offer several advantages: their scale-up is straightforward, it is an energy efficient technology, and the operational parameters can be relatively easy optimized (Patil et al., 2014). However, several experimental trials must be conducted for establishing proper operational conditions. Since oligosaccharides with prebiotic capacity are mainly those having a DP higher than two (or three) saccharide units, it is expected that these compounds will remain in the retentate, while monosaccharides and other impurities of smaller molecular size will pass through the membrane to be collected in the permeate stream (Fig. 4.1). Over the past decade, efforts were aimed toward the use of membrane technology as a strategy for the purification of oligosaccharides. A

118

Separation of Functional Molecules in Food by Membrane Technology

FIGURE 4.1 Schematic representation of the oligosaccharides purification by membrane technology. The circles represent a unit of any monosaccharide (glucose, fructose, galactose, xylose, etc.) which may be in free form or linked to another monosaccharide as disaccharide or as well as forming oligosaccharide with linear or branched glycosidic bonds.

preliminary conclusion was that membrane technology had not reached yet a state of maturity for such purpose, and its industrial feasibility was doubtful because of the lack of information about the effect of operational variables on purification yield and productivity. Also, membrane separation was reputedly expensive and caused problems during operation due to fouling. Nevertheless, progresses in material sciences and membrane technology during the last few years have increased the availability of different support materials, configurations, and knowledge about process conditions for purification of targeted compounds. As a result, membrane technology is being considered an attractive option for fractionation, purification, and concentration of prebiotic oligosaccharides mixtures as well as in the processing of other beverages (Salehi, 2014). This chapter reports the main technological challenges, mechanisms, advances, trends, and perspectives regarding the use of membrane technology as a strategy for the purification of enzymatically produced oligosaccharides.

4.2 SYNTHESIS OF OLIGOSACCHARIDES IN MEMBRANE BIOREACTORS Membrane bioreactors (MBRs) are systems in which a chemical conversion promoted by a biocatalyst is implemented by a membrane operation. The bioconversion can occur in a vessel that is combined with a membrane module or take place at the membrane level itself. Therefore, the membrane can either compartmentalize the biocatalyst in a circuit of the membrane module (lumen or shell) or support the biocatalyst in its micro/nanostructured matrix (Mazzei et al., 2017). A schematic representation of the different

Membrane Technology for the Purification Chapter | 4

119

FIGURE 4.2 Schematic classification of membrane bioreactor on the basis of enzyme compartmentalization with respect to the membrane: A) membrane bioreactor with free biocatalyst and B) membrane bioreactor with immobilized biocatalyst. Mazzei, R., Piacentini, E., Gebreyohannes, A.Y., Giorno, L., 2017. Membrane bioreactors in food, pharmaceutical and biofuel applications: state of the art, progresses and perspectives, Curr. Org. Chem., reproduced by permission of Bentham Science.

configuration alternatives of MBRs is shown in Fig. 4.2. This topic has been covered in depth in a recently published review by Mazzei et al. (2017). Since MBRs simultaneously combine mass transfer with chemical reactions, they allow the selective removal of products from the reaction site, which results in an increase in conversion, especially in reactions inhibited by the product or those that are thermodynamically disfavored (Giorno et al., 2003). This is a very important issue during the synthesis of oligosaccharides, since it is a kinetically controlled reaction where glycosyl hydrolases are strongly inhibited by the hydrolysis products (Nishizawa et al., 2001; Vera et al., 2012). Also, MBRs allow performing the synthesis in continuous operations. It provides an excellent setup for increasing the availability of enzymes to the substrate with greater turnover during a considerable period of time, due to the continuous removal of product saccharides and their simultaneous replacement with fresh substrate in quasi-steady state, as long as the integrity of the bioreactor is maintained (Rehman et al., 2016). These facts are significant from a processing cost perspective. Synthesis of galactooligosaccharides (GOS) is the most studied application of MBR in the enzymatic production of oligosaccharides. Regarding the MBR operating with free biocatalyst, Czermak et al. (2004) were one of the first to introduce the continuous synthesis of GOS by using different commercial sources of β-galactosidases from Kluyveromyces lactis and

120

Separation of Functional Molecules in Food by Membrane Technology

Aspergillus oryzae, in a membrane-coupled reactor. They used both flat polymer and inorganic cylindrical membranes with nominal MWCO of 50 and 20 kDa, respectively. The results of this investigation showed that it is possible to retain and recycle the enzyme, while the byproducts and unreacted substrate are collected in the permeate. The maximum concentration of GOS was close to 28% w/w of the total carbohydrate mass, obtained using 24% w/w lactose as substrate, and β-galactosidase from K. lactis in a 19-channel ceramic membrane module with an effective area of 0.1 m2 (a stable flux in the range of 20 L/m2/h was achieved). Subsequently, Matella et al. (2006) reported a maximum GOS yield of 20% w/w using a MBR fed with 25% lactose (w/v) and β-galactosidase from A. oryzae dosed at a concentration of 4.5 g/L. Procedikova et al. (2010) used the commercial β-galactosidase Maxilact LX 5000, dosed at 6 IU/mL, using lactose substrate at 19.8% w/v in MBR, using ultrafiltration membranes of 50 and 150 kDa nominal MWCO. The results showed that the concentration of GOS obtained in the permeate was lower than that obtained in a conventional batch process, but the total amount of GOS obtained with the same biocatalyst mass was higher in MBR due to the continuous processing of a greater volume of substrate. Similarly, Co´rdova et al. (2016a) showed that production of GOS is feasible in an ultrafiltration MBR when using a high lactose concentration of 40% w/ w (470 g/L). In general, higher conversion and slightly higher GOS concentration were achieved than in a conventional batch operation, while stable flux values ranged from 20 to 60 kg/m2/h depending on the processing conditions. Long-term operation was limited by flux decay rather than inactivation of the A. oryzae β-galactosidase. However, this inconvenience can be overcome by increasing the ratio of membrane area to reaction volume, thus reducing the substrate residence time. Therefore, an increased membrane area necessarily demands increasing the enzyme/substrate ratio. If this consideration is taken into account, the time required to obtain the maximum reaction yield is reduced, a higher substrate mass can be processed, and it avoids carrying out the operation for long times. This intervention helps in reducing membrane fouling and the appearance of microbial contamination (biofouling). Synthesis of GOS has been also assessed by using MBR with immobilized biocatalyst on nanofiltration (NF) membranes. Immobilization of the biocatalyst enhances its stability against physicochemical stresses, increases productivity, and improves the quality of the final product (Galanakis et al., 2012). Using β-galactosidase of Bacillus circulans immobilized in polyvinylidene fluoride membranes, Palai and Bhattacharya (2013) demonstrated that this enzyme can be immobilized up to three times in the same membrane, performing the synthesis of GOS continuously during 30 hours each time. While this demonstrates the reusability of the biocatalyst immobilized on the membrane, membrane tests showed a fourfold increase in total resistance at the third reimmobilization, reducing the initial flux by 73.8%, and thus also

Membrane Technology for the Purification Chapter | 4

121

the percentage of immobilized enzyme on the membrane. Similar results were reported by Nath et al. (2013), who designed a membrane reactor in dead-end configuration, in which the β-galactosidase from B. circulans was immobilized on the surface of the UF membrane (flat sheet, polyethersulfone, MWCO 5 1 kDa), by adsorption with polyethylenimine. As expected, the increase in transmembrane pressure to a certain limit level increased flux and lactose conversion, but then decreased with the application of a higher pressure. Such effects were attributed to the fact that the use of a higher pressure generates a greater driving force on the surface of the membrane that facilitates the transfer of the inhibitory compounds of the reaction and its byproducts through the channels of the membrane. However, the same increase in the passage of such molecules through the membrane occurred with lactose, which resulted in a shorter contact time of the substrate with the enzyme and, therefore, in a lower conversion. With respect to the production of fructooligosaccharides (FOS) in MBRs, Nishizawa et al. (2001) reported it using a β-fructofuranosidase from A. niger immobilized by physical adsorption and chemical coupling, and sucrose as substrate. The enzyme was best retained when immobilized by chemical coupling at an immobilization percentage of 64% on tubular ceramic membrane with pore size of 0.2 μm. This high immobilization percentage allowed maximum volumetric productivities (around 3.87 kg/m3/s), which were 560-fold higher than in a conventional batch. This situation was mainly due to the short residence time of the substrate (11 seconds) in the MBR. In this regard, the use of lower substrate concentrations (30% w/w) than usually used (50% w/w) allowed a higher permeation rate, thus increasing the overall productivity. More recently, Rehman et al. (2016), reported the synthesis of (FOS) in MBR but using Pectinex Ultra SP-L (a pectinase commercial preparation) and beet molasses as substrate. Despite this commercial biocatalyst is mostly used for juice clarification because of their pectinolytic and cellulolytic activities, it also has a minor fructosyltransferase activity sufficient to produce FOS (Vega-Paulino and Zu´niga-Hansen, 2012). Furthermore, by using molasses solution at 50% (w/w) with a residence time of 1 hour, it was possible to produce 2 kg of FOS/m2 of membrane/h, continuously in a system composed by an ultrafiltration ceramic membrane with a MWCO of 20 kDa. The main drawback was the membrane fouling caused by the molasses, which was double than when using sucrose as substrate. However, this is not a limiting factor since a simple cleaning procedure was enough to overcome such problem. Even more, previous filtration of the raw substrate is advised for this kind of applications of MBR, to reduce flux decay due to fast fouling. The use of MBR in enzymatic reactions of hydrolysis has been tested to produce xylooligosaccharides (XOS), too. Simultaneous hydrolysis of xylan and its separation from lignocelullosic material was used by Freixo and De Pinho (2002) for the enzymatic hydrolysis of beechwood xylan. In this

122

Separation of Functional Molecules in Food by Membrane Technology

case, the enzyme was continuously recirculated through an ultrafiltration membrane in cross-flow configuration. During the operation, the substrate (B10 kDa) was progressively hydrolyzed and recovered in the permeate stream as a mixture of oligosaccharides, which presented two main molecular size ranges: one from 6 to 1.5 kDa and the other from 0.6 to 0.45 kDa. Despite the advantages of having simultaneous hydrolysis and fractionation, residence time was not enough for further hydrolysis of the larger polymers. The results presented above, and others not listed here, show that synthesis of prebiotic oligosaccharides in MBR can be an interesting alternative for their large-scale production. Some facts that apparently can explain the lack of development on MBR technology for oligosaccharides production at industrial scale are: G

G

G

The lack of knowledge of processing suitable conditions must be set in each specific case to match a high threshold flux (see Section 4.3.3) with a high biocatalyst stability. The progress made in biotechnology during the last few years that has reduced the costs of enzymes production, thus reducing their commercial value. However, for some food applications, biocatalysts are still considered expensive, and also unstable at operating conditions. Sealing problems as well as mechanical and thermal fragility of some membranes (Stankiewicz, 2003).

Therefore, the use of MBR for the synthesis of oligosaccharides opens several research opportunities, not only as a strategy for improving the performance of enzymes and their continuous reuse, but also as a strategy for process intensification (Satyawali et al., 2017). To the best of our knowledge, information related to investment and production costs, energy efficiency, and exergy analysis related to the use of MBRs, in comparison to the conventional strategies currently used in the large scale production of oligosaccharides, has not been reported yet.

4.3 STRATEGIES AND MECHANISMS INVOLVED IN THE NANOFILTRATION OF OLIGOSACCHARIDES 4.3.1 Retention Due to Concentration Polarization One of the main challenges during NF of oligosaccharides is the selective fractionation of solutions having compounds with small differences in molecular size. There are several membrane indicators used to characterize their performance in this regard. One of the most used and easy to determine experimentally is the apparent rejection coefficient (Rap): Rap 5 1 2

Cp Cb

ð4:1Þ

Membrane Technology for the Purification Chapter | 4

123

where Cp and Cb are the concentrations of the solute in the permeate and the bulk of the solutions, respectively. Also quite important is the concentration polarization phenomenon, which occurs in a boundary layer adjacent to the membrane surface. Assuming a steady state mass balance for a solute arriving to the membrane surface by convection, while back-diffusion of solutes occurs simultaneously from membrane surface to the bulk, the film theory model can be written as:   Cm 2 Cp Jv 5 exp ð4:2Þ Cb 2 Cp k where Cm is the concentration of a given solute on the membrane surface, Jv is the volumetric flux, and k is the mass transfer coefficient. As stated above, Cp and Cb are the concentrations in the permeate and in the bulk, respectively. Usually, Cm is quite difficult to determine experimentally quantified, hence it must be mathematically predicted by alternative expressions. The extended Nernst
Planck equation, which is the fundamental relationship governing the transport of ionic species through membrane pores (Bhattacharjee et al., 2001), can be adapted for electroneutrality conditions (such in the case of oligosaccharides) in the well-known Kedem
Speigler equation (Chaabane et al., 2007): J i 5 2 Pi

dCi 1 Jv Ci ð1 2 σÞ dx

ð4:3Þ

where Ji is the presence of the solute (i) in the permeate, Ci is their concentration, Pi is the permeability of the solute, and σ is the reflection coefficient of the solute. The two latter are constant characteristics for each solute (Chaabane et al., 2002), and particularly, σ is fitted mathematically. Thus, by integrating Eq. (4.3) over the thickness of the membrane (0 , x , Δx), and considering the boundary conditions Ci 5 Cm at x 5 0 and Ci 5 Cp at x 5 Δx, the following expression for Cm can be obtained: 2 3 C σ h i5U p ð4:4Þ C m 5 41 2 Jv Uð1 2 σÞ 1 2σ exp Pi

Therefore, with Eqs. (4.2) and (4.4), the apparent retention coefficient (Eq. 4.1) can be arranged as follows: Rap 5 1 2

11

σ 12σ

n exp


2 Jv  k

1

h  io σ 1 1 2 exp 2Jv 1 2 Pi k

ð4:5Þ

Thus, Eq. (4.5) shows how Rap is expected to increase as a higher volume accumulates in the permeate. The practical importance of such expression is to highlight how Rap is increased as filtration progress, especially with membranes of higher MWCO. Normally NF may have a wide distribution pore

124

Separation of Functional Molecules in Food by Membrane Technology

size; hence, some oligosaccharides molecules with DP $ 2 will not be retained by the membrane, decreasing the yield of recovery (Yr), which is defined as the ratio between the mass of solute i in the retentate at the end of batch filtration (mr,i), and the mass of that compound in the inlet feed solution (mf,i): mr;i Yr 5 ð4:6Þ mf;i A controlled development of the boundary layer thickness, that is, the appearence of concentration polarization (without fouling itself), may increase the retention of oligosaccharides along operation time, improving the membrane selectivity. The cost to be paid is a higher retention of monosaccharides and other unwanted compounds. However, the evidence suggests that such an increase in Rap of monosaccharides occurs to a much lesser extent. Eq. (4.5) has proved to be useful in the prediction of Rap at different transmembrane pressures during the NF of XOS (Hua et al., 2010). A detailed explanation and the theoretical basis of the expressions shown in this section can be found in the works reported by Dresner (1972), Bhattacharjee et al. (2001), Chaabane et al. (2002), and Chaabane et al. (2007).

4.3.2 Effect of Temperature As stated previously, oligosaccharides have different technological properties such as impairing viscosity in food formulations. Additionally, most oligosaccharides are produced by enzymatic synthesis that must be conducted at high substrate concentrations (see Section 4.3.4). As a consequence, the downstream processing of these kinds of reacted mixtures may lead to a high resistance to flow inside the membrane, hindering mass transfer. Indeed, it is well known that permeate flux is inversely proportional to viscosity, hence membrane process performance may be poor. An alternative to overcome this problem is diluting the oligosaccharide solutions, but this may imply the use of large volumes of water that should be subsequently removed. However, viscosity is decreased by increasing the temperature. Fig. 4.3 shows the dependence of viscosity on temperature for a highly concentrated solution of GOS (40 Brix). According to the Arrenhius equation, the positive slope of the natural logarithm of the viscosity (ln μ) versus the inverse of the absolute temperature (1/T) straight line indicates the reduction of viscosity with increasing temperature. Hence, the permeability of the solution through the membrane increases, which is reflected in higher flux values (Cordova et al., 2016b). It must be considered that an increase in temperature may improve filtration due to decreased viscosity, but not all the membrane materials withstand

Membrane Technology for the Purification Chapter | 4

125

1/T 0.0030 –5.60

ln µ

–5.80

0.0031

0.0031

0.0032

0.0032

y = 2635.5x - 14.183 R² = 0.9717

–6.00

–6.20

–6.40

FIGURE 4.3 Relationship between temperature and viscosity for galactooligosaccharide (GOS) solutions at 40 Brix (Co´rdova et al., 2016b).

a wide operation temperature range. Table 4.2 shows the main characteristics of several NF membranes used at industrial scale. Although some membranes can withstand high temperatures close to 100 C, generally, commercial NF membranes have a maximum operating temperature that does not exceed 60 C. Such characteristics depend fundamentally on the type of polymer used in its manufacture. Temperature not only affects viscosity, but also the selectivity of the membranes, due to changes in the effective pore size of polymeric membranes (Sharma et al., 2003; Pruksasri et al., 2015). A generalized correlation of the linear effect of the temperature (T) on pore size radius (rp) can be described as follows: rp 5 mUT 1 b

ð4:7Þ

where T is the absolute temperature, m is the slope and b the intercept, both parameters depending of the membrane material properties. Thus, an increase in temperature involves an increase in the effective pore size diameter. Mechanisms that may explain this situation are: 1. a reduction of the adsorbed water layer on the hydrophilic pore surface when temperature increases, resulting in a decrease of the surface tension and an increase of the effective pore size 2. a higher thermal energy of the solvent that reduces the energy barrier generated in the micro/nanopores due to the pore wall forces (Tsuru et al., 2000) In this regard, an aspect to be considered in the mass transfer phenomena of NF of oligosaccharides is the so-called hindrance factor (ki,d), in what is called the steric pore model (Bowen and Mukhtar, 1996). Depending on the

TABLE 4.2 Characteristics of Nanofiltration Membranes Used at Industrial Scale (Data According to Manufacturers) Membrane Type and Brand

Material

Configuration

Range pH

Max. Temperature ( C)

Max. Pressure (bar)

Rap (%)

Hydraulic Permeability/Max Water Flux

AFC99 (PCI Membranes Ltd, UK)

PA

Tubular

1.5
12

80

64

99% NaCl




AFC80 (PCI Membranes Ltd, UK)

PA

Tubular

1.5
10

70

60

80% NaCl




AFC40 (PCI Membranes Ltd, UK)

PA

Tubular

1.5
9.5

60

60

60% CaCl2




AFC30 (PCI Membranes Ltd, UK)

PA

Spiral

1.5
9.5

60

60

75% CaCl2




NF90 (FilmTec Membranes, USA)

PA

Spiral

2
11

45

41

92% NaCl

8.3 L/(h  m2  bar)

NF200
400 (FilmTec Membranes, USA)

PA

Spiral

3
10

45

41

35
50% CaCl2

15.9 m3/h

NF2540 (FilmTec Membranes, USA)

Polypirazine

Spiral

3
10

45

41

98% MgSO4

1.4 m3/h

ESNA-1-LF2 (Nitto Denko, Japan)

TFC-PA

Spiral

3
10

45

41.6

73
92% CaCl2

17 m3/h

ESNA-1-LF (Nitto Denko, Japan)

TFC-PA

Spiral

3
10

45

41.6

84
96% CaCl2

17 m3/h

ESNA-1-LF-LD (Nitto Denko, Japan)

TFC-PA

Spiral

3
10

45

41.6

86
95% CaCl2

17 m3/h

MPS-34 (Koch Membranes, USA)

TFC-PS

Spiral

0
14

70




200 Da




MPF-44 (Koch Membranes, USA)

PDMS







40




98% lactose

1.3 L/(h m2 bar)

MPF-55 (Koch Membranes, USA)

PDMS







40




700 Da

1 L/(h  m2  bar)

NP030 (Microdyn Nadir, Germany)

PES

Spiral

0
14

95




26% NaCl

3.9 L/(h  m2  bar)

NP010 (Microdyn Nadir, Germany)

PES

Spiral

0
14

95




15% NaCl

13 L/(h  m2  bar)

Desal-5-DK (Osmonics, USA)

PA







50




98% MgSO4

5.4 L/(h  m2  bar)

Desal-5-DL (Osmonics, USA)

PA







50




96% MgSO4

9 L/(h  m2  bar)

SS-01 (Solsep BV, Netherlands)










150




97% 1000 Da

10 L/(h  m2  bar)

NF (Alfalaval, Denmark)

PP-PE

Plate

1.5
11

50

55

98% MgSO4




PA, polyamide; PDMS, polydimethylsiloxane; PES, polyether sulfone; PP-PE, polypropylene
polyester; TFC-PA, thin film composite covered with polyamides.

128

Separation of Functional Molecules in Food by Membrane Technology

solute molecules, ki,d normally is a potential function of the parameter λ, which is defined as: rs λ5 ð4:8Þ rp where rs and rp are the radio of the solute and pore size, respectively. The above shows that by changing the temperature, the coefficient of partition of the solute will be affected. An interesting experimental assessment and a deeper explanation of this model have been recently published (Bandini and Morelli, 2017). In practical terms, this means that an increased filtration temperature will increase rp, and therefore, will increase the flux. This will allow reducing the length of the operation, but may also increase the permeation of larger compounds (such as oligosaccharides with DP $ 3) that must be retained by the membrane, reducing the effectiveness of the purification process. Then, choosing the temperature for filtration of oligosaccharides must be a compromise between a reduced viscosity, an improved flux, and a controlled effect on the pore size diameter. Additionally, oligosaccharide manufacturers must be aware that if temperature is highly increased (i.e., above 55 C) during extended filtration processes, the presence of reducing sugars as well as other compounds may lead to caramelization and/or Maillard reactions, yielding other unwanted byproducts (called neocontaminants). In this regard, it is known that some neocontaminants may alter not only the color of the solution, but also the safety of the product. For instance, production of neocontaminants, such as 5-hydroxymethylfurfural, must be properly controlled since their presence at high concentrations is cytotoxic, and irritating to the eyes, upper respiratory tract, skin, and mucous membranes (Capuano & Fogliano, 2011). This is an important aspect to keep in mind, since the most important applications in which oligosaccharides are used as prebiotics are the formulation of products destined for infant and baby foods (Arslanoglu et al., 2008) or elderly people (Surakka et al., 2009).

4.3.3 Effect of Transmembrane Pressure It is well known that when the transmembrane pressure (TMP; the difference between the applied and the osmotic pressure) is raised, the flux is raised linearly, until a certain value. This TMP value is the critical transmembrane pressure (TMPC) and it refers to a flux value called the critical flux (Jc). If TMP is increased over the TMPC value, the J versus TMP curve will start to deviate from linearity, but the flux will increase until reaching a maximum value, the limiting flux (JL), which occurs at limiting TMP (TMPL). The theory implies than an operation under critical condition for avoiding fouling, leads to long-term operation where fouling is undetectable, also called zero fouling (Howell, 1995). Other concepts, such as sustainable and threshold

Membrane Technology for the Purification Chapter | 4

129

(Field and Pearce, 2011) fluxes, have been introduced in recent years. In the beginning, there was no evident difference between threshold and sustainable flux concepts. The first one is the flux value where the fouling given for the rate of d(TMP)/dt had a sharp increase in operations under constant flux. The second one involves operations under constant flux, which are preferred in industry when the key parameter to keep constant is the system productivity. However, that definition of sustainable flux has evolved to an “economic type definition,” where the membrane system remains at a flux value for minimizing fouling and avoid frequent chemical cleaning. The sustainable flux value depends on plant designer and operator criteria and it will change if the costs of water, membranes, or energy change (Field and Pearce, 2011). Meanwhile, the threshold flux is the maximum flux value remaining steady and low where fouling takes place at almost constant rate. If an operation takes place at a flux value above the threshold flux, the rate of fouling increases rapidly. The behavior of J versus TMP curves showing critical fluxes was first observed in microfiltration (Field et al., 1995) and ultrafiltration, and later on in NF. In NF stages, this phenomenon has been observed during coffee extract concentration (Vincze and Vatai, 2004), paper effluent filtration (Ma¨ntta¨ri et al., 1997; Ma¨ntta¨ri and Nystro¨m, 2000), and whey demineralization (Pan et al., 2011). The critical flux theory was originally developed for colloids filtration, because these particles are the main ones responsible for fouling. Colloidal fouling in NF could be produced by filtration of streams containing bacteria, viruses, RNA, DNA, and proteins, as occurs in wastewater treatment or in the recovery of valuable compounds from agroindustrial wastes. Even though oligosaccharides are molecules, not colloids, there is little evidence about deviation from linearity of the J versus TMP curve in NF of oligosaccharides (Nabarlatz et al., 2007; Vegas et al., 2006), therefore that the critical flux theory can be applied to oligosaccharides purification by NF. In the particular case of GOS, the critical fluxes were computed according to the model proposed by Astudillo-Castro (2015) (Co´rdova et al., 2016c, 2017). However, the use of critical, threshold, and limiting flux theories is still lacking in NF for oligosaccharides purification. Future research in this field will contribute to develop a proper strategy for fouling control in the purification of oligosaccharides by NF. The concepts of critical, limiting, sustainable, and threshold fluxes are gaining attention as a strategy for fouling control because “the lack of knowledge and control of membrane fouling is an additional cost for the industry which should be minimized to permit successful competiveness of the membrane technology” (Stoller et al., 2013). If batch systems are used, the feed stream will change their properties during the operation as consequence of an increased concentration. In fact, hydrodynamic conditions may change quickly, especially when using high solute concentrations as in the case of oligosaccharide solutions. This situation leads to changes in critical,

130

Separation of Functional Molecules in Food by Membrane Technology

FIGURE 4.4 Displacement of the TMPc depending to the concentration factor (FC) in a solution of galactooligosaccharide (GOS) subjected to NF: starting from FC 5 1 3 (20 Brix) to FC 5 2 3 (40 Brix). Assay performed in batch filtration (Co´rdova et al., 2017).

threshold, or limiting values along the time of operation and represents a major operational difficulty because it is necessary to control and vary the TMP during the operation to avoid undesirable effects due to membrane fouling (Astudillo-Castro, 2015). For instance, Fig. 4.4 shows the displacement of the TMPc depending to the concentration factor (FC) during batch NF of GOS (Co´rdova et al., 2017). The displacement starts from a TMPc1 point for a diluted solution (FC 5 1), that is, 20◦Brix, to a TMPc2 point that corresponds to the same solution, but whose solute concentration has been doubled (FC 5 2), that is, solution concentrated at 40◦Brix. Therefore, TMPc must be decreased over time to make concentration polarization equivalent to that obtained in the initial stages of the filtration. Assuming a linear displacement from FC 5 1 to FC 5 2, a model can be obtained for the variation of TMPc throughout batch NF in such a way to keep the Rap constant and decrease the impact on membrane performance due to appearance of fouling. The implications of the above are important because it is the degree of concentration polarization which also contributes to regulate the retention coefficient of a membrane. For small pore size membranes, such as those used in NF, larger molecules than the pore size are expected to be highly rejected independent of the degree of polarization. Smaller molecules are allowed to pass almost freely at low degree of polarization, but they will be rejected as the degree of polarization increases (Cheryan, 1998). If the cross-flow velocity and the feed concentration are steady, the transmembrane pressure is the parameter that allows to change the degree of polarization. Botelho-Cunha et al. (2010) observed both behaviors described by Cheryan (1998), that is, the

Membrane Technology for the Purification Chapter | 4

131

rejection of small and large molecules during GOS fractionation by NF using membranes made of cellulose acetate at 25 bar and 40 C. They determined the effect of the effective transmembrane pressure (TMPe) on the Rap. For GOS-3 the rejection was near to 100% regardless the TMPe applied, but for smaller molecules such as GOS-2, lactose, glucose, and galactose, the rejection was increasing with TMPe. This effect was also observed during the fractionation of enzymatically produced GOS and model solutions containing disaccharides and monosaccharides (sucrose and glucose, respectively), using 5 different membranes: NP010 and NP030 from Microdyn Nadir, GE from Desalogics, and NFA and ATF from Parker (Co´rdova et al., 2016c). The effect of the TMPe on the selectivity of the membrane is valid as long as the filtration is carried out in a temperature range such that the pore size does not increase significantly, so that larger oligosaccharides are not able to cross the membrane.

4.3.4 Effect of Solute Concentration and Feed Composition One of the main known limitations of membrane technology is their poor separation performance on highly concentrated streams (Cheryan, 1998). As matter of fact, the first efforts related to the purification of oligosaccharides by NF were mostly conducted at low solute concentrations (Goulas et al., 2002, 2003; Li et al., 2004; Feng et al., 2009). Despite the fact that these works are quite interesting and useful, most of the sugar solutions at industrial scale are rather concentrated, with total dry solids concentrations in the range 30
60% w/w (Sjo¨man et al., 2007). This situation is also extensive to oligosaccharides, since their syntheses are kinetically controlled reactions, where the main strategy for increasing the transglycosylation over hydrolysis reaction rates is performing the synthesis at high substrate concentrations. Thus, in the presence of a suitable nucleophile other than water, the enzyme can transfer the glycosyl residues to an acceptor molecule forming glycosidic linkages (hence oligosaccharide yield is increased). Considering the facts already described, it seems that NF of oligosaccharides in highly concentrated solutions results in an almost impracticable challenge. However, research conducted during the last 10 years shows the opposite: if the simultaneous effects of temperature, cross-flow velocity, and effective transmembrane pressure are properly addressed, fractionation of concentrated oligosaccharides by NF is quite possible. For instance, Co´rdova et al. (2016c) showed that at 20 bar and 53.5 C, GOS were fractionated with steady fluxes (B28 kg/m2/h) and a good selectivity, even when highly concentrated solutions of 40 Brix (approximately 460 g/L) were treated. The main problem was the removal of lactose, because the raw GOS product obtained by enzymatic transgalactosylation that was fed still contained a major proportion of residual lactose (45.5%), 28% monosaccharides, and 25.5% trisaccharides (GOS), the ratio between one compound and another

132

Separation of Functional Molecules in Food by Membrane Technology

having a direct impact in fractionation performance. In this regard, Sjo¨man et al. (2007) showed that the separation of molecules of similar molecular size, such as xylose (150 Da) and glucose (180 Da), is possible, but the ratio of xylose to glucose affected both the total sugar flux and the water flux. At high glucose proportion (feed xylose to glucose mass ratio 1:9), practically all the xylose was permeated. This effect was increased at the higher total solid concentration tested (30% w/w) suggesting that glucose “pushed” xylose through the membrane. Later on, Zhao et al. (2012) showed that fractionation of multicomponent solutions containing monosaccharides, sucrose, raffinose, and stachyose can be properly handled by increasing flux but in a very limited range. In that work, retention of all saccharides was increased with flux, until reaching a saturation point in which the increase in flux produced a decrease in raffinose and stachyose retention, while retention of monosaccharides and sucrose was virtually unchanged. According to the film theory, as the flux increased, the solute concentration at the surface of the membrane increased. Thus, the solutes concentration in the permeate increased, leading to the decrease of the retentions of raffinose and stachyose. The same trend was observed by Sjo¨man et al. (2007) in the separation of xylose from glucose, concluding that at high fluxes, the water flow through the pores is strong and might diminish the size difference between one compound and another. The cost to be paid for using more concentrated oligosaccharide solutions is the possible impact on membrane compression and a higher osmotic pressure, since higher applied pressures must be used than in a more diluted solution. Hua et al. (2010) demonstrated that increasing solute concentration of XOS solutions from 4.2% to 22% (w/w), at a high pressure (B25 bar), selectivity towards the permeation of xylose and arabinose could be achieved, while retention of xylobiose (a disaccharide with prebiotic effect) remained over 90%. A higher solute concentration generates higher osmotic pressure, which leads to a lower effective TMP. Thus, molecules of lower size may pass through the membrane preferentially by diffusion rather than by a driving force effect caused by the pressure difference. It must be pointed out that when using highly concentrated oligosaccharide solutions, a critical issue is choosing the suitable level of transmembrane pressure that must be applied. This can be determined by experimental tests that are easy to implement, by conducting flux versus transmembrane curves at different cross-flow velocities and temperatures. In general, fractionation of concentrated oligosaccharide solutions performs better at lower effective transmembrane pressures levels, since in this case a lower, but sustainable flux is ensured for a longer time with an appropriate selectivity for the retention of the larger compounds (oligosaccharides with DP $ 3), while retention of compounds in higher concentrations such as mono and disaccharides is lower.

Membrane Technology for the Purification Chapter | 4

133

4.4 CURRENT STATUS OF THE PURIFICATION OF SPECIFIC OLIGOSACCHARIDES BY MEMBRANE TECHNOLOGY Regardless of the usefulness of the predictive equations for Rap shown above, the practical importance for a user of a NF system that needs to purify oligosaccharides will be, fundamentally, to establish a strategy allowing the selective removal of monosaccharides and the other unwanted compounds. Selective permeation of monosaccharides may be improved if the operation is conducted under specific conditions. In this sense, it must be kept in mind that for NF, there is no a single way to achieve this goal: it will vary from case to case according to the nature of the treated stream.

4.4.1 Purification of Xylooligosaccharides Agroindustrial and forest byproducts are plant tissues rich in hemicelluloses. The degradation and fractionation of such complex materials may result in xylose and differently substituted XOS (Kabel et al., 2002). XOS have shown to have great potential as agents for maintaining and improving a balanced intestinal microbiota for enhanced health and well-being, and can be incorporated into many food products (Aachary and Prapulla, 2011). Sources of agricultural byproducts, such as cotton stalks, corn cobs, wheat straw, almond shells, sunflower stalks, and others, have been tested for XOS production. During the manufacturing of XOS for food grade applications, the lignocellulosic material must be subjected to chemical methods using hot water or steam in media catalyzed with externally added mineral acids (a process called autohydrolysis). Autohydrolysis consists of the deacetylation of D-xylan at elevated temperatures in the presence of water to produce acetic acid, which promotes the hydrolysis of the hemicelluloses. Then, the obtained liquors are subjected to the specific action of xylanases, yielding mixtures of monosaccharides and nonsaccharide compounds that must be removed to increase the concentration and purity of XOS as much as possible. Typically, the usual purity of commercial XOS lies in the range 75%
95% (Moure et al., 2006). The whole set of soluble fragments from xylan hydrolysis has a wide broad range of polymerization degrees, but generally polymers from 2 to 20 units are considered XOS with prebiotic effects (Moure et al., 2006). Xylan is the polysaccharide accounting for 25%
35% of the dry biomass of woody tissues of dicots and lignified tissues of monocots, and comprises up to 50% weight in some grasses and cereal grain tissues. The structure of xylan depends on its source and fractionation processes are applied for its recovery (Samanta et al., 2015). Fig. 4.5 shows some chemical structure examples of xylan and other derived mono-, di-, and XOS. Xylan typically consists of a β-D-(1,4)-linked xylopyranosyl backbone that can be substituted on O2, O3,

134

Separation of Functional Molecules in Food by Membrane Technology O

HO HO

OH

OH

D-Xylose

CH2OH O HO OH

D-Arabinofuranose

OH OH O HO OH

OH

L-Arabinose

O

HO HO

OH

D-Arabinopyranose

OH

OCH3 OH HO O

COOH

OH O O

HO

HO

OH OH

O

O

HO

O

O OH

OH

n

O

HO O

OH HO O

O

O OH

O

O HO

O

O OH

Xylan

OH

OH

HO OH HO

O

HO

di-L-Arabinofuranosyl-xylotriose

OH

HO

OH HO O

OH O

Xylobiose

HO

O

HO HO HO

OH HO O

O

O HO

O OH OH

Xylotriose

FIGURE 4.5 Chemical structure of xylose, arabinose, xylan, and other derived xylooligosaccharides (XOS).

or O4 with different configurations of α-L-arabinofuranosyl, α-D-glucropyranosyl, β-D-xylapyranosyl, and other groups where the origin of the raw material strongly determines the amount, position and distribution of these sidechains (Ishii, 1997). Therefore, a proper design for the downstream processing of XOS by membrane technology must consider the origin of the plant material from which it was obtained, as well the conditions of their extraction.

Membrane Technology for the Purification Chapter | 4

135

The information regarding the purification of XOS by using membrane technology is rather scarce. Due to the complexity of the compounds obtained during the hydrolysis of the lignocellulosic material, the use of ultrafiltration as pretreatment may be considered as part of the downstream processing. Swennen et al. (2005) compared ethanol precipitation with ultrafiltration for the isolation of long-chain arabino-XOS (up to 8 kDa) obtained from wheat flour. They concluded that gradual ethanol precipitation yielded arabino-XOS with narrow molecular weight distribution, while more polydisperse fractions were obtained by ultrafiltration with membranes of MWCO from 5 to 30 kDa. Thus, membranes with narrow pore size distribution should be used to minimize the dispersion of molecular sizes. Nevertheless, depending on the fractions of interest, one- or two-step ultrafiltration systems could be used for the purification of arabino-XOS. This work was performed in a dead-end system, which is not designed for long-term operations; furthermore, agitation rate and temperature used were not reported. Another issue to consider in ultrafiltration is that membrane apparent MWCO does not always match exactly with the nominal MWCO. Nabarlatz et al. (2007) evaluated different ultrafiltration membranes (MWCO from 1 to 8 kDa) for the purification of XOS obtained from almond shells, using flat-sheet membranes in cross-flow array. Despite being a linear correlation between the nominal MWCO and the apparent MWCO experimentally observed, these later were higher in all cases than the nominal values declared by the membrane supplier. This explains the presence of some oligomers (from 0.5 to 5 kDa) in the permeate, even though at much lower concentrations than those in the feed. Additionally, since crude extracts from autohydrolysis were subjected to enzymatic hydrolysis, extracted compounds included a wide molar mass distribution (from 0.1 to 70 kDa). In all cases, selectivity towards permeation of lignin-derived compounds (impurities) decreased at higher fluxes regardless of the applied pressure, and concentration of XOS was higher with membranes having lower MWCO. Nevertheless, due to the wide molecular size distribution obtained by this method, an additional purification step must be considered and, in the case of XOS, ultrafiltration seems to be an alternative for pretreatment rather than for obtaining a highly purified product with short-chain oligosaccharides. Regarding to the NF of XOS, Yuan et al. (2004) developed a pilot-plant production for the production of XOS from corncobs. Production stages included steaming extraction, enzymatic hydrolysis, and NF. For NF, four different membranes were tested (processing conditions and membrane materials were not specified). A compromise between permeate flux and rejection coefficient of XOS was adopted. Although XOS (polymerization degree from 2 to 5 units) were concentrated 10-fold after filtration, monosaccharides (xylose and glucose) were also retained and concentrated during filtration. Therefore, a marginal increase in the purity of the XOS product was achieved (from 69.6% to 73.8%). This reflects one of the main drawbacks of

136

Separation of Functional Molecules in Food by Membrane Technology

using NF for oligosaccharides purification: a nonselective separation between mono- and disaccharides. This situation occurs because the retention of a target compound by the membrane is not directly related to its MWCO, and depends strongly on not only the hydrodynamics of the solution (the influence of the temperature on the viscosity, the cross-flow velocity, transmembrane pressure, solute concentration in the feed), but also on the chemical structure of the compound. For instance, Hua et al. (2010) reported that during NF of XOS from corncobs, using thin film composite membranes (MWCO 250 Da), the retention of arabinose was higher than xylose, despite having the same molecular weight (150 Da). This was attributed to the different hydroxyl configurations around C2 and C3 in these isomers, resulting in unequal polarity between them. This is why MWCO is a more appropriate concept for ultrafiltration, while apparent retention coefficient is more appropriate for NF. Alternatively, the use of a higher temperature during filtration may improve NF performance for XOS concentration in a reduced time due to enhanced permeation. The cost to be paid for such increase in temperature is the permeation of xylobiose, thus decreasing the total mass of concentrated XOS obtained. The degree of substitution has shown to have a direct effect on membrane filtration of XOS (Swennen et al., 2005). Therefore, as filtration proceeds, deposition of compounds having higher molecular size on the membrane surface will raise the rejection of xylobiose, improving the retention of total XOS. Interesting results can also be obtained using membranes of higher MWCO. Vegas et al. (2006) used a ceramic membrane with molecular mass cut-off of 1000 Da in two processes for processing liquors from rice husks where NF was used in combination with solvent extraction and ion exchange. By operating transmembrane pressure at 14 bar and 34 C, the simultaneous concentration and purification of XOS was achieved. The success of the process was mainly due to the preferential removal of monosaccharides and nonsaccharide compounds (mainly phenolic compounds), as well as by the high retention of XOS (92%). NF of XOS can be conducted with tighter membranes (MWCO B 250 Da) using continuous and discontinuous diafiltration (Zhao et al., 2012). In such system, monosaccharides can be washed while xylobiose passes through the membrane during the first minutes of filtration. Then, xylobiose begins to be retained (with the other XOS of higher degree of polymerization), reaching a purity for total XOS of B94%. Fig. 4.6 summarizes the different alternatives for the extraction and purification of XOS using a membrane technology approach. From the results shown above, it can be concluded that membrane filtration of XOS is a feasible technology for their purification. One advantage deserving mention is that typical concentrations of liquors extracted have initial purities of XOS ranging from 65% to 83%, thus products with very high purities can be achieved by

Membrane Technology for the Purification Chapter | 4

137

FIGURE 4.6 Process diagram for enzymatically produced xylooligosaccharides (XOS) in a membrane technology approach. Dotted lines representing an alternative step for enzyme hydrolysis in ultrafiltration membrane bioreactor (MBR).

simple partial removal of monosaccharides. However, this is not always fulfilled; there are cases where much lower purities have been also reported.

4.4.2 Purification of Isomaltooligosaccharides Consist of α-D-glucose residues linked mostly by α(1-6) glycosidic bonds. Some also contain mixtures of both α(1-6) and α(1-4) linked glucose units, such as in the case of the trisaccharide panose. Unlike other oligosaccharides’ laxative effect, isomaltooligosaccharides (IMOS) are well tolerated at higher doses. IMOS are produced using a combination of immobilized enzymes in a two-stage reactor: in the first stage, starch is liquefied using α-amylases. The liquefied starch is then processed in a second stage that involves reactions catalyzed by both β-amylase and α-glucosidase: the β-amylase first hydrolyses the liquefied starch to maltose and then IMOS are synthesized by the transglucosidase activity of α-glucosidase (Crittenden and Playne, 1996). Commercial IMOS preparations consist of isomaltose, isomaltotriose, panose, and isomaltotetraose as major compounds, but substantial proportions of maltose and glucose are also present, so that the raw IMOS must be purified. Methods for the purification of IMOS, such as chromatography (Demuth et al., 2002), have been applied, but selective yeast fermentation appears as the most used alternative (Crittenden and Playne, 1996; Pan and Lee, 2005; Chockchaisawasdeea and Poosaran, 2013). On the other hand, few studies have been done for the appraisal of membrane technology as a strategy for the purification of IMOS.

138

Separation of Functional Molecules in Food by Membrane Technology

Goulas et al. (2004) studied the synthesis of IMOS in a MBR with combined use of dextrasucrase and dextranase. The system included a polyethersulfone/polysulfone membrane (MWCO 10 kDa), operating at 30 C. Results showed that by increasing substrate concentration flux decayed, but by increasing of dextranase concentration flux was increased because it affected the molecular size of the sugars produced, having direct effect on the concentration polarization. The IMOS produced in the system had molecular weights lower than in batch synthesis reactions. Despite this, the IMOS formed in the MBR had a degree of polymerization greater than 5, depending on the reaction conditions. Similarly, Liaw et al. (2000) developed a patent for producing for IMOS based different saccharification steps followed by sequential membrane separations at high temperatures (micro and ultrafiltration) where these last ones are carried out as an integral part of the saccharification process; that is, by the continuous recirculation of thermostable α and β glucoamylases. Specifically regarding the fractionation of IMOS by NF, to the best of our knowledge, only the work of Goulas et al. (2003) has compared the performance of different membranes. Two NF membranes—NF-CA-50. composed of cellulose acetate, and NF-TFC-50, a thin film trilaminate membrane composed of polyethersulfone (both from Intersep Ltd)—were used for purifying commercial syrup of IMOS (Panorich). The composition of the syrup was 230 g/kg glucose, 260 g/kg maltose and isomaltose, 300 g/ kg panose and other branched oligosaccharides. Additionally, an ultrafiltration membrane (1 kDa MWCO) of cellulose acetate UF-CA-1 was also tested. The system consisted of dead-end filtration cell operated in discontinuous diafiltration at 40 bar, stirring rate at 150 rpm at room temperature (20 C to 25 C). Diafiltration was conducted in 4 repeated steps. Although 95% of the monosaccharides were removed by membrane UF-CA-1, its larger MWCO gave low purification yield for di- and oligosaccharides, which are responsible for the prebiotic effect. The best results were achieved with membrane NF-TFC-50 where around of 80% of monosaccharides were removed, while 88% of the di- and oligosaccharides were rejected by the membrane. Despite the contribution of this work, its main drawback was that carbohydrate profile, specifically for disaccharides, was not reported. IMOS have DP value from 2 to 10, linkages types α-1-2, α 1-3, α 1-4, or α 1-6, and a defined proportion and position of each type of linkage (Goffin et al., 2011). In particular, isomaltose (α 1-4) constitutes an IMO, but not maltose (α 1-6). Hence, a suitable purification of IMOS implies the difficult goal of being able to selectively separate both disaccharides in such a way that the maltose crosses the membrane, while the isomaltose is retained. Separation of molecules having same molecular mass is a tough challenge in NF but not impossible (Hua et al., 2010). Other variables such as the concentration ratio between the isomers, the difference between its stoke radius, as well as their solubilities have to be taken into account.

Membrane Technology for the Purification Chapter | 4

139

4.4.3 Purification of Galactooligosaccharide and Other Lactose-Derived Prebiotics Synthesis of GOS normally occurs with low yields ranging from 20% to 45%, which correspond to lactose conversions between 40% and 60% (Torres et al., 2010). Due to the kinetically controlled nature of the reaction, the raw GOS solutions include a mixture of hydrolysis products (glucose and galactose), unreacted lactose, and the resulting transgalactosylation GOS products. Fig. 4.7 shows schematically the reaction mechanism of GOS synthesis (Vera et al., 2011). The most abundant transgalactosylation products obtained with commercial β-galactosidases are low molecular weight oligosaccharides, mostly trisaccharides (GOS-3) and tetrasaccharides (GOS-4), and much lower amounts of penta(GOS-5) and hexaoligosaccharides (GOS-6). The main glycosidic bonds in these structures are β1-4 and β1-6 (Coulier et al., 2009). Prebiotic effect is mostly associated to GOS-3 and GOS-4, but more recently disaccharides containing β (1-6) bonds, that is, allolactose and galactobiose, have been reported to present similar bifidogenic effect to GOS-3 (Rodriguez-Colinas et al., 2014). For producing GOS with high purity, removal of residual lactose is mandatory for avoiding crystallization and allowing consumption by lactose intolerant people; removal of monosaccharides is important for reducing its caloric contribution and making the product available for diabetic people (Vera et al., 2016). Goulas et al. (2002) conducted a pioneering study that evaluated the effect of transmembrane pressure and feed concentration on GOS purification. They used model solutions simulating the composition of raw GOS. From the results obtained with a commercial galactooligosaccharide product (Vivinal GOS) using several membranes, those that offered the best fractionation performance were selected. The NF system consisted of a high pressure cell test

FIGURE 4.7 Mechanism of synthesis of galactooligosaccharides (GOS) with β-galactosidase from A. oryzae. Modulation by glucose has not been considered (Vera et al., 2011).

140

Separation of Functional Molecules in Food by Membrane Technology

unit (Osmonics Desal) in cross-flow operated at 25 C, in a pressure range from 7 to 27 bar, using total solute concentrations of 80 g/L in full recirculation mode. As expected, the increase in transmembrane pressure allowed increasing the flux. However, this increase caused a greater compaction of the solutes in the membrane, which resulted in an increase in the retention coefficient not only of oligosaccharides, but also of mono- and disaccharides. The best condition was achieved at 60 C with the membrane DS-5DL (MWCO not specified but MgSO4 rejection $ 96%), where 80% of the monosaccharides, but only 11% of the lactose were removed. Probably, the used membrane was too tight for such purpose. Facing such difficulty, an alternative that can be useful in a NF system is performing it in diafiltration mode. Herna´ndez et al. (2009) purified commercial Vivinal GOS diluted to 50 g/L in continuous diafiltration during 5 hours, but similar results to those of Goulas et al. (2002) were obtained, even though the used cellulose acetate membranes had higher MWCO (500 and 1000 Da). One aspect to keep in mind is that, despite the values reported in these works, Vivinal GOS is a prepurified product containing a high concentration of oligosaccharides (around 60%). As described above, reaction yields obtained in the synthesis of GOS are low, usually in the range from 25% to 40% depending on the enzyme source. The maximum purity of GOS to be obtained by NF depends on processing conditions, but also the origin of the β-galactosidase is a very important variable to consider because of its effect in the reaction yield. Therefore, the biocatalyst used in the stage of GOS synthesis will determine the initial composition of the reaction mixture to be purified. NF studies of GOS synthesized with different enzymes have been conducted, normally using low solute concentrations. For instance, Feng et al. (2009) used spiral wound membranes in continuous diafiltrations, treating GOS solutions at 50 g/L. Later on, Michelon et al. (2014) used flatsheets membranes in a dead-end NF unit to purify GOS at 100 g/L produced by Kluyveromyces marxianus β-galactosidase. Ren et al. (2015) produced GOS by continuous diafiltration coupled to a MBR, using 50 g/L of lactose and β-galactosidase from K. lactis. More concentrated solutions were studied by Pruksasri et al. (2015) using raw GOS solutions at 216 g/L synthesized with Lactobacillus delbrueckii bulgaricus β-galactosidase, while Co´rdova et al. (2016c) demonstrated the NF potential with GOS solutions, obtained with A. oryzae β-galactosidase, at very high concentrations (460 g/L). Despite to the variability of results obtained GOS purity attainable by NF does not exceed 65%. The facts that explain such operational limitation are: G

G

the small difference in molecular size between lactose (342 Da) and GOS-3 (504 Da); the high lactose concentration remaining at the end reaction, which can be two- to threefold higher than that of GOS, depending on the biocatalyst used;

Membrane Technology for the Purification Chapter | 4 G

G

141

the poor selectivity that membranes offer to this fractionation where the use of tight membranes leads to a high rejection of lactose, while the use of membranes with larger pore sizes increase the permeation not only of lactose, but also of GOS-3, thus reducing the concentration of purified GOS in the retentate; the wide distribution of pore size that some polymeric membranes exhibit.

Despite the limitations described above, there are some alternatives that can contribute to the increased product yield in the NF of GOS. Maximum purification yield implies that during the operation, maximum rejection of GOS is achieved with minimum rejection of lactose and monosaccharides. Precisely, Pruksasri et al. (2015) showed that by NF at 5 C the difference between the rejection coefficient of lactose and trisaccharides was much higher than at temperatures used in NF of GOS (normally from 30 C to 60 C). Such effect was observed from 25 to 45 bar, thus improving the membrane selectivity towards the recovery of GOS (80.5%). From an economic point of view, higher production costs will result due to the low permeate flux, which in that case was around 3 L/m2/h because of the higher viscosity at lower temperatures. Alternatively, the incorporation of a previous step of lactose hydrolysis was proposed by Santiba´n˜ez et al. (2016) for the purification of raw GOS. This later strategy, by converting lactose into its monosaccharide components, may improve purification due to the higher difference in molecular size between oligosaccharides and monosaccharides. Recently, Co´rdova et al. (2017) compared the NF of GOS with and without prehydrolysis of lactose. After three sequential batches, similar purities (B55%) were obtained with both GOS solutions, but the NF time necessary for achieving the same volume reduction factor was considerably shorter for GOS subjected to lactose prehydrolysis. Thus, the use of lactose prehydrolysis had two opposite effects in GOS purification by NF: an increased flux due to the lower lactose deposition on the membrane, and a lower GOS retention by the membrane, resulting in a marginal gain in purification. However, this last effect was mostly due to the temperatures used during NF (from 53.5 C to 65 C), which implies an increase in pore size that allows an easy permeation of the oligosaccharides, affecting the membrane selectivity. Hence, low temperatures are recommended for this application. Another strategy to produce higher purities of GOS is subjecting the solutions to a continuous multistage operation. Patil et al. (2014) evaluated the potential of membrane cascading on downstream processing of GOS concluding that a five-stage cascade system would be optimal for such purpose. A valid questioning against this option would be the possible increase in investment and operating costs, but this has to be properly evaluated at production scale. For instance, Vanneste et al. (2011) conducted a technoeconomic evaluation

142

Separation of Functional Molecules in Food by Membrane Technology

where membrane cascades were compared with simulated moving bed chromatography for the purification of mono- and oligosaccharides. Separation of raffinose (trisaccharide) from sucrose (disaccharide) by membrane cascade yielded similar costs than simulated moving bed chromatography. However, competitiveness of membrane cascade system increased with the plant size. In conclusion operational aspects of NF of GOS still require further technological implementation. It is also worthwhile to make a parallel with the purification by NF of other lactose-derived oligosaccharides, such as lactulose and lactosucrose. Lactulose is a disaccharide composed of galactose and fructose that has been used for treating constipation and hepatic encephalopathy. Also, lactulose has gained interest as a functional food ingredient because it can be properly considered as prebiotic (Guerrero et al., 2011). Conventionally, lactulose has been produced by chemical isomerization, but in recent years the alternative of producing it by enzymatic synthesis with β-galactosidase using lactose and fructose as substrate has gained interest, due to the possibility of using mild reaction conditions making the process compliant with the principles of green chemistry (Panesar and Kumari, 2011). As in the case of GOS, this is a kinetically controlled reaction where transglycosylation and hydrolysis reactions will occur, but the resulting product is even more heterogeneous as a result of the use of two different substrates in the reaction. Thus a mixture of unreacted fructose and lactose, glucose, lactulose, GOS, and FOS will result at end of the synthesis. This wide variety of resulting products can be an interesting and challenging opportunity for the use of NF processes to obtain fractionation. However, this technology has been poorly exploited for such purpose. For instance, Sitanggang et al. (2014) studied the use of an enzyme MBR to produce lactulose. The continuous removal of lactulose restricted their secondary hydrolysis, improving the productivity and reducing the enzyme consumption in comparison with conventional batch systems. Zhang et al. (2011) studied the purification of lactulose by using NF in diafiltration mode, but since the lactulose syrup was produced by chemical catalysis, the main goal of that work was the separation of lactulose from the reaction catalysts (NaCl and H3BO3). Despite the successful removal of both catalysts ($ 96.5%), separation between both disaccharides (residual lactose and lactulose) was not achieved. Lactosucrose is a synthetic trisaccharide composed of galactose, glucose, and fructose with high potential though not yet considered as prebiotic. Lactosucrose is also highlighted because of its properties conferring technological potential (Silve´rio et al., 2015). At the industrial level, lactosucrose can be obtained in a batch reaction process by a transfructosylation reaction catalyzed by β-fructofuranosidase using sucrose and lactose as substrates. However, maximum lactosucrose content does not exceed 70%
75% with respect to total sugars. Other than lactosucrose compounds

Membrane Technology for the Purification Chapter | 4

143

are glucose, fructose and FOS, as well as unreacted lactose and sucrose. Purification strategies such as selective fermentation, simulating moving bed chromatography, and activated charcoal adsorption have been proposed for the purification of lactosucrose, but no research has been done on the use of NF for this purpose. In this sense, several research questions may arise: since lactosucrose is a trisaccharide and glucose and fructose are monosaccharides, is NF a suitable technology to purify lactosucrose? Is it possible to separate FOS from lactosucrose to produce the simultaneous purification of two different types of prebiotics produced in the same enzymatic reactor? What kind of membrane arrays could make possible such fractionations? Is NF a competitive technology in terms of costs and productivity for the purification of lactosucrose, compared with previously evaluated technologies? All these questions and others may open several interesting new research opportunities to this field to which attention has not been paid yet.

4.4.4 Purification of Fructooligosaccharides Such as in the purification of other prebiotics, the literature about FOS purification by membrane technology is scarce. Actually, until 2015 there were very few reports available regarding FOS purification (Nobre et al., 2015). Li et al. (2004) studied NF in diafiltration mode in two different modes considering constant and variable volume diafiltration, concluding that for a given purity, the mode of variable volume diafiltration needs a lower dilution and water consumption, being an interesting achievement with respect to sustainable water management. According to their results, the assumption that the concentration polarization is low enough to be considered negligible is suitable for a FOS system using spiral wound elements of 1.7 m2. Besides, in both modes, they achieved purities up to 90% with yields up to B50%) from solutions with 30.9% glucose and fructose, 14.4% sucrose, 26.0% 1-kestose (DP3), nystose (DP4), and 1-β-frusctofuranosyl nystose (DP5) with rejection coefficient of 0.81, 0.55, 0.24, 0.13, and 0, respectively, using membranes GH-NF (GE Water & Process Technologies). The higher the difference among rejections of FOS and sugars (mono- and disaccharides) the better will be the separation process. Also, Li et al. (2005) determined an extended pore model for reflection coefficient and solute permeability taking into account the pressure gradient, steric hindrance, and wall friction. Earlier, Nishizawa et al. (2001) tested several NF membranes. NTR-7410 and NTR 7450 showed low rejection for glucose and sucrose (in the range of 1%
2%), and low rejection for DP3, DP4, and DP5, in the range of 3%
28%; meanwhile NF-45, NTR-7250, and NTR 729HP showed over 99% rejection for DP3, DP4, and DP5, but high rejection for sucrose and glucose too (in the range from 75% to 97%). Despite these results, the NF-45 membrane, the one allowing the higher flux, was selected for FOS production

144

Separation of Functional Molecules in Food by Membrane Technology

using β-fructofuranosidase from A. niger in a MBR. After 12 hours of operation the composition of the concentrate stream was 93% of FOS and 5% of sucrose, better than in the batch system where a product with 55%
60% FOS was obtained. Kuhn et al. (2011) tested several membranes to purify FOS, among them NP10 and NP030 (Microdyn Nadir) Desal-5 DL and Desal-HL (GE Water & Process Technologies) in dead-end cell and tangential cell devices. The best results were achieved using the membrane NP030 with observed retention of 0.71 for FOS, 0.31 for sucrose, 0.02 for glucose, and 0.001 for fructose in tangential membrane cell filtration. Also, diafiltration was performed reaching a purity of 80%, but yield was not reported. In another work, Kuhn et al. (2010) published an interesting two-stage purification process. The first one was a NF in diafiltration mode followed by a concentration stage and in the second stage, the permeate of the first one was concentrated to increase the overall yield of the process. Six membranes were tested (NP010, NP030, NF 270 from Dow-Filmtec and DL, HL, and DK from Desal-Osmonics). Again, NP030 showed the best performance: rejection coefficient of 87% for FOS, 58% for sucrose, 9% for glucose, and 1% for fructose. Although NF030 was not the membrane allowing higher flux, it was selected for meeting the following criteria: the observed retention for FOS should exceed 50% and the rest of the sugars must be significantly below this value, and steady observed retention during time is preferred. The NF030 was selected for being used in both stages in a spiral wound membrane with a 1.8 m2 of filtration area. Purity over 90% with overall process yield around 80% were obtained. It can be concluded from those studies that NF in diafiltration mode is a sound strategy for FOS purification, and high purities ($90%) have been achieved already. However, not only purity must be taken into account for developing a suitable NF operation; other key parameters, such as yield, diluting water consumption, operation time, and productivity, need more attention from researchers to make this technology applicable at industrial scale.

4.4.5 Purification of Other Oligosaccharides Knowledge related to the purification of other oligosaccharides different than GOS, FOS, and XOS is still lacking. No information in the scientific literature has been found dealing with purification by NF of gentiooligosaccharides. Fructooligosteviol, which is a novel prebiotic with sweetener properties, was produced by enzymatic synthesis (Spohner and Czermak, 2016). Purification of lactosucrose by NF has not been reported to date. Soybean oligosaccharides are not produced by enzyme technology, but recently their fractionation by NF has been reported (Li et al., 2018). Other compounds to be considered as prebiotic are pectin-oligosaccharides (POS). Partial hydrolysis of pectin by chemical and/or enzymatic methods leads to the production of pectin-derived oligosaccharides (POS), which have been

Membrane Technology for the Purification Chapter | 4

145

proposed as a new class of prebiotics. However, limited information exists on experimental studies assessing the prebiotic effects caused by POS in humans (Gullo´n et al., 2013). The most usual fractionation technologies used in POS are ion exchange chromatography and gel filtration chromatography (GF). Regarding purification of POS by membrane technology, Iwasaki and Matsubara (2000) used ultrafiltration and NF for purifying POS obtained by enzymatic hydrolysis of citrus pectin. The first stage of UF was used to remove nonhydrolyzed pectate while NF removed large amounts of monogalacturonic acid and other contaminants. To our knowledge there has been no further progress for this latest application. Therefore, there is a clear opportunity for future research in this novel field of knowledge.

4.5 CHALLENGES AND PERSPECTIVES As to types of membranes, there are two main types: ceramic and polymeric. The first category belongs to the porous membranes while polymeric are dense membranes. Ceramic membranes have several advantages, such as high chemical, mechanical, and thermal stability, compared with their polymeric counterparts (Baruah et al., 2006). Ceramic membranes are compatible with the use of steam sterilization and back flushing, and exhibit high abrasion resistance. Also, they can be regenerated by chemical cleaning and stored dry after it. The durability of ceramic membranes properly used is up to 6 years (Ge´san-Guiziou et al., 2007). Their disadvantages are brittleness, they require high cross-flow velocities, and their high cost, which makes them more expensive than polymeric membranes. Ceramic membranes are well developed for microfiltration and ultrafiltration, but until now, it has been difficult to find in the market ceramic membranes for NF. This is because the slurry method used for synthesis of MF and UF membranes is not able to generate pore size below 1 nm (Duscher, 2014). For that task NF ceramic membranes have to be produced by the sol
gel process (Van Gestel et al., 2006). Today, Inopor GmbH sells Inopor Nano series of TiO2, in three different MWCO: 750, 450, and 200 Da with a porosity of 30%
40% and length from 100 to 1200 mm being applicable for laboratory- and large-scale applications. As stated above, oligosaccharides produced enzymatically normally are highly concentrated solutions. Hence, relative high temperatures (40
60 C) must be used to reduce viscosity for increasing permeate flux. However, such level of temperature changes the membrane selectivity by increasing the pore size, decreasing the rejection of oligosaccharides with DP $ 3. Therefore, the use of ceramic membranes will avoid the effect of temperature on the rejection of oligosaccharides because their pore size remains unaffected to temperature, contrary to the case of polymeric membranes. Moreover, the nominal MWCO of the few available NF ceramic membranes is within the molecular size range suitable for the fractionation of

146

Separation of Functional Molecules in Food by Membrane Technology

oligosaccharides. It will be interesting to compare the performance of these novel membranes in the fractionation of oligosaccharide solutions of different nature. Finally, other novel alternative such as the use of vibrating modules, electrofiltration, and backshocking stated by Pinelo et al. (2009) have not been yet used systematically for oligosaccharide NF as well as for other NF applications. In the case of vibrating modules, only oligoglucuronans have been reported as being purified by high shear rate dynamic NF (Mellal et al., 2008). Despite that, vibrating modules are being used widely in MF and UF applications (Al-Akoum et al., 2002; Akoum et al., 2005). Other uses of vibrating modules have been applied for NF and reverse osmosis in the treatment of dairy effluents (Ding and Jaffrin, 2014). Additionally, microfluidic contactors (Kolfschoten et al., 2011) are a novel application, but until now have only been used in laboratory studies. Among the challenges still pending for the use of membrane technology for purification purposes at large scale are to avoid fouling, and to increase the lifetime of membranes and their chemical resistance. The first one can be addressed to a certain extent by performing filtration with critical or threshold flux concept, while longer membrane lifetimes can be obtained using ceramic membranes.

4.6 CONCLUSIONS From the state of the art it can be concluded that the purity that can be obtained in oligosaccharides NF as a function of the proportion of the constituents of the reaction mixture. An example is the higher purity that can be achieved with NF in FOS and XOS, compared with GOS. In the last one, the percentage of residual lactose is prominent compared with the produced oligosaccharides. In this sense, the type of biocatalyst used will have a direct effect on the final purity achieved by NF, due to the effect of the enzyme in the reaction yield and the conversion. On the other hand, evidence from recent years demonstrates that the selectivity of the membrane can be improved using concentrated oligosaccharide solutions. In fact, problems associated with the generation of fouling because of using highly concentrated solutions can be remedied by a series of easy-to-implement actions, such as operating under critical transmembrane pressure as well as by defining the effect of temperature and cross-flow velocity on the membrane selectivity and flux stability. Also, NF in diafiltration mode seems to be the most widely used fractionation strategy; but studies owing to investment, processing, water consumption, and energy costs must be properly assessed. Finally, it can be concluded that NF remains advantageously competitive over other purification strategies such as chromatographic or ion exchange technology, as long as maximum purity desired in the product does not

Membrane Technology for the Purification Chapter | 4

147

exceed 70%
80%. To date, greater purities than 85% can be obtained in specific cases such as XOS and FOS, but it is not yet a generalization that can be extended for NF to all types of enzymatically produced oligosaccharides.

REFERENCES Aachary, A.A., Prapulla, S.G., 2011. Xylooligosaccharides (XOS) as an emerging prebiotic: microbial synthesis, utilization, structural characterization, bioactive properties, and applications. Compr. Rev. Food Sci. Food Saf. 10, 2
16. Abdul Manas, N.H., Md Illias, R., Mahadi, N.M., 2018. Strategy in manipulating transglycosylation activity of glycosyl hydrolase for oligosaccharide production. Crit. Rev. Biotechnol. 38, 272
293. Akoum, O., Jaffrin, M.Y., Ding, L.-H., 2005. Concentration of total milk proteins by high shear ultrafiltration in a vibrating membrane module. J. Memb. Sci. 247, 211
220. Al-Akoum, O., Ding, L.H., Jaffrin, M.Y., 2002. Microfiltration and ultrafiltration of UHT skim milk with a vibrating membrane module. Sep. Purif. Technol. 28, 219
234. Al-Sheraji, S.H., Ismail, A., Manap, M.Y., Mustafa, S., Yusof, R.M., Hassan, F.A., 2013. Prebiotics as functional foods: a review. J. Funct. Foods 5, 1542
1553. Arslanoglu, S., Moro, G., Schmitt, J., Tandoi, L., Rizzardi, S., Boehm, G., 2008. Early dietary intervention with a mixture of prebiotic oligosaccharides reduces the incidence of allergic manifestations and infections during the first two years of life. J. Nutr. 138, 1091
1095. Astudillo-Castro, C.L., 2015. Limiting flux and critical transmembrane pressure determination using an exponential model: the effect of concentration factor, temperature, and cross-flow velocity during casein micelle concentration by microfiltration. Ind. Eng. Chem. Res. 54, 414
425. Bandini, S., Morelli, V., 2017. Effect of temperature, pH and composition on nanofiltration of mono/disaccharides: experiments and modeling assessment. J. Memb. Sci. 533, 57
74. Baruah, G.L., Nayak, A., Belfort, G., 2006. Scale-up from laboratory microfiltration to a ceramic pilot plant: design and performance. J. Memb. Sci. 274, 56
63. Bhattacharjee, S., Chen, J.C., Elimelech, M., 2001. Coupled model of concentration polarization and pore transport in crossflow nanofiltration. AIChE J. 47, 2733
2745. Boeckner, L.S., Schnepf, M.I., Tungland, B.C., 2001. Inulin: a review of nutritional and health implications. Adv. Food. Nutr. Res. 43, 1
63. Botelho-Cunha, V.A., Mateus, M., Petrus, J.C.C., de Pinho, M.N., 2010. Tailoring the enzymatic synthesis and nanofiltration fractionation of galacto-oligosaccharides. Biochem. Eng. J. 50, 29
36. Bowen, W.R., Mukhtar, H., 1996. Characterisation and prediction of separation performance of nanofiltration membranes. J. Memb. Sci. 112, 263
274. Capuano, E., Fogliano, V., 2011. Acrylamide and 5-hydroxymethylfurfural (HMF): a review on metabolism, toxicity, occurrence in food and mitigation strategies. LWT - Food Sci. Technol. 44, 793
810. Chaabane, T., Taha, S., Cabon, J., Dorange, G., Maachi, R., 2002. Dynamic modeling of mass transfer through a nanofiltration membrane using calcium salt in drinking water. Desalination 152, 275
280. Chaabane, T., Taha, S., Taleb Ahmed, M., Maachi, R., Dorange, G., 2007. Coupled model of film theory and the Nernst-Planck equation in nanofiltration. Desalination 206, 424
432. Cheryan, M., 1998. Ultrafiltration and Microfiltration Handbook, Vasa.

148

Separation of Functional Molecules in Food by Membrane Technology

Chockchaisawasdeea, S., Poosaran, N., 2013. Production of isomaltooligosaccharides from banana flour. J. Sci. Food Agric. 93, 180
186. Co´rdova, A., Astudillo, C., Vera, C., Guerrero, C., Illanes, A., 2016a. Performance of an ultrafiltration membrane bioreactor (UF-MBR) as a processing strategy for the synthesis of galacto-oligosaccharides at high substrate concentrations. J. Biotechnol. 223, 26
35. Co´rdova, A., Astudillo, C., Guerrero, C., Vera, C., Illanes, A., 2016b. Assessment of the fouling mechanisms of an ultrafiltration membrane bioreactor during synthesis of galacto-oligosaccharides: effect of the operational variables. Desalination 393, 78
89. Co´rdova, A., Astudillo, C., Giorno, L., Guerrero, C., Conidi, C., Illanes, A., et al., 2016c. Nanofiltration potential for the purification of highly concentrated enzymatically produced oligosaccharides. Food Bioprod. Process 98, 50
61. Co´rdova, A., Astudillo, C., Santiban˜ez, L., Cassano, A., Ruby-Figueroa, R., Illanes, A., 2017. Purification of galacto-oligosaccharides (GOS) by three-stage serial nanofiltration units under critical transmembrane pressure conditions. Chem. Eng. Res. Des. 117, 488
499. Coulier, L., Timmermans, J., Richard, B., Van Den Dool, R., Haaksman, I., Klarenbeek, B., et al., 2009. In-depth characterization of prebiotic galactooligosaccharides by a combination of analytical techniques. J. Agric. Food Chem. 57, 8488
8495. Crittenden, R.G., Playne, M.J., 1996. Production, properties and applications of food-grade oligosaccharides. Trends Food Sci. Technol. 7, 353
361. Czermak, P., Ebrahimi, M., Grau, K., Netz, S., Sawatzki, G., Pfromm, P., 2004. Membraneassisted enzymatic production of galactosyl-oligosaccharides from lactose in a continuous process. J. Memb. Sci. 232, 85
91. Demuth, K., Jordening, H.J., Buchholz, K., 2002. Oligosaccharide synthesis by dextransucrase: new unconventional acceptors. Carbohyd. Res. 337, 1811
1820. Ding, L., Jaffrin, M.Y., 2014. Benefits of high shear rate dynamic nanofiltration and reverse osmosis: a review. Sep. Sci. Technol. 49, 1953
1967. Dresner, L., 1972. Some remarks on the integration of the extended Nernst-Planck equations in the hyperfiltration of multicomponent solutions. Desalination 10, 27
46. Duscher, S., 2014. Ceramic membranes for the filtration of liquids : an actual overview. Filtr. Sep. Int. Ed. 13
21. Feng, Y.M., Chang, X.L., Wang, W.H., Ma, R.Y., 2009. Separation of galacto-oligosaccharides mixture by nanofiltration. J. Taiwan Inst. Chem. Eng. 40, 326
332. Field, R.W., Pearce, G.K., 2011. Critical, sustainable and threshold fluxes for membrane filtration with water industry applications. Adv. Colloid Interface Sci. 164, 38
44. Field, R.W., Wu, D., Howell, J.A., Gupta, B.B., 1995. Critical flux concept for microfiltration fouling. J. Memb. Sci. 100, 259
272. Freixo, M.R., De Pinho, M.N., 2002. Enzymatic hydrolysis of beechwood xylan in a membrane reactor. Desalination 149, 237
242. Fujimoto, Y., Hattori, T., Uno, S., Murata, T., Usui, T., 2009. Enzymatic synthesis of gentiooligosaccharides by transglycosylation with β-glycosidases from Penicillium multicolor. Carbohydr. Res. 344, 972
978. Galanakis, C.M., Kanellaki, M., Koutinas, A.A., Bekatorou, A., Lycourghiotis, A., Kordoulis, C. H., 2012. Effect of pressure and temperature on alcoholic fermentation by Saccharomyces cerevisiae immobilized on γ-alumina pellets. Bioresour. Technol. 114, 492
498. Ge´san-Guiziou, G., Alvarez, N., Jacob, D., Daufin, G., 2007. Cleaning-in-place coupled with membrane regeneration for re-using caustic soda solutions. Sep. Purif. Technol. 54, 329
339. Gibson, G.R., Probert, H.M., Van Loo, J., Rastall, R.A., Roberfroid, M.B., 2004. Dietary modulation of the human colonic microbiota: updating the concept of prebiotics. Nut. Res. Rev. 17, 259
275.

Membrane Technology for the Purification Chapter | 4

149

Giorno, L., De Bartolo, L., Drioli, E., 2003. Membrane bioreactors for biotechnologyand medical applications. In: Bhattacharyya, D., Butterfield, D.A. (Eds.), New Insights Into Membrane Science and Technology: Polymeric and Biofunctional Membranes. Elsevier Science, pp. 187
217. Goffin, D., Delzenne, N., Blecker, C., Hanon, E., Deroanne, C., Paquot, M., 2011. Will Isomalto-oligosaccharides, a well-established functional food in Asia, Break through the European and American Market?. The status of knowledge on these prebiotics. Crit. Rev Food Sci. Nut. 51, 394
409. Goulas, A.K., Kapasakalidis, P.G., Sinclair, H.R., Rastall, R.A., Grandison, A.S., 2002. Purification of oligosaccharides by nanofiltration. J. Memb. Sci. 209, 321
335. Goulas, A.K., Grandison, A.S., Rastall, R.A., 2003. Fractionation of oligosaccharides by nanofiltration. J. Sci. Food Agric. 83, 675
680. Goulas, A.K., Fisher, D.A., Grimble, G.K., Grandison, A.S., Rastall, R.A., 2004. Synthesis of isomaltooligosaccharides and oligodextrans by the combined use of dextransucrase and dextranase. Enzyme Microb. Technol. 35, 327
338. Guerrero, C., Vera, C., Plou, F., Illanes, A., 2011. Influence of reaction conditions on the selectivity of the synthesis of lactulose with microbial β-galactosidases. J. Mol. Catal. B Enzym. 72, 206
212. Gullo´n, B., Go´mez, B., Martı´nez-Sabajanes, M., Ya´n˜ez, R., Parajo´, J.C., Alonso, J.L., 2013. Pectic oligosaccharides: manufucture and functional properties. Trends Food Sci. Tech. 30, 153
161. Herna´ndez, O., Ruiz-Matute, A.I., Olano, A., Moreno, F.J., Sanz, M.L., 2009. Comparison of fractionation techniques to obtain prebiotic galactooligosaccharides. Int. Dairy J. 19, 531
536. Howell, J., 1995. Sub-critical flux operation of microfiltration. J. Memb. Sci. 107, 165
171. Hua, X., Zhao, H., Yang, R., Zhang, W., Zhao, W., 2010. Coupled model of extended NernstPlanck equation and film theory in nanofiltration for xylo-oligosaccharide syrup. J. Food Eng. 100, 302
309. Ishii, T., 1997. Structure and functions of feruloylated polysaccharides. Plant Sci. 127, 111
127. Iwasaki, K., Matsubara, Y., 2000. Purification of pectate oligosaccharides showing root-growthpromoting activity in lettuce using ultrafiltration and nanofiltration membranes. J. Biosci. Bioeng. 89, 495
497. Kabel, M.A., Carvalheiro, F., Garrote, G., Avgerinos, E., Koukios, E., Parajo´, J.C., et al., 2002. Hydrothermally treated xylan rich by-products yield different classes of xylooligosaccharides. Carbohydr. Polym. 50, 47
56. Kim, Y.S., Oh, D.K., 2012. Lactulose production from lactose as a single substrate by a thermostable cellobiose 2-epimerase from Caldicellulosiruptor saccharolyticus. Bioresour. Technol. 104, 668
672. Kizawa, H., Shinoyama, H., Yasui, T., 1991. The synthesis of new xylosyloligosaccharides by transxylosylation with Aspergillus niger β-xylosidase. Agric. Biol. Chem. 55, 671
678. Kolfschoten, R.C., Janssen, A.E.M., Boom, R.M., 2011. Mass diffusion-based separation of sugars in a microfluidic contactor with nanofiltration membranes. J. Sep. Sci. 34, 1338
1346. Kuhn, R.C., Maugeri Filho, F., Silva, V., Palacio, L., Herna´ndez, A., Pra´danos, P., 2010. Mass transfer and transport during purification of fructooligosaccharides by nanofiltration. J. Memb. Sci. 365, 356
365. Kuhn, R.C., Palacio, L., Pra´danos, P., Herna´ndez, A., 2011. Selection of membranes for purification of fructooligosaccharides. Desalin. Water Treat. 27, 18
24.

150

Separation of Functional Molecules in Food by Membrane Technology

Kuriki, T., Yanase, M., Takata, H., Takesada, Y., Imanaka, T., Okada, S., 1993. A new way of producing isomalto-oligosaccharide syrup by using the transglycosylation reaction of neopullulanase. Appl. Environ. Microbiol. 59, 953
959. Lee, J.-W., Kwon, T.-O., Moon, I.-S., 2004. Adsorption of monosaccharides, disaccharides, and maltooligosaccharides on activated carbon for separation of maltopentaose. Carbon N. Y. 42, 371
380. Li, W., Li, J., Chen, T., Chen, C., 2004. Study on nanofiltration for purifying fructooligosaccharides I. Operation modes. J. Memb. Sci. 245, 123
129. Li, W., Li, J., Chen, T., Zhao, Z., Chen, C., 2005. Study on nanofiltration for purifying fructooligosaccharides II. Extended pore model. J. Memb. Sci. 258, 8
15. Li, Z., Wang, K., Zhang, Y., 2018. Eco-friendly separation and purification of soybean oligosaccharides via nanofiltration technology. Sep. Sci. Technol. 53, 777
785. Liaw, G.C., Pedersen, S., Hendriksen, H.V., Svendsen, A., Nielsen, B.R., Nielsen, R.I., 2000. Method of producing saccharide preparations. US6136571 A. Loman, A.A., Ju, L.K., 2016. Soybean carbohydrate as fermentation feedstock for production of biofuels and value-added chemicals. Process Biochem. 51, 1046
1057. Ma¨ntta¨ri, M., Nystro¨m, M., 2000. Critical flux in NF of high molar mass polysaccharides and effluents from the paper industry. J. Memb. Sci. 170, 257
273. Ma¨ntta¨ri, M., Nuortila-Jokinen, J., Nystro¨m, M., 1997. Influence of filtration conditions on the performance of NF membranes in the filtration of paper mill total effluent. J. Memb. Sci. 137, 187
199. Matella, N.J., Dolan, K.D., Lee, Y.S., 2006. Comparison of galactooligosaccharide production in free-enzyme ultrafiltration and in immobilized-enzyme systems. J. Food Sci. 71, 363
368. Mazzei, R., Piacentini, E., Gebreyohannes, A.Y., Giorno, L., 2017. Membrane bioreactors in food, pharmaceutical and biofuel applications: state of the art, progresses and perspectives. Current Organic Chemistry . Mellal, M., Jaffrin, M.Y., Ding, L.H., Delattre, C., Michaud, P., Courtois, J., 2008. Separation of oligoglucuronans of low degrees of polymerization by using a high shear rotating disk filtration module. Sep. Purif. Technol. 60, 22
29. ˆ .S.M., Ferna´ndez Rodrı´guez, D.E., Ortiz, G.M.D., 2015. Meyer, T.S.M., Miguel, A Biotechnological production of oligosaccharides — applications in the food industry. In: Eissa, A.H.A. (Ed.), Food Production and Industry. INTECH. Michelon, M., Manera, A.P., Carvalho, A.L., Maugeri Filho, F., 2014. Concentration and purification of galacto-oligosaccharides using nanofiltration membranes. Int. J. Food Sci. Technol. 49, 1953
1961. Moure, A., Gullo´n, P., Domı´nguez, H., Parajo´, J.C., 2006. Advances in the manufacture, purification and applications of xylo-oligosaccharides as food additives and nutraceuticals. Process Biochem. 41, 1913
1923. Nabarlatz, D., Torras, C., Garcia-Valls, R., Montane´, D., 2007. Purification of xylooligosaccharides from almond shells by ultrafiltration. Sep. Purif. Technol. 53, 235
243. Nath, A., Bhattacharjee, C., Chowdhury, R., 2013. Synthesis and separation of galactooligosaccharides using membrane bioreactor. Desalination 316, 31
41. Nishizawa, K., Nakajima, M., Nabetani, H., 2001. Kinetic study on transfructosylation by. BETA.-Fructofuranosidase from Aspergillus niger ATCC 20611 and availability of a membrane reactor for fructooligosaccharide production. Food Sci. Technol. Res. 7, 39
44. Nobre, C., Teixeira, J.A., Rodrigues, L.R., 2015. New trends and technological challenges in the industrial production and purification of fructo-oligosaccharides. Crit. Rev. Food Sci. Nutr. 55, 1444
1455.

Membrane Technology for the Purification Chapter | 4

151

Palai, T., Bhattacharya, P.K., 2013. Kinetics of lactose conversion to galacto-oligosaccharides by β-galactosidase immobilized on PVDF membrane. J. Biosci. Bioeng. 115, 668
673. Pan, Y.C., Lee, W.C., 2005. Production of high-purity isomaltooligosaccharides syrup by the enzymatic conversion of transglucosidase and fermentation of yeast cells. Biotechnol. Bioeng. 89, 797
804. Pan, K., Song, Q., Wang, L., Cao, B., 2011. A study of demineralization of whey by nanofiltration membrane. Desalination . Panesar, P.S., Kumari, S., 2011. Lactulose: production, purification and potential applications. Biotechnol. Adv. 29, 940
948. Patel, S., Goyal, A., 2011. Functional oligosaccharides: production, properties and applications. World J. Microbiol. Biotechnol. 27, 1119
1128. Patil, N.V., Janssen, A.E.M., Boom, R.M., 2014. The potential impact of membrane cascading on downstream processing of oligosaccharides. Chem. Eng. Sci. 106, 86
98. Pinelo, M., Jonsson, G., Meyer, A.S., 2009. Membrane technology for purification of enzymatically produced oligosaccharides: molecular and operational features affecting performance. Sep. Purif. Technol. 70, 1
11. Procedikova, K., Curda, L., Misun, D., Dryakova, A., Diblı´kova, L., 2010. Preparation of galacto-oligosaccharides using membrane reactor. J. Food Eng. 99, 479
484. Pruksasri, S., Nguyen, T.-H., Haltrich, D., Novalin, S., 2015. Fractionation of a galactooligosaccharides solution at low and high temperature using nanofiltration. Sep. Purif. Technol. 151, 124
130. Rastall, R.A., 2010. Functional oligosaccharides: application and manufacture. Annu. Rev. Food Sci. Technol. 1, 305
339. Rehman, A.U., Kovacs, Z., Quitmann, H., Ebrahimi, M., Czermak, P., 2016. Enzymatic production of fructooligosaccharides from inexpensive and abundant substrates using membrane bioreactor system. Sep. Sci. Technol. 51 (9), 1537
1545. Ren, H., Fei, J., Shi, X., Zhao, T., Cheng, H., Zhao, N., et al., 2015. Continuous ultrafiltration membrane reactor coupled with nanofiltration for the enzymatic synthesis and purification of galactosyl-oligosaccharides. Sep. Purif. Technol. 144, 70
79. Roberfroid, M.B., 2008. Prebiotics: concept, definition, criteria, methodologies and products. In: Gibson, G., Roberfroid, M.B. (Eds.), Handbook of Prebiotics. CRC Press, Boca Raton, pp. 39
68. Rodriguez-Colinas, B., Fernandez-Arrojo, L., Ballesteros, A.O., Plou, F.J., 2014. Galactooligosaccharides formation during enzymatic hydrolysis of lactose: towards a prebiotic-enriched milk. Food Chem. 145, 388
394. Salehi, F., 2014. Current and future applications for nanofiltration technology in the food processing. Food Bioprod. Process. 92, 161
177. Samanta, A.K., Jayapal, N., Jayaram, C., Roy, S., Kolte, A.P., Senani, S., et al., 2015. Xylooligosaccharides as prebiotics from agricultural by-products: production and applications. Bioact. Carbohydrates Diet. Fibre 5, 62
71. Sangeetha, P.T., Ramesh, M.N., Prapulla, S.G., 2005. Recent trends in the microbial production, analysis and application of Fructooligosaccharides. Trends Food Sci. Technol. 16, 442
457. Santiba´n˜ez, L., Ferna´ndez-Arrojo, L., Guerrero, C., Plou, F.J., Illanes, A., 2016. Removal of lactose in crude galacto-oligosaccharides by β-galactosidase from Kluyveromyces lactis. J. Mol. Catal. B Enzym. 133, 85
91. Satyawali, Y., Vanbroekhoven, K., Dejonghe, W., 2017. Process intensification: the future for enzymatic processes? Biochem. Eng. J. 121, 196
223.

152

Separation of Functional Molecules in Food by Membrane Technology

Sharma, R.R., Agrawal, R., Chellam, S., 2003. Temperature effects on sieving characteristics of thin-film composite nanofiltration membranes: pore size distributions and transport parameters. J. Memb. Sci. 223, 69
87. Silve´rio, S.C., Macedo, E.A., Teixeira, J.A., Rodrigues, L.R., 2015. Perspectives on the biotechnological production and potential applications of lactosucrose: a review. J. Funct. Foods 19, 74
90. Sitanggang, A.B., Drews, A., Kraume, M., 2014. Continuous synthesis of lactulose in an enzymatic membrane reactor reduces lactulose secondary hydrolysis. Bioresour. Technol. 167, 108
115. Sjo¨man, E., Ma¨ntta¨ri, M., Nystro¨m, M., Koivikko, H., Heikkila¨, H., 2007. Separation of xylose from glucose by nanofiltration from concentrated monosaccharide solutions. J. Membr. Sci. 292, 106
115. Spohner, S.C., Czermak, P., 2016. Enzymatic production of prebiotic fructo-oligosteviol glycosides. J. Mol. Catal. B Enzym. 131, 79
84. Stankiewicz, A., 2003. Reactive separations for process intensification: an industrial perspective. Chem. Eng. Process. 42, 137
144. Stoller, M., Bravi, M., Chianese, A., 2013. Threshold flux measurements of a nanofiltration membrane module by critical flux data conversion. Desalination 315, 142
148. Surakka, A., Kajander, K., Rajili´c-Stojanovi´c, M., Karjalainen, H., Hatakka, K., Vapaatalo, H., et al., 2009. Yoghurt containing galactooligosaccharides facilitates defecation among elderly subjects and selectively increases the number of Bifidobacteria. Int. J. Probiotics Prebiotics 4, 65
74. Swati, O., Mishra, S., Chand, S., 2015. Production of isomalto-oligosaccharides by cell bound α-glucosidase of Microbacterium sp. LWT—Food Sci. Technol. 60, 486
494. Swennen, K., Courtin, C.M., Van Der Bruggen, B., Vandecasteele, C., Delcour, J.A., 2005. Ultrafiltration and ethanol precipitation for isolation of arabinoxylooligosaccharides with different structures. Carbohydr. Polym. 62, 283
292. Takiura, K., Honda, S., Endo, T., Kakehi, K., 1972. Studies of oligosaccharides. IX. Synthesis of gentiooligosaccharides by block condensation. Chem. Pharm. Bull. (Tokyo). 20, 438
442. Torres, D.P.M., Gonc¸alves, M., do, P.F., Teixeira, J.A., Rodrigues, L.R., 2010. Galacto-oligosaccharides: production, properties, applications, and significance as prebiotics. Compr. Rev. Food Sci. Food Saf. 9, 438
454. Tsuru, T., Izumi, S., Yoshioka, T., Asaeda, M., 2000. Temperature effect on transport performance by inorganic nanofiltration membranes. AIChE J. 46, 565
574. Van Gestel, T., Kruidhof, H., Blank, D.H.A., Bouwmeester, H.J.M., 2006. ZrO2 and TiO2 membranes for nanofiltration and pervaporation. Part 1. Preparation and characterization of a corrosion-resistant ZrO2 nanofiltration membrane with a MWCO , 300. J. Memb. Sci. 284, 128
136. Vanneste, J., De Ron, S., Vandecruys, S., Soare, S.A., Darvishmanesh, S., Van Der Bruggen, B., 2011. Techno-economic evaluation of membrane cascades relative to simulated moving bed chromatography for the purification of mono- and oligosaccharides. Sep. Purif. Technol. 80, 600
609. Va´zquez, M., Alonso, J., Domı´nguez, H., Parajo´, J., 2000. Xylooligosaccharides: manufacture and applications. Trends Food Sci. Technol. 11, 387
393. Vega-Paulino, R.J., Zu´niga-Hansen, M.E., 2012. Potential application of commercial enzyme preparations for industrial production of short-chain fructooligosaccharides. J. Mol. Catal. B Enzym. 76, 44
51.

Membrane Technology for the Purification Chapter | 4

153

Vegas, R., Luque, S., Alvarez, J.R., Alonso, J.L., Domı´nguez, H., Parajo´, J.C., 2006. Membraneassisted processing of xylooligosaccharide-containing liquors. J. Agric. Food Chem. 54, 5430
5436. Vera, C., Guerrero, C., Illanes, A., 2011. Determination of the transgalactosylation activity of Aspergillus oryzae β-galactosidase: effect of pH, temperature, and galactose and glucose concentrations. Carbohydr. Res. 346, 745
752. Vera, C., Guerrero, C., Conejeros, R., Illanes, A., 2012. Synthesis of galacto-oligosaccharides by β-galactosidase from Aspergillus oryzae using partially dissolved and supersaturated solution of lactose. Enzyme Microb. Technol. 50, 188
194. Vera, C., Co´rdova, A., Aburto, C., Guerrero, C., Sua´rez, S., Illanes, A., 2016. Synthesis and purification of galacto-oligosaccharides: state of the art. World J. Microbiol. Biotechnol. 32. Vincze, I., Vatai, G., 2004. Application of nanofiltration for coffee extract concentration. Desalination 162, 287
294. Yuan, Q.P., Zhang, H., Qian, Z.M., Yang, X.J., 2004. Pilot-plant production of xylooligosaccharides from corncob by steaming, enzymatic hydrolysis and nanofiltration. J. Chem. Technol. Biotechnol. 79, 1073
1079. Yun, J.W., 1996. Fructooligosaccharides-Occurrence, preparation, and application 0229, 107
117. Zhang, Z., Yang, R., Zhang, S., Zhao, H., Hua, X., 2011. Purification of lactulose syrup by using nanofiltration in a diafiltration mode. J. Food Eng. 105, 112
118. Zhao, H., Hua, X., Yang, R., Zhao, L., Zhao, W., Zhang, Z., 2012. Diafiltration process on xylooligosaccharides syrup using nanofiltration and its modelling. Int. J. Food Sci. Technol. 47, 32
39.

FURTHER READING Ebrahimi, M., Placido, L., Engel, L., Ashaghi, K.S., Czermak, P., 2010. A novel ceramic membrane reactor system for the continuous enzymatic synthesis of oligosaccharides. Desalination 250, 1105
1108. Foda, M.I., Lopez-Leiva, M., 2000. Continuous production of oligosaccharides from whey using a membrane reactor. Process Biochem. 35, 581
587.

This page intentionally left blank

Chapter 5

Pectin Removal and Clarification of Juices Sankha Karmakar and Sirshendu De Department of Chemical Engineering, Indian Institute of Technology, Kharagpur, West Bengal, India

Chapter Outline 5.1 Introduction 155 5.2 Various Processing or Preservation Techniques Used for Fruit and Vegetable Juice 157 5.2.1 Depectinization of Fruit Juices 163 5.3 Other Treatment Methods for Fruit Juice Clarification/ Depectinization 163 5.3.1 Membrane 164 5.3.2 Membrane Classification 166 5.3.3 Membrane Modules 166

5.4 Membrane Based Separation Processes 168 5.5 Depectinization and Membrane Based Clarification of Some Typical Fruit and Vegetable Juices 168 5.5.1 Enzymatic Depectinization of Juices 168 5.5.2 Membrane Based Clarification Process for Some Typical Fruit and Vegetable Juice 172 5.6 Conclusion 182 References 184

5.1 INTRODUCTION Fruits and vegetables are natural food products that have been quite popular since early times. In fact, production and preservation of fruits and vegetables are almost as old as human civilization (Skolnik, 1968). They are very susceptible to decay due to high moisture content and are characterized as easily perishable commodities. Preservation of fruits and vegetables dates back to 6000 BC, when people avoided microbial decay by drying and smoking the fruits. Ancient Chinese civilization started preservation of fruits around the first century and Plinius the Elder reportedly preserved white cabbage in earthenware pots in Italy (Derieux, 1988). In the modern era, people prefer to consume ready-made food products due to their busy lifestyle. Hence, fruit juice processing industries have experienced significant growth in the last few decades. Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00005-X © 2019 Elsevier Inc. All rights reserved.

155

156

Separation of Functional Molecules in Food by Membrane Technology

In the juice processing industry, for most of fruits such as orange (Citrus sinensis), apple (Malus pumila), guava (Psidium guajava), litchi (Litchi chinensis), cranberry (Vaccinium oxycoccos), pineapple (Ananas comosus), mango (Mangifera indica), plum (Prunus domestica), mosambi (Citrus limetta), pomegranate (Punica granatum), peach (Prunus persica), banana (Musa acuminata), grape (Vitis vinifera), etc., the fruit pulp is extracted and mashed to create a homogenous mixture. The juice from plants like sugarcane (Saccharum officinarum) is taken out by crushing the plant using a mechanical crusher and in some cases, like coconut (Cocos nucifera), water is directly taken out and consumed. All these fruit juices have their own therapeutic values: G G G G G G

G

G

Orange juice helps in brain development. Apple aids in the growth of red blood cells. Guava boosts immunity. Litchi is very rich in vitamins. Cranberry helps in bladder problems. Pineapple is a rich source of vitamin C, vitamin A, fiber, calcium, potassium, and other essential minerals. Mango contains huge amount of proteins, fibers, vitamins, iron, and other essential minerals. Plum, pomegranate, and mosambi make the digestive system strong; peach improves vision in growing children.

Also, tender coconut is rendered as the most effective sports drink for young athletes (Karmakar and De, 2017). Fruit juices are rich in fibers, polysaccharides, pectin, gum, lignin, cellulose, starch, protein, and many other constituents. The presence of such composition makes the fruit juice very cloudy and viscous. Some of the major factors for consumer acceptability of the processed juice are: 1. juice with higher clarity and reduced viscosity, 2. natural flavor of the juice, 3. long shelf life of the juice. Keeping these things in mind, various separation and preservation methods have been employed to make the processed fruit juice more attractive to the consumers. Maintaining food quality according to the prescribed regulatory board while making the process economically viable is a major challenge in the food processing industry. As stated earlier, one of the key factors for consumer acceptability of a fruit juice is clear appearance of the juice in terms of color and clarity. Several conventional methods are already in use to ensure processed juice with higher clarity, reduced viscosity, and original flavor. Some of the conventional methods to achieve clear juice are centrifugation, enzymatic treatment, or addition of fining agents. But, these processes are generally

Pectin Removal and Clarification of Juices Chapter | 5

157

operated manually and in batch mode, thus, making the process more labor intensive and less economically viable. Moreover, addition of external chemicals in the processed fruit juice is inadvisable since it may leave a slight after taste, thereby, lowering its acceptability. Many fruit juice regulatory boards have approved the addition of sugar and certain natural additives, like ascorbic and citric acid. Sugar acts as natural preservative when added in sufficiently high amount. Excess use of sugar causes diabetes and obesity (Guthrie and Morton, 2000; Ludwig et al., 2001; Malik et al., 2010; Imamura et al., 2015). Also, addition of many coloring agents and artificial flavors to maintain the sensory and aesthetic properties of the fruit juice is a common practice among the fruit juice processing industries. In recent decades, serious awareness has been raised in public due to the carcinogenic nature of additives. Hence, fruit juice with high shelf life without any additive or preservative is creating its market and the challenge lies in its economic production. Shelf life of the processed juice plays a key role in economical viability of the production process, since consumers nowadays are more attracted to natural juice without any additive or preservative. Generally, fruit juices are very susceptible to fouling even under refrigeration due to the presence of many potent sites for bacterial growth. The shelf life of some of the natural juices is presented in Table 5.1.

5.2 VARIOUS PROCESSING OR PRESERVATION TECHNIQUES USED FOR FRUIT AND VEGETABLE JUICE In recent decades, many conventional and nonconventional techniques have been developed to increase the shelf life of the naturally processed juices. As discussed earlier, the market for natural fruit juice without any external additive or preservative is growing and many new technologies have been developed for economic production of natural fruit juice. These technologies aim to produce safe, fresh, and nutritive beverages with good sensory properties in a cost effective method. Various thermal and nonthermal technologies have been developed and adopted to make the method less energy consuming, and have high production yield and low production cost to increase the profit margin. Some common technologies and their advantages and disadvantages are presented in Table 5.2. Apart from these, other conventional nonthermal fruit juice processing techniques include ultrasonic vibration, high-pressure homogenization, cold plasma technique, electroheating, and others (Aneja et al., 2014; Yıkmı¸s, 2016; Misra et al., 2017). Thermal processes like drying, blanching, pasteurization, or sterilization increase the shelf life of the juice at the expense of its sensory properties and nutritional values. Although nonthermal technologies improve the shelf life of the juice, they have other side effects, such as contamination, loss of

158

Separation of Functional Molecules in Food by Membrane Technology

TABLE 5.1 Shelf Life of Some Typical Natural Juices Fruit Name

Shelf Life

References

Natural Juice

Refrigerated Natural Juice (B4 C)

Processed Juice in Refrigeration (B4 C)

Orange

2h

10
16 days

.90 days

Fellers (1988), Polydera et al. (2003)

Apple

2h

14 days

.67 days

Evrendilek et al. (2000), Sua´rezJacobo et al. (2010)

Guava puree

,2 days

10 days

40 days

Yen and Lin (1996), Ninga et al. (2018)

Litchi

2
3 days

15 days

45 days

Kumar et al. (2012)

Pomegranate




,7 days

28
35 days

Varela-Santos et al. (2012), Vegara et al. (2013)

Mosambi




2 weeks

10 weeks

Khandpur and Gogate (2016)

Mango

Few hours

2 weeks

4 weeks

Mkandawire et al. (2016)

Grape

Few hours

7 days

161 days

Siricururatana et al. (2013)

Coconut water

Few hours

2 days

18 weeks

Karmakar and De (2017)

sensory properties, and formation of toxic compounds. Processes like UV-irradiation and ozone processing are not very effective and also the operational cost is huge. The addition of preservatives or additives (natural or synthetic) increases the shelf life of the juice but this scenario is not so attractive to the consumers. Hence, an economic clarification process would be preferable. In this regard, membrane based separation techniques have been proved to be one of the most profitable solutions, as they do not involve the addition of any chemical agents or thermal treatment, while they need low maintenance and unskilled labor, and they maintain the original flavor, aroma, smell, taste, and nutritional parameters of the fruit. The cost analysis comparison for membrane based processes and conventional fruit

TABLE 5.2 Conventional Processes for Fruit Juice Processing Process

Fruit Juice Processed

Advantages

Disadvantages

References

Thermal Process Blanching

Blueberry, strawberry

G G G

Preserve color and texture Tones down the strong taste Stops enzyme action responsible for loss of flavor or color

G

G

G

Thermal drying

Mushrooms, green chilies, tomato, kiwifruit

G

G

Pasteurization and sterilization

Apple, orange, white grape

G

Less storage space than canned or frozen foods No special equipment is required

G

All pathogenic (harmful) bacteria and most nonpathogenic bacteria are killed

G

G

G

Cannot be used for all types of fruits Also often needs to couple with other processes Loss of vitamins

Wrolstad et al. (1980), Rossi et al. (2003)

Loss of sensory properties Loss of nutritive values

Sharma et al. (1995), Maskan (2001)

Loss of sensory properties Loss of naturally occurring nutrients

Mazzotta (2001), Lee and Coates (2003), Fiore et al. (2005)

Nonthermal Process High hydrostatic pressure

Passion, apple, orange, pineapple, cranberry, grape, pomegranate

G

Causes no significant losses of functional compounds in foods

G

Several toxic compounds of abiotic origin can be present or formed in foods processed by pressure processing technologies

Raso et al. (2006), Laboissie`re et al. (2007), Ferrari et al. (2010), Escobedo-Avellaneda et al. (2011), VarelaSantos et al. (2012)

Pulsed electric field

Grapefruit, lemon, orange, tangerine, cranberry, pineapple, apple, tomato

G

Causes no significant losses of functional compounds in foods

G

Contamination of food products due to chemical products from electrolysis Disintegration of food particles due to shock waves

Raso et al. (1998, 2006), Cserhalmi et al. (2006), Aguilar-Rosas et al. (2007)

G

(Continued )

TABLE 5.2 (Continued) Process

Fruit Juice Processed

Advantages

Ultraviolet irradiation

Orange, guava, pineapple, apple

G

Ozone processing

Orange, kiwifruit, blackberry

G

Disadvantages

Does not produce chemical residues, byproducts, or radiation

G

Effective against various kinds of microorganism

G

G

G

G

Use of preservatives

Apple, raspberry, cranberry, grapefruit, mango, kiwi, pineapple peach, apricot, orange, tomato

G

High shelf life

G G

References

Generally effective in surface sterilization Not cost effective

Guerrero-Beltrn and Barbosa-Cnovas (2004), Keyser et al. (2008), Char et al. (2010)

Loss of sensory properties Sometimes produce bromate (a carcinogen) Not cost effective

Karaca and Velioglu (2007), Tiwari et al. (2008, 2009), Barboni et al. (2010)

Altered sensory properties Less attractive to consumers

Beuchat (1982), Komitopoulou et al. (1999), Kabasakalis (2000), Jordan et al. (2001), Shui and Leong (2002)

Pectin Removal and Clarification of Juices Chapter | 5

161

´ lvarez et al. (2000) for juice clarification processes has been conducted by A apple juice clarification. This was performed on the basis of the annual maintenance cost, new profit, gross margin, and payback period. The cost comparison is presented in Fig. 5.1. As evident from this figure, the membrane based process was found to be more profitable with less payback period since the annual maintenance cost was lower than the conventional process, making it a viable economical alternative. However, membrane based processes are limited due to fouling. Because the operation is based on physical separation by size exclusion, formation of a fouling layer over the membrane surface reduces the productivity, thereby limiting its operational time and life. One of the major gel forming agents present in various fruits and vegetables is pectin (Huber and Lee, 1986; Thakur et al., 1997; Zhou et al., 2000; Willats et al., 2006; Sagu et al., 2014; Ninga et al., 2018). Pectin is found in all types of plants. It is a class of heteropolysaccharide found primarily in the cell walls of the plants (Mualikrishna and Tharanathan, 1994). It was first isolated in 1825 by a French scientist and pharmacist, Henri Braconnot (Keppler et al., 2006). Pectin is predominantly found in citrus fruits like apple, guava, oranges, pears, plums, peach, mosambi (Zhou et al., 2000; Rai et al., 2004; Willats et al., 2006; Ninga et al., 2018). Apart from these, pectin content is also high in banana (Vı´quez et al., 2007; Sagu et al., 2014). Pectins are a group of pectic polysaccharides rich in galacturonic acid (Prade et al., 1999; Ninga et al., 2018). Among the several polysaccharides, homogalacturonas are linear chains of α-(1
4)-linked D-galacturonic acids (Ridley et al., 2001; Golovchenko et al., 2002). D-Xylose or D-apiose in case 5.0

Conventional method Membrane based separation technique

4.5 4.0 3.5

4.1

3.716 3.432 3.0

3.0 2.5 2.0 1.5

1.365 1.175

1.0 0.5 0.0

AMC Net profit (million Euro/year) (million Euro/year)

0.32 0.37 Gross margin

Payback period (years)

FIGURE 5.1 Cost comparison of membrane based process and conventional method for apple juice processing.

162

Separation of Functional Molecules in Food by Membrane Technology

of xylogalacturonan and apiogalacturonan, respectively, can be present as substituted galacturonans (Schols et al., 1995). Many neutral sugars like D-galactose, L-arabinose, and D-xylose branch off from Rhamnogalacturonan I pectins (RG-I). RG-I pectin contains a backbone of repeating disaccharide, 4-α-D-galacturonic acid-(1,2)-α-L-rhamnose-1. Rhamnogalacturonan II pectins (RG-II) are a less frequently occurring pectin that are primarily made of D-galacturonic acid units (Willats et al., 2001; Deng et al., 2006). Pectins have molecular weights in the range of 30,000
130,000 Da, depending on the source and extraction conditions. They can be classified into high or low ester pectins depending upon the amount of carboxyl groups present in galacturonic acid that can be esterified by methanol. The nonesterified galacturonic acid can be classified into: 1. Pectinates: salts of partially esterified pectin. 2. Pectates: if the degree of esterification is below 5%. 3. Pectic acid: it is the insoluble acid form. Acetylated galacturonic acid is present in some plants, such as potatoes and pears, along with pectin. Acetylation increases the stability and emulsifying effect of pectin while preventing gel formation. In some plants, amidated pectin, a modified form of pectin, is present. For amidated pectins, galacturonic acid is converted to carboxylic acid amide with the help of ammonia. These types of pectins are more resistant to fluctuating calcium concentration while forming gels. Pectin is a typical gelling agent that easily forms gel in the presence of Ca21 ions at low pH (Thakur et al., 1997). This property has found huge application in the fruit industry, typically in the preparation of jam and jelly. Depending upon the type of pectin and the gelling conditions, gels can be categorized as hard and soft gels. If the gel formation is too strong it results in a granular texture, whereas, if the gel is too soft it makes a weak gel. In case of high ester pectins (more than 60% solid content with pH 2.8
3.6), the individual chains are bound by hydrogen bonds and hydrophobic interactions. Sugar present in pectin binds water and compels the pectin strands to attach to each other resulting in a three-dimensional macromolecular gel. It is also known as low water activity gel or sugar acid pectin gel. In case of pectin with low ester content, the gelling mechanism is due to the ionic bridges formed between calcium ions and the ionized carboxyl groups of the galacturonic acid. This type of gel can form in pH range 2.6
7 and at a low soluble solid content (10%
70%). Amidated pectins form reversible gels that are dependent on temperature. Also, it needs less calcium ions to form gels. The presence of pectin is undesirable in fruit juice due to following reasons: G

pectin interacts with protein in the juice forming a complex that acts as a potent source of microbial growth, thereby spoiling the juice fast;

Pectin Removal and Clarification of Juices Chapter | 5 G

G

163

pectin makes the juice viscous, thereby affecting flowability of juice making the processing difficult; pectin, being a natural gel forming agent, forms a thick cake type of layer, thereby, affecting the filtration performance.

Thus, complete depectinization of fruit juice is essential in view of its long shelf life.

5.2.1 Depectinization of Fruit Juices Pectin content is pretty high in most of the citrus fruits, guava, banana, and others (Rai et al., 2004; Sagu et al., 2014; Ninga et al., 2018). Conventionally, enzymes like pectinase and amylase are used to depectinize the fruit juices (Dingle et al., 1953; Alkorta et al., 1998; Prade et al., 1999; Hoondal et al., 2002; Gummadi and Panda, 2003; Jayani et al., 2005; Pedrolli et al., 2009; Visser and Voragen, 2009; Sagu et al., 2014; Ninga et al., 2018). Pectinase is usually obtained from microbial sources (Hoondal et al., 2002; Gummadi and Panda, 2003; Jayani et al., 2005; Pedrolli et al., 2009) and it is also used industrially for depectinization (Alkorta et al., 1998). Details of enzymatic depectinization are presented in subsequent sections.

5.3 OTHER TREATMENT METHODS FOR FRUIT JUICE CLARIFICATION/DEPECTINIZATION Fruit juices contain many constituents that affect the subsequent filtration efficiency. Suspended solids, proteins, and fibers are also present in fruit juice along with pectin. As already discussed, pectin concentration can be reduced significantly by treatment with enzymes. However, other pretreatment methods are also employed to devoid the feed of any solutes that might hamper the filtration process. The suspended solids generally contain insoluble pectinous solids, cellulose, hemicelluloses, and lignins (Kirk et al., 1983) and these constituents also affect the filtration performance (Girard and Fukumoto, 2000; Sagu et al., 2014). In most of the cases, centrifugation at high speed is employed as a suitable pretreatment method (Kratchanova et al., 2004; Rai et al., 2007; Chen et al., 2008; Chhaya et al., 2013; Domingues et al., 2014; Sagu et al., 2014; Biswas et al., 2016). Some fruit products that contain low suspended solids (like coconut water) can be pretreated using a fine food grade mesh (Karmakar and De, 2017). However, juice treatment with commercial pectinase enzyme is much more costly and it accounts for almost 30% of the total processing cost (Porter, 1990). Although the centrifugation cost is less than that of pectinase treatment (Rai et al., 2006), in most cases it is coupled with pectinase treatment to obtain the pretreated juice (Sagu et al., 2014). Treatment by fining agents, like

164

Separation of Functional Molecules in Food by Membrane Technology

bentonite (acts on pectin) and gelatin (acts on protein) and their combination, is also common for pectin and protein removal (Girard and Fukumoto, 2000). However, fining treatment causes addition of external chemicals that need to be removed by coarse filtration followed by fine filtration. These additional steps make the processing more complex and expensive. Membrane based processes can be an attractive alternative in this regard. Membrane based processes offer a wide range of separation processes starting from pretreatment to isolation of any specific bioactive agent from the fruit juice. Because these processes increase the shelf life of the juice, keeping the sensory and nutritional profile of the juice intact, they are revolutionizing the fruit juice processing industry. The processed juice, devoid of any external chemical agents, is also very appealing to the consumers.

5.3.1 Membrane Membrane is an interface separating two phases, restricting the transport of various species through it. Fruit juice clarification using membranes is a physical, rate-governed separation process based solely on size exclusion. Various grades of membranes (based on pore size) can be used for pretreatment, purification, extraction, and also typical ultrafiltration or nanofiltration membranes can be used to fractionate a specific bioactive agent. Porous membranes can be classified into two broad categories based on their structure, namely, symmetric and asymmetric membrane (Ladewig and Al-Shaeli, 2017). Symmetric membranes have uniform structure throughout the membrane cross-section. The thickness of the membrane governs the resistance to mass transfer thereby controlling the permeation rate. Asymmetric membranes have a 0.1
5-μm-thin skin layer at the top followed by a highly porous structure. This thin skin serves as the primary selective barrier for solute transport and the membrane characteristics are defined particularly by the property of the skin layer. In most of the pressure-driven membrane based operations, such as nanofiltration, ultrafiltration, and reverse osmosis (RO), asymmetric membranes are employed to achieve high throughput. Asymmetric membranes can be prepared by a simple phase inversion process leading to integral structure (Kesting, 1971). A schematic of symmetric and asymmetric porous membrane is presented in Fig. 5.2. (A)

(B)

Thick dense layer

FIGURE 5.2 Representative diagram of (A) symmetric and (B) asymmetric membrane.

Pectin Removal and Clarification of Juices Chapter | 5

165

Based on the material of construction, membranes can be organic, inorganic, or a blend of both. Organic membranes are prepared from different organic polymers, like polysulfone (PSF), polyacrylonitrile (PAN), polyethylene (PE), etc. Organic membranes are more sensitive to heat and less stable at high temperature. Also, organic membranes are susceptible to mechanical damage and are very sensitive to harsh operating conditions. However, organic membranes can be easily tuned to obtain varying pore size ranging from angstrom to microns. These membranes can be backwashed easily to recover permeability. Hence, they find wide application in various fields of wastewater treatment, fruit juice clarification, purification of drinking water, desalination, and many more. Inorganic membranes are usually made from aluminum oxide (Al2O3), zirconium oxide (ZrO2), zeolites, porous glass, molybdenum, and stainless steel. Inorganic membranes have long-term stability at high temperature, are resistant to harsh chemical environments, and are resistant to high pressure drops. However, they are expensive and it is very difficult to prepare inorganic membranes with submicron pore size, thereby limiting their application in separation of fine solutes of low molecular weights. Another class of polymeric membranes is mixed matrix membranes. These are polymeric membranes incorporated with inorganic additives. Using these membranes, smaller sized molecules, such as fluoride, arsenic, phenol, etc., are removed by specific adsorption by inorganic compounds and at the same time, larger sized suspended solids and bacteria can be removed by size exclusion at higher throughput at lower pressure (Karmakar et al., 2017, 2018).

5.3.1.1 Membrane Based Processes: Advantages, Disadvantages, and Challenges There are several advantages of membrane based process including: 1. Operation is carried out at room temperature, thereby requiring less energy. 2. Filtration is conducted at room temperature, so heat sensitive solutes, like proteins, fruit juices, etc., can be easily processed. 3. Membrane setups are very simple in design and do not require any complex controls or auxiliary equipment, making them less labor intensive, easy to operate, and require less maintenance. 4. Membrane based separations are mostly physical in nature, hence, they do not produce any harmful byproducts. 5. Membrane based processes are easily scalable for industrial applications. Disadvantages of membrane based processes: 1. Membrane based processes are often coupled with various pretreatment processes, such as depectinization, centrifugation, cloth filtration, etc., to reduce the membrane load, thereby increasing the treatment cost.

166

Separation of Functional Molecules in Food by Membrane Technology

2. Major limitation of these processes is membrane fouling, hampering performance, and life of membrane. 3. Designing and implementing a membrane based system to handle high throughput at high pressure is often very complex and expensive. The major challenge for application of membranes in the fruit juice processing industry lies in tackling the membrane fouling, scaling up of the system for better productivity, and making the process economical with respect to commercial processing techniques.

5.3.2 Membrane Classification Based on pore size, molecular weight of solutes to be separated, and operating pressure, membranes can be categorized into four major classes, as presented in Table 5.3. Pervaporation (PV), osmotic dehydration, and membrane distillation (MD) are some other membrane based methods employed for fruit juice clarification. Some typical grades of membrane obtained from different base polymers are: G

G

G

G

Microfiltration: Obtained from cellulose acetate (CA), cellulose nitrate (CN), cellulose triacetate (CTA), blend of CA and CTA, polyvinyl chloride (PVC), polyvinyl alcohol (PVA), PAN, PAN-PVC copolymer, polyamide-aromatic, polycarbonate (PC), polyamide (PA), polypropylene (PP), PSF, polytetrafluoroethylene (PTFE), PTFE. Ultrafiltration: CA, CTA, PVA, PAN, PAN-PVC copolymer, polyamidearomatic, PA, PSF, polyvinylidene fluoride (PVDF), cellulose acetate phthalate (CAP). Nanofiltration: PVDF, polyethersulfone (PES), and blend membranes with different organic and inorganic additives. Reverse osmosis: CA, CTA, polyamide-aromatic, PA, PES.

5.3.3 Membrane Modules The equipment in which membranes are housed is known as the membrane module. The membrane modules are designed to have maximum filtration area in a relatively smaller volume for a compact design. Various configurations of membrane modules are used to achieve maximum productivity (flux). According to the configuration, membrane modules can be classified into four types: 1. 2. 3. 4.

plate and frame module spiral wound module tubular modules hollow fiber modules

TABLE 5.3 Comparison of Various Grades of Membranes Membrane Type

Pore Size (A˚)

MWCO (kDa)

Particle size of the Solutes Rejected (μm)

Operating Pressure (atm)

Characteristic of Permeate

Characteristic of Retentate

Applications

Microfiltration (MF)

.1000

100
500

0.1
10

2
4

Solvent (water) 1 dissolved solutes

Large suspended particles, bacteria, clay, emulsion

Filtration of clay solution, latex, paint, pretreatment step in fruit juice processing

Ultrafiltration (UF)

20
1000

1
100

0.001
0.1

6
8

Water 1 ultra fine solutes dissolved in water

Mostly organic compound, macromolecules, proteins, viruses, bacteria, enzymes, polysaccharides and colloids

Filtration of protein, red blood cells and wide application in fruit juice processing

Nanofiltration (NF)

5
20

0.2
1

0.0001
0.001

15
25

Water 1 nanomolecules of solute, glucose (if present)

High molecular weight compound, polyvalent anions, oligosaccharides

Filtration of dyes, small molecular weight organics and concentrating different clarified fruit juices

Reverse osmosis (RO)

2
10

,0.1

0.0001

.25

Water

Ions and most organic compounds, amino acids

Desalination

168

Separation of Functional Molecules in Food by Membrane Technology

Details of each module type are available in the literature (Morigami et al., 2001; Gu et al., 2011; Wang et al., 2011; Thakur and De, 2012).

5.4 MEMBRANE BASED SEPARATION PROCESSES Herein, a cold sterilization technique that produces fruit juices with longer shelf life is desired. The technology should be economical, having easy maintenance, as well as yielding processed fruit juice at superior quality and quantity. Membrane based separation processes (MSP) have emerged and gained popularity in the last decade to address this issue. Different grades of the membrane can be employed for fruit juice clarification starting from pretreatment (MF) to concentrate the clarified juice (RO). Size exclusion mechanism of the membranes is employed to retain suspended solids, colloids, lipids, high molecular weight proteins, and microorganisms. Smaller solutes, like salts, polyphenols, minerals, sugar, and vitamins are allowed to permeate, producing clarified juice with longer shelf life after being coupled with aseptic packaging along with full nutritional content. From the preceding sections, it can be inferred that MSPs have an edge over other conventional clarification technologies for fruit and vegetable juices. Several advantages of MSP can be exploited to obtain processed juice having similar nutritional profile and sensory properties as the natural one, having long shelf life. Since no external chemicals or preservatives are added in case of MSP, the processed juice is far more attractive to consumers. MSP not only improves the shelf life of the juice, but it also enhances its physicochemical characteristics in terms of clarity, total solid content, and protein concentration. Moreover, different configurations of the membrane separation modules can be utilized to improve product quality and quantity in a very economical fashion. NF and RO based membrane processes are often used to concentrate the juice, which is less energy intensive than traditional processes like multiple effect evaporation. Thus, MSP offers a better and greener alternative to the conventional processes and has huge potential in the sector of fruit and vegetable juice processing.

5.5 DEPECTINIZATION AND MEMBRANE BASED CLARIFICATION OF SOME TYPICAL FRUIT AND VEGETABLE JUICES 5.5.1 Enzymatic Depectinization of Juices As discussed earlier, pectin is present in almost all fruits. In citrus fruits, apple, pear, guava, banana, etc., it is more significant than in the others. The presence of negatively charged rhamnogalacturonans and neutral arabinogalactans imparts viscosity to the juices, but they also form haze in

Pectin Removal and Clarification of Juices Chapter | 5

169

combination with different proteins. Haze formation not only retards the maximum recovery of the juice (Whitaker, 1984) but also acts as a potent site for microbial growth spoiling the juice. Due to its sticky nature and gel forming characteristics, many treatment methods have been employed to remove pectin from various fruit juices. One of the most widely used treatment methods for pectin removal is enzyme treatment. Enzymes like pectin methylesterase, pectate lyase, polygalacturonase, etc., obtained from Aspergillus niger can hydrolyze the pectin present in the juice (Whitaker, 1984). Tangerine juice had been depectinized at constant temperature using pectinase by Chamchong et al. (1991) before subjecting it to subsequent microfiltration (MF) and ultrafiltration (UF) process. Chang et al. (1995) used five different commercial pectinase enzymes to treat Stanley plum juice. They varied the enzyme concentration keeping the temperature and the time of operation constant. It was noted that the taste turns bitter with enzyme concentration. Banana puree has been depectinized by commercial pectinase along with other commercial enzymes, like pectinol and pectinol D by Viquez et al. (2007) at varying enzyme concentration at 45 C for 2 hours. Similar work was reported by Sagu et al. (2014). Another citrus fruit, mosambi, was subjected to primary treatment by pectinase by Rai et al. (2004). The optimum condition was selected based on minimum viscosity and alcohol insoluble solids (AIS) and maximum clarity. Ninga et al. (2018) had depectinized guava juice using commercial pectinase enzyme obtained from Aspergillus niger. The optimizing criteria for enzyme treatment are enzyme concentration, time, and temperature. For their study, the optimized time, temperature, and enzyme concentration were 45 minutes, 45 C, and 0.078% w/w, respectively. The hydrolysis of pectinous matter was also modeled using the Hill equation: v5

Vmax ½Sn K 1 ½Sn

ð5:1Þ

where Vmax and n refer to the maximum reaction rate and Hill coefficient, respectively. K is expressed as K 5 [S50]n (S50 is the substrate concentration for which the reaction rate is half of its maximum value). The value of n decreased from 7 to 4 with enzyme concentration. This is due to decrease in the binding sites of the enzyme. The enzymatic treatment increased the reducing sugar content by twofold and decreased the viscosity significantly. The developed depectinization kinetics provided insights about the molecular mechanism and helped in scaling up the system. Depectinization of apple juice was carried out by Alvarez et al. (1998) using commercial pectinase enzyme, Pectinex-3XL (pectin esterase, polygalacturonase, pectate lyase, pectin lyase, and hemicellulases). The process was carried out in a batch reactor with varying enzyme concentration and fixed time and temperature, 120 minutes and 55 C, respectively. It was found that 300 mg/L of pectinase dose was optimum for the process. The viscosity was reduced drastically (by 81%). Around 80%

170

Separation of Functional Molecules in Food by Membrane Technology

of reduction in pectin and turbidity was attained. Also, due to depectinization, flux of the subsequent membrane based process increased significantly. Influence of depectinization on fruit juices like West Indian cherry (Malpighia glabra) and pineapple was studied by Barrosi, Mendes, and Peres (2004). Citrozym Ultra-L, a mixture of pectinase, hemicellulase, and cellulase was used for enzyme treatment that was carried out in a batch reactor at 40 C for 90 minutes. At an enzyme concentration of 300 mg/L, around 82% and 26% of reduction in galacturonic acid concentration was obtained from West Indian cherry puree and pineapple juice, respectively. However, enzyme concentration of 20 mg/L, operated at 40 C for 60 minutes, was considered to be the optimum pretreatment step prior to UF. Mosambi juice was depectinized using commercial pectinase enzyme obtained from Aspergillus niger by Rai et al. (2004). Optimization of time, temperature, and concentration of enzyme was carried out by response surface methodology (RSM) based on apparent viscosity, AIS, and clarity. The optimized operating conditions were obtained as 99.27 minutes, 42 C, and 0.0004% w/v, for time, temperature, and enzyme concentration, respectively. Banana contains high amount of pectin making the juice viscous and turbid. Sagu et al. (2014) developed optimum low temperature extraction process of banana juice using commercial pectinase enzyme. The operating conditions were optimized using RSM and it was found to be 33 C temperature, 108 minutes of treatment at enzyme concentration of 0.03% v/w. The physicochemical parameters obtained at the optimum conditions were, viscosity: 1.42 6 0.04 mPa; clarity: 55.5 6 5%T; AIS: 0.52 6 0.02% w/w; total polyphenol: 14.8 6 0.4 mg GAE/100 mL, and protein concentration: 1341 6 71 mg/L. The extract also contained high amount of potassium and sugar. A first order kinetics of degradation of the substrate was proposed based on the amount of AIS: dC 5 2 kC dt

ð5:2Þ

where C is the concentration of AIS at any time point t with kinetic constant k. It was observed that the kinetic constant varied linearly with substrate concentration. Apart from this, depectinization is common in many other fruit juices, like grape, orange, grapefruit, tangerine, lemon, etc. Depectinization study of some typical fruit juices is summarized in Table 5.4. Depectinization not only reduces the viscosity and arrests haze formation, but it also increases the productivity of the subsequent membrane based processes significantly (Liew Abdullah et al., 2007; Pap et al., 2009; Domingues et al., 2014). These studies confirm the improvement in clarity, reduction in viscosity, and lowered pectin concentration using commercial pectinase enzyme. Hence, pectinase treatment is a viable pretreatment method before subjecting the fruit juice to subsequent membrane based operations. However, to remove the left over pectin, a suitably selected UF membrane has to be used to obtain a pectin free clarified juice.

TABLE 5.4 Depectinization Study of Some Typical Fruit Juice Fruit

Enzyme Used

Time (min)

Temperature ( C)

Enzyme Concentration

Properties of the Clarified Juice

References

Sour cherry juice (Prunes cerasm)

Pectinex-3XL

120

50




Reduced viscosity

S¸ ahin and Bayindirli (1993)

Black carrots (Daucus carota)

Panzym P5

120

50




Increase in shelf life

¨ zkan and Kırca, O Cemeroglu (2006)

Apple, grape, and lemon

Pectinex 3XL

90

50




Increase in shelf life

¨ zkan and Kırca, O Cemeroglu (2006)

Apple, pear, grape

Pectinmethylesterase, polygalacturonases, pectinlyase










Clear juice

Grassin and Fauquembergue (1996)

Cranberry

Crystalzyme Cran

60

45

0.12% w/v

Negative alcohol precipitation test

White, Howard and Prior (2011)

0.004% v/v

Increase in flux in subsequent MSP

Pap et al. (2009)

Blackcurrant (Ribes nigrum)

Panzym Super E and

720

25

Trenolin Rot DF

1440

25

Starfruit (Averrhoa carambola)

Pectinex Ultra SP-L

20

30

0.10% v/v

Reduction in viscosity, effectively increasing flux in subsequent MSP

Liew Abdullah et al. (2007)

Peach

Pectinex AFPL3 and Ultra SP WOP

60

25

240 mg/L

Reduction in pulp content and viscosity

Santin et al. (2008)

Passionfruit (Passiflora edulis f. flavicarpa)

Pectinex 3 XL

90

44

1 mL/L

Reduction in viscosity, effectively increasing flux in subsequent MSP

Domingues et al. (2014)

Orange

Pectinex Ultra SPL

240

Room temperature

20 mg/L

Reduction in viscosity, effectively increasing flux in subsequent MSP

Qaid et al. (2017)

172

Separation of Functional Molecules in Food by Membrane Technology

5.5.2 Membrane Based Clarification Process for Some Typical Fruit and Vegetable Juice Fruits and vegetables are excellent source of nutrition and can be consumed in its raw form. The juice form of the fruits is preferred all around the world because it is easy to consume, easily digested, and the nutrients are easily assimilated in the body. Moreover, since many fruits are not available yearround, juice with long shelf life is preferred.

5.5.2.1 Apple Juice Apple (Malus pumila) is a well-known fruit, popular all over the world due to its health benefits. The plant, mainly originating from central Asia, has spread all over the world. The whole fruit is edible including the outer skin. Apple contains high amount of vitamins like vitamin C and vitamin B12, minerals like calcium and phosphorous, and a rich source of carbohydrate. Apple juice in various forms is popular among the masses and many studies have been conducted to improve the shelf life and nutritional parameters of this juice. Some of those studies are presented herein. An integrated membrane process including UF, MD, and osmotic distillation (OD) had been employed by Onsekizoglu et al. (2010) for clarification of apple juice. The feed apple juice containing total soluble sugar (TSS) of 12oBrix was clarified by a series of UF membrane (100 and 10 kDa) and % concentrated using an integrated MD and OD process. TSS of the final product was 65oBrix. Their physicochemical properties like, aroma, taste, texture, % color, and nutritional properties like, phenolic compounds, organic acids, and sugar were recovered after the integrated MSP. Moreover, formation of 5-hydroxymethylfurfural was significantly arrested in the case of the combined OD and MD process. Koza´k et al. (2006) also developed an integrated membrane process for the clarification of apple juice. The raw juice was initially subjected to MF followed by concentration by RO method. The concentrated juice was then subjected to RO-OD and RO-MD process using hydrophobic porous hollow fiber module. The feed juice had a TSS of 17.5oBrix, which was concentrated to 62 and 71oBrix, for RO-OD and RO-MD%process, respectively. The % physicochemical property of the final juice was also comparable with the raw juice in terms of aroma and flavor. Reverse osmosis is the most common MSP for concentrating fruit juices. Al-Obaidi et al. (2017) had developed a multistage RO using spiral wound modules. A series configuration of 12 elements of 1.03 m2 area was optimized for this multistage process. The product was concentrated to about 142% compared with the feed with a final TSS of 25.5oBrix. Moreover, the % study also revealed that the transmembrane pressure (TMP) and cross-flow rate play a major role in percentage rejection of sugar.

Pectin Removal and Clarification of Juices Chapter | 5

173

UF is commonly used in MSP for secondary clarification of fruit juices and it serves as a posttreatment step. Gulec et al. (2017) used three polymeric UF membrane of molecular weight cutoff of 100, 50, and 30 kDa, respectively, for apple juice clarification. It was observed that 30 kDa membrane showed better recovery of hydraulic permeability compared with others. Also, the flux decline in case of 30 kDa membrane was quite less compared with the other two. Ninety percent TSS was recovered (9.7oBrix % of in permeate compared with 11.3oBrix in raw juice). The pore size % 30 kDa, being smaller in size, promoted cake filtration, which was removed easily. Often the MSP is coupled with enzymatic pretreatment of the juice. Alvarez et al. (1998) had developed a UF process for depectinized apple juice using a 15 kDa tubular module with an area of 0.023 m2. The process was carried out at TMP of 0.4 MPa. It was observed that the highest flux was obtained for the sample, which was depectinized using an enzyme concentration of 300 FDU/g pectin, where FDU is the ferment depectinization unit. Aguiar et al. (2012) had developed a coupled RO and osmotic evaporation (OE) method to treat depectinized apple juice. Initial pretreatment was carried out by a 0.3-μm MF membrane. RO and OE were carried out by plate and frame module and polymeric flat sheet module, respectively. The pretreated juice was successfully concentrated using both RO and OE yielding total solid concentration of 29 and 53 g/100 g, respectively. Hence, membrane based processes provide a viable alternative for clarification, treatment, sterilization, and concentration of the final product. Onsekizoglu et al. (2010) had shown a fivefold increase in the TSS removal using coupled membrane based processes while arresting the formation of undesired 5-hydroxymethylfurfural. Similar work has been demonstrated by Koza´k et al. (2006) where the TSS was concentrated fourfold compared with the value of the initial one while keeping the sensory profiles intact. Reverse osmosis process was also employed for apple juice concentration and TSS enhancement. Besides, UF membranes have been used as post or tertiary treatment methods to enhance the texture and smoothness of the juice. The polymeric membranes selected by Gulec et al. (2017) were composed of PSF, PES, and cellulose. The membranes with higher pore size and higher hydrophobicity experienced higher resistance due to reversible fouling whereas for the 30 kDa membrane cake filtration was the dominant mechanism, thereby yielding higher flux than more open pore size membranes. Moreover, the essential nutrients like TSS, salts, polyphenol, antioxidant fractions, amino acids, and others were easily permeated through the UF membrane pores due to their small size. As evident from their study, Gulec et al. (2017) had shown that there is only 14% decrease in the TSS after filtering through UF. Many pectinous fruits like apple, banana, guava, etc., tend to foul the membranes more due to their high pectin content. As

174

Separation of Functional Molecules in Food by Membrane Technology

discussed earlier, pectin is a gel forming material that tends to develop a cake type layer over the membrane surface and thus imparting additional resistance to the permeate flow. It is a common practice nowadays to deploy depectinization methods to reduce the pectin content prior to membrane based process. Alvarez et al. (1998) had demonstrated filtration of depectinized apple juice using a 15 kDa membrane. Giovanelli and Ravasini (1993) processed laccase treatment of the apple juice to reduce its phenolic content and decolorize it subsequently by passing it through polyvinylpyrrolidone and activated charcoal. The processed apple juice was then passed through a cross-flow MF membrane, which stabilized the juice. Similar treatment had been adopted by Aguiar et al. (2012), where MF was followed by coupled reverse osmosis and OE to concentrate the juice. A comparative analysis between MF and UF of apple juice was carried out by Wu et al. (1990). The study concluded that based on the sensory properties MF was preferred rather than UF although the apple juice was more stable after UF. Mondor et al. (2000) had shown that in a batch process apple juice can be clarified using dead end ceramic filters with pore size ranging from 0.02 to 0.2 μm and the process throughput can be correlated to the volume concentration factor. For apple juice depectinization sometimes additives are used to reduce fouling enhancing the performance of ensuing membrane based processes. All these help in upscaling membrane based systems to industrial level in place of conventional treatment processes.

5.5.2.2 Orange Juice Orange (Citrus sinensis) is a citrus fruit and a hybrid of pomelo and mandarin orange. Relative ratio of sugar and acid determines the taste of the orange and the aroma is governed by volatile organic compounds like alcohols, ketones, terpenes, esters, etc. Orange juice is pulpy in nature with a pH of 2.9
4. Orange is a rich source of vitamin C along with potassium, carotenoids, flavonoids, and other phytoactive compounds. However, like most of the citrus juices orange contains a huge amount of pectin (Qaid et al., 2017). Hence, orange being a highly pectinous fruit is usually depectinized prior to MSP. A study by Mizrahi and Berk (1970) showed that orange juice exhibits non-Newtonian flow behavior due to the presence of pectin and suspended solids. Depectinized orange juice exhibits Newtonian flow-like (water like property and reduced viscosity) behavior. The following are some studies where depectinized orange juice was further treated with MSP. Destani et al. (2013) had clarified depectinized blood orange juice using 100 kDa cutoff UF membrane. Prior to clarification, the raw juice was depectinized for 4 hours at room temperature using 20 mg/L Pectinex Ultra SPL. TSS of the ultrafiltered juice had increased significantly (10.5
61.4oBrix). The UF membrane showed nominal rejection for phenolic %

Pectin Removal and Clarification of Juices Chapter | 5

175

compounds. Following UF, the juice was evaporated by OD process but the phenolic compounds were significantly preserved. Blood orange juice was clarified using UF membrane by Conidi et al. (2015). The raw juice was depectinized using 1% w/w commercial pectinase enzyme (Pectinex Ultra SP-L) at room temperature for 4 hours. The resultant depectinized juice was subjected to UF using polymeric hollow fiber membranes of 100 and 50 kDa. Depending upon the physicochemical properties, nutritional quality, and the productivity of the membrane, 100 kDa PSF based UF membrane was found to be the most suitable. Moroccan Valencia orange juice was clarified by Qaid et al. (2017). Twenty mg/L of commercial pectinase enzyme (Pectinex Ultra SP-L) was used to depectinize the raw juice at room temperature and the process was carried out for 4 hours. Two UF membranes having molecular weight cutoff 30 and 20 kDa were used. Thirty kDa PES membrane was found to be the most suitable in terms of selectivity and productivity. PES membrane removed all the suspended solids and pectin while exhibiting a steady state flux of 27.43 l/m2.h. However, due to presence of pectinous material the flux decline was around 61%. Orange also contains huge amounts of anthocyanins, which are responsible for their red color and are widely used as food colorants. However, anthocyanins are highly reactive and form undesirable colorless or brown colored compounds. Many studies have been carried out to stabilize the anthocyanins and observe their properties dependent on temperature (Jackman et al., 1987; Mazza and Brouillard, 1987; Fabroni et al., 2016). Anthocyanins have also been recovered and concentrated using membrane based technology by various researchers (Gilewiczlukasik et al., 2007; Martı´n et al., 2018). The molecular structures as well as the intra- and intermolecular interaction of anthocyanins dictate their stability and coloring attributes. Anthocyanins are significantly affected by change in environmental conditions, for example, pH, temperature, chemical composition, etc. Hence, a safe, nonthermal technology has been required for the successful concentration of anthocyanins. Membrane based processes, being nonthermal, provide several advantages for the recovery of delicate nutritional and bioactive agents; thus they are widely used to recapture anthocyanins from pomegranate, grape, radish, raspberry, cranberry, blackberry, potato, orange, etc. Blood orange juice was subjected to UF using a 15-kDa tubular PVDF membrane and the resultant permeate was analyzed in terms of TSS, antioxidant activity, anthocyanins, and flavonoids (Cassano et al., 2007). Permeate and feed streams were almost identical in properties except for insoluble solids, attributed to high pectin content. More than 90% of the anthocyanins and around 94% TSS were recovered in permeate and the insoluble solids rejected completely. Membrane fouling was attributed to partial and complete pore blocking. Similar work has been carried out by Galaverna et al. (2008). In their work, an MSP was employed as an alternative to thermal

176

Separation of Functional Molecules in Food by Membrane Technology

evaporation to concentrate blood orange juice. Initially the orange juice was subjected to UF using a 15-kDa membrane. The ultrafiltered juice was then concentrated in two stages, using RO followed by OD. The recovery of anthocyanin and antioxidant fraction was more than that of thermal treatment. Therefore, membrane based processes can be aptly utilized for the clarification of orange juice and recovery of the nutritional parameters. However, due to the high pectin content of the juices it is preferable to be coupled with enzymatic treatment. As discussed above, depectinized orange juice yields higher productivity than raw juice. In most cases, MF is used as a pretreatment step before applying a suitable cutoff UF membrane. The processed juice can be concentrated using RO or OD to enhance its nutritional properties.

5.5.2.3 Pineapple Juice Pineapple (Ananas comosus) is a tropical fruit and it is the most economically significant plant in the Bromeliaceae family. The largest producer of pineapple in the world is Costa Rica, closely followed by countries like, Brazil, Philippines, Indonesia, and India. It is a rich source of vitamins, primarily vitamin C and vitamin B6. Pineapple juice also contains a huge array of bioactive compounds, like gallic acid, tyrosine, genistin, chlorogenic acid, epicatechin, chavicol, and many others (Sopie Edwige Salome et al., 2011). Some studies for combined depectinization and subsequent membrane based clarification are presented herein. Barrosi, Mendes and Peres (2004) clarified wild cherry and pineapple juice using Citrozym Ultra-L enzyme. The optimum enzyme concentration for pineapple juice was found to be 20 mg/L, at 40 C for 60 minutes. The resulted juice was then further processed using a ceramic tubular membrane and a PSF hollow fiber membrane of 100 kDa MWCO. It was observed that the recovery of ascorbic acid was much higher in the case of PSF membrane. Also, depectinization led to higher clarity and lower viscosity in case of pineapple and hence, the productivity had increased manifold in case of both hollow fiber and tubular membranes. A combined cold sterilization process coupled with clarification by MF membrane was developed by Carneiro et al. (2002). The raw juice was depectinized using a 0.03%v/v connotation of Pectinex SP-L and Celuclast 1.5-L. Depectinization was carried out for 60 minutes at 30 C. The enzymatic treatment reduced the viscosity drastically and the pulpy nature of the juice, yielding higher productivity in the subsequent MF process. PES based MF membrane, with pore size 0.3 μm, was used for further clarification of the depectinized juice. MF was carried out at 100 kPa TMP resulting in a flux of 100 l/m2.h. The treated juice had an extended shelf life (28 days) compared with the raw juice.

Pectin Removal and Clarification of Juices Chapter | 5

177

De Barros et al. (2003) used ceramic tubular module and PSF based hollow fiber (MWCO: 100 kDa) to clarify depectinized pineapple juice. Depectinization was carried out using 20 3 1023 kg/m3 Citrozym Ultra-L enzyme at 40 C for 90 minutes. The productivity of the ceramic membrane was higher than that of the PSF based hollow fibers. The temperature and TMP were selected as 40 C and 80 kPa, respectively. Carvalho and Silva (2010) hydrolyzed the pineapple juice using commercial pectinase (Ultrazym). A tubular PES MF membrane with 0.3 μm was employed for the postdepectinization step. At low TMP, the productivity was higher, since, at higher TMP, the formation of gel layer over the membrane surface was more compact imparting additional resistance to solute permeation. Formation of a dynamic layer decreased the productivity significantly. Retention of sugar from pineapple juice was reported by de Carvalho et al. (2008). The raw juice was initially hydrolyzed using commercial pectinase and a combination of cellulase and pectinase. Plate and frame and tubular modules were used for UF of the depectinized juice. The best recovery of sugar was observed in juices clarified by 50 kDa PSF membrane operating at 7.5 bar.

5.5.2.4 Banana Juice Banana (Musa acuminata) is one of the most widely consumed fruit in the world. India is one of the major producers of banana with total production volume of 29.1 million tons per year. This fruit is a staple food in developing countries. Due to its high potassium, magnesium, and sugar content, it acts as an alternative to a full course meal. However, storage of banana is a problem as it starts deteriorating in 2
3 days. Pectin is another major constituent in banana, for example, unripe fruits contain huge amount of pectin in their peels. The cell wall degradation of the fruit during ripening is primarily caused by solublization and depolymerization of pectins and hemicelluloses (Asif and Nath, 2005). Hence, during ripening of the banana the pectin content decreases. On the other hand, during ripening water soluble pectin concentration increases whereas acid soluble pectin content gradually reduces (Duan et al., 2008). Also, hot water homogenization of banana results in greater pectin content in ripe bananas (Prabha and Bhagyalakshmi, 1998). As discussed in the earlier sections, pectin being a gel forming agent is undesired for MSPs. Although the addition of fining agents like bentonite, gelatin, and others can stop browning and turbidity as well as can maintain the physicochemical properties of the juice (Lee et al., 2007), these external additives are not desired during processing. One of the major reasons for the browning of banana juice and its increment in turbidity is the presence of polyphenoloxidases (PPO) (Koffi et al., 1991). The latest enzymes or tyrosinases are tetramers with four copper atoms per molecule and provide binding sites for aromatic compounds and oxygen. PPO produce black, brown, or red

178

Separation of Functional Molecules in Food by Membrane Technology

pigments by rapid polymerization of o-quinones. Suitable membrane based processes coupled with proper depectinization treatment are able to remove PPO and ensure high throughput. As discussed in the earlier section, depectinization of banana can be performed using various enzymes, like pectinase, amylase, cellulase, hemicellulase, and others. A judicious combination of pectinase, cellulose, and hemicellulase was adapted by Koffi et al. (1991) who studied the effects of other additives, too. It was seen that the right combination of enzymes and the addition of preservatives like sulfite completely stops the browning. However, the addition of external preservative is not desired in the current scenario of fruit juice processing industry. MF was used by Valliant et al. (2008) to process pulpy fruits like, banana, pineapple, and blackberry. The turbidity of the feed was reduced after MF, but the physicochemical properties of the juice remain almost the same. Therefore, MF can only be used as a pretreatment method after depectinization to reduce membrane fouling in the subsequent membrane processes. Mature banana juice has also been subjected to UF and heat treatment to compare the two processes by evaluating physicochemical properties like color, PPO activity, sensory property of the juice, and browning (Sims et al., 1994). It was observed that 10 kDa UF membrane successfully removed all PPO and stopped the browning of the juice. The heat treatment increased the browning and also altered the flavor significantly. A complete formulation from extraction of banana juice from raw banana, and comparisons between preclarification methods such as MF and centrifugation, and selection of suitable UF membrane and optimized operating conditions, were undertaken by Sagu and his coworkers. Enzymatic pretreatment of the banana pulp was done using commercial pectinase obtained from Aspergillus niger. Optimized time, temperature, and enzyme concentration were 108 minutes, 33 C, and 0.03% (v/w), respectively (Sagu et al., 2014). Depectinized pulp was then processed by centrifugation and MF. Centrifugation was carried out in a batch centrifuge whereas MF was done using a PAN based hollow fiber module. The optimum operating conditions of both the processes were determined by RSM based on the physicochemical properties, such as viscosity, clarity, AIS, polyphenol, and protein. The optimized centrifugal speed and time were obtained as 46 minutes and 6065 g, respectively. The optimum TMP and cross-flow rate for MF were 76 kPa and 20 L/h, respectively. MF was found to be more economical, less power consuming, and produced better results in terms of physicochemical properties compared with centrifugation (Sagu et al., 2014). The microfiltered product was then subjected to UF using polymeric hollow fiber membranes of different MWCO ranging from 10 to 44 kDa. UF with MWCO 27 kDa was found to be the most suitable. The flux obtained at 104 kPa TMP was 82 L/m2 h. The processed juice had a shelf life of more than 1 month (Sagu et al., 2014).

Pectin Removal and Clarification of Juices Chapter | 5

179

Since membrane based processes do not require temperature for the clarification of juice’s bioactive fractions, a depectinization process followed by pre- and posttreatment by subsequent MF or UF process ensures higher shelf life for the juice without sacrificing its aroma, flavor, or sensory properties.

5.5.2.5 Mosambi Juice Mosambi (Citrus limetta) is a cultivar of sweet lime variation. Due to high vitamin C content and sweet taste, this fruit is gradually gaining popularity all over the world. Extraction, depectinization, pretreatment, and final clarification of the preprocessed juice has been studied by Rai and coworkers. Pretreatment of mosambi was undertaken using pectinase from Aspergillus niger with activity 3.5
7 units/mg. The process parameters were optimized by RSM taking into account the apparent viscosity, clarity, and AIS. The optimum process parameters were obtained as 42 C temperature, 0.0004% (w/v) enzyme concentration, and 99 minutes of operation (Rai et al., 2004). The best pretreatment step prior to UF was found to be enzymatic treatment followed by adsorption using bentonite (Rai et al., 2007). UF experiments were carried out using commercial membranes with MWCO ranging from 10 to 100 kDa. Prior to UF, the juice was subjected to MF using a 0.2-μm membrane. Due to partial or complete pore blocking, the permeate flux for MF was less than that of any subsequent UF process. The microfiltered juice was subjected to further clarification using UF membranes and 50-kDa UF membrane was found to be the most suitable (Rai et al., 2006). The clarified juice had a high shelf life of 30 days in refrigerated condition (Rai et al., 2008). 5.5.2.6 Other Juices Other than the aforementioned fruit juices, many studies revealed that combined depectinization and membrane based operation improved the quality, texture, and flavor of the juice. Moreover, coupled depectinization followed by MSP increased the shelf life of the juice significantly. A combined process of depectinization and subsequent membrane based separation for some fruit juice is presented in Table 5.5. Most of the citrus fruits are rich in pectin content and require depectinization prior to tertiary treatments like MF or UF. As stated in Table 5.5, many studies had been conducted for the depectinization and clarification of citrus fruits like lemon, kiwi, passionfruit, tangerine, pomegranate, and blackcurrant. For instance, lemon juice has been subjected to depectinization using the commercial enzyme Novozyme 33095 (Uc¸an et al., 2014). At the optimized condition of 35 C and concentration of 40 μL/100 mL depectinization was carried out for 15 minutes followed by separation using a commercial filter paper plate MinifiltroF6 from Italy. It was observed that the depectinized and clarified juice had zero pectin content after the process. The

TABLE 5.5 Coupled Depectinization of Some Fruit Juices and Subsequent MSP Fruit Juice

Enzyme Treatment

Membrane Based Clarification Process

References

Enzyme Used

Time (min)

Temperature ( C)

Concentration

Filtration Process

MWCO (kDa)

TMP (kPa)

Cross-Flow Rate

Lemon

Novozym 33095

15

35

40 μL/100 mL

Filter paper




By gravity




Uc¸an et al. (2014)

Passionfruit

Pectinex Ultra SP-L

60

50

150 mg/L

0.3 μm MF




100




de Oliveira et al. (2012)

Kiwifruit

Pectinex Ultra SP-L

240

Room temperature

10 g/kg

PVDF UF membrane

15

85

800 L/h

Cassano et al. (2004)

Tangerine

Pectinex Ultra SP-L

180

Room temperature

130 mg/L

0.1 μm MF




194

0.9 m/s

Chamchong and Noomhorm (1991)

Pomegranate

Fungus (F. fomentarius)

300

20

5 units/mL

Carbosep M2

15

200

1 m/s

Neifar et al. (2011)

Blackcurrant

Panzym Super E

720

25

0.004% (v/v)




Pap et al. (2009)

1440

25

Reverse osmosis membrane (salt rejection .80%)




Trenolin Rot DF

Tubular AFC80 PA membrane

Pectin Removal and Clarification of Juices Chapter | 5

181

clarified juice was then stored for 180 days at
25 C. Passionfruit is a rich source of carbohydrates, carotenes, vitamin C, iron, and other minerals. Its juice was depectinized using commercial enzyme Pectinex Ultra SP-L (de Oliveira et al., 2012). The operating parameters were 150 mg/L of enzyme at 50 C, whereas the depectinization was carried out for 60 minutes. Depectinized passionfruit juice was then subjected to MF using a 0.3-μm ceramic tubular module. The processed juice was analyzed in terms of turbidity, color, sensory properties, etc. It was observed that the permeate possessed low turbidity, low suspended solids content and was readily acceptable in terms of sensory qualities. As evident from earlier discussions, membrane based processes could be used to concentrate the fruit juice and it is a better alternative to thermal treatment. In another approach, kiwi fruit juice had been concentrated using OD process by Cassano et al. (2004). Prior to concentration the kiwi fruit juice was depectinized using 10 g/kg of Pectinex Ultra SP-L at room temperature and the process was carried out for 4 hours. The resultant depectinized juice was subjected to UF using a 15-kDa PVDF membrane. TMP and cross-flow rate were varied as 85 kPa and 800 L/h, respectively. The clarified kiwi fruit juice was concentrated using OD to a TSS of more than 60oBrix. Similar works have been carried out for tangerine, pomegranate, and % blackcurrant fruit juices (Chamchong and Noomhorm, 1991; Pap et al., 2009; Neifar et al., 2011). Tangerine juice was subjected to UF with MWCO ranging from 25 to 100 kDa and MF with pore size 0.1
0.2 μm. Enzymatic pretreatment was performed using Pectinex Ultra SP-L at acidic condition (pH 2). By analyzing the permeate quality and membrane productivity, 0.1-μm membrane was selected as optimum. It was evident that by increasing TMP the productivity increases without compromising the quality. However, with y increasing the cross-flow velocity the productivity increased by 9%; however the quality of permeate was deteriorated. Pomegranate juice on the other hand has been ultrafiltered using a 15-kDa Carbosep membrane. The raw juice was depectinized using laccase treatment to prevent browning or haze formation during storage. The ultrafiltered juice was processed at 200 kPa TMP and a clear, stable liquid was obtained. Similarly, blackcurrant juice was concentrated using a polyamide RO membrane. Prior to RO process, the juice was subjected to enzymatic depectinization using two different commercial enzymes, Panzym Super E and Trenolin Rot DF. The raw juice, depectinized by Panzym Super E and concentrated by RO yielded a TSS of 28.6oBrix. Depectinization is one of the necessary% pretreatment steps to reduce the pectin content in the fruit juice thereby reducing the fouling of the membranes effectively. Coupled depectinization and membrane based clarification have been successfully employed for several fruit juices, namely apple, orange, banana, mosambi, pineapple, kiwi, passion, lemon, etc. Since membrane based processes are easily scalable at industrial scale, they possess a huge potential as alternative to conventional fruit juice remediation. Hence,

182

Separation of Functional Molecules in Food by Membrane Technology

it is quite evident that membrane based systems could be widely used for the clarification and concentration of different fruit juices.

5.6 CONCLUSION Membrane based processes are more economical and less energy intensive than conventional juice clarification technologies. One of the conventional methods to concentrate fruit juices is thermally intensive multieffect evaporation operation. It basically utilizes heat from steam to evaporate water, thus making the juice concentrated. MSP, particularly RO, is an economical alternative to this. Since RO can be conducted at normal room temperature, the processed juice does not undergo any thermal degradation. In case of RO, most of the times the stream of importance is the retentate stream, which gets gradually concentrated with time of operation. pH and ionic characteristics of the juice are also preserved. Although RO based processes require high pressure, their energy requirements are relatively fewer compared with multiple effect evaporator systems, leading to significantly lower operating costs. Nanofiltration based MSP are generally used to concentrate the clarified juices. Also, nanofiltration is often used to fractionate and purify bioactive fractions of the juice. The MWCO of the NF membranes vary from 0.2 to ˚ . Nanofiltration was used to 1 kDa with the pore size ranging from 10 to 20 A clarify and concentrate apple and pear juice (Warczok et al., 2004), green tea extract (Ekanayake et al., 1999), cactus pear juice (Cassano et al., 2007), sea buckthorn juice (Vincze et al., 2006), red fruit juice (Koroknai et al., 2008), sugarcane juice (Nene et al., 2002), kiwifruit juice (Cassano et al., 2004), and others (Girard and Fukumoto, 2000; Cisse´ et al., 2011; Echavarrı´a et al., 2011). NF is also known for separation of sugar (Black and Bray, 1995). UF of fruit juices is the most common membrane based clarification process. Since the MWCO of the UF membranes vary from 1 to 100 kDa, a wide range of fruit juices can be processed using UF technology. Some typical applications of UF in clarification of juices for fruits and vegetables include passionfruit (Jiraratananon and Chanachai, 1996), potato (Zwijnenberg et al., 2002), mosambi (Rai et al., 2005), starfruit (Liew Abdullah et al., 2007), pineapple (de Barros et al., 2003), pear (Kirk et al., 1983), apple (Vladisavljevi´c et al., 2003), grape, lemon, orange (Chatterjee et al., 2004), banana (Sagu et al., 2014), bottle gourd (Mondal et al., 2016), coconut water (Karmakar and De, 2017), and many others. UF has also been applied on the other way around, for example, to clarify a pectin-rich extract recovered from olive mill wastewater (i.e., collected in the retentate stream) from potassium and other coextracted compounds (Galanakis et al., 2010). UF based MSP effectively eliminates microorganisms, improves the clarity and texture of the juice, and also removes potent sites for bacterial growth, thereby enhancing the shelf life of the processed juice.

Pectin Removal and Clarification of Juices Chapter | 5

183

MF is generally carried out by membranes having pore size of 0.2 μm and above. Generally, MF is used as a pretreatment step prior to subsequent polished filtration. MF eliminates all the suspended solids, high molecular weight proteins, and bacteria, improving the clarity of the juice significantly. Many fruit and vegetable juices from passionfruit (Vaillant et al., 1999), mango, pineapple, naranjilla, blackberry (Vaillant et al., 2001), cashew apple (Campos et al., 2002), acerola (Matta et al., 2004), pineapple (Carneiro et al., 2002), cactus pear (Cassano et al., 2010), orange (Cisse et al., 2005), pomegranate (Mirsaeedghazi et al., 2010), watermelon (Rai et al., 2010), banana (Sagu et al., 2014), bottle gourd (Biswas et al., 2016) and many others (Chatterjee et al., 2004; Vaillant et al., 2008) were pretreated using MF. Also, in most cases, MF is more effective and economical than other commonly used pretreatment methods like centrifugation. One of the major advantages of MSP is that the process and the technology are more economical than the majority of conventional methods and they also yield better results in terms of product quality and quantity. As dis´ lvarez et al. (2000) had shown that a membrane cussed in section 5.2, A based process is much more economical than conventional methods and it also results in less payback period (refer to Fig. 5.1). Sotoft et al. (2012) had developed a pilot scale plant for blackcurrant juice. The integrated membrane system includes vacuum MD, RO, NF, and direct contact MD. It was perceived that for the membrane based process the operational cost is 43% less than that of the conventional method. In view of the above discussion, it can be envisaged that membrane based processes can be utilized for processing of fruit juice for two purposes, clarification and concentration. But, the major bottleneck of the technology is membrane fouling leading to severe flux decline and making the system operationally unviable to commercial scale. Presence of pectinous materials (natural gelling agents) in raw juice causes this fouling. Using suitably selected enzymes (e.g., pectinase) at optimum operating conditions (reaction time and temperature), depectinization can be undertaken effectively to enhance membranes’ performance. However, the incomplete removal of pectin is harmful for fermentation of the juice affecting its shelf life adversely. Thus, leftover pectin (even after depectinization) should be ultrafiltered with an appropriately selected UF membrane to be completely removed. The processed juice is then coupled with an aseptic packaging to have a high shelf life. The selection of UF membrane depends on the juice’s nature, its content, and pectin concentration after depectinization. Thus, the role of UF is mainly the clarification. Next, important application of membrane based processes is concentration. For concentration purposes, reverse osmosis, low cutoff nanofiltration or osmotic dehydration could be implemented. The major purpose is to remove water from juice without any thermal treatment; thereby preventing degradation of essential nutritional and sensory components of the juice. However, in most cases, the UF-clarified juice is used for

184

Separation of Functional Molecules in Food by Membrane Technology

RO, NF, or OD processes to reduce the fouling of the membrane, enhancing its performance by processing large volume in short time. In such cases, UF is considered as a pretreatment prior to RO/NF/OD. The selection of operating conditions is quite critical, too. Two main operating parameters are TMP drop and cross-flow rate. Beyond a particular TMP drop, the permeate flux is independent of pressure (known as limiting flux) due to the formation of a gel layer. Thus, the identification of TMP drop corresponding to limiting permeate flux is also of vital importance. The actual operating pressure should not exceed the limiting TMP drop. Increase in cross-flow rate improves the mass transfer, thereby pushing up the limiting conditions realizing higher permeate flux. However, the positive effect of increase in cross-flow rate would not enhance the limiting flux beyond 20%
30%. In fact, there exists a phase space that clearly shows the envelope of TMP drop and cross-flow rate under limiting conditions (Roy and De, 2015; Karmakar and De, 2017). Selection of membrane modules is another important consideration. Although high TMP can be generated in spiral wound membrane modules, hollow fiber modules are recently gaining more attention. There are three distinct advantages of hollow fibers. First, the performance of laboratory scale experiments is directly scalable due to similarity in flow geometry. In case of spiral wound module, laboratory experiments in flat sheet membranes may not be a direct representative to actual module performance due to complication of flow geometry and flow path by spacer and winding of the membranes about a central axis. Secondly, hollow fibers are operated at much lower operating pressure compared with NF and RO, thereby reducing the power requirement tremendously, lowering the operating cost significantly. High throughput in the hollow fiber modules can be realized by packing large number of fibers or placing a large number of modules in parallel configuration. Thirdly, the fouling of the membrane is also significantly less at lower operating pressure. Thus, the selection of membrane modules and operating conditions are important issues in membrane based processes. With a prudent and judicial selection of membrane, module, and operating conditions, MSPs offer better alternative to conventional thermal and nonthermal processing technologies. They have distinct advantages over other conventional methods in terms of product quality, quantity, economical viability, and maintenance. Therefore, it is envisaged that MSP will have a lasting impact on the fruit juice processing industry in the future.

REFERENCES Aguiar, I.B., et al., 2012. Physicochemical and sensory properties of apple juice concentrated by reverse osmosis and osmotic evaporation. Innovative Food Sci. Emerg. Technol. 16, 137
142. Available from: https://doi.org/10.1016/j.ifset.2012.05.003.

Pectin Removal and Clarification of Juices Chapter | 5

185

Aguilar-Rosas, S.F., et al., 2007. Thermal and pulsed electric fields pasteurization of apple juice: effects on physicochemical properties and flavour compounds. J. Food Eng. 83 (1), 41
46. Available from: https://doi.org/10.1016/j.jfoodeng.2006.12.011. Al-Obaidi, M.A., Kara- Zaı¨tri, C., Mujtaba, I.M., 2017. Optimum design of a multi-stage reverse osmosis process for the production of highly concentrated apple juice. J. Food Eng. 214, 47
59. Available from: https://doi.org/10.1016/j.jfoodeng.2017.06.020. Alkorta, I., et al., 1998. Industrial applications of pectic enzymes: a review. Process Biochem. 33 (1), 21
28. Available from: https://doi.org/10.1016/S0032-9592(97)00046-0. ´ lvarez, S., et al., 1998. Influence of depectinization on apple juice ultrafiltration. Colloids A Surfaces A Physicochem. Eng. Aspects 138 (2
3), 377
382. Available from: https://doi. org/10.1016/S0927-7757(98)00235-0. ´ lvarez, S., et al., 2000. A new integrated membrane process for producing clarified apple juice A and apple juice aroma concentrate. J. Food Eng. 46 (2), 109
125. Available from: https:// doi.org/10.1016/S0308-8146(00)00139-4. Aneja, K.R., et al., 2014. Emerging preservation techniques for controlling spoilage and pathogenic microorganisms in fruit juices. Int. J. Microbiol. 2014, 1
14. Available from: https:// doi.org/10.1155/2014/758942. Asif, M.H., Nath, P., 2005. Expression of multiple forms of polygalacturonase gene during ripening in banana fruit. Plant. Physiol. Biochem. 43 (2), 177
184. Available from: https://doi. org/10.1016/j.plaphy.2005.01.011. Barboni, T., Cannac, M., Chiaramonti, N., 2010. Effect of cold storage and ozone treatment on physicochemical parameters, soluble sugars and organic acids in Actinidia deliciosa. Food. Chem. 121 (4), 946
951. Available from: https://doi.org/10.1016/j. foodchem.2010.01.024. Barrosi, S.T.D., Mendes, E.S., Peres, L., 2004. Influence of depectinization in the ultrafiltration of West Indian cherry (Malpighia glabra L.) and pineapple (Ananas comosus (L.) Meer) juices. Cieˆncia e Tecnologia de Alimentos 24 (2), 194
201. Available from: https://doi.org/ 10.1590/S0101-20612004000200006. Beuchat, L.R., 1982. Thermal inactivation of yeasts in fruit juices supplemented with food preservatives and sucrose. J. Food. Sci. 47 (5), 1679
1682. Available from: https://doi.org/ 10.1111/j.1365-2621.1982.tb05010.x. Biswas, P.P., Mondal, M., De, S., 2016. Comparison between centrifugation and microfiltration as primary clarification of bottle gourd (Lagenaria siceraria) juice. J. Food Proces. Preservation 40 (2), 226
238. Available from: https://doi.org/10.1111/jfpp.12599. Black, H.F., Bray, R.G., 1995. Sugar separation from juices and product thereof. Google Patents. Available at: https://www.google.com/patents/US5403604. Campos, D.C.P., et al., 2002. Cashew apple juice stabilization by microfiltration. Desalination 148 (1
3), 61
65. Available from: https://doi.org/10.1016/S0011-9164(02)00654-9. Carneiro, L., et al., 2002. Cold sterilization and clarification of pineapple juice by tangential microfiltration. Desalination 148 (1
3), 93
98. Available from: https://doi.org/10.1016/ S0011-9164(02)00659-8. Carvalho, L.M.Jde, Silva, C.A.Bda, 2010. Clarification of pineapple juice by microfiltration. Cieˆncia e Tecnologia de Alimentos 30 (3), 828
832. Available from: https://doi.org/ 10.1590/S0101-20612010000300040. Cassano, A., Jiao, B., Drioli, E., 2004. Production of concentrated kiwifruit juice by integrated membrane process. Food Res. Int. 37 (2), 139
148. Available from: https://doi.org/10.1016/ j.foodres.2003.08.009.

186

Separation of Functional Molecules in Food by Membrane Technology

Cassano, A., et al., 2007. A membrane-based process for the clarification and the concentration of the cactus pear juice. J. Food Eng. 80 (3), 914
921. Available from: https://doi.org/ 10.1016/j.jfoodeng.2006.08.005. Cassano, A., Marchio, M., Drioli, E., 2007. Clarification of blood orange juice by ultrafiltration: analyses of operating parameters, membrane fouling and juice quality. Desalination 212 (1
3), 15
27. Available from: https://doi.org/10.1016/j.desal.2006.08.013. Cassano, A., Conidi, C., Drioli, E., 2010. Physico-chemical parameters of cactus pear (Opuntia ficus-indica) juice clarified by microfiltration and ultrafiltration processes. Desalination 250 (3), 1101
1104. Available from: https://doi.org/10.1016/j.desal.2009.09.117. Chamchong, M., Noomhorm, A., 1991. Effect of pH and enzymatic treatment on microfiltration and ultrafiltration of tangerine juice. J. Food Process Eng. 14 (1), 21
34. Available from: https://doi.org/10.1111/j.1745-4530.1991.tb00079.x. Chang, T., et al., 1995. Commercial pectinases and the yield and quality of stanley plum juice. J. Food Process. Preservation 19 (2), 89
101. Available from: https://doi.org/10.1111/j.17454549.1995.tb00280.x. Char, C.D., et al., 2010. Use of high-intensity ultrasound and UV-C light to inactivate some microorganisms in fruit juices. Food Bioprocess Technol. 3 (6), 797
803. Available from: https://doi.org/10.1007/s11947-009-0307-7. Chatterjee, S., et al., 2004. Clarification of fruit juice with chitosan. Process Biochem. 39 (12), 2229
2232. Available from: https://doi.org/10.1016/j.procbio.2003.11.024. Chen, J., et al., 2008. Role of phenylalanine ammonia-lyase in heat pretreatment-induced chilling tolerance in banana fruit. Physiol. Plant. 132 (3), 318
328. Available from: https://doi.org/ 10.1111/j.1399-3054.2007.01013.x. Chhaya, Majumdar, G.C., De, S., 2013. Primary clarification of stevia extract: a comparison between centrifugation and microfiltration. Sep. Sci. Technol. 48 (1), 113
121. Available from: https://doi.org/10.1080/01496395.2012.674605. Cisse, M., et al., 2005. The quality of orange juice processed by coupling crossflow microfiltration and osmotic evaporation. Int. J. Food Sci. Technol. 40 (1), 105
116. Available from: https://doi.org/10.1111/j.1365-2621.2004.00914.x. Cisse´, M., et al., 2011. ) Selecting ultrafiltration and nanofiltration membranes to concentrate anthocyanins from roselle extract (Hibiscus sabdariffa L.). Food Res. Int. 44 (9), 2607
2614. Available from: https://doi.org/10.1016/j.foodres.2011.04.046. Conidi, C., Destani, F., Cassano, A., 2015. Performance of hollow fiber ultrafiltration membranes in the clarification of blood orange juice. Beverages 1 (4), 341
353. Available from: https://doi.org/10.3390/beverages1040341. Cserhalmi, Z., et al., 2006. Study of pulsed electric field treated citrus juices. Innovative Food Sci. Em. Technol. 7 (1
2), 49
54. Available from: https://doi.org/10.1016/j.ifset.2005.07.001. de Barros, S., et al., 2003. Study of fouling mechanism in pineapple juice clarification by ultrafiltration. J. Memb. Sci. 215 (1
2), 213
224. Available from: https://doi.org/10.1016/ S0376-7388(02)00615-4. de Carvalho, L.M.J., de Castro, I.M., da Silva, C.A.B., 2008. A study of retention of sugars in the process of clarification of pineapple juice (Ananas comosus, L. Merril) by micro- and ultra-filtration. J. Food Eng. 87 (4), 447
454. Available from: https://doi.org/10.1016/j. jfoodeng.2007.12.015. Deng, C., O’Neill, M.A., York, W.S., 2006. Selective chemical depolymerization of rhamnogalacturonans. Carbohydr. Res. 341 (4), 474
484. Available from: https://doi.org/10.1016/j. carres.2005.12.004.

Pectin Removal and Clarification of Juices Chapter | 5

187

de Oliveira, R.C., Doceˆ, R.C., de Barros, S.T.D., 2012. Clarification of passion fruit juice by microfiltration: analyses of operating parameters, study of membrane fouling and juice quality. J. Food Eng. 111 (2), 432
439. Available from: https://doi.org/10.1016/j. jfoodeng.2012.01.021. Derieux, J., 1988. Histoire de la panification et de levure dans Levure et panification Fould. Springer. Tecno-Nathant, Arbrissel, France. Destani, F., et al., 2013. Recovery and concentration of phenolic compounds in blood orange juice by membrane operations. J. Food Eng. 117 (3), 263
271. Available from: https://doi. org/10.1016/j.jfoodeng.2013.03.001. Dingle, J., Reid, W.W., Solomons, G.L., 1953. The enzymic degradation of pectin and other polysaccharides. II—Application of the “Cup-plate” assay to the estimation of enzymes. J. Sci. Food Agric. 4 (3), 149
155. Available from: https://doi.org/10.1002/jsfa.2740040305. Domingues, R.C.C., et al., 2014. Microfiltration of passion fruit juice using hollow fibre membranes and evaluation of fouling mechanisms. J. Food Eng. 121, 73
79. Available from: https://doi.org/10.1016/j.jfoodeng.2013.07.037. Duan, X., et al., 2008. Modification of pectin polysaccharides during ripening of postharvest banana fruit. Food. Chem. 111 (1), 144
149. Available from: https://doi.org/10.1016/j. foodchem.2008.03.049. Echavarrı´a, A.P., et al., 2011. Fruit juice processing and membrane technology application. Food Eng. Rev. 3 (3
4), 136
158. Available from: https://doi.org/10.1007/s12393-0119042-8. Ekanayake, A., Bunger, J.R., Mohlenkamp, M.J., 1999. Green tea extract subjected to cation exchange treatment and nanofiltration to improve clarity and color. Google Patents. Available at: https://www.google.com/patents/US5879733. Escobedo-Avellaneda, Z., et al., 2011. Benefits and limitations of food processing by highpressure technologies: effects on functional compounds and abiotic contaminants Beneficios y limitaciones del procesamiento de alimentos por tecnologı´as de alta presio´n: efectos en componentes funcionale. CyTA - J. Food 9 (4), 351
364. Available from: https://doi.org/ 10.1080/19476337.2011.616959. Evrendilek, G.A., et al., 2000. Microbial safety and shelf-life of apple juice and cider processed by bench and pilot scale PEF systems. Innovative Food Sci. Emerg. Technol. 1 (1), 77
86. Available from: https://doi.org/10.1016/S1466-8564(99)00004-1. Fabroni, S., et al., 2016. Anthocyanins in different Citrus species: an UHPLC-PDA-ESI/MS n -assisted qualitative and quantitative investigation. J. Sci. Food. Agric. 96 (14), 4797
4808. Available from: https://doi.org/10.1002/jsfa.7916. Fellers, P.J., 1988. Shelf life and quality of freshly squeezed, unpasteurized, polyethylene-bottled citrus juice. J. Food. Sci. 53 (6), 1699
1702. Available from: https://doi.org/10.1111/j.13652621.1988.tb07819.x. Ferrari, G., Maresca, P., Ciccarone, R., 2010. The application of high hydrostatic pressure for the stabilization of functional foods: pomegranate juice. J. Food Eng. 100 (2), 245
253. Available from: https://doi.org/10.1016/j.jfoodeng.2010.04.006. Fiore, A., et al., 2005. Antioxidant activity of pasteurized and sterilized commercial red orange juices. Mol. Nutr. Food. Res. 49 (12), 1129
1135. Available from: https://doi.org/10.1002/ mnfr.200500139. Galanakis, C.M., Tornberg, E., Gekas, V., 2010. Clarification of high-added value products from olive mill wastewater. J. Food Eng. 99 (2), 190
197. Available from: https://doi.org/ 10.1016/j.jfoodeng.2010.02.018.

188

Separation of Functional Molecules in Food by Membrane Technology

Galaverna, G., et al., 2008. A new integrated membrane process for the production of concentrated blood orange juice: effect on bioactive compounds and antioxidant activity. Food. Chem. 106 (3), 1021
1030. Available from: https://doi.org/10.1016/j.foodchem.2007.07.018. Gilewiczlukasik, B., Koter, S., Kurzawa, J., 2007. Concentration of anthocyanins by the membrane filtration. Sep. Purif. Technol. 57 (3), 418
424. Available from: https://doi.org/ 10.1016/j.seppur.2006.03.026. Giovanelli, G., Ravasini, G., 1993. Apple juice stabilization by combined enzyme—membrane filtration process. LWT - Food Sci. Technol. 26 (1), 1
7. Available from: https://doi.org/ 10.1006/fstl.1993.1001. Girard, B., Fukumoto, L.R., 2000. Membrane processing of fruit juices and beverages: a review. Crit. Rev. Biotechnol. 20 (2), 109
175. Available from: https://doi.org/10.1080/ 07388550008984168. Golovchenko, V.V., et al., 2002. Structural studies of the pectic polysaccharide from duckweed Lemna minor L. Phytochemistry 60 (1), 89
97. Available at: http://www.ncbi.nlm.nih.gov/ pubmed/11985856. Grassin, C., Fauquembergue, P., 1996. Application of pectinases in beverages. In pp. 453
462. doi: 10.1016/S0921-0423(96)80275-9. Gu, B., et al., 2011. Mathematical model of flat sheet membrane modules for FO process: plateand-frame module and spiral-wound module. J. Memb. Sci. 379 (1
2), 403
415. Available from: https://doi.org/10.1016/j.memsci.2011.06.012. Guerrero-Beltrn, J.A., Barbosa-Cnovas, G.V., 2004. Advantages and limitations on processing foods by UV light. Food Sci. Technol. Int. 10 (3), 137
147. Available from: https://doi.org/ 10.1177/1082013204044359. Gulec, H.A., Bagci, P.O., Bagci, U., 2017. Clarification of apple juice using polymeric ultrafiltration membranes: a comparative evaluation of membrane fouling and juice quality. Food Bioprocess Technol. 10 (5), 875
885. Available from: https://doi.org/10.1007/s11947-0171871-x. Gummadi, S.N., Panda, T., 2003. Purification and biochemical properties of microbial pectinases—a review. Process Biochem. 38 (7), 987
996. Available from: https://doi.org/ 10.1016/S0032-9592(02)00203-0. Guthrie, J.F., Morton, J.F., 2000. Food sources of added sweeteners in the diets of Americans. J. Am. Diet. Assoc. 100 (1), 43
51. Available from: https://doi.org/10.1016/S0002-8223(00)00018-3. Hoondal, G., et al., 2002. Microbial alkaline pectinases and their industrial applications: a review. Appl. Microbiol. Biotechnol. 59 (4
5), 409
418. Available from: https://doi.org/ 10.1007/s00253-002-1061-1. Huber, D.J., Lee, J.H., 1986. Comparative analysis of pectins from Pericarp and Locular Gel in developing tomato fruit. In pp. 141
156. https://doi.org/10.1021/bk-1986-0310.ch012. Imamura, F., et al., 2015. Consumption of sugar sweetened beverages, artificially sweetened beverages, and fruit juice and incidence of type 2 diabetes: systematic review, metaanalysis, and estimation of population attributable fraction. BMJ h3576. Available from: https://doi.org/10.1136/bmj.h3576. Jackman, R.L., et al., 1987. Anthocyanins as food colorants
a review. J. Food Biochem. 11 (3), 201
247. Available from: https://doi.org/10.1111/j.1745-4514.1987.tb00123.x. Jayani, R.S., Saxena, S., Gupta, R., 2005. Microbial pectinolytic enzymes: a review. Process Biochem. 40 (9), 2931
2944. Available from: https://doi.org/10.1016/j.procbio.2005.03.026. Jiraratananon, R., Chanachai, A., 1996. A study of fouling in the ultrafiltration of passion fruit juice. J. Memb. Sci. 111 (1), 39
48. Available from: https://doi.org/10.1016/0376-7388(95) 00270-7.

Pectin Removal and Clarification of Juices Chapter | 5

189

Jordan, S.L., et al., 2001. Inactivation and injury of pressure-resistant strains of Escherichia coli O157 and Listeria monocytogenes in fruit juices. J. Appl. Microbiol. 91 (3), 463
469. Available from: https://doi.org/10.1046/j.1365-2672.2001.01402.x. Kabasakalis, V., 2000. Ascorbic acid content of commercial fruit juices and its rate of loss upon storage. Food. Chem. 70 (3), 325
328. Available from: https://doi.org/10.1016/S0308-8146 (00)00093-5. Karaca, H., Velioglu, Y.S., 2007. Ozone applications in fruit and vegetable processing. Food Rev. Int. 23 (1), 91
106. Available from: https://doi.org/10.1080/87559120600998221. Karmakar, S., De, S., 2017. Cold sterilization and process modeling of tender coconut water by hollow fibers. J. Food Eng. 200, 70
80. Available from: https://doi.org/10.1016/j.jfoodeng.2016.12.021. Karmakar, S., Bhattacharjee, S., De, S., 2017. Experimental and modeling of fluoride removal using aluminum fumarate (AlFu) metal organic framework incorporated cellulose acetate phthalate mixed matrix membrane. J. Environ. Chem. Eng. 5 (6), 6087
6097. Available from: https://doi.org/10.1016/j.jece.2017.11.035. Karmakar, S., Bhattacharjee, S., De, S., 2018. Aluminium fumarate metal organic framework incorporated polyacrylonitrile hollow fiber membranes: spinning, characterization and application in fluoride removal from groundwater. Chem. Eng. J. 334, 41
53. Available from: https://doi.org/10.1016/j.cej.2017.10.021. Keppler, F., et al., 2006. Methane emissions from terrestrial plants under aerobic conditions. Nature 439 (7073), 187
191. Available from: https://doi.org/10.1038/nature04420. Kesting, R., 1971. Synthetic Polymeric Membranes. McGraw-Hill, New York. Keyser, M., et al., 2008. Ultraviolet radiation as a non-thermal treatment for the inactivation of microorganisms in fruit juice. Innovative Food Sci. Emerg. Technol. 9 (3), 348
354. Available from: https://doi.org/10.1016/j.ifset.2007.09.002. Khandpur, P., Gogate, P.R., 2016. Evaluation of ultrasound based sterilization approaches in terms of shelf life and quality parameters of fruit and vegetable juices. Ultrason. Sonochem. 29, 337
353. Available from: https://doi.org/10.1016/j.ultsonch.2015.10.008. Kirk, D.E., Montgomery, M.W., Kortekaas, M.G., 1983. Clarification of pear juice by hollow fiber ultrafiltration. J. Food. Sci. 48 (6), 1663
1667. Available from: https://doi.org/ 10.1111/j.1365-2621.1983.tb05055.x. Koffi, E.K., Sims, C.A., Bates, R.P., 1991. Viscosity reduction and prevention of browning in the preparation of clarified banana juice. J. Food. Qual. 14 (3), 209
218. Available from: https://doi.org/10.1111/j.1745-4557.1991.tb00062.x. Komitopoulou, E., et al., 1999. Alicyclobacillus acidoterrestris in fruit juices and its control by nisin. Int. J. Food Sci. Technol. 34 (1), 81
85. Available from: https://doi.org/10.1046/ j.1365-2621.1999.00243.x. Koroknai, B., et al., 2008. Preservation of antioxidant capacity and flux enhancement in concentration of red fruit juices by membrane processes. Desalination 228 (1
3), 295
301. Available from: https://doi.org/10.1016/j.desal.2007.11.010. ´ ., Rektor, A., Vatai, G., 2006. Integrated large-scale membrane process for producing Koza´k, A concentrated fruit juices. Desalination 200 (1
3), 540
542. Available from: https://doi.org/ 10.1016/j.desal.2006.03.428. Kratchanova, M., Pavlova, E., Panchev, I., 2004. The effect of microwave heating of fresh orange peels on the fruit tissue and quality of extracted pectin. Carbohyd. Polym. 56 (2), 181
185. Available from: https://doi.org/10.1016/j.carbpol.2004.01.009. Kumar, S., et al., 2012. Inhibition of pericarp browning and shelf life extension of litchi by combination dip treatment and radiation processing. Food. Chem. 131 (4), 1223
1232. Available from: https://doi.org/10.1016/j.foodchem.2011.09.108.

190

Separation of Functional Molecules in Food by Membrane Technology

¨ zkan, M., Cemeroglu, B., 2006. Stability of black carrot anthocyanins in various Kırca, A., O fruit juices and nectars. Food. Chem. 97 (4), 598
605. Available from: https://doi.org/ 10.1016/j.foodchem.2005.05.036. Laboissie`re, L.H.E.S., et al., 2007. Effects of high hydrostatic pressure (HHP) on sensory characteristics of yellow passion fruit juice. Innovative Food Sci. Emerg. Technol. 8 (4), 469
477. Available from: https://doi.org/10.1016/j.ifset.2007.04.001. Ladewig, B., Al-Shaeli, M.N.Z., 2017. Fundamentals of membrane processes. In pp. 13
37. https://doi.org/10.1007/978-981-10-2014-8_2. Lee, H.S., Coates, G.A., 2003. Effect of thermal pasteurization on Valencia orange juice color and pigments. LWT - Food Sci. Technol. 36 (1), 153
156. Available from: https://doi.org/ 10.1016/S0023-6438(02)00087-7. Lee, W.C., et al., 2007. Effects of fining treatment and storage temperature on the quality of clarified banana juice. LWT - Food Sci. Technol. 40 (10), 1755
1764. Available from: https://doi.org/10.1016/j.lwt.2006.12.008. Liew Abdullah, A.G., et al., 2007. Response surface optimization of conditions for clarification of carambola fruit juice using a commercial enzyme. J. Food Eng. 81 (1), 65
71. Available from: https://doi.org/10.1016/j.jfoodeng.2006.10.013. Ludwig, D.S., Peterson, K.E., Gortmaker, S.L., 2001. Relation between consumption of sugarsweetened drinks and childhood obesity: a prospective, observational analysis. Lancet 357 (9255), 505
508. Available from: https://doi.org/10.1016/S0140-6736(00)04041-1. Malik, V.S., et al., 2010. Sugar-sweetened beverages, obesity, type 2 diabetes mellitus, and cardiovascular disease risk. Circulation 121 (11), 1356
1364. Available from: https://doi.org/ 10.1161/CIRCULATIONAHA.109.876185. Martı´n, J., Dı´az-Montan˜a, E.J., Asuero, A.G., 2018. Recovery of anthocyanins using membrane technologies: a review. Crit. Rev. Analyt. Chem. 48 (3), 143
175. Available from: https:// doi.org/10.1080/10408347.2017.1411249. Maskan, M., 2001. Kinetics of colour change of kiwifruits during hot air and microwave drying. J. Food Eng. 48 (2), 169
175. Available from: https://doi.org/10.1016/S0260-8774(00)00154-0. Matta, V.M., Moretti, R.H., Cabral, L.M.C., 2004. Microfiltration and reverse osmosis for clarification and concentration of acerola juice. J. Food Eng. 61 (3), 477
482. Available from: https://doi.org/10.1016/S0260-8774(03)00154-7. Mazza, G., Brouillard, R., 1987. Recent developments in the stabilization of anthocyanins in food products. Food. Chem. 25 (3), 207
225. Available from: https://doi.org/10.1016/03088146(87)90147-6. Mazzotta, A.S., 2001. Thermal inactivation of stationary-phase and acid-adapted Escherichia coli O157:H7, Salmonella, and Listeria monocytogenes in fruit juices. J. Food. Prot. 64 (3), 315
320. Available from: https://doi.org/10.4315/0362-028X-64.3.315. Mirsaeedghazi, H., et al., 2010. Clarification of pomegranate juice by microfiltration with PVDF membranes. Desalination 264 (3), 243
248. Available from: https://doi.org/10.1016/j. desal.2010.03.031. Misra, N.N., et al., 2017. Landmarks in the historical development of twenty first century food processing technologies. Food Res. Int. 97, 318
339. Available from: https://doi.org/ 10.1016/j.foodres.2017.05.001. Mizrahi, S., Berk, Z., 1970. Flow behaviour of concentrated orange juice. J. Texture Stud. 1 (3), 342
355. Available from: https://doi.org/10.1111/j.1745-4603.1970.tb00735.x. Mkandawire, W., et al., 2016. Estimation of shelf life of mango juice produced using small-scale processing techniques. J. Food Res. 5 (6), 13. Available from: https://doi.org/10.5539/jfr. v5n6p13.

Pectin Removal and Clarification of Juices Chapter | 5

191

Mondal, M., Biswas, P.P., De, S., 2016. Clarification and storage study of bottle gourd (Lagenaria siceraria) juice by hollow fiber ultrafiltration. Food Bioprod. Process. 100, 1
15. Available from: https://doi.org/10.1016/j.fbp.2016.06.010. Mondor, M., Girard, B., Moresoli, C., 2000. Modeling flux behavior for membrane filtration of apple juice. Food Res. Int. 33 (7), 539
548. Available from: https://doi.org/10.1016/S09639969(00)00089-2. Morigami, Y., et al., 2001. The first large-scale pervaporation plant using tubular-type module with zeolite NaA membrane. Sep. Purif. Technol. 25 (1
3), 251
260. Available from: https://doi.org/10.1016/S1383-5866(01)00109-5. Mualikrishna, G., Tharanathan, R.N., 1994. Characterization of pectic polysaccharides from pulse husks. Food. Chem. 50 (1), 87
89. Available from: https://doi.org/10.1016/0308-8146 (94)90098-1. Neifar, M., et al., 2011. Effective clarification of pomegranate juice using laccase treatment optimized by response surface methodology followed by ultrafiltration. J. Food Process Eng. 34 (4), 1199
1219. Available from: https://doi.org/10.1111/j.1745-4530.2009.00523.x. Nene, S., et al., 2002. Membrane distillation for the concentration of raw cane-sugar syrup and membrane clarified sugarcane juice. Desalination 147 (1
3), 157
160. Available from: https://doi.org/10.1016/S0011-9164(02)00604-5. Ninga, K.A., et al., 2018. Kinetics of enzymatic hydrolysis of pectinaceous matter in guava juice. J. Food Eng. 221, 158
166. Available from: https://doi.org/10.1016/j. jfoodeng.2017.10.022. Onsekizoglu, P., Bahceci, K.S., Acar, M.J., 2010. Clarification and the concentration of apple juice using membrane processes: a comparative quality assessment. J. Memb. Sci. 352 (1
2), 160
165. Available from: https://doi.org/10.1016/j.memsci.2010.02.004. Pap, N., et al., 2009. Concentration of blackcurrant juice by reverse osmosis. Desalination 241 (1
3), 256
264. Available from: https://doi.org/10.1016/j.desal.2008.01.069. Pedrolli, D.B., et al., 2009. Pectin and pectinases: production, characterization and industrial application of microbial pectinolytic enzymes. Open Biotechnol. J. 3 (1), 9
18. Available from: https://doi.org/10.2174/1874070700903010009. Polydera, A., Stoforos, N., Taoukis, P., 2003. Comparative shelf life study and vitamin C loss kinetics in pasteurised and high pressure processed reconstituted orange juice. J. Food Eng. 60 (1), 21
29. Available from: https://doi.org/10.1016/S0260-8774(03)00006-2. Porter, M.C., 1990. Handbook of Industrial Membrane Technology. Noyes Publication, IL, Park Ridge, New Jersey. Prabha, T., Bhagyalakshmi, N., 1998. Carbohydrate metabolism in ripening banana fruit. Phytochemistry 48 (6), 915
919. Available from: https://doi.org/10.1016/S0031-9422(97) 00931-X. Prade, R.A., et al., 1999. Pectins, pectinases and plant-microbe interactions. Biotechnol. Genetic Eng. Rev. 16 (1), 361
392. Available from: https://doi.org/10.1080/ 02648725.1999.10647984. Qaid, S., 2017. Ultrafiltration for clarification of Valencia orange juice: Comparison of two flat sheet membranes on quality of juice production. J. Mater. Environ. Sci. 8, 1186
1194. Rai, C., et al., 2010. Mechanism of permeate flux decline during microfiltration of watermelon (Citrullus lanatus) juice. Food Bioprocess Technol. 3 (4), 545
553. Available from: https:// doi.org/10.1007/s11947-008-0118-2. Rai, P., et al., 2004. Optimizing pectinase usage in pretreatment of mosambi juice for clarification by response surface methodology. J. Food Eng. 64 (3), 397
403. Available from: https://doi.org/10.1016/j.jfoodeng.2003.11.008.

192

Separation of Functional Molecules in Food by Membrane Technology

Rai, P., et al., 2005. Modeling the performance of batch ultrafiltration of synthetic fruit juice and mosambi juice using artificial neural network. J. Food Eng. 71 (3), 273
281. Available from: https://doi.org/10.1016/j.jfoodeng.2005.02.003. Rai, P., Majumdar, G.C., Jayanti, V.K., et al., 2006. Alternative pretreatment methods to enzymatic treatment for clarification of mosambi juice using ultrafiltration. J. Food Process Eng. 29 (2), 202
218. Available from: https://doi.org/10.1111/j.1745-4530.2006.00058.x. Rai, P., Majumdar, G.C., Sharma, G., et al., 2006. Effect of various cutoff membranes on permeate flux and quality during filtration of mosambi (Citrus sinensis (L.) Osbeck) juice. Food Bioprod. Process. 84 (3), 213
219. Available from: https://doi.org/10.1205/ fbp.05181. Rai, P., et al., 2007. Effect of various pretreatment methods on permeate flux and quality during ultrafiltration of mosambi juice. J. Food Eng. 78 (2), 561
568. Available from: https://doi. org/10.1016/j.jfoodeng.2005.10.024. Rai, P., et al., 2008. Storage study of ultrafiltered mosambi (Citrus sinensis (l.) Osbeck) juice. J. Food Process. Preservation 32 (6), 923
934. Available from: https://doi.org/10.1111/j.17454549.2008.00222.x. Raso, J., et al., 1998. Inactivation of mold ascospores and conidiospores suspended in fruit juices by pulsed electric fields. LWT - Food Sci. Technol. 31 (7
8), 668
672. Available from: https://doi.org/10.1006/fstl.1998.0426. Raso, J., et al., 2006. Inactivation of Zygosaccharomyces bailii in fruit juices by heat, high hydrostatic pressure and pulsed electric fields. J. Food. Sci. 63 (6), 1042
1044. Available from: https://doi.org/10.1111/j.1365-2621.1998.tb15850.x. Ridley, B.L., O’Neill, M.A., Mohnen, D., 2001. Pectins: structure, biosynthesis, and oligogalacturonide-related signaling.Phytochemistry 57 (6), 929
967. Available at. Available from: http://www.ncbi.nlm.nih.gov/pubmed/11423142. Rossi, M., et al., 2003. Effect of fruit blanching on phenolics and radical scavenging activity of highbush blueberry juice. Food Res. Int. 36 (9
10), 999
1005. Available from: https://doi. org/10.1016/j.foodres.2003.07.002. Roy, A., De, S., 2015. Resistance-in-series model for flux decline and optimal conditions of Stevia extract during ultrafiltration using novel CAP-PAN blend membranes. Food Bioprod. Process. 94, 489
499. Available from: https://doi.org/10.1016/j.fbp.2014.07.006. Sagu, S.T., Nso, E.J., et al., 2014. Optimisation of low temperature extraction of banana juice using commercial pectinase. Food. Chem. 151, 182
190. Sagu, S.T., Karmakar, S., Nso, E.J., De, S., 2014. Primary clarification of banana juice extract by centrifugation and microfiltration. Sep. Sci. Technol. (Philadelphia) 49 (8), 1156
1169. Available from: https://doi.org/10.1080/01496395.2013.877932. Sagu, S.T., Karmakar, S., Nso, E.J., Kapseu, C., et al., 2014. Ultrafiltration of banana (Musa acuminata) juice using hollow fibers for enhanced shelf life. Food Bioprocess Technol. 7 (9), 2711
2722. Available from: https://doi.org/10.1007/s11947-014-1309-7. Sahin, ¸ S., Bayindirli, L., 1993. The effect of depectinization and clarification on the filtration of sour cherry juice. J. Food Eng. 19 (3), 237
245. Available from: https://doi.org/10.1016/ 0260-8774(93)90045-L. Santin, M.M., et al., 2008. Evaluation of enzymatic treatment of peach juice using response surface methodology. J. Sci. Food. Agric. 88 (3), 507
512. Available from: https://doi.org/ 10.1002/jsfa.3114. Schols, H.A., et al., 1995. A xylogalacturonan subunit present in the modified hairy regions of apple pectin. Carbohydr. Res. 279, 265
279. Available from: https://doi.org/10.1016/00086215(95)00287-1.

Pectin Removal and Clarification of Juices Chapter | 5

193

Sharma, V.K., Colangelo, A., Spagna, G., 1995. Experimental investigation of different solar dryers suitable for fruit and vegetable drying. Renew. Energy 6 (4), 413
424. Available from: https://doi.org/10.1016/0960-1481(94)00075-H. Shui, G., Leong, L.P., 2002. Separation and determination of organic acids and phenolic compounds in fruit juices and drinks by high-performance liquid chromatography. J. Chromatogr. A. 977 (1), 89
96. Available from: https://doi.org/10.1016/S0021-9673(02)01345-6. Sims, C.A., Bates, R.P., Arreola, A.G., 1994. Color, polyphenoloxidase, and sensory changes in banana juice as affected by heat and ultrafiltration. J. Food. Qual. 17 (5), 371
379. Available from: https://doi.org/10.1111/j.1745-4557.1994.tb00158.x. Siricururatana, P., et al., 2013. Shelf-life evaluation of natural antimicrobials for concord and niagara grape juices. J. Food. Prot. 76 (1), 72
78. Available from: https://doi.org/10.4315/ 0362-028X.JFP-12-144. Skolnik, H., 1968. History, evolution, and status of agriculture and food science and technology. J. Chem. Doc. 8 (2), 95
98. Sopie Edwige Salome, Y., et al., 2011. Phenolic profiles of pineapple fruits (Ananas comosus L. Merrill) Influence of the origin of suckers. Aust. J. Basic Appl. Sci. 5, 1372
1378. Sotoft, L.F., et al., 2012. Full scale plant with membrane based concentration of blackcurrant juice on the basis of laboratory and pilot scale tests. Chem. Eng. Processing: Process Intensif. 54, 12
21. Available from: https://doi.org/10.1016/j.cep.2012.01.007. ´ ., et al., 2010. Effect of UHPH on indigenous microbiota of apple juice. Int. J. Food Sua´rez-Jacobo, A Microbiol. 136 (3), 261
267. Available from: https://doi.org/10.1016/j.ijfoodmicro.2009.11.011. Thakur, B.K., De, S., 2012. A novel method for spinning hollow fiber membrane and its application for treatment of turbid water. Sep. Purif. Technol. 93, 67
74. Available from: https:// doi.org/10.1016/j.seppur.2012.03.032. Thakur, B.R., et al., 1997. Chemistry and uses of pectin — a review. Crit. Rev. Food. Sci. Nutr. 37 (1), 47
73. Available from: https://doi.org/10.1080/10408399709527767. Tiwari, B.K., et al., 2008. Modelling colour degradation of orange juice by ozone treatment using response surface methodology. J. Food Eng. 88 (4), 553
560. Available from: https:// doi.org/10.1016/j.jfoodeng.2008.03.021. Tiwari, B.K., et al., 2009. Anthocyanin and colour degradation in ozone treated blackberry juice. Innovative Food Sci. Emerg. Technol. 10 (1), 70
75. Available from: https://doi.org/ 10.1016/j.ifset.2008.08.002. Uc¸an, F., Akyildiz, A., A˘gc¸am, E., 2014. Effects of different enzymes and concentrations in the production of clarified lemon juice. J. Food Process. 2014, 1
14. Available from: https:// doi.org/10.1155/2014/215854. Vaillant, F., et al., 1999. Crossflow microfiltration of passion fruit juice after partial enzymatic liquefaction. J. Food Eng. 42 (4), 215
224. Available from: https://doi.org/10.1016/S02608774(99)00124-7. Vaillant, F., et al., 2001. Strategy for economical optimisation of the clarification of pulpy fruit juices using crossflow microfiltration. J. Food Eng. 48 (1), 83
90. Available from: https:// doi.org/10.1016/S0260-8774(00)00152-7. Vaillant, F., et al., 2008. Turbidity of pulpy fruit juice: a key factor for predicting cross-flow microfiltration performance. J. Memb. Sci. 325 (1), 404
412. Available from: https://doi. org/10.1016/j.memsci.2008.08.003. Varela-Santos, E., et al., 2012. Effect of high hydrostatic pressure (HHP) processing on physicochemical properties, bioactive compounds and shelf-life of pomegranate juice. Innovative Food Sci. Emerg. Technol. 13, 13
22. Available from: https://doi.org/10.1016/j. ifset.2011.10.009.

194

Separation of Functional Molecules in Food by Membrane Technology

Vegara, S., et al., 2013. Effect of pasteurization process and storage on color and shelf-life of pomegranate juices. LWT - Food Sci. Technol. 54 (2), 592
596. Available from: https:// doi.org/10.1016/j.lwt.2013.06.022. Vincze, I., Stefanovits-Ba´nyai, E´., Vatai, G., 2006. Using nanofiltration and reverse osmosis for the concentration of seabuckthorn (Hippophae rhamnoides L.) juice. Desalination 200 (1
3), 528
530. Available from: https://doi.org/10.1016/j.desal.2006.03.423. Vı´quez, F., Lastreto, C., Cooke, R.D., 2007. A study of the production of clarified banana juice using pectinolytic enzymes. Inter. J. Food Sci. Technol. 16 (2), 115
125. Available from: https://doi.org/10.1111/j.1365-2621.1981.tb01003.x. Visser, J., Voragen, A.G., 2009. In: Schols, H.A., Visser, R.G.F., Voragen, A.G.J. (Eds.), Pectins and Pectinases. Wageningen Academic Publishers, The Netherlands. Available from: https:// doi.org/10.3920/978-90-8686-677-9. Vladisavljevi´c, G.T., Vukosavljevi´c, P., Bukvi´c, B., 2003. Permeate flux and fouling resistance in ultrafiltration of depectinized apple juice using ceramic membranes. J. Food Eng. 60 (3), 241
247. Available from: https://doi.org/10.1016/S0260-8774(03)00044-X. Wang, L.K., et al., (Eds.), 2011. Membrane and Desalination Technologies. Humana Press, Totowa, NJ. Available from: https://doi.org/10.1007/978-1-59745-278-6. Warczok, J., et al., 2004. Concentration of apple and pear juices by nanofiltration at low pressures. J. Food Eng. 63 (1), 63
70. Available from: https://doi.org/10.1016/S0260-8774(03) 00283-8. Whitaker, J.R., 1984. Pectic substances, pectic enzymes and haze formation in fruit juices. Enzyme Microb. Technol. 6 (8), 341
349. Available from: https://doi.org/10.1016/01410229(84)90046-2. White, B.L., Howard, L.R., Prior, R.L., 2011. Impact of different stages of juice processing on the anthocyanin, flavonol, and procyanidin contents of cranberries. J. Agric. Food. Chem. 59 (9), 4692
4698. Available from: https://doi.org/10.1021/jf200149a. Willats, W.G., et al., 2001. Pectin: cell biology and prospects for functional analysis. Plant Mol. Biol. 47 (1
2), 9
27. Available at: http://www.ncbi.nlm.nih.gov/pubmed/11554482. Willats, W.G., Knox, J.P., Mikkelsen, J.D., 2006. Pectin: new insights into an old polymer are starting to gel. Trends Food Sci. Technol. 17 (3), 97
104. Available from: https://doi.org/ 10.1016/j.tifs.2005.10.008. Wrolstad, R.E., Lee, D.D., Poei, M.S., 1980. Effect of microwave blanching on the color and composition of strawberry concentrate. J. Food. Sci. 45 (6), 1573
1577. Available from: https://doi.org/10.1111/j.1365-2621.1980.tb07566.x. Wu, M.L., Zall, R.R., Tzeng, W.C., 1990. Microfiltration and ultrafiltration comparison for apple juice clarification. J. Food. Sci. 55 (4), 1162
1163. Available from: https://doi.org/10.1111/ j.1365-2621.1990.tb01622.x. Yen, G.-C., Lin, H.-T., 1996. Comparison of high pressure treatment and thermal pasteurization effects on the quality and shelf life of guava puree. Int. J. Food Sci. Technol. 31 (2), 205
213. Available from: https://doi.org/10.1111/j.1365-2621.1996.331-32.x. Yıkmı¸s, S., 2016. New approaches in non-thermal processes in the food industry. Int. J. Nutr. Food Sci. 5 (5), 344. Available from: https://doi.org/10.11648/j.ijnfs.20160505.15. Zhou, H.-W., Ben-Arie, R., Lurie, S., 2000. Pectin esterase, polygalacturonase and gel formation in peach pectin fractions. Phytochemistry 55 (3), 191
195. Available from: https://doi.org/ 10.1016/S0031-9422(00)00271-5. Zwijnenberg, H.J., et al., 2002. Native protein recovery from potato fruit juice by ultrafiltration. Desalination 144 (1
3), 331
334. Available from: https://doi.org/10.1016/S0011-9164(02) 00338-7.

Chapter 6

Recovery of Phenolic-Based Compounds From Agro-Food Wastewaters Through PressureDriven Membrane Technologies Roberto Castro-Mun˜oz1,2,3, Carmela Conidi2 and Alfredo Cassano2 1

University of Chemistry and Technology Prague, Prague, Czech Republic, 2Institute on Membrane Technology, ITM-CNR, Rende, Italy, 3Nanoscience Institute of Aragon (INA), University of Zaragoza, Zaragoza, Spain

Chapter Outline Abbreviations 6.1 Introduction 6.2 Membrane-Based Technologies as Emerging Tools for Food Wastewaters Valorization 6.3 Recovery of Phenolic Compounds From Agro-Food Wastewaters 6.3.1 Olive Mill Wastewaters 6.3.2 Artichoke Wastewaters 6.3.3 Wastewaters From Winemaking Industry

195 196

197 199 199 205

6.3.4 Other Food Wastewaters and By-Products 209 6.4 Current Uses of Phenolic-Based Compounds Extracted From Food Wastewaters 214 6.5 Economic Overview of Membrane-Based Technologies in Phenolic Recovery 220 6.6 Conclusions and Future Trends 222 References 222

206

ABBREVIATIONS MF MWCO NF OD RO UF

microfiltration molecular weight cutoff nanofiltration osmotic distillation reverse osmosis ultrafiltration

Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00006-1 © 2019 Elsevier Inc. All rights reserved.

195

196

Separation of Functional Molecules in Food by Membrane Technology

6.1 INTRODUCTION Agro-food wastewaters are mainly organic in nature being constituted by a complex mixture of carbohydrates, proteins, lipids, nutraceuticals, and many other compounds. They are characterized by high biological oxygen demand and chemical oxygen demand (COD) and variations in composition and pH owing to seasonal variations and handling processes. Nowadays, the final disposal of agro-food wastewaters has become a major challenge for food processing industries due to its potential impact on the environment (Ravindran and Jaiswal, 2016). Several methodologies have been proposed for the management of these wastes including decantation, separation, dissolved air flotation, deemulsification, centrifugation, coagulation, flocculation, adsorption, advanced oxidation processes, bioreactors, ozonation, enzymatic treatments, and coupled processes (i.e., electrocoagulation) (Azzam et al., 2004; Cheryan and Rajagopalan, 1998; Inan et al., 2003; Mert et al., 2010; Chedeville et al., 2009; Dammak et al., 2016), all of which aim to reduce the organic matter from the downstream processing wastes. The efficiency, complexity, and cost-effectiveness of these methods may vary significantly. Combined physicochemical and biological systems seem to guarantee high efficiency in terms of pollution control but the great amount of sludge produced remains a significant problem in the management of agro-food wastewaters. Limited resources and increasing interest in the use of bioactive compounds play an important role in the development of sustainable waste management practices. In contrast to the depolluting approaches, the recovery of bioactive compounds from agro-food wastes is one of the most important challenges for sustainable industrial processes. Phenolic-based compounds are of particular interest in food and pharmaceutical industries due to their benefits to human health (antioxidant activity, protection against cardiovascular diseases, diabetes, osteoporosis, and neurodegenerative conditions) (Pandey and Rizvi, 2009); they are common constituents in foods of plant origin such as vegetables (artichoke, olive, maize, etc.), fruits (grapes, apple, pear, cherries, berries, etc.), and other foodstuffs (Scalbert et al., 2005). Nevertheless, there is strong evidence that such valuable compounds are also contained in several agro-food wastewaters (CastroMun˜oz et al., 2016a), including olive mill, artichoke, nixtamalization, and citrus wastewaters, to mention just a few. These wastewaters could represent a new source of phenolic-based compounds and their derivatives that are commonly leached. Target compounds of agro-food wastewaters include dietary fibers, antioxidants, carbohydrates, betalains, anthocyanins, flavonoids, sugars, pectins, proteins, and different phenolic-based compounds (Cassano et al., 2016; Castro-Mun˜oz et al., 2015a, 2017a; Galanakis, 2015). Particularly, for phenolic-based compounds and their derivatives, there is great interest in

Recovery of Phenolic-Based Compounds Chapter | 6

197

identifying new sources and tangible methods for extracting them from such sources. Indeed, some techniques, such as hot-water extraction, solvent extraction, irradiation-assisted extraction, adsorption, ultrasound-assisted extraction, enzyme-assisted extraction, supercritical fluid extraction, and crystallization, can be considered as possible recovery tools. However, their efficiency is limited by the stability of the target compounds (phenolic degradation usually occurs at high temperatures), long extraction times, and safety problems due to the use of organic solvents. Membrane-based technologies, such as microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), and reverse osmosis (RO), represent useful tools for the separation, concentration, and recovery of high-added value compounds, including pectin and polyphenols, from agro-food by-products (Bazinet and Doyen, 2017; Tylkowski et al., 2017). High separation efficiency, easy scale-up, simple operation, high productivity, and absence of phase transition are key advantages of membrane processes over conventional separation technologies. The fractionation method based on the use of membrane technology has been well developed over the past decade in food technology. Indeed, compounds with varying molecular sizes can be retained by different membranes ranging from a macroscopic pretreatment (MF) to the separation and concentration of micromolecules (e.g., by UF and NF) (Galanakis, 2012, 2015; Castro-Mun˜oz et al., 2018). The aim of this chapter is to provide an outlook on the use of membranebased processes for the recovery of phenolic compounds from agro-food wastewaters in the light of recent studies carried out on both laboratory and pilot scale. Case studies referred to the implementation of multistage pressure-driven membrane processes in selected areas of the agro-food production are analyzed and discussed. The proposed purification schemes provide a clear outlook on the potential of these systems toward a sustainable industrial growth within the logic of a process intensification strategy and a zero discharge approach.

6.2 MEMBRANE-BASED TECHNOLOGIES AS EMERGING TOOLS FOR FOOD WASTEWATERS VALORIZATION Technological innovations, including those in clean technologies and processes, aim at introducing in food processing factories advanced wastewater treatment practices able to recycle spent process waters onsite and to reduce the amount of wastewater to be discharged in municipal sewage treatment plants. In addition, the recovery of bioactive compounds from these wastewaters offer new interesting perspectives for their valorization and new market trends in relation to their beneficial effect on human health. Conventional extraction techniques for the purification of natural products include solvent extraction, ultrasound-assisted extraction, pressurized-liquid

198

Separation of Functional Molecules in Food by Membrane Technology

extraction, enzyme-aided extraction, supercritical fluid extraction, resin-based extraction, and alkaline extraction. These extraction methods are characterized by some drawbacks, including the degradation of the target compounds due to high temperatures and long extraction times (as in solvent extractions) and health-related risks. Membrane processes offer very interesting perspectives over conventional technologies in the treatment of wastewaters from food processing industries. In particular, pressure-driven membrane operations have become wellestablished technologies for the separation, fractionation, and concentration of phenolic-based components from agro-food wastewaters (Galanakis, 2012). They provide more advantages for separation than typical methods, such as thermal processes and chromatographic applications, which give low yields at high operational costs. In addition, these processes allow not only the recovery of bioactive compounds, but also their reutilization inside foods. Particularly, UF and NF processes seem to be the most viable technology for application in the food processing industry in the coming future. Many applications have been reported such as in the water softening, vegetable oil processing, wastewater treatment, beverage, dairy (whey processing, lactose recovery, lactic acid separation) and sugar (sugar beet press water, oligosaccharide filtration) industries (Bennett, 2015; Salehi, 2014). In principle, UF and NF processes are able to separate specific compounds through a sieving mechanism based on the molecular weight cutoff (MWCO). However, the membrane’s MWCO is not the only criterion that has to be taken into account. For instance, the asymmetric manufacture of membrane pores does not always reflect a narrow MWCO range. In addition, some other phenomena (e.g., concentration polarization, membrane fouling, coulombic and hydrophobic interactions) also contribute to the phenolic retention (Galanakis, 2013, 2015). Membrane materials and operating parameters are other key factors that affect the membrane performance in terms of productivity (permeation flux) and retention of specific compounds. The combination of different membrane unit operations or between membrane operations and conventional separation technologies (i.e., adsorption, centrifugation, evaporation) offers new and interesting perspectives in this field. This approach has been particularly investigated in recent years for the valorization of agro-food wastewaters (Cassano et al., 2016; Galanakis et al., 2016; Castro-Mun˜oz et al., 2016a). Typically, this recovery strategy involves fractionation using MF, UF, and NF membranes in a sequential design. MF membranes are typically used to remove the suspended solids that can produce operational issues (mainly membrane fouling) in the subsequent operations. While UF supports the removal of the macromolecules in the retentate stream, conserving the phenolic-based compounds and their derivatives in the permeate stream. The NF membranes enhance the recovery through the concentration of the solutes in the retentate (Cassano et al., 2016). This approach enables the recovery of at least 70% of the water volume of the

Recovery of Phenolic-Based Compounds Chapter | 6

199

starting total volume of raw effluents. It has been proposed that, following concentration, the permeate from these narrow membranes can be reused in the industrial process (i.e., as process water, membrane cleaning, and diafiltration operations). Phenolic compounds can be separated, recovered, or concentrated depending on their molecular weight and the membrane used; low molecular polyphenols are normally recovered from NF retentate, displaying phenolic recovery rates from 65% up to 100% (Castro-Mun˜oz et al., 2016a).

6.3 RECOVERY OF PHENOLIC COMPOUNDS FROM AGRO-FOOD WASTEWATERS 6.3.1 Olive Mill Wastewaters Olive mill wastewaters are generated during press, two-phase centrifugal, or three-phase centrifugal olive oil extraction. They appear like a dark liquid with acid reaction containing great quantities of organic substances (sugars, tannins, organic acids, nitrogen substances, and phenolic compounds including polyphenols, polyalcohols, pectins, and lipids) and minerals. These effluents are well known for their significant negative impact on the environment because of their high organic load and also phytotoxic and antibacterial phenolic substances resistant to biological degradation. Intensive research in the field of olive mill wastewater management suggests that these effluents are a useful resource for the recovery of fine chemicals and for different biotechnological applications such as the production of important metabolites (Federici et al., 2009). In particular, olive mill wastewaters are characterized by the presence of more than 30 different types of biophenols and related compounds with tyrosol, hydroxytyrosol, and oleuropein the most important ones (Roig et al., 2006). The chemical structure and molar mass of the main phenolic compounds contained in olive mill wastewaters is illustrated in Fig. 6.1. These compounds are well recognized for their antioxidant, cardioprotective (inhibition of LDL oxidation and platelets aggregation), antiatherogenic, chemopreventive, and antiinflammatory activities (Bendini et al., 2007; Obied et al., 2005). Therefore, biophenols recovered from olive mill wastewaters may be considered a promising nutraceutical compounds for the treatment of oxidative stress-related diseases, which might have interesting applications in cosmetics, nutraceuticals, or functional foods (Cardinali et al., 2012). Until now, different methods for the recovery of polyphenols from olive mill wastewaters, including solvent extraction (Kalogerakis et al., 2013), adsorption onto resins (Frascari et al., 2016; Scoma et al., 2011), cooling crystallization (Kontos et al., 2014), and liquid
liquid solvent extraction (Rahmanian et al., 2014), have been suggested. Unfortunately, almost all tested processes suffer from serious inconveniences such as high cost and low efficiency. Therefore, emerging technologies, such as membrane

200

Separation of Functional Molecules in Food by Membrane Technology

FIGURE 6.1 Chemical structure and molar mass of the main phenolic compounds contained in OMWs.

separation processes, have gained a great attention as promising technologies in the treatment of olive mill wastewaters due to their several advantages such as low energy consumption, no additive requirements, and no phase change (Mudimu et al., 2012; Rahmanian et al., 2014; Conidi et al., 2014a; Cassano et al., 2016). In the last decade the application of sequential membrane processes in decreasing order of MWCO has been strongly investigated for the selective recovery, purification, fractionation, and concentration of biophenols from olive mill wastewaters (Gebreyohannes et al., 2016). A sequential membrane design based on the use of MF, UF, and RO membranes for total recovery of polyphenols, water, and organic substances from olive mill wastewaters was investigated by Russo (2007). The raw wastewaters were directly submitted to MF without a preliminary centrifugation step. This step was considered as critical section of the process due to severe fouling and difficulties in the cleaning procedure. UF on the MF permeates was realized using polymeric membranes with MWCO of 80, 20, and 6 kDa and on UF 6 kDa permeate with a ceramic membrane of 1 kDa. The processing resulted in permeates with different phenolic fractions containing mainly hydroxytyrosol (134,879
266,679 ppm), tyrosol (7968
11,218 ppm), oleuropein (7765
26,698 ppm), caffeic acid (10,570
21,982 ppm), and protocatechuic acid (8871
22,601 ppm). The

Recovery of Phenolic-Based Compounds Chapter | 6

201

ceramic membrane of 1 kDa and the polymeric one of 6 kDa showed no differences on the selectivity of hydroxytyrosol. MF/UF permeates were concentrated by RO. The RO concentrate, containing enriched and purified low MW polyphenols, was considered of interest for food, pharmaceutical, or cosmetic applications while the RO permeate (about 50%
60% of the initial volume) was suggested as suitable for beverage formulations. Paraskeva et al. (2007) evaluated the use of UF, NF, and RO operations for the treatment and fractionation of olive mill wastewaters. The UF process allowed to remove high molecular weight constituents, including suspended solid particles, from the raw wastewater pretreated with an 80-μm polypropylene screen. More than 95% of phenolic compounds were removed through the following NF step. Better efficiency of the olive mill wastewater treatment was achieved applying RO after UF. A sequential combination of rough filtration, MF, UF, NF, and RO, was implemented by Villanova et al. (2009) for preparing tyrosol and/or hydroxytyrosol from olive mill wastewaters. The process allows recovering, after concentration, at least 1 g/L of hydroxytyrosol and 0.6 g/L of tyrosol. These components can then be isolated with purity higher than 98% by chromatography in inverted phase on a preparatory column. The process also allows to recover at least 70% of the water volume as to the starting total volume of the raw effluent, with a quality within the legal limits (lower than 100 mg O2/L of COD), which allows for its agricultural or civil reuse. Galanakis et al. (2010) clarified olive mill wastewaters by using four different UF membranes in the range of 2
100 kDa, showing that a 25 kDa membrane was highly efficient for the removal of the heavier fractions of hydroxycinnamic acids and flavonols. By using this membrane, almost all of the initial phenolic compounds were separated and recovered in the permeate stream (retention 10%). In particular, o-diphenols, hydroxycinnamic derivatives, and flavonols retentions were equal to 6%, 32%, and 37%, respectively, for the 25-kDa membrane, and significantly higher (32%, 44%, and 56% as well as 48%, 53%, and 62%, respectively) for membranes of 2 and 10 kDa; while the use of an NF membrane with a MWCO of 120 Da led to reject around 70% of the phenolic compounds (including 99% recovery of the initial hydroxycinnamic acids and flavonols). Flat-sheet UF membranes with different MWCO (4, 5, and 10 kDa) and polymeric material (regenerated cellulose and polyethersulfone) have been evaluated for their retention coefficients toward phenolic compounds, total antioxidant activity, and total organic carbon (Cassano et al., 2011). All selected membranes showed lower rejection toward free low molecular weight phenolic compounds in comparison with values observed for total polyphenols. This was in agreement with the molecular weight of the investigated phenols, which is in the range of 138
284 g/mol and hence lower than the MWCO of each UF membrane.

202

Separation of Functional Molecules in Food by Membrane Technology

Regenerated cellulose membranes exhibited lower rejections toward phenolic compounds, higher permeate fluxes, and lower fouling index when compared with polyethersulfone membranes. Indeed polyphenols, also aggregated with polysaccharides, have a higher affinity for the polyethersulfone membranes leading to severe fouling by pore narrowing and blocking under UF conditions. Polar interactions, such as van der Waals and electron donor
acceptor interactions, and multiple hydrogen bonds toward the additive polyvinylpyrrolidone used in the manufacture of polyethersulfone membranes, seem to be the main mechanisms involved in the adsorption of polyphenols to polyethersulfone membranes (Ulbricht et al., 2009). The fractionation of olive mill wastewaters by combining UF and NF membranes has been investigated by Cassano et al. (2013). In particular, a sequence of two UF processes with membranes of 0.02 μm and 1 kDa, respectively, followed by a final NF step with a thin-film composite membrane in spiral-wound configuration was investigated on laboratory scale operating in selected process parameters. Suspended solids were completely removed in the first UF step, while most of the organic compounds were removed in the following UF step. Phenolic compounds were recovered in the permeate stream of both UF processes. Their yields can be increased working at high volume reduction factors and through a diafiltration step, consisting of the addition of water to UF concentrates. The proposed design was capable to fractionate the wastewater in three valuable fractions: 1. a concentrated fraction (UF retentate) with a high content of organic substances suitable for other biotechnological applications (e.g., biogas production); 2. a concentrated fraction (NF retentate) rich in phenolic substances (c. 960 mg/L) suitable for food, cosmetic, and pharmaceutical applications; and 3. an aqueous stream (NF permeate) with a low total organic carbon content (95 mg/L) that can be reused in different ways within the olive oil extraction process (process water) or in the membrane process treatment (cleaning of membranes, diafiltration) (Fig. 6.2). The separation of phenolic micromolecules depends on the narrow pore size of the membrane used, but the nature of the molecules is also crucial for such recovery efficiency. Phenolic-based compounds present aromatic rings and aliphatic chains that produce a hydrophobic profile. This leads to the attraction of water molecules, an increased volume of the molecules, and the restriction of permeation through membrane pores due to the “polarity resistance” phenomenon (Galanakis, 2015). Garcia-Castello et al. (2010) analyzed the potential of an integrated membrane system for the recovery of phenolics from olive mill wastewaters through a combination of MF and NF membranes followed by a concentration step performed by using osmotic distillation (OD). In their study, almost

Recovery of Phenolic-Based Compounds Chapter | 6

203

Flocculant

Olive mill wastewater

Prefiltration (5 μm) Solids

pH adjustment

Enzymatic treatment

UF (Etna 01PP)

UF permeate

Process water Water for diafiltration Membrane cleaning

NF permeate

Composting Anaerobic digestion

UF permeate

UF retentate (suspended solids, polysaccharides, proteins, etc.)

UF (HSF)

NF (NF90) NF retentate

Nutraceuticals Food supplements Pharmaceuticals Cosmetics

FIGURE 6.2 Integrated membrane process used for the fractionation of OMWs. Adapted from Cassano, A., Conidi, C., Giorno, L., Drioli, E., 2013. Fractionation of olive mill wastewaters by membrane separation techniques. J. Hazard. Mater., 248, 185
193.

all of the initial phenolics were recovered (319 mg/L) in the permeate of the NF step after a preliminary MF of raw wastewaters devoted to the removal of suspended solids. The rejection of the NF membrane toward low molecular weight polyphenols was about 5%. The lowest rejection, of about 1%, was observed for oleuropein, while the rejection for the protocatechuic acid was of about 21%. A concentrated solution containing about 0.5 g/L free low molecular weight polyphenols, with hydroxytyrosol representing 56% of the total, was obtained by treating the NF permeate by OD. In the three-phase membrane systems investigated by Servili et al. (2011a), fresh olive mill wastewaters were pretreated by MF and then the MF permeate was submitted to an UF treatment with a 7-kDa membrane. The final concentration of the UF permeate was obtained by RO. The membrane treatment produced a crude phenolic concentrate with final volumes between 20% and 25% of the original olive mill wastewater. The main phenols found in the final concentrate were the same as those in the original raw wastewater, but their concentrations were fourfold higher in the concentrate stream. The most abundant compounds in the RO retentate were

204

Separation of Functional Molecules in Food by Membrane Technology

verbascoside and 3,4-DHPEA-EDA, a dialdehydic form of decarboxymethyl elenolic acid linked to hydroxytyrosol. The membrane treatment produced a strong reduction of the organic load, with a drawdown of 98%. Similarly, Zagklis et al. (2015) developed a complex integrated system for the separation of low molecular weight polyphenols based on the use of UF, NF, and RO operations in a sequential design. The NF permeate, containing 246 mg/L of hydroxytyrosol, was concentrated up to 558 mg/L through RO. This final retentate was purified by adsorption, which led to obtaining an extract with a high hydroxytyrosol content (around 974 mg/L). Bazzarelli et al. (2016) processed olive mill wastewaters through an integrated membrane process which included MF, UF, NF, OD, and membrane emulsification (ME) steps. Particularly, the NF retentate (with phenolic content of 12.5 g/L) was concentrated by OD unit, obtaining a phenolic-enriched OD retentate (87.5 g/L). The final ME step was used to encapsulate the final recovered product (encapsulation efficiency 90%). According to the estimated mass balance of the process, 80 L of MF retentate, 800 L of purified water stream as NF permeate and 16.7 L of OD retentate, containing an enriched fraction of phenolic compounds, are produced from 1 m3 of raw wastewaters. In a previous work, Bazzarelli et al. (2015) developed a new pretreatment approach of olive mill wastewaters based on their acidification until the isoelectric point. Ideal destabilization conditions occurred at pH 1.8, corresponding to the lowest zeta potential value and to the formation of large agglomerates with a mean particle size equal to 21 μm. This method allowed obtaining of higher MF and UF permeate fluxes respect to severe and/or expensive conventional pretreatment processes, using very mild operating conditions. Zirehpour et al. (2015) evaluated the filtration performance of three spiral-wound NF membranes with different properties (NF270 and NF90 from Dow-Filmtec and an NF self-made membrane) for the purification of olive mill wastewaters previously prefiltered by MF and UF membranes. Among the selected membranes, the NF-270 exhibited the highest permeate flux at all operating pressures, except 20 bar. On the other hand, both NF-90 and self-made NF membranes showed COD rejection efficiencies better than the NF-270. In addition, the contributions of both irreversible and reversible fouling were different for all selected membranes and operating pressures investigated. The higher flux decline and contribution of irreversible fouling of the NF-self membrane was attributed to its higher MWCO. The control of fouling is one of the problems that still slow down large-scale membrane applications with respect to olive mill wastewater management. Properly tailored pretreatment processes and the use of the critical and threshold flux theories are important factors to ensure the cost-effectiveness of the membrane treatment when transferred to the industrial scale (Ochando-Pulido and Martinez-Ferez, 2015).

Recovery of Phenolic-Based Compounds Chapter | 6

205

pH
temperature flocculation, either stand-alone or followed by photocatalysis by novel lab-made titanium dioxide nanoparticles under irradiation of UV light (Ochando-Pulido et al., 2012a) and a sequence of Fenton-like advanced oxidation, flocculation
sedimentation and filtration through olive stones (Ochando-Pulido et al., 2012b), have been proposed as specific pretreatments essential for fouling inhibition protocols.

6.3.2 Artichoke Wastewaters Artichoke by-products are characterized by marked antioxidative effects due to their special chemical composition, which includes high levels of phenolic compounds with a wide range of caffeoylquinic acid derivatives and flavonoids such as apigenin-7-O-glucoside and luteolin (Christaki et al., 2012). Conidi et al. (2014b) investigated the use of UF and NF membranes in the treatment of artichoke wastewaters to separate and concentrate phenolic compounds from sugars and other impurities. The UF pretreatment allowed to remove suspended solids from the raw extract. The clarified solution was treated with two different NF membranes (Desal DL, cross-linked aromatic polyamide, 150
300 Da, from GE Osmonics and NP030, polyethersulfone, 400 Da, from Microdyn Nadir). Both membranes showed high rejections toward phenolic compounds (chlorogenic acid, cynarin, and apigenin-7-Oglucoside). On the other hand, the Desal DL membrane exhibited a total rejection toward sugar compounds (glucose, fructose, and sucrose) when compared with the NP030 membrane (4.0%). On the basis of experimental results, an integrated membrane process was proposed and validated on laboratory scale by using a sequential combination of UF and NF membranes. The proposed system allowed to produce an enriched fraction of phenolic compounds (retentate of NP030 membrane), an enriched fraction of sugars (retentate of Desal DK membrane, from GE) and an aqueous stream (permeate of Desal DK membrane) suitable for membrane cleaning or as process water (Cassano et al., 2015). Higher fouling indexes detected for polyethersulfone membranes were attributed to a greater adsorption of phenolic compounds on these membranes due to benzene ring
benzene ring interactions as well as to the formation of hydrogen bonds between the additive polyvinylpyrrolidone usually used in the manufacturing process and phenolic compounds (Susanto et al., 2009). In another approach Conidi et al. (2015) evaluated the performance of an integrated process based on the use of membrane operations such as UF and NF and adsorbents resins, in a sequential form, for the concentration and purification of phenolic compounds from artichoke wastewaters coming from the blanching step. Artichoke wastewaters were previously pretreated by using a 15-kDa ceramic membrane producing a clarified solution depleted in suspended solids and macromolecular compounds. The UF permeate was

206

Separation of Functional Molecules in Food by Membrane Technology

processed by NF producing a retentate stream containing about 1.6 g/L of chlorogenic acid and 0.3 g/L of apigenin 7-O-glucoside. The final retentate of the NF process was processed by using three different macroporous resins based on polystyrene (Lewatit S 6328 A, Lewatit S 2328, and Lewatit S 7968, all from Lanxess). Among the three different tested resins, the S 7968 offered the best performance in terms of adsorption/desorption ratio for chlorogenic acid, while for the apigenin 7-O-glucoside the S 7968 and the S 2328 resins showed a total adsorption
desorption yield in the range 68.31%
78.45%. Sugar compounds were quite totally recovered in the desorbed fraction independently of the type of resin. The global results indicated the combination of UF and NF membranes with an adsorption/desorption system produces a more purified fraction of phenolic compounds if compared with an integrated system fully based on the use of membranes. Recently, a sustainable extractive technology followed by membrane separation methods has been applied to olive and artichoke by-products (leaves, pitted olive pulp, and stems) to obtain standardized commercial extracts for application in the functional food industry, pharmaceutical, and cosmetic fields (Romani et al., 2016). The process is based on a water extraction of the raw material followed by MF (only for olive residues), UF (only for artichoke residues), NF (only for olive residues), RO and final concentration by evaporation at low temperature or spray-dried technique.

6.3.3 Wastewaters From Winemaking Industry Grape is a rich source of numerous classes of natural products including polyphenols such as flavonols (quercetin, kaempferol, myricetin), flavanols (catechins, epicatechins), anthocyanins (malvidin 3-O-glucoside, peonidin 3-O-glucoside, petunidin 3-O-glucoside), and stilbenes (resveratrol) (Xia et al., 2010). During winemaking, only a small part of these compounds is transferred from grape to wine, while large quantities remain in the pomace, the by-product consisting of pressed grape leftovers (e.g., seeds, skin, stems). The annual disposal of grape pomace worldwide represents a serious environmental and economic problem. On the other hand, it is an inexpensive source of beneficial phytochemicals, which could be successfully used in the food, cosmetic, and pharmaceutical industries (Ðilas et al., 2009). Dı´az-Reinoso et al. (2009) evaluated the performance of different UF and NF membranes with MWCO in the range 150
1000 Da for obtaining fractions enriched in antioxidant compounds from aqueous extracts of pressed distilled grape pomace. All selected membranes presented similar rejections of total phenolics and sugars, and were suitable for concentration purposes. The phenolic content in retentates was increased by factors of 3
6 with respect to the feed solution.

Recovery of Phenolic-Based Compounds Chapter | 6

207

A combination of membrane processes (UF and NF) and polymeric resins was also investigated for obtaining isolates of high antioxidant activity from distilled grape pomace pressing liquors (Dı´az-Reinoso et al., 2010). Commercial polymeric resins (Sepabeads SP700) were employed in the adsorption step, and desorption was carried out with aqueous ethanol solutions. Experimental results indicated that the adsorption percentage of phenolic compounds was higher for permeates than for original liquors. Winery sludge is another derivative by-product generated by winemaking industry during wine decanting. Santamarı´a et al. (2002) assayed the combination of membranes with different MWCO to fractionate phenolics (gallic acid, catechin, gallates, etc.) from defatted milled grape seeds, using acetone
water mixtures. A fractionation sequence based on the use of UF and NF membranes was developed for purification of fractions of proanthocyanidins with different MW. UF membranes were also used to separate phenolic fractions from pectins contained in winery sludge. Polysulfone membranes with MWCO of 20 and 100 kDa were able to separate phenolic compounds from pectin fractions. However, they were not able to fractionate different phenolic classes and sugars (reducing or not) which were retained in rather high percentages. On the other hand, fluoropolymer membranes of 1 kDa separated different phenolic classes like hydroxycinnamic acids, flavonols, and anthocyanins on the basis of polarity. These results confirmed that the polarity of the solutes plays an important role in their separation. For example, o-diphenols are more polar (negative) molecules than other polyphenols due to the presence of more hydroxyl groups (Galanakis, 2015). Phenolic-based compounds were also extracted from grape seeds by using a solvent extraction method based on the use of 50% ethanol and 50% water and then concentrated by UF (Nawaz et al., 2006). In this work, 0.2 g/mL solid to liquid ratio, double stage extraction, and 0.22-μm membrane pore size were found as optimal conditions of extraction and concentration. Under these conditions, the maximum amounts of polyphenols (11.4% of the total seeds weight) were recovered from grape seeds. The procedure provided high extraction rates, high extraction selectivity, short extraction time, and significant labor savings. An extraction and purification scheme of phenolic compounds from grape seeds is depicted in Fig. 6.3. In this approach the purification process devoted to the removal of undesirable compounds such as fats, proteins, and sugars, traditionally performed by adsorption chromatography, is substituted with a sequential combination of membrane operations. The use of UF and NF membranes with different MWCO allows to obtain fractions enriched in target phenolic compounds with different degree of polymerization (i.e., recovery of monomeric and oligomeric compounds). The removal of solvent can be performed by RO, which is economically favorable up to an osmotic pressure difference of about 40 bar, above which

208

Separation of Functional Molecules in Food by Membrane Technology

FIGURE 6.3 Extraction and purification scheme of phenolic compounds from grape seeds. Adapted from Crespo, J.G., Brazinha, C., 2010. Membrane processing: natural antioxidants from winemaking by-products. Filtration & Separation, 47(2), 32
35.

evaporation is more competitive. The hydroalcoholic stream, enriched in the target phenolic compounds, is dried by evaporative techniques to obtain a final product with the specified characteristics (Crespo and Brazinha, 2010). An integrated sedimentation/UF process for the recovery of polyphenols and polysaccharides from winery effluent generated in the second racking was investigated by Giacobbo et al. (2013a). The UF membrane (GR95PP, supplied by Alfa Laval), with a MWCO of 7600 Da exhibited high rejection coefficients to polysaccharides and to polyphenols independently by the operating pressure. At pH 3.6 rejections for these compounds were of 99.3 and more than 93.5, respectively. The authors also evaluated the performance of different NF membranes (both laboratory-made and commercial membranes) in the fractionation of the UF permeate (Giacobbo et al., 2013b). Among the selected membranes the NF270 membrane (200
300 Da, from Filmtec) presented the highest rejection coefficients for all investigated parameters and the ETNA01PP membrane (1 kDa, from Alfa-Laval) the lowest ones. This membrane presented also the largest gap between the rejection coefficients to polysaccharides and polyphenols.

Recovery of Phenolic-Based Compounds Chapter | 6

209

The use of MF membranes was also investigated as pretreatment of winery effluents for the removal of suspended solids and the achievement of a limpid permeate enriched in phenolic compounds to be concentrated by NF (Giacobbo et al., 2015, 2017a). The use of diluted effluents allowed to increase both the permeation flux through MF membranes and the recovery rate of polyphenols and polysaccharides in the permeate. This phenomenon was attributed to the intermolecular interactions between polyphenols and polysaccharides, which improve the polyphenols’ solubility in water, increasing the aqueous extraction and, consequently, the recovery in the permeate. MF membranes with larger pore size showed lower permeate fluxes and consequently, more severe fouling. The MF permeate was fractionated by using a sequence of UF and NF membranes with decreasing MWCO. UF membranes with MWCO of 1 and 10 kDa (Etna 01PP and 10PP, from Alfa-Laval) presented rejections to polysaccharides higher than 77% and to polyphenols nearby 50%, showing that the polyphenols preferably permeate these UF membranes while the polysaccharides are mainly retained. On the other hand, the NF270 membrane showed rejections greater than 90% to polyphenols, 99% to polysaccharides, and full rejection to anthocyanins with lower rejections to total organic carbon (42%) and conductivity (75%). Based on these results an integrated membrane process was proposed for the fractionation of winery effluents generated in the second racking (Giacobbo et al., 2017b). The proposed design is capable to fractionate the effluent in three valuable fractions: 1. a concentrated UF fraction enriched in polysaccharides; 2. a concentrated NF fraction enriched in bioactive compounds, with potential applications in the pharmaceutical, cosmetic and food industries; and 3. an NF permeate stream reusable as process water (Fig. 6.4).

6.3.4 Other Food Wastewaters and By-Products The separation, purification, and concentration of bioactive compounds through the use of membrane operations has been investigated also for other types of agro-food wastewaters. Aguiar Prudeˆncio et al. (2012) investigated the concentration of phenolic compounds from mate bark aqueous extract by using a spiral-wound NF membrane with a MWCO of 150
300 Da (HL2521TF, from GE Osmonics). The major compounds detected in the extract and its concentrates (volume reduction factor of 4 and 6) were chlorogenic acid and epigallocatechin gallate, which are strongly correlated to the high antioxidant activity of mate. A combination of UF and diafiltration was investigated by Xu et al. (2004) to separate isoflavones from proteins and other components of soy processing waste stream. UF was performed at 50 C by using a regenerated cellulose membrane with a MWCO of 30 kDa in (Millipore Ltd). The

210

Separation of Functional Molecules in Food by Membrane Technology

Wine lees from 2° racking

Retentate (suspended colloidal matter)

Retentate (polyphenols)

Retentate (polysaccharides) NF (150–300 Da)

MF (0.4 μm)

Permeate

UF (1–10 kDa)

Permeate (reuse on dilution process)

FIGURE 6.4 Integrated membrane process for the fractionation of wine lees. Adapted from Giacobbo, A., Meneguzzi, A., Bernardes, A.M., De Pinho, M.N., 2017a. Pressure-driven membrane processes for the recovery of antioxidant compounds from winery effluents. Journal of Cleaner Production, 155, 172
178.

concentrated soymilk was then continuously diafiltered with water to improve the isoflavone yield in the permeate. The UF permeate was then ultrafiltered by using a much tighter regenerated cellulose membrane with a MWCO of 1 kDa. Isoflavones were finally concentrated (10
15 times) by treating the permeate stream with a thin-film RO membrane (Desal-3, from GE Osmonics). The yield of the total amount of isoflavones in the RO retentate was of approximately 50%. A sustainable process for the production of a natural extract from spent coffee using membrane technology, with no organic solvents or adsorbents involved, was developed by Brazinha et al. (2015). Aqueous extracts were produced in optimized extraction conditions aimed at achieving extracts with maximized concentrations and yields of phenolic compounds and then submitted to a fractionation process based on the use of UF (GH, polyamide, 2.5 kDa from GE Osmonics) and NF (GE, polyamide, 1 kDa from GE Osmonics and NP010, polyethersulfone, 1 kDa from Microdyn Nadir) membranes. The observed rejections of the total phenolic compounds were high with the NP010 membrane exhibiting the highest value (about 95%) indicating that most of the total phenolic compounds correspond to large molecules. Low rejection values were detected for caffeine (in the range 13%
26%) with the GH membrane exhibiting the lowest one. A schematic diagram of the process is depicted in Fig. 6.5. An integrated process based on the use of NF and RO membranes was developed by Almanasrah et al. (2015) for recovering, fractionating, and purifying valuable phenolic compounds (mainly catechin and its derivatives) from small sugars (glucose, fructose, and sucrose) present in carob byproducts. Diananofiltration was used to enable the removal of sugars and gallic acid from higher molecular catechin and its derivatives. According to the proposed flowchart two different valuable products were obtained: a purified extract rich in catechin for the profitable nutraceuticals market and a purified extract rich in small sugars for the food industry. This approach, which

Recovery of Phenolic-Based Compounds Chapter | 6

211

FIGURE 6.5 Implementation of membrane operations in the production of fractionated extracts from wet spent coffee. Adapted from Brazinha, C., Cadima, M., Crespo, J.G., 2015. Valorisation of spent coffee through membrane processing. J. Food Eng., 149, 123
130.

fulfills the zero discharge principle, may be easily extended to other agroindustrial by-products. Ruby-Figueroa et al. (2012) analyzed the effect of operating conditions on the polyphenols rejection of UF membranes in the clarification of orange press liquor, a by-product of the citrus processing industry. The liquor was

212

Separation of Functional Molecules in Food by Membrane Technology

clarified by using a polysulfone hollow fiber membrane with a MWCO of 100 kDa and the response surface methodology was used to analyze the interaction between the investigated parameters. Polyphenols rejection of 28.45% and total antioxidant activity of 32.28 mM Trolox in the clarified liquor were estimated, respectively, in optimized operating conditions of transmembrane pressure (0.2 bar), temperature (19.85 C), and feed flowrate (244.64 L/h). These conditions are able to minimize concentration polarization and fouling phenomena. In a previous study, the authors estimated permeate fluxes of 23.7 kg/m2h and fouling index of 48.0%, respectively, in optimized operating variables of 1.4 bar, 15 C, and 167.7 L/h through the same response surface methodology approach (Ruby-Figueroa et al., 2011). Different spiral-wound NF membranes with MWCO ranging from 250 to 1000 Da were evaluated for their performance in the treatment of orange press liquors (Conidi et al., 2012). In particular, the rejection of these membranes toward anthocyanins, flavonoids, and sugars was evaluated to identify a suitable membrane to separate phenolic compounds from sugars. A strong reduction of the average rejection toward sugar compounds was observed by increasing the MWCO of the selected membranes, while for anthocyanins rejections were higher than 89%, independent of the pore size. Among the investigated membranes, the NFPES10 membrane (1000 Da, from MicrodynNadir) gave the lowest average rejection toward sugar compounds and high rejections toward both anthocyanins (89.2%) and flavonoids (70%). This behavior was attributed to the positive charge of anthocyanins at the acidic pH of the orange press liquor (3.4) and the positive charge of the NF membrane at this pH value. Indeed, the zeta potential of the NF PES10 membrane at pH 3 was estimated to be 1 mV (Boussu et al., 2008). Consequently, the electrostatic repulsion, independent of the pore size of the NF membrane, contributes to the high average rejection of the membrane toward anthocyanins. An integrated membrane process for the recovery of flavonoids from orange press liquors was also investigated by Cassano et al. (2014). The liquor was previously clarified by UF and then preconcentrated by NF up to 32 Brix by using the NFPES10 membrane. A concentrated solution at 47 Brix was produced from the NF retentate by using an OD apparatus equipped with polypropylene hollow fiber membranes. Flavanones and anthocyanins were highly rejected by the NF membrane, producing a permeate stream of 4.5 Brix. The concentration of flavanones and anthocyanins in the NF retentate increased by increasing the volume reduction factor of the process. The final concentration of flavonoids by OD produced a concentrated solution of interest for nutraceutical and pharmaceutical applications. Membrane processes have also been investigated for the recovery of phenolic compounds from bergamot juice (Conidi et al., 2011; Conidi and Cassano, 2015). Bergamot is a citrus fruit derived from bitter orange and lemon essentially used for the production of essential oil widely used in the

Recovery of Phenolic-Based Compounds Chapter | 6

213

pharmaceutical, cosmetic, and food industries. The juice due to its bitter taste has not found a real use in the food industry and it is considered a waste of the essential oil production with serious problems for its disposal. On the other hand, phenolic compounds of the juice have a great potential as active principles in the pharmaceutical industry and as antioxidant compounds in the food industry. In particular, some flavonoids (i.e., brutierdin and melitidin) have been recognized for their anticholesterolemic activity (Di Donna et al., 2009). In the work of Cassano et al. (2011), authors investigated the potential of UF and NF membranes with MWCO from 450 to 1000 Da in the separation and concentration of phenolic compounds from the depectinized juice previously clarified by UF. The best separation of phenolic compounds and sugars occurred by using an NF membrane with a nominal MWCO of 450 Da (a monotubular ceramic membrane from Inopor). The NF permeate was a clear solution enriched in sugar and organic acids; phenolic compounds were recovered, instead, on the retentate side as also confirmed by the high total antioxidant activity of the NF retentate. In another work, Conidi and Cassano (2015) evaluated the performance of three polymeric membranes (NFPES10 and N30F, both from Microdyn Nadir and NF270 from Dow-Filmtec) in terms of productivity and selectivity toward target compounds (including flavonoids and total phenolic compounds) of ultrafiltered bergamot juice. Experimental results indicated that the NFPES10 membrane gave the lowest average rejection toward sugar compounds (35%) and high rejections toward both phenolic compounds (71%) and flavonoids (between 88.4% and 90.1%). This behavior was attributed to the formation of fouling layers as confirmed by the high fouling index measured for this membrane (75.4% in comparison to 21.4% of the NF270 membrane). Recently, Castro-Mun˜oz and Ya´n˜ez-Ferna´ndez (2015) analyzed the potential of an integrated membrane system for the fractionation of nixtamalization wastewaters (also known as nejayote) generated by the maize processing industry (Dı´az-Montes et al., 2016; Castro-Mun˜oz et al., 2017b). Typically, these wastewaters contain significant amounts of soluble and insoluble solids including arabinoxylans, dietary fiber, and phenolic-based compounds (e.g., hydroxycinnamic, p-cumaric, ferulic, dehydrodiferulic, and dehydrotriferulic acids) (Ayala-Soto et al., 2014). The nejayote extract was preliminarily submitted to an MF process to recover parts of the grain (fiber, endosperm, pericarp) and suspended solids of high molecular weight. The MF retentate was suggested for its reuse as carbon source for a biotechnological process to produce biogas, bioethanol, or other types of biotechnological products (Castro-Mun˜oz et al., 2015b). The MF permeate was processed by using a 100 kDa UF membrane (UFP-100-E-4A, Amersham Biosciences Corp.) in hollow fiber configuration. The concentrated solution enriched in carbohydrates is of potential interest as food additive (Castro-Mun˜oz et al.,

214

Separation of Functional Molecules in Food by Membrane Technology

FIGURE 6.6 Integrated membrane process proposed for the fractionation of NWs. Adapted from Castro-Mun˜oz, R., Ya´n˜ez-Ferna´ndez, J., 2015. Valorization of nixtamalization wastewaters by integrated membrane process. Food Bioprod. Process., 95, 7
18.

2015c). The UF permeate was then submitted to a final UF process performed by using a hollow fiber membrane module with a MWCO of 1 kDa (UFP-1-E-4A, Amersham Biosciences Corp.) producing two different streams: a concentrated solution containing components of soluble calcium (retentate), which can be reused in the nixtamalization process of maize and a clear solution enriched in polyphenols (permeate) with high antioxidant activity of interest for cosmetic, food, and/or pharmaceutical industries (Castro-Mun˜oz et al., 2016b). Samples obtained in the investigated process are depicted in Fig. 6.6. Table 6.1 summarizes the main phenolic-based compounds recovered from different agro-food wastewaters by using membrane-based technologies; it also shows some specifications of the process type (single or integrated membrane process) and membrane types used for such recovery.

6.4 CURRENT USES OF PHENOLIC-BASED COMPOUNDS EXTRACTED FROM FOOD WASTEWATERS Despite their limited postrecovery applications, some studies have been addressed to the potential reuse of phenolic compounds recovered from agrofood wastewaters. Esposto et al. (2015) evaluated the effect of phenolic extracts recovered from a by-product of the olive oil process on the quality

TABLE 6.1 Phenolic-Based Compounds Recovered From Agro-Food Wastewaters Using Membrane-Based Technologies Target Compounds

Wastewater

Membrane Process

Pore Dimension/Material/Configuration

References

Phenolic compounds

Table olive processing wastewaters

UF

30 kDa/PES/flat sheet

Garcia-Ivars et al. (2015)

Phenolic compounds

Winery effluents

MF

0.5 μm/PVDF/flat sheet; 0.5 μm/fluoropolymer/flat sheet; 0.4 μm/PI/hollow fiber

Giacobbo et al. (2015)

Phenolic compounds

Winery effluents

UF/UF/NF

UF: 10 and 1 kDa/composite fluoropolymer/flat sheet; NF: 200
300 Da/Polypiperazine thin-film composite/flat sheet

Giacobbo et al. (2017b)

Phenolic compounds

Orange press liquor

UF

100 kDa/PS/hollow fiber

Ruby-Figueroa et al. (2011, 2012)

Phenolic compounds

Nixtamalization wastewaters

MF/UF/UF

MF: 0.2 μm/PS/hollow fiber; UF: 100 and 1 kDa/PS/hollow fiber

Castro-Mun˜oz and Ya´n˜ez-Ferna´ndez (2015)

Phenolic compounds

Olive mill wastewaters

UF/NF/RO

UF: 100 nm/zirconia/tubular; NF: 200 Da/ polymeric/spiral-wound; RO: 200 Da/polymeric/ spiral-wound

Paraskeva et al. (2007)

Phenolic compounds

Grape seeds

MF

0.45 and 0.22 μm/mixed cellulose esters/flat sheet

Nawaz et al. (2006)

Phenolic compounds

Fermented grape pomace

UF/NF

UF: 1000 Da/TFC/spiral-wound; UF: 1000 Da/ titania/tubular; NF: 350 and 250 Da/PA
PS/spiralwound; NF: 150
300 Da/thin-film/spiral-wound

Dı´az-Reinoso et al. (2009) (Continued )

TABLE 6.1 (Continued) Target Compounds

Wastewater

Membrane Process

Pore Dimension/Material/Configuration

References

Hydroxytyrosol, protocatechuic acid, caffeic acid, tyrosol, and p-cumaric acid

Olive mill wastewaters

MF/UF

MF: 0.2 μm/PP/tubular; UF: 10 and 4 kDa/PES/flat sheet; UF: 10 and 5 kDa/regenerated cellulose/flat sheet

Cassano et al. (2011)

Hydroxycinnamic acids, o-diphenols

Winery sludge

UF

100 and 20 kDa/PS/flat sheet; 1 kDa/composite fluoropolymer/flat sheet

Galanakis et al. (2013)

3,4-DHPEA, p-HPEA, 3,4-DHPEAEDA, verbascoside, and total phenols

Olive mill wastewater

MF/UF/RO

MF: 0.3 μm/PP/tubular; UF: 7 kDa/PA
PS/spiralwound; RO: 100 Da/PS/spiral-wound

Servili et al. (2011a,b)

Hydroxytyrosol, protocatechuic acid, caffeic acid, tyrosol, oleuropein, and p-cumaric acid

Olive mill wastewaters

MF/UF/MBR

MF: 0.2 and 0.45 μm/CA/flat sheet; MF: 0.2 μm/ nylon/flat sheet; MF: 0.2 μm/PVDF/flat sheet; UF: 10 kDa/regenerated cellulose/flat sheet; MBR: 30 kDa/PS/hollow fiber

Conidi et al. (2014a)

Chlorogenic acid, cynarin, apigenin-7-O-glucoside

Artichoke wastewaters

UF/NF

UF: 50 kDa/PS/hollow fiber; NF: 400 Da/PES/spiralwound; NF: 150
300 Da/PA/spiral-wound

Conidi et al. (2014b)

Gallic acid, chlorogenic acid, and epigallocatechin gallate

Residues from mate tree

NF

150
300 Da/TFC/spiral-wound

Aguiar Prudeˆncio et al. (2012)

Free low MW polyphenols, hydroxytyrosol, protocatechuic acid, tyrosol, oleuropein, and caffeic acid

Olive mill wastewaters

MF/UF/RO

MF: 0.45 and 0.8 μm/zirconium oxide/tubular; MF: 500 kDa/PES/spiral-wound; UF: 80 and 20 kDa/PS/ spiral-wound; UF: 6 kDa/PES/spiral-wound; UF: 1 kDa/zirconium oxide/tubular; RO: 99.5% salt rejection/PA/spiral-wound

Russo (2007)

Proanthocyanidins

Defatted milled grape seeds

UF/NF

UF: 200 kDa/PVDF/tubular; UF: 20 and 8 kDa/PS/ tubular; UF: 4 kDa/PES/tubular; RO: 60% CaCl2 rejection/PA/tubular

Santamarı´a et al. (2002)

Hydroxytyrosol, procatechin acid, catechol, tyrosol, caffeic acid, p-cumaric acid, and rutin

Olive mill wastewaters

UF/NF

0.02 μm/PVDF/hollow fiber; 1 kDa/composite fluoropolymer/flat sheet; NF: salt rejection .97%/thin-film composite/spiral-wound

Cassano et al. (2013)

Isoflavones (aglycone and glucoside)

Soy processing waste

UF

30 and 1 kDa/regenerated cellulose/spiral-wound

Xu et al. (2004)

Hydroxytyrosol, procatechin acid, tyrosol, caffeic acid, p-cumaric acid, oleuropein, and low MW polyphenols

Olive mill wastewaters

MF/NF/OD

MF: 200 nm/Al2O3/tubular; NF: 578 Da/PES/spiralwound; OD: 0.2 μm/PP/hollow fiber

Garcia-Castello et al. (2010)

Total phenols, o-diphenols, hydroxycinnamic acid derivatives and flavonols

Olive mill wastewaters

UF/NF

UF: 100 and 25 kDa/PS/flat sheet; UF: 10 and 2 kDa/PES/flat sheet; NF: 120 Da/polypiperazine amide/flat sheet

Galanakis et al. (2010)

Anthocyanins, flavonoids

Orange press liquor

NF

180 Da/PA
PS/spiral-wound; 300 Da/ polypiperazine amide/spiral-wound; 1000 and 400 Da/PES/spiral-wound

Conidi et al. (2012)

Anthocyanins (cyanidin-3-glucoside chloride, myrtillin chloride and peonidin-3-glucoside chloride), flavanones

Orange press liquor

UF/NF/OD

UF: 100 kDa/PS/hollow fiber; NF: Na2SO4 rejection .25%
50%/PES/spiral-wound; OD: 0.2 μm/PP/ hollow fiber

Cassano et al. (2014)

Chlorogenic acid, apigenin-7-Oglucoside

Artichoke wastewaters

UF/NF

UF: 15 kDa/TiO2/tubular; NF: 200
300 Da/PA/ spiral-wound

Conidi et al. (2015)

Catechin and its derivatives

Carob byproducts

NF/RO

NF: 150
300 Da/PA-TFC/flat sheet; RO: 99.4% salt rejection/PA-TFC/flat sheet

Almanasrah et al. (2015)

Total phenolics, caffeine, chlorogenic acid

Spent coffee

UF/NF

UF: 2.5 kDa/PA/spiral-wound; NF: 1 kDa/PA/spiralwound; NF: 1 kDa Da/PES/spiral-wound

Brazinha et al. (2015) (Continued )

TABLE 6.1 (Continued) Target Compounds

Wastewater

Membrane Process

Pore Dimension/Material/Configuration

References

Phenolic compounds

Olive mill wastewaters

UF/NF/RO

UF: 100 nm/zirconia/tubular; NF: 150
300 Da/ composite noncellulosic membrane/spiral-wound; RO: NaCl rejection 99%/composite noncellulosic membrane/spiral-wound

Zagklis and Paraskeva (2014)

Total phenols, catechin, quercetin, epicatechin, rutin

Grape marc

UF/NF

UF: 100 nm/zirconia/tubular; NF: 470 Da/PA/spiralwound

Zagklis and Paraskeva (2015)

Catechol, hydroxytyrosol, tyrosol, caffeic acid, and vanillic acid

Olive mill wastewaters

MF/NF/OD/ ME

MF: 140 nm/TiO2/tubular; NF: 2 nm/TiO2/tubular; NF: 150
300 Da/PA/spiral-wound; NF: 200 Da/ PA-TFC/spiral-wound; OD: 0.2 μm/PP/hollow fiber; ME: 3100 nm/Al2O3 SiO2 glass/tubular

Bazzarelli et al. (2016)

CA, Cellulose acetate; MBR, membrane bioreactor; ME, membrane emulsification; MF, microfiltration; NF, nanofiltration; OD, osmotic distillation; PA, polyamide; PES, polyethersulfone; PI, polyimide; PP, polypropylene; PS, polysulfone; PVDF, polyvinylidene fluoride; RO, reverse osmosis; TFC, thin film composite; UF, ultrafiltration. Source: Adapted from Castro-Mun˜oz, R., Ya´n˜ez-Ferna´ndez, J., Fı´la, V., 2016a. Phenolic compounds recovered from agro-food by-products using membrane technologies: an overview. Food. Chem. 213, 753
762; Castro-Mun˜oz, R., Barraga´n-Huerta, B.E., Fı´la, V., Denis, P.C., Ruby-Figueroa, R., 2017a. Current role of membrane technology: from the treatment of agro-industrial by-products up to valorization of valuable compounds. Waste and Biomass Valorization, in press. DOI: 10.1007/s12649-017-0003-1.

Recovery of Phenolic-Based Compounds Chapter | 6

219

of olive oil during deep frying at 180 C. Results indicated the capability of the extract in preserving the α-tocopherol content and reducing the production of negative volatile compounds in the olive oil during the frying process. Similarly, Servili et al. (2011b) used phenolic compounds extracted from olive mill wastewaters to fortify milk beverages fermented with γ-amino butyric acid and autochthonous human gastrointestinal lactic acid bacteria. Phenolic compounds did not interfere with the fermentation process and with the activities and survival of functional lactic acid bacteria. A crude phenolic concentrate obtained by membrane treatment was also used with the aim of improving the virgin olive oil phenolic content (Servili et al., 2011a). Results obtained with four different olive cultivars showed that the crude concentrate increased the phenolic content of olive oils without any alteration of the aroma profile. Recently, phenolic compounds recovered from olive mill wastewater have been tested against ascorbic acid, tocopherol mixtures, and α-tocopherol for the prevention in oil oxidation (Galanakis et al., 2018a). Selected antioxidants were mixed or emulsified at different concentrations with extra virgin and refined kernel olive oils, prior to being heated in the oven at 100 C (30 minutes) and 160 C (120 minutes). The oxidation of both heated oils was reduced by olive phenolics at concentration of 500 and 3000 mg/L. Ascorbic acid resulted more efficient than the olive polyphenols, at 2000 and 3000 mg/L, especially in the case of olive kernel oil. Adversely, the tocopherol formulations showed low effectiveness, probably due to their instability at high cooking temperatures. The addition of olive phenols to bakery products to enhance the oxidative stability of these products and prolong their shelf life was tested, too (Galanakis et al., 2018b). In particular, a concentration of 200 mg polyphenols/kg resulted in the most efficient antimicrobial formulation in bread making and extended the preservation of rusks from 6 to 12 weeks. On the other hand, ascorbic acid and tocopherol mixture did not affect significantly the overall bread preservation in both assayed concentrations (500 and 1000 mg/kg). These results suggest the potential use of olive polyphenols as antioxidant and antimicrobial agents in food products that are sensitive to oxidative deterioration during cooking. Olive phenols were also more active as UV filters in a broader region of UVB and UVA when compared with ascorbic acid and α-tocopherol suggesting their application in sunscreen formulations to complement UV filter photoprotection of synthetic compounds (Galanakis et al., 2018c). In addition, the entrapment of olive phenols in silica particles, prior their emulsification in cosmetics, improved their water resistance, revealing the potential use of phenols from olive mill wastewater as UV booster in cosmetics (Galanakis et al., 2018d).

220

Separation of Functional Molecules in Food by Membrane Technology

It is important to point out that food rich in biologically active compounds has become an important choice for consumers aiming to reduce the risk of contracting specific diseases or to treat certain minor illnesses. Phenolic compounds can be also used to improve the utilization of food and agricultural products. El-Shourbagy and El-Zahar (2014) evaluated the efficiency of using bioactive compounds extracted from peanut skins, pomegranate peels, and olive pomace in improving the quality and oxidative stability of ghee during storage under thermal oxidative conditions. Authors confirmed as these extracts could be used as preservative ingredients in the food and/or pharmaceutical industries. Furthermore, phenolic-based fractions recovered from olive mill (Cassano et al., 2013), nixtamalization (CastroMun˜oz et al., 2016b), and artichoke wastewaters (Cassano et al., 2015) by membrane operations have also been tested for their antioxidant activity. This potential of phenolic-enriched extracts increases the interest of pharmaceutical companies that are looking for new sources to produce nutraceuticals, cosmetics, and food supplements (Conidi et al., 2014b). In addition, some other phenolic-based compounds are also attractive for the production of foodstuffs: this is the case of betalains (Moβhammer et al., 2005) and anthocyanins (He and Giusti, 2010), which are considered as potential natural colorants for food and pharmaceutical or cosmetic uses. It is important to note that food companies are currently searching for colorants from natural sources according to the restriction of the use of synthetic colorants, making the anthocyanins’ use even more popular. Anthocyanin-derived colorants offer key advantages over synthetic ones in terms of greater tinctorial strength, reduction of off-flavors, greater uniformity, and increased stability (Wrolstad, 2004). Finally, betalains or anthocyanins have been proposed to replace FD&C red #40 (allura red, E129) in foods and beverages (He and Giusti, 2010).

6.5 ECONOMIC OVERVIEW OF MEMBRANE-BASED TECHNOLOGIES IN PHENOLIC RECOVERY As it is well known, pressure-driven membrane technologies are characterized by high energy consumption when compared with other separation methods (Strathmann et al., 2006). The membranes, as well as energy requirement, represent the main cost of these processes. Generally, the high energy requirement is due to the high driving force needed to perform the separation. Furthermore, investment and maintenance-related costs contribute often significantly to the overall process costs. Nevertheless, the benefit
cost relation has to be considered in these processes. The costs of phenolic compounds and their derivatives, such as anthocyanins, betalains, etc., are high based on the traditional methods applied for obtaining them; whereas their benefits into human health are a high priority. The nonuse of additional phases and heating sources in membrane processes can be an advantage for

Recovery of Phenolic-Based Compounds Chapter | 6

221

biologically active compounds aimed at human consumption (Conidi et al., 2014b). Similarly, the increasing demand for active compounds stems from the growing consumer concern with quality of life (Brazinha and Crespo, 2014). Furthermore, the real impact of food wastewater disposition on water and environmental pollution has to be taken into account. On the other hand, these pressure-driven membrane processes can be reused as long as the initial properties of the membranes are kept. In this view, many chemical cleaning agents such as alkalis, acids, metal chelating agents, surfactants, oxidation agents, and enzymes can be used to restore the initial hydraulic membrane permeability (Al-Amoudi and Lovitt, 2007; Shi et al., 2014). Certainly, the membrane costs to carry out the separation, fractionation, and concentration are considered as high but the cost of the recovered product usually tends to be higher. Brazinha and Crespo (2014) recently reported valuable data about the world market for flavors and nutraceutical ingredients which was estimated around h13 billion in 2006; the US market was projected around h5.5 billion in 2014 with the segments of the market of food 36%, cosmetics and toiletries 27%, beverages 15%, and a forecast to rise 3% per year. In food formulations, the market tends to grow to compensate the present reformulation of food products toward reduced sodium, sugar, and fat products. According to the analysis of Zagklis et al. (2013), membrane processes are the most effective processes in terms of organics reduction together with electrolysis, supercritical water oxidation, and photo-Fenton methodologies when compared with other technologies in the treatment of olive mill wastewaters. A technoeconomical study shows that the cost of treatment with membrane filtration can be covered by exploitation of the phenolic fraction as well as of the organic fractions to be reused as fertilizers, leading to possible profit (Arvaniti et al., 2012). The operational cost for the treatment of 50,000 tons of waste was estimated to be of about 1,535,740h, which is equivalent to 30.71h per treated m3 of olive mill wastewater. The possible profit for the same amount of waste was estimated to be 250,000h for the nutrient fraction and 1,875,000h for the phytotoxic fraction, with a profit of 42.5h per treated m3 of waste, which allows a net profit of 11.79h per m3. Finally, the commercial success can be a considerable indicator of the importance of the membrane-based technologies in the industry and its market growth can suggest that membrane cost may be rather low in future providing better membrane availability. Strathmann (2001) reported that the sales of membranes and modules for applications in water purification (wastewater treatment) and food processing were of about US$400 million and US$200 million, respectively. Indeed, an increase (8%
10% per year) of membrane market was expected. UF and NF processes seem to be the most profitable technology for membrane industrialists for their multiple applications as recovery of high-added value compounds from agro-food wastes (Galanakis et al., 2016), contributing to meet the zero waste economy

222

Separation of Functional Molecules in Food by Membrane Technology

concept, in which waste is used as a feedstock material for new products and applications (Mirabella et al., 2014). Finally, this recovery would allow the companies to diminish the wastewater treatment cost. Nevertheless, it is a difficult task to provide a cost estimation of the total process because studies found in the literature are focused on investigating particular recovery stages in lab scale experiments.

6.6 CONCLUSIONS AND FUTURE TRENDS Pressure-driven membrane processes have demonstrated their ability to recover different classes of phenolic-based compounds from several food processing wastewaters. Particularly, UF and NF can be used to separate, fractionate, and concentrate specific phenolic compounds that, according to their biological activity, have potential applications in the food and pharmaceutical industries. Furthermore, compared with traditional methodologies, these pressure-driven processes are economically profitable in terms of recovery and productivity (high permeate fluxes). Thus, the recovery of phenolic-based solutes from agro-food wastewaters is both industrially sustainable and environmentally friendly. On the other hand, the high costs of waste disposal will make the use of large-scale production processes necessary for industries to focus on waste recycling. Surely, in coming years, it is quite possible that governments will legislate to ensure the use of approaches such as those described herein to reduce water and environmental pollution. It is likely that research and development will be focused on new implementations of NF technology as the primary tool for the recovery and concentration of phenolic compounds. If their purification is required, the use of other membrane technologies or adsorption processes can be useful. Today, though, market opportunities for the natural extracts obtained from such processes are missing. Thus, it is high time that industry started to address this challenge to achieve the fifth stage of the universal recovery process.

REFERENCES Aguiar Prudeˆncio, A.P., Schwinden Prudeˆncio, E., Castanho Amboni, R.D.M., Negra˜o Murakami, A.N., Maraschin, M., Cunha Petrus, J.C., et al., 2012. Phenolic composition and antioxidant activity of the aqueous of bark from residues from mate tree (Ilex paraguariensis St. Hil.) bark harvesting concentrated by nanofiltration. Food Bioprod. Process. 90 (C3), 399
405. Al-Amoudi, A., Lovitt, R.W., 2007. Fouling strategies and the cleaning system of NF membranes and factors affecting cleaning efficiency. J. Memb. Sci. 303 (1
2), 4
28. Almanasrah, M., Brazinha, C., Kallioinen, M., Duarte, L.C., Roseiro, L.B., Bogel-Lukasik, R., et al., 2015. Nanofiltration and reverse osmosis as a platform production of natural botanic extracts: the case study of carob by-products. Sep. Purif. Technol. 149, 389
397.

Recovery of Phenolic-Based Compounds Chapter | 6

223

Arvaniti, E.C., Zagklis, D.P., Papadakis, V.P., Paraskeva, C.A., 2012. High-added value materials production from OMW: a technical and economical optimization. Int. J. Chem. Eng. 2012, Article ID 607219. Ayala-Soto, F.E., Serna-Saldı´var, S.O., Garcı´a-Lara, S., Pe´rez-Carrillo, E., 2014. Hydroxycinnamic acids, sugar composition and antioxidant capacity of arabinoxylans extracted from different maize fiber sources. Food Hydrocolloids 35, 471
475. Azzam, M.O.J., Al-Malah, K.I., Abu-Lail, N.I., 2004. Dynamic post treatment response of olive mill effluent wastewater using activated carbon. J. Environ. Sci. Health Part A 39, 269
280. Bazinet, L., Doyen, A., 2017. Antioxidants, mechanism, and recovery by membrane processes. Crit. Rev. Food. Sci. Nutr. 56 (4), 677
700. Bazzarelli, F., Poerio, T., Mazzei, R., D’Agostino, N., Giorno, L., 2015. Study of OMWWs suspended solids destabilization to improve membrane processes performance. Sep. Purif. Technol. 149, 183
189. Bazzarelli, F., Piacentini, E., Poerio, T., Mazzei, R., Cassano, A., Giorno, L., 2016. Advances in membrane operations for water purification and biophenols recovery/valorization from OMWWs. J. Memb. Sci. 497, 402
409. Bendini, A., Cerretani, L., Carrasco-Pancorbo, A., Go´mez-Caravaca, A.M., Segura-Carretero, A., Ferna´ndez-Gutie´rrez, A., et al., 2007. Phenolic molecules in virgin olive oils: a survey of their sensory properties, health effects, antioxidant activity and analytical methods. An overview of the last decade. Molecules 12 (8), 1679
1719. Bennett, A., 2015. Advances in filtration systems for wastewater treatment. Filtr. Sep. 52 (5), 28
33. Boussu, K., Vandecasteele, C., Van der Bruggen, B., 2008. Relation between membrane characteristics and performance in nanofiltration. J. Memb. Sci. 310 (1
2), 51
65. Brazinha, C., Crespo, J.G., 2014. Valorization of food processing streams for obtaining extracts enriched in biologically active compounds. In: Cassano, A., Drioli, E. (Eds.), Integrated Membrane Operations in the food Production. De Gruyter, Berlin, pp. 295
309. Brazinha, C., Cadima, M., Crespo, J.G., 2015. Valorisation of spent coffee through membrane processing. J. Food Eng. 149, 123
130. Cardinali, A., Pati, S., Minervini, F., D’Antuono, I., Linsalata, V., Lattanzio, V., 2012. Verbascoside, isoverbascoside, and their derivatives recovered from olive mill wastewater as possible food antioxidants. J. Agric. Food. Chem. 60 (7), 1822
1829. Cassano, A., Conidi, C., Drioli, E., 2011. Comparison of the performance of UF membranes in olive mill wastewaters treatment. Water Res. 45 (10), 3197
3204. Cassano, A., Conidi, C., Giorno, L., Drioli, E., 2013. Fractionation of olive mill wastewaters by membrane separation techniques. J. Hazard. Mater. 248, 185
193. Cassano, A., Conidi, C., Ruby Figueroa, R., 2014. Recovery of flavonoids from orange press liquor by an integrated membrane process. Membranes 4 (3), 509
524. Cassano, A., Conidi, C., Ruby Figueroa, R., Castro-Mun˜oz, R., 2015. A two-step nanofiltration process for the production of phenolic-rich fractions from artichoke aqueous extracts. Int. J. Mol. Sci. 16, 8968
8987. Cassano, A., Conidi, C., Galanakis, C.M., Castro-Mun˜oz, R., 2016. Recovery of polyphenols from olive mill wastewaters by membrane operations. In: Figoli, A., Cassano, A., Basile, A. (Eds.), Membrane Technologies for Biorefining. Woodhead Publishing, Cambridge, pp. 163
187. Castro-Mun˜oz, R., Ya´n˜ez-Ferna´ndez, J., 2015. Valorization of nixtamalization wastewaters by integrated membrane process. Food Bioprod. Process. 95, 7
18.

224

Separation of Functional Molecules in Food by Membrane Technology

´ lvarez, C., Ya´n˜ez-Ferna´ndez, J., 2015a. Recovery of bioactive comCastro-Mun˜oz, R., Orozco-A pounds from food processing wastewaters by ultra and nanofiltration: a review. Adv. Biores. 6 (3), 152
158. ´ lvarez, C., Cero´n-Montes, G.I., Ya´n˜ez-Ferna´ndez, J., 2015b. Castro-Mun˜oz, R., Orozco-A Characterization of the microfiltration process for the treatment of nixtamalization wastewaters. Ingenierı´a Agricola y Biosistemas 7 (1), 23
34. Castro-Mun˜oz, R., Cero´n-Montes, G.I., Barraga´n-Huerta, B.E., Ya´n˜ez-Ferna´ndez, J., 2015c. Recovery of carbohydrates from nixtamalization wastewaters (Nejayote) by ultrafiltration. Revista Mexicana de Ingenierı´a Quı´mica 14 (3), 735
744. Castro-Mun˜oz, R., Ya´n˜ez-Ferna´ndez, J., Fı´la, V., 2016a. Phenolic compounds recovered from agrofood by-products using membrane technologies: an overview. Food. Chem. 213, 753
762. Castro-Mun˜oz, R., Barraga´n-Huerta, B.E., Ya´n˜ez-Ferna´ndez, J., 2016b. The use of nixtamalization waste waters clarified by ultrafiltration for production of a fraction rich in phenolic compounds. Waste Biomass Valorizat. 7, 1167
1176. Castro-Mun˜oz, R., Barraga´n-Huerta, B.E., Fı´la, V., Denis, P.C., Ruby-Figueroa, R., 2017a. Current role of membrane technology: from the treatment of agro-industrial by-products up to valorization of valuable compounds. Waste Biomass Valorization . Available from: https://doi.org/10.1007/s12649-017-0003-1, in press. Castro-Mun˜oz, R., Fı´la, V., Dura´n-Pa´ramo, E., 2017b. A review of the primary by-product (Nejayote) of the nixtamalization during maize processing: potential reuses. Waste Biomass Valorization . Available from: https://doi.org/10.1007/s12649-017-0029-4, in press. Castro-Mun˜oz, R., Fı´la, V., Barraga´n-Huerta, B.E., Ya´n˜ez-Ferna´ndez, J., Pin˜a-Rosas, J.A., Arboleda-Mejı´a, J., 2018. Processing of Xoconostle fruit (Opuntia joconostle) juice for improving its membrane commercialization using membrane filtration. J. Food Process. Preservation 42 (1), Article number e13394. Chedeville, O., Debacq, M., Porte, C., 2009. Removal of phenolic compounds present in olive mill wastewaters by ozonation. Desalination 249 (2), 865
869. Cheryan, M., Rajagopalan, N., 1998. Membrane processing of oily streams. Wastewater treatment and waste reduction. J. Membr. Sci. 151 (1), 13
28. Christaki, E., Bonos, E., Florou-Paneri, P., 2012. Nutritional and functional properties of Cynara crops (globe artichoke and cardoon) and their potential applications: a review. Int. J. Appl. Sci. Technol. 2 (2), 64
70. Conidi, C., Cassano, A., Drioli, E., 2011. A membrane-based study for the recovery of polyphenols from bergamot juice. J. Memb. Sci. 375 (1
2), 182
190. Conidi, C., Cassano, A., Drioli, E., 2012. Recovery of phenolic compounds from orange press liquor by nanofiltration. Food Bioprod. Process. 90 (4), 867
874. Conidi, C., Mazzei, R., Cassano, A., Giorno, L., 2014a. Integrated membrane system for the production of phytotherapics from olive mill wastewaters. J. Memb. Sci. 454, 322
329. Conidi, C., Cassano, A., Garcia-Castello, E., 2014b. Valorization of artichoke wastewaters by integrated membrane process. Water Res. 48, 363
374. Conidi, C., Cassano, A., 2015. Recovery of phenolic compounds from bergamot juice by nanofiltration membranes. Desalination Water Treatment 56, 3510
3518. Conidi, C., Rodriguez-Lopez, A.D., Garcia-Castello, E.M., Cassano, A., 2015. Purification of artichoke polyphenols by using membrane filtration and polymeric resins. Sep. Purif. Technol. 144, 153
161. Crespo, J.G., Brazinha, C., 2010. Membrane processing: natural antioxidants from winemaking by-products. Filtr. Sep. 47 (2), 32
35.

Recovery of Phenolic-Based Compounds Chapter | 6

225

Dammak, I., Khoufi, S., Sayadi, S., 2016. A performance comparison of olive oil mill wastewater enzymatic treatments. Food Bioprod. Process. 100 (Part A), 61
71. Dı´az-Montes, E., Castro-Mun˜oz, R., Ya´n˜ez-Ferna´ndez, J., 2016. An overview of Nejayote, a nixtamalization by product. Ingenierı´a Agrı´cola y Biosistemas 8 (2), 41
60. Dı´az-Reinoso, B., Moure, A., Domı´nguez, H., Parajo´, J.C., 2009. Ultra- and nanofiltration of aqueous extract from distilled fermented grape pomace. J. Food Eng. 91 (4), 587
593. Dı´az-Reinoso, B., Gonza´lez-Lo´pez, N., Moure, A., Domı´nguez, H., Parajo´, J.C., 2010. Recovery of antioxidants from industrial waste liquors using membranes and polymeric resins. J. Food Eng. 96 (1), 127
133. Di Donna, L., De Luca, G., Mazzotti, F., Napoli, A., Salerno, R., Taverna, D., et al., 2009. Statinlike principles of Bergamot fruit (Citrus bergamia): isolation of 3-hydroxymethylglutaryl flavonoid glycosides. J. Nat. Prod. 72 (7), 1352
1354. ˇ ˇ Ðilas, S., Canadanovi´ c-Brunet, J., Cetkovi´ c, G., 2009. By-products of fruits processing as a source of phytochemicals. Chem. Ind. Chem. Eng. Quat. 15 (4), 191
202. El-Shourbagy, G.A., El-Zahar, K.M., 2014. Oxidative stability of ghee as affected by natural antioxidants extracted from food processing wastes. Ann. Agric. Sci. 59 (2), 213
220. Esposto, S., Taticchi, A., Di Maio, I., Urbani, S., Veneziani, G., Selvaggini, R., et al., 2015. Effect of an olive phenolic extract on the quality of vegetable oils during frying. Food. Chem. 176, 184
192. Federici, F., Fava, F., Kalogerakis, N., Mantzavinos, D., 2009. Valorisation of agro-industrial by-products, effluents and waste: concept, opportunities and the case of olive mill wastewaters. J. Chem. Technol. Biotechnol. 84 (6), 895
900. Frascari, D., Bacca, A.E.M., Zama, F., Bertin, L., Fava, F., Pinelli, D., 2016. Olive mill wastewater valorisation through phenolic compounds adsorption in a continuous flow column. Chem. Eng. J. 283, 293
303. Galanakis, C.M., 2012. Recovery of high added-value components from food wastes: conventional, emerging technologies and commercialized applications. Trends Food Sci. Technol. 26 (2), 68
87. Galanakis, C.M., 2013. Emerging technologies for the production of nutraceuticals from agricultural by-products: a viewpoint of opportunities and challenges. Food Bioprod. Process. 91 (4), 575
579. Galanakis, C.M., 2015. Separation of functional macromolecules and micromolecules: from ultrafiltration to the border to nanofiltration. Trends Food Sci. Technol. 42 (1), 44
63. Galanakis, C.M., Tornberg, E., Gekas, V., 2010. Clarification of high-added value products from olive mill wastewaters. J. Food Eng. 99 (2), 190
197. Galanakis, C.M., Markouli, E., Gekas, V., 2013. Recovery and fractionation of different phenolic classes from winery sludge using ultrafiltration. Sep. Purif. Technol. 107, 245
251. Galanakis, C.M., Castro-Mun˜oz, R., Cassano, A., Conidi, C., 2016. Recovery of high-addedvalue compounds from food waste by membrane technology. In: Figoli, A., Cassano, A., Basile, A. (Eds.), Membrane Technologies for Biorefining. Woodhead Publishing, Cambridge, pp. 189
215. Galanakis, C.M., Tsatalas, P., Charalambous, Z., Galanakis, I.M., 2018a. Polyphenols recovered from olive mill wastewater as natural preservatives in extra virgin olive oils and refined olive kernel oils. Environ. Technol. Innovation 10, 62
70. Galanakis, C.M., Tsatalas, P., Charalambous, Z., Galanakis, I.M., 2018b. Control of microbial growth in bakery products fortified with polyphenols recovered from olive mill wastewater. Environ. Technol. Innovation 10, 1
15.

226

Separation of Functional Molecules in Food by Membrane Technology

Galanakis, C.M., Tsatalas, P., Galanakis, I.M., 2018c. Phenols from olive mill wastewater and other natural antioxidants as UV filters in sunscreens. Environ. Technol. Innovation 9, 160
168. Galanakis, C.M., Tsatalas, P., Galanakis, I.M., 2018d. Implementation of phenols recovered from olive mill wastewater as UV booster in cosmetics. Ind. Crops Products 111, 30
37. Garcia-Castello, E., Cassano, A., Criscuoli, A., Conidi, C., Drioli, E., 2010. Recovery and concentration of polyphenols from olive mill wastewaters by integrated membrane system. Water Res. 44 (13), 3883
3892. Garcia-Ivars, J., Iborra-Clar, M.I., Alcaina-Miranda, M.I., Mendoza-Roca, J.A., Pastor-Alcan˜iz, L., 2015. Treatment of table olive processing wastewaters using novel photomodified ultrafiltration membranes as first step for recovering phenolic compounds. J. Hazard. Mater. 290, 51
59. Gebreyohannes, A.Y., Mazzei, R., Giorno, L., 2016. Trends and current practices of olive mill wastewater treatment: application of integrated membrane process and its future perspective. Sep. Purif. Technol. 162, 45
60. Giacobbo, A., Oliveira, M., Duarte, E.C.N.F., Mira, H.M.C., Moura Bernardes, A., De Pinho, M.N., 2013a. Ultrafiltration based process for the recovery of polysaccharides and polyphenols from winery effluents. Sep. Sci. Technol. 48 (3), 438
444. Giacobbo, A., Oliveira, M., Duarte, E.C.N.F., Mira, H.M.C., Moura Bernardes, A., De Pinho, M.N., 2013b. Nanofiltration for the recovery of low molecular weight polysaccharides and polyphenols from winery effluents. Sep. Sci. Technol. 48 (17), 2524
2530. Giacobbo, A., Do Prado, J.M., Meneguzzi, A., Moura Bernardes, A., De Pinho, M.N., 2015. Microfiltration for the recovery of polyphenols from winery effluents. Sep. Purif. Technol. 143, 12
18. Giacobbo, A., Meneguzzi, A., Bernardes, A.M., De Pinho, M.N., 2017a. Pressure-driven membrane processes for the recovery of antioxidant compounds from winery effluents. J. Cleaner Production 155, 172
178. Giacobbo, A., Moura Bernardes, A., De Pinho, M.N., 2017b. Sequential pressure-driven membrane operations to recover and fractionate polyphenols and polysaccharides from second racking wine lees. Sep. Purif. Technol. 173, 49
54. He, J., Giusti, M.M., 2010. Anthocyanins: natural colorants with health-promoting properties. Ann. Rev. Food Sci. Technol. 1 (1), 163
187. Inan, H., Dimoglo, A., Simsek, H., Karpuzku, M., 2003. Olive oil mill wastewater treatment by means of electrocoagulation. Sep. Purif. Technol. 36 (1), 23
31. Kalogerakis, N., Politi, M., Foteinis, S., Chatzisymeon, E., Mantzavinos, D., 2013. Recovery of antioxidants from olive mill wastewaters: a viable solution that promotes their overall sustainable management. J. Environ. Manage. 128, 749
758. Kontos, S.S., Koutsoukos, P.G., Paraskeva, C.A., 2014. Removal and recovery of phenolic compounds from olive mill wastewater by cooling crystallization. Chem. Eng. J. 251, 319
328. Mert, B.K., Yonar, T., Kilic, M.Y., Kestioglu, K., 2010. Pre-treatment studies on olive oil mill effluent using physicochemical Fenton and Fenton-like oxidations process. J. Hazard. Mater. 174 (1
3), 122
128. Mirabella, N., Castellani, V., Sala, S., 2014. Current options for the valorization of food manufacturing waste: a review. J. Cleaner Production 65, 28
41. Moβhammer, M.R., Stintzing, F.C., Carle, R., 2005. Development of a process for the production of a betalain-based colouring foodstuff from cactus pear. Innovative Food Sci. Emerg. Technol. 6 (2), 221
231.

Recovery of Phenolic-Based Compounds Chapter | 6

227

Mudimu, O.A., Peters, M., Brauner, F., Braun, G., 2012. Overview of membrane processes for the recovery of polyphenols from olive mill wastewater. Am. J. Environ. Sci. 8 (3), 195
201. Nawaz, H., Shi, J., Mittal, G.S., Kakuda, Y., 2006. Extraction of polyphenols from grape seeds and concentration by ultrafiltration. Sep. Purif. Technol. 48 (2), 176
181. Obied, H.K., Allen, M.S., Bedgood, D.R., Prenzler, P.D., Robards, K., Stockmann, R., 2005. Bioactivity and analysis of biophenols recovered from olive mill waste. J. Agric. Food. Chem. 53 (4), 823
837. Ochando-Pulido, J.M., Stoller, M., Bravi, M., Martinez-Ferez, A., Chianese, A., 2012a. Batch membrane treatment of olive vegetation wastewater from two-phase olive oil production process by threshold flux based methods. Sep. Purif. Technol. 101, 34
41. Ochando-Pulido, J.M., Hodaifa, G., Martinez-Ferez, A., 2012b. Fouling inhibition upon Fentonlike oxidation pretreatment for olive mill wastewater reclamation by membrane process. Chem. Eng. Process. Process Intensif. 62, 89
98. Ochando-Pulido, J.M., Martinez-Ferez, A., 2015. On the recent use of membrane technology for olive mill wastewater purification. Membranes 5, 513
531. Pandey, K.B., Rizvi, S.I., 2009. Plant polyphenols as dietary antioxidants in human health and disease. Oxid. Med. Cell Longev. 2 (5), 270
278. Paraskeva, C.A., Papadakis, V.G., Tsarouchi, E., Kanellopoulou, D.G., Koutsoukos, P.G., 2007. Membrane processing for olive mill wastewater fractionation. Desalination 213 (1-3), 218
229. Rahmanian, N., Jafari, S.M., Galanakis, C.M., 2014. Recovery and removal of phenolic compounds from olive mill wastewater. J. Am. Oil Chem. Soc. 91 (1), 1
18. Ravindran, R., Jaiswal, A.K., 2016. Exploitation of food industry waste for high-value products. Trends Biotechnol. 34 (1), 58
69. Roig, A., Cayuela, M.L., Sa´nchez-Monedero, M.A., 2006. An overview on olive mill wastes and their valorisation methods. Waste Manage. 26 (9), 960
969. Romani, A., Scardigli, A., Pinelli, P., 2016. An environmentally friendly process for the production of extracts rich in phenolic antioxidants from Olea europaea L. and Cynara scolymus L. matrices. Europ. Food Res. Technol. 243 (7), 1229
1238. Ruby-Figueroa, R., Cassano, A., Drioli, E., 2011. Ultrafiltration of orange press liquor: optimization for permeate flux and fouling index by response surface methodology. Sep. Purif. Technol. 80 (1), 1
10. Ruby-Figueroa, R., Cassano, A., Drioli, E., 2012. Ultrafiltration of orange press liquor: optimization of operating conditions for the recovery of antioxidant compounds by response surface methodology. Sep. Purif. Technol. 98, 255
261. Russo, C., 2007. A new membrane process for the selective fractionation and total recovery of polyphenols, water and organic substances from vegetation waters (VW). J. Memb. Sci. 288 (1-2), 239
246. Salehi, F., 2014. Current and future applications for nanofiltration technology in the food processing. Food Bioprod. Process. 92 (C2), 161
177. Santamarı´a, B., Salazar, G., Beltra´n, S., Cabezas, J.L., 2002. Membrane sequences for fractionation of polyphenolic extracts from defatted milled grape seeds. Desalination 148 (1
3), 103
109. Scalbert, A., Manach, C., Morand, C., Re´me´sy, C., Jimene´z, L., 2005. Dietary polyphenols and the prevention of diseases. Crit. Rev. Food. Sci. Nutr. 45 (4), 287
306. Scoma, A., Bertin, L., Zanaroli, G., Zanaroli, G., Fraraccio, S., Faba, F., 2011. A physicochemical-biotechnological approach for an integrated valorization of olive mill wastewater. Bioresour. Technol. 102 (22), 10273
10279.

228

Separation of Functional Molecules in Food by Membrane Technology

Servili, M., Esposto, S., Veneziani, G., Urbani, S., Taticchi, A., Di Maio, I., et al., 2011a. Improvement of bioactive phenol content in virgin olive oil with an olive-vegetation water concentrate produced by membrane treatment. Food. Chem. 124 (4), 1308
1315. Servili, M., Rizzello, C.G., Taticchi, A., Esposto, S., Urbani, S., Mazzacane, F., et al., 2011b. Functional milk beverage fortified with phenolic compounds extracted from olive vegetation water, and fermented with functional lactic acid bacteria. Int. J. Food Microbiol. 147 (1), 45
52. Shi, X., Tal, G., Hankins, N.P., Gitis, V., 2014. Fouling and cleaning of ultrafiltration membranes: a review. J. Water Process Eng. 1, 121
138. Strathmann, H., 2001. Membrane separation processes: current relevance and future opportunities. AIChE J. 47 (5), 1077
1087. Strathmann, H., Giorno, L., Drioli, E., 2006. An introduction to Membrane Science and Technology. Consiglio Nazionale delle Ricerche, Rome. Susanto, H., Feng, Y., Ulbricht, M., 2009. Fouling behavior during ultrafiltration of aqueous solutions of polyphenolic compounds during ultrafiltration. J. Food Eng. 91 (2), 333
340. Tylkowski, B., Nowak, M., Tsibranska, I., Trojanowska, A., Marciniak, L., Garcia Valls, R., et al., 2017. Concentration and fractionation of polyphenols by membrane operations. Curr. Pharm. Des. 23 (2), 231
241. Ulbricht, M., Ansorge, W., Danielzik, I., Ko¨nig, M., Schuster, O., 2009. Fouling in microfiltration of wine: the influence of the membrane polymer on adsorption of polyphenols and polysaccharides. Sep. Purif. Technol. 68 (3), 335
342. Villanova, L., Villanova, L., Fasiello, G., Merendino, A., 2009. Process for the recovery of tyrosol and hydroxytyrosol from oil mill wastewaters and catalytic oxidation method in order to convert tyrosol in hydroxytyrosol. US Patent 20090023815 A1. Wrolstad, R.E., 2004. Symposium 12: interaction of natural colors with other ingredients
anthocyanin pigments—Bioactivity and coloring properties. J. Food. Sci. 69 (5), C419
C421. Xia, E.Q., Deng, G.F., Guo, Y.J., Li, H.B., 2010. Biological activities of polyphenols from grapes. Int. J. Mol. Sci. 11 (2), 622
646. Xu, L., Lamb, K., Layton, L., Kumar, A., 2004. A membrane-based process for recovering isoflavones from a waste stream of soy processing. Food Res. Int. 37 (9), 867
874. Zagklis, D.P., Paraskeva, C.A., 2014. Membrane filtration of agro-industrial wastewaters and isolation of organic compounds with high added values. Water Sci. Technol. 69 (1), 202
207. Zagklis, D.P., Paraskeva, C.A., 2015. Purification of grape marc phenolic compounds through membrane filtration and resin adsorption/desorption. Sep. Purif. Technol. 156, 328
335. Zagklis, D.P., Arvaniti, E.C., Papadakis, V.P., Paraskeva, C.A., 2013. Sustainability analysis and benchmarking of olive mill wastewater treatment methods. J. Chem. Technol. Biotechnol. 88 (5), 742
750. Zagklis, D.P., Vavouraki, A.I., Kornaros, M.E., Paraskeva, C.A., 2015. Purification of olive mill wastewater phenols through membrane filtration and resin adsorption/desorption. J. Hazard. Mater. 285, 69
76. Zirehpour, A., Rahimpour, A., Jahanshahi, M., 2015. The filtration performance and efficiency of olive mill wastewater treatment by integrated membrane process. Desalinat. Water Treatment 53, 1254
1262.

Chapter 7

Lignin Separation and Fractionation by Ultrafiltration Javier Ferna´ndez-Rodrı´guez, Xabier Erdocia, Fabio Herna´ndez-Ramos, Marı´a Gonza´lez Alriols and Jalel Labidi Biorefinery Processes Research Group (BioRP), Chemical and Environmental Engineering Department, University of the Basque Country UPV/EHU, Plaza Europa, Donostia-San Sebastia´n, Spain

Chapter Outline 7.1 Introduction 7.2 Lignin Chemistry 7.3 Lignin Isolation 7.3.1 Sulfite Process 7.3.2 Kraft Process 7.3.3 Alkaline Treatments 7.3.4 Organosolv Process 7.4 Lignin Application in the Food Industry 7.4.1 Lignin as Part of Dietary Fiber 7.4.2 Lignin as Feed Additive 7.4.3 Antimicrobial Effects 7.4.4 Prebiotic Effects of Lignin and Weight Gain 7.4.5 Packaging and Films 7.5 Separation of Lignin by Ultrafiltration

229 230 232 232 233 234 235 236 237 238 238 239 240

7.5.1 Kraft Lignin Purification 7.5.2 Lignosulfonates Purification 7.5.3 Sulfur-Free Lignin Fractionation: Alkaline and Organosolv Lignin 7.5.4 Hydrolysis Lignin Recovery 7.5.5 Membrane Flux Decay in Lignin Separation Process 7.5.6 Ultrafiltration Purification as Source for Food Industry 7.6 Conclusions References Further Reading

245 247

248 249 250

253 257 258 265

241

7.1 INTRODUCTION Lignocellulosic biomass is described as a three-dimensional polymeric composite material synthesized by plants and has been used as energy resource since prehistory by its direct combustion. More recently, lignocellulosic materials have been used primarily in the chemical industry and the furniture Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00007-3 © 2019 Elsevier Inc. All rights reserved.

229

230

Separation of Functional Molecules in Food by Membrane Technology

and construction sector due to their abundance, ubiquity, and availability. In the last few decades, lignocellulosic biomass has aroused interest as a promising alternative to petrochemical fossil resources because new technological advances have enabled the development of more efficient and cleaner processes for the conversion of biomass into energy, biofuels, chemicals, new materials, and food additives. Lignocellulosic biomass consists, mainly, of structural components along with other minor nonstructural compounds (Octave and Thomas, 2009). The proportion of these compounds in the different lignocellulose materials depends primarily on the species, the plant tissue, and the growth conditions (Brandt et al., 2013). The structural compounds of the plant’s cell wall are synthesized by photosynthetic process and they are formed by three biopolymers: cellulose, hemicelluloses, and lignin. These compounds give strength and protection to the cell and represent almost 90% of the dry weight of the total lignocellulosic material (Brandt et al., 2013). Among the structural compounds that make up lignocellulosic biomass, cellulose and hemicelluloses are very well-known polysaccharides, which can be used for many different applications in a variety of areas, such as medicine, energy obtaining, in the chemical industry, in the synthesis of polymeric materials and biosurfactants, as papermaking additives, or in the food industry (Barakat et al., 2013). However, lignin, which is mostly generated as a side product in the pulp and paper production process, is mainly burned in industrial boilers underestimating its enormous potential (Rinaldi et al., 2016). The main reason of this limited use of lignin is due to its high dispersity; the difference in type and number of functional groups and its thermal properties, which vary in accordance with the lignin’s origin and applied isolation process (Zinovyev et al., 2017). Despite all the limitations concerning a wide exploitation of lignin, its chemical composition is so rich in different functional groups that it is worthwhile to dedicate efforts to establish its potential downstream valorization pathways, for example, different separation methods, for the production of different bulk chemicals, which could be used in a variety of industries and, especially, in the food industry.

7.2 LIGNIN CHEMISTRY Lignin comprises 20%
35% of biomass. Lignin is a phenolic biopolymer composed by combination of three phenylpropanoid units: p-coumaryl alcohol, coniferyl alcohol, and synapyl alcohol. These three monolignols are also identified as p-hydroxyphenyl (H) guaiacyl (G), and syringyl (S) units, respectively. These units are randomly connected by several interunit ether and carbon
carbon linkages in different proportions, including β-O-4, α-O-4, β-5, β-1, β 2 β, 4-O-5, 5-5 bonds (Bauer et al., 2012), which are formed by dehydrogenation, cross-coupling, and dehydrodimerization reactions during the biosynthesis process of macromolecular lignin. The contribution of a particular monomer to the polymerization process determines the percentage

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

231

TABLE 7.1 Different Linkages Proportions and BDE in Lignin Molecule (Li et al., 2015; Parthasarathi et al., 2011) Linkage Type

Softwoods (%)

Hardwoods (%)

BDE (kcal/mol)

43
50

50
65

53.94
72.30

Ether Linkages β-O-4 α-O-4

6
8

4
8

48.31
57.28

4-O-5

4

6
7

77.74
82.54

β-5

9
12

4
6

125.2
127.6

5-5

10
25

4
10

114.9
118.4

β-1

3
7

5
7

64.7
165.8

β-β

2
4

3
7




Others

16

7
8




C
C Linkages

of each linkage in lignin molecule. As shown in Table 7.1, ether linkages, particularly β-O-4, are, by far, the most frequent linkages in lignin, followed by carbon
carbon bonds, as 5-5 type. It is also remarkable that ether linkages have lower bond dissociation enthalpy (BDE) than C
C linkages, which are, overall, more stable. The initial oxidation reaction of the monolignol (H, G, or S) is catalyzed by peroxidases and laccases (Weng et al., 2008) and consists of the removal of the phenolic hydrogen atom from the precursors leading to phenoxyl radicals. These radicals are linked together by cross-coupling reactions and produce a three-dimensional amorphous polymer, which forms a randomized structure inside the cell wall (Garcı´a et al., 2009). In addition, native lignin chemistry can vary depending upon its origin. Plant species, plant tissue type, and the external environment all have roles in determining the chemistry of a specific lignin sample. Softwoods are known to contain higher contents of lignin, followed by hardwoods and, finally, grasses (Zakzeski et al., 2010). On the other hand, lignin from hardwood contains roughly equal guaiacyl and syringyl units, while in softwood guaiacyl units account for around 90% of the total units (Azadi et al., 2013). Grass lignins are also classified as guaiacyl
syringyl lignin. However, unlike hardwood lignins, grass lignins additionally contain small but significant amounts of structural elements derived from p-coumaryl alcohol (H). Despite the numerous studies on lignin, its real chemical structure is still unknown as it cannot yet be isolated in its native unaltered state.

232

Separation of Functional Molecules in Food by Membrane Technology

Nevertheless, some authors have tried to define its structure and have proposed different models that represent the lignin macromolecule. In 1977, Adler presented a structure for softwood lignin that is still widely accepted today (Adler, 1977), but recently some other models have also been proposed (Vanholme et al., 2010). Within complex lignin structure different functional groups can be found: carbonyl, carboxyl, aliphatic and phenolic hydroxyl, methoxyl groups, etc. (Tejado et al., 2007). Their amount, position, and relative ratio strongly determine the lignin physicochemical behavior, in terms of solubility, activity, and reactivity, which subsequently influences its commercial utilization.

7.3 LIGNIN ISOLATION The isolation of lignin from its lignocellulosic source is one of the most exigent processes due to the complicated physical and chemical linkages between lignin and the other components present in plants. Lignin is usually isolated in the pretreatment step. For this purpose, several different methods have been studied. These processes could be grouped in two main blocks: (1) processes in which lignin is solubilized from the lignocellulosic biomass and removed by separating the solid residue from the spent liquor, and (2) processes where polysaccharides are selectively hydrolyzed leaving lignin along with some condensed carbohydrate deconstruction products as a solid residue. A good example of the former group is a pulping process, such as Kraft, sulfite, alkali, or organosolv. An example of the latter would be dilute acid hydrolysis of lignocellulosic biomass to yield sugar monomers, furfural, and levulinic acid (Azadi et al., 2013). However, the most typical methods for lignin isolation and recovery belong to the first group, which will be described later.

7.3.1 Sulfite Process In 1867, B.C. Tilghman patented a process in which the decomposition of wood with sulfurous acid or its salts was used for the production of fibrous pulp for the manufacture of paper. However, it was not until 1874 when the Swedish chemist Ekman marketed the first sulfite pulp. Almost simultaneously with Ekman, A. Mitscherlich in Germany worked on the sulfite cooking process of wood and, in 1880, he started a sulfite pulp mill in Zell in south Germany. In the sulfite process, usually conducted under acid or neutral conditions, sulfur dioxide and/or bisulfite ions react with lignin to produce water-soluble sulfonated lignins, called lignosulfonates, which are degraded by acid hydrolysis reactions in large pressure vessels called digesters (Hintz, 2001). The employed salts are either sulfites (SO322) or bisulfites (HSO32) and even sulfur dioxide (SO2) or sulfuric acid (H2SO4) could be used. The content of

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

233

these different forms of sulfur species varies as the sulfite pulping can be conducted in a wide range of pH. The most common cation of the sulfite and bisulfite salts is sodium (Na1) but calcium (Ca21), potassium (K1), magnesium (Mg21), or ammonium (NH41) cations have been employed as well. The sulfite process produces wood pulp, which is almost pure cellulose. The residual lignin in sulfite pulps is sulfonated and it is relatively easy to remove in pulp bleaching processes (Hintz, 2001). Sulfite pulping has been used for a long time in the industry but, nowadays, it has been replaced by the Kraft process, which is capable of processing different wood species to produce strong fibers for paper manufacturing (Zhu et al., 2009). However, around 1 million tons per year of lignosulfonates are still produced as dry solids. Typically, in lignosulfonate lignins a high quantity of impurities (around 30% in weight), like ash or carbohydrates, can be found (Galkin and Samec, 2016). These lignins usually have very high average molecular weight, even higher than Kraft lignins, as sulfonate groups are added to lignosulfonate lignins during the pulping process (Galkin and Samec, 2016). As a consequence of the incorporation of these groups, the catalytic transformation of lignosulfonates lignins is more difficult (Patil and Argyropoulos, 2017). Moreover, in the sulfite pulping processes, ether bonds are cleaved, methoxyl groups are destroyed, and new carbon
carbon bonds are formed (Li et al., 2015). Nevertheless, the diversity of functional groups (phenolic, carboxylic, and sulfur containing groups) on lignosulfonate lignins offers exceptional colloidal properties, which allows its use in application such as dispersants, additives (plasticizers), surfactants, flocculants, etc. (Aro and Fatehi, 2017). In addition, these lignins can be used in the food industry as they are easily depolymerized to produce vanillin (Rahimi et al., 2014).

7.3.2 Kraft Process Kraft process is the foremost pulping process in the board and paper industry and the biggest lignin supplier. About 130 million tons per year of Kraft pulp are produced worldwide (Bruijnincx et al., 2015), which is around 80% of the produced chemical pulp and about 50
55 million metric tons of lignin in the form of black liquor (Mahmood et al., 2016). The Kraft pulping process involves the digestion of wood chips at high temperature (145
170 C) and pressure in “white liquor,” which is a water solution of sodium sulfide (Na2S) and sodium hydroxide (NaOH), for a few hours (Huber et al., 2014). During this treatment, the hydroxide and hydrosulfide anions react with the lignin, causing the polymer to fragment into smaller water/alkali-soluble fragments and isolating the cellulose fibers. About 90% of lignin is removed in the Kraft process. The cooked chips are then washed, and the filtrate (weak black liquor) is then concentrated to 65% solids in a direct-contact evaporator, by bringing

234

Separation of Functional Molecules in Food by Membrane Technology

the liquor into contact with the outlet gases from the recovery boiler, or in an indirect-contact concentrator. Afterward, the strong black liquor (where lignin is dissolved) is usually burned in the recovery boiler, to recover heat and chemicals. The combustion of the organics dissolved in the black liquor (mainly lignin) provides heat that is used to generate steam that is used as hot utility in the process, and for converting sodium sulfate to sodium sulfide. The high quantity of steam produced in this step makes the black liquor the fifth most important fuel in the world (Tran and Vakkilainnen, 2012) Otherwise, inorganic chemicals present in the black liquor are collected at the bottom of the furnace. In the Kraft process, the action of strong aqueous alkali with the simultaneous presence of hydrosulfide ions alters and depolymerizes the lignin structure (Gellerstedt, 2015). These structural changes comprise the addition of thiol groups (Galkin and Samec, 2016), the cleavage of the interunit ether linkages, and the increment of both, recalcitrant C
C bonds and the number of phenolic hydroxyl groups (Gillet et al., 2017). One of the major disadvantages of Kraft lignin is that it is greatly contaminated with carbohydrates coming from hemicelluloses as well as of some fatty acids (Gellerstedt, 2015). In addition, substantial amount of sulfur (1
3 wt.%) is usually covalently bound to Kraft lignin in form of thiols. Considering all these disadvantages and the great potential of the Kraft lignin as fuel, only 2% of all industrial lignins and only 100,000 tons of available Kraft lignins are valorized per year instead of being burned (Schorr et al., 2014). Kraft lignins can be used as dispersants (dye dispersants, agrochemical dispersants) and emulsifiers (asphalt emulsifiers). Moreover, low molecular weight chemicals, like dimethylsulfoxide, can be produced from Kraft lignin (Lora, 2008). Furthermore, the antioxidant capacity of Kraft lignin is also used for cosmetic formulations and pharmaceutical preparations (Vinardell et al., 2008). In contrast to lignosulfonates or Kraft lignins, alkaline and organosolv lignins are sulfur free, which makes them suitable for higher added value applications (Zhang, 2011). The revalorization of these lignins is the key point for producing small aromatic building blocks to satisfy the enormous and diverse industrial demand for these types of chemicals. Lignin catalytic transformations provide routes to obtain cresols, catechols, resorsinols, quinones, vanillin, guaiacols, and so on, considered as potential chemicals that can be used in the food industry (Li et al., 2015).

7.3.3 Alkaline Treatments Nearly the 5% of the total pulp production is conducted by alkaline treatments (Patt et al., 2012). The use of alkali agents, which could be NaOH, KOH, Ca(OH)2, hydrazine, and anhydrous ammonia (Agbor et al., 2011), causes the swelling of biomass and disrupts the lignin structure. In alkaline

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

235

pretreatments, the linkage between lignin and the other carbohydrate fractions is broken, which increases the reactivity of the remaining polysaccharides as delignification occurs (Pedersen and Meyer, 2010). The main reactions occurring during the alkaline treatment of lignocellulosic biomass are the dissolution of lignin and hemicellulose and the saponification of the intermolecular ester bonds (Kim et al., 2016). The mentioned reactions influence the polymerization degree of lignin and hemicelluloses. Among all the basic catalyst that can be used in alkaline treatments, sodium hydroxide is the strongest, allowing a higher dissolution degree of lignin than any other alkaline treatment. Sodium hydroxide cleaves the ether and ester bonds between lignin and hemicelluloses in the lignin
carbohydrate structure. In addition, it attacks ester and C
C bonds in lignin molecules (Kim et al., 2016). During the NaOH pretreatment reaction, sodium hydroxide is dissociated into hydroxide anion (OH2) and sodium cation (Na1) and, as the hydroxide ion concentration increases, the rate of the hydrolysis reaction increases accordingly. The lignins obtained from this process are usually considered relatively free of impurities but they can have high ash content due to the presence of soda in the reaction media (Gillet et al., 2017). It has been shown that the addition of catalytic quantities of anthraquinone has a marked effect on the stabilization of carbohydrates and the dissolution of lignin. However, the rate of lignin removal is still low compared with that of the Kraft process (Azadi et al., 2013). This is the reason why the alkaline pretreatment only has effective results for certain raw materials with low lignin content, such as agricultural residues (bagasse and straw) and herbaceous crops (Galbe and Zacchi, 2012). It is expected that alkaline treatment will become more important in the near future due to its sulfur free nature and the low cost and easy recovery of the employed reagents. Nevertheless, alkaline treatment is not very selective and along with lignin, extractives and carbohydrates are dissolved and several nondesirable reactions occur. Therefore, expensive purification steps are required for the use of the resulting solid and liquid fractions (Egu¨e´s et al., 2012).

7.3.4 Organosolv Process Organosolv processes are chemical treatments where lignin is extracted from lignocellulosic feedstock by its dissolution using organic solvents or their aqueous solutions. The most commonly used solvents for organosolv delignification are alcohols, such as methanol and ethanol (usually with 50% water); organic acids, such as formic and acetic acid; mixed organic solvent
inorganic alkali chemicals (Azadi et al., 2013); acetone (Capolupo and Faraco, 2016); or ethylene glycol (Gonza´lez Alriols et al., 2009). The main drawback of this process is the high cost of these solvents, though this

236

Separation of Functional Molecules in Food by Membrane Technology

disadvantage can be minimized by recovering and recycling them through evaporation and condensation (Kumar and Sharma, 2017). Among all the employed organosolvents, ethanol and methanol are the most commonly used ones mainly due to their lower boiling points (Xu and Huang, 2014). Organosolv processes can be applied in combination with catalysts (usually inorganic bases or acids), which can help in the dissolution of lignin by breaking the hemicellulose bonds and lignin linkages (Hallac et al., 2010). The treatment can be carried out in a wide range of temperatures (100
250 C) and, normally, under high pressures. In contrast with other methods, the reaction media effectively solubilize lignin without the need of its structural modifications so, organosolv pulping can be considered as a more efficient option to preserve the native structure of lignin during extraction (Pandey and Kim, 2011). However, some lignin internal bonds are cleaved along with lignin
hemicellulose bonds (Capolupo and Faraco, 2016). Typically, in the organosolv pretreatment, very high lignin removal (.70%) could be achieved while the cellulose loss is almost negligible (,2%) (Xu and Huang, 2014). Besides lignin dissolution, some cleavage of β-O-4 linkages inside the lignin molecule could happen, which causes a partial depolymerization of lignin (Xu and Huang, 2014). Furthermore, lignin condensation reactions have been observed as well under organosolv pretreatments (Wildschut et al., 2013). The advantages of organosolv lignin with respect to Kraft and sulfonated lignins include the absence of sulfur, greater ability to be derivatized, lower ash content, a higher purity degree (due to lower content of carbohydrates), generally lower molecular weight, and hydrophobicity (Lora and Glasser, 2002). These characteristics make organosolv lignin very suitable for the production of high added-value chemicals (Zakzeski et al., 2010) and it can be used to formulate glues, binders, in polymer substitutions, or as phenolic precursor (Zhao et al., 2017). Thanks to all these advantages, to the benefit of obtaining separate cellulose, hemicelluloses, and lignin streams, and due to the growing concern about the environmental impact of traditional pulping processes, it is expected that the organosolv methods will find more interest in the future for biomass fractionation (Azadi et al., 2013). This way, it is expected that there will be more organosolv lignin in the future, which could be very useful for applications in the food industry.

7.4 LIGNIN APPLICATION IN THE FOOD INDUSTRY Research on lignin is a broad area where many researchers are developing new practices beyond its use as an energy source in the paper industry. One of these possible new fields of application of lignin is its use in the food industry. The research that is being carried out is focused on a better

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

237

understanding of lignin’s behavior, in its purified form, as part of the dietary fiber in human gastrointestinal tract and the associated effects on health. In this context, the application of lignin is related to the production of food additives, as a component of the plant’s cell wall is part of the human and animal diet. Other studies are devoted to the use of lignin in the manufacture of packaging and films in the food industry (Kirk and Farrell, 1987).

7.4.1 Lignin as Part of Dietary Fiber According to the definition adopted in 2001 by the American Association of Cereal Chemists (Dietary Fiber Definition Committee, 2001), dietary fiber is “the edible part of the plants or analogous carbohydrates that are resistant to digestion and absorption in the human small intestine with complete or partial fermentation in the large intestine. Dietary fiber includes polysaccharides, oligosaccharides, lignin and associated plants substances. Dietary fiber promotes physiological benefits such as laxative effects, decrease blood cholesterol level and decreased blood sugar.” According to the European Union (EU), lignin is included as one of the components of dietary fiber when it is closely associated to the original polysaccharides of plants, but not when it is as an isolated compound added to food (Commission Directive 2008/100/EC, 2008). Lignin can be found in many foods of human diet, such as cereals, fruits, and vegetables, so it could be said that lignin has an important part in human diet (Bunzel et al., 2005) with an approximate daily intake of lignin of 1.6
2 g/day (Bunzel and Ralph, 2006). Bunzel et al. (2011) determined that, in foods such as grain wheat, kale, and pear, the lignin content was 5%, 7%, and 16% of the dietary fiber fraction, respectively. However, despite the important presence of lignin in food for human consumption, the possible degradation of lignin in the human gastrointestinal tract and its possible effects have not been sufficiently studied, neither as a dietary fiber nor in its purified form (Niemi et al., 2013). Therefore, more studies are being carried out to understand how lignin acts in the human body, before using this compound as a food additive. These studies are being performed in vivo and in vitro, although in vitro assays are more used since in vivo ones are more expensive and require more time. However, these types of tests are static and they do not consider the physiological responses of the body in digestion (Thuenemann, 2015). In recent years, some progress has been made with the development of dynamic gastrointestinal digestion simulators as, for instance SHIME (Ghent, Belgium) or Simgi (Madrid, Spain), which have included the physiological responses of the body, such as the increase in the pH, the gradual transit of compounds in the digestive tract, etc. These types of equipment could allow a better understanding of the behavior of lignin in the gastrointestinal tract.

238

Separation of Functional Molecules in Food by Membrane Technology

Due to this polymeric nature, lignin cannot be absorbed in the human gastrointestinal tract and has been generally assumed to have no nutritional value. However, although it does not act as an energy source, lignin does not remain inert and it is able to induce other effects in the human gastrointestinal tract interacting with the food components. In this sense, the studies developed by Funk et al. (2006, 2007) concluded that lignin from dietary fiber can absorb carcinogenic compounds in the upper intestine and colon, preventing them from being absorbed by the body and, thus, reducing the risk of cancer. In addition, Sakagami et al. (2010) concluded that lignin
carbohydrate complexes, in experiments carried out in vitro and in vivo, have various pharmacological activities such as antitumor, antimicrobial, anti-HIV, and antioxidant properties thanks to their phenolic nature. Dhingra et al. (2012) affirmed that lignin has several therapeutic functions as well; for example, it reduces the risk of heart disease, decreases blood sugar level, etc. The research of Zhang et al. (2013) revealed that lignin could be a new type of activator that increases the activity of α-amylase, an enzyme responsible for catalyzing the hydrolysis of α-glucosidic bonds of high
molecularweight polysaccharides as starch. In a more recent study, Zhang et al. (2014) reported that lignin might have an inhibitory or activation effect of the pancreatic lipase enzyme depending on the reaction medium and the used substrate.

7.4.2 Lignin as Feed Additive The effects of lignin degradation in animals have not been extensively studied. In fact, the most common use that has been given to purified lignin, as lignosulfonate, is as a binder in feed for livestock (Baurhoo et al., 2008). In one of the studies conducted by Kajikawa et al. (2000), it was concluded that monogastric animals cannot digest the lignin molecule while the bacterial flora of ruminants facilitates the degradation of the benzyl
ether bonds of lignin. In the cattle industry, lignin has been considered as a barrier in the digestion of nutrients. Probably for this reason, purified lignin has not received more scientific interest and its potential as a food additive is not well recognized. However, there are some studies carried out in vivo and in vitro, which have shown that purified lignins could have beneficial effects on animal health, such as antimicrobial properties, prebiotic effects, and positive effects on the weight gain of these (Baurhoo et al., 2008).

7.4.3 Antimicrobial Effects One potential beneficial effect of lignin on animal health is due to the wellknown antimicrobial effects of its phenolic fragments, which endow lignin with the ability to eliminate microorganisms, such as bacteria, fungi, or

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

239

parasites, or inhibit their growth (Baurhoo et al., 2008). Therefore, due to the presence of various low-molecular-weight phenolic monomers in purified lignin that are not present in native lignin, the use of the former could be effective as a natural antibiotic and could completely or partially replace the use of the current antibiotics used in farming. Despite the great potential that purified lignin can offer in animal health and welfare, there are not enough studies on the possible use of this resource in the livestock industry (Baurhoo et al., 2008). In tests conducted in vitro by Zemek et al. (1979) the ability of these phenolic fragments to inhibit the growth of different microorganisms, such as Escherichia coli, Saccharomyces cerevisiae, Bacillus licheniformis, and Aspergillus niger, was demonstrated. In other research, Nelson et al. (1994) concluded that the phenolic fragments of Alcell lignin reduced the growth of E. coli, Staphylococcus aureus, and Pseudomonas in liquid cultures, and that there existed a correlation between antibacterial effect and Alcell lignin dose (Phillip et al., 2000). Additional studies (Baurhoo et al., 2007a,b) have suggested the ability of Alcell lignin to reduce the content of E. coli in the intestines of chickens and in poultry litter. Baurhoo et al. (2007a,b) indicated that greater inhibition of this bacteria was found at higher concentrations of lignin, suggesting that the antibacterial effect of Alcell lignin occurred mostly at higher doses. However, not only the amount of lignin used is important; the method used in the separation of lignin from the other components of the cell wall is important as well, since the presence of impurities could influence the antimicrobial properties of the lignin (Niemi, 2016).

7.4.4 Prebiotic Effects of Lignin and Weight Gain Prebiotics are a class of foods defined as “nondigestible ingredients that benefit the body, through the growth and/or activity of various bacteria in the colon, improving health” (Gibson and Nutrition, 1995). As previously mentioned, lignin is included in the definition that the EU and the American Association of Cereal Chemists have given to dietary fiber. This fact involves that lignin can present prebiotic effects that could be beneficial to be used in livestocks as dietary fiber additive. Ricke et al. (1982) reported that the addition of induline, a purified form of Kraft lignin from the paper industry, in the diet of poultry improved the weight gain of the animals and greater efficiency was obtained in the consumption of food. However, in similar studies, the addition of Alcell lignin in the diet of chickens (Baurhoo et al., 2007a,b) and in pigs (Valencia and Chavez, 1997) did not show an increase of weight or better efficiency in food consumption, which could indicate that the source of the lignin has an effect on the fattening of the animals. Furthermore, it has been shown that purified lignin could have prebiotic effects in the organism. In this sense, Baurhoo et al. (2007a,b) determined an

240

Separation of Functional Molecules in Food by Membrane Technology

increase in the concentration of Lactobacilli and Bifidobacteria in the intestines of chickens when using Alcell lignin as a supplement in their diet. In this same study, it was seen that the number of goblet cells increased, as well as villi height, which resulted in better intestinal health and, therefore, better animal health. These studies could contribute to extend the application of lignin and its derivatives in the food industry. There are, although, some products derived from lignin that are present in the market and are commonly used in the food industry such as lignosulfonates and perhaps the best known of all, vanillin. Calcium lignosulfonate is a light yellow
brown powder-shaped compound, soluble in water but practically insoluble in organic solvents. It is a compound that has generally been used in the food industry as an emulsifier in animal feed, as a raw material for the production of vanillin, and as a boiler additive. The calcium lignosulfonate, with a range of molecular weight of 40,000
65,000 Da, is one of the different types of lignosulfonate that exist in the market (Toledo and Kuznesof, 2008). Calcium lignosulfonate also offers other alternative to be used in the food industry as an encapsulating agent for fat-soluble vitamins (A, D, E, and K), carotenoids (e.g., β-carotene, β-apo-80 -carotenal, zeaxanthin, canthaxanthin, lutein, and lycopene), and in other functional ingredients. Its mission in this application is to facilitate the introduction into water-based foods, for example, beverages based of fruit, drinkable vitamins, and hard candies (Toledo and Kuznesof, 2008). The level of calcium lignosulfonate in foods depends on the application and the limits allowed for the use of food colors and nutrients. These levels are usually low so there is no risk of interaction between the additive and the fat-soluble compounds that are present in the diet (Toledo and Kuznesof, 2008).

7.4.5 Packaging and Films The appearance of plastic materials was a revolution in many industrial sectors due to plastic’s flexible characteristics. In fact, today almost everything that is commercialized is made of plastic or comes packed in plastic containers or films. This fact generates an annual consumption of plastics that, in 2015, reached 322 million tons, of which 39.9% were destined for packaging uses (Plastics Europe—Plastics—the Facts 2016, 2016). The main objective of food packaging is the protection of food from the environment. Another purpose is to maintain the properties of foods. Therefore, packaging materials must act as a physical barrier to protect food from contamination and preserve nutrients, avoiding contact with the gases (O2, CO2, etc.), moisture and light, maintaining good mechanical, optical, and thermal properties (Rhim et al., 2013). Not all conventional plastic materials or bioplastics have all these properties so, sometimes, it is necessary to use reinforcements that complement them. One of these reinforcements is lignin, thanks to its

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

241

phenolic nature. Lignin contains chromophore components that act as UV absorbers. These chromophores make lignin a broad-spectrum sunblock. In addition to this ability to absorb UV, lignin has excellent antioxidant properties and can increase the thermal and oxidation stability of polymers in mixtures (Glasser et al., 1999). However, the use of lignin as a reinforcement in biopolymer composites has not been widely accepted in modern industry. This is because the properties of lignin vary depending on the extraction method, the particle size, the molecular weight, and the degree of polymerization (Mamun and Bledzki, 2013). On the other hand, when formulating copolymers, some aspects need to be considered, such as the nature of the matrix, and lignin’s incompatibilities with some other components, which may require modifications to ensure compatibility. Despite this fact, many studies are being carried out where lignin appears as a suitable option as an additive in the manufacture of new bio-based and biodegradable plastic materials for the food industry. For example, the addition of lignin to an agar film improved the barrier properties against UV and the mechanical properties of the film (Shankar et al., 2015). Sadeghifar et al. (2017) observed that the addition of a 2% of Kraft lignin in a cellulose film allowed the absorption of the 100% of UV-B and a 90% of UV-A while maintaining mechanical and thermal properties in an acceptable range. A study developed by Rai et al. (2017) demonstrated the antibacterial capacity of chitosan films with lignin as reinforcement, which allowed a better conservation of food. Therefore, the use of lignin can have great importance in the manufacture of new bio-based plastics due to the boom experienced by this class of materials that, in 2017, reached about 1% of the global production of plastics. This figure means about 2.05 million tons of which 58% (1.189 million tons) were destined to packaging.

7.5 SEPARATION OF LIGNIN BY ULTRAFILTRATION The pulp and paper industry, the principal sector for lignin production, has used membrane processes since the early 1970s due to the rising demand for environmental protection, reducing the high amount of water consumption, recovery of by-products, and energy savings (Wallberg et al., 2003a). Industrially, the introduction of these methods has depended on the lignin extraction process (Kraft, sulfite or organosolv) and/or the aim of the separation. In the Kraft process, the most widely established process for cellulose production, the use of membrane processes was initially introduced as wastewater treatment for reducing mainly the quantity of water consumption in the global process, and diminishing the environmental impact in the paper mills (Thompson et al., 2001). However, the trend from the last few decades in the utilization of lignin as source of phenolic compounds as chemical building blocks in value-added applications, such as small phenolic molecules production, carbon fibers, and many others (Ragauskas et al., 2014),

242

Separation of Functional Molecules in Food by Membrane Technology

FIGURE 7.1 Number of publications per year on lignin separation by membrane technology. Scopus.

has encouraged the implementation of UF as lignin separation and fractionation process to enable its further valorization. In this sense, the number of scientific publications carried out about the concentration and purification of spent liquors by UF has increased in the recent years, as can be seen in Fig. 7.1. Besides the numerous studies reported in literature, the evolution of patent publications for membranes in biorefinery is increasing indicating the high attention given to this topic (Abels et al., 2013). As an example, 32 patents about UF and NF were filled out in 2015, most of them focused on the lignin purification from different cooking liquors (Servaes et al., 2017). As was mentioned earlier, lignin can be obtained by several methods that lead to different spent liquors and different lignin composition and properties. Not only does lignin vary with the pulping process, but also the sugars obtained from the hemicelluloses that are contained in the raw material. In alkaline and organosolv liquors, hemicelluloses are basically dissolved as oligosaccharides with high molecular weights (up to 80,000 g/mol) (Lundqvist et al., 2002), whereas the strong acid conditions in sulfite pulping involve the cleavage of the oligomeric hemicelluloses into their monomeric sugars (150
180 g/mol) (Rueda et al., 2015a,b). A summary of compositions and molecular weight of lignin of technical lignins from different extraction method can be observed in Table 7.2 (Vishtal and Kraslawski, 2011). Membrane technology essentially consists of a separation process that employs a semipermeable membrane, which acts as a selective barrier for

TABLE 7.2 Characteristics of Lignin in Function of Its Pulping Process Parameter

KL

AS

HL

OSL

LS

Carbohydrates (%)

1.0
2.3

1.5
3.0

10.0
22.4

1
3




Acid soluble lignin (%)

1.0
4.9

1.0
11

2.9

1.9




Sulfur (%)

1.0
3.0

0

0
1.0

0

3.5
8.0

Ashes (%)

0.5
3.0

0.5
3.0

1.0
3.0

1.7

Molecular weight (g/mol)

1500
5000 (up to 25,000)

1000
3000 (up to 15,000)

5000
10,000

500
5000

1000
50,000 (up to 150,000)

Polydispersity

2.5
3.5

2.5
3.5

4.0
11.0

1.5

4.2
7.0

AL, alkaline lignin; HL, hydrolysis lignin; KL, Kraft lignin; LS, lignosulfonates; OSL, organosolv lignin. Source: Adapted from Vishtal, A., Kraslawski, A., 2011. Challenges in industrial applications of technical lignins. BioResources 6, 3547
3568.

244

Separation of Functional Molecules in Food by Membrane Technology

splitting two or more components from each other. The major difference with respect to the conventional filtration is the cross-flow mode of operation instead perpendicular flow; that is, the feed stream flows tangentially over the membrane surface. In comparison with other traditional processes of separation (evaporation or precipitation), membrane technology offers great advantages, such as continuous operation, no requirement of additional agents, low energy consumption, and possibility of being combined with other techniques, among many others. However, negative aspects as the permeate flux decay, the lifetime of the membranes, and the low selectivity could emerge in these processes (Mulder, 1991). The separation by molecular size depends on the material of the membrane (polymeric or ceramic), and the cutoff of particles that are wished to be retained (Kova´cs et al., 2009). Ceramic membranes allow more extreme operation conditions (temperature, pH range, and lifetime) than polymeric membranes and, therefore, their acquisitions costs are greater than the polymeric ones, though their selectivity is lower than that of the polymeric membranes (Baker, 2000). According to these factors, four types of processes can be distinguished: microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), and reversible osmosis. All the lignin molecular weights given in Table 7.2 are inside the range of UF technology (100,000
1000 kDa), as the difference in size from the rest of liquor components is big enough to approach the separation and fractionation of lignin by UF technology. The importance of the lignin molecular weight and its distribution to facilitate its further conversion into value-added products has been highlighted in several works. The survey of Tejado et al. (2007) suggested that the fractionation of lignin could facilitate its conversion into phenolic resins, selecting the most suitable fraction for its transformation in terms of their molecular weight and distribution. Zhou et al. (2006) studied the influence of lignin molecular weight on the viscosity of coal-water slurry. UF was also used to obtain several lignin fractions with different molecular sizes. In the same line, Yang et al. (2015) experimented with the different behavior of lignin fractionated by UF in several streams based on its molecular weight as a dispersing agent in dye preparations. The research of Li et al. (2017) remarked that the molecular uniformity contributes to improve the mechanical performance of the lignin-based carbon fibers. In this context, membrane separation has been presented as a suitable technology to separate and fractionate lignin to facilitate its transformation or its direct use in other applications. The main factors that should be considered in the process design are the lignin source, the pulping method it was obtained from, and the main aim of the process, for example, lignin fractionation, carbohydrates separation, monomeric sugars purification, etc. (Jo¨nsson, 2016). In all cases, membrane technologies can be employed for lignin

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

245

separation due to the difference in molecular size of the components involved in the liquors, as well as the high flexibility that this technology offers. Furthermore, there is not any chemical addition in these membrane processes, a fact that tremendously facilitates its further utilization in food applications in comparison with other separation or fractionation treatments (Zinovyev et al., 2017). Hereafter, different examples of UF processes to separate and fractionate lignin are depicted according to the previous pulping method used, which shows the wide range of combinations that can be used for this target.

7.5.1 Kraft Lignin Purification The main methods to isolate Kraft lignin from the black liquor produced during the pulping stage are precipitation, selective solvents, and membrane processes. The most widely used method is the precipitation of lignin by acidification of the liquor. LignoBoost, a process developed by Chalmers University of Technology and Innventia, Sweden, now owned by Metso, Finland, is the most applied method in industry to precipitate lignin based on the acidification with carbon dioxide and sulfuric acid in different stages ¨ hman et al., 2007). However, lignin from this process presents higher ash (O content than in other processes, such as UF (Ziesig et al., 2014). In addition, no pH modification is needed when using the UF. Generally, to extract lignin from Kraft black liquor, the precipitation method is used, whereas UF has been employed mainly to purify the lignin fraction in cooking liquor from sulfite and Kraft pulp mills to use it as a value-added chemical precursor (Bhat et al., 2015). Many research studies have been focused on the lignin isolation and purification from the Kraft black liquor, as is the most employed method to extract lignin. As an example, lignin was separated from the hemicellulose contained in the black liquor by UF technology (Wallberg et al. 2003b). The flexibility of the UF processes is highlighted by the possibility of inserting the membrane separation in different points of the process, before or after evaporation stage and even without cooling nor adjusting the pH (Wallberg and Jo¨nsson, 2006). A complete study about black Kraft liquor fractionation was carried out by Persson et al. (2010), who used the combination of MF, UF, and NF in a sequential mode. MF was used as a pretreatment stage, focused on the rejection of the suspended matter with 100% of effectiveness using 0.2 μm cutoff membrane. Then, a stage of UF was performed with a 10 kDa membrane, splitting hemicelluloses in the retentate stream and lignin in the permeate stream. The rejections reached for hemicelluloses and lignin were greater than 90% and 15%, respectively, meaning that the concentration of the lignin achieved in the permeate was around 50% even if it still presented high

246

Separation of Functional Molecules in Food by Membrane Technology

amount of salts. The NF stage was conducted with a 300 Da membrane for recovering fresh water that could be reutilized in the pulping process. Separation of the lignin from the salts was achieved employing a higher cutoff membrane in the NF stage, by a similar work conducted by Jo¨nsson et al. (2008) with 1 kDa membrane after the MF and UF stages. The isolation of lignin from salts is fairly important for its further valorization, since a high amount of salts hinders these treatments, for example, reducing the energy production by its burning. Then, the salts could be recovered in a further stage using a lower molecular weight cutoff (MWCO) membrane as was described earlier. UF used as a pretreatment of Lignoboost process was proposed by Ziesig et al. (2014). A sequential process where MF was followed by a UF stage using ceramic membranes with 300 and 15 kDa cutoff was applied to the black liquor in diafiltration mode. As a result, the purity of the final precipitated lignin was increased by reduction of carbohydrate content and inorganic matter (50% lower in ash content). Diafiltration consists of the recirculation of the retentate stream whereas the permeate is driven to further stages. In this way, retentate stream can be deeply concentrated in its main component, lignin in this case; reducing small molecular size components. Wallberg et al. (2003a) also applied this principle to obtain a high purity lignin stream with an increase in its purity from 36% to 78% after the concentration of lignin by diafiltration. A technical comparison of UF lignin against LignoBoost lignin was studied by Giummarella et al. (2016). In this research, UF lignin obtained from a pilot process designed previously by Keyoumu et al. (2004), using ceramic membranes of 5 kDa cutoff, presented better solubility in different organic solvents than LignoBoost lignin, as well as a lower impact over the final viscosity than unfiltered lignins. The cost estimation of the implementation of membrane separation to Kraft black liquors has been extensively studied by Jo¨nsson and Wallberg (2009), who established a cost of about 60h per ton of lignin recovered from a Kraft cooking liquor, and 33h per ton of lignin when separation was made from the black liquor withdrawn from the evaporation system. In any case, the final cost was found to be extremely dependent on the initial volume treated and the desired level of volume reduction (Jo¨nsson et al., 2008). The fractionation of lignin in different streams with narrower molecular weight distribution (MWD) has been broadly studied in recent years. Sevastyanova et al. (2014) studied the fractionation of lignin in different streams according to its different molecular weight. Cross-flow filtration of a Kraft black liquor was performed using ceramic membranes of 10, 5, and 1 kDa membrane. Several streams were obtained with the molecular weight and MWD as their significant difference. The lower molecular weight of the fraction, the narrower the MWD of the lignin was, whose influence on the subsequent functionality and thermal properties was demonstrated. Similar

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

247

work was carried out by Shao et al. (2017) who proposed a fractionation of Kraft lignin in several streams using 10, 5, and 1 kDa membranes. Their reports showed the reduction on carbohydrate content for the smaller size streams as well as their reduction in molecular weight and polydispersity index. Another Kraft lignin fraction was developed by Zinovyev et al. (2017) using organic membranes with different cutoffs (100, 30, 10, 5, 3, and 1 kDa). In this way, final lignin samples showed noticeable narrow molecular size distributions. A linear correlation was established among the MWCO and membrane cutoff, which allows predicting molecular weight using a specific membrane. Carboxyl group content was increased as molecular weight decreased, which makes these fractions more suitable for its further conversion by chemical processes.

7.5.2 Lignosulfonates Purification Currently, lignosulfonates (the main by-product of the sulfite process) reach a world production of around 1.5 million tons per year, with Borregaard LignoTech as the major manufacture, achieving 500,000 tons per year of total production (dry basis) (Lora, 2008). Lignosulfonates present higher molecular weight and polydispersity index than other lignins such as Kraft lignin, according to Table 7.2. However, the smaller size of the sugars from hemicelluloses in this liquor, mainly in monomeric form, enables the separation between lignosulfonates and monomeric sugars, components with enormous interest for producing high valueadded products from renewable resources (Koutinas et al., 2014). Traditionally, spent sulfite liquors are concentrated by evaporation and finally dried by centrifugation. Therefore, no separation is produced regarding their dissolved components and they are commercialized as lignosulfonates because of their high content in this component (55%
70%) (Rueda et al., 2015a,b). The principal use of these unpurified lignosulfonates is as a binder in concretes or as surfactant (Duval et al., 2013). As it can be observed, high amount of sugars (around 25%
35% of the commercialized lignosulfonates) integrated in the lignosulfonates matrix is not being exploited. Hence, a separation of lignosulfonates and monomeric sugars not only would allow the sugars’ valorization, it would also allow to obtain more purified lignosulfonates. This fact could open new alternatives for being used in more innovative and valuable applications (Rueda et al., 2015a,b). In this sense, membrane technology emerges as a suitable method for this purification stage based on the difference in molecular weight between lignosulfonates (1000
50,000 g/mol) and monomeric sugars (150
180 g/mol). UF has been applied to this type of separation and it has even been adopted by the industries. Borregaard Industries (Norway) have possessed the most important plant for lignosulfonates isolation by UF since 1981. This plant allows treating 50 m3/h of spent liquor with 12% of dissolved solids,

248

Separation of Functional Molecules in Food by Membrane Technology

producing a concentrate stream enriched in lignosulfonates (approximately 95% of purity) with a low percentage of sugars and salts. Polysulfone membranes are used for this treatment, with a total membrane area of 1120 m2. High value-added products as vanillin can be obtained by this company from these lignosulfonates after using an UF stage, demonstrating the benefits of this technique (Pabby et al., 2015). In addition, a higher fractionation degree of the lignosulfonates can be carried out with a sequential system using membranes with different cutoff. In this sense, Restolho et al. (2009) performed a comparison of different membrane processes to fractionate a spent sulfite liquor. Plate polymeric membranes were tested for each type of process, however the low molecular weight of lignosulfonates, make more suitable the use of NF. In other work, Ferna´ndez-Rodrı´guez et al. (2015) performed a study of the splitting of spent sulfite liquor in different fractions with three UF membranes (15, 5, and 1 kDa). A cascade fractionation process led to generate a permeate stream enriched in monomeric sugars and reducing the polydispersity index of the rest of the streams, which enabled the conversion of these components into value-added products.

7.5.3 Sulfur-Free Lignin Fractionation: Alkaline and Organosolv Lignin As alternative to the Kraft process, the alkaline delignification method has currently emerged to avoid the use of sulfur compounds that pollute the final products (both cellulose and lignin products). Alkaline pulping leads to worse fiber properties than Kraft pulping if the application is the paper industry, the most traditional application. However, new applications for cellulosic pulp are being utilized nowadays, such as micro- or nanocellulose products, biofuels productions, and chemical precursors (sorbitol, glycerol, etc.), where quality of cellulosic fiber is not important. In this frame, valorization of alkaline lignin could open new opportunities to be used in value-added applications. Thus, application of lignin in food industry can be facilitated due to the reduction in hazard compounds that other traditional lignins have. However, alkaline lignin also presents high polydispersity index and high ash content that hinders its further utilization (Ferna´ndez-Rodrı´guez et al., 2017). The separation principle by membranes is equal to that in Kraft liquors due to the components that belong to the alkaline black liquor are similar to the ones obtained by the Kraft method. Toledano et al. (2010) presented the comparison of acid precipitation of lignin from an alkaline process (NaOH 7.5 wt.%) against UF fractionation, using 15, 10, and 5 kDa ceramic membranes. UF technology allowed controlling the molecular weight of the final lignin with lower carbohydrate pollution. A combination of UF and NF was proposed by Servaes et al. (2017) to fractionate lignin from an alkaline process (AlkOx method). Polymeric membranes of 5, 4, 3, 2 kDa and 750 Da

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

249

were employed to gradually remove Na2CO3 and other organic impurities, for example, organic acids. A preliminary survey of cost estimation indicated the high cost of the liquor treatment that make these processes industrially not implemented so far in high indexes. The studies about the use of membrane for organosolv liquor fractions have only been conducted in a research scale and in very low number of publications, since this pulping process is barely used industrially (Kautto et al., 2013). In comparison with the soda and Kraft process, significant lower amounts of hemicelluloses are dissolved into the organosolv liquor as a consequence of the major selectivity of this method for the lignin dissolution (Bozell et al., 2007). Hence, the use of membranes over this liquor is more focused on the purification of the obtained lignin than on its separation from other compounds. A comparison between organosolv and alkaline methods was performed by Garcı´a et al. (2010). The fractionation of lignin by UF system was applied to soda and organosolv liquors. The antioxidant activity of the different streams produced after the UF was assessed, pointing out the increase in this important property for food conversation as an example after this treatment in comparison with lignin that was not ultrafiltrated. The influence of lignin molecular weight in further valorization processes was demonstrated by Toledano et al. (2013), who developed a UF system for splitting the organosolv liquor in different streams obtaining a lignin with a narrower polydispersity index (from 4.18 up to 2.89) that was intended for the production of small phenolic compounds by means of its depolymerization. This study demonstrated that the smallest molecular weight fractions allow reaching higher yields for monomer phenolic compounds, such as phenol, catechols or cresols. Therefore, the fractionation of lignin by UF could allow dividing the valorization processes of lignin according to their properties in order to obtain suitable yields in each process of transformation. Hussin et al. (2015) studied the improvement in phenolic
OH content and solubility of low molecular weight lignin obtained from a UF process using polymeric membranes of 5 kDa cutoff. A cost estimation was also included in the work presented by Gonza´lez Alriols et al. (2010), where an organosolv method using ethanol as the main reagent was used to obtain a black liquor. Lignin was separated from this liquor by UF treatment prior to the recirculation of ethanol by its distillation. The final cost of the produced lignin was estimated in 52 h/ton, higher than the one presented in an above-mentioned previous work (Jo¨nsson and Wallberg, 2009), but related to lignin samples with enhanced properties that broaden the possibilities for further applications.

7.5.4 Hydrolysis Lignin Recovery Prehydrolyzed liquors have mainly come from the acid hydrolysis for producing cellulosic ethanol and from the modification of the Kraft process to

250

Separation of Functional Molecules in Food by Membrane Technology

produce “dissolving degree” cellulose as well (Saeed et al., 2012). The modification of the Kraft process consists in the addition of a prehydrolysis stage to remove hemicelluloses, although a significant amount of lignin is also extracted. The most commonly used methods for this purpose have been autohydrolysis, steam explosion, dilute acid hydrolysis, or alkaline hydrolysis, due to their weak conditions to carry out the extraction. In any case, the prehydrolyzed stream could be valorized exploiting its sugar and lignin content boosting the integrated biorefinery concept. Hence, a separation among these compounds would need to be conducted, the membrane technology being a suitable method for this aim. Although several studies have been developed using this type of separation (Ferna´ndez-Rodrı´guez et al., 2017), the main goal is the concentration of hemicelluloses and not the lignin separation due to the greater content in hemicelluloses present in these streams (Vegas et al., 2006).

7.5.5 Membrane Flux Decay in Lignin Separation Process Flux decline can be caused by several factors, such as concentration polarization or permanent fouling. The pressure-driven fluid flow through the membrane convectively transports solutes/particles toward the upstream surface of the membrane. This fact causes an accumulation of partially (or completely) retained solute at the upstream surface of the membrane. This phenomenon, referred to as concentration polarization, reduces the diffusion of the flux because of the resistance generated by these boundary layers in which the diffusivity of the solutes is lower than in the bulk. The solute concentration profile in the solution adjacent to the membrane varies then from its value at the membrane surface (Cm) to that in the bulk solution (Cb) over the thickness of the concentration boundary layer (Zydney, 1997). In the worst case, a gel layer can be formed by the rejected macromolecules over the membrane adding an additional resistance. The gel concentration depends on the size, shape, chemical structure and degree of salvation but it is independent of the bulk concentration. The appearance of this gel layer justified the occurrence of the limiting flux. The effects of these phenomena over the concentration gradients in the membrane surface are represented in Fig. 7.2. In addition, membrane fouling is defined as the irreversible deposition of retained particles, colloids, macromolecules, salts, etc., at the membrane surface or inside the pore wall, which causes a continuous flux decline, that is, after the cleaning of the membrane the value of the permeate flux does not achieve the initial value (Noble and Stern, 1995) as it can been observed in Fig. 7.3, where the characteristic evolution of the flux when fouling has appeared in the process is shown. In lignin separation processes, the original liquors, regardless of the process they have been obtained by, induce the fouling generation that becomes a significant cost factor when the processes are scaled up to the industrial level.

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

251

FIGURE 7.2 Schematic representation of polarization concentration and gel layer effect over the solute concentration gradients adjacent to the membrane surface. Cb, Bulk concentration; Cm, membrane surface concentration; Cg, gel layer concentration; Cp, permeate concentration; δ, length of the boundary layer.

FIGURE 7.3 Example of fouling influence on permeate flux.

The main consequence of polarization concentration and fouling appearance is the reduction of the separation performance. The level of the decrease is specific and depends mostly on the application. Hence, the methods to avoid or to reduce these negative events can be only described in a general way due to the complexity of these phenomena. Despite this fact,

252

Separation of Functional Molecules in Food by Membrane Technology

some strategies could be proposed, such as membrane cleaning, pretreatment of feed stream, and optimization of the operating conditions. The development of effective cleaning strategies is necessary to increase the performance and to extend the lifetime of the membranes as well as to improve the economic feasibility of this technology (Humpert et al., 2016). A combination of a rinsing step with the permeate stream and a chemical cleaning was proposed by Wallberg et al. (2003b) to treat a ceramic membrane of 15 kDa cutoff. The rinsing was conducted at 0.5 bar of TMP with the permeate stream because the pH was similar to the feed stream. The chemical cleaning was accomplished at the same TMP, 60 C using 0.25 wt.% Ultrasil 11 solution (Henkel, Germany) until the initial pure water flux was restored. Ultrasil 11 is one of the most commonly used agents in membrane cleaning, composed by an alkaline agent free of ethylenediaminetetraacetic acid and nitrilotriacetic acid. The rinsing stage showed a recovery of 70%
80% of the initial pure water flux, whereas the combination of both cleanings restored the 90%
95% of the initial flux value (Wallberg et al., 2003a). Another technique for ceramic membrane cleaning has been proposed by Liu et al. (2004) using alkaline and acid cleaning stages with an additional treatment in a furnace at 550 C. The initial water flux was completely restored in this study. For organic membranes the same approach can be used without the furnace stage. However, the flux recovering only reached a 50%
80% of the initial pure water flux. A physical cleaning can be also carried out to increase the water flux by a back-pulse or back-flush. Although the application made by Krawczyk and Jo¨nsson (2011) was approached for MF this principle can be applied to UF as well. This process consists of the momentary change in the liquid flux direction throughout the membrane. The permeate flows through the membrane in reverse, usually for less than a second at regular intervals. In this sense, the cake or gel layer is broken or partially reduced, decreasing the negative effect of the flux decay for the polarization concentration phenomenon. Concretely, a flux increasing from 300 to 420 L/m2h could be obtained. Besides the cleaning strategies and maintenance labors of the membrane to enhance the performance, the selection of the best operating variables is a key factor. Several studies have been carried out aimed at the process optimization of UF showing the influence of the difference variables of the process in the final performance of the global process. In this is sense, one example is the work carried out by Kekana et al. (2016), who, by using different polymeric membranes, tested the influence of different parameters, such as the membrane cutoff, lignin concentration in the feed stream, and transmembrane pressure (TMP). The results of this work are summarized in Fig. 7.4. In this example, the lignin rejection yield is increased when the membrane cutoff was decreased from 20 to 1 kDa. However, the permeate flux decreased when lower membrane cutoff was used. The feed concentration also had a big influence on the process since the most concentrated was the feed stream; the highest was the lignin rejection yield at the UF membrane.

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

253

FIGURE 7.4 General influence of the membrane cutoff, feed concentration, and TMP on the lignin rejection yield and permeate flux.

In any case, in membrane processes a compromise between the retention yield and the appearance of flux decay phenomenon has to be defined beforehand. The TMP, which is the difference between the pressures in both sides of the membrane, also played an important role in the permeate flux. The biggest the TMP value, the highest permeate flux is obtained. The limitation of the TMP is the energy required to increase this parameter that could increase the operating cost of the process. Some works have been previously mentioned where the operation approach is the cascading process, that is, membranes with different cutoff are subsequently displayed, not only to obtain lignin streams with narrower polydispersity, but also to enhance the operating performance. In the work by Ferna´ndez-Rodrı´guez et al. (2015), a sequence of three membranes was proposed: 15, 5, and 1 kDa. The change in the approach from individual membrane use to a cascade process led to an increase in the permeate flux for the membrane of 5 kDa around the double. In addition, the first stages of 15 and 5 kDa made that the last membrane of 1 kDa, which in the individual evaluation had logically presented the lowest permeate flux; presented even higher permeate flux than the previous stage (5 kDa membrane). This fact involves that the bottleneck of the process could even not be the membrane with lowest cutoff. Furthermore, in this example the bottleneck stage (5 kDa membrane) doubled the permeate flux, which led to an important process advantage in terms of cost savings. Therefore, the cascading approach has been demonstrated to be a suitable strategy to reduce the negative events of flux decay that take place in the membrane process. A summary of several examples of different UF process configurations for lignin separation are described in Table 7.3. The type of feed stream, material of the membrane used in the work, and operating conditions are detailed in this table.

7.5.6 Ultrafiltration Purification as Source for Food Industry The spectra of industrial applications in which filtration processes with membranes are used has been continuously increasing for 40 years. The food

TABLE 7.3 Characteristics and Operating Conditions of Different UF Processes Applied to the Lignin Separation Feed Stream

Membrane Material

Filtration Mode

TMP (kPa)

Permeate Flux (L/m2h)

References

Kraft liquor

PS, PES

Cross-flow

100
700

100

Wallberg et al. (2003a)

Kraft liquor

Al2O3
TiO2

Cross-flow

20
200

120

Wallberg et al. (2003b)

Kraft liquor

Al2O3
TiO2

Cross-flow

100
400

50
120

Jo¨nsson and Wallberg (2009)

Kraft liquor

PES

Cross-flow

600

2150

Persson et al. (2010)

Kraft liquor

TiO2
ZrO2

Cross-flow

350




Sevastyanova et al. (2014)

Kraft liquor

PS

Dead-end

150
350

,50

Kekana et al. (2016)

Kraft liquor

RCvsPP

Dead-end

150
200




Zinovyev et al. (2017)

Sulfite liquor

PS, PES

Dead-end

400

27
83

Restolho et al. (2009)

Sulfite liquor

TiO2

Cross-flow

200

,55

Ferna´ndez-Rodrı´guez et al. (2015)

Alkaline liquor

PES

Dead-end

500

,84

Servaes et al. (2017)

TiO2
ZrO2

Cross-flow

PES, Polyethersulfone; PS, polysulfone; RCvsPP, regenerated cellulose with polypropylene.

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

255

FIGURE 7.5 (A) Methyl vanillin molecule. (B) Ethyl vanillin molecule.

industry is a very good example of the use of this separation technique (Sondhi et al., 2003). The UF membrane technique fits very well with food industry processes as the working conditions of this technique can retain macromolecular species while small molecules pass through. Typical macromolecules present in food processes are, among others, polysaccharides or proteins (Zabkova et al., 2007a,b).

7.5.6.1 Vanillin Purification by Ultrafiltration The main application of lignin-derived compounds in food processes is related to the use of vanillin, as a flavor agent, as antioxidant or due to its antimicrobial capacity (Cerruti et al., 1997). As a flavoring agent, it presents some interesting characteristics, as good compatibility with fruit flavors when concentrations do not exceed 3000 ppm (Cerrutti and Alzamora, 1996). There are two types of vanillin, depending on the alkoxy substituent, 3-methoxy-4-hydroxybenzaldehyde and 3-ethoxy-4-hydroxybenzaldehyde, represented in Fig. 7.5A and B. The first type, methyl vanillin, with a mild vanilla flavor, is mainly used in flavored foods. The second type, ethyl vanillin, which is exclusively synthetic, presents a stronger vanilla flavor (Sinha et al., 2008). Two different production routes can be followed for obtaining vanillin: extraction from natural resources or chemical synthesis. The production of synthetic vanillin can be achieved by using petrochemical origin guaiacol or lignin and represents the majority of the total vanillin market to fulfill the current demand of this product (Walton et al., 2000). Guaiacol based vanillin represents 85% of world supply while synthetic vanillin form lignin accounts for the remaining 15% (Silva et al., 2009). Nevertheless, even if the synthetic production of vanillin is an economical route, as are many other chemical processes, severe drawbacks related to the associated environmental pollution due to the generation of effluents makes it necessary to explore other production routes (Kaur and Chakraborty, 2013). Natural vanillin, extracted from the beans of the vanilla orchid, accounts for 3% of the total vanillin production and, even if its extraction technology is much more expensive than the synthetic route and the yield is very low, it is preferred over synthetic vanillin, especially for applications in the food industry (Sinha et al., 2008). This fact can be explained as European and US legislation forbids the use of chemically synthesized flavor chemicals for the production of natural flavors (Zheng et al., 2007). These legislations, on the contrary, accept as

256

Separation of Functional Molecules in Food by Membrane Technology

“natural” the vanillin produced from microbiological transformations of precursors from natural origin (Gallage and Møller, 2015). Thus, biotechnological production of vanillin using microorganisms such as yeasts, fungi, bacteria, and plant cells is attracting more and more attention as an alternative to extraction from vanilla orchid (Zheng et al., 2007). Several biotechnological routes for bio vanillin production have been developed recently, starting from phenolic compounds. The main substrates for bio vanillin production are lignin (Havkin-Frenkel and Belanger, 2008), eugenol (Overhage et al., 2003), isoeugenol (Furuya et al., 2017), and ferulic acid (Kaur and Chakraborty, 2013). Two processes comprise the main demand of lignin in the food industry: the manufacture of chocolate and the production of ice cream. Both applications are very sensitive to the aroma and fragrance of vanillin and, thus, even if the obtaining process from lignin is more expensive, the higher aromatic intensity of the obtained product, compared with the guaiacol-based one, compensates for the cost difference (Arau´jo, 2008). Apart from the previously mentioned synthetic and biotechnological routes for obtaining vanillin from lignin, other technologies have been studied as well, such as the use of microwaves (Clark et al., 2006). Furthermore, the separation techniques to be applied after reaction for further purification of the product have been the core of many research works. The solution from which vanillin must be extracted will have a variable composition depending on the starting raw material, applied process, and conditions (Humpert et al., 2016). Nevertheless, in different proportions, lignin-derived oligomers, phenol-based low-molecular-weight aromatic compounds, aromatic aldehydes, guaiacol, and syringol will mainly compose the mixture (Rodrigues Pinto et al., 2012). Indeed, the purity of the isolated vanillin directly determines the cost of the process, which has driven the research of several techniques in this field (Fache et al., 2016). Among these technologies are ion-exchange (Zabkova et al., 2007a,b), macroporous resin (Wang et al., 2010), adsorbent resins (Hua et al., 2007), solvent extraction (Kaygorodov et al., 2010), crystallization (Ibrahim et al., 2009), membrane contactor technology (Sciubba et al., 2009), and UF (Silva et al., 2009). Concretely in UF processes, Zabkova et al. (2007a,b) studied the influence of different parameters, such as lignin and vanillin concentration or the pH modification in a UF process of vanillin recovery from Kraft lignin. Tubular ceramic membranes of 1, 5, and 15 kDa of cutoffs were used to separate, on the one hand, the lowest molecular weight compound in the permeate (in this case, vanillin); and, on the other hand, the higher molecular weight compounds as lignin in the retentate. The retentate flow-rate, the permeate flow-rate, and the generated delta pressures were the parameters measured in the experimental scheme. They concluded that an equilibrium optimum between the purity of vanillin and the permeate flux should be achieved and that this combination of factors should be controlled by the

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

257

membrane cutoff; lower cutoffs allowed the obtaining of high purity vanillin although the permeate flux was noticeably decreased. The possibility of introducing more UF stages starting from larger cutoff membrane in a cascade model could help to solve the flux decay phenomenon. Regarding the pH influence, a considerable flux decline was observed at higher pH of the filtrated solution, which was associated to the hydrophobicity of the membrane surface and solute. In summary, the importance of the separation and purification techniques for the obtaining of high quality vanillin has governed the developed techniques for this process, especially if vanillin is going to be used as an additive in food industry processes. For this propose, UF technology has been proved to be a suitable option that can provide a compromise solution between vanillin purity and obtained flux, that is, UF has shown a suitable compromise between quality and productivity.

7.6 CONCLUSIONS Lignin, a biopolymer composed by combination of three phenylpropanoid units (p-coumaryl alcohol, coniferyl alcohol, and synapyl alcohol), is one of the main constituents of the plant cell wall and represents 20%
35% of biomass. There are two main ways to isolate lignin from lignocellulosic feedstocks: 1. processes in which lignin is solubilized from the lignocellulosic biomass and is removed by separating the solid residue from the spent liquor (Kraft, sulfite, alkali, and organosolv) and 2. processes where polysaccharides are selectively hydrolyzed leaving lignin along with some condensed carbohydrate deconstruction products as a solid residue (hydrolysis of lignocellulosic biomass). The Kraft process is the most used process in the pulp and paper industry and is the biggest lignin supplier. However, lignin is generally considered as a side product in the pulp and paper industry and is mainly burned in boilers to produce energy underestimating its enormous potential as a raw material to obtain high value-added products such as vanillin, widely used in the food industry. Lignin, as a component of the plant’s cell wall, is part of the human and animal diet. Due to the polymeric nature of lignin, the human gastrointestinal tract is not able to absorb it. The same happens in the case of other monogastric animals and only ruminants are able to degrade the benzyl
ether bonds of lignin. However, lignin does not remain inert and can have healthpromoting effects acting as a prebiotic. It can also be attributed some therapeutic functions, pharmaceutical activities (antitumor, antimicrobial, anti-HIV), and antioxidant properties due to its phenolic nature. Despite these interesting applications, more studies are needed to describe the effects

258

Separation of Functional Molecules in Food by Membrane Technology

of lignin in the organism both in its fiber and in its purified forms. In the food industry, the lignin purification processes could play a very important role for the use of lignin as a dietary supplement for humans and/or animals in its purified form. On the other hand, it has been demonstrated that, in the production of vanillin, the most widely used lignin-derived product in the food industry, the UF process, offers a good compromise between quality and productivity.

REFERENCES Abels, C., Carstensen, F., Wessling, M., 2013. Membrane processes in biorefinery applications. J. Memb. Sci. 444, 285
317. Adler, E., 1977. Lignin chemistry—past, present and future. Wood Sci. Technol. 11, 169
218. Agbor, V.B., Cicek, N., Sparling, R., Berlin, A., Levin, D.B., 2011. Biomass pretreatment: fundamentals toward application. Biotechnol. Adv. 29, 675
685. Arau´jo, J., 2008. Production of vanillin From Lignin Present in the Kraft Black Liquor of the Pulp and Paper Industry. University of Porto, Portugal. Aro, T., Fatehi, P., 2017. Production and application of lignosulfonates and sulfonated lignin. ChemSusChem 10, 1861
1877. Azadi, P., Inderwildi, O.R., Farnood, R., King, D.A., 2013. Liquid fuels, hydrogen and chemicals from lignin: a critical review. Renew. Sustain. Energy Rev. 21, 506
523. Baker, R.W., 2000. Membrane technology. Kirk-Othmer Encyclopedia of Chemical Technology. John Wiley & Sons, New York. Barakat, A., de Vries, H., Rouau, X., 2013. Dry fractionation process as an important step in current and future lignocellulose biorefineries: a review. Bioresour. Technol. 134, 362
373. Bauer, S., Sorek, H., Mitchell, V.D., Iba´n˜ez, A.B., Wemmer, D.E., 2012. Characterization of Miscanthus giganteus lignin isolated by ethanol organosolv process under reflux condition. J. Agric. Food Chem. 60, 8203
8212. Baurhoo, B., Letellier, A., Zhao, X., Ruiz-Feria, C.A., 2007a. Cecal populations of lactobacilli and bifidobacteria and Escherichia coli populations after in vivo Escherichia coli challenge in birds fed diets with purified lignin or mannan oligosaccharides. Poult. Sci. 86, 2509
2515. Baurhoo, B., Phillip, L., Ruiz-Feria, C.A., 2007b. Effects of purified lignin and mannan oligosaccharides on intestinal integrity and microbial populations in the ceca and litter of broiler chickens. Poult. Sci. 86, 1070
1078. Baurhoo, B., Ruiz-Feria, C.A., Zhao, X., 2008. Purified lignin: nutritional and health impacts on farm animals—a review. Anim. Feed Sci. Technol. 144, 175
184. Bhat, A.H., Dasa, Y.K., Khan, I., 2015. Extraction of lignin from biomass for biodiesel production. In: Hakeem, K.R. (Ed.), Agricultural Biomass Based Potential Materials. Springer International Publishing, Switzerland, pp. 155
179. Bozell, J., Holladay, J., Johnson, D., White, J., 2007. Top Value-Added Chemicals from Biomass—Volume II: Results of Screening for Potential Candidates from Biorefinery Lignin. Report No. PNNL-16983. Evaluation II, 87. Brandt, A., Gra¨svik, J., Hallett, J.P., Welton, T., 2013. Deconstruction of lignocellulosic biomass with ionic liquids. Green Chem. 15, 550. Bruijnincx, P.C.A., Rinaldi, R., Weckhuysen, B.M., 2015. Unlocking the potential of a sleeping giant: lignins as sustainable raw materials for renewable fuels, chemicals and materials. Green Chem. 17, 4860
4861.

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

259

Bunzel, M., Ralph, J., 2006. NMR characterization of lignins isolated from fruit and vegetable insoluble dietary fiber. J. Agric. Food Chem. 54, 8352
8361. Bunzel, M., Seiler, A., Steinhart, H., 2005. Characterization of dietary fiber lignins from fruits and vegetables using the DFRC method. J. Agric. Food Chem. 53, 9553
9559. Bunzel, M., Schu¨ßler, A., Tchetseubu Saha, G., 2011. Chemical characterization of klason lignin preparations from plant-based foods. J. Agric. Food Chem. 59, 12506
12513. Capolupo, L., Faraco, V., 2016. Green methods of lignocellulose pretreatment for biorefinery development. Appl. Microbiol. Biotechnol. 100, 9451
9467. Cerrutti, P., Alzamora, S.M., 1996. Inhibitory effects of vanillin on some food spoilage yeasts in laboratory media and fruit pure´es. Int. J. Food Microbiol. 29, 379
386. Cerruti, P., Alzamora, S.M., Vidales, S.L., 1997. Vanillin as an antimicrobial for producing shelf-stable strawberry puree. J. Food Sci. 62, 608
610. Clark, J.H., Budarin, V., Deswarte, F.E.I., Hardy, J.J.E., Kerton, F.M., Hunt, A.J., et al., 2006. Green chemistry and the biorefinery: a partnership for a sustainable future. Green Chem. 8, 853. Commission Directive 2008/100/EC [WWW Document], 2008. Dhingra, D., Michael, M., Rajput, H., Patil, R.T., 2012. Dietary fibre in foods: a review. J. Food Sci. Technol. 49, 255
266. Dietary Fiber Definition Committee, 2001. The definition of dietary fiber. Cereal Foods World 46, 112
126. Duval, A., Molina-Boisseau, S., Chirat, C., 2013. Comparison of Kraft lignin and lignosulfonates addition to wheat gluten-based materials: mechanical and thermal properties. Ind. Crops Prod. 49, 66
74. Egu¨e´s, I., Sanchez, C., Mondragon, I., Labidi, J., 2012. Separation and purification of hemicellulose by ultrafiltration. Ind. Eng. Chem. Res. 51, 523
530. Fache, M., Boutevin, B., Caillol, S., 2016. Vanillin production from lignin and its use as a renewable chemical. ACS Sustain. Chem. Eng. 4, 35
46. Ferna´ndez-Rodrı´guez, J., Garcı´a, A., Coz, A., Labidi, J., 2015. Spent sulphite liquor fractionation into lignosulphonates and fermentable sugars by ultrafiltration. Sep. Purif. Technol. 152, 172
179. Ferna´ndez-Rodrı´guez, J., Erdocia, X., Sa´nchez, C., Gonza´lez Alriols, M., Labidi, J., 2017. Lignin depolymerization for phenolic monomers production by sustainable processes. J. Energy Chem. 26, 622
631. Funk, C., Weber, P., Thilker, J., Grabber, J.H., Steinhart, H., Bunzel, M., 2006. Influence of lignification and feruloylation of maize cell walls on the adsorption of heterocyclic aromatic amines. J. Agric. Food Chem. 54, 1860
1867. Funk, C., Braune, A., Grabber, J.H., Steinhart, H., Bunzel, M., 2007. Model studies of lignified fiber fermentation by human fecal microbiota and its impact on heterocyclic aromatic amine adsorption. Mutat. Res. Fundam. Mol. Mech. Mutagen. 624, 41
48. Furuya, T., Kuroiwa, M., Kino, K., 2017. Biotechnological production of vanillin using immobilized enzymes. J. Biotechnol. 243, 25
28. Galbe, M., Zacchi, G., 2012. Pretreatment: the key to efficient utilization of lignocellulosic materials. Biomass Bioenergy 46, 70
78. Galkin, M.V., Samec, J.S.M., 2016. Lignin valorization through catalytic lignocellulose fractionation: a fundamental platform for the future biorefinery. ChemSusChem 9, 1544
1558. Gallage, N.J., Møller, B.L., 2015. Vanillin
bioconversion and bioengineering of the most popular plant flavor and its de novo biosynthesis in the vanilla orchid. Mol. Plant 8, 40
57.

260

Separation of Functional Molecules in Food by Membrane Technology

Garcı´a, A., Toledano, A., Serrano, L., Egu¨e´s, I., Gonza´lez, M., Marı´n, F., et al., 2009. Characterization of lignins obtained by selective precipitation. Sep. Purif. Technol. 68, 193
198. ´ ., Labidi, J., 2010. Study of the antioxidant capacity of Garcı´a, A., Toledano, A., Andre´s, M.A Miscanthus sinensis lignins. Process Biochem. 45, 935
940. Gellerstedt, G., 2015. Softwood kraft lignin: raw material for the future. Ind. Crops Prod. 77, 845
854. Gibson, G., Nutrition, M.R., 1995. Dietary modulation of the human colonic microbiota: introducing the concept of prebiotics. J. Nutr. 125, 1401
1412. Gillet, S., Aguedo, M., Petitjean, L., Morais, A.R.C., da Costa Lopes, A.M., Łukasik, R.M., et al., 2017. Lignin transformations for high value applications: towards targeted modifications using green chemistry. Green Chem. 19, 4200
4233. Giummarella, N., Lindgren, C., Lindstro¨m, M.E., Henriksson, G., 2016. Lignin prepared by ultrafiltration of black liquor: investigation of solubility, viscosity, and ash content. BioResources 11, 3494
3510. Glasser, W., Northey, R., Schultz, T. (Eds.), 1999. Lignin: Historical, Biological, and Materials Perspectives, ACS Symposium. American Chemical Society, Washington, DC. Gonza´lez Alriols, M., Tejado, A., Blanco, M., Mondragon, I., Labidi, J., 2009. Agricultural palm oil tree residues as raw material for cellulose, lignin and hemicelluloses production by ethylene glycol pulping process. Chem. Eng. J. 148, 106
114. Gonza´lez Alriols, M., Garcı´a, A., Llano-ponte, R., Labidi, J., 2010. Combined organosolv and ultrafiltration lignocellulosic biorefinery process. Chem. Eng. J. 157, 113
120. Hallac, B.B., Pu, Y., Ragauskas, A.J., 2010. Chemical transformations of Buddleja davidii lignin during ethanol organosolv pretreatment. Energy & Fuels 24, 2723
2732. Havkin-Frenkel, D., Belanger, F.C., 2008. Biotechnological production of vanillin. In: HavkinFrenkel, D., Belanger, F.C. (Eds.), Biotechnology in Flavor Production. Blackwell, Oxford, pp. 83
103. Hintz, H.L., 2001. Paper: pulping and bleaching. Encyclopedia of Materials: Science and Technology. Elsevier, Amsterdam, pp. 6707
6711. Hua, D., Ma, C., Song, L., Lin, S., Zhang, Z., Deng, Z., et al., 2007. Enhanced vanillin production from ferulic acid using adsorbent resin. Appl. Microbiol. Biotechnol. 74, 783
790. Huber, P., Burnet, A., Petit-Conil, M., 2014. Scale deposits in kraft pulp bleach plants with reduced water consumption: a review. J. Environ. Manage. 141, 36
50. Humpert, D., Ebrahimi, M., Czermak, P., 2016. Membrane technology for the recovery of lignin: a review. Membranes (Basel) 6, 42. Hussin, M.H., Rahim, A.A., Mohamad Ibrahim, M.N., Perrin, D., Brosse, N., 2015. Enhanced properties of oil palm fronds (OPF) lignin fractions produced via tangential ultrafiltration technique. Ind. Crops Prod. 66, 1
10. Ibrahim, M.N.M., Sipaut, C.S., Yusof, N.N.M., 2009. Purification of vanillin by a molecular imprinting polymer technique. Sep. Purif. Technol. 66, 450
456. Jo¨nsson, A.S., 2016. Membranes for lignin and hemicellulose recovery in pulp mills. In: Figoli, A., Cassano, A., Basile, A. (Eds.), Membrane Technologies for Biorefining, 2016, Elsevier, Amsterdam, 105
133. Jo¨nsson, A.S., Wallberg, O., 2009. Cost estimates of kraft lignin recovery by ultrafiltration. Desalination 237, 254
267. Jo¨nsson, A.S., Nordin, A.K., Wallberg, O., 2008. Concentration and purification of lignin in hardwood kraft pulping liquor by ultrafiltration and nanofiltration. Chem. Eng. Res. Des. 86, 1271
1280.

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

261

Kajikawa, H., Kudo, H., Kondo, T., Jodai, K., Honda, Y., Kuwahara, M., et al., 2000. Degradation of benzyl ether bonds of lignin by ruminal microbes. FEMS Microbiol. Lett. 187, 15
20. Kaur, B., Chakraborty, D., 2013. Biotechnological and molecular approaches for vanillin production: a review. Appl. Biochem. Biotechnol. 169, 1353
1372. Kautto, J., Realff, M.J., Ragauskas, A.J., 2013. Design and simulation of an organosolv process for bioethanol production. Biomass Convers. Biorefinery 3, 199
212. Kaygorodov, K.L., Chelbina, Y.V., Tarabanko, V.E., Tarabanko, N.V., 2010. Extraction of vanillin by aliphatic alcohols. J. Sib. Fed. Univ. Chem. 3, 228
233. Kekana, P., Sithole, B., Ramjugernath, D., 2016. Stirred cell ultrafiltration of lignin from black liquor generated from South African kraft mills. South African J. Sci. 112 (11/12), 1
7. Keyoumu, A., Sjo¨dahl, R., Henriksson, G., Ek, M., Gellerstedt, G., Lindstro¨m, M.E., 2004. Continuous nano- and ultra-filtration of kraft pulping black liquor with ceramic filters: a method for lowering the load on the recovery boiler while generating valuable sideproducts. Ind. Crops Prod. 20, 143
150. Kim, J.S., Lee, Y.Y., Kim, T.H., 2016. A review on alkaline pretreatment technology for bioconversion of lignocellulosic biomass. Bioresour. Technol. 199, 42
48. Kirk, T.K., Farrell, R.L., 1987. Enzymatic “combustion”: the microbial degradation of lignin. Annu. Rev. Microbiol. 41, 465
501. Koutinas, A.A., Vlysidis, A., Pleissner, D., Kopsahelis, N., Lopez Garcia, I., Kookos, I.K., et al., 2014. Valorization of industrial waste and by-product streams via fermentation for the production of chemicals and biopolymers. Chem. Soc. Rev. 43, 2587. Kova´cs, Z., Discacciati, M., Samhaber, W., 2009. Modeling of batch and semi-batch membrane filtration processes. J. Memb. Sci. 327, 164
173. Krawczyk, H., Jo¨nsson, A., 2011. Separation of dispersed substances and galactoglucomannan in thermomechanical pulp process water by microfiltration. Separ. Purif. Tech. 79 (1), 43
49. Kumar, A.K., Sharma, S., 2017. Recent updates on different methods of pretreatment of lignocellulosic feedstocks: a review. Bioresour. Bioprocess. 4, 7. Li, C., Zhao, X., Wang, A., Huber, G.W., Zhang, T., 2015. Catalytic transformation of lignin for the production of chemicals and fuels. Chem. Rev. 115, 11559
11624. Li, Q., Serem, W.K., Dai, W., Yue, Y., Naik, M.T., Xie, S., et al., 2017. Molecular weight and uniformity define the mechanical performance of lignin-based carbon fiber. J. Mater. Chem. A 5, 12740
12746. Liu, G., Liu, Y., Ni, J., Shi, H., Qian, Y., 2004. Treatability of kraft spent liquor by microfiltration and ultrafiltration. Desalination. 160 (2), 131
141. Lora, J., 2008. Industrial commercial lignins: sources, properties and applications. In: Belgacem, M.N., Gandini, A. (Eds.), Monomers, Polymers and Composites from Renewable Resources. Elsevier, Amsterdam, pp. 225
241. Lora, J.H., Glasser, W.G., 2002. Recent industrial applications of lignin: a sustainable alternative to nonrenewable materials. J. Polym. Environ. 10, 39
48. Lundqvist, J., Teleman, A., Junel, L., Zacchi, G., Dahlman, O., Tjerneld, F., et al., 2002. Isolation and characterization of galactoglucomannan from spruce (Picea abies). Carbohydr. Polym. 48, 29
39. Mahmood, N., Yuan, Z., Schmidt, J., Xu, C.C., 2016. Depolymerization of lignins and their applications for the preparation of polyols and rigid polyurethane foams: a review. Renew. Sustain. Energy Rev. 60, 317
329. Mamun, A.A., Bledzki, A.K., 2013. Micro fibre reinforced PLA and PP composites: enzyme modification, mechanical and thermal properties. Compos. Sci. Technol. 78, 10
17.

262

Separation of Functional Molecules in Food by Membrane Technology

Mulder, M., 1991. Basic Principles of Membrane Technology. Kluwer Academic Publishers, Dordrecht, The Netherlands. Nelson, J.L., Alexander, J.W., Gianotti, L., Chalk, C.L., Pyles, T., 1994. Influence of dietary fiber on microbial growth in vitro and bacterial translocation after burn injury in mice. Nutrition 10, 32
36. Niemi, P., 2016. Enzymatic Fractionation of Brewer’s Spent Grain and Bioconversion of LigninRich Fractions in a Colon Model In Vitro. Aalto University, Finland. Niemi, P., Aura, A.-M., Maukonen, J., Smeds, A.I., Mattila, I., Niemela¨, K., et al., 2013. Interactions of a lignin-rich fraction from Brewer’s spent grain with gut microbiota in vitro. J. Agric. Food Chem. 61, 6754
6762. Noble, R.D., Stern, S.A. (Eds.), 1995. Membrane Separations Technology: Principles and Applications (Vol. 2). Elsevier, Amsterdam. Octave, S., Thomas, D., 2009. Biorefinery: toward an industrial metabolism. Biochimie 91, 659
664. ¨ hman, F., Wallmo, H., Theliander, H., 2007. An improved method for washing lignin precipiO tated from kraft black liquor—the key to a new biofuel. Filtration 7, 309
315. Overhage, J., Steinbu¨chel, A., Priefert, H., 2003. Highly efficient biotransformation of eugenol to ferulic acid and further conversion to vanillin in recombinant strains of Escherichia coli. Appl. Environ. Microbiol. 69, 6569
6576. Pabby, A.K., Rizvi, S.S.H., Sastre, A.M., 2015. Handbook of membrane separation: chemical, pharmaceutical. Food, and Biotechnological Applications. CRC Press, Boca Raton, FL. Pandey, M.P., Kim, C.S., 2011. Lignin depolymerization and conversion: a review of thermochemical methods. Chem. Eng. Technol. 34, 29
41. Parthasarathi, R., Romero, R.A., Redondo, A., Gnanakaran, S., 2011. Theoretical study of the remarkably diverse linkages in lignin. J. Phys. Chem. Lett. 2, 2660
2666. Patil, S.V., Argyropoulos, D.S., 2017. Stable organic radicals in lignin: a review. ChemSusChem 10, 3284
3303. Patt, R., Kordsachia, O., Suttinger, R., 2012. Pulp. Ullmann’s Encyclopedia of Industrial Chemistry. Wiley-VCH Verlag, Weinheim, pp. 476
539. Pedersen, M., Meyer, A.S., 2010. Lignocellulose pretreatment severity
relating pH to biomatrix opening. N. Biotechnol. 27, 739
750. Persson, T., Krawczyk, H., Nordin, A.-K., Jo¨nsson, A.-S., 2010. Fractionation of process water in thermomechanical pulp mills. Bioresour. Technol. 101, 3884
3892. Phillip, L., Idziak, E.S., Kubow, S., 2000. The potential use of lignin in animal nutrition, and in modifying microbial ecology of the gut. Eastern Nutrition Conference. Animal Nutrition Association of Canada, Montreal (Canada), pp. 165
184. Plastics Europe—Plastics—the Facts 2016 [WWW Document], 2016. Ragauskas, A.J., Beckham, G.T., Biddy, M.J., Chandra, R., Chen, F., Davis, M.F., et al., 2014. Lignin valorization: improving lignin processing in the biorefinery. Science 344 (6185), 1246843. Available from: https://doi.org/10.1126/science.1246843. Rahimi, A., Ulbrich, A., Coon, J.J., Stahl, S.S., 2014. Formic-acid-induced depolymerization of oxidized lignin to aromatics. Nature 515, 249
252. Rai, S., Duta, P.K., Mehrotra, G.K., 2017. Lignin incorporated antimicrobial chitosan film for food packaging application. J. Polym. Mater. 34, 171
183. Restolho, J.A., Prates, A., de Pinho, M.N., Afonso, M.D., 2009. Sugars and lignosulphonates recovery from eucalyptus spent sulphite liquor by membrane processes. Biomass Bioenergy 33, 1558
1566.

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

263

Rhim, J.-W., Park, H.-M., Ha, C.-S., 2013. Bio-nanocomposites for food packaging applications. Prog. Polym. Sci. 38, 1629
1652. Ricke, S.C., Van Der Aar, P.J., Fahey, G.C., Berger, L.L., 1982. Influence of dietary fibers on performance and fermentation characteristics of gut contents from growing chicks. Poult. Sci. 61, 1335
1343. Rinaldi, R., Jastrzebski, R., Clough, M.T., Ralph, J., Kennema, M., Bruijnincx, P.C.A., Weckhuysen, B.M., et al., 2016. Paving the way for lignin valorisation: recent advances in bioengineering, biorefining and catalysis. Angew. Chemie—Int. Ed. 55 (2016), 8164
8215. Rodrigues Pinto, P.C., Borges da Silva, E.A., Rodrigues, A.E., 2012. Lignin as source of fine chemicals: vanillin and syringaldehyde BT. In: Baskar, C., Baskar, S., Dhillon, R.S. (Eds.), Biomass Conversion: The Interface of Biotechnology, Chemistry and Materials Science. Springer, Berlin, Heidelberg, pp. 381
420. Rueda, C., Calvo, P.A., Moncalia´n, G., Ruiz, G., Coz, A., 2015a. Biorefinery options to valorize the spent liquor from sulfite pulping. J. Chem. Technol. Biotechnol. 90, 2218
2226. Rueda, C., Ferna´ndez-Rodrı´guez, J., Ruiz, G., Llano, T., Coz, A., 2015b. Monosaccharide production in an acid sulfite process: kinetic modeling. Carbohydr. Polym. 116, 18
25. Sadeghifar, H., Venditti, R., Jur, J., Gorga, R.E., Pawlak, J.J., 2017. Cellulose
lignin biodegradable and flexible UV protection film. ACS Sustain. Chem. Eng. 5, 625
631. Saeed, A., Jahan, M.S., Li, H., Liu, Z., Ni, Y., van Heiningen, A., 2012. Mass balances of components dissolved in the pre-hydrolysis liquor of kraft-based dissolving pulp production process from Canadian hardwoods. Biomass Bioenergy 39, 14
19. Sakagami, H., Kushida, T., Oizumi, T., Nakashima, H., Makino, T., 2010. Distribution of lignin
carbohydrate complex in plant kingdom and its functionality as alternative medicine. Pharmacol. Ther. 128, 91
105. Schorr, D., Diouf, P.N., Stevanovic, T., 2014. Evaluation of industrial lignins for bio composites production. Ind. Crops Prod. 52, 65
73. Sciubba, L., Di Gioia, D., Fava, F., Gostoli, C., 2009. Membrane-based solvent extraction of vanillin in hollow fiber contactors. Desalination 241, 357
364. Servaes, K., Varhimo, A., Dubreuil, M., Bulut, M., Vandezande, P., Siika-aho, M., et al., 2017. Purification and concentration of lignin from the spent liquor of the alkaline oxidation of woody biomass through membrane separation technology. Ind. Crops Prod. 106, 86
96. Sevastyanova, O., Helander, M., Chowdhury, S., Lange, H., Wedin, H., Zhang, L., et al., 2014. Tailoring the molecular and thermo-mechanical properties of kraft lignin by ultrafiltration. J. Appl. Polym. Sci. 131 (18), 9505
9515. Shankar, S., Reddy, J.P., Rhim, J.W., 2015. Effect of lignin on water vapor barrier, mechanical, and structural properties of agar/lignin composite films. Int. J. Biol. Macromol. 81, 267
273. Shao, L., Zhang, X., Chen, F., Xu, F., 2017. Fast pyrolysis of Kraft lignins fractionated by ultrafiltration. J. Anal. Appl. Pyrolysis 128, 27
34. Silva, E.A.B., da, Zabkova, M., Arau´jo, J.D., Cateto, C.A., Barreiro, M.F., Belgacem, M.N., et al., 2009. An integrated process to produce vanillin and lignin-based polyurethanes from Kraft lignin. Chem. Eng. Res. Des. 87, 1276
1292. Sinha, A.K., Sharma, U.K., Sharma, N., 2008. A comprehensive review on vanilla flavor: extraction, isolation and quantification of vanillin and others constituents. Int. J. Food Sci. Nutr. 59, 299
326. Sondhi, R., Bhave, R., Jung, G., 2003. Applications and benefits of ceramic membranes. Membr. Technol. 2003 (11), 5
8.

264

Separation of Functional Molecules in Food by Membrane Technology

Tejado, A., Pen˜a, C., Labidi, J., Echeverria, J.M., Mondragon, I., 2007. Physico-chemical characterization of lignins from different sources for use in phenol
formaldehyde resin synthesis. Bioresour. Technol. 98, 1655
1663. Thompson, G., Swain, J., Kay, M., Forster, C.F., 2001. The treatment of pulp and paper mill effluent: a review. Bioresour. Technol. 77, 275
286. Thuenemann, E.C., 2015. Dynamic digestion models: general introduction. In: Verhoeckx, K., Cotter, P., Lo´pez-Expo´sito, I., Kleiveland, C., Lea, T., Mackie, A., et al.,The Impact of Food Bioactives on Health: In Vitro and Ex Vivo Models. Springer International Publishing, Cham, pp. 33
36. Toledano, A., Serrano, L., Garcia, A., Mondragon, I., Labidi, J., 2010. Comparative study of lignin fractionation by ultrafiltration and selective precipitation. Chem. Eng. J. 157, 93
99. Toledano, A., Serrano, L., Balu, A.M., Luque, R., Pineda, A., Labidi, J., 2013. Fractionation of organosolv lignin from olive tree clippings and its valorization to simple phenolic compounds. ChemSusChem 6, 529
536. Toledo, M.C.F., Kuznesof, P.M., 2008. Calcium Lignosulfonate (40
65). In: 69th JECFA, pp. 1
8. Tran, H., Vakkilainnen, E.K., 2012. The Kraft chemical recovery process. In: TAPPI Kraft Recover. Valencia, Z., Chavez, E.R., 1997. Lignin as a purified dietary fiber supplement for piglets. Nutr. Res. 17, 1517
1527. Vanholme, R., Demedts, B., Morreel, K., Ralph, J., Boerjan, W., 2010. Lignin biosynthesis and structure. Plant Physiol. 153, 895
905. Vegas, R., Luque, S., Alvarez, J.R., Alonso, J.L., Domı´nguez, H., Parajo´, J.C., 2006. Membraneassisted processing of xylooligosaccharide-containing liquors. J. Agric. Food Chem. 54, 5430
5436. Vinardell, M.P., Ugartondo, V., Mitjans, M., 2008. Potential applications of antioxidant lignins from different sources. Ind. Crops Prod. 27, 220
223. Vishtal, A., Kraslawski, A., 2011. Challenges in industrial applications of technical lignins. BioResources 6, 3547
3568. Wallberg, O., Jo¨nsson, A.S., Wimmerstedt, R., 2003a. Fractionation and concentration of kraft black liquor lignin with ultrafiltration. Desalination 154, 187
199. Wallberg, O., Jarwon, A.-S., Wimmerstedt, R., 2003b. Ultrafiltration of kraft black liquor with a ceramic membrane. Desalination 156, 145
153. Wallberg, O., Jo¨nsson, A.-S., 2006. Separation of lignin in kraft cooking liquor from a continuous digester by ultrafiltration at temperatures above 100 C. Desalination 195, 187
200. Walton, N.J., Narbad, A., Faulds, C., Williamson, G., 2000. Novel approaches to the biosynthesis of vanillin. Curr. Opin. Biotechnol. 11, 490
496. Wang, Z., Chen, K., Li, J., Wang, Q., Guo, J., 2010. Separation of vanillin and syringaldehyde from oxygen delignification spent liquor by macroporous resin adsorption. CLEAN Soil, Air, Water 38, 1074
1079. Weng, J.-K., Li, X., Bonawitz, N.D., Chapple, C., 2008. Emerging strategies of lignin engineering and degradation for cellulosic biofuel production. Curr. Opin. Biotechnol. 19, 166
172. Wildschut, J., Smit, A.T., Reith, J.H., Huijgen, W.J.J., 2013. Ethanol-based organosolv fractionation of wheat straw for the production of lignin and enzymatically digestible cellulose. Bioresour. Technol. 135, 58
66. Xu, Z., Huang, F., 2014. Pretreatment methods for bioethanol production. Appl. Biochem. Biotechnol. 174, 43
62.

Lignin Separation and Fractionation by Ultrafiltration Chapter | 7

265

Yang, D., Li, H., Qin, Y., Zhong, R., Bai, M., Qiu, X., 2015. Structure and properties of sodium lignosulfonate with different molecular weight used as dye dispersant. J. Dispers. Sci. Technol. 36, 532
539. Zabkova, M., Borges da Silva, E.A., Rodrigues, A.E., 2007a. Recovery of vanillin from Kraft lignin oxidation by ion-exchange with neutralization. Sep. Purif. Technol. 55, 56
68. ˇ ´ , M., da Silva, E.A.B., Rodrigues, A.E., 2007b. Recovery of vanillin from lignin/vanillin Zabkova mixture by using tubular ceramic ultrafiltration membranes. J. Memb. Sci. 301, 221
237. Zakzeski, J., Bruijnincx, P.C.A., Jongerius, A.L., Weckhuysen, B.M., 2010. The catalytic valorization of lignin for the production of renewable chemicals. Chem. Rev. 110, 3552
3599. Zemek, J., Koˇs´ıkova´, B., Augustı´n, J., Joniak, D., 1979. Antibiotic properties of lignin components. Folia Microbiol. (Praha). 24, 483
486. Zhang, Y.H.P., 2011. What is vital (and not vital) to advance economically-competitive biofuels production. Process Biochem. 46 (2011), 2091
2110. Zhang, J., Cui, J., Yin, T., Sun, L., 2013. Activated effect of lignin on α-amylase. Food Chem. 141, 2229
2237. Zhang, J., Xiao, L., Yang, Y., Wang, Z., Li, G., 2014. Lignin binding to pancreatic lipase and its influence on enzymatic activity. Food Chem. 149, 99
106. Zhao, X., Li, S., Wu, R., Liu, D., 2017. Organosolv fractionating pre-treatment of lignocellulosic biomass for efficient enzymatic saccharification: chemistry, kinetics, and substrate structures. Biofuels, Bioprod. Biorefining 11, 567
590. Zheng, L., Zheng, P., Sun, Z., Bai, Y., Wang, J., Guo, X., 2007. Production of vanillin from waste residue of rice bran oil by Aspergillus niger and Pycnoporus cinnabarinus. Bioresour. Technol. 98, 1115
1119. Zhou, M., Qiu, X., Yang, D., Lou, H., 2006. Properties of different molecular weight sodium lignosulfonate fractions as dispersant of coal-water slurry. J. Dispers. Sci. Technol. 27, 851
856. Zhu, J.Y., Pan, X.J., Wang, G.S., Gleisner, R., 2009. Sulfite pretreatment (SPORL) for robust enzymatic saccharification of spruce and red pine. Bioresour. Technol. 100, 2411
2418. Ziesig, R., Tomani, P., Theliander, H., 2014. Production of a pure lignin product. Part 2: separation of lignin from membrane filtration permeates of black liquor. Cellul. Chem. Technol. Cellul. Chem. Technol. 48, 805
811. Zinovyev, G., Sumerskii, I., Korntner, P., Sulaeva, I., Rosenau, T., Potthast, A., 2017. Molar mass-dependent profiles of functional groups and carbohydrates in kraft lignin. J. Wood Chem. Technol. 37, 171
183. Zydney, A.L., 1997. Stagnant film model for concentration polarization in membrane systems. J. Memb. Sci. 130 (1), 275
281.

FURTHER READING Boerjan, W., Ralph, J., Baucher, M., 2003. Lignin biosynthesis. Annu. Rev. Plant Biol. 54, 519
546.

This page intentionally left blank

Chapter 8

Membrane Separations in the Dairy Industry George Q. Chen1,2, Thomas S.H. Leong1,2,3, Sandra E. Kentish1,2, Muthupandian Ashokkumar1,3 and Gregory J.O. Martin1,2 1

ARC Dairy Innovation Hub, The University of Melbourne, Parkville, VIC, Australia, Department of Chemical Engineering, The University of Melbourne, Parkville, VIC, Australia, 3 School of Chemistry, The University of Melbourne, Parkville, VIC, Australia 2

Chapter Outline 8.1 Introduction 8.2 Milk and Whey 8.2.1 Milk Components and Composition 8.2.2 Milk Proteins 8.2.3 Casein Micelles 8.3 Milk Processing 8.3.1 On Farm Concentration 8.3.2 Control of Microbial Growth 8.3.3 Milk Protein Fractionation 8.3.4 Milk Fat Fractionation 8.4 Cheese Processing 8.4.1 Cheese Milk Standardization 8.5 Whey Processing 8.5.1 Whey Protein Concentration

267 270 270 272 272 273 273 273 275 278 278 279 280 280

8.5.2 Whey Protein Fractionation 281 8.5.3 Whey Concentration 282 8.5.4 Whey Demineralization 283 8.5.5 Cheese Brine Purification 284 8.6 Waste Treatment 285 8.7 Fouling 287 8.8 Application of Sonication 289 8.8.1 Physical and Chemical Effects of Ultrasound 289 8.8.2 Ultrasonic Reduction of Fouling Buildup 290 8.8.3 Mechanisms of Ultrasonic Membrane Cleaning in Dairy Applications 291 8.8.4 Application of Ultrasound to Membrane Cleaning 292 8.9 Conclusions and Future Trends 295 References 295

8.1 INTRODUCTION The dairy industry has successfully incorporated membrane technologies throughout different manufacturing stages, from on-farm raw milk concentration, to milk processing and to effluent treatment. Milk contains a wide range

Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00008-5 © 2019 Elsevier Inc. All rights reserved.

267

268

Separation of Functional Molecules in Food by Membrane Technology

of functional molecules that can be best utilized when made available separately. Milk is a complex colloidal system, containing species spanning a broad size distribution from ions up to large fat globules (see Fig. 8.1 and Section 8.2). This makes membrane separation a particularly suitable technology for fractionating and concentrating the various functional components in milk. For example, ensuring a consistent protein composition in milk leads to a higher quality and consistency of processed consumer products (cheese and yogurt). The ability to recover specific milk components makes possible the development of new products with desired functionalities (e.g., bioactive peptides and edible coatings). For more than 40 years, conventional pressure-driven membranes including microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), and reverse osmosis (RO) have been applied to dairy processing. UF is the most widely used membrane type for dairy, mainly for adjusting and standardizing the protein level in milk and whey. UF can retain suspended solids and solutes of molecular weight higher than about 8000 Da (i.e., fat globules and all proteins), while allowing the passage of lactose and soluble salts (Fig. 8.1). This is followed by microfiltration, which has been successfully used for bacterial removal from milk and casein micelle enrichment in cheese milk (Brans et al., 2004). Microfiltration is capable of removing colloidal and suspended particles in the range of 0.05
10 μm, while allowing the smaller soluble proteins, lactose, and minerals to permeate through. The main application of NF in the dairy industry is partial whey demineralization. NF retains molecules with molecular weight greater than 100
500 Da. In addition to molecular size, components are separated by NF membranes based on charge. For instance, most multivalent ions can be rejected but the removal of monovalent ions by NF membranes varies

FIGURE 8.1 Examples of applications of membrane processes in the dairy industry.

Membrane Separations in the Dairy Industry Chapter | 8

269

between 50% and 90% depending on the membrane material and the operating parameters (Nath, 2008; Oatley-Radcliffe et al., 2017). Calcium and magnesium, which are arguably the most important ions in milk/whey, can hence be effectively retained in the retentate. RO membranes only allow passage of water molecules, making this an ideal candidate for concentrating milk and whey prior to evaporation and drying, thereby reducing energy consumption. Some examples of membrane applications in the dairy industry are presented in Fig. 8.2. In recent years, emerging membrane technologies such as electrodialysis, membrane distillation, and forward osmosis have been developed and advanced, making their application in the dairy industry more promising. Electrodialysis can be used to remove ionic species (e.g., organic acids and salts) from milk and whey by using an electrical driving force across ion exchange membranes. Molecules migrate through these membranes based on their charge and molecular size. Proteins are completely rejected by ion exchange membranes, making electrodialysis an effective alternative technology for whey demineralization. Membrane distillation is an emerging membrane process that is thermally driven and uses a vapor pressure gradient between the feed and permeate side to achieve the separation. A hydrophobic membrane is used as the barrier to the liquid phase, allowing the vapor phase (i.e., water vapor) to pass through. Another emerging membrane technology, forward osmosis, has recently been developed to remove water from a contaminated stream. It is a diffusive process that uses a high osmotic potential created by a concentrated draw solution (typically a salt solution) to “draw” water from the less

FIGURE 8.2 Some examples of commercial applications of membrane filtration processes for powder production of milk and whey derivatives in the dairy industry. Heat treatment processes are not included. D40, demineralized whey with 40% mineral removal; MF, microfiltration; NF, nanofiltration; UF ultrafiltration; WPC, whey protein concentrate.

270

Separation of Functional Molecules in Food by Membrane Technology

concentrated contaminated feed stream. To reduce the osmotic potential difference, water migrates from the contaminated feed stream into the permeate, thereby diluting the salty permeate stream, and concentrating the feed. Since no transmembrane pressure (TMP) gradient is needed for membrane distillation and forward osmosis processes, they are less susceptible to membrane fouling at low concentration factors. Both technologies have been investigated recently for concentrating milk and whey. This chapter provides an overview of the applications of membrane technology in the dairy industry, including milk processing, cheese production, whey processing, and waste treatment. It begins with a basic description of the key physicochemical characteristics of milk and whey, which will be further discussed in relation to membrane performance, fouling and products. The use of ultrasonic technology for membrane cleaning and other potential applications will also be explored.

8.2 MILK AND WHEY Milk is a highly complex fluid, comprising an emulsion of fat globules, a colloidal suspension of casein micelles and whey proteins, and a solution of lactose and minerals. The physicochemical properties of milk are dependent on the concentration, temperature, and pH; all of which are important parameters during membrane filtration. As such, milk chemistry can play an important role in the performance of membrane filtration processes and must be understood in relation to operational parameters.

8.2.1 Milk Components and Composition Milk is essentially a delivery vehicle for nutrients to young mammals. While the composition varies between species, the main components of milk are water, fat, protein, lactose, and minerals. This chapter considers cows’ milk, the most important globally (Table 8.1). The typical composition of this milk is shown in Table 8.1. The composition of milk varies in relation to lactation periods (Fox and McSweeney, 1998; Rattray and Jelen, 1996). One of the important advances achieved by membrane filtration has been the standardization of milk across such periods, in particular in relation to the protein and fat composition, that allows the production of more consistent products. Crucial to membrane processing, the key components of milk vary greatly in size, allowing effective separation. Fat is present in milk as fat globules—droplets of triacylglycerides surrounded by a protein- and phospholipid-rich membrane (Michalski et al., 2002). These globules are the largest species in milk, typically in the range 1
10 μm in diameter (Michalski et al., 2001). The removal of most of the fat by centrifugation results in skim milk, which typically has ,0.1 g/L of residual fat.

Membrane Separations in the Dairy Industry Chapter | 8

271

TABLE 8.1 Average Composition of Bovine Milk: Concentration and Size Distribution Description

Concentration (g/100 g)

Size/Molecular Weight Range

Water

87.1

Dry matter

12.9

Fat globules

4

0.1
15 μm, average 3.4 μm

Casein (in micelles)

2.6

20
300 nm, average 110 nm

Serum proteins

0.7

3
6 nm

α-Lactalbumin

0.12

14 kDa

β-Lactoglobulin

0.32

18 kDa

Bovine serum albumin

0.04

66 kDa

Proteose-pepton

0.08

4
40 kDa

Immunoglobulins

0.08

150
900 kDa

Lactoferrin

0.01

86 kDa

Transferrin

0.01

76 kDa

Others

0.04

Lactose

4.6

Ash

0.7

Organic acids

0.17

Other

0.15

0.35 kDa

Source: Modified from Brans, G., et al., 2004. Membrane fractionation of milk: state of the art and challenges. J. Membr. Sci. 243(1): 263
272.

After water, lactose is the most abundant component of cow milk, with a concentration of approximately 4.6 g/L (Brans et al., 2004). While lactose is important as a substrate for fermentation in the production of some dairy products, it typically plays a relatively minor role in membrane filtration apart from its contribution to bulk viscosity (Ng et al., 2017, 2018). The proteins and minerals have a much greater influence on membrane processes, as they interact dynamically with each other, and are the key components in fouling (Ng et al., 2017, 2018). Milk contains an array of mineral salts and ions that form complex species with each other and the proteins (Holt, 2004). Of particular importance are the calcium and phosphate ions in milk, which as described in Section 8.2.3, are a key component of the casein micelles.

272

Separation of Functional Molecules in Food by Membrane Technology

8.2.2 Milk Proteins Milk contains two broad groups of proteins—the caseins and the whey proteins. The caseins represent about 80% of the total protein in milk and consist of four different proteins: αs1-CN, αs2-CN, β-CN, and κ-CN, present at a ratio of roughly 4:1:4:1 (Fox and McSweeney, 1998). Most of the casein in milk is present as large protein aggregates called casein micelles, which are described in detail in the following subsection. The casein proteins are phosphorylated (Farrell et al., 2004) and can be considered to have a relative open and flexible structure (Holt and Sawyer, 1993). The whey proteins include β-lactoglobulin (β-LG), α-lactalbumin (α-LA), bovine serum albumin (BSA), immunoglobulins (IG), and other minor but potentially important proteins, for example, lactoferrin (Farrell et al., 2004). In contrast to the casein micelles, the whey proteins can be considered as small, soluble, globular proteins of approximately 3
6 nm in diameter that have a defined tertiary structure. Importantly for milk processing, and again in contrast to the caseins, the whey proteins are sensitive to denaturation and aggregation (Hillier and Lyster, 1979; McKenna and O’Sullivan, 1971; Oldfield et al., 2005). This is particularly relevant to thermal processing, as denaturation occurs above about 63 C. One useful attribute of membrane processing is that it can be conducted below whey protein denaturation temperatures to retain the proteins in their native state (Ng et al., 2017).

8.2.3 Casein Micelles Most of the casein in milk is present in the form of approximately spherical aggregates with a diameter in the range 50
250 nm. Although having been studied for many decades, the exact structure of casein micelles is still debated (Horne, 2006; Dalgleish, 2011). It is well known however, that κ-CN is present on the external surface, with the hydrophilic, glycosylated section of the protein protruding from the micelles to form a so-called hairy layer that provides steric and electrostatic stabilization (Walstra, 1990). As such, the casein micelles are highly stable unless the hairy layer is removed (i.e., by rennet gelation during cheese making) (Walstra, 1990) or collapsed (i.e., by reducing the pH to 4.6) (Lucey and Singh, 1997). In addition to the casein protein, casein micelles include about 8% (w/w dry matter) of colloidal calcium phosphate (Bouchoux et al., 2010). This mineral component is present in nanosized clusters that are believed to contribute to the structural integrity of the casein micelles (Dalgleish, 2011). The calcium and phosphate are in equilibrium with the serum minerals, which is dependent on the pH and temperature (Davies and White, 1960; Liu et al., 2013). The casein micelles are also highly hydrated, with water representing about 75% of their overall mass (de Kruif, 1999; Walstra, 1979; Liu et al., 2012). As large colloidal particles these casein micelles can be

Membrane Separations in the Dairy Industry Chapter | 8

273

retained by UF and microfiltration membranes, forming a gel layer (RabillerBaudry et al., 2005). Although most of the casein ( . 95%) is present as casein micelles, the rest is present as soluble casein in the serum (Rose, 1968). The equilibrium between serum and micellar casein is tied to the equilibrium between calcium and phosphate species and is temperature and pH dependent (Liu et al., 2013). This is of relevance to membrane filtration, particularly when microfiltration is used to separate casein (micelles) from whey proteins. The equilibria of calcium and casein between the serum and the casein micelles are also important for the composition of filtrate and retentates (Liu et al., 2014) and therefore the properties and quality of the resulting products.

8.3 MILK PROCESSING 8.3.1 On Farm Concentration The concept of concentrating milk on the farm before transportation to dairy processors arose in the early 1970s, aiming to lower transportation costs. UF or RO, and sometimes NF, is used for this purpose. Despite the fact that these membrane technologies were already well developed for on-farm concentration of milk, it was not until 1996 that the first UF plant started producing concentrated milk on a farm in New Mexico, USA (Fleming, 1999). UF and RO membranes can typically achieve a concentration factor of 3.5 and 2.5, respectively. For this approach to be economically viable, it was estimated that on farm concentration of milk with a factor of three using UF should only be introduced to large farms with at least 1000 cows (Slack et al., 1982). Like many membrane processes in the dairy industry, these plants operate below 8 C to limit bacterial growth, minimize disruption to fat globules, and maintain the dispersion of milk fat. However, strict sanitary control must be in place to meet the regulatory requirements imposed by relevant agencies. For example, on-farm concentrated milk without pasteurization must meet the total bacterial count limit of Grade A dairy foods in the United States (i.e., , 300 3 103 colony forming units (CFU) per mL) (Mistry and Maubois, 2004).

8.3.2 Control of Microbial Growth Microbial growth control is one of the most important aspects of milk processing, as milk and dairy systems are susceptible to spoilage by microorganisms. Heat treatment (e.g., thermization, pasteurization, ultra high temperature (UHT)) is the most commonly used process to destroy pathogens and deactivate spoilage organisms (Muir, 2011). Tight control over the microbial content is necessary during downstream processing and for extending the shelf-life of milk products. However, heat treatment processes can

274

Separation of Functional Molecules in Food by Membrane Technology

result in heat-induced chemical and physico-chemical changes in milk, such as protein denaturation (Anema and Li, 2003), damage to creaming properties and sensory attributes (Fox et al., 2015; Meersohn, 1989), and nonenzymatic Maillard browning. Bacterial food-borne pathogens such as Salmonellae and Listeria, as well as natural nonpathogenic flora can be removed from milk by MF prior to protein fractionation or cheese making (Madec et al., 1992). Somatic cells and spores such as Clostridium spp. that would survive conventional heat treatment process (e.g., pasteurization) can cause issues in semihard and hard cheese manufacture. Microfiltration offers a non-thermal alternative for removing bacteria and spores from milk and milk related products. Inorganic membranes (usually ceramic) are used in the dairy industry because of their excellent thermal stability, durability, and ability to tolerate a wide range of chemicals and pH (Saboyainsta and Maubois, 2000; Baruah et al., 2006). The permeate flux of a typical MF membrane with a 1.4-μm pore size ranges from 500 to 700 LMH (L m-2 h-1) at a uniform transmembrane pressure (TMP) of around 0.5 bar (Saboyainsta and Maubois, 2000). A 4
5 decimal reduction factor of bacteria and spores can be achieved, with retention of over 99% of the bacteria. The first industrial plant using microfiltration for shelf life extension of pasteurized milk was built in Sweden. The effective retention of B. cereus successfully increased the shelf life from 6
8 to 16
21 days (Malmberg and Holm, 1988). The removal of bacteria from milk by MF caused little compositional change, with a negligible decrease in total protein (, 0.03%) (Hoffmann et al., 2006). Bactocatch systems developed by Alfa-Laval in the 1990s are commercial systems widely adopted for the production of drinking milk (Daufin et al., 2001). Fouling can occur during MF of milk due to pore blocking by bacteria and spores, partial pore blocking by bridging of casein micelles, surface adsorption of whey proteins, and in-pore fouling by whey proteins. To mitigate membrane fouling and maintain high permeate flux under low TMPs and high cross-flow velocities (typically 6
9 m/s (Saboyainsta and Maubois, 2000)), these MF plants were the first to be equipped with a uniform transmembrane pressure (UTMP) system (Sandblom, 1978). The concept of UTMP involves recirculating the permeate in a cocurrent direction to the retentate to create a homogeneous TMP along the membrane length, leading to less compact fouling layers and variation in membrane selectivity along the tube. A more recent and novel approach to achieving a uniform permeate flux is to produce inhomogeneous ceramic membranes with a gradient of membrane resistance. A high hydraulic resistance is required at the membrane inlet where the operational TMP is high. The resistance is gradually lowered along the membrane length. These are “third generation” graded permeability (GP) membranes (Ferna´ndez Garcı´a and Riera Rodrı´guez, 2015) that work either by varying the porosity of the membrane support (Membralox GP membranes, Pall-Exekia France) (Garcera and Toujas, 2000),

Membrane Separations in the Dairy Industry Chapter | 8

275

or by varying the thickness of the selective front layer of the membrane (Isoflux membranes, Tami-Industries France) (Skrzypek and Burger, 2010). Industrial MF plants with two stages can typically achieve a final volume reduction ratio (VRR) of 200, resulting in a retentate loss of less than 0.5% of the incoming milk. Microfiltered drinking milk is mainly found in the drinking milk market in Germany, Switzerland, and Austria (Ferna´ndez Garcı´a et al., 2013). As the size of large fat globules and small bacteria overlap (see Fig. 8.1), microfiltration is often applied to skim milk after cream is separated by centrifugation to enhance the separation efficiency. In addition, the presence of fat in whole milk would also make membrane fouling more pronounced. To adjust the fat-to-protein ratio, the cream is pasteurized and added to the microfiltered skim milk to produce microfiltered whole milk. In this approach, heat treatment is only applied to a small portion of milk (i.e., the cream) so that the organoleptic and sensory attributes of milk are retained. This extends the shelf life from 12 to 45 days at refrigerated temperatures (Saboyainsta and Maubois, 2000; Kumar et al., 2013; Goff and Griffiths, 2006). The use of ceramic membranes of a smaller pore diameter (0.8 μm (Fauquant et al., 2016) or 0.5 μm (Lindquist, 1998)) leads to an increase in bacterial removal by 2
3 decimal reductions compared with that achieved by a 1.4-μm pore diameter membrane. The smaller pore size results in a slight decrease in the casein micelle permeation rate, but the remaining bacterial count can be less than 1 CFU/mL. The sterility of the filtered milk obtained from smaller pore diameters was shown to decrease the heat treatment parameters required to produce drinking milk with a long shelf life to 96 C for 6 seconds (0.8 μm MF) (Fauquant et al., 2016) and ,100 C for 2 seconds (0.5 μm MF) (Saboyainsta and Maubois, 2000). Typical UHT milk requires a holding time of 4 seconds at 140 C. To reduce membrane hydrodynamic resistance, microsieves with a narrow pore size distribution (PSD) and a smooth inert silicon nitride surface have been developed using micromachining technology (Brans et al., 2004). The narrow PSD results in high reduction of bacteria, while the reduced hydrodynamic resistance lowers the TMP by two orders of magnitude, hence reducing fouling tendency and improving permeate flux (Kromkamp and Van, 2002). To date, flux values greater than 5000 LMH had been obtained in cross-flow filtration systems with back pulsing (Brito-de la Fuente et al., 2010; Carpintero-Tepole et al., 2014).

8.3.3 Milk Protein Fractionation Skim milk contains approximately 3.2% w/w protein, about 80% of which is casein and the remainder “serum” or whey protein. As casein micelles are much larger than the whey proteins (see Section 8.2), casein and whey protein can be fractionated by microfiltration membranes with a pore size

276

Separation of Functional Molecules in Food by Membrane Technology

diameter of 0.1
0.5 μm. This can be achieved without using chemical precipitation and neutralization to produce casein-enriched products and un-denatured “native” whey proteins. Both the permeate (rich in whey protein) and retentate (rich in casein micelles) from this process are of great commercial interest. The permeate of skim milk microfiltration is a much better feed for recovering and purifying serum proteins than cheese whey (Brans et al., 2004). Concentrated casein micelles in the retentate can be used to recombine with cream for cheese production, to standardize milk and to produce dried native casein for various food applications. The permeate from skim milk microfiltration contains whey proteins (also often referred to as serum proteins) that are in a native, functional state. The MF permeate is sterile and contains no rennet or starter culture, and is low in lipid content (,0.01%) (Britten and Pouliot, 1996). Hence, clarification and pasteurization of MF permeates is not required as is the case for cheese whey generated from cheese making processes, minimizing the risks of denaturing the whey proteins or impairing their functionality. The whey proteins can be further concentrated by UF to produce serum protein concentrate (SPC, 34%
80% total protein) or serum protein isolate (SPI, .80% total protein). There are only subtle differences between serum protein concentrate obtained from casein fractionation and whey protein concentrate (WPC) produced from cheese whey. The SPC is lower in fat and higher in pH. Evans et al. (2010) found that beverages made with SPC scored higher by a sensory panel for aroma, appearance, and mouthfeel compared with those made with WPC, but the flavor and overall liking scores were lower. Polymeric spiral wound membranes with pore diameter in the rage of 0.3
0.5 μm and tubular ceramic membranes with pore diameter of 0.1
0.2 μm have been used for the fractionation of casein micelles from skim milk. However, a number of research groups (Lawrence et al., 2008; Beckman et al., 2010; Zulewska and Barbano, 2013) have found that polymeric membranes retain significant levels of serum protein (e.g., β-lactoglobulin), due to the formation of a non-selective membrane fouling layer. This fouling layer is referred to as a “dynamic membrane” because of the role it plays in blocking serum protein transmission. A UTMP approach (see Section 8.3.2) is required to overcome this issue, which is best achieved using ceramic membranes (Karasu et al., 2010). While ceramic membranes are typically an order of magnitude more expensive than polymeric membranes, their lifespan is more than three times longer (10 vs 3 years) (Cheryan, 1998). Microfiltration of skim milk using ceramic membranes with a pore diameter of 0.1
0.2 μm can produce a permeate with a composition close to that of sweet whey (Fauguant et al., 1988) and a casein-rich retentate. It has been shown that to prevent the formation of an irreversible fouling cake that will limit serum transmission, the ratio of TMP to wall shear stress, which in turn relates to cross-flow velocity must be tightly controlled (Le Berre and Daufin, 1996; Ge´san-Guiziou et al., 1999). The operational characteristics of

Membrane Separations in the Dairy Industry Chapter | 8

277

the MF processes thus include a high cross-flow velocity ( . 7 m/s) and an operating temperature of 50
55 C to maximize flux, as discussed previously. Up to 90% casein concentration can be achieved when diafiltration (subsequent diluting and refiltration of the retentate) is employed (from 2.74 g/100 g initial skim milk to 17.8 g/100 g retentate after diafiltration (Saboyainsta and Maubois, 2000)), making it suitable for industrial, pharmaceutical, and food and beverage applications. Various commercial ceramic membranes, including a 0.22-μm Ceraflo membrane (Pouliot et al., 1996), a 0.2-μm Membralox membrane (Vadi and Rizvi, 2001), and a 0.05-μm asymmetric Ceramen membrane (Punidadas and Rizvi, 1998), have been tested for concentrating skim milk to a wide range of concentration factors (2
10). The flux in these processes was observed to be an order of magnitude lower than bacterial removal of skim milk using MF, ranging from 47 to 90 LMH (1.3
2.5 3 1025 m/s) under a TMP of 138 kPa (Punidadas and Rizvi, 1998) to 193 kPa (Vadi and Rizvi, 2001). Membralox GP ceramic membranes with a 0.1-μm pore diameter were used by Tremblay-Marchand et al. (2016) to assess the effects of various operating parameters on the process efficiency of serum protein removal from skim milk. The flux ranged from 120 to 80 LMH at 50 C for volume concentration factors (VCFs) from 2 to 3, which is in the same range reported for the commercial ceramic membranes above. As expected, the energy consumption of the process increased with increasing VCF, along with serum protein rejection and total casein loss. However, the variation of TMP from 124 to 207 kPa had no significant effect on retentate and permeate composition. Optimization of pore size and filtration temperature using ceramic membranes was recently investigated by Jørgensen et al. (2016) using a UTMP to achieve a VCF of 2.5 at constant permeate fluxes (44
59 LMH). A permeate free from casein could not be achieved with membranes of an average pore size of 0.2 μm, as 1.4% of the caseins permeated through the membranes. Optimal protein fractionation of skim milk into a retentate rich in casein and a permeate with native serum proteins and negligible caseins was achieved by membranes with an average pore size of 0.10 μm at 50 C. The fractionation of whole casein into individual caseins is of great commercial interest, and can potentially be achieved from the MF retentates from skim milk or whole milk. For example, β-casein can be separated based on its preferential solubilization at low temperature (Creamer et al., 1977). It contains numerous peptide sequences with beneficial physiological properties and is the main component of human casein (Ge´san-Guiziou, 2013). Lucey and Smith (2009) investigated a two-stage MF process using polymer membranes to separate β-casein directly from whole milk. The first MF step operated at low temperatures, separating the β-casein from the retentate that contained casein micelles and fat. The β-casein-rich permeate from the first MF step was then processed at room temperature to allow the β-casein to aggregate so it could be recovered in the retentate from a second MF step.

278

Separation of Functional Molecules in Food by Membrane Technology

8.3.4 Milk Fat Fractionation Membranes can also be used to separate cream from milk and fractionate fat globules. The separation of cream from whole milk is traditionally done by centrifugation, which is an energy intensive process. The size of fat globules is known to have a profound impact on the stability and the textural and sensory characteristics of dairy products (Timmen and Patton, 1988). Highpressure homogenization is therefore often used to reduce the size of the fat droplets, resulting in a homogenized milk that is more consistent and stable. Unlike homogenization, membrane processes are able to fractionate fat globules with different fatty acid composition according to their size without disrupting the milk fat globule membranes. The fraction containing smaller fat globules is claimed to yield improved products with a finer texture than those made from cream with larger globules (Goude´dranche et al., 2000). A patented ceramic microfiltration membrane process was developed by Goudedranche et al. (2000, 2003) for fractionating fat globules into two fractions (,2 μm and .2 μm). Microfiltration membranes with a large range of pore sizes (1.8
10 μm) were tested. With a ceramic membrane of 2 μm pore diameter, whole milk could be fractionated into retentate and permeate streams with fat contents of up to 197 and 17 g/kg, respectively, at a permeate flux of 700 LMH. Increasing the fat content of the feed stream from 39 (whole milk) to 120 g/kg, decreased permeate flux (to 250 LMH) but the fat content in the retentate and the permeate streams increased to 297 and 69 g/kg, respectively.

8.4 CHEESE PROCESSING Cheese is conventionally made from high quality raw or pasteurized milk by casein coagulation using rennet enzymes and/or lactic acid bacteria. The curd is then separated from the whey before further processing. For ripened cheese, the curd is molded, compressed, and stored under suitable conditions with well-tuned temperature and humidity for a defined period of time. For semihard and hard cheeses, salt must be added to draw water from the cheese and so lower its moisture content. Membrane processing has been used in cheese manufacture to improve the nutritive value, enhance compositional consistency, increase the yield of cheese, and reduce the amounts of rennet and starter culture required (Kumar et al., 2013). Using microfiltration to remove bacteria and spores from milk and fractionate casein micelles, as discussed in the previous sections, are two of the examples for milk processing prior to cheese making. Studies (St-Gelais et al., 1998; Caron et al., 1997; Heino et al., 2009; Maubois, 2002) have found that cheese milk pretreated by MF membranes with 0.1
0.2 μm pore size can shorten coagulation time, improve curd firmness, increase cheese yield, and accelerate ripening. The casein fraction can

Membrane Separations in the Dairy Industry Chapter | 8

279

also be used to produce semihard and hard cheeses without using cheese vats, removing the cheese cutting and washing steps (Thomann et al., 2008; Govindasamy-Lucey et al., 2007). The casein-enriched milk from microfiltration retentates contains a reduced amount of native whey protein. Using microfiltration retentate for cheese making reduces flavor and texture defects caused by whey proteins during cheese making, and eliminates any detrimental effects of heat treatment on rennet coagulability (Ge´san-Guiziou, 2013). Native whey protein is found to inhibit the enzymatic action of chymosin and present a physical barrier to para-casein aggregation (Gamlath et al., 2018), leading to a shorter gelation time. Microfiltration has great potential to add value to conventional cheese making processes, as evaluated in an economic study on the feasibility of MF in cheese making in Northern America in 2003 (Papadatos et al., 2003). However, industrial scale demonstration has not been reported in terms of cost effectiveness (TremblayMarchand et al., 2016).

8.4.1 Cheese Milk Standardization For a number of decades, UF has been commonly used in cheese production to adjust and control the protein content. This can improve the cheese yield, as well as provide better compositional and quality control, while requiring fewer processing steps (Guinee et al., 1995, 2006; Johnson and Lucey, 2006). UF retains all the milk proteins, while simultaneously removing lactose and minerals. UF can be used to concentrate the protein in cheese milk to a desired concentration. The UF permeate can also be used to reduce the protein concentration of the cheese milk. Through the standardization of protein, ultrafiltered milk is not affected by any inherent variations in the composition of raw milk caused by factors such as weather, seasons, stage of lactation, feeding, and breed (Rattray and Jelen, 1996). UF milk standardization offers opportunities not only for cheese making, but also for manufacturing yogurt, ice cream, and drinking milk because thermal evaporation is no longer required to standardize the protein content. This avoids protein denaturation, and any associated alternation in the functionality and nutritional value of the milk proteins (Rattray and Jelen, 1996; Johnson and Lucey, 2006). UF can concentrate milk by a factor of 1.2
6 (Mistry and Maubois, 2004; Rosenberg, 1995; Henning et al., 2006; van Leeuwen et al., 1990). Spiral wound polymeric membranes (typically made with polyethersulfone and polyvinylidene fluoride) and tubular ceramic membranes (less often), with a molecular weight cut-off (MWCO) of 10
50 kDa are used at a TMP of 2
4 bar. The flux typically ranges from 30 to 120 LMH (Ge´san-Guiziou, 2013). UF can be performed either at around 10 C or 50 C to minimize the growth of pathogenic bacteria. While the flux is greater at higher

280

Separation of Functional Molecules in Food by Membrane Technology

temperatures due to reduced fluid viscosity, the rate of flux decline is higher due to proteinaceous fouling (Ng et al., 2017). Low concentration factors (1.2
2) can increase the protein content by B5% for various types of cheese, including cheddar, brick, cottage cheese, Colby, mozzarella, Edam, Saint Paulin, and quarg. Medium concentration factors (2
6) can increase cheese yield by 6%
8% for cheddar, feta, Havarti, Gouda, and blue cheese, but special cheese making equipment is required. Higher volume reduction factor (VRF) (5
7) results in a UF concentrate that is also called “pre-cheese,” which has the same protein concentration as in the final cheese. The MMV process (Maubois et al., 1969), named after its inventors Maubois, Mocquot, and Vassal, is widely used to process pre-cheese for manufacturing fresh unripened cheeses (e.g., quarg, ricotta, and cream cheese) and soft and semihard cheeses (e.g., mozzarella, Saint Paulin, and feta). The overall cheese yield is about 10%
30% higher than the traditional processes, due to the effective rejection of whey proteins and subsequent reduction in enzyme usage. This process does not require the use of a cheese vat and minimizes the need for whey drainage, resulting in a great saving in capital investment and operating costs. However, some researchers have reported that UF milk causes compromised sensory and functional properties of semihard and hard cheese, due to impairment of ripening. The reduction in ripening rate relates to a reduction in proteolysis resulting from a decrease in chymosin action, and the presence of more whey proteins (Mistry and Maubois, 1993; Neocleous et al., 2002).

8.5 WHEY PROCESSING Whey is a byproduct generated during cheese, yogurt, and casein production. The coagulation of milk proteins in these products leaves behind this nutritional stream that is rich in proteins, lactose, and dairy minerals. While large dairy processors recover these nutritional components from whey, it is treated as a waste by smaller processors and in most developing countries, where further processing for valorization is not deemed economically viable. Membrane processing can be used to produce value-added products from whey that can be used to improve the functional properties of other food products.

8.5.1 Whey Protein Concentration In whey processing facilities, microfiltration can be used to pretreat sweet whey by removing bacteria and fat prior to UF. During UF, whey proteins are separated from the lactose-rich permeate stream, producing WPC. WPC containing 30%
80% protein is obtained using UF membranes with a MWCO of 10
20 kDa, by removing both the lactose and salts in the permeate stream. The UF retentate is further processed in evaporators and dryers to

281

Membrane Separations in the Dairy Industry Chapter | 8

TABLE 8.2 Composition of Ultrafiltration Retentate of Whey With Different Concentration Factors Concentration Factor

1

5

10

20

35

35a

20

Dry matter (%)

6.6

10

14

20

25

22

17

6.6

6.6

6.6

6.6

6.6

3.2

pH during ultrafiltration Protein/dry matter (%)

12

34

45

58

70

82

59

Lactose/dry matter (%)

74

51

39

27

17

7

27

“Ash”/dry matter (%)

8

6

5

4

3.5

2.5

2.7

Citrate/dry matter (%)

2.5

1.8

1.7

1.6

1.4

Fat/dry matter (%)

1

2

3

4

5

6

4

a Followed by diafiltration, that is, water is added to increase the volume by a factor of 3, the mixture is processed again by ultrafiltration. Source: Reproduced from Walstra, P., 1979. The voluminosity of bovine casein micelles and some of its implications. J. Dairy Res., 46 (2), 317
323.

produce WPC powder. The VRF is important in determining the protein content in the final WPC powder (see Table 8.2 below). A VRF of 5
10 is required for WPC35 powder, and a VCF .35 is needed to produce a WPC with a purity of 75%
85% using diafiltration. The presence of residual lipids originated from cheese making, the precipitation of calcium phosphate at high temperature, and the accumulation of whey proteins on the membrane surface are common causes of the fouling of UF membranes. One approach to mitigating this is to operate UF processes at a low temperature (,10 C), although the flux (,40 LMH at 3 bar TMP (Meyer and Kulozik, 2016)) is half of the flux obtained at B50 C. This has become a favorable practice in dairy processing because it also maintains the microbiological quality of the end product, by avoiding the growth of thermophilic microbes.

8.5.2 Whey Protein Fractionation Whey proteins such as IGs, α-lactalbumin, β-lactoglobulin, lactoferrin, and other minor proteins have useful functional and nutraceutical properties (Ge´san-Guiziou, 2013). α-Lactalbumin is high in the essential amino acid tryptophan and can be used for nutraceutical foods and in infant formula. β-Lactoglobulin can be used in emulsification, foaming, and gelling, as well as in the manufacture of protein hydrolysates for ingredient formulation. α-Lactalbumin and β-lactoglobulin are very similar in size (14 and 18 kDa, respectively), making their fractionation difficult in a single-step membrane filtration process. To enhance the yield and purity, the protein molecules can

282

Separation of Functional Molecules in Food by Membrane Technology

be manipulated before membrane filtration is applied (Slack et al., 1986). α-Lactalbumin loses its bound calcium and its stability when heated at 55 C for 30 minutes at pH 3.8. Under this condition, α-lactalbumin unfolds and precipitates as calcium dissolves in the solution (Bramaud et al., 1995). Because of this property, α-lactalbumin can be precipitated (together with BSA and IGs) from UF whey concentrate. This allows the recovery of β-lactoglobulin from the UF permeate using diafiltration processes employing UF or MF membranes (Pearce, 1983; Maubois et al., 1987). More recent studies have focused on optimizing operational modes (e.g., batch, continuous, and diafiltration) (Muller et al., 1999), operating parameters (e.g., VRR (Espina et al., 2009) and TMP (Marella et al., 2011)), and membrane molecular weight cut-off (MWCO) (Marella et al., 2011) to achieve better separation of these two proteins using polyvinylidene fluoride (PVDF) and polyethersulfone (PES UF) membranes. However, the purity of α-lactalbumin achieved among these studies ranges from 0.4 to 0.6, indicating that membrane selectivity of these proteins is the limiting factor. To improve the selectivity, modification of PES membranes has been attempted (Cowan and Ritchie, 2007). The membranes were treated to create an open pore structure with charged sulfonated grafted polymer chains to enhance the electrostatic repulsion between the charged membrane and the proteins. The selectivity was reported to be improved five-fold compared with the untreated membrane. IGs are comparatively large (Table 8.1) and can be separated from whey using UF membranes. However, cow colostrum has a significantly higher concentration of IGs (20
200 g/L) than milk or whey (0.15
0.8 and ,0.1 g/L, respectively). To recover IGs from colostrum, microfiltration (0.1 μm) is firstly used to remove blood, somatic cells, fat globules, and casein micelles, producing a permeate containing 80% of the initial IGs. The permeate is further processed using UF membranes with MWCO of 100 kDa (Cowan and Ritchie, 2007).

8.5.3 Whey Concentration RO is often used to pre-concentrate milk and whey streams prior to evaporation and drying, to increase the total solids concentration. Its energy consumption is significantly lower than vacuum evaporation (9
20 vs 100
150 kWh/m3 water removed (Daufin et al., 2001; Ge´san-Guiziou, 2013)). Milk concentration using RO is rarely found in the dairy industry because of the low flux (,15 LMH at 10 C (Meyer and Kulozik, 2016)), the low concentration factor that can be achieved, and severe membrane fouling (Glover, 1971). Whey concentration using RO can be achieved only up to a concentration factor of B4, from 5%
6% TS to B20% TS, due to the limited hydraulic pressure (40
60 bar) that can be applied to the system. The permeate flux during RO of whey is twice as high as for skim milk at 10 C.

Membrane Separations in the Dairy Industry Chapter | 8

283

At 20% TS concentration, the permeate flux is diminished because of the high osmotic pressure and viscosity of the concentrated whey, as well as the crystallization of lactose and precipitation of calcium phosphate on the membrane surface. Membrane distillation is an alternative for concentrating dairy streams. Hydrophobic polytetrafluoroethylene (PTFE) membranes are mostly used for this emerging technology. A few applications in dairy processing have been investigated, including the concentration of whey protein (Christensen et al., 2006), milk, whey, and lactose (Hausmann et al., 2011, 2014). At 20% TS concentration, a flux of B12 and B20 LMH has been achieved for membrane distillation of skim milk and whey, respectively. Forward osmosis has not been investigated for any potential dairy applications until recently. HydrOxSys, established in 2012, has been commercializing a proprietary thin film composite membrane suitable for forward osmosis systems that is intended to concentrate milk at or near milk farms before it is transported to dairy processing facilities (Adams, 2013). Integrated (Phuntsho et al., 2012) systems incorporating forward osmosis and RO were studied for water recovery and whey powder production from whey by Aydiner et al. (2014). The typical flux range of FO membranes was 10
25 LMH using an NaCl draw solution at low whey concentrations (Aydiner et al., 2014, 2013). It was found that a FO/RO system using a 2M NaCl draw solution could be successfully applied, with the total solids content increasing from 6.75% TS (unprocessed whey) to 28% TS. Membrane distillation can be coupled with forward osmosis to recover water from the diluted draw solution. In 2011, Wang et al. (2011) demonstrated this technology for the concentration of protein solutions using concentrated NaCl as the draw solution Ge et al. (2012) similarly used this approach for the treatment of dye wastewater using a poly(acrylic acid) sodium (PAA-Na) salt as the draw solution. They concluded that the FO/MD hybrid system was stable, repeatable, controllable, and predictable, and performed better than an individual FO process.

8.5.4 Whey Demineralization Demineralization of sweet whey is necessary in some instances to meet the ingredient requirements in human food applications and to reduce the mineral levels for downstream processing. The production of ice cream and bakery products requires partially demineralized whey concentrate, whereas highly demineralized whey concentrate/powder is used for making infant formula. Depending on the degree of desalination required, NF, electrodialysis, or ion exchange can be employed. NF membranes with a MWCO between 200 and 1000 Da are used in the dairy industry to desalinate dairy process streams such as skim milk, whey, and UF permeates. Proteins and lactose are retained by these membranes.

284

Separation of Functional Molecules in Food by Membrane Technology

Lactose is an important ingredient in infant formula and in food processing. It can also be used as a filler or coating agent in the pharmaceutical industry. During the concentration of whey protein, the permeate from the UF process is a salty lactose solution containing B4%
8% lactose and B0.5% salt. To recover lactose, the UF permeate is concentrated to 60% solids in multiple effect evaporators. Lactose is then crystallized from the concentrate, separated, and dried into lactose powder. To enhance the yield and purity, the UF permeate is also often processed by NF to partially remove the salts and to increase the lactose concentration. Partial demineralization of sweet whey can be achieved by the partial removal of monovalent ions using NF to produce D40 whey (40% of whey demineralization). Whey is concentrated to 20%
22% at a VRF of B4. Although NF is the most economical technology for whey demineralization, a higher degree of demineralization requires electrodialysis or ion exchange, due to the insufficient removal of monovalent ions (typically ,70%) by NF. Electrodialysis is found to be more cost-effective than ion exchange for demineralization to levels below 70% (Bylund and Hellman, 2015). Preconcentration of whey (to 18%
24%) by RO or evaporation can enhance the efficiency of electrodialysis. Acid whey is a special type of whey resulting from Greek yogurt and cream cheese manufacturing. If acid whey is processed to produce powder in the presence of lactic acid, the resulting powder is highly susceptible to moisture absorption, due to the hygroscopic nature of the lactate ions. This leads to the formation of powder agglomerates and sticky deposits within the dryer that cannot be tolerated in normal operation. NF was recently demonstrated at both laboratory (Chandrapala et al., 2016a,b) and pilot scale (Be´das et al., 2017), for removing lactic acid from acid whey, to enable the recovery of whey proteins and lactose in a downstream drying unit. Effective removal of lactic acid from acid whey using electrodialysis was also demonstrated by Chen et al. (2016), resulting in a demineralized whey stream that can be further processed before drying. The energy consumption for removing 80% of the minerals from acid whey or sweet whey by electrodialysis was B13 Wh/kg whey processed.

8.5.5 Cheese Brine Purification Salting/brining is an important step for manufacturing semihard and hard cheeses. It involves the extraction of water from the cheese curd using the high osmotic pressure driving force offered by either dry salt or a cheese brine. The cheese brine is replenished on a regular basis, but this is dependent on maintaining the brine quality. Cheese brine may contain microorganisms such as gas-producing lactobacilli, pathogenic bacteria (e.g., Staphylococci and Listeria), yeast, and molds (Pedersen, 1992). Heat treatment can be performed to sanitize cheese brine; however, this alters the

Membrane Separations in the Dairy Industry Chapter | 8

285

calcium phosphate equilibrium in the brine and requires expensive plate heat exchangers to avoid corrosion (Saboyainsta and Maubois, 2000). Hence, microfiltration is an effective way to sanitize and maintain cheese brine. Microfiltration can remove bacteria and spores, and maintain the chemical balance between NaCl (18
26%wt), calcium salts, lactose, lactic acid, and soluble and denaturated whey proteins (Ottosen and Kønigsfeldt, 1999). Ceramic membranes with 0.8 or 1.4 μm pore size with a UTMP system can be used to reject all yeasts and molds, 99.9% of the bacteria, and only 7% of the calcium salts and 2%
3% of the nitrogen matter, with a flux of 600 LMH at 20 C (Pedersen, 1992). Since microfiltration cannot separate water and salt, salts need to be added to the brine to maintain the required salinity. Recently, the application of membrane distillation to regenerate cheese brine was demonstrated by Eykens et al. (2018) at both laboratory and pilot scale. Microfiltration membranes (0.1 μm pore size PVDF) and UF membranes (50 kDa MWCO) were used to pretreat cheese brine by removing suspended solids. Comparable flux (B350 LMH) was obtained at a TMP of B0.9 bar. The permeate was then processed by direct contact membrane distillation (DCMD) to remove water from the brine to achieve a salt concentration of up to 220 g/L. The flux for DCMD ranges from 4 to 6 LMH for the polyethylene membranes (Lydall, USA) used for the lab scale tests, compared to around 0.8 LMH for a standard air gap membrane distillation (AGMD) module (Aquastill, The Netherlands). Protein and calcium phosphate fouling was observed, but the treatment cost was estimated to be at least 35 euro/m3 feed water for the MD process without using waste heat. This could be reduced to 12 euro/m3 feed water if waste heat was utilized.

8.6 WASTE TREATMENT Dairy processing produces 0.2
11 L of effluent per liter of processed milk with a typical biological oxygen demand of 0.2
2.5 g/L (OECD, 2012). These waste streams include milk spills, cleaning chemicals, product rejects, biological treatment sludge, and plant wastewater (Wilkinson, 2007; Watkins and Nash, 2010). Dairy wastewater is typically treated onsite in biological treatment plants to lower the organic load and associated trade waste charges. A number of these waste streams can benefit from the application of membrane technology. Dairy processors rely heavily upon clean-in-place (CIP) to meet the stringent hygiene requirements. CIP is a method of cleaning and sanitizing the interior surfaces of pipes, vessels, equipment, filters, and fittings without taking them apart. CIP operations consume a considerable amount of process water and are responsible for 50%
95% of the overall volume of the waste streams (Durham and Hourigan, 2009). A combination of membrane technologies such as microfiltration, UF, and NF have proven effective for

286

Separation of Functional Molecules in Food by Membrane Technology

regenerating cleaning solutions and reducing water, chemical, and energy consumption (Daufin et al., 2001; Eide et al., 2003; Merin et al., 2002; Ra¨sa¨nen et al., 2002). The pre-rinsing water from the first step of CIP, for example, is often filtered using UF, NF, or RO to recover the milk components for re-processing or animal feed production, simultaneously reducing the organic loads in the plant effluent (Delbeke, 1981; Blanchard, 1991). NF permeate of milk, whey and evaporator condensates can be treated by RO to recover “food grade” water that can be used for rinsing and cleaning (Horton, 1997; Yorgun et al., 2008; Vourch et al., 2008). When microfiltration is used to filter spent CIP caustic soda solution, the small molecules responsible for the low surface tension are not retained (Vourch et al., 2008). The presence of these molecules in the recycled solution (i.e., the MF permeate) has been found to accelerate the cleaning process, making the regenerated caustic solution as effective as commercial alkaline detergents (Alvarez et al., 2007). NF membranes can also be efficient at polishing spent caustic CIP solutions, with around 98% chemical oxygen demand (COD) rejection and up to 95% recovery of the acid or caustic. The clean permeate stream is then ready to be reused in subsequent cleaning cycles. The SelRO Caustic Recovery System using SelRO MPS-34 NF membranes was designed by KOCH Membrane Systems to remove more than 90% of the brown, burnt-colored contaminants and COD from spent caustic or acid, accompanied with substantial reduction of calcium and carbonates (i.e., ineffective alkalinity) (KOCH Membrane Systems, 2017). Membrane technology can also be effective at treating dairy wastewater with high salinity. Salty streams originating from dairy processing operations include chromatography wastes, CIP wastewater, acid whey, and waste generated from whey demineralization processes including NF, electrodialysis and ion exchange. Currently, the high salinity waste streams are commonly disposed of in evaporation ponds. However, environmental issues of land degradation, odor, and dust have prevented the construction of further evaporation ponds in some areas of Australia. As a result, there is great interest in controlling and monitoring the load of salinity in dairy trade waste, to avoid penalties imposed by the local water authorities. As already mentioned, salt (sodium chloride) is added to protein-rich cheese curds when making semihard or hard cheese (e.g., cheddar and Colby). A significant amount (50%
65%) of the added salt forms a brine stream called salty whey. A process called Ultra-Osmosis (Heldman et al., 2006) was reported for selectively removing the dissolved salts from salty whey, achieving a retentate stream close to the normal sweet whey mineral composition. The retentate can then be fed into the ordinary whey processing operations. Recently, direct membrane distillation was successfully demonstrated for concentrating salty whey using flat sheet PTFE membranes, achieving a final total dissolved solids (TDS) concentration of B30%w/w, with up to B83% of the water recovered (Kezia et al., 2015). This can

Membrane Separations in the Dairy Industry Chapter | 8

287

reduce the volume of waste brine, providing the opportunity to recover the salt via crystallization processes.

8.7 FOULING Fouling is a limiting factor in membrane filtration processes. It must be minimized due to its effect of reducing permeate flux and membrane selectivity, and the cost of implementing cleaning cycles required to restore productivity. Fouling takes place when particles deposit on the membrane surface or in the pores of membranes. Fouling mechanisms include adsorption, pore blockage, precipitation, and cake formation. In particular, porous membranes, such as microfiltration and UF membranes, are highly susceptible to fouling. During membrane filtration, the flux increases linearly with increasing TMP until it passes a critical flux (Fig. 8.3), which is an indication of the onset of irreversible fouling (Field et al., 1995; Rice, 2008). As mentioned in Section 8.3.3, above this flux the protein fouling layer can totally control membrane selectivity, by acting as a dynamic or secondary membrane. It is desirable to operate below this critical TMP, in the pressure-controlled (linear) region due to a reduced cleaning demand. However, a large membrane area is needed due to the low capacity and low flux. A higher cross-flow velocity or higher operating temperature can enhance the capacity. As pressure increases beyond the critical pressure, the flux increases at a slower rate with respect to TMP, and reaches a maximum value known as the limiting flux. This is caused by the build-up of concentration of the rejected solutes near the membrane surface due to concentration polarization. The back-diffusion model is used to describe this limiting flux (Mulder, 1996), where the mass transfer of solute from the bulk solution to the membrane is in equilibrium with the diffusive back-flow of solute from the

FIGURE 8.3 Pressure and mass transfer controlled regions for cross-flow membrane filtration. TMP, transmembrane pressure. Reprinted from Cheryan, M., 1998. Ultrafiltration and Microfiltration Handbook. CRC Press, Copyright 1998, with permission from Taylor & Francis USA.

288

Separation of Functional Molecules in Food by Membrane Technology

membrane to the bulk solution. If the concentration of solute at the membrane surface reaches the solubility limit, the solute precipitates out, forming a gel layer, in which case the gel polarization model may be used. Due to the presence of the concentration polarization layer, the flux is independent of the TMP and the pore size of the membrane. The main foulants in dairy processing are proteins and calcium phosphate (Ng et al., 2017, 2018). Uncharged solutes such as lactose are not found to cause membrane fouling (Nystro¨m et al., 1995). Lipids have no impact on membrane fouling unless the concentration is high in the feed solution (Marshall and Daufin, 1995). Protein adsorption onto the membrane surface is controlled by the hydrophobicity of the membrane surface and the electrostatic interaction between the charged proteins and the membrane (Marshall and Daufin, 1995). This interaction is responsible for the formation of the initial fouling layer. However, protein
protein interactions determine the fouling behavior once the membrane surface is no longer available to interact with the proteins deposited on top of the first fouling layer. The presence of calcium in dairy streams is an important factor in protein cake formation, because of its ability to form protein
membrane and protein
protein bridges. Precipitation of salts and scaling on the membrane surface can occur if the soluble salts in the feed have reached the solubility limit in the polarization layer adjacent to the membrane surface. Examples include NF of whey UF permeate (Rice et al., 2006) and membrane distillation of salty whey (Kezia et al., 2015) where polyvalent salts are concentrated as water is removed. Calcium phosphate is the dominant salt foulant in dairy systems and displays reverse solubility (i.e., solubility decrease with increasing temperature (Rice, 2008)). An increase in pH also leads to a greater calcium phosphate precipitation. Hence, operating membrane systems at low temperatures and pH in the dairy industry is favorable for reducing calcium phosphate scaling. The precipitation of calcium phosphate is less problematic for milk streams containing casein micelles that absorb calcium phosphate as colloidal calcium phosphate (CCP), preventing precipitation on the membrane (Ng et al., 2018). Different methods have been developed to reduce membrane fouling during long-term operation and to minimize the number of cleaning cycles. Membrane cleaning is a costly way to restore membrane performance, as it interrupts the production, requires use of chemical cleaning agents, and accelerates the need for membrane replacement. For example, high cross-flow velocity is always used during microfiltration of milk to minimize the formation of a casein micelle gel layer by increasing wall shear stress (Cheryan and Alvarez, 1995). UTMP and GP microfiltration membranes, as discussed in the relevant previous sections, are examples of strategies for fouling control that have been implemented at an industrial scale.

Membrane Separations in the Dairy Industry Chapter | 8

289

Other means of increasing shear force close to the membrane surface to control fouling include the use of vibrating modules (Al-Akoum et al., 2002), rotating disks (Engler and Wiesner, 2000; Ding et al., 2002), or air slugs (Cui and Wright, 1996; Cui and Taha, 2003). Back-pulsing (Redkar et al., 1996; Mores and Davis, 2003) can be used in hollow fiber membranes to remove the fouling cake by reversing the TMP. Electric fields (Visvanathan and Aim, 1989; Wakeman, 1998) have been used to keep charged particles away from the membrane surface. To enhance back transport of the deposits, turbulence promoters (Winzeler and Belfort, 1993; Krsti´c et al., 2002) and pulsating/ reversed cross-flow (Jaffrin et al., 1994; Curcio et al., 2002) can be used, respectively, by generating microturbulence near the membrane surface, and rapidly changing the velocity in the flow channel. Another approach to promoting back-transport of the foulants is the application of ultrasonic waves (Wakeman and Tarleton, 1991; Shu et al., 2007) that create vibrations and shear from cavitation, as discussed in the next section.

8.8 APPLICATION OF SONICATION The significant downtime required for cleaning, and the use of potentially large amounts of chemicals, reduce the overall economics of membrane processing. In the past few decades, researchers have reported the use of ultrasonic waves to control membrane fouling in situ, enhance cleaning, and reduce required usage of chemicals.

8.8.1 Physical and Chemical Effects of Ultrasound Ultrasound (US) refers to oscillating pressure waves that vibrate at a frequency above the human hearing threshold. Typically, the frequency of ultrasound suitable for fluid processing applications ranges between 20 and 3000 kHz. The frequency of ultrasound governs the effects of the ultrasound applied to the system, and can be broadly divided into two main regimes: low-frequency highpower (LFHP) ultrasound and high-frequency low-power (HFLP) ultrasound. LFHP ultrasound typically refers to ultrasound with a characteristic frequency between 20
100 kHz. At these frequencies, the oscillating pressure wave induces the formation, growth, and high-intensity collapse of bubble nuclei within the fluids. The promotion of bubble collapse events, known as acoustic cavitation, is accompanied by the release of large amounts of energy in the form of heat (localized temperatures can increase up to B5000K) and physical shear. The shear forces may manifest in the form of pressure shockwaves, fluid microstreaming, and microjets. The temperature evolved during bubble collapse is sufficient to induce the formation of highly reactive radical species, particularly at higher frequencies between 300 and 500 kHz (Ashokkumar et al., 2008). In an aqueous medium such as milk, hydrogen and hydroxide radicals are produced

290

Separation of Functional Molecules in Food by Membrane Technology

(Weissler, 1959), which are responsible for ultrasonically enhanced chemical reactions, that is, sonochemistry. The use of these radicals to enhance oxidation reactions, for example, has been reported to be a highly complementary technique in membrane-based wastewater treatment (Naddeo et al., 2015). As the frequency of oscillation increases above 500 kHz, the collapse intensity of the bubbles declines correspondingly. At these frequencies, the resonance size that a bubble grows to before its collapse is reduced and so less physical energy is evolved from their collapse (Mason et al., 2011). At these frequencies there tends to be a larger population of sonochemically active bubbles, and so sonochemical efficiency is often optimal in this frequency range.

8.8.2 Ultrasonic Reduction of Fouling Buildup One of the problems of membrane fouling is the decline in flux as the foulant is deposited on the membrane. Ultrasound has been shown to be a useful method by which fouling build-up can be slowed down to improve the effective filtration time before more extensive cleaning of the membrane module is required. The effectiveness is mostly due to the physical effects of ultrasound, and as such, the use of lower frequency ultrasound tends to result in better performance (Alventosa-deLara et al., 2014). When used in combination with chemicals, ultrasonic membrane cleaning often results in a synergistic improvement. For example, in the cleaning of dye particles from an UF membrane, the use of NaOH resulted in a tripling of the cleaning efficiency (Alventosa-deLara et al., 2014). Ultrasound enhances the mass transfer of chemicals onto the surface of the membrane, and the shockwaves produced by collapsing bubbles near the surface of membranes assist in dislodging material from the membrane, which enables more effective attack of the foulant by the cleaning product. As the efficacy of the chemicals is enhanced, the amount required for cleaning can be reduced. One consideration that needs to be carefully made when using ultrasound to improve membrane performance, particularly in cross-flow membranes, is the selection of the TMP. An increase in the pressure has two effects on the acoustic cavitation generated by the applied ultrasound. As the pressure of the system increases, so too does the threshold for cavitation, which consequently reduces the total number of cavitation bubbles. However, at higher pressures, the effects of cavitation collapse become more intense, which in turn enhances the detachment of foulant from the membrane due to increased turbulence. As such, there tends to be some optimal pressure whereby there is a balance between the number of cavitation bubbles generated and the intensity of the cavitation itself. If this critical pressure is exceeded, cavitation can in some cases become suppressed and no cavitation activity occurs (Leong et al., 2009).

Membrane Separations in the Dairy Industry Chapter | 8

291

8.8.3 Mechanisms of Ultrasonic Membrane Cleaning in Dairy Applications Muthukumaran conducted a series of studies (Muthukumaran et al., 2005a,b, 2007, 2004) to understand the mechanisms behind the ultrasonic enhancement of cleaning efficiency for dairy fouled (e.g., whey protein) UF membranes. In general, ultrasound can provide vigorous mixing both at a macroscopic and microscopic level. At the macroscopic level, ultrasound can induce strong convective currents known as acoustic streaming, which bring about increased fluid motion and turbulent mixing. At a microscopic level, there are physical effects from the collapse of cavitation bubbles that generate micromixing, pressure shockwaves, and microjetting. With regard to whey protein fouling, four main effects were attributed to the ultrasonic enhancement of membrane filtration (Muthukumaran et al., 2005a). Firstly, sonication can promote the agglomeration of fine particles, which will reduce pore blockage and cake compaction. Secondly, sonication supplies mechanical vibrational energy to the system that keeps particles suspended. Thirdly, cavitation bubbles will scour the membrane surface with the advantage of being able to reach areas not accessible to conventional cleaning methods. Finally, acoustic streaming causes turbulence and more intense mixing, which will cause bulk fluid movement toward and away from the membrane cake layer. When cleaning membranes with detergents, Muthukumaran (Muthukumaran et al., 2005b, 2004) found that the use of ultrasound in combination with a surfactant resulted in a greater improvement than when either was used individually. The reason for this synergistic effect is that ultrasound can improve the effectiveness of cleaning agents through the enhancement of turbulence and mass transfer, while the surfactants can increase the cavitation activity (Ashokkumar et al., 1997; Lee et al., 2005a,b; Leong et al., 2011). The applied power level is critical from an energy efficiency standpoint. Excessive energy use would outweigh the benefit of improved flux or cleaning efficiency and can readily lead to membrane damage and a decrease in membrane life. Muthukumaran et al. (2005a) showed that the use of relatively low power levels (B2 W/L) could enhance the permeate flux during whey protein concentration, with an enhancement factor of between 1.2 and 1.7. Further, it was postulated that the use of intermittently applied ultrasound, that is, pulsed ultrasound, could reduce the required energy. However, this was shown to offer little flux improvement over long operations, with continuous ultrasound observed to be more effective in enhancing membrane flux (Muthukumaran et al., 2007). The application of continuous ultrasound at B2 W/L for a relatively short duration of 10 minutes during a cleaning cycle was sufficient to enhance membrane cleaning efficiency, with little gained by sonicating for longer durations.

292

Separation of Functional Molecules in Food by Membrane Technology

As the effect is primarily due to physical phenomena, it is not surprising that lower frequency ultrasound (i.e., 20 kHz) is more effective. At higher frequencies, the intensity of cavitation decreases (Mason et al., 2011) while the attenuation of the acoustic energy also increases, so there is less effective delivery of ultrasound into the membrane. Muthukumaran et al. (2007) showed that operation at 100 kHz resulted in no ultrasonic flux enhancement, and that operation at 1 MHz ultrasound caused the long-term flux to actually fall below the level provided in the absence of ultrasound. As mentioned, the TMP that is applied plays a role in the acoustic cavitation intensity and hence cleaning efficiency. Muthukumaran et al. (2005b) found that the overall effect of TMP was not particularly strong, which indicates that the effect of ultrasound enhancement is predominantly acoustic streaming and increased turbulence rather than cavitation.

8.8.4 Application of Ultrasound to Membrane Cleaning 8.8.4.1 Preventing Membrane Damage From Ultrasound Due to the intense physical forces generated during acoustic cavitation, the application of intense ultrasound may damage the membrane. Masselin et al. (2001) reported that PES membranes could be damaged after as little as 5 minutes exposure to ultrasound (47 kHz). However, this damage could be mitigated if the membrane was operated under cross-flow operation (Muthukumaran et al., 2004). Alternatively, the use of reduced power as reported by Wang et al. (2016) of 1 W/cm2 can be used to avoid membrane damage during ultrasonic application. 8.8.4.2 Membrane Modules Submerged in an Ultrasonic Bath One method to deliver ultrasound to a membrane is to submerge the entire module into an ultrasonic bath (e.g., Fig. 8.4 (Muthukumaran et al., 2005b)). While straightforward, the main downside of this application method is that a large portion of the acoustic energy can be absorbed by the module surrounding the membrane with only a small portion of the applied ultrasound energy reaching the membrane surface. Reducing the thickness of the membrane housing can improve ultrasound transmission, as reported by Michaud et al. (2015). 8.8.4.3 In Situ Ultrasound The need for ultrasound waves to travel from the transducer, through the membrane module to the actual membrane surface in a submerged system, results in a large loss of acoustic energy. Hengl et al. (2014) reported a new method to apply the ultrasound in situ for a cross-flow UF system, relevant to dairy applications. In their system, ultrasound was coupled directly to the membrane surface using a vibrating blade that oscillated at

Membrane Separations in the Dairy Industry Chapter | 8

293

FIGURE 8.4 An experimental setup whereby a membrane module is submerged in an ultrasonic bath. Reprinted from Lee, J., S. Kentish, and M. Ashokkumar, 2005b. Effect of surfactants on the rate of growth of an air bubble by rectified diffusion. J. Phys. Chem. B 109 (30), 14595
14598 Copyright 2005, with permission from Elsevier.

20 kHz (Fig. 8.5). This blade, located in the feed channel across the surface of the membrane, was able to effectively reduce mass transfer resistance induced by fouling and concentration polarization during operation so that less frequent cleaning downtime was required. This method of application was almost five times more effective at enhancing the permeate flux compared with when the module was submerged in an ultrasonic bath (Muthukumaran et al., 2005a).

8.8.4.4 Ultrasound Applied to Recirculated Cleaning Solution or as Pretreatment In most laboratory studies, ultrasound is applied directly to the module or via submergence in an ultrasound bath. In contrast to this approach, LujanFacundo et al. (2013, 2016) reported the effective use of ultrasound applied to a cleaning solution that was externally recirculated through the membrane. By applying an ultrasound frequency of 20 kHz to the cleaning solution, an increase in flux recovery of approximately 10% was achieved compared with the absence of ultrasound. The likely mechanism for the ultrasonic enhancement when treating the recirculated cleaning solution is that foulant particles removed from the membrane and into the cleaning solution are processed in such a way that they are no longer able to foul the membrane upon recirculation, for example, via size reduction or size aggregation. As higher frequency ultrasound is not effective at breaking apart aggregates, improved performance was primarily achieved using low frequency ultrasound (Luja´nFacundo et al., 2013).

294

Separation of Functional Molecules in Food by Membrane Technology

FIGURE 8.5 Schematic diagram of in situ application of ultrasound for membrane processing. Reprinted from Hengl, N., et al., 2014. A new way to apply ultrasound in cross-flow ultrafiltration: application to colloidal suspensions. Ultrason. Sonochem. 21 (3), 1018
1025. Copyright 2014, with permission from Elsevier.

Apart from using ultrasound for cleaning, studies have also considered the use of ultrasound as a pretreatment to reduce membrane fouling during operation. This can be particularly useful, for example, in the pretreatment of dairy protein streams, for example, whey protein. Koh et al. (2014) used ultrasound to improve downstream UF performance. Although sonication by itself reduced membrane fouling slightly, it was found that the use of US following a heating step significantly reduced membrane pore blockage and fouling cake formation of whey protein streams. This is a highly relevant outcome for the dairy industry as heat treatment is often applied to dairy whey protein streams during concentration of the protein to produce powders

Membrane Separations in the Dairy Industry Chapter | 8

295

with 35%
80% protein content. When heated above 70 C, whey protein denatures and aggregates, increasing the viscosity, which can limit the solids concentration that can be viably achieved using membrane concentration upstream of spray drying. A combination of heat and ultrasound can effectively control and mitigate heat instability of whey protein (Ashokkumar et al., 2009; Chandrapala et al., 2012). Application of ultrasound can break apart any aggregates formed by heating by disrupting the hydrophobic interactions through the shear forces from acoustic cavitation. The key outcome is that once formed and broken apart, the disrupted whey protein aggregates are no longer able to reform as easily upon further heating. This effectively delays the “gelling” that would normally occur with the protein, enabling more effective control of the viscosity of the protein stream with heating.

8.9 CONCLUSIONS AND FUTURE TRENDS Membrane technology provides the dairy industry with robust, reliable, and safe processes. They provide alternatives to conventional evaporative operations, and enhance productivity and product quality. Using a variety of membrane processes, the natural components in milk can be fractionated, separated, and purified to produce value-added products that have great commercial potential. Membrane technology can be used to create highly functional ingredients and nutraceutical products with desired characteristics, as well as to recover value added components from byproducts and plant effluents. Membrane performance in dairy processing is limited by the flux decline caused by fouling. This results in high water and chemical consumption for frequent CIP cycles, and high energy demand for producing the cross-flow velocities required for fouling mitigation. Continuing efforts are being made to develop inexpensive membrane materials and to improve membrane performance through improved membrane and module design and the application of ultrasound. Membrane integration has been seen in large dairy processing facilities for cheese milk standardization and whey processing, but is expected to be further advanced in the future. The ability of membrane processes to provide commercial opportunities as well as to mitigate environmental impact in the dairy industry will further promote innovative applications.

REFERENCES Adams, C., 2013. “Magic” to revolutionise dairy industry. The New Zealand Herald, 04 July [cited2014 02 September], Available from: ,http://www.nzherald.co.nz/business/news/article.cfm?c_id 5 3&objectid 5 10894642.. Al-Akoum, O., et al., 2002. Casein micelles separation from skimmed milk using a VSEP dynamic filtration module. Desalination 144 (1), 325
330.

296

Separation of Functional Molecules in Food by Membrane Technology

Alvarez, N., Ge´san-Guiziou, G., Daufin, G., 2007. The role of surface tension of re-used caustic soda on the cleaning efficiency in dairy plants. Int. Dairy J. 17 (4), 403
411. Alventosa-deLara, E., et al., 2014. Study and optimization of the ultrasound-enhanced cleaning of an ultrafiltration ceramic membrane through a combined experimental
statistical approach. Ultrason. Sonochem. 21 (3), 1222
1234. Anema, S.G., Li, Y., 2003. Association of denatured whey proteins with casein micelles in heated reconstituted skim milk and its effect on casein micelle size. J. Dairy. Res. 70 (1), 73
83. Ashokkumar, M., et al., 1997. Sonoluminescence from aqueous alcohol and surfactant solutions. J. Phys. Chem. B 101 (50), 10845
10850. Ashokkumar, M., et al., 2008. Modification of food ingredients by ultrasound to improve functionality: a preliminary study on a model system. Innovat. Food Sci. Emerg. Technol. 9 (2), 155
160. Ashokkumar, M., et al., 2009. Hot topic: sonication increases the heat stability of whey proteins. J. Dairy Sci. 92 (11), 5353
5356. Aydiner, C., et al., 2013. A novel implementation of water recovery from whey: “forward
reverse osmosis” integrated membrane system. Desalin. Water Treat. 51 (4
6), 786
799. Aydiner, C., et al., 2014. Techno-economic investigation of water recovery and whey powder production from whey using UF/RO and FO/RO integrated membrane systems. Desalin. Water Treat. 52 (1
3), 123
133. Baruah, G.L., Nayak, A., Belfort, G., 2006. Scale-up from laboratory microfiltration to a ceramic pilot plant: design and performance. J. Memb. Sci. 274 (1), 56
63. Beckman, S.L., et al., 2010. Production efficiency of micellar casein concentrate using polymeric spiral-wound microfiltration membranes1. J. Dairy Sci. 93 (10), 4506
4517. Be´das, M., et al., 2017. Nanofiltration of lactic acid whey prior to spray drying: scaling up to a semi-industrial scale. LWT - Food Sci. Technol. 79 (Suppl. C), 355
360. Blanchard, B., 1991. Plant effluents dairy waste streams recovery. Dairy, food and environmental sanitation (USA). Bouchoux, A., et al., 2010. How to squeeze a sponge: casein micelles under osmotic stress, a SAXS study. Biophys. J. 99, 3754
3762. Bramaud, C., Aimar, P., Daufin, G., 1995. Thermal isoelectric precipitation of α-lactalbumin from a whey protein concentrate: influence of protein
calcium complexation. Biotechnol. Bioeng. 47 (2), 121
130. Brans, G., et al., 2004. Membrane fractionation of milk: state of the art and challenges. J. Memb. Sci. 243 (1), 263
272. Brito-de la Fuente, E., et al., 2010. Microfiltration of whole milk with silicon microsieves: effect of process variables. Chem. Eng. Res. Design 88 (5), 653
660. Britten, M., Pouliot, Y., 1996. Characterization of whey protein isolate obtained from milk microfiltration permeate. Lait 76 (3), 255
265. Bylund, G., Hellman, M., 2015. Dairy Processing Handbook. Tetra Pak Processing Systems, Lund, Sweden. Caron, A., St-Gelais, D., Pouliot, Y., 1997. Coagulation of milk enriched with ultrafiltered or diafiltered microfiltered milk retentate powders. Int. Dairy J. 7 (6), 445
451. Carpintero-Tepole, V., et al., 2014. Microfiltration of concentrated milk protein dispersions: the role of pH and minerals on the performance of silicon nitride microsieves. LWT - Food Sci. Technol. 59 (2, Part1), 827
833.

Membrane Separations in the Dairy Industry Chapter | 8

297

Chandrapala, J., et al. A possible mechanism to understand the ultrasound induced heat stability of whey protein concentrates. In: Int. Non. thermal Workshop, Melbourne. 2012. Chandrapala, J., et al., 2016a. Nanofiltration and nanodiafiltration of acid whey as a function of pH and temperature. Sep. Purif. Technol. 160 (Suppl. C), 18
27. Chandrapala, J., et al., 2016b. Removal of lactate from acid whey using nanofiltration. J. Food Eng. 177, 59
64. Chen, G.Q., et al., 2016. Removal of lactic acid from acid whey using electrodialysis. Sep. Purif. Technol. 158, 230
237. Cheryan, M., 1998. Ultrafiltration and Microfiltration Handbook. CRC Press, Cleveland, OH. Cheryan, M., Alvarez, J.R., 1995. Chapter 9 food and beverage industry applications. In: Noble, R.D., Stern, S.A. (Eds.), Membrane Science and Technology. Elsevier, Amsterdam, pp. 415
465. Christensen, K., et al., 2006. Using direct contact membrane distillation for whey protein concentration. Desalination 200 (1
3), 523
525. Cowan, S., Ritchie, S., 2007. Modified polyethersulfone (PES) ultrafiltration membranes for enhanced filtration of whey proteins. Sep. Sci. Technol. 42 (11), 2405
2418. Creamer, L.K., Berry, G.P., Mills, O.E., 1977. A study of the dissociation b-casein from the bovine casein micelle at low temperature. NZ J. Dairy Sci. Technol. 12 (1), 58
66. Cui, Z., Taha, T., 2003. Enhancement of ultrafiltration using gas sparging: a comparison of different membrane modules. J. Chem. Technol. Biotechnol. 78 (2
3), 249
253. Cui, Z.F., Wright, K.I.T., 1996. Flux enhancements with gas sparging in downwards crossflow ultrafiltration: performance and mechanism. J. Memb. Sci. 117 (1), 109
116. Curcio, S., Calabro`, V., Iorio, G., 2002. Monitoring and control of TMP and feed flow rate pulsatile operations during ultrafiltration in a membrane module. Desalination 145 (1), 217
222. Dalgleish, D.G., 2011. On the structural models of bovine casein micelles - review and possible improvements. Soft Matter 7, 2265
2272. Daufin, G., et al., 2001. Recent and emerging applications of membrane processes in the food and dairy industry. Food Bioprod. Process. 79 (2), 89
102. Davies, D.T., White, J.C.D., 1960. The use of ultrafiltration and dialysis in isolating the aqueous phase of milk and in determining the partition of milk constituents between the aqueous and disperse phases. J. Dairy Res. 27, 171
190. de Kruif, C.G., 1999. Casein micelle interactions. Int. Dairy J. 9, 183
188. Delbeke, R., 1981. Recovery of milk solids by hyperfiltration. Milchwissenschaft. Ding, L., et al., 2002. Milk protein concentration by ultrafiltration with rotating disk modules. Desalination 144 (1), 307
311. Durham, R.J., Hourigan, J.A., 2009. Waste management and co-product recovery in dairy processing. In: Waldron, K. (Ed.), Handbook of Waste Management and Co-Product Recovery in Food Processing, Volume 1. Woodhead Publishing Limited and CRC Press LLC, Cambridge, pp. 332
387. Eide, M.H., Homleid, J.P., Mattsson, B., 2003. Life cycle assessment (LCA) of cleaning-in-place processes in dairies. LWT - Food Sci. Technol. 36 (3), 303
314. Engler, J., Wiesner, M.R., 2000. Particle fouling of a rotating membrane disk. Water Res. 34 (2), 557
565. Espina, V., Jaffrin, M.Y., Ding, L., 2009. Extraction and separation of α-lactalbumin and β-lactoglobulin from skim milk by microfiltration and ultrafiltration at high shear rates: a feasibility study. Sep. Sci. Technol. 44 (16), 3832
3853.

298

Separation of Functional Molecules in Food by Membrane Technology

Evans, J., et al., 2010. Comparison of composition and sensory properties of 80% whey protein and milk serum protein concentrates1. J. Dairy Sci. 93 (5), 1824
1843. Eykens, L., et al., 2018. Recovery of brines from cheese making using membrane distillation at lab and pilot scale. J. Food Eng. 219 (Suppl. C), 52
61. Farrell Jr., H.M., et al., 2004. Nomenclature of the proteins of cows’ milk - Sixth revision. J. Dairy Sci. 87 (6), 1641
1674. Fauguant, J., Maubois, J.L., Pierre, A., 1988. Microfiltration du lait sur membrane mine´rale. Tech. Lait 1028, 21
23. Fauquant, J., Robert, B., Lopez, C., 2016. Method for Reducing the Bacterial Content of a Food Medium Containing Lipid Droplets. Google Patents. Ferna´ndez Garcı´a, L., Riera Rodrı´guez, F.A., 2015. Microfiltration of milk with third generation ceramic membranes. Chem. Eng. Commun. 202 (11), 1455
1462. ´ lvarez Blanco, S., Riera Rodrı´guez, F.A., 2013. Microfiltration applied Ferna´ndez Garcı´a, L., A to dairy streams: removal of bacteria. J. Sci. Food. Agric. 93 (2), 187
196. Field, R., et al., 1995. Critical flux concept for microfiltration fouling. J. Memb. Sci. 100 (3), 259
272. Fleming, R., 1999. On-farm Concentration of Milk. U.S. Department ofAgriculture. Fox, P.F., McSweeney, P.L.H., 1998. Dairy Chemistry and Biochemistry, first ed. UK Academic & Professional, London: Blackie. Fox, P.F., et al., 2015. Heat-induced changes in milk. Dairy Chemistry and Biochemistry. Springer International Publishing, Cham, pp. 345
375. Gamlath, C.J., et al., 2018. The inhibitory roles of native whey protein on the rennet gelation of bovine milk. Food. Chem. 244, 36
43. Garcera, D., Toujas, E., 2000. Macroporous Support Having a Permeability Gradient and Method for Its Production. Google Patents. Ge, Q., et al., 2012. Polyelectrolyte-promoted forward osmosis
membrane distillation (FO
MD) hybrid process for dye wastewater treatment. Environ. Sci. Technol. 46 (11), 6236
6243. Ge´san-Guiziou, G., 2013. 12 - Separation technologies in dairy and egg processing A2. In: Rizvi, Syed S.H. (Ed.), Separation, Extraction and Concentration Processes in the Food, Beverage and Nutraceutical Industries. Woodhead Publishing, Cambridge, pp. 341
380. Ge´san-Guiziou, G., Boyaval, E., Daufin, G., 1999. Critical stability conditions in crossflow microfiltration of skimmed milk: transition to irreversible deposition. J. Memb. Sci. 158 (1-2), 211
222. Glover, F.A., 1971. Concentration of milk by ultrafiltration and reverse osmosis. J. Dairy Res. 38 (3), 373
379. Goff, H.D., Griffiths, M.W., 2006. Major advances in fresh milk and milk products: fluid milk products and frozen desserts. J. Dairy Sci. 89 (4), 1163
1173. Goude´dranche, H., Fauquant, J., Maubois, J.-L., 2000. Fractionation of globular milk fat by membrane microfiltration. Lait 80 (1), 93
98. Goudedranche, H.Y.M., Maubois, J.L.J., Fauquant, J.E., 2003. Method for Obtaining Milk Products Containing Selected Fractions of Fat Globules. Google Patents. Govindasamy-Lucey, S., et al., 2007. Use of cold microfiltration retentates produced with polymeric membranes for standardization of milks for manufacture of pizza cheese. J. Dairy Sci. 90 (10), 4552
4568. Guinee, T.P., et al., 1995. Composition, microstructure and maturation of semi-hard cheeses from high protein ultrafiltered milk retentates with different levels of denatured whey protein. Int. Dairy J. 5 (6), 543
568.

Membrane Separations in the Dairy Industry Chapter | 8

299

Guinee, T.P., O’Kennedy, B.T., Kelly, P.M., 2006. Effect of milk protein standardization using different methods on the composition and yields of cheddar cheese. J. Dairy Sci. 89 (2), 468
482. Hausmann, A., et al., 2011. Direct contact membrane distillation of dairy process streams. Membranes 1 (1), 48
58. Hausmann, A., et al., 2014. Performance assessment of membrane distillation for skim milk and whey processing. J. Dairy Sci. 97 (1), 56
71. Heino, A., Uusi-Rauva, J., Outinen, M., 2009. Microfiltration of milk III: impact of milk modification on milk coagulation kinetics. Milchwissenschaft 64 (2), 128
131. Heldman, D.R., Lund, D.B., Sabliov, C., 2006. Handbook of Food Engineering, second ed. Taylor & Francis, Boca Raton, FL. Hengl, N., et al., 2014. A new way to apply ultrasound in cross-flow ultrafiltration: application to colloidal suspensions. Ultrason. Sonochem. 21 (3), 1018
1025. Henning, D.R., et al., 2006. Major advances in concentrated and dry milk products, cheese, and milk fat-based spreads1. J. Dairy Sci. 89 (4), 1179
1188. Hillier, R.M., Lyster, R.L.J., 1979. Whey protein denaturation in heated milk and cheese whey. J. Dairy Res. 46, 95
102. Hoffmann, W., et al., 2006. Processing of extended shelf life milk using microfiltration. Int. J. Dairy Technol. 59 (4), 229
235. Holt, C., 2004. An equilibrium thermodynamic model of the sequestration of calcium phosphate by casein micelles and its application to the calculation of the partition of salts in milk. Eur. Biophys. J. 33 (5), 421
434. Holt, C., Sawyer, L., 1993. Caseins as rheomorphic proteins: interpretation of primary and secondary structures of the αs1-, β- and κ-caseins. J. Chem. Soc. Faraday Trans. 89, 2683
2692. Horne, D.S., 2006. Casein micelle structure: models and muddles. Curr. Opin. Colloid Interface Sci. 11 (2,3), 148
153. Horton, B., 1997. Water, chemical and brine recycle or reuse-Applying membrane processes. Austr. J. Dairy Technol. 52 (1), 68. Jaffrin, M.Y., Gupta, B.B., Paullier, P., 1994. Energy saving pulsatile mode cross flow filtration. J. Memb. Sci. 86 (3), 281
290. Johnson, M.E., Lucey, J.A., 2006. Major technological advances and trends in cheese. J. Dairy Sci. 89 (4), 1174
1178. Jørgensen, C.E., et al., 2016. Optimization of protein fractionation by skim milk microfiltration: choice of ceramic membrane pore size and filtration temperature. J. Dairy Sci. 99 (8), 6164
6179. KOCH Membrane Systems, 2017. Recovery of Caustic and Acids in the Dairy Industry [cited 2017 09 September], Available from: http://www.kochmembrane.com/resources/applicationbulletins/recovery-of-caustic-and-acids-in-the-dairy-industr.aspx. Karasu, K., et al., 2010. A comparison between ceramic and polymeric membrane systems for casein concentrate manufacture. Int. J. Dairy Technol. 63 (2), 284
289. Kezia, K., et al., 2015. Direct contact membrane distillation for the concentration of saline dairy effluent. Water Res. 81, 167
177. Koh, L.L.A., et al., 2014. The use of ultrasonic feed pre-treatment to reduce membrane fouling in whey ultrafiltration. J. Memb. Sci. 453, 230
239. Kromkamp, I.J., Van, R.D.C.J., 2002. Method for Filtering Milk. Google Patents. Krsti´c, D.M., et al., 2002. The effect of turbulence promoter on cross-flow microfiltration of skim milk. J. Memb. Sci. 208 (1), 303
314.

300

Separation of Functional Molecules in Food by Membrane Technology

Kumar, P., et al., 2013. Perspective of membrane technology in dairy industry: a review. AsianAustralas. J. Anim. Sci. 26 (9), 1347
1358. Lawrence, N.D., et al., 2008. Microfiltration of skim milk using polymeric membranes for casein concentrate manufacture. Sep. Purif. Technol. 60 (3), 237
244. Le Berre, O., Daufin, G., 1996. Skimmilk crossflow microfiltration performance versus permeation flux to wall shear stress ratio. J. Memb. Sci. 117 (1-2), 261
270. Lee, J., et al., 2005a. Determination of the size distribution of sonoluminescence bubbles in a pulsed acoustic field. J. Am. Chem. Soc. 127 (48), 16810
16811. Lee, J., Kentish, S., Ashokkumar, M., 2005b. Effect of surfactants on the rate of growth of an air bubble by rectified diffusion. J. Phys. Chem. B 109 (30), 14595
14598. Leong, T., et al., 2011. The role of surfactant headgroup, chain length, and cavitation microstreaming on the growth of bubbles by rectified diffusion. J. Phys. Chem. C 115 (49), 24310
24316. Leong, T.S.H., et al., 2009. Minimising oil droplet size using ultrasonic emulsification. Ultrason. Sonochem. 16 (6), 721
727. Lindquist, A., 1998. A Method for the Production of Sterile Skimmed Milk. Google Patents. Liu, D.Z., Dunstan, D.E., Martin, G.J.O., 2012. Evaporative concentration of skim milk: effect on casein micelle hydration, composition, and size. Food. Chem. 134, 1446
1452. Liu, D.Z., et al., 2013. Temperature-dependent dynamics of bovine casein micelles in the range 10-40 C. Food. Chem. 141, 4081
4086. Liu, D.Z., et al., 2014. Alterations to the composition of casein micelles and retentate serum during ultrafiltration of skim milk at 10 and 40 C. Int. Dairy J. 35 (1), 63
69. Lucey, J.A., Singh, H., 1997. Formation and physical properties of acid milk gels: a review. Food Res. Int. 30 (7), 529
542. Lucey, J.A., Smith, K., 2009. An integrated processing system to produce beta-casein, native serum protein and casein concentrates from whole milk. J. Dairy Sci. 92 (E-Suppl. 1), 164. Luja´n-Facundo, M., et al., 2013. Ultrasonic cleaning of ultrafiltration membranes fouled with BSA solution. Sep. Purif. Technol. 120, 275
281. Luja´n-Facundo, M., et al., 2016. Cleaning efficiency enhancement by ultrasounds for membranes used in dairy industries. Ultrason. Sonochem. 33, 18
25. Madec, M.N., Mejean, S., Maubois, J.L., 1992. Retention of Listeria and Salmonella cells contaminating skim milk by tangential membrane microfiltration (“Bactocatch” process). Lait 72 (3), 327
332. Malmberg, R., Holm, S., 1988. Low bacteria skim milk by microfiltration. North Eur. Food Dairy J. 54, 75
78. Marella, C., Muthukumarappan, K., Metzger, L.E., 2011. Evaluation of commercially available, wide-pore ultrafiltration membranes for production of α-lactalbumin-enriched whey protein concentrate. J. Dairy Sci. 94 (3), 1165
1175. Marshall, A., Daufin, G., 1995. Physico-chemical aspects of membrane fouling by dairy fluids. Int. Dairy J. 9504, 8
35. Mason, T.J., et al., 2011. New evidence for the inverse dependence of mechanical and chemical effects on the frequency of ultrasound. Ultrason. Sonochem. 18 (1), 226
230. Masselin, I., et al., 2001. Effect of sonication on polymeric membranes. J. Memb. Sci. 181 (2), 213
220. Maubois, J., 2002. Membrane microfiltration: a tool for a new approach in dairy technology. Austr. J. Dairy Technol. 57 (2), 92. Maubois, J.L., Mocquot, G., Vassal, L., 1969. A Method for Processing Milk and Dairy Products. FR Patent 2052121.

Membrane Separations in the Dairy Industry Chapter | 8

301

Maubois, J.L., et al., 1987. Industrial fractionation of main whey proteins. Bullet. Int. Dairy Federat. J. 212, 154
159. McKenna, B.M., O’Sullivan, A.C., 1971. Whey protein denaturation in concentrated skim milks. J. Dairy Sci. 54, 1075
1077. Meersohn, M., 1989. Nitrate free cheese making with Bactocatch. North Eur. Food Dairy J. 55, 108
113. Merin, U., et al., 2002. Cleaning-in-place in the dairy industry: criteria for reuse of caustic (NaOH) solutions. Lait 82, 357
366. Meyer, P., Kulozik, U., 2016. Impact of protein removal by an upstream ultrafiltration on the reverse osmosis of skim milk and sweet whey. Chemie Ingenieur Technik 88 (5), 585
590. Michalski, M., et al., 2002. Native vs damaged milk fat globules: membrane properties affect the viscoelasticity of milk gels. J. Dairy Sci. 85, 2451
2461. Michalski, M.-C., Briard, V., Michel, F., 2001. Optical parameters of milk fat globules for laser light scattering measurements. Lait 81 (6), 787
796. Michaud, M., et al., 2015. Design parameters of stainless steel plates for maximizing high frequency ultrasound wave transmission. Ultrason. Sonochem. 26, 56
63. Mistry, V.V., Maubois, J.-L., 1993. Application of membrane separation technology to cheese production. In: Fox, P.F. (Ed.), Cheese: Chemistry, Physics and Microbiology: Volume 1 General Aspects. Springer US, Boston, MA, pp. 493
522. Mistry, V.V., Maubois, J.L., 2004. Application of membrane separation technology to cheese production. In: Fox, P.F. (Ed.), Cheese: Chemistry, Physics, and Microbiology. Elsevier, Amsterdam, pp. 261
285. Mores, W.D., Davis, R.H., 2003. Yeast-fouling effects in cross-flow microfiltration with periodic reverse filtration. Ind. Eng. Chem. Res. 42 (1), 130
139. Muir, D.D., 2011. 27 - The stability and shelf life of milk and milk products. Food and Beverage Stability and ShelfLife. Woodhead Publishing, Cambridge, pp. 755
778. Mulder, M., 1996. Basic Principles of Membrane Technology, second ed. Kluwer, Dordrecht. Muller, A., Daufin, G., Chaufer, B., 1999. Ultrafiltration modes of operation for the separation of α-lactalbumin from acid casein whey. J. Memb. Sci. 153 (1), 9
21. Muthukumaran, S., et al., 2004. The use of ultrasonic cleaning for ultrafiltration membranes in the dairy industry. Sep. Purif. Technol. 39 (1), 99
107. Muthukumaran, S., et al., 2005a. Mechanisms for the ultrasonic enhancement of dairy whey ultrafiltration. J. Memb. Sci. 258 (1), 106
114. Muthukumaran, S., et al., 2005b. The optimisation of ultrasonic cleaning procedures for dairy fouled ultrafiltration membranes. Ultrason. Sonochem. 12 (1), 29
35. Muthukumaran, S., et al., 2007. The application of ultrasound to dairy ultrafiltration: the influence of operating conditions. J. Food Eng. 81 (2), 364
373. Naddeo, V., Borea, L., Belgiorno, V., 2015. Sonochemical control of fouling formation in membrane ultrafiltration of wastewater: effect of ultrasonic frequency. J. Water Process Eng. 8, e92
e97. Nath, K., 2008. Membrane Separation Processes. PHI Learning Pvt. Ltd, New Delhi. Neocleous, M., Barbano, D.M., Rudan, M.A., 2002. Impact of low concentration factor microfiltration on the composition and aging of cheddar cheese. J. Dairy Sci. 85 (10), 2425
2437. Ng, K.S.Y., et al., 2017. Mechanisms of flux decline in skim milk ultrafiltration: a review. J. Memb. Sci. 523, 144
162. Ng, K.S.Y., Dunstan, D.E., Martin, G.J.O., 2018. Influence of processing temperature on flux decline during skim milk ultrafiltration. Sep. Purif. Technol. 195, 322
331.

302

Separation of Functional Molecules in Food by Membrane Technology

Nystro¨m, M., Kaipia, L., Luque, S., 1995. Fouling and retention of nanofiltration membranes. J. Memb. Sci. 98 (3), 249
262. Oatley-Radcliffe, D.L., et al., 2017. Nanofiltration membranes and processes: a review of research trends over the past decade. J. Water Process Eng. 19 (Suppl. C), 164
171. OECD, O.f.E.C.-o.a.D. and F.a.A.O. FAO, OECD-FAO Agricultural Outlook 2011-2022, Chapter 9. 2012. Oldfield, D.J., Singh, H., Taylor, M.W., 2005. Kinetics of heat-induced whey protein denaturation and aggregation in skim milks with adjusted whey protein concentration. J. Dairy Res. 72 (3), 369
378. Ottosen, N., Kønigsfeldt, P., 1999. Microfiltration of cheese brine. SDI. Scandinavian Dairy Information (Denmark). Papadatos, A., et al., 2003. Economic feasibility evaluation of microfiltration of milk prior to cheesemaking. J. Dairy Sci. 86 (5), 1564
1577. Pearce, R., 1983. Thermal separation of [beta]-Lactoglobulin and [alpha]-Lactalbumin in bovine cheddar cheese whey. Austr. J. Dairy Technol. 38 (4), 144. Pedersen, P., 1992. Microfiltration for the reduction of bacteria in milk and brine. Bull. Int. Dairy Federat. 1, 33
50. Phuntsho, S., et al., 2012. Fertiliser drawn forward osmosis desalination: the concept, performance and limitations for fertigation. Rev. Environ. Sci.d Bio/Technol. 11 (2), 147
168. Pouliot, M., Pouliot, Y., Britten, M., 1996. On the conventional cross-flow microfiltration of skim milk for the production of native phosphocaseinate. Int. Dairy J. 6 (1), 105
111. Punidadas, P., Rizvi, S.S.H., 1998. Separation of milk proteins into fractions rich in casein or whey proteins by cross flow filtration. Food Res. Int. 31 (4), 265
272. Rabiller-Baudry, M., et al., 2005. Limiting flux in skimmed milk ultrafiltration: impact of electrostatic repulsion due to casein micelles. Desalination 175, 49
59. Ra¨sa¨nen, E., et al., 2002. Purification and regeneration of diluted caustic and acidic washing solutions by membrane filtration. Desalination 149 (1
3), 185
190. Rattray, W., Jelen, P., 1996. Protein standardization of milk and dairy products. Trends Food Sci. Technol. 7 (7), 227
234. Redkar, S., Kuberkar, V., Davis, R.H., 1996. Modeling of concentration polarization and depolarization with high-frequency backpulsing. J. Memb. Sci. 121 (2), 229
242. Rice, G., 2008. Membrane separation of calcium salts from dairy ultrafiltraion permeates. Department of Chemical and Biomolecular Engineering. The University of Melbourne, Melbourne. Rice, G., et al., 2006. Fouling behaviour during the nanofiltration of dairy ultrafiltration permeate. Desalination 199 (1), 239
241. Rice, G.S., 2008. Membrane Separation of Calcium Salts from Dairy Ultrafiltration Permeates. University of Melbourne, Department of Chemical & Biomolecular Engineering, Melbourne. Rose, D., 1968. Relation between micellar and serum casein in bovine milk. J. Dairy Sci. 51, 1897
1902. Rosenberg, M., 1995. Current and future applications for membrane processes in the dairy industry. Trends Food Sci. Technol. 6 (1), 12
19. Saboyainsta, L.V., Maubois, J.-L., 2000. Current developments of microfiltration technology in the dairy industry. Lait 80 (6), 541
553. Sandblom, R.M., 1978. Filtering Process. Google Patents.

Membrane Separations in the Dairy Industry Chapter | 8

303

Shu, L., Xing, W., Xu, N., 2007. Effect of ultrasound on the treatment of emulsification wastewater by ceramic membranes. Supported by the Special Funds for Major State Basic Research Program of China (973 Program, No.2003CB615707) and the National Natural Science Foundation of China (No.20436030). Chinese J. Chem. Eng. 15 (6), 855
860. Skrzypek, M., Burger, M., 2010. Isofluxs ceramic membranes — practical experiences in dairy industry. Desalination 250 (3), 1095
1100. Slack, A.W., Amundson, C.H., Hill Jr., C.G., 1982. On-farm ultrafiltration of milk: parteconomic analysis. Process Biochem. 17, 23
25. 30,33. Slack, A.W., Amundson, C.H., Hill, C.G., 1986. Production of enriched β-lactoglobulin and A-lactalbumin whey protein fractions. J. Food Process. Preservation 10 (1), 19
30. St-Gelais, D., Roy, D., Audet, P., 1998. Manufacture and composition of low fat Cheddar cheese from milk enriched with different protein concentrate powders. Food Res. Int. 31 (2), 137
145. Thomann, S., Schenkel, P., Hinrichs, J., 2008. Effect of homogenization, microfiltration and pH on curd firmness and syneresis of curd grains. LWT - Food Sci. Technol. 41 (5), 826
835. Timmen, H., Patton, S., 1988. Milk fat globules: fatty acid composition, size and in vivo regulation of fat liquidity. Lipids 23 (7), 685
689. Tremblay-Marchand, D., et al., 2016. A process efficiency assessment of serum protein removal from milk using ceramic graded permeability microfiltration membrane. J. Dairy Sci. 99 (7), 5230
5243. Vadi, P.K., Rizvi, S.S.H., 2001. Experimental evaluation of a uniform transmembrane pressure crossflow microfiltration unit for the concentration of micellar casein from skim milk. J. Memb. Sci. 189 (1), 69
82. van Leeuwen, H.J., et al., 1990. Preparation of Hard Cheese From Concentrated Milk. Google Patents. Visvanathan, C., Aim, R.B., 1989. Application of an electric field for the reduction of particle and colloidal membrane fouling in crossflow microfiltration. Sep. Sci. Technol. 24 (5
6), 383
398. Vourch, M., et al., 2008. Treatment of dairy industry wastewater by reverse osmosis for water reuse. Desalination 219 (1-3), 190
202. Wakeman, R., Tarleton, E., 1991. An experimental study of electroacoustic crossflow microfiltration. Chem. Eng. Re-search Design 69a, 386
397. Wakeman, R.J., 1998. Electrically enhanced microfiltration of albumin suspensions. Food Bioprod. Process. 76 (1), 53
59. Walstra, P., 1979. The voluminosity of bovine casein micelles and some of its implications. J. Dairy Res. 46 (2), 317
323. Walstra, P., 1990. On the stability of casein micelles. J. Dairy Sci. 73 (8), 1965
1979. Walstra, P., 1999. Dairy Technology: Principles of Milk Properties and Processes. CRC Press, New York. Wang, J., et al., 2016. Ultrasonic-assisted acid cleaning of nanofiltration membranes fouled by inorganic scales in arsenic-rich brackish water. Desalination 377, 172
177. Wang, K.Y., et al., 2011. Integrated forward osmosis
membrane distillation (FO
MD) hybrid system for the concentration of protein solutions. Chem. Eng. Sci. 66 (11), 2421
2430. Watkins, M., Nash, D., 2010. Dairy factory wastewaters, their use on land and possible environmental impacts: a mini review. Open Agric. J. 4. Weissler, A., 1959. Formation of hydrogen peroxide by ultrasonic waves: free radicals. J. Am. Chem. Soc. 81 (5), 1077
1081.

304

Separation of Functional Molecules in Food by Membrane Technology

Wilkinson, K.G., 2007. Characterisation of selected dairy processing waste streams from Victoria, Australia. Austr. J. Dairy Technol. 62 (3), 159. Winzeler, H.B., Belfort, G., 1993. Enhanced performance for pressure-driven membrane processes: the argument for fluid instabilities. J. Memb. Sci. 80 (1), 35
47. Yorgun, M., Balcioglu, I.A., Saygin, O., 2008. Performance comparison of ultrafiltration, nanofiltration and reverse osmosis on whey treatment. Desalination 229 (1-3), 204
216. Zulewska, J., Barbano, D.M., 2013. Influence of casein on flux and passage of serum proteins during microfiltration using polymeric spiral-wound membranes at 50 C. J. Dairy Sci. 96 (4), 2048
2060.

Chapter 9

Current and Future Applications of Nanofiltration in Food Processing Alfredo Cassano1, Carmela Conidi1 and Roberto Castro-Mun˜oz1,2,3 1

Institute on Membrane Technology, ITM-CNR, Rende, Italy, 2University of Chemistry and Technology, Prague, Prague, Czech Republic, 3Nanoscience Institute of Aragon (INA), University of Zaragoza, Zaragoza, Spain

Chapter Outline Abbreviations 9.1 Introduction 9.2 General Aspects of NF Membranes 9.3 Fruit Juice Processing 9.4 Wine and Must Processing 9.5 Dairy Industry 9.5.1 Concentration and Demineralization of Whey

305 306 307 310 314 319 319

9.5.2 Concentration and Demineralization of UF Whey Permeate 9.5.3 Recovery of Lactic Acid 9.6 Sugar Industry 9.7 Recovery of Functional Compounds From Food Processing Byproducts 9.8 Conclusions and Future Trends References Further Reading

320 322 325

329 342 342 348

ABBREVIATIONS MF MWCO NF RO TMP UF VRF

microfiltration molecular weight cut-off nanofiltration reverse osmosis transmembrane pressure ultrafiltration volume reduction factor

Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00009-7 © 2019 Elsevier Inc. All rights reserved.

305

306

Separation of Functional Molecules in Food by Membrane Technology

9.1 INTRODUCTION Today, one of the most important trends in the food industry is the demand for “natural foods” and ingredients that are free from toxic chemical additives and with high quality grade. To meet food safety and quality, new processes have been introduced following the concept of “green food processing.” “Green food processing” is based on the discovery and design of technical processes that will reduce energy and water consumption, allowing recycling of byproducts through biorefinery and ensuring a safe and high quality product (Boye and Arcand, 2013; Chemat et al., 2017). This is a new concept introduced in the 21st century, to protect both the environment and consumers, and in the meantime enhance competition of industries to be more ecological, economical, and innovative. During the last decade, the interest in the use of membrane technology has emerged in different fields of food processing as alternative technology to conventional separation methods. This growth can be explained by the continuous demand of both manufacturers and consumers for safer products, minimally processed foods free of external contaminants, and healthpromoting foods. Other important issues are those related to the health and environmental impact of food processing (Sen˜orans et al., 2003). In this contest, membrane processes offer several advantages over conventional methodologies such as mild operating conditions of temperature and pressure, therefore preserving the functional properties of food products, as well as the possibility to avoid the use of chemical agents or solvents, and, consequently, product contamination. In addition, they are characterized by high separation efficiency, simple equipment, easy scale-up, reduced number of processing steps, and low energy consumption. Based on these properties membrane processes represent innovative and economical systems for purification and/or fractionation of food-based fluids (Daufin et al., 2001; Malik et al., 2013; Kotsanopoulos and Arvanitoyannis, 2015). These processes are based on the use of permselective barriers that allow the diffusion of specific feed components while retaining the others. In particular, pressure-driven membrane processes, such as microfiltration (MF), ultrafiltration (UF), nanofiltration (NF), and reverse osmosis (RO) together with electrodialysis (ED) and osmotic membrane distillation have been widely investigated as green technologies, alternatives to conventional ones, for processing new ingredients and foods (Dhineshkumar and Ramasamy, 2017). NF is a relatively new process among the pressure-driven membrane separation technologies. It is still evolving for more challenging applications in agro-food processing, involving fractionation, water softening, wastewater treatment, vegetable oil processing, and treatment of products from the dairy, beverage, and sugar industry (Salehi, 2014).

Current and Future Applications of Nanofiltration Chapter | 9

307

Most industrialized countries have strict regulations governing hygiene of food products that must be adhered during processing. In addition, low-fat and low-calorie products as well as products suitable for special diets require more and improved separations. In this view, NF membranes have a great potential due to their capacity to separate monovalent and multivalent ions as well as organic solutes with different size from other species. NF membranes have been explored in different areas of food processing, including sugar and water purification, juice concentration, concentration of whey protein, as well as fractionation and selective removal of solutes from complex process streams making the technology commercially attractive (Van der Bruggen et al., 2003; Salehi, 2014). The growing demand for NF membranes is one of the key emerging trends in the market of membranes for agro-food applications. According to data reported by Allied Market Research (2017), the global NF membranes market was valued at $565 million in 2016, and is expected to reach at $813 million by 2023, registering a compound annual growth rate of 5.3% from 2017 to 2023. This chapter provides an overview of the recent development of NF membranes in food processing. After summarizing the general aspects of the process, specific applications of NF in different areas of the agro-food processing industry (i.e., fruit juice, wine, dairy and sugars), as well as for the recovery of high-added value compounds from food processing byproducts are presented and discussed, highlighting their key advantages over conventional technologies.

9.2 GENERAL ASPECTS OF NF MEMBRANES In pressure-driven membrane processes the membrane acts as a selective barrier through which solvent fluids with permeable solutes are selectively transported under a hydrostatic pressure applied on the feed side. As a result, the feed solution is divided into a permeate fraction containing all components that have permeated the membrane and a retentate fraction containing all compounds rejected by the membrane, within the retained solvent. NF is a relatively new pressure-driven membrane process, introduced in the late 1980s. The separation capabilities of NF membranes are situated between those of UF and RO membranes, with pore size typically of 0.5
2 nm, corresponding to molecular weight cut-off (MWCO) of 200
1000 Da (Paul and Jons, 2016). Polymeric NF membranes contain ionizable groups, such as carboxylic and sulfonic acid groups, which result in charged surfaces in the presence of a feeding solution. As a consequence, the NF separation characteristics are not based only on size exclusion but also on the so-called Donnan exclusion, which postulates that ions carrying the same charge as the membrane (socalled coions) will be excluded by the membrane (Mohammad et al., 2015).

308

Separation of Functional Molecules in Food by Membrane Technology

NF membranes can efficiently remove small organic molecules and inorganic salt just like RO membranes. However, higher rejection of divalent ions, lower rejection of monovalent ions, and higher flux are the key features of distinction between NF and RO membranes. In addition, NF processes are characterized by lower energy consumption and lower applied pressures than RO (in the range 10
40 bar) and better retention than UF for low molar mass molecules such as sugars, amino acids, peptides, and ions (Luo and Wan, 2013). This makes the separation process highly competitive in terms of selectivity and cost benefit when compared with traditional technologies. As a result, NF represents a promising process in the food industry due to its ability to separate and fractionate ionic and relatively low molecular weight (LMW) organic species. Membrane fouling is the main drawback of NF applications. It can be generated by binding, accumulation, or absorption of materials on the surfaces of the membrane and/or within the porous structure, causing deleterious effects including flux decline, increased costs due to increased energy demand, chemicals and frequency of cleaning, and shorter membrane lifespan. Therefore the selection of a suitable membrane in terms of MWCO, material (functional groups, charge, hydrophobicity), and morphology (i.e., surface roughness) is very important in determining the success of a specific application, meaning the retained and permeable compounds, the stability in the presence of the main solvent and the adsorption of solutes. Commonly used NF membranes are polymeric in nature, including cellulose acetate, polyamide, polyimide, polysulfone and polyethersulfone or ceramic such as zirconia, titania, silica
zirconia, and alumina. Compared with polymeric NF membranes, ceramic membranes generally show a higher chemical, structural, and thermal stability. Polymeric membranes generally present a thinfilm composite (TFC) structure (made of polyamides or polyethersulfone) that combines high selectivity with high permeability (Van der Bruggen, 2009). NF membranes are produced in different configurations such as plateand-frame, spiral-wound, and tubular. The choice of a specific configuration is based on economic considerations associated with performance, ease of cleaning, and replacement. The predominant configuration of NF membranes used today for industrial applications is the spiral-wound one. The most common process configurations used in NF processes are total recycle, batch concentration, feed-and-bleed, diafiltration (DF), and multistage operation. In the total recycle configuration permeate and retentate streams are returned to the feed reservoir and no net concentration of particles occurs. This configuration is mainly used to study the effect of different operating parameters on the permeate flux. In the batch concentration the retentate is partially or totally recycled back to the feed reservoir and the permeate is collected separately. The feed-and-bleed configuration is commonly used for continuous full-scale operation. It is characterized by the

Current and Future Applications of Nanofiltration Chapter | 9

309

removal of the permeate from the system together with a small part of the retentate. The latest is recycled and mixed with the feed solution to maintain a high tangential velocity in the membrane module. A feed pump and a recirculation pump are required to provide the required transmembrane pressure and the cross-flow, respectively. In diafiltration the retentate is recycled to the feed reservoir and the permeate is replaced by an equal volume of pure solvent. An overall DF process may involve a preconcentration step, the DF step and a postconcentration step. This configuration is used to overcome low permeate fluxes at high concentrations or to improve the separation of permeable compounds. Single stage NF is usually limited by the relatively low purity that can be achieved. On the other hand, multistage or cascade NF are used for the fractionation of a feed mixture to the desired level of purity, when this is not achieved in a single step. A cascade is a combination of different units, where the permeate or retentate of one stage is used as feed for the subsequent one. The selection of a specific process configuration depends on the specific goal and applications. The membrane performance in NF is characterized in terms of separation efficiency and recovery factor. The separation efficiency is generally expressed by the rejection (R) of a given compound, which is given by:  Cp R5 12  100 ð9:1Þ Cf where Cp is the solute concentration in the permeate and Cf the solute concentration in the feed. Rejection values range between 0% (for solutes that completely pass through the NF membrane) and 100% (for solutes completely rejected by the membrane) (Mulder, 1991). The recovery factor (Δ) is defined as the fraction of feed flow which passes through the membrane: Δ5

Vp Vf

ð9:2Þ

where Vp and Vf are the volumes of permeate and feed solution. The recovery ranges from 0 and 1 and it is a parameter of economic importance. Commercial NF membrane processes are often designed with a recovery value as high as possible. The volume reduction factor (VRF) in NF processes is defined as the ratio between the initial feed volume and the volume of the resulting retentate given by: VRF 5

Vf Vp 511 Vr Vr

ð9:3Þ

where Vf, Vp, and Vr are the volume of feed, permeate, and retentate, respectively.

310

Separation of Functional Molecules in Food by Membrane Technology

9.3 FRUIT JUICE PROCESSING Fruit juices are recognized as important components of the human diet, providing a range of key nutrients as well as many nonnutrient phytochemicals, which are important for their role in preventing chronic diseases such as cancer, cardiovascular, and neurological disorders. Currently, their manufacturing process has attracted much attention in both consumers and researchers, becoming a critical issue in the food industry. During the production of fruit juices, the application of processes designed to concentrate the content of biologically active compounds present in the juice and to preserve the physical, chemical, and biological properties of these compounds is extremely important. Traditional methods of processing fruits limit the possibility to obtain products able to retain as much as possible the peculiarity of the fresh fruit and its high content in health-biologically active compounds (Cassano et al., 2011). Membrane technology has been used in the manufacture of fruit juices since the 1970s. It represents one of the technological answers to the problem of producing additive-free juices with standard organoleptic quality and natural fresh taste. As previously reported, the main advantages offered by membrane processes are in terms of energy savings, selectivity, and simplicity of operation. Additionally, the separation can be carried out at room or mild temperatures, allowing thermolabile compounds to be processed without chemical changes and losses in their functional properties (Bhattacharjee et al., 2017; Echavarrı´a et al., 2011; Jiao et al., 2004). Pressure-driven membrane processes have been successfully explored, also in integrated systems, for juice clarification and concentration. Juice clarification, concentration, and deacidification are typical steps where NF can be explored in association with MF or UF (Nath et al., 2018). NF offers new possibilities due to its advantages, which are mainly associated with the retention of lower particles when compared with UF and MF processes. In particular, the main advantage of employing NF membranes in the treatment of fruit juice is the possibility of fractionating molecules of similar molecular weight (in the range 100
1000 Da) by the selection of membranes with suitable MWCO. In addition, this process is advantageous when compared with RO for juice concentration: indeed it is less energyconsuming and allows to better preserve the juice quality since it operates at lower pressures. The NF process has also a great potential for the separation and concentration of bioactive compounds of fruit juices with potential exploitation in the preparation and formulation of functional food and beverages (Conidi et al., 2017a). The potential of NF in the concentration of bioactive compounds in watermelon juice has been investigated by Arriola et al. (2014) using a spiral-wound polyvinylidene difluoride (PVDF) membrane module with a

Current and Future Applications of Nanofiltration Chapter | 9

311

MWCO ranging between 150 and 300 Da (HL2521TF, from GE Osmonics). Most of the bioactive compounds were concentrated in the retentate side due to the high retention coefficient of the membrane. At a VRF of 3 rejection values detected for lycopene, flavonoids, and phenolic compounds were of 99%, 96%, and 65%, respectively. A strict correlation between the concentration level and the increasing of the antioxidant activity was also observed. Acosta et al. (2017) evaluated the performance of different NF flat-sheet membranes with MWCO of around 200 Da in the concentration of anthocyanins and ellagitannins from blackberry juice. All tested membranes exhibited a total rejection toward ellagitannin independently by their composition and structure. Among the selected membranes, the NF270 membrane (semiaromatic polypiperazineamide TFC, from Dow-Filmtec) showed the highest productivity (permeate flux) and solute retention when operating at a transmembrane pressure (TMP) of 30 bar. Sugars were also completely retained at high pressures, while the observed retention of titratable acidity was under 90% indicating the suitability of the membrane in juice deacidification. In a similar approach, Arend et al. (2017) evaluated the feasibility of the NF process in the concentration of the bioactive compounds from strawberry juice for the production of beverages with improved nutritional and sensorial properties. At this study, a NF membrane in PVDF with a MWCO of 150
300 Da (from GE Osmonics) was used to process both untreated and microfiltered juices at an operating pressure of 6 bar. The results showed an increase of the phenolic content and antioxidant activity by increasing the concentration factor of both juices due to the high rejection of the membrane toward polyphenols and anthocyanins (higher than 95% and 97%, respectively). In particular, the antioxidant activity of the NF retentate presented an increase of 99% and 51% starting from the natural and microfiltered juice, respectively. A correlation between the content of pelargonidin-3-O-glucoside and total antioxidant activity was also observed. The NF process allowed also to preserve the red color of the juice according to the global color variation (ΔE ) determination. Polyamide TFC NF membranes with sodium chloride rejection higher than 98% (from Alfa Laval) were used to concentrate chokeberry juice at different pressures (45, 50, and 55 bar) and temperatures (with cooling and without cooling of the retentate) (Popovic et al., 2016). Highest permeate fluxes (between 9.61 and 19.43 L/m2h) were obtained at 55 bar without cooling the retentate. This operating pressure provided also the highest concentration of phenolic compounds and the lowest loss of aromatic compounds. The NF process was considered suitable to concentrate the juice to a certain level and partially replace the conventional thermal evaporation process. The use of UF and NF flat-sheet membrane, with MWCO ranging from 1000 to 4000 Da, was tested to purify biologically active compounds from clarified pomegranate juice (Conidi et al., 2017b). The composition of the clarified juice produced in the treatment of the juice with a TFC membrane

312

Separation of Functional Molecules in Food by Membrane Technology

TABLE 9.1 Analyses of Biologically Active Compounds in Pomegranate Juice Samples Concentrated by NF Parameter

Fresh Juice

Ultrafiltered Juice

NF Permeate

NF Retentate

TSS ( Brix)

17.03 6 0.04

14 6 0.12

5.23 6 0.09

18.36 6 0.15

Glucose (g/L)

12.9 6 0.5

12.5 6 0.5

11.7 6 0.5

12.5 6 0.5

Fructose (g/L)

21.2 6 4.0

19.4 6 0.6

19.1 6 0.4

21.2 6 2.4

TAA (mM Trolox)

26.8 6 2.9

26.0 6 2.8

1.2 6 0.4

41.9 6 5.8

Total polyphenols (mg GAE/L)

2636.0 6 12.8

2457.5 6 15.3

65.2 6 0.7

3589.0 6 21.2

Cyanidin-3,5O-diglucoside (mg/L)

150.9 6 3.2

136.1 6 5.38

0.5 6 0.01

186.3 6 4.3

Cyanidin-3-Oglucoside (mg/L)

57.6 6 2.5

53.7 6 2.1

0.5 6 0.06

70.8 6 2.3

Pelargonidin3,5-Odiglucoside (mg/L)

4.6 6 0.9

3.5 6 0.9

0.65 6 0.01

6.2 6 0.3

Delphinidin3-O-glucoside

17.8 6 0.1

14.6 6 0.5

0.16 6 0.09

25.2 6 1.3

NF, nanofiltration; TAA, total antioxidant activity; TSS, total soluble solids.

having an approximate MWCO of 1000 Da (SelRO MPF-36, from Koch Membrane Systems) is reported in Table 9.1. According to the analytical results, rejection coefficients higher than 95% were measured for total antioxidant activity (TAA) and total polyphenols, while the retention for different anthocyanins was in the range of 80.4%
99.5%. As expected, a strict correlation was observed between the rejection toward phenolic compounds and TAA, since phenolic compounds mainly contribute to the TAA of the juice (Tezcan et al., 2009). On the other hand, lower rejections were measured for glucose and fructose (in the range 3.86
9.92) indicating the suitability of the NF process for the removal of sugar moieties in the permeate stream. NF membranes have been successfully employed for the fractionation and concentration of phenolic compounds from bergamot juice as alternative to conventional methodologies, too. For instance, tubular ceramic

Current and Future Applications of Nanofiltration Chapter | 9

313

membranes in titania with MWCO of 450 and 750 Da (from Inopor GmbH) were tested for the purification and concentration of polyphenols from depectinized juice previously clarified by UF (Conidi et al., 2011). The NF 750-Da membrane showed a rejection toward TAA and total polyphenols of 54% and 44%, respectively, while flavonoids identified in the juice (narirutin, naringin, hesperidin, and neohesperidin), with molecular weight from 550 to 610 Da, were only partially retained by the membrane (the rejection was in the range 43%
62%). This behavior was attributed to a reduction of the membrane pore size due to fouling phenomena. The NF 450-Da membrane showed higher rejections toward flavonoids (in the range 91%
99%) and TAA (90%). Therefore, antioxidant compounds, including flavonoids, were recovered in the retentate fraction. On the other hand, this membrane offered the best performance in terms of separation between sugars and flavonoids. Operating at a TMP of 33 bar and at a temperature of 24 C, a steady-state value of 18 L/m2h was reached after 80 minutes starting from an initial permeate flux of about 40 L/m2h. Conidi and Cassano (2015) evaluated also the performance of different NF polymeric spiral-wound membranes for the purification and concentration of polyphenols from ultrafiltered bergamot juice. Polyethersulfone membranes with MWCO of 400 and 1000 Da (NF PES10 and N30 F, from Microdyn Nadir) and a composite membrane composed by a polyamide top layer and a polysulfone microporous support with MWCO of 150
250 Da (NF 270 from Dow-Filmtec) were tested. For all selected membranes the steady-state permeate flux increased linearly by increasing the operating pressure (in the range 4
16 bar) with the NF 270 membrane exhibiting the highest value. The experimental results confirmed the key role of the chemical nature of the membrane in membrane performance. In addition, the hydrophobic character of PES membranes accounted for their higher fouling index when compared with the NF 270 membrane. A significant increase in the content of flavonoids (naringin, hesperidin, and neohesperidin) in the NF retentate of the selected membranes was observed due to the high rejection (in the range 85%
97.7%) measured toward these compounds. However, the NF PES 10 membrane showed the lowest average rejection toward sugar compounds (35%) providing the best separation of phenolic compounds from sugars (Fig. 9.1). Therefore, the produced retentate fractions, characterized by high antioxidant potential, can be considered of interest for nutraceutical applications. The recovery of anthocyanins from microfiltered ac¸ai juice by NF was investigated by Couto et al. (2011). Among the investigated membranes, the NF 270 exhibited the highest permeate flux value with anthocyanins retention higher than 99%. Flat-sheet NF membranes (MPT-34, from Koch Membrane Systems and Desal-5DK, from Osmonics) were used to concentrate pear juice (Warczoc et al., 2004). The experimental results showed that the Desal-5DK membrane produced a very high permeate flux, sufficiently high retention, and a higher

314

Separation of Functional Molecules in Food by Membrane Technology

NF PES10 N30F NF270

100

Rejection (%)

80

60

40

20

0 Naringin Neohesperidin Hesperidin

TAA

Sugars

FIGURE 9.1 Rejection of NF membranes toward analyzed compounds in the treatment of clarified bergamot juice by NF. NF, nanofiltration; TAA, total antioxidant activity. Adapted from Conidi, C., Cassano, A., 2015. Recovery of phenolic compounds from bergamot juice by nanofiltration membranes. Desalin. Water Treat., 56(13), 3510
3518.

concentration degree than the MPT-34 membrane. The decrease in permeate flux was significantly greater in juice solutions than in fructose solutions due to the complex composition of the juice. Rejection mechanisms of NF membranes for phenolic compounds were analyzed by Cai et al. (2017a) through a general quantitative structure activity relationship model. Results indicated that the phenolic retention is affected by operating conditions, membrane properties, and molecular parameters. In another work (Cai et al., 2017b) the authors found that fouling resistances and mechanisms were significantly different among several phenolic compounds due to molecular parameters such as acidity coefficient, molecular refractive index, octanol
water partition coefficient, and lipo
hydro partition coefficient. A cake/gel layer was found to be the main fouling resistance accounting for the reversible component of fouling, while the adsorption played a significant role in the irreversible fouling resistance.

9.4 WINE AND MUST PROCESSING Wine is one of the most popular drinks in the world, especially in Mediterranean countries such as Spain, France, or Italy, which are the most important producers in the world (about 46% of worldwide production)

Current and Future Applications of Nanofiltration Chapter | 9

315

(Barba et al., 2016). Wine is a complex alcoholic beverage with more than 800 organic compounds that contribute to the flavor and its specific aroma (Ortega-Heras et al., 2002). Therefore, any wine processing should preserve the original aroma minimizing the formation of undesired characteristics (i.e., plastic flavor). Membrane processes have replaced several conventional operations in winemaking process. In particular, cross-flow MF is largely used in the wine industry for wine clarification and stabilization as alternative to the use of fining agents. Other membrane processes, such as NF and RO, as well as emerging processes such as ED and membrane contactors, have been introduced for tartaric stabilization, reduction of alcohol content in wine, control of the must’s sugar content, management of dissolved gases, and reduction of volatile acidity (Daufin et al., 2001; El Rayess and MiettonPeuchot, 2016; Massot et al., 2008). In the last few years, the demand for low-alcohol beverages has risen in several countries as a result of health and social concerns. In addition, the alcoholic content has a strong impact on the quality of wine affecting the volatility of aroma compounds. Several strategies have been developed to partially reduce the alcohol content in wines without lowering the concentration of other compounds involved in wine quality, including viticultural or prefermentation practices, microbiological techniques (i.e., use of novel yeast strains), postfermentation practices (i.e., blending of high and low wine), cryoconcentration, and spinning cone column (SCC) (Schmidtke et al., 2012; Lo´pez et al., 2002). These technologies present different drawbacks including high energy, high pressure (such as RO), and high working temperature (i.e., SCC) that entail strong alteration of the aroma profile (Pilipovik and Riverol, 2005). In this contest, NF seems to be a more promising technology for obtaining low-alcohol wines. This process provides higher alcohol flow rates together with greater permeation rates than RO. In addition, it can be carried out at low pressures and temperatures thus preserving the sensory characteristics of the original product. Garcı´a-Martı´na et al. (2009) analyzed the retention characteristics of different NF and tight UF membranes toward glucose and fructose in synthetic mixtures, with the aim of controlling the sugar in grape musts. The final objective was to decrease the alcohol content of wine at traditional ranges accepted by the consumer without altering the specific organoleptic balance of the final wines. The selected membranes (UF-GH, NF-HL, and NF-DK, all from GE Osmonics) showed different rejections toward glucose and fructose (in the range 15%
38% and 80%
90%, for UF and NF membranes, respectively). A higher rejection of NF membranes toward polyphenols, anthocyanins and tartaric acid, was also observed. The authors proposed a two-step NF process aimed at producing wines with a low amount of alcohols but enriched with low and high molecular weight compounds (HMW). According to this process, the unfermented must is submitted to a first NF step producing a solution with a moderate content of sugars. HMW

316

Separation of Functional Molecules in Food by Membrane Technology

Unfermented must

NF1

Retentate UF permeate

NF2

Retentate (high sugar content) Permeate (low sugar content)

Fermentation

Low alcohol wine FIGURE 9.2 Two-stage NF process for the production of low alcohol wine. NF, nanofiltration; UF, ultrafiltration. Adapted from Garcı´a-Martı´na, N., Palacio, L., Pra´danos, P., Herna´ndez, A., Ortega-Heras, M., Pe´rez-Magarin˜o, S., et al., 2009. Evaluation of several ultra- and nanofiltration membranes for sugar control in winemaking. Desalination, 245(1
3), 554-558.

(i.e., polyphenols and anthocyanins) are retained by the membrane. The NF permeate is then forwarded to a second NF unit to concentrate all sugars in the retentate and at the same time to preserve LMW compounds (ions and tartaric acid) in the permeate. A low sugar content must, enriched in LMW and HMW compounds, is obtained by remixing the NF retentate and the NF permeate from the first and second step, respectively (Fig. 9.2). The NF retentate of the second step was considered as suitable for the production of sweet wines, liquors, or additives for functional foods thus increasing the profitability of the process. A significant decline in permeate flux appeared during the first NF step due to the presence of larger molecules. In addition, the contribution of the gel layer, the increasing viscosity, and the cake formation effects was much more significant in the case of red musts due to the presence of polyphenols (Garcı´a-Martı´na et al., 2010).

Current and Future Applications of Nanofiltration Chapter | 9

Un-withdrawn original juice

317

Grape juice

UF Low sugar content juice

UF retentate

UF permeate

NF permeate

NF

FIGURE 9.3 Production of low sugar content fruit juices by UF and NF membranes. NF, nanofiltration; UF, ultrafiltration. Adapted from Bonnet, J., De Vilmorin, H., 2004. Process for the controlled reduction of the sugar content of fruit juice and device for practicing this process. US Patent 0234658 A1.

Partial sugar reduction can also be obtained through a combination of UF and NF membranes according to a process patented by Bonnet and De Vilmorin (2004) and marketed under the name of REDUX. It consists of a first UF step producing a clear must with the same sugar concentration of the initial must. The UF retentate is reincorporated into the must that is being treated. The UF permeate is then concentrated by a selective NF membrane substantially impermeable to sugars, producing a permeate poor in sugars and relatively rich in acids and other qualitative constituents of the initial juice. The twice-filtered NF permeate is mixed with UF retentate and, if desired, with a portion of the untreated original juice to produce a juice with a low sugar content (Fig. 9.3). More recently, Salgado et al. (2015) evaluated the performance of both single-stage and two-stage NF processes to reduce the alcohol content of wines by controlling the sugar content in grape must before its fermentation. Grape must filtrations were performed in a pilot plant scale unit equipped with a spiral-wound membrane module (KMS SR3, from Koch Membrane Systems) by using two different grape must varieties, one white (Verdejo) and one red (Garnacha), respectively. Both processes allowed the reduction of sugar content in grape must without a significant alteration of important compounds such as polyphenols and malic and tartaric acids. This allows the production of wines with sensorial and chemical characteristics similar to those of wines obtained through the fermentation of untreated musts. Therefore, this technique could be applied at a larger scale for the production of low alcohol content wines. Catarino and Mendes (2011) tested several RO (CA995PE from Alfa Laval) and NF membranes (NF99 HF, NF99, NF97 from Alfa Laval and

318

Separation of Functional Molecules in Food by Membrane Technology

YMHLSP1905 from Osmonics) for removing ethanol in red wine from c. 12 to 7
8 vol.% operating in DF mode. All the selected NF membranes showed higher effectiveness in alcohol removal from wine, due to their good permeability to ethanol and high rejection toward aroma compounds, resulting in dealcoholized wine samples with promising organoleptic properties. Selected NF membranes were used to produce dealcoholized wines (up to c. 5 vol.%) that were blended with the original wine to produce reconstituted wine samples. In another work, Labanda et al. (2009) studied the reduction of alcoholic degree of a white model wine by using RO and NF membranes. Their analysis was directed basically to the permeation of aroma compounds during the concentration process. Among the investigated membranes the UB70 (from Toray) showed the highest permeability to model wine and the lowest solute rejection values. NF membranes can also be used to increase the sugar concentration of must, when grape musts do not have sufficient potential alcohol content, in order to have a well-balanced wine. Traditional techniques to increase the natural alcohol content of wine include the treatment of grape must with additives (i.e., addition of sugars or ethanol), must concentrate (MC), or rectified must concentrate (RMC). However, the addition of MC could affect the quality of wine, while the addition of RMC causes a dilution effect. NF membranes allow increasing the sugar contents in wine without addition of nongrape components at room temperature, preserving the heat-sensitive substances and volatile compounds without affecting the marked character of wine (Banvolgyi et al., 2006). Versari et al. (2003) tested two different NF membranes with the aim of increasing the sugar content of grape must used for wine production. In particular, NF membranes showed a higher rejection toward sugars (in the range of 77%
97%) and lower rejection toward malic acid (in the range 2%
14%). The grape must treated by NF showed the composition of the initial grape must, with minimal modification of malic and tartaric acids. This is critical to preserve the quality of the wines. A preservation of the total polyphenols content was also observed. The concentration of grape must with NF membranes has been investigated by Santos et al. (2008), too. Among the investigated membranes the NF270 membrane presented the highest permeate flux with the most efficient fractionation between sugars and organic acids (higher rejection coefficients of sugars than organic acids). The efficiency of different RO and NF membranes in must concentration from different grape varieties from southern Italy, was also assayed by Pati et al. (2014). Both membrane processes allowed to obtain a high-quality wine, in terms of total polyphenols, sugars, color intensity, and acidity when compared with wines produced by addition of RCM.

Current and Future Applications of Nanofiltration Chapter | 9

319

9.5 DAIRY INDUSTRY In the dairy industry, NF is mainly used for special applications such as partial demineralization of whey, lactose-free milk, or volume reduction of whey (Kumar et al., 2013). NF reached industrial scale and has been used worldwide by the dairy industry to desalt whey, mother liquors, and brines since 1980 (Pouliot, 2008). NF membranes due to their intermediate selectivity between UF and RO membranes represent an interesting alternative to ion exchange and ED if moderate demineralization is required. This technology demineralizes whey at the time of concentrating, and saves cost, time, and water (in terms of disposal). In another approach, NF has been applied for the removal of salt from salty whey (up to 84%), the partial removal of acid from acid whey (42%), and the concentration of whey protein (up to 20%
22% of dry matter) allowing simultaneously the reduction of minerals by 20%
50%. Other applications of NF membranes in the dairy industry include the concentration and demineralization of UF whey permeate containing the desired content of lactose, the concentration of milk in yogurt manufacture (as alternative to vacuum evaporation), and the selective demineralization of yogurt.

9.5.1 Concentration and Demineralization of Whey Whey is a byproduct of the cheese making and casein industry with a low content of solids (up to 5%
6%) and high biological oxygen demand (BOD5 5 30
50 g/L), which make its disposal difficult and costly (Prazeres et al., 2012). For many years, whey was considered a waste material and used as animal feed (postdrying) and partly as a fertilizer. However, over the past few decades there was an increased interest in the utilization of whey for the production of value-added products like whey proteins due to their nutritional, biological, and functional properties (Daufin et al., 2001; Ganju and Gogate, 2017). Processing of whey to yield valuable products helps to reduce environmental pollution and also provides the dairy industry with an added economic incentive due to the possible sale of these recovered or processed products. In the traditional process of whey demineralization, whey is concentrated by evaporation or RO followed by demineralization of the concentrated whey using ED and/or ion-exchange (Alkhatim et al., 1998; Chen et al., 2016). These processes are widely used on industrial scale but generate high investment and cleaning costs due to the high volume of effluent that needs to be treated in a wastewater treatment plant. As previously reported, the NF process allows concentration (up to 20%
22%) and partial demineralization of the whey in one step (between 25% and 60% or 90% with DF). Particularly, due to the selectivity of the membranes most of the monovalent

320

Separation of Functional Molecules in Food by Membrane Technology

ions, organic acids, and some of the lactose pass through the membrane (Daufin et al., 2001). Additional advantages offered by NF over conventional processes include the reduction of overall costs, lower energy consumption, and the production of a product with better taste, properties, and viscosity (Van der Horst et al., 1995). Moreover, the amount of effluent is greatly reduced if compared with other demineralization processes. Due to these advantages NF is increasingly replacing the traditional processes for concentration and desalination of whey and its potential in this field has been largely investigated. Magueijo et al. (2005) tested two different NF membranes (NFT50 from Alfa-Laval and HR-95-PP from DDS) in plate and frame configuration for the recovery of valuable components from second cheese whey, a byproduct of Serpa cheese and curd production. In particular, the NF process was used to fractionate cheese whey into a rich lactose fraction (retentate) and a process water with a high salt content (permeate). At an operating pressure of 30 bar the NFT750 membrane allowed to obtain a water recovery of approximately 80%, with a significant reduction of the wastewater organic load, and a fivefold concentration of valuable whey nutrients. In a previous work, Nguyen et al. (2003) investigated the use of NF to concentrate the solid content of cottage cheese whey. The pretreated whey (decanting, heating to 65 C, and filtering through a cheese bag of 5 μm) was treated through a NF unit fitted with plate-and-frame membranes (XP-45, from APV Company) for a total membrane surface area of 0.29 m2. A large pilot plant unit equipped with two spiral-wound membrane modules (Desal 5, from Osmonics) was also tested. The process allowed to concentrate cottage cheese whey (from 7 to 26 Brix on the large plant unit by increasing the operating pressure in three steps from 20 up to 34 bar) while removing about three-quarters of the sodium and potassium salts and some acid (Table 9.2). The retentate fraction, enriched in fat, proteins, and lactose, was considered of interest in the dairy industry for the production of ice cream and yogurt. Pan et al. (2011) studied the simultaneous concentration and demineralization of acidified whey through the use of a spiral-wound NF membrane (TFC 2540 SR2, from Koch membrane Systems) in aromatic polyamide. The desalination of whey was greatly affected by the pH of the feed solution. The best desalination rate was reached at the isoelectric point of whey (pH 5 4.60). An improvement of the demineralization rate, up to 72%, was obtained through the DF process.

9.5.2 Concentration and Demineralization of UF Whey Permeate NF is also successfully used for the demineralization of UF whey permeate streams for the production of lactose, minimization of volume prior to

Current and Future Applications of Nanofiltration Chapter | 9

321

TABLE 9.2 Composition of Cottage Cheese Before and After Nanofiltration (NF) Parameter

Feed

Permeate

Retentate

Total solids (%)

6.23

1.46

23.8

Ash (%)

0.70

0.76

1.55

Fat (%)

0.38

0.28

1.20

Protein (%)

0.66

0.14

2.41

Lactose (%)

4.80

0.29

19.7

Sodium (ppm)

609

548

510

Potassium (ppm)

1818

2044

1715

Calcium (ppm)

766

324

2410

Magnesium (ppm)

88

30

361

pH

4.54

4.40

4.28

Acidity (%)

0.55

0.46

1.47

Source: Adapted from Nguyen, M., Reynolds, N., Vigneswaran, S., 2003. By-product recovery from cottage cheese production by nanofiltration. J. Cleaner Production, 11(7), 803
807.

disposal, and production of water for other process streams. The benefits of using NF as opposed to evaporation is the reduction of concentration costs, the removal of ions that impact the crystallization step, reduction of postcrystallization lactose processing costs, and higher lactose yields (Bargeman et al., 2005). In proposed processes for lactose production, UF is firstly used to concentrate proteins for whey powder production, followed by a NF step for demineralization and lactose concentration, and a crystallization or spraydrying step for the production of lactose. Atra et al. (2005) evaluated the application of UF for milk and whey protein concentration, followed by the lactose concentration of the UF permeate by NF. Two flat-sheet UF membranes (FS10, PVDF, MWCO 6
8 kDa and SP015, polyethersulfone, MWCO 15
20 kDa, both from Zoltek Rt Mavibran) and a spiral-wound NF membrane (RA55, polyamide, MWCO 400 Da, from Millipore) were used. The UF of fresh milk and whey produced whey protein concentrates with a protein content of 12%
14% and 8%
10%, respectively, that can be used in the cheese production improving its nutritional value. The UF permeate, containing about 0.1%
0.5% of proteins and 5% of lactose, was submitted to the NF process. At an operating pressure of 20 bar a concentrated solution containing higher than 25% of lactose was obtained. By selecting proper operating parameters (operating

322

Separation of Functional Molecules in Food by Membrane Technology

temperature of 30 C and VRF 5) the lactose yield was higher than 90%. The NF retentate can be reused in the sweets industry, while the permeate stream, containing only 0.1%
0.3% of lactose, was considered suitable for cleaning, irrigation, or other purposes. Sua´rez et al. (2006) analyzed the performance of an aromatic polyamide spiral-wound membrane (DK 2540C, from Osmonics) for the partial demineralization of whey and milk UF permeate. Results showed a reduction of total salt content in the range 27%
36% depending on the operating conditions and VRF. As expected, a lower retention for monovalent ions was observed in comparison with that of divalent ions. Permeate fluxes were higher when treating milk UF permeate due to the lower amount of proteins, which can form a gel layer on the membrane surface. On the other hand, the salt rejection was lower when compared to that measured in the treatment of whey. The combination of concentration and DF modes can result in relatively high demineralization with minimal lactose losses. In the work investigated by Cuartas-Uribe et al. (2009) the concentration of ultrafiltered whey was studied by using a TFC NF spiral-wound membrane (DS-5 DL membrane, from GEOsmonics) according to three different operating configurations: recirculation, concentration, and continuous DF. Whey mineralization and lactose concentration were obtained through a combination of concentration and continuous DF. An operating pressure of 20 bar and a volume dilution factor (VDF) of 2 were estimated as the best operating conditions to minimize the loss of lactose into the permeate. In these conditions the removal of monovalent ions was about 80%. Working at higher VDF implied higher chloride removal from whey but at the same time an increasing of lactose loss. Rice et al. (2006) found that calcium phosphate is the main cause of flux decline of NF membranes in the treatment of dairy UF permeate, while whey protein and lactose play a lesser role. In addition, a modification of the pH in the range 5
9 had a significant influence on the rate of flux decline. In particular, an increase in feed pH to 8.3 produced a significant decline of permeate flux (less than 40% of the initial flux in the first hour of filtration), while at pH 5.5 the flux decline during the entire filtration resulted of 20%. At the natural pH of the UF permeate (6.9) fouling increased by increasing the operating temperature.

9.5.3 Recovery of Lactic Acid Lactic acid is a natural organic acid widely used in the food industry as acidulant and preservative as well as in the production of many pharmaceutical, cosmetic, and biotechnological products (Datta and Henry, 2006). It can be manufactured either by chemical synthesis or by carbohydrate fermentation. The conventional process for fermentative production of lactic acid is a batch process with low productivity and high capital and operating costs mainly

Current and Future Applications of Nanofiltration Chapter | 9

323

due to the further separation and purification steps that are necessary to achieve the quality grade requirements for food grade. Particularly, the traditional downstream separation and purification involves a series of different steps like filtration, acidification, neutralization, crystallization, carbon adsorption, evaporation, ion exchange, etc. These processes are time consuming, energy-intensive, and require expensive chemicals, which account for 50% of the production costs and the generation of waste, such as gypsum. New techniques have been used to recover lactic acid from fermentation broths including extraction, adsorption, and membrane separation (Bernardo et al., 2016). Membrane separation has been extensively used in the last few decades for lactic acid separation (Wang et al., 2013). For instance, NF in downstream purification can replace the multiple purification steps by a single step while yielding monomer grade lactic acid. A combined process of NF and RO was developed by Li et al. (2008) for the separation and concentration of lactic acid from cheese whey fermentation broth. The process involved a preliminary NF step to separate lactose and cells from lactic acid in the fermentation broth, followed by RO of the NF permeate (Fig. 9.4). Five different NF membranes (CK, DK, DL, GE, and HL, all from GE Osmonics) were studied for this purpose. Among the selected membranes, the HL membrane (a TFC membrane with an approximate MWCO of 400 Da) was the most efficient with almost 97% lactose retention (Table 9.3). For this membrane permeate fluxes were of the order of 33 L/m2h at 37 C and operating pressure of 21 bar. The NF permeate, containing mainly lactic acid and water, was then processed by using two different RO membranes (DS 11 AG and ADF, from GE Osmonics) to concentrate lactic acid. The ADF membrane showed a total rejection of lactic acid with a production of a permeate containing only water. Chandrapala et al. (2016) compared the performance of three flat-sheet NF membranes in TFC (DK, DL, and HL, from GE Osmonics) in the processing of cream cheese acid whey at different pH values (3, 4, 5, or 7.3) and two different processing temperatures (25 C and 40 C). The separation efficiency of NF membranes for lactose, lactic acid, proteins, and minerals, was also investigated working in a DF mode. Particularly, the aim was to assess the removal of lactic acid and minerals with respect to proteins and lactose. Among the investigated membranes the HL membrane exhibited higher permeate fluxes in both NF and DF processes. At a temperature of 40 C and pH 3, this membrane showed the highest permeate transmission for lactic acid (about 50%) and a lactose retention of 93%. The obtained results indicated that NF is a promising process in the valorization of acid whey allowing the production of spray-dried whey powders that can be used in different food ingredient applications. This is accompanied by the reduction of the waste disposal costs and environmental impact currently associated with acid whey production.

Cheese whey

Whey protein

UF Fermentation

Lactose, cells) NF

Permeate (2%-10% lactic acid)

RO

Retentate (5%-30% lactic acid) Vacuum evaporation

Water

Lactic acid 90%-98% Polymerization

Polylactide (PLA) FIGURE 9.4 Flow diagram for lactic acid production and separation from cheese whey. NF, nanofiltration; RO, reverse osmosis; UF, ultrafiltration. Adapted from Li, Y., Shahbazi, A., Williams, K., Wan, C., 2008. Separate and concentrate lactic acid using combination of nanofiltration and reverse osmosis membranes. Appl. Biochem. Biotechnol., 147(1
3), 1-9.

TABLE 9.3 Rejection of NF Membranes Toward Lactose and Lactic Acid in the Treatment of Cheese Whey Type of Membrane

MWCO (Da) Lactose

Lactic Acid

CK

200

19.0

18.0

DK

150
300

88.2

38.0

GE

500

40.0

21.0

DL

150
300

69.8

26.1

HL

150
300

96.8

43.0

MWCO, molecular weight cut-off.

Rejection (%)

Current and Future Applications of Nanofiltration Chapter | 9

325

A semiindustrial NF plant fitted with two spiral-wound membranes (NF245-3840/30FF with a MWCO of 200 Da, from Dow-Filmtec) to demineralize and concentrate lactic acid whey prior to vacuum evaporation and spray drying has been tested recently by Beda´s et al. (2017). The NF process improved the dryability of the NF concentrate and the hygroscopicity of the resulting lactic acid whey powder without affecting the particle size distribution. Moreover, NF reduced the overall industrial energy cost of the lactic acid powder production process in comparison with standard processes. These results can contribute significantly to improve the application of NF in the processing of these byproducts and the production of lactic acid powders at industrial scale.

9.6 SUGAR INDUSTRY In the sugar industry, concentration and purification processes play an important role in the production of sugars. Traditional processes employed in this sector include evaporation, crystallization, and ion-exchange chromatography. These processes are energy intensive resulting in high operating costs, environmental problems, and low yield products for large-scale applications. Increasing fuel prices and stringent environmental regulations are pushing sugar producers to find alternative energy efficient and environmentally friendly processes (Gul and Harasek, 2012). Membrane technology can play an important role in the sugar industry by supporting the development of sustainable processes. Among membrane processes, NF is a promising industrial-scale method for sugars purification and concentration from natural carbohydrates mixtures, offering considerable economic advantages over conventional processes, such as chromatography and vacuum distillation, since it requires lower energy consumption (Feng et al., 2009). In spite of the relatively low differences among the molar mass of monosaccharides and di-/trisaccharides, separation by NF occurs with interesting yields, for example, of glucose from sucrose, as well as for glucose from lactose, or from raffinose (Pontalier et al., 1997). Different studies report the suitability of NF membranes for the separation of xylose from glucose that is important in the commercial purification of xylose for xylitol production. Sjo¨man et al. (2007) compared three different NF membranes (Desal-5 DK and Desal-5 DL, from GE Osmonics, and NF 270 from Dow-Filmtec) with MWCO in the range 150
300 Da in the treatment of solutions made of xylose and glucose in different mass ratios and total monosaccharide concentration, in order to separate both sugars. The selected membranes gave comparable fractionation results. Sieving was considered as the main mechanism involved in the separation of these uncharged molecules. However, changes in membrane structure (i.e., swelling and compaction) due to operating pressure and temperature can also play an important role in the separation. The

326

Separation of Functional Molecules in Food by Membrane Technology

retention of monosaccharides was strongly affected by the permeate flux and, consequently, by the operating pressure. Xylose retentions resulted lower than glucose retentions and the differences between these rejections decreased by increasing the permeate flux. All selected membranes provided a xylose separation factor over 2 and the most favorable xylose separation from glucose was achieved with a concentrated monosaccharide solution at rather high pressure. According to the experimental results, the NF process allows to enhance the yield and partially replace chromatographic methods in xylose production. NF have been successfully used in the concentration of dextrose syrup produced from starch as alternative process to traditional gelatinization, dextrinization, and enzymatic conversion of the dextrin process. According to the process patented by Binder et al. (1999) NF is used to treat a glucose syrup containing 95% dextrose and 5% di-trisaccharides to produce inexpensive high purity dextrose. The process was studied at pilot and industrial scale, by using different spiral-wound NF membranes such as NF-70 and NF-40 (from Dow-Filmtec), Desal 5-DL (from GE Osmonics), and FE 700-0022 (from Filtration Engineering). During the NF process di- and tri-saccharides are retained by the NF membrane, while a solution with a dextrose purity higher than 99% is obtained as permeate. The recovery of dextrose from crystallization mother liquors (a byproduct coming from dextrose manufacturing) with commercial NF membranes has been recently investigated by Bandini and Nataloni (2015). Four spiral-wound NF membranes (Desal-DK and Desal-DL, from GE Osmonics, and K-SR2 and K-MPS34 from Koch Membrane Systems) were tested to obtain, at industrial scale, a purity of dextrose higher than 95%. The selected membranes showed different separation efficiencies; only the Desal-DL membrane fitted the industrial requirements for process application. For this membrane dextrose purities in the permeate stream higher than 97% were achieved in optimized operating conditions of temperature (50 C), pressure (30 bar), and feed flow rates (from 2300 to 3600 dm3/h). NF is also attractive and competitive with traditional technologies, for the purification of disaccharides or monosaccharides from disaccharides and other mixtures of carbohydrates (Catarino et al., 2008). For example, Xu et al. (2005) employed NF membranes to purify and condense maltitol from maltitol syrup containing sorbitol, maltitol, multisugar alcohol, and others. NF is also an effective method for the purification of xylooligosaccharides (XOs), characterized by specific health benefits, from impurities of raw XOs syrups including monosaccharides (mainly xylose and arabinose), salts, and organic acids, which can cause off-flavors and raise safety risks (Mellal et al., 2008). A theoretical model based on the extended Nernst
Planck equation and film theory was developed by Hua et al. (2010) to evaluate the retention for XOs syrup in a NF process performed using a TFC spiralwound membrane module (HDS-12-2540, from Hyflux Corporation). It was

Current and Future Applications of Nanofiltration Chapter | 9

327

found that the reflection coefficient (σ) for arabinose (0.895) was higher than that for xylose (0.776), which indicated a stronger retention of the membrane for arabinose. Patil et al. (2015) studied the separation of mono- and disaccharides from a mixture of inulin using an experimental pilot-scale NF cascade system. Two different spiral-wound NF membranes (GE-1812C-34D and GH-1812C34D, from GE Osmonics) with MWCO of 1000 and 2500 Da, were used and tested according two different cascade configurations: a countercurrent product recycle cascade (configuration 1) and an adapted cascade (configuration 2). The cascade systems showed great potential for continuous enrichment of inulin over conventional concentration and DF with significant advantages in terms of minimum solvent consumption and product loss as well as higher productivity without compromising the purity. Both configurations allowed to obtain a yield of inulin of about 99%
99.7%, with a purity in the range 93.8%
96%. Moreno-Vilet et al. (2014) investigated the potential of NF in the purification of inulin-type fructans from model solutions containing LMW sugars like glucose, fructose, and sucrose. A pilot-scale NF unit equipped with a spiral-wound membrane (HPA-600, MWCO 600 Da, hydrophilized polyamide material, from Permionics) was used for cross-flow experiments. The effect of TMP, feed concentration, and retentate flux over the inulin-type fructan separation was estimated by employing the design of experiments methodology by using single, binary, and multicomponent carbohydrate model solutions. The selected membrane produced a good separation of inulin, while eliminating LMW carbohydrates. Response surface response analysis (P , .05) revealed inulin rejection values over 90%, for high pressures (14 bar) and low feed concentrations (5 g/100 mL). The food industry has recently shown increased interest in oligosaccharides such as fructooligosaccharides (FOS) and galactooligosaccharides (GOS) due to their prebiotic properties beneficial to the human health. NF appears to be an interesting approach for the purification and concentration of oligosaccharide mixtures as alternative to chromatographic techniques (Lo´pez Leiva and Guzman, 1995). Goulas et al. (2002) evaluated the potential of NF in the purification of a commercial galactooligosaccharide mixture by using different flat-sheet membranes. Continuous DF purification by using a NF-CA-50 membrane (from Intersep Ltd) at 25 C and a DS-5-DL membrane (from GE Osmonics) at 60 C, gave yields of 14%
18% for monosaccharides, 59%
89% for disaccharides, and 81%
98% for oligosaccharides, respectively. In another work, Goulas et al. (2003) used two NF membranes NF-CA50 and NF-TCF-50 (from Intersep Ltd) in cellulose acetate and TFC, respectively, to fractionate commercial oligosaccharide mixtures working in a DF mode. According to the experimental results, the NF TFC-50 produced the best purification of a commercial mixture of oligosaccharides with yields of

328

Separation of Functional Molecules in Food by Membrane Technology

19% for monosaccharides and 88% for di- and oligosaccharides, after four DF steps, indicating that removal of monosaccharides is possible with only slight loss of di- and oligosaccharides. Two different spiral-wound NF membranes (GH-NF and GK-NF) were tested by Li et al. (2004) for the purification and concentration of FOS under constant volume diafiltration (CVD) and variable volume diafiltration (VVD) mode. Both modes allowed to obtain a FOS syrup purity above 90%. Feng et al. (2009) selected a spiral-wound NF membrane (NF-3, supplied by Sepro Co.) with a MWCO of 800
1000 Da for the separation of a commercial galactooligosaccharide mixture with low content of oligosaccharides in CVD process at 50 C and operating pressure of 6 bar. This membrane produced higher permeate flux and best results in terms of rejections between macromolecular and micromolecular sugars when compared with other NF membranes with removal of monosaccharides and lactose of 90.5% and 52.5%, respectively, and oligosaccharides purity of 54.5%. Similar results were recently obtained by Pruksasri et al. (2015) in the purification of GOS mixtures by using a flat-sheet NF membrane (NP030, from Microdyn-Nadir). In this case, a product purity of 85% (based on monosaccharide content) and a oligosaccharide recovery yield of 82% were achieved. Working in DF mode, a higher improved product purity (of about 90%) was obtained. More recently, Co´rdova et al. (2016) evaluated the potential of NF flat-sheet membranes of different membrane material (TFC and polyethersulfone) and MWCO in the purification of GOS in solutions at high concentrations. According to the results polyethersulfone membranes (NP010 and NP030, both from Microdyn-Nadir) showed low rejection values for monosaccharides and disaccharides at TMP values lower than 25 bar. In particular, the NP010 membrane allowed the fractionation of GOS with sustainable flux values (28 kg/m2/h) and a satisfactory selectivity working at an operating pressure of 20 bar and with highly concentrated solutions (40 Brix), demonstrating the possibility to use NF for treating highly concentrated GOS solutions. A purification process of GOS, using threestage serial NF units, operating under critical TMP was also designed (Fig. 9.5) by Co´rdova et al. (2017). Despite using concentrated GOS solutions, the membrane permeability for NF1 and NF2 steps remained virtually unchanged. This phenomenon was attributed to the reduction of critical TMP with time by using the proposed model. GOS purity was of the order of 55%, quite similar to that of some commercial products currently purified by conventional processes. Table 9.4 presents an outline of NF applications in selected areas of food processing with detailed information on the aims of the treatment, membrane typology, and operating conditions.

Current and Future Applications of Nanofiltration Chapter | 9

329

Hydrolized raw GOS

UF

Water

Water

UF permeate (20°Brix)

NF1

NF2

NF3

Mixture of permeates

Purified GOS FIGURE 9.5 Three-stage serial NF array for purification of GOS solutions (NF1: membrane NP010, PES, 1000 Da, 53.5 C; NF2: membrane NP010, PES, 1000 Da, 65 C; NF3: membrane GE, PA, 1000 Da, 65 C). GOS, galactooligosaccharides; NF, nanofiltration; UF, ultrafiltration. Adapted from Co´rdova, A., Astudillo, C., Santiban˜ez, L., Cassano, A., Ruby-Figueroa, A., & Illanes, A. (2017). Purification of galacto-oligosaccharides (GOS) by three-stage serial nanofiltration units under critical transmembrane pressure conditions. Chem. Eng. Res. Design, 117, 488-499.

9.7 RECOVERY OF FUNCTIONAL COMPOUNDS FROM FOOD PROCESSING BYPRODUCTS Natural antioxidants, including polyphenols, carotenoids, tocopherols, and ascorbic acid as well as proteins and dietary fibers are well known for their nutritional and functional properties (Boskou, 2006; Elleuch et al., 2011; Kristinsson and Rasco, 2000). The recovery of these compounds from undervalued bioresources is a major current research challenge, in the light of their beneficial properties on human health associated to the reduction of environmental pollution and disposal costs of agro-food wastes (Schieber et al., 2001). According to the “5-Stages Universal Recovery Processing” strategy proposed by Galanakis (2012), the recovery of high-added value components from food wastes follows the principles of analytical chemistry and can be accomplished through processing steps that include (1) a macroscopic

TABLE 9.4 Application of NF in Food Processing Products and Byproducts Source

Objective

Type of Membrane

Operating Conditions

References

Bergamot juice

Polyphenols concentration and purification

Inopor nano (Inopor), monotubular, TiO2, 750 Da

TMP, 7.5 bar; Qf, 97.6 L/h; T, 24 C

Conidi et al. (2011)

Inopor nano (Inopor), monotubular, TiO2, 450 Da

TMP, 33 bar; Qf, 100 L/h; T, 24 C

NF-270 (Dow-Filmtec), spiral-wound, PA, 200
300 Da

TMP, 6 bar; T, 20 C

NF PES 10 (Microdyn Nadir), spiral-wound, PES, 1000 Da

TMP, 6 bar; T, 20 C

N30F (Microdyn Nadir), spiral-wound, PES, 400 Da

TMP, 6 bar; T, 20 C

NF270 (Dow-Filmtec), flat-sheet, polypiperazine TFC, 150
300 Da

TMP, 5
30 bar; v, 0.3 m/s; T, 30 C

UTC60 (Toray), flat-sheet, TFC

TMP, 5
30 bar; v, 0.3 m/s; T, 30 C

MPF36 (Koch Membrane Systems), flat-sheet, TFC, 1000 Da

TMP, 5
30 bar; v, 0.3 m/s; T, 30 C

DL (GE Osmonics), flat-sheet, PA, 150
300 Da

TMP, 5
30 bar; v, 0.3 m/s; T, 30 C;

DK (GE Osmonics), flat-sheet, PA, 150
300 Da

TMP, 5
30 bar; v, 0.3 m/s; T, 30 C

NP010 (Microdyn-Nadir), flat-sheet, PES, 1000 Da

TMP, 5
30 bar; v, 0.3 m/s; T, 30 C

NP030 (Microdyn-Nadir), flat-sheet, PES, 400 Da

TMP, 5
30 bar; v, 0.3 m/s; T, 30 C

Bergamot juice

Blackberry juice

Polyphenols concentration and purification

Polyphenols concentration

Conidi and Cassano (2015)

Acosta et al. (2017)

Pomegranate juice

Polyphenols separation and purification

Selro MPF 36 (KochMembrane), flat-sheet, TFC, 1000 Da

TMP, 6 bar; T, 25 6 1 C

Conidi et al. (2017a,b)

Strawberry juice

Polyphenols concentration

NF membrane (GE Osmonics), PVDF, 150
300 Da

TMP, 6 bar; T, 20 6 2 C

Arend et al. (2017)

Watermelon juice

Bioactive compounds concentration

L2521 TF (GE Osmonics), spiral-wound, PVDF, 150
300 Da

TMP, 6 bar; v, 1 m/s; T, 25 6 2 C

Arriola et al. (2014)

Red and white wine

Alcohol reduction

KMS SR3 (Koch Membrane Systems), spiralwound, TFC PA, 200 Da

TMP, 33 bar; Qf, 540 L/h; T, 25 6 2 C

Salgado et al. (2015)

Red and white must

Sugar reduction

HL (GE Osmonics), spiral-wound, TFC

TMP, 24 bar; Qf, 60 L/h; T, 3 C (white must)

Garcı´a-Martı´na et al. (2010)

TMP, 24 bar; Qf, 90 L/h; T, 6 C (red must) Red wine

Grape white must

Alcohol removal

Sugar content increasing

NF99 HF (Alfa Laval), flat-sheet, TFC, 200 Da

TMP, 16 bar; Qf, 120 L/h; T, 30 C

NF99 (Alfa Laval), flat-sheet, TFC, 200 Da

TMP, 16 bar; Qf, 120 L/h; T, 30 C

NF97(Alfa Laval), flat-sheet, TFC, 200 Da

TMP, 16 bar; Qf, 120 L/h; T, 30 C

YMHLSP1905 (GE Osmonics), flat-sheet

TMP, 16 bar; Qf, 120 L/h; T, 30 C

DS (GE Osmonics), spiral-wound, PA

TMP, 40 bar; Qf, 1900 L/h; T, 15 C

DK (GE Osmonics), spiral-wound, PA

TMP, 40 bar; Qf, 1900 L/h; T, 15 C

Catarino and Mendes (2011)

Versari et al. (2003)

(Continued )

TABLE 9.4 (Continued) Source

Objective

Type of Membrane

Operating Conditions

References

Wine (Verdeca and Bombino nero varieties)

Wine quality improving

DK (GE Osmonics), spiral-wound, PA, salt rejection 98%

TMP, 39 bar

Pati et al. (2014)

VinoPro (GE Osmonics), spiral-wound, salt rejection 98%

TMP, 39 bar

TFC-S4v (Koch Membrane Systems), spiralwound, salt rejection 99%

TMP, 39 bar

NF270 (Dow-Filmtec), spiral-wound, polypiperazine TFC, 150
300 Da

TMP, 39 bar

Milk and whey

Milk and whey protein concentration

RA55 (Millipore), spiral-wound, PA, 400 Da

TMP, 10
20 bar; Qf, 100
200 L/h; T, 30
50 C

Atra et al. (2005)

Whey

Separation and concentration of lactic acid

DK (GE Osmonics), flat-sheet, TFC, 150
300 Da

TMP, 14, 21, 28 bar; T, 37 C

Li et al. (2008)

DL (GE Osmonics), flat-sheet, TFC, 150
300 Da

TMP, 14, 21, 28 bar; T, 37 C

HL (GE Osmonics), flat-sheet, TFC, 150
300 Da

TMP, 14, 21, 28 bar; T, 37 C

CK (GE Osmonics), flat-sheet, CA, 200 Da

TMP, 14, 21, 28 bar; T, 37 C

GE (Ge-Osmonics), flat-sheet, PA, 500 Da

TMP, 14, 21, 28 bar; T, 37 C

Whey

Separation of lactose

DS-5 DL (GE Osmonics), spiral-wound, TFC, 150
300 Da

TMP, 5
25 bar; T, 16
18 C

Cuartas-Uribe et al. (2009)

Whey

Removal of lactic acid

Desal DL (GE Osmonics), flat-sheet, TFC, 150
300 Da

TMP, 21 bar; Qf, 174 L/h; T, 25, 40 C

Chandrapala et al. (2016)

Desal HL (GE Osmonics), flat-sheet, TFC, 150
300 Da

TMP, 21 bar; Qf, 174 L/h ; T, 25, 40 C

Desal DK (GE Osmonics), flat-sheet, TFC, 150
300 Da

TMP, 21 bar; Qf, 174 L/h ; T, 25, 40 C

Desal 5- DK (GE Osmonics), flat-sheet, TFC, 150
300 Da

TMP, 2
40 bar; v, 0.7
1 m/s; T, 50 C

Desal 5- DL (GE Osmonics), flat-sheet, TFC, 150
300 Da

TMP, 2
40 bar; v, 0.7
1 m/s; T, 50 C

NF270 (Dow-Filmtec), flat-sheet, polypiperazine TFC, 150
300 Da

TMP, 2
40 bar; v, 0.7
1 m/s; T, 50 C

GE (GE Osmonics), spiral-wound, TFC, 1000 Da

TMP, 0
24 bar; T, 50 C

GH (GE Osmonics), spiral-wound, TFC, 2500 Da

TMP, 0
24 bar; T, 50 C

HPA-600 (Permionics), spiral-wound, PA, 600 Da

TMP, 14 bar

Model solutions of sugars

Solution of inulin of different polymer sizes

Model carbohydrates solution

Separation of xylose from glucose

Separation of mono- and disaccharides from inulin

Separation of sugars

Sjo¨man et al. (2007)

Patil et al. (2015)

Moreno-Vilet et al. (2014) (Continued )

TABLE 9.4 (Continued) Source

Objective

Type of Membrane

Operating Conditions

References

Model sugar solution

Purification of oligosaccharide

NF-CA-50 (Intersep Ltd), flat-sheet, CA, 50% NaCl rejection

TMP, 6.9
27.6 bar; T, 25 C

Goulas et al. (2002)

DS-5-DL, (Osmonics Desal) flat-sheet, TFC, 96% MgSO4 rejection

TMP, 6.9
27.6 bar; T, 25
60 C

DS-51-HL (Osmonics Desal), flat-sheet, TFC, 96% MgSO4 rejection

TMP, 6.9
27.6 bar; T, 25 C

NF-CA-50 (Intersep Ltd), flat-sheet, CA, 50 % NaCl rejection

TMP, 40 bar

NFTFC-50 (Intersep Ltd), flat-sheet, TFC, 50% NaCl rejection

TMP, 40 bar

NF-2 (Sepro Membranes Inc.), spiral-wound, CA, 500
600 Da

TMP, 2
8 bar; T, 25 C

NF-3 (Sepro Membranes Inc.), spiral-wound, CA, 800
1000 Da

TMP, 2
8 bar; T, 25
50 C

NF-1812-50 (Dow-Filmtec), spiral-wound, PA, 150
300

TMP, 2
8 bar; T, 25 C

HBRO-1812-2 (Hebei R.O. Environment Tech. Co.), spiral-wound, CA, 800-1000 Da

TMP, 2
8 bar; T, 25 C

NP030 (Microdyn-Nadir), flat-sheet, PES, 400 Da

TMP, 25, 35, 45 bar; T, 5, 25, 60 C

Desal-5-DL (GE Osmonics), flat-sheet, PA, 150
300 Da

TMP, 25, 35, 45 bar; T, 5, 25, 60 C

Model sugar solution

Sugar solutions

Sugar solutions

Fractionation oligosaccharide

Separation of galactooligosaccharides mixture

Fractionation of galactooligosaccharides

Goulas et al. (2003)

Feng et al. (2009)

Pruksasri et al. (2015)

Model solution of sugars

Olive mill wastewaters

Artichoke extracts

Artichoke brines

Purification of oligosaccharides at high concentrations

GE (Desalogics), flat-sheet, PA, 1000 Da

TMP, 5
25 bar; T, 53.5 C 

NP030 (Microdyn-Nadir), flat-sheet, PES, 400 Da

TMP, 5
40 bar; T, 53.5 C

NP010 (Microdyn-Nadir), flat-sheet, PES, 1000 Da

TMP, 5
40 bar; T, 53.5 C

NFA (Parker), flat-sheet, TFC, 500 Da

TMP, 5
30 bar; T, 53.5 C

ATF (Parker), flat-sheet, TFC, 200 Da

TMP, 5
30 bar; T, 53.5 C

Purification and concentration of phenolic compounds

ETNA 01PP (Alfa Laval), flat-sheet, composite fluoropolymer, 1000 Da

TMP, 5, 9 bar; Qf, 600 L/h T, 30 C

NF90 (Dow-Filmtec), spiral-wound, TFC, salt rejection .97%

TMP, 12 bar; T, 22 C

Purification and concentration of phenolic compounds

NP030 (Microdyn Nadir), spiral-wound, PES, 400 Da

TMP, 4 bar; T, 25 C

Desal DK2540 (GE Osmonics), spiral-wound, PA, 150
300 Da

TMP, 10 bar; T, 25 C

Purification and concentration of phenolic compounds

NP010 (Microdyn Nadir), spiral-wound, PES, 1000 Da

TMP, 6 bar; T, 20 6 2 C

NP030 (Microdyn Nadir), spiral-wound, PES, 400 Da

TMP, 6 bar; T, 20 6 2 C

NF200 (Dow-Filmtec), spiral-wound, semiaromatic piperazineamide, 300
400 Da

TMP, 6 bar; T, 20 6 2 C

Desal DL (GE Osmonics), spiral-wound, PA, 150
300 Da

TMP, 6 bar; T, 20 6 2 C

Desal DK (GE Osmonics), spiral-wound, PA, 150
300 Da

TMP, 4, 6, 8, 12 bar; T, 15, 20, 21, 30 6 2 C

Co´rdova et al. (2016)

Cassano et al. (2013)

Cassano et al. (2015)

Cassano et al. (2016)

(Continued )

TABLE 9.4 (Continued) Source

Objective

Type of Membrane

Operating Conditions

References

Artichoke extracts

Concentration of oligosaccharides

NP010 (Microdyn Nadir), flat-sheet, PES, 1000 Da

TMP, 10, 15, 20 bar; T, 25 6 2 C

Machado et al. (2016)

NP030 (Microdyn Nadir), flat-sheet, PES, 400 Da

TMP, 10, 15, 20 bar; T, 25 6 2 C

NF270 (Dow-Filmtec), flat-sheet, polypiperazine TFC, 150
300 Da

TMP, 10, 15, 20 bar; T, 25 6 2 C

NF-70 (Dow-Filmtec), spiral-wound, PA/PS, 180 Da

TMP, 20 bar; T, 20 C

NF-200 (Dow-Filmtec), spiral-wound, semiaromatic piperazineamide, 300
400 Da

TMP, 20 bar; T, 20 C

N30F (Microdyn Nadir), spiral-wound, PES, 400 Da

TMP, 6 bar; T, 20 C

NFPES10 (Microdyn Nadir), spiral-wound, PES, 1000 Da

TMP, 6 bar; T, 20 C

Orange press liquor

Purification and concentration of phenolic compounds

Conidi et al. (2012)

Defatted milled grape seeds

Fractionation of phenolic compounds

AFC40 (PCI Membrane Systems), tubular, PA, CaCl2 rejection 60%

TMP, 6 bar; T, 16 6 2 C

Santamarı´a et al. (2002)

Racking wine lees

Fractionation of polyphenols and polysaccharides

ETNA 01PP (Alfa Laval), flat-sheet, composite fluoropolymer, 1000 Da

TMP, 3, 5, 7, 10 bar; Qf, 150 L/h

Giacobbo et al. (2017)

NF270 (Dow-Filmtec), flat-sheet, polypiperazine TFC, 150
300 Da

TMP, 3, 5, 7, 10 bar; Qf, 150 L/h

Distilled fermented grape pomace

Concentration of phenolic compounds

Nanomax 95 (Millipore)

TMP, 2, 4, 6, 8 bar

PA/PA, 250 Da Nanomax 50 (Millipore)

Dı´az-Reinoso et al. (2009)

TMP, 2, 4, 6, 8 bar

PA/PA, 350 Da DL2540 (GE Osmonics)

TMP, 2, 4, 6, 8 bar

spiral-wound, TFC, 150
300 Da GE2540 (GE Osmonics)

TMP, 2, 4, 6, 8 bar

spiral-wound, TFC, 1000 Da Inside Ce´ram (Tani)

TMP, 2, 4, 6, 8 bar

tubular, Titania, 1000 Da White vinasses

Concentration of phenolic compounds

(Iberlacht) spiral-wound, PES, 200 Da

TMP, 8 bar; T, 23 6 4 C

Diaz-Reinoso et al. (2017)

CA, cellulose acetate; NF, nanofiltration; PA, polyamide; PES, polyethersulfone; PS, polysulfone; PVDF, polyvinylidene fluoride; Qf, feed flow rate; T, temperature; TFC, thin-film composite; TMP, transmembrane pressure; v, cross flow velocity.

338

Separation of Functional Molecules in Food by Membrane Technology

pre-treatment, (2) separation of macro- and micromolecules, (3) extraction, (4) isolation purification, and (5) product formation or encapsulation. Pressure-driven membrane operations represent well-established technologies covering most of this strategy since they can be used for macroscopic pretreatment (MF), macro- and micromolecules separation (UF), purification (NF), and concentration (RO). UF and NF membranes, also through the use of membranes with different MWCO in sequential design, offer the possibility of obtaining different fractions enriched in diverse target compounds, for specific applications and markets. At the same time, they allow to control the removal of undesirable products (i.e., exclusion of heavy metals is assured due to the ability of NF membranes to reject di- and multivalent ions) as well as to obtain fractions of interest for other applications (i.e., purified water or fractions enriched in sugars, etc.). Tailor made processes for specific byproducts can be identified through an appropriated selection of membrane typology as well as optimization of operating and fluid-dynamic conditions (e.g., pressure, temperature, feed flow rate, VRF, etc.) which are key parameters influencing both productivity and selectivity toward target compounds (Cassano et al., 2018). The fractionation of olive mill wastewaters (OMWs) by using UF and NF membranes in sequential design allows to produce a concentrated fraction enriched in phenolic compounds suitable for cosmetic, food, and pharmaceutical industries as liquid, frozen, dried, or lyophilized formulations (Cassano et al., 2013). In particular, a two-step UF process (a first UF step with a PVDF membrane of 0.02 μm in hollow fiber configuration followed by a treatment with a flat-sheet polymeric membrane with a MWCO of 1000 Da) was used to remove suspended solids and total organic carbon from raw and preultrafiltered wastewaters, respectively. Polyphenols were mainly recovered in the UF permeate of the second UF treatment (the rejection of UF membranes toward polyphenols were of 26.7% and 31.8%, respectively). UF membranes exhibited lower rejections toward LMW phenolic compounds in comparison with values observed for total polyphenols (2.1% and 17.6%, respectively). The final NF step, performed with a spiral-wound membrane (NF90, from Dow-Filmtec), produced a concentrated fraction containing more than 85 mg/L of low MW polyphenols. The membrane rejection toward these compounds was of 100%, in agreement with the estimated MWCO of this membrane (90 Da) and the MW of the investigated phenols (in the range 138
284 g/mol). The NF permeate, depleted in phenolic compounds, was considered suitable for reuse as process water or as membrane cleaning solution. Integrated membrane processes based on the use of NF membranes for recovering phenolic compounds from OMWs have been also proposed by Russo (2007), Paraskeva et al. (2007), Garcia-Castello et al. (2010), and Bazzarelli et al. (2016).

Current and Future Applications of Nanofiltration Chapter | 9

339

Concentrated NF fractions containing 38.84 g/L of polyphenols were obtained by Romani et al. (2017), who investigated a sustainable extractive technology followed by membrane separation methods to obtain standardized commercial extracts for application in the functional food industry, pharmaceutical, and cosmetic fields from Olea europaea L. leaves and pitted olive pulp. The treatment was based on a water extraction of the vegetal material followed by MF, NF, RO, and final concentration by evaporation at low temperature or spray-dried technique. Artichoke byproducts, such as leaves, external bracts, and stems represent about 80% of the plant biomass and are a promising and cheap source of health-promoting phenolic compounds and inulin (Lattanzio et al., 2009). A two-step NF process for the production of phenolic-rich fractions from artichoke aqueous extracts was investigated by Cassano et al. (2015). In this approach, a clarified artichoke extract was processed through a first NF spiral-wound membrane (NP030, MWCO 400 Da, from Microdyn Nadir) at a TMP of 4 bar and a temperature of 25 C up to a VRF of 10, producing a retentate enriched in phenolic compounds (rejections of apigenin, cynarin, and chlorogenic acid were higher than 85%) and a permeate containing mainly sugars (rejections of fructose, glucose, and sucrose were lower than 4%). The content of phenolic compounds in the NF retentate increased by increasing the VRF of the process. Accordingly, the total antioxidant activity of the retentate stream was increased by nine times in comparison with the clarified extract. A permeate stream free of phenolic and sugar compounds was produced by processing the NF permeate through a NF spiral-wound membrane with a MWCO of 150
300 Da (Desal DK, from GE Osmonics). NF membranes with different polymeric material (polyethersulfone, polyamide) and MWCO (from 200 to 1000 Da) were also tested to recover and concentrate phenolic compounds from artichoke brines (Cassano et al., 2016) after a preliminary UF step. NF membranes with MWCO of 150
300 Da (NF200 and Desal DL, from Dow-Filmtec and GE Osmonics, respectively) produced better separation of salt compounds from caffeoylquinic acid derivatives with low rejection values toward dry residue (between 14% and 18%) and higher than 92% toward total caffeoylquinic acids, flavonoids, and cynarin. The concentrated stream was considered a good source of possible ingredients to functionalize foodstuffs. The conceptual design proposed for the recovery of phenolic compounds from artichoke brines is illustrated in Fig. 9.6. Similarly, polyamide NF membranes with MWCO of 150
300 Da (such as NF270 from Dow-Filmtec) showed 100% of retention of prebiotic sugars in the treatment of microfiltered artichoke extract (Machado et al., 2016). Conidi et al. (2012) investigated the use of different NF membranes for the separation and concentration of bioactive compounds from press liquors, a byproduct of citrus juice production. Among the investigated membranes, the NFPES10 membrane (a polyethersulfone membrane with a MWCO of

340

Separation of Functional Molecules in Food by Membrane Technology

Reuse in the blanching step

Permeate (salt solution)

UF Artichoke brines

prefiltration

NF Permeate

Retentate (suspended solids, high molecular weight substances)

Retentate (caffeoylquinic acids, flavonoids, etc.)

Formulations for food or nutraceutical applications

FIGURE 9.6 Recovery of phenolic compounds from artichoke brines by integrated membrane process. NF, nanofiltration; UF, ultrafiltration. Adapted from Cassano, A., Cabri, W., Mombelli, G., Peterlongo, F., Giorno, L., 2016. Recovery of bioactive compounds from artichoke brines by nanofiltration. Food Bioprod. Process., 98, 257
265.

1000 Da, from Microdyn-Nadir) gave the lowest average rejection toward sugar compounds and high rejections toward both anthocyanins (89.2%) and flavonoids (70%). The rejection of NF membranes toward the analyzed compounds decreased as a function of the MWCO of the selected membranes. However, a different behavior was observed for anthocyanins. For these compounds rejections were higher than 89%, independently by MWCO. This behavior was attributed to the positive charge of anthocyanins at the acidic pH of the orange press liquor (3.4). At this pH the selected membranes exhibit a positive charge: therefore, the observed retention was attributed to an electrostatic repulsion rather than molecular sieving phenomena. In agreement with data reported by Galanakis (2015), for NF and tight UF membranes (MWCO 1
2 kDa), the electrostatic interactions between membrane surface and solutes can improve membrane selectivity, if the chemical environment is wisely selected. Grape processing residues such as pomace, grape seeds, and lees contain valuable phenolic compounds, such as flavonoids (catechins, epicatechins, proanthocyanidins, etc.) and stilbenzenes (i.e., trans-resveratrol) with recognized antioxidant properties. The traditional extraction of these compounds

Current and Future Applications of Nanofiltration Chapter | 9

341

from their sources is based on the use of organic solvents that are not safe for human consumption due to potential toxic effects from the residual solvent. At the same time, conventional purification processes of bioactive compounds from grape extracts is based on the use of adsorption chromatography. Huge amounts of solvent are needed with significant impact on economic and environmental aspects of the process. Additional drawbacks are represented by the low flexibility of the process and the production of a large volume of brine solutions used for regeneration of adsorbent materials between consecutive cycles (Crespo and Brazinha, 2010). Membrane processes offer a very powerful alternative for the production of natural grape extracts, due to their flexibility, mild operating conditions, low energy requirement, no additive, greater separation efficiency, and easy scaling up. In this contest, NF membranes can contribute significantly to fractionation of wine byproducts, also in integrated membrane systems, for the recovery of high-added value compounds. A fractionation sequence for the purification of proanthocyanic fractions with different degrees of polymerization as developed by Santamarı´a et al. (2002). In the investigated process, a tubular NF polymeric membrane (AFC40, rejection 60% CaCl2, from PCI Membrane Systems) allowed to remove acids and aldehydes from defatted milled grape seed extracts, while monomers were almost completely rejected. DF was employed to increase the purity of proanthocyanidins in the retentate. A combination of pressure-driven membrane operations in sequential design was proposed by Giacobbo et al. (2017) to recover and fractionate polyphenols and polysaccharides from second racking wine lees. According to the conceptual proposed design the diluted wine lees from the second racking is firstly microfiltered to remove suspended colloidal particles. The MF permeate is processed by UF producing a concentrated stream enriched in polysaccharides and a permeate stream enriched in polyphenols. The phenolic fraction is then concentrated by NF. Among the investigated membranes the NF 270 membrane (polypiperazine TFC, MWCO 200
300 Da, from Dow-Filmtec) exhibited a rejection coefficient toward polyphenols higher than 90% independently by the applied operating pressure (from 3 to 5 bar). Fractions enriched in compounds with antioxidant activity were also obtained by Dı´az-Reinoso et al. (2009) by processing aqueous extracts from pressed distilled grape pomace with UF and NF membranes. All selected membranes presented similar rejections of total phenolics and sugars, and were considered suitable for concentration purposes. Concentrated fractions were extracted with ethyl ether to obtain extracts with high antioxidant activity. The highest yields (1.60 g solid/100 g retentate) corresponded to retentates obtained with Nanomax 50 (polyamide/polysulfone, 250 Da, from Millipore) and Inside Ce´ram (Titania, 1 kDa, from Tami) membranes.

342

Separation of Functional Molecules in Food by Membrane Technology

Recently, Diaz-Reinoso et al. evaluated the performance of a membrane process for the production of a phenolic-rich product with high antioxidant activity from white vinasses (Diaz-Reinoso et al., 2017). A sequence of centrifugation and membrane filtration to selectively remove suspended and soluble solids and phenolic compounds was considered an optimal approach. The final dried product, obtained from the NF retentate, contained 45% GAE (gallic acid equivalent) and presented an ABTS (2,20-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid)) radical scavenging capacity equivalent to almost 2 g of Trolox. The recovery of phenolic compounds from agro-food byproducts by using membrane technology has been recently reviewed by Castro-Mun˜oz et al. (2016). This emerging approach can provide a better outlook on the utilization of valuable solutes in the food industry as well as for the productions of new nutraceuticals, cosmetics, and food supplements. In this view, the authors concluded that research and development will be focused on new implementations of NF technology as the primary tool for the recovery and concentration of these valuable compounds.

9.8 CONCLUSIONS AND FUTURE TRENDS In this chapter an overview of potential and well-established applications of NF in food processing have been revised and discussed. Results on laboratory and pilot scale units demonstrate that NF membranes meet the requirement of the “green food processing” strategy. Case studies in different areas of food processing (fruit juice, wine and must, dairy, and sugars) underline the key advantages of NF membranes over conventional technologies due to their intrinsic properties in terms of high degree of selectivity, minimal thermal damage of the treated solutions, reduction of energetic consumption, and environmental impact. Better control of the production process, great flexibility, compactness and possibility of automation, water saving, and low energy costs are additional advantages. The increased interest in NF over the past 20 years indicates that with further membrane development research, the application of NF in food processing will increase significantly in the near future.

REFERENCES Acosta, O., Vaillant, F., Perez, A.P., Dornier, M., 2017. Concentration of polyphenolic compounds in blackberry (Rubus Adenotrichos Schltdl.) juice by nanofiltration. J. Food Process Eng. 40 (1), 1
7. Alkhatim, H.S., Alcaina, M.I., Soriano, E., Iborra, M.I., Lora, J., Arnal, J., 1998. Treatment of whey effluents from dairy industries by nanofiltration membranes. Desalination 119 (1
3), 177
184. Allied Market Research, 2017. Nanofiltration Membranes Market: Global Opportunity Analysis and Industry Forecast, 2017
2023.

Current and Future Applications of Nanofiltration Chapter | 9

343

Arend, G.D., Adorno, T.W., Rezzadori, K., Di Luccio, M., Chaves, C.V., Reginatto, F.H., et al., 2017. Concentration of phenolic compounds from strawberry (Fragaria Xananassa Duch) juice by nanofiltration membrane. J. Food Eng. 201, 36
41. Arriola, N.A., dos Santos, G.D., Prudencio, E.S., Vitali, L., Petrus, C.C., Amboni, R.D.M., 2014. Potential of nanofiltration for the concentration of bioactive compounds from watermelon juice. Int. J. Food Sci. Technol. 49 (9), 2052
2060. Atra, R., Vatai, G., Bekassy-Molnar, E., Balint, A., 2005. Investigation of ultra- and nanofiltration for utilization of whey protein and lactose. J. Food Eng. 67 (3), 325
332. Bandini, S., Nataloni, L., 2015. Nanofiltration for dextrose recovery from crystallization mother liquors: a feasibility study. Sep. Purif. Technol. 139, 53
62. Banvolgyi, S., Kiss, I., Bekassy-Molnar, E., Vatai, G., 2006. Concentration of red wine by nanofiltration. Desalination 198 (1
3), 8
15. Barba, F.J., Zhu, Z., Koubaa, M., Sant’Ana, A.S., 2016. Green alternative methods for the extraction of antioxidant bioactive compounds from winery wastes and by-products: a review. Trend Food Sci. Technol. 49, 96
109. Bargeman, G., Timmer, M., van der Horst, C., 2005. Nanofiltration in the food industry. In: Schaefer, A.I., Fane, A.G., Waite, T.D. (Eds.), Nanofiltration: Principles and Applications. Elsevier, Oxford, pp. 305
328. Bazzarelli, F., Piacentini, E., Poerio, T., Mazzei, R., Cassano, A., Giorno, L., 2016. Advances in membrane operations for water purification and biophenols recovery/valorization from OMWWs. J. Memb. Sci. 497, 402
409. Beda´s, M., Tanguy, G., Dolivet, A., Serge Me´jean, S., Gaucheron, F., Garric, G., et al., 2017. Nanofiltration of lactic acid whey prior to spray drying: scaling up to a semi-industrial scale. LWT - Food Sci. Technol. 79, 355
360. Bernardo, M.P., Coelho, L.F., Cristina Sassa, D.C., Contiero, J., 2016. L-(1)-Lactic acid production by Lactobacillus rhamnosus B103 from dairy industry waste. Braz. J. Microbiol. 47 (3), 640
646. Bhattacharjee, C., Saxena, V.K., Dutta, S., 2017. Fruit juice processing using membrane technology: a review. Innovat. Food Sci. Emerg. Technol. 43, 136
153. Binder, T.P., Hadden, D.K., Sievers, L.J., 1999. Nanofiltration process for making dextrose. US Patent 5,869,297. Bonnet, J., De Vilmorin, H., 2004. Process for the controlled reduction of the sugar content of fruit juice and device for practicing this process. US Patent 0234658 A1. Boskou, D., 2006. Sources of natural phenolic antioxidants. Trends Food Sci. Technol. 17, 505
512. Boye, J.I., Arcand, Y., 2013. Current trends in green technologies in food production and processing. Food Eng. Rev. 5, 1
17. Cai, M., Hou, W.Z., Lv, Y.Q., Sun, P.L., 2017a. Behavior and rejection mechanisms of fruit juice phenolic compounds in model solution during nanofiltration. J. Food Eng. 195, 97
104. Cai, M., Hou, W.Z., Lv, Y.Q., Sun, P.L., 2017b. Understanding nanofiltration fouling of phenolic compounds in model juice solution with two membranes. Food Bioprocess Technol. 10 (12), 2123
2131. Cassano, A., Conidi, C., Drioli, E., 2011. Clarification and concentration of pomegranate juice (Punica granatum L.) using membrane processes. J. Food Eng. 107 (3
4), 366
373. Cassano, A., Conidi, C., Giorno, L., Drioli, E., 2013. Fractionation of olive mill wastewaters by membrane separation techniques. J. Hazard. Mater. 248, 185
193. Cassano, A., Conidi, C., Ruby-Figueroa, R., Castro-Mun˜oz, R., 2015. A two-step nanofiltration process for the production of phenolic-rich fractions from artichoke aqueous extracts. Int. J. Mol. Sci. 16 (4), 8968
8987.

344

Separation of Functional Molecules in Food by Membrane Technology

Cassano, A., Cabri, W., Mombelli, G., Peterlongo, F., Giorno, L., 2016. Recovery of bioactive compounds from artichoke brines by nanofiltration. Food Bioprod. Process. 98, 257
265. Cassano, A., Conidi, C., Ruby-Figueroa, R., Castro-Mun˜oz, R., 2018. Nanofiltration and tight ultrafiltration membranes for the recovery of polyphenols from agro-food by-products. Int. J. Mol. Sci. 19 (2), article 351. Castro-Mun˜oz, R., Ya´n˜ez-Ferna´ndez, J., Fı´la, V., 2016. Phenolic compounds recovered from agro-food by-products using membrane technologies: an overview. Food. Chem. 213, 753
762. Catarino, I., Minhalma, M., Beal, L.L., Mateus, M., de Pinho, M.N., 2008. Assessment of saccharide fractionation by ultrafiltration and nanofiltration. J. Memb. Sci. 312 (1
2), 34
40. Catarino, M., Mendes, A., 2011. Dealcoholizing wine by membrane separation processes. Innov. Food Sci. Emerg. Technol. 12 (3), 330
337. Chandrapala, J., Duke, M.C., Gray, S.R., Weeks, M., Palmer, M., Vasiljevic, T., 2016. Nanofiltration and nanodiafiltration of acid whey as a function of pH and temperature. Sep. Purif. Technol. 160, 18
27. Chemat, F., Rombaut, N., Meullemiestre, A., Turk, M., Perino, S., Fabiano-Tixier, A.S., et al., 2017. Review of green food processing techniques. Preservation, transformation and extraction. Innovat. Food Sci. Emerg. Technol. 41, 357
377. Chen, G.Q., Eschbach, F.I.I., Weeks, M., Gras, S.L., Kentish, S.E., 2016. Removal of lactic acid from acid whey using electrodialysis. Sep. Purif. Technol. 158, 230
237. Conidi, C., Cassano, A., Drioli, E., 2011. A membrane-based study for the recovery of polyphenols from bergamot juice. J. Memb. Sci. 375 (1
2), 182
190. Conidi, C., Cassano, A., 2015. Recovery of phenolic compounds from bergamot juice by nanofiltration membranes. Desalin. Water Treat. 56 (13), 3510
3518. Conidi, C., Cassano, A., Drioli, E., 2012. Recovery of phenolic compounds from orange press liquor by nanofiltration. Food Bioprod. Process. 90 (4), 867
874. Conidi, C., Cassano, A., Drioli, E., 2017a. Membrane-based agro-food production processes for polyphenol separation, purification and concentration. Curr. Opin. Food Sci. 17. Available from: https://doi.org/10.1016/j.cofs.2017.10.009. Conidi, C., Cassano, A., Caiazzo, F., Drioli, E., 2017b. Separation and purification of phenolic compounds from pomegranate juice by ultrafiltration and nanofiltration membranes. J. Food Eng. 195, 1
13. Co´rdova, A., Astudillo, C., Giorno, L., Guerrero, C., Conidi, C., Illanes, A., et al., 2016. Nanofiltration potential for the purification of highly concentrated enzymatically produced oligosaccharides. Food Bioprod. Process. 98, 50
61. Co´rdova, A., Astudillo, C., Santiban˜ez, L., Cassano, A., Ruby-Figueroa, A., Illanes, A., 2017. Purification of galacto-oligosaccharides (GOS) by three-stage serial nanofiltration units under critical transmembrane pressure conditions. Chem. Eng. Res. Design 117, 488
499. Couto, D.S., Dornier, M., Pallet, D., Reynes, M., Dijoux, D., Freitas, S.P., et al., 2011. Evaluation of nanofiltration membranes for the retention of anthocyanins of ac¸ai (Euterpe oleracea Mart.) juice. Desalin. Water Treat. 27 (1
3), 108
113. Crespo, J.G., Brazinha, C., 2010. Membrane processing: natural antioxidants from winemaking by-products. Filtrat. Sep. 47 (2), 32
35. Cuartas-Uribe, B., Alcaina-Miranda, M.I., Soriano-Costa, E., Mendoza-Roca, J.A., Iborra-Clar, M.I., Lora-Garcı´a, J., 2009. A study of the separation of lactose from whey ultrafiltration permeate using nanofiltration. Desalination 241 (1
3), 244
255. Datta, R., Henry, M., 2006. Lactic acid: recent advances in products, processes and technologies: a review. J. Chem. Technol. Biotechnol. 81 (7), 1119
1129.

Current and Future Applications of Nanofiltration Chapter | 9

345

Daufin, G., Escudier, J.P., Carre`re, H., Be´rot, S., Fillaudeau, L., Decloux, M., 2001. Recent and emerging applications of membrane processes in the food and dairy industry. Trans. IChemE 79 (C2), 89
102. Dhineshkumar, V., Ramasamy, D., 2017. Review on membrane technology applications in food and dairy processing. Appl. Biotechnol. Bioeng. 3 (5), 1
10. Dı´az-Reinoso, B., Moure, A., Domı´nguez, H., Parajo´, J.C., 2009. Ultra- and nanofiltration of aqueous extracts from distilled fermented grape pomace. J. Food Eng. 91 (4), 587
593. Diaz-Reinoso, B., Moure, A., Gonzalez, J., Dominguez, H., 2017. A membrane process for the recovery of a concentrated phenolic product from white vinasses. Chem. Eng. J. 327, 210
217. Echavarrı´a, A.P., Torra´s, C., Paga`n, J., Ibarz, A., 2011. Fruit juice processing and membrane technology application. Food Eng. Rev. 3 (3
4), 136
158. Elleuch, M., Bedigian, D., Roiseux, O., Besbes, S., Blecker, C., Attia, H., 2011. Dietary fibre and fibre-rich by-products of food processing: characterization, technological functionality and commercial applications: a review. Food. Chem. 124, 411
421. El Rayess, Y., Mietton-Peuchot, M., 2016. Membrane technologies in wine industry: an overview. Crit. Rev. Food. Sci. Nutr. 56 (12), 2005
2020. Feng, Y.M., Chang, X.L., Wang, W.H., Ma, R.Y., 2009. Separation of galacto-oligosaccharides mixture by nanofiltration. J. Taiwan Inst. Chem. Eng. 40 (3), 326
332. Galanakis, C.M., 2012. Recovery of high added-value components from food wastes: conventional, emerging technologies and commercialized applications. Trends Food Sci. Technol. 26, 68
87. Galanakis, C.M., 2015. Separation of functional macromolecules and micromolecules: from ultrafiltration to the border to nanofiltration. Trends Food Sci. Technol. 42 (1), 44
63. Ganju, S., Gogate, P.R., 2017. A review on approaches for efficient recovery of whey proteins from dairy industry effluents. J. Food Eng. 215, 84
96. Garcia-Castello, E., Cassano, A., Criscuoli, A., Conidi, C., Drioli, E., 2010. Recovery and concentration of polyphenols from olive mill wastewaters by integrated membrane system. Water Res. 44 (13), 388
3892. Garcı´a-Martı´na, N., Palacio, L., Pra´danos, P., Herna´ndez, A., Ortega-Heras, M., Pe´rez-Magarin˜o, S., et al., 2009. Evaluation of several ultra- and nanofiltration membranes for sugar control in winemaking. Desalination 245 (1
3), 554
558. Garcı´a-Martı´na, N., Perez-Magarin˜o, S., Ortega-Heras, M., Gonza´lez-Huerta, C., Mihnea, M., Gonza´lez-Sanjose´, M.L., et al., 2010. Sugar reduction in musts with nanofiltration membranes to obtain low alcohol-content wines. Sep. Purif. Technol. 76 (2), 158
170. Giacobbo, A., Bernardes, A.M., de Pinho, M.N., 2017. Sequential pressure-driven membrane operations to recover and fractionate polyphenols and polysaccharides from second racking wine lees. Sep. Purif. Technol. 173, 49
54. Goulas, A.K., Kapasakalidis, P.G., Sinclair, H.R., Rastall, R.A., Grandison, A.S., 2002. Purification of oligosaccharides by nanofiltration. J. Memb. Sci. 209 (1), 321
335. Goulas, A.K., Grandison, A.S., Rastall, R.A., 2003. Fractionation of oligosaccharides by nanofiltration. J. Sci. Food. Agric. 83 (7), 675
680. Gul, S., Harasek, M., 2012. Energy saving in sugar manufacturing through the integration of environmental friendly new membrane processes for thin juice pre-concentration. Appl. Thermal Eng. 43, 128
133. Hua, X., Zhao, H., Yang, R., Zhang, W., Zhao, W., 2010. Coupled model of extended NernstPlanck equation and film theory in nanofiltration for xylooligosaccharides syrup. J. Food Eng. 100 (2), 302
309.

346

Separation of Functional Molecules in Food by Membrane Technology

Jiao, B., Cassano, A., Drioli, E., 2004. Recent advances on membrane processes for the concentration of fruit juices: a review. J. Food Eng. 63 (3), 303
324. Kotsanopoulos, K.V., Arvanitoyannis, I.S., 2015. Membrane processing technology in the food industry: food processing, wastewater treatment, and effects on physical, microbiological, organoleptic, and nutritional properties of foods. Crit. Rev. Food. Sci. Nutr. 55 (9), 1147
1175. Kristinsson, H.G., Rasco, B.A., 2000. Fish protein hydrolysates: production, biochemical, and functional properties. Crit. Rev. Food. Sci. Nutr. 40, 43
81. Kumar, P., Sharma, N., Ranjan, R., Kumar, S., Bhat, Z.F., Jeong, D.K., 2013. Perspective of membrane technology in dairy industry: a review. Asian-Australas. J. Anim. Sci. 26 (9), 1347
1358. Labanda, J., Vichi, S., Llorens, J., Lopez-Tamames, E., 2009. Membrane separation technology for the reduction of alcoholic degree of a white model wine. LWT - Food Sci. Technol. 42 (8), 1390
1395. Lattanzio, V., Kroon, P.A., Linsalata, V., Cardinali, A., 2009. Globe artichoke: a functional food and source of nutraceutical ingredients. J. Funct. Foods 1 (2), 131
144. Li, W., Li, J., Chen, T., Chen, C., 2004. Study on nanofiltration for purifying fructooligosaccharides I. Operation modes. J. Memb. Sci. 245, 123
129. Li, Y., Shahbazi, A., Williams, K., Wan, C., 2008. Separate and concentrate lactic acid using combination of nanofiltration and reverse osmosis membranes. Appl. Biochem. Biotechnol. 147 (1
3), 1
9. Lo´pez, M., Alvarez, S., Riera, F.A., Alvarez, R., 2002. Production of low alcohol content apple cider by reverse osmosis. Ind. Eng. Chem. Res. 41 (25), 6600
6606. Lo´pez Leiva, M.H., Guzman, M., 1995. Formation of oligosaccharides during enzymic hydrolysis of milk whey permeates. Process Biochem. 30 (8), 757
762. Luo, J., Wan, Y., 2013. Effects of pH and salt on nanofiltration-a critical review. J. Memb. Sci. 438, 18
28. Machado, M.T.C., Trevisan, S., Pimentel-Souza, J.D.R., Pastore, G.M., Hubinger, M.D., 2016. Clarification and concentration of oligosaccharides from artichoke extract by a sequential process with microfiltration and nanofiltration membranes. J. Food Eng. 180, 120
128. Magueijo, V., Minhalma, M., Queiroz, D., Geraldes, V., Macedo, A., de Pinho, M.N., 2005. Reduction of wastewaters and valorisation of by-products from “Serpa” cheese manufacture using nanofiltration. Water Sci. Technol. 52 (10
11), 393
399. Malik, A.A., Kour, H., Bhat, A., Kaul, R.K., Khan, S., Khan, S.U., 2013. Commercial utilization of membranes in food industry. Int. J.Food Nutr. Safety 3 (3), 147
170. Massot, A., Mietton-Peuchot, M., Peuchot, C., Milisica, V., 2008. Nanofiltration and reverse osmosis in winemaking. Desalination 231 (1
3), 283
289. Mellal, M., Jaffrin, M.Y., Ding, L.H., Delattre, C., Michaud, P., Courtois, J., 2008. Separation of oligoglucuronans of low degrees of polymerization by using a high shear rotating disk filtration module. Sep. Purif. Technol. 60 (1), 22
29. Mohammad, A.W., Teow, Y.H., Ang, W.L., Chung, Y.T., Oatley-Radcliffe, D.L., Hilal, N., 2015. Nanofiltration membranes review: recent advances and future prospects. Desalination 356, 226
254. Moreno-Vilet, L., Bonnin-Paris, J., Bostyn, S., Ruiz-Cabrera, M.A., Moscosa-Santilla´n, M., 2014. Assessment of sugars separation from a model carbohydrates solution by nanofiltration using a design of experiments (DoE) methodology. Sep. Purif. Technol. 131, 84
93. Mulder, M., 1991. Basic Principles of Membrane Technology. Kluwer Academic Publishers, Dordrecht.

Current and Future Applications of Nanofiltration Chapter | 9

347

Nath, K., Dave, H.K., Patel, T.M., 2018. Revisiting the recent applications of nanofiltration in food processing industries: progress and prognosis. Trend Food Sci. Technol. 73, 12
24. Nguyen, M., Reynolds, N., Vigneswaran, S., 2003. By-product recovery from cottage cheese production by nanofiltration. J. Cleaner Production 11 (7), 803
807. Ortega-Heras, M., Gonza´lez-SanJose´, M.L., Beltra´n, S., 2002. Aroma composition of wine studied by different extraction methods. Anal. Chim. Acta 458 (1), 85
93. Pan, K., Song, Q., Wang, L., Cao, B., 2011. A study of demineralization of whey by nanofiltration membrane. Desalination 267 (2
3), 217
221. Paraskeva, C.A., Papadakis, V.G., Tsarouchi, E., Kanellopoulou, D.G., Koutsoukos, P.G., 2007. Membrane processing for olive mill wastewater fractionation. Desalination 213 (1
3), 218
229. Pati, S., La Notte, D., Clodoveo, M.L., Cicco, G., Esti, M., 2014. Reverse osmosis and nanofiltration membranes for the improvement of must quality. Europ. Food Res. Technol. 239 (4), 595
602. Patil, N.V., Feng, X., Sewalt, J.J.W., Boom, R.M., Anja, E.M., Janssen, A.E.M., 2015. Separation of an inulin mixture using cascaded nanofiltration. Sep. Purif. Technol. 146, 261
267. Paul, M., Jons, S.D., 2016. Chemistry and fabrication of polymeric nanofiltration membranes: a review. Polymer. (Guildf). 103, 417
456. Pilipovik, M.V., Riverol, C., 2005. Assessing dealcoholization systems based on reverse osmosis. J. Food Eng. 69 (4), 437
441. Pontalier, P.Y., Ismail, A., Ghoul, M., 1997. Mechanisms for the selective rejection of solutes in nanofiltration membranes. Sep. Purif. Technol. 12 (2), 175
181. Popovic, K., Pozderovic, A., Jakobek, L., Rukavina, J., Pichler, A., 2016. Concentration of chokeberry (Aronia melanocarpa) juice by nanofiltration. J. Food Nutr. Res. 55 (2), 159
170. Pouliot, Y., 2008. Membrane processes in dairy technology - From a simple idea to worldwide panacea. Int. Dairy J. 18 (7), 735
740. Prazeres, A.R., Carvalho, F., Rivas, J., 2012. Cheese whey management: a review. J. Environ. Manage. 110, 48
68. Pruksasri, S., Nguyen, T.H., Haltrich, D., Novalin, S., 2015. Fractionation of a galactooligosaccharides solution at low and high temperature using nanofiltration. Sep. Purif. Technol. 151, 124
130. Rice, G., Kentish, S., O’Connor, A., Stevens, G., Lawrence, N., Barber, A., 2006. Fouling behaviour during the nanofiltration of dairy ultrafiltration permeate. Desalination 99 (1
3), 239
241. Romani, A., Scardigli, A., Pinelli, P., 2017. An environmentally friendly process for the production of extracts rich in phenolic antioxidants from Olea europaea L. and Cynara scolymus L. matrices. Europ. Food Res. Technol. 243 (7), 1229
1238. Russo, C., 2007. A new membrane process for the selective fractionation and total recovery of polyphenols, water and organic substances from vegetation waters (VW). J. Memb. Sci. 288 (1
2), 239
246. Salehi, F., 2014. Current and future applications for nanofiltration technology in the food processing. Food Bioprod. Process. 92 (C2), 161
177. Salgado, C.M., Ferna´ndez-Ferna´ndez, E., Palacio, L., Herna´ndez, A., Pra´danos, P., 2015. Alcohol reduction in red and white wines by nanofiltration of musts before fermentation. Food Bioprod. Process. 96, 285
295.

348

Separation of Functional Molecules in Food by Membrane Technology

Santamarı´a, B., Salazar, G., Beltra´n, S., Cabezas, J.L., 2002. Membrane sequences for fractionation of polyphenolic extracts from defatted milled grape seeds. Desalination 148 (1
3), 103
109. Santos, F.R., Catarino, I., Geraldes, V., de Pinho, M.N., 2008. Concentration and rectification of grape must by nanofiltration. Am. J. Enol. Viticult. 59 (4), 446
450. Schieber, A., Stintzing, F.C., Carle, R., 2001. By-products of plant food processing as a source of functional compounds
recent developments. Trends Food Sci. Technol. 12, 401
413. Schmidtke, L.M., Blackman, J.W., Agboola, S.O., 2012. Production technologies for reduced alcoholic wines. J. Food. Sci. 77 (1), R25
R41. Sen˜orans, F.J., Iban˜ez, E., Cifuentes, A., 2003. New trends in food processing. Crit. Rev. Food. Sci. Nutr. 43 (5), 507
526. Sjo¨man, E., Ma¨ntta¨ri, M., Nystro¨m, M., Koivikko, H., Heikkila¨, H., 2007. Separation of xylose from glucose by nanofiltration from concentrated monosaccharide. J. Memb. Sci. 292 (1
2), 106
115. ´ lvarez, S., Riera, F.R., A ´ lvarez, R., 2006. Partial demineralization of Sua´rez, E., Lobo, A., A whey and milk ultrafiltration permeate by nanofiltration at pilot-plant scale. Desalination 198 (1
3), 274
281. ¨ zgu¨ven, M., Diken, T., O ¨ zc¸elic, B., Erim, F.D., 2009. Antioxidant activity Tezcan, F., Gu¨ltekin-O and total phenolics, organic acid and sugar content in commercial pomegranate juices. Food. Chem. 115 (3), 873
877. Van der Bruggen, B., 2009. Chemical modification of polyethersulfone nanofiltration membranes: a review. J. Appl. Polym. Sci. 114 (1), 630
642. Van der Bruggen, B., Vandecasteele, C., Van Gestel, T., Doyen, W., Leysen, R., 2003. A review of pressure-driven membrane processes in wastewater treatment and drinking water production. Environ. Progress 22 (1), 46
56. Van der Horst, H.C., Timmer, J.M.K., Robbertsen, T., Leenders, J., 1995. Use of nanofiltration for concentration and demineralization in the dairy industry: model for mass transport. J. Memb. Sci. 104 (3), 205
218. Versari, A., Ferrarini, R., Parpinello, G.P., Galassi, S., 2003. Concentration of grape must by nanofiltration membranes. Food Bioproduct Processing 81 (C3), 275
278. Wang, K., Li, W., Fan, Y., Xing, W., 2013. Integrated membrane process for the purification of lactic acid from a fermentation broth neutralized with sodium hydroxide. Ind. Eng. Chem. Res. 52 (6), 2412
2417. Warczoc, J., Ferrando, M., Lo´pez, F., Gu¨ell, C., 2004. Concentration of apple and pear juices by nanofiltration at low pressures. J. Food Eng. 63 (1), 63
70. Xu, L., Wang, S., Zeng, X., 2005. The maltitol purification and concentration by nanofiltration. Desalination 184 (1
3), 295
303.

FURTHER READING ,https://www.alliedmarketresearch.com/nanofiltration-membranes-market?nanofiltration-membranes-market., Accessed 13.02.18.

Chapter 10

Electrodialysis-Based Separation Technologies in the Food Industry Yaoming Wang1, Chenxiao Jiang1, Laurent Bazinet2 and Tongwen Xu1 1

CAS Key Laboratory of Soft Matter Chemistry, Collaborative Innovation Centre of Chemistry for Energy Materials, School of Chemistry and Materials Science, University of Science and Technology of China, Hefei, P.R. China, 2Institute of Nutrition and Functional Foods (INAF), Dairy Research Center (STELA), Department of Food Sciences, Laboratory of Food Processing and ElectroMembrane Processes (LTAPEM), Laval University, Quebec City, QC, Canada

Chapter Outline 10.1 General Introduction of Electrodialysis and Electrodialysis With Bipolar Membrane 10.2 Electrodialysis in Food Processing 10.2.1 Whey 10.2.2 Fruit Juice 10.2.3 Wine 10.2.4 Sauce 10.2.5 Sugars 10.2.6 Amino Acids 10.3 Electrodialysis With Bipolar Membrane Applications in Food Processing 10.3.1 Applications to Fruit Juices 10.3.2 Applications to Proteins

350 351 351 353 354 355 356 357

359 359 361

10.3.3 Application to Phospholipids 364 10.4 Electrodialysis With Filtration Membrane Applications in Food Processing 366 10.4.1 Applications to Protein Hydrolysates 367 10.4.2 Separation of Protein 370 10.4.3 Separation of Chitosan Oligomers 371 10.4.4 Concentration of Cranberry Juice in Phenolic Compounds 372 10.5 Conclusion 373 Acknowledgments 374 References 374 Further Reading 381

Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00010-3 © 2019 Elsevier Inc. All rights reserved.

349

350

Separation of Functional Molecules in Food by Membrane Technology

10.1 GENERAL INTRODUCTION OF ELECTRODIALYSIS AND ELECTRODIALYSIS WITH BIPOLAR MEMBRANE Electrodialysis (ED) is a membrane separation process based on ion exchange (IE) membranes. It is used to transport ionic compounds from one solution to another under the influence of an applied potential difference. For a typical ED stack, as shown in Fig. 10.1, anion and cation exchange membranes (CEM) are alternating arranged between two electrodes. Dilute stream, concentrate stream, and electrode stream are subjected to flow through the cell compartments formed by the IE membranes. When an electrical current field is applied, anions migrate toward the anode, and pass across the anion exchange membranes (AEM), but are prevented from further migration toward the anode by the CEM. Similarly, cations migrate toward the cathode, and pass across the CEM, but are prevented from further migration toward the cathode by the CEM. As a result, all the ionic

FIGURE 10.1 The main principles of ED and ED with bipolar membranes (EDBM). A, Anion exchange membrane; BP, bipolar membrane; C, cation exchange membrane.

Electrodialysis-Based Separation Technologies Chapter | 10

351

components are concentrated in the concentrate stream with a depletion of ions in the dilute solution feed stream. Owning to its distinguished functions, ED has been widely applied to deionization or condensation of aqueous solutions, including desalination of brackish water, water reuse, food processing, etc. (Strathmann, 2010; Ran et al., 2017). Especially the applications in the food and chemical industry such as the demineralization of whey or deacidification of fruit juice are becoming increasingly more important (Huang et al., 2007; Xu and Huang, 2008). ED with bipolar membrane (EDBM) is considered as the state-of-the-art of ED, because it integrates the functions of ED the water splitting of bipolar membranes (Huang and Xu, 2006). In a typical EDBM stack (BP
A
C
BP), as shown in Fig. 10.1, bipolar membranes, CEM, and AEM are alternating arranged between two electrodes. Four cell compartments, that is, acid compartment, base compartment, salt compartment, and electrode compartment are formed. Under the influence of an electrical potential difference, anions migrate toward the anode, and pass across the AEM, and then combine with protons generated from the bipolar membrane. Therefore, acids are generated in the acid compartment. Cations migrate toward the cathode, and pass across the CEM, and combine with hydroxyls generated from the bipolar membrane. Just like in the ED process, ionic components are depleted in the salt compartment. This allows the conversion of salts from the corresponding salts. The use of bipolar membranes, in contrast to solvent electrolysis for alkali generation, has the advantages of no gas or byproduct generation, lower voltage drop, maximal energy utilization, space saving, and easier installation and operation (Xu, 2005). Therefore, this technique has extensively applications in chemical and biochemical synthesis, food processing, and pollution control. For food processing, EDBMs are developing as a very promising substitute to traditional acidification or basification technologies. The advantages of EDBM over traditional methods are lower thermal damage to product, less energy consumption, and lower equipment costs (Nagarale et al., 2006). The involvement of IE membranes not only makes the process cleaner and more energy-efficient, but also increases the added value of the products.

10.2 ELECTRODIALYSIS IN FOOD PROCESSING 10.2.1 Whey Whey, the liquid fraction drained from the curd, is a valuable byproduct of the manufacture of cheese or casein. There are two types of whey, sweet whey and acid whey. The former is the byproduct of rennet-coagulated cheese and has a pH $ 5.6. The latter is the byproduct of acid-coagulated cheese and has pH # 5.1. For cheese making, 1 kg milk results in 0.1
0.2 kg cheese and 0.8
0.9 kg whey (Akpinar-Bayizit et al., 2009). In

352

Separation of Functional Molecules in Food by Membrane Technology

general, sweet whey contains a high concentration of whey protein, lactose, vitamins, and minerals (Benitez and Ortero, 2012). Acid whey has similar components but more citric acid because it uses citric acid as a coagulation agent. Whey proteins consist of α-lactalbumin (α-LA), β-lactoglobulin (β-LG), serum albumin, immunoglobulins, and proteose-peptones. These proteins are believed to be involved in some biological functions such as antioxidant activity and have anticarcinogenic effects (Madureira et al., 2007; Chico´n et al., 2009; Smithers, 2008). Whey has increasingly been a focus of commercial interest as potential ingredients in functional or healthpromoting foods. Therefore, recovery and reuse of whey is of significance in the food industry, because it can achieve not only the recycling of waste but also alleviating the pollution problem arising from discharge of whey. Whey is usually further processed to make whey powder, lactose, whey protein concentrate (WPC), and demineralized whey powder. But in practical conditions, it is difficult to utilize it due to its unfavorable lactose to protein ratio and high-salt content in the dry matter of whey (Kinsella and Whitehead, 1989). The salt content has negative effects on functionality, organoleptic properties, and the value of the final products, most of which are related to infant food. Traditionally, IE resin was attempted to demineralize whey, but this technology was not successful because of the severe fouling of the resins and denaturation of the protein arisen from the substantial pH perturbation in the resin. Numerous studies have proven the high effectiveness of ED for demineralization of whey (Okada et al., 1975; Bleha et al., 1992; Greiter et al., 2002). Chen et al. (2016) used ED for the removal of lactic acid from acid whey because high lactic acid concentration causes operational problems in downstream spray-drying operations due to increased powder stickiness. The energy consumption was low as B0.014 kW h/kg whey with achieving 90% demineralization. They found that the processing time was shortened at high temperature. Greiter et al. (2002) investigated the comparison between ED and IE with regard to sustainability by calculating their cumulative energy demand. For a flow a 45 m3 per day of whey, IE achieved a 99% desalination efficiency, yielded 3.7 m3 wastewater, and an organic charge of 26 kg chemical oxygen demand (COD)/m3 feed whey; ED achieved a 90% desalination efficiency, yielded 1.25 m3 wastewater, and an organic charge of 8.4 kg COD/m3 whey. They concluded that ED utilizes substantially more energy than IE. But there are also several limitations for ED for demineralization of whey. First, ED cannot achieve a 100% demineralization due to the rapid increase in resistance as the ions are being depleted. The energy consumption increases sharply at the later period of demineralization. To address this problem, efficiency is increased by a combination of ED and IE for whey demineralization. ED is used first to demineralize until 65%
75%, then IE is used to demineralize up to the required amount. Furthermore, ED demonstrated no selectivity for different charged ions in the demineralization process, monovalent and

Electrodialysis-Based Separation Technologies Chapter | 10

353

divalent ions are simultaneously desalted that the removal of chloride, and sodium and potassium ions are accomplished by partial elimination of other interesting ions such as calcium. To overcome these shortcomings, Andres et al. (1995) developed a selective ED (SED) stack with monoselective IE membrane. SED resulted in high calcium and low sodium content in the final product that is appropriate for infant formula. The precipitation of insoluble salt (calcium phosphate) and attachment of proteins on the surface of IE causes not only fouling of the membranes but also results in polarized current that is above the limiting current density of the membranes. To address the fouling and scaling problems, the ED stack is desirable to operate at high flow velocity to prevent the precipitation of proteins and calcium phosphate on the surface of the membranes. Finally, the robustness of ED for whey demineralization shall be improved to increase the life span of the membranes. The desirable membranes for the demineralization process of whey shall exhibit some special features, such as antiorganic fouling, and antiacid and antialkaline circumstances for chemical cleaning, and high permeability of larger organic ions.

10.2.2 Fruit Juice Raw fruit juice, which contains flavor and aroma compounds and noteworthy polysaccharides (pectins, cellulose, hemicellulose, and starch), is widely used in various food preparations such as beverages, ice creams, marmalades, cocktails, or pies. But the high acidity of some fruit juices reduces their quality, limits their addition in food preparation, and gives poor aroma and flavor products. Citric and malic acids are generally responsible for the high acidity of fruit juices. Therefore, deacidification is required to remove the acidic components. Traditionally, the deacidification was achieved by neutralization with a base such as calcium lime or IE resins. The former method not only involves the high consumption of chemicals but also affects the flavor and natural taste of juices, resulting in poor palatability. IE resins have the disadvantages of changes in the organoleptic characteristics of juice treated and large amounts of effluent produced during the regeneration phase of resins. ED is a promising alternative for these traditional methods. The advantages of ED over precipitation and IE resins include: G G G

providing high quality with no changes of the tastes and flavors; reducing the consumption of chemicals; and reducing the effluent production (Vera et al., 2003a,b)

Calle et al. (2002) investigated the deacidification of the clarified passion fruit juice (Passiflora edulis flavicarpa) using electrodialysis. The pH of the juice was successfully increased from 2.9 to 4.0. In another case, Vera et al. (2007a) investigated the deacidification of four tropical fruit juices, passion

354

Separation of Functional Molecules in Food by Membrane Technology

fruit, castilla mulberry, najanrilla, and araza. The fruit juice pHs increase from 3 to 7.9 with satisfactory performances. Membrane fouling was observed for four kinds of juices, which is particularly significant with the mulberry juice probably because of its high content in phenolic compounds. The authors also found that the juices whose acidity was caused by malic acid exhibit low performance compared with that caused by citric acid. Earlier, Vera et al. (2003b) compared the physicochemical and sensory analyses of juices with the fresh juices. They found that at a final pH of 4, the acidity and organic anion concentration decreased by about 70% and 50%
60%, respectively, for all the juices. But it should be stated that for sauce juice deacidification using ED, three cell compartments besides the electrode compartment were usually used (Vera et al., 2003a, b), which is different from the commonly used two-cell compartment ED stack. An extra base stream was provided by a soda solution circulating in an adjacent compartment of juice. If the bipolar membranes were introduced, the deacidification of fruit juice could be achieved without reagent additions. The application of EDBM for the fruit juice deacidification will be discussed in the following section.

10.2.3 Wine Wine is one of the most consumed alcoholic drinks in the world. The major physical instability in bottled wines remains the precipitation of the tartaric salts, including potassium hydrogen tartrate (KHT) and calcium tartrate (CaT). KHT is a natural constituent of grapes but the solubility of the tartrate salts decreases in the presence of ethanol and low temperatures. These tartrate salt deposits are not appreciated by consumer. To overcome the problem, traditional stabilization methods were performed by cooling the wine to very low temperature (
4 C) over several days (Lasanta and Go´mez, 2012). To accelerate the cold stabilization, crystal seeds are usually added (Dunsford and Boulton, 1981). But this is time and energy consuming, and the stabilization performance is dependent on the wine composition. ED is a highly effective technology for stabilizing tartaric precipitation. Gonc¸alvesa et al. (2003) investigated the stabilization of wine tartaric by ED. They found that the extent of KHT removal is varied depending on the duration of the ED up to 24%. In another study, Soares et al. (2009) reported that white and rose´ wines desalted by ED are only stable after a demineralization of 25% and 30%, respectively. In comparison with cold stabilization method, ED has some unique advantageous features (Lasanta and Go´mez, 2012). First, the energy cost of tartaric stabilization by ED is especially low. The total energy consumption was in the range of 0.5
1.0 kWh/m3, that is, about 10 times less than the energy required for cold stabilization method (Daufin et al., 2001). Second, ED does not alter the organoleptic properties of wine. The other compounds of wine including polyphenols,

Electrodialysis-Based Separation Technologies Chapter | 10

355

polysaccharides, amino acids, and volatile compounds are not affected by ED treatment. Romanov and Zelentsov (2007) reported that the chemical and physiological of wines have no taste or smell changes after ED processing (Romanov and Zelentsov, 2007). Third, ED offers flexibility in the removal rate of tartrate by the variation of operating time. This makes possible automatic control of the stabilization process by monitoring the conductivity of the wine stream (Youssef et al., 2016). Finally, some other ionic components, such as calcium ions and sulfate ions can be additionally removed along with the ED stabilization process. The additional removal of calcium is of great interest in wines with risks of CaT precipitation, because normally there is no removal of calcium salts during the traditional cold stabilization process. The reduction of sulfate ions by ED is of particular interest in wines acidified with gypsum, because generally there are no additional procedures to remove sulfate. In summary, ED offers a viable alternative method to tartrate stabilization in wines with significant advantage in the power consumption and wine losses. But the robustness of the membranes, especially the AEM, should be further improved. In Europe the ED system is authorized for table and country wines. To meet more countries’ food wine stabilization regulations, membranes must strictly comply with regulations governing additives, materials in contact with foodstuffs.

10.2.4 Sauce Soy sauce is traditional condiment of Chinese origin and largely used in East Asia to impart an appetizing flavor to several cooked foods, as well as to help digestion. Soy sauce is made from fermented paste of soybeans, roasted grain, brine, and Aspergillus oryzae or Aspergillus sojae mold. To avoid microbial contamination during the production process, 16%
20% (w/v) of sodium chloride is commonly added. High-salt food causes health risks such as high blood pressure, heart, and kidney disease. Therefore, the sodium chloride content of soy sauce should be reduced to 5%
8%. Several desalting technologies such as extraction, freezing, IE, nanofiltration, and reverse osmosis have been proposed to remove salt from soy sauce. But the high capital and operating cost, as well as the reduction of taste and aromabearing components, limited the practical applications of these technologies. ED has also been applied to remove salt from soy sauce. Fidaleo et al. (2012) studied the desalting of soy sauce by ED. They found that ED is technically feasible for the removal of salt content in the soy sauce. The recovery of total acids and amino nitrogen in desalted soy sauce was (80 6 4)% and (70 6 3)%, respectively. The final concentrations of chloride ions and total solids were (14 6 2)% and (73 6 6)% of the corresponding initial sauce, respectively. The taste of the desalted soy sauce is not significant from that of raw soy sauce. In the same way, Mok et al. (2001) obtained 75% desalted

356

Separation of Functional Molecules in Food by Membrane Technology

soy sauce using ED with 16.5% amino nitrogen loss and no deterioration in its flavor, color, and taste. Fish sauce is a popular condiment made from fish coated in salt and fermented for several weeks to up to 2 years. It is usually used to prepare various dishes in Thai and other Asian cuisines. Despite its desirable flavor and aroma characteristics, fish sauce also contains a high concentration of salt. ED has been proven to be an effective and efficient desalting method to remove the abundant salt (Chindapan et al., 2013). Chindapan et al. (2009) constructed an ED system for the removal of salt from fish sauce and evaluated the physicochemical properties of desalted sauce. The salt content could be reduced to a low concentration of 2% w/w. They found that the physicochemical properties of the treated fish sauce, in terms of the total soluble solids, density, viscosity, ion concentrations (i.e., Na 1 , K 1 ), total nitrogen, and color were significantly affected by the input voltage and the saltremoval level. In a further work, Chindapan et al. (2011) investigated the selected aroma compounds and amino acid compositions after ED. They found the amounts of all amino acids decreased with the increased saltremoval level. Meanwhile, the amounts of amounts of trimethylamine, 2,6dimethylpyrazine, phenols, and all carboxylic acids except for hexanoic acid significantly decreased, whereas benzaldehyde increased significantly when the salt-removal level was higher. Even though ED is an efficient method for the removal of salt content from fish sauce, there are also losses of important aroma compounds and amino acids. This led to a significant flavor difference. Apart from flavor changes, one more thing that should be noticed for this application is that the IE membranes must meet food hygiene requirements and a system must be easy in place for cleaning and sanitation by strong acid/base solutions.

10.2.5 Sugars Sugar is a common food additive. In the sugar manufacturing process, the majority of organic nonsugar components consisting primarily of large molecules are usually removed by means of defecation, carbonation, and adsorption (bone char, active carbon, ion exchanger, etc.). The inorganic components are not removed and usually transferred to the following evaporation system, in which residual organic nonsugar components are separated from sugar crystals and remain in molasses. In this case, the inorganic and organic nonsugar components are gradually accumulated in the molasses and are usually discharged as waste molasses. ED is successfully applied to increase the sugar recovering ratio by removing the residual inorganic components. Kokubu et al. (1983) used ED for the demineralization of molasses to increase sugar recovery. To avoid the water dissociation and organic fouling on AEM, they used porous neutral membranes instead of AEM. In that

Electrodialysis-Based Separation Technologies Chapter | 10

357

case, organic fouling was decreased but the current efficiencies are very low (9.5%
43%). In addition to conventional cane sugar and beet sugar, there are more and more applications of ED in oligosaccharides (Yang et al., 2006), such as inulin, xylose, pectose, etc. Some of them have been applied at industrial scale. Xylitol is an important sugar as a sweetener. It is usually made from corn cob following by acid hydrolysis. A residue 0.6%
0.8% H2SO4 would be generated and must be deacidified before further processing. The conventional neutralization method with lime has many drawbacks such as high energy consumption, large consumption of chemicals, and environmental pollution. To overcome these problems, Wang et al. (2015) used ED for the selective removal of residue acid. ED obtained a demineralization rate higher than 99% and xylose recovery rate of 84.9%. In another case, Liu et al. (2014) claimed the use of ED for preparing of a soybean oligosaccharide product, which is prepared by extracting a soybean raw material with water under a pH of 3
6 and at a temperature of 50
70 C to obtain an extract containing acidic soluble saccharides of soybean. They found the conductivity of the acidic solution from 32 to 10 mS/cm, and the salts concentration was reduced from 1.34% to 0.07%. The removal rate of the salts was estimated to be 95%, while the retention rate of the soluble solids (Brix) was 69.4%.

10.2.6 Amino Acids The desalination of amino acids containing salts is an important application of ED that has been commercialized. ED is widely used to remove NaCI, NH4Cl, Na2SO4, etc., from the mother liquid for the separation of amino acids. For these applications, ED can only increase the yield rate but also reduce the cost of wastewater treatment. Amino acids are zwitterionic molecules and their losses during the ED process can be minimized by adjusting the pH values of feed to the isoelectric points of the amino acids. There are numerous studies on ED for amino acid separation and purification (Chen et al., 1995; Montiel et al., 1998; Bukhovets et al., 2009; Zhang et al., 2011; Yuan et al., 2016; Kattan Readi et al., 2013a,b; Wang et al., 2016). In the industrial synthesis of D-α-p-hydroxyphenylglycine, a mother liquor containing a high-salt content (phosphates and sulfates) and an amino acid concentration of B0.12 6 0.15 M were discharged. The disposal causes an environmental problem. Montiel et al. (1998) used ED for the recovery of this amino acid with 85% recovery rate. The authors claimed that ED could achieve not only avoiding the discharge of high salinity and COD wastewater but also valorizing the high value product. Zhang et al. (2011) theoretically and experimentally investigated the separation of small organic ions from salts in the ED process. They found that the selectivity of the IE membrane can be influenced by the applied current density. Even though separation of

358

Separation of Functional Molecules in Food by Membrane Technology

inorganic ions from organic solutes was feasible, the separation performances were highly related to the size, charge, and functional groups of the organic ions. Gamma-aminobutyric acid (GABA) is an important nonprotein amino acid and has many important physiological functions. In particular, GABA is prepared by the reaction between 4-chlorobutryic acid with ammonia to obtain GABA mother liquid. The separation of GABA from reaction mixtures containing ammonium chloride is a problem, since GABA and byproduct NH4Cl have a similar molecular weight (MW) and a similar solubility to most organic solvents. Hence, Wang et al. (2016) investigated ED for the separation of GABA from the mixture salts and obtained a highest desalination rate of 99.29% and GABA loss rate ,3%. To increase the product recovery rate, the lost GABA in the concentrate stream was fed back to dissolve the raw GABA dry product. An industry ED plant was built in Zhejiang Province of China for GABA production with production ability around 3000 t/a, as shown in Fig. 10.2. The residue chloride content in the dry GABA product after ED could be controlled ,5 ppm, which can be used as active pharmaceutical ingredients for treatment of epilepsy. The limitation of ED for amino acid production is that the migration of amino acid ions across IE membranes is usually unsatisfactory due to their weak dissociation and complex chemical structure (Bukhovets et al., 2009; Kattan Readi et al., 2013a). To decrease the loss of products during the separation of amino acids from the salt mixtures, two things are noticed. For one thing, the IE membranes for small amino acid separation or purification shall exhibit dense structure, because there will be molecular diffusion of amino acids from the dilute stream into the stream. If the membranes have much denser structure, there will be less loss of the amino acids. For another thing, the pH values of the feed solution should be carefully monitored and adjusted around isoelectric points of the amino acids. Nevertheless, ED holds high flexibility and strong adaptability for amino acids production.

FIGURE 10.2 An industrial ED plant was built in Zhejiang province of China for GABA production with production ability around 3000 t/a.

Electrodialysis-Based Separation Technologies Chapter | 10

359

10.3 ELECTRODIALYSIS WITH BIPOLAR MEMBRANE APPLICATIONS IN FOOD PROCESSING EDBM is a technology resulting from the combination of ED and specific properties of bipolar membranes. EDBM acts in two ways: electroacidification and electroalkalinization. For a solution circulated on the cationic side of the bipolar membrane, where the H1 are generated, the pH of the solution will decrease. Similarly, a solution circulated on the anionic side of the bipolar membrane, where OH2 are generated, will experience an increase in pH. EDBM has been applied among others for inhibition of enzymatic browning in cloudy juices, deacidification of fruit juices, purification and fractionation of proteins, and separation of phospholipids.

10.3.1 Applications to Fruit Juices 10.3.1.1 Inhibition of Enzymatic Browning in Cloudy Apple Juice During its production, cloudy or unclarified apple juice experiences enzymatic changes that modify its taste and color. These enzymatic changes are related to enzymatic browning reactions, which confer a dark color to the juice and has a negative impact on sales (Tronc, 1996). Tronc et al. (1997) were the first to use bipolar membranes to lower the pH of cloudy apple juice. The apple juice was circulated on the cationic side of the bipolar membrane where the H1 ions are generated. The pH of apple juice was reduced temporarily from 3.5 to 2.0 in a small-scale unit at a constant current of 40 mA/cm2. Reducing the pH of the juice to 2.0 completely inhibited polyphenol oxidase (PPO) activity compared with the control. Following acidification, the pH of the juice was returned to its initial value by introducing OH2 produced by water splitting; the juice was circulated on the anionic side of the bipolar membrane where the OH2 are generated. The pH readjustment of the juice partially reactivated the PPO, but browning inhibition was complete and irreversible. The treatment enhanced the color of cloudy apple juice during storage without modifying the flavor (Tronc et al., 1997). However, the total process was too lengthy, about 90 minutes, and required the addition of exogenous KCl to the juice to reach pH 2.0. Furthermore, the voltage applied greatly exceeded the 2 V/membrane average and was not compatible with industrial constraints. Lam Quoc et al. (2000) replaced cationic membranes in the stack by anionic membranes and the KCl solution with 0.1 M HCl. In this configuration, acidification was still triggered by the introduction of protons generated by the bipolar membranes, but the retention of these protons was more effective owing to the continuous introduction of Cl2 counterions from the HCl compartment. The Cl2 ions accumulated in the juice will subsequently serve as counterions for the introduction of OH2, generated by the anionic side of the bipolar membranes, during the return of the juice pH to its initial value. The authors demonstrated that with a 4-membrane stack, the juice pH decreased from 3.35 to 2.0 in 4.7 minutes. EDBM can be

360

Separation of Functional Molecules in Food by Membrane Technology

considered as a nonthermal stabilization that is simple, efficient, and applicable to the stabilization of food liquids.

10.3.1.2 Deacidification of Acid Juices The yellow passion fruit has intense and special aroma and flavor, which make it a desirable ingredient in the formulation of various food products, while cranberry juice is recognized for its health properties (Raz et al., 2004; Vasileiou et al., 2013). However, these particular juices have high organic acid contents, limiting their uses (Adhikary et al., 1983; Couture and Rouseff, 1992) or causing undesirable effects (diarrhea, vomiting, and bloating) (Wing et al., 2008; Vasileiou et al., 2013; McMurdo et al., 2005). Deacidification of fruit juice is possible by EDBM when the feed stream is circulated in the base compartment. Hence, Vera et al. (2007a,b) and Serre et al. (2016a,b) used an EDBM configuration to deacidify respectively passion fruit and cranberry juices up to 75%
80%. The deacidification of cranberry juice by EDBM was recently tested successfully at an industrial scale during 100 hours (unpublished results). Furthermore, Serre et al. (2016b) demonstrated that in the presence of deacidified cranberry juice, the integrity of Caco-2 cell monolayers (representative of the intestinal cells) was increased by 56% in comparison with raw cranberry juice but a minimal deacidification rate of 37% was necessary to reach this level of protection (Fig. 10.3). Finally, it was demonstrated that EDBM treatments do not have Electrodialysis

Deacidified juice

0 19 37 50 77 % % % % %

0

HB S Ra S w 19 juic e 37 % % 50 % 77 % Tr ito n 1%

Δ TEER ± S.E.M (Ω*cm2)

Raw cranberry juice In vitro digestion (oral, gastric and intestinal)

–500

–1000

56% *** n.s * *

0 19 37 50 77 In vitro tests % % % % % Integrity of epithelial cells

** FIGURE 10.3 Effect of different deacidification rates of cranberry juice on variation of transepithelial electrical resistance (ΔTEER (Ω cm2)) of Caco-2 monolayers ( P , .05,  P , .01,  P , .001, n.s., P..05): Hank’s Balanced Salt Solution (HBSS) corresponding to negative control while Triton 1% is a positive control.

Electrodialysis-Based Separation Technologies Chapter | 10

361

significant influences on physicochemical characteristics of deacidified juices (Serre et al., 2016a,b; Vera et al., 2007a). The EDBM method seems to be advantageous due to the possibility of additional organic acids recovery from the acid compartment that can be reused in other industrial steps or used as preservative agents for food.

10.3.2 Applications to Proteins 10.3.2.1 Production of Isolates A large proportion of the soybean and dairy proteins used in the food industry are in the form of protein isolates. Separation of proteins by isoelectric precipitation at the isoelectric pH range 4.2
4.6 is the recognized industrial process. The isoelectric point is the pH value at which the net global charge of the protein is neutral, and can result in complete precipitation. The conventional procedure involves 4
5 steps: extraction (for soybean only from defatted soy flakes), precipitation, washing, resolubilization, and drying. The defatted soy flakes are dissolved in water at pH 9 6 2 in a ratio ranging from 6:1 to 20:1 at ,80 C. The extraction step usually takes B30 minutes. Soy or dairy proteins are then precipitated by lowering the pH of the solution to the isoelectric point, B4.5, using hydrochloric acid. Centrifugation is used to separate the protein-containing curd from the supernatant (or whey) containing the soluble materials and low MW compounds. The curd is then washed with water to remove soluble impurities. After washing, it is neutralized with NaOH to obtain the protein as proteinate, meaning that it is soluble when placed in solution. The final product is spray dried, and packaged as dry material. Most commercial soy protein products have been developed from the acid-precipitated fraction (Kilara and Sharkasi, 1986). The disadvantages of this method include denaturation of protein on exposure to alkali and acid treatment, high ash content, and alteration of protein solubility after rehydration (Nash and Wolf, 1967; Bazinet et al., 1997a). Local extremes in pH can cause irreversible denaturation of the proteins (Kilara and Sharkasi, 1986; Fisher et al., 1986; Bazinet et al., 1997a). In this context, a procedure using bipolar membranes was developed to precipitate soy proteins (Bazinet et al., 1997b, 1998, 1999a) and caseins from milk (Bazinet et al. 1999b, 2000a, 2002). This technique uses the acid stream from the EDBM stack to decrease the pH of the feed solution until the isoelectric point, where the protein net charge is neutral, leading to protein precipitation (Fig. 10.4). Centrifugation can then be used, as in the conventional process for separation of the proteins. To lower the pH of the protein solution, this solution was circulated on the cationic side of the bipolar membrane. The pH of the protein solution was lowered from 6.8
8.0 to 4.5. Lowering the pH to 4.5 allowed a precipitation of 95%
100% protein. The chemical composition of bipolar membrane electrodialyzed samples was

362

Separation of Functional Molecules in Food by Membrane Technology

NaOH KOH solution pH 12

Neutralization pH 7.5

Soy protein isolate 95% protein

HCl By-product

Electro-acidification

CEM BM

CEM BM

–

OH K

+

CEM BM CEM

H

+

+

K

+

K

Soy flakes 52% protein Extraction pH 8 NaOH

KCl solution

By-product

FIGURE 10.4 ED with bipolar membrane cell configuration and process for soy isolate production. BM, Bipolar membrane; CEM, cation exchange membrane.

demonstrated to be superior or equal to that of commercial standards for soybean protein isolate, with functional properties comparable with these standards (Bazinet et al., 1997a). After washing the precipitate, it is possible to resolubilize the proteins by reusing the sodium hydroxide generated on the anionic side of the bipolar membrane (Fig. 10.4), or by recirculating the precipitated protein solution on the anionic side during the acidification of another protein solution on the cationic side of the bipolar membrane. Ali et al. (2010, 2011) also proposed an approach coupling EDBM with ultrafiltration (UF) membranes for the production of a low-phytate soy protein isolate. However, in the case of milk casein, in spite of the attractiveness of EDBM, precipitation of casein inside the stack and scaling on cation exchange membrane affect the process performance hampering industrial application of this technique (Bazinet et al., 1999b). To answer this problem, Balster et al. (2007) proposed a complex approach avoiding clogging of the EDBM stack by caseins. This approach consists of a classical chemical acidification (for the first batch) of milk in a precipitator followed by separation of casein from whey. The whey flux is further directed to the EDBM stack for demineralization and neutralization. The acid generated in the acidification compartment of EDBM is then used in the precipitator for further milk acidification. Mier et al. (2008) placed an online basket centrifuge allowing separation of whey from casein behind the EDBM cell and in front of the milk reservoir. In spite of promising results of the above studies, the presence of scaling, organic fouling by whey proteins or by casein curd was reported. Very recently, a hybrid approach was proposed (Mikhaylin et al., 2016, 2018) including bipolar membrane ED coupled with ultrafiltration

Electrodialysis-Based Separation Technologies Chapter | 10

363

(EDBM-UF). The crucial problem during EDBM-UF of milk, that is, protein fouling and scaling of membrane, was successfully solved. Moreover, the life cycle assessment of the novel EDBM-UF protein production process demonstrated that the proposed electromembrane process has significant environmental benefits compared with the conventional process using chemicals independently from the electricity supply mix from all considered geographical locations (Mikhaylin et al., 2018). Separation of proteins by EDBM has specific advantages over the conventional isoelectric precipitation used industrially for the production of protein isolates. The protein isolate from EDBM electroacidification has a lower ash content compared with conventional methods with the addition of inorganic acid (Bazinet, 2005; Mier et al., 2008). The higher protein purity relates to the fact that there is no addition of anions such as Cl2 and SO422 during electroacidification and there is an additional solution demineralization (Mikhaylin and Bazinet, 2015). This technology does not use any added acids or bases during the process to adjust the pH of the protein solution, and the chemical effluents generated during the process could be reused at different stages in the process. Indeed, EDBM technology provides an alkali solution, which could be used for further protein neutralization to render protein isolates soluble in water solutions (Fig. 10.4). Thus, the use of inorganic alkali (e.g., NaOH) applying during conventional protein neutralization can be avoided. Moreover, the water consumption is decreased by reusing the effluents generated, and the cell electrical energy consumption for protein precipitation is low (B0.3
0.7 kWh/kg of isolate). Furthermore, EDBM-UF could become a prospective industrial technology taking into account environmental concerns and promoting the development of healthy human society.

10.3.2.2 Fractionation of Proteins Adjusting the temperature to 10 C during EDBM allowed a selective fractionation of soybean protein fractions (Bazinet et al., 2000b). In practice, if a sample of precipitated protein was taken at pH 4.2, the protein composition will be composed of 33.8%, 92.9%, 16.2%, and 8.7% of the original content of the 15S, 11S, 7S, and 2S fractions respectively. This leads to a solution enriched in the 11S fraction in the precipitate (71.8% of 11S and 10.8% of 7S), and to a solution enriched in the 7S fraction in the supernatant (46.6% of 7S and 4.6% of 11S) (Bazinet et al., 2000b). The selective fractionation at low temperature is explained by the fact that proteins, which have a high proportion of hydrophobic to polar amino acids, and therefore have structures that depend on hydrophobic interactions, are particularly sensitive to denaturation at the freezing point (Cheftel et al., 1985). According to Cheftel et al. (1985), soy fractions 11S and 7S have relatively high average hydrophobicity values as calculated using Bigelow’s equation (Bigelow, 1967).

364

Separation of Functional Molecules in Food by Membrane Technology

Consequently, the combination of the differences in low temperature sensitivity and in the isoelectric point of both the 11S and 7S fractions, the main protein soybean fractions, would allow their selective separation by bipolar membrane electroacidification. In the same way, Bazinet et al. (2004) used EDBM for fraction of whey proteins, however, the separation process was based on the ionic strength of the solution as well as the concentration in proteins. Hence at 5% whey protein isolate (WPI) initial concentration, this technology allowed the separation of 98% pure β-LG fraction with a 44.0% recovery yield. At 10% WPI initial concentration, EDBM allowed the production of a β-LG-enriched fraction containing 97.3 of β-LG and 2.7% α-LA, for a 98% total protein purity. EDBM has numerous advantages in comparison with alternative methods for protein fractionation. The process can be precisely controlled, as electroacidification rate is regulated by the effective current density in the cell. The in situ generation and reuse of dangerous chemicals for the environment (acids and bases) suppress the drawbacks and the risks linked to the handling, transportation, use, and elimination of these products (Bazinet et al., 2004). However, the main advantage of this technology is the high protein recovery yield, up to 53% in comparison with 33% and 17% respectively from acid whey and sweet whey proteins (Amundson et al., 1982; Slack et al., 1986).

10.3.3 Application to Phospholipids Cheese whey contains 7
10 g/L of proteins and about 2
8 g/L of lipids (Lehmann and Wasen, 1990). These lipids are composed of almost 66% nonpolar lipids and 33% polar lipids. Polar lipids are mainly phospholipids with 34% phosphatidylethanolamine, 31% phosphatidylcholine, 15% sphingomyelins, 12% phosphatidylinositol, and 8% phosphatidylserine (The´odet and Gandemer, 1994). Whey phospholipids have different particularities and functionalities of great interest for the industry and mainly the health sector. According to Guo et al. (2005), new discoveries in biochemistry have demonstrated that phospholipids can play pivotal roles, at least in three major aspects, which are to provide structural assurance for the integrity of membranes and help carry out their functions (Bezrukov, 2000; White et al., 2001), to involve many metabolism-related and neurological diseases (Schlegel et al., 2003), and to regulate basic biological processes as signaling compounds (Hannun et al., 2001; Irvine and Schell, 2001). EDBM was used for acidification and decreasing the ionic strength of cheddar cheese whey to separate phospholipids (Lin Teng Shee et al., 2005). In this study, EDBM process was carried out with or without preliminary decrease of whey mineral salts content by ED. A 32.1% whey lipid precipitation was obtained without ED (54% increase of lipids precipitation level in comparison with a centrifugation step alone) while the demineralization step

Electrodialysis-Based Separation Technologies Chapter | 10

365

prior to electroacidification had only small effect on the precipitation level. Based on these observations, further works were carried out (Lin Teng Shee et al., 2005) to evaluate the effect of protein concentration on lipid precipitation. Hence a WPC with 55% protein on dry basis was used. WPC electroacidification was carried out with or without preliminary demineralization by ED as previously. The effect of the ionic strength on precipitation rates of fats was also evaluated by a sixfold water dilution of the WPC samples. The electroacidification process without dilution of the WPC resulted in a precipitation of 35%
39% of initial lipids content (vs 18% for the centrifugation alone) and confirmed previous results obtained on cheese whey (Lin Teng Shee et al., 2005). However, by additionally decreasing the ionic strength of acidified WPC solutions by dilution of the supernatant, significant drops in protein and lipid levels were observed in comparison with the initial product composition. In a general manner, the acidification by EDBM followed by a dilution resulted in the preservation of more than 90% of the protein present initially. The combination of the bipolar membrane acidification process with the dilution of the WPC allowed the decrease of the WPC lipid contents by 73% and 66% for EDBM/dilution and EDBM/ED/dilution respectively. A dilution step is necessary to increase the lipid precipitation rate. A sixfold dilution contributed to decrease significantly the ionic strength of the medium while the ED process at a demineralization level of 50% was not sufficient to increase lipid precipitation (Lin Teng Shee et al., 2007). The conductivity decreased by nearly 70% for all products because of the sixfold dilution. However, the final conductivities of acidified and diluted samples vary from 0.5 to 0.9 mS/cm in comparison with the control sample (1.1 mS/ cm). These results show that it is necessary to decrease the conductivity of the acidified WPC below 1 mS/cm to obtain a precipitation level of WPC lipids between 45% and 73% (Lin Teng Shee et al., 2007). It also resulted in clarified supernatants with a low level in lipids and with the majority of the proteins present initially. The increase in lipid precipitation was due to a combination of acidification and to a decrease in the ionic strength of the medium affecting the formation of lipids/proteins complexes (De Wit and Klarenbeek, 1978). Acid pH enhances the electrostatic interactions between negatively charged compounds for lipids and positively charged compounds for proteins. The decrease in mineral salts, particularly in magnesium and calcium ions, during EDBM and ED 1 EDBM promotes the formation of lipid/protein complexes. Lau et al. (1981) have shown that calcium ions are linked to phospholipids and consequently the binding of calcium with lipids inhibits the formation of lipid/protein complexes. The decrease in mineralization level allows the liberation of ionized zones in phospholipids and thus enhances the lipoprotein complex formation. The EDBM process may be the first step of WPC treatment for whey valorization in various lipid fractions. In comparison with the chemical acidification, the EDBM process has the advantage to offer a continuous

366

Separation of Functional Molecules in Food by Membrane Technology

acidification without salt addition. This could be useful, in an industrial process, to maintain a low ionic strength in the solution prior to or following WPC dilution and to produce a purified WPC with a low ash level. This process would have two advantages, the production of a phospholipid-enriched fraction, which could be used in cosmeceutics and nutraceuticals and a purified (demineralized and defatted) and more valuable protein fraction after concentration of the whey.

10.4 ELECTRODIALYSIS WITH FILTRATION MEMBRANE APPLICATIONS IN FOOD PROCESSING A new technology called ED with ultrafiltration membrane (EDUF) or more generically ED with filtration membrane (EDFM) was developed and patented by Bazinet et al. (2013). This technique couples size exclusion capabilities of porous membranes with the charge selectivity of ED. In this process, a conventional ED cell is used, in which some IE membranes are replaced by filtration ones (microfiltration, UF, nanofiltration); compounds of MWs higher than the membrane cutoff can be separated so as to extend the field of application of ED to biological charged molecules. No pressure is applied in the ED cell, only the charged molecules migrate under the effect of the electric field and the neutral molecules stay theoretically in the feed solution and do not reach or pass the filtration membrane (FM) (Bazinet and Firdaous, 2013). The advantages of this unique technology are numerous: G

G

G G

G

G

Preservation of the commercial value of the nonbioactive fractions for more conventional low-value uses, since only the aimed fraction are extracted and no solvent is used. Double utilization of the unprocessed protein hydrolysate as well as the peptide of interest. High selectivity of the process due to the unique size/charge separation. Reduced membrane fouling since in the ED cell, only charged molecules migrate under the effect of the electric field. The EDFM process can be adapted to very large-scale production simply by stacking of membranes to increase the exchange surface. Can be easily integrated in an existing industrial production line (after the pressure-driven raw separation process step) or can be established in a small industrial unit close to the source of raw material.

This technology constitutes a major breakthrough in purification of bioactive peptides and other fine chemicals, since it stands as a reliable and costeffective separation technique, when compared with traditional methods such as chromatography. The applications are potentially numerous, as already demonstrated by the very different nature of molecules on which the

Electrodialysis-Based Separation Technologies Chapter | 10

367

technology was tested: bioactive peptides, proteins, polysaccharide oligomers, and polyphenols.

10.4.1 Applications to Protein Hydrolysates 10.4.1.1 Separation of Bioactive Peptide Fractions The use of EDFM for bioactive peptide separation was carried out first by Poulin et al. (2006). They performed the separation of a β-LG tryptic hydrolysate adjusted at pH 5.0 or 9.0 for the recovery of cationic or anionic peptides respectively. The EDFM cell, for this simple separation of both charged peptides, was composed of one AEM, one CEM, and one UF membrane with a molecular weight cutoff (MWCO) of 20 kDa. A second experiment involved the simultaneous separation of anionic and cationic peptides from β-LG hydrolysate with pH value adjusted at 5.0, 7.0, and 9.0. The main difference with the previous configurations was the insertion of a second 20kDa UFM MWCO (Fig. 10.5). After RP-HPLC analyses, the authors confirmed that EDFM appeared to be a selective method of separation since among a total of 24 identified peptides in the raw hydrolysate, only 13 were recovered in the cationic and anionic peptide-enriched compartments. Among these 13 peptides, the authors detected 3 acid peptides that migrated

A–RC

NaCl

β-LG hydrolysate

C+RC

AEM UFM UFM CEM +

Recovery of anionic peptides

–

Recovery of cationic peptides

AEM UFM UFM CEM P– Na+

P+

P–

+ + Na

P+ P–

Cl–

P–

Cl–

P–

+ P+ + P+/–

Na+ Na+

P+

Cl–

P+

Cl–

–

FIGURE 10.5 Electrodialytic configuration used for the simultaneous separation of anionic and cationic peptides. AEM, Anion exchange membrane; A2RC, anionic peptide recovery compartment; CEM, cation exchange membrane; C1RC, cationic peptide recovery compartment; UFM, ultrafiltration membrane; β-LG, β-lactoglobulin protein.

368

Separation of Functional Molecules in Food by Membrane Technology

only in the anionic peptide-enriched compartment, while 3 basic peptides migrated only in the cationic peptide-enriched compartment. Finally, in the cationic enriched-peptide compartment, the bioactive peptide-lactokinin (
LG 142
148), an antihypertensive peptide, reached a migration rate of 10.5%. Recently, the lactokinin migration was drastically increased to reach up to 66% in still nonoptimal conditions (Doyen et al., 2013). Since 2006, this innovative technology has been used among others to purify an antihypertensive peptide from an alfalfa white protein hydrolysate (Firdaous et al., 2009), to separate an anionic antimicrobial peptide from snow crab byproduct hydrolysate (Doyen et al., 2013), to produce peptide fractions having antidiabetic and/or antihypertensive effect (Doyen et al., 2014; Roblet et al., 2014, 2016), to purify a peptide fraction enriched in histidine having anticancer activities on different cancerous cell lines (Doyen et al., 2011), and to produce an antimicrobial and antioxidant peptide fraction used for increasing meat shelf life.

10.4.1.2 Improvement of Protein Hydrolysate Bioactivity Numerous studies have reported the potential beneficial effects of bioactive peptides recovered from different food sources including fish (van Woudenbergh et al., 2009), some of which reduced the symptoms of diabetes and related medical complications (Anderson et al., 1998; Ishihara et al., 2003). Hence, in a study published in 2016 on salmon frame protein hydrolysate, the authors observed that the final feed compartment (FFC) after EDFM treatment at pH 6 was able to significantly enhance glucose uptake in L6 muscle cells grown in tissue culture (Roblet et al., 2016). The reported bioactivity of FFC at pH 6 would be due to one or more peptides in the MW range of 301
400 Da lowered by 5% in FFC (inhibiting effect) or because of other peptides or their fractions in the MW range of 601
700 Da being higher by 3.4% in FFC (enhancing effect) at same pH conditions enhancing its bioactivity as compared with initial feed compartment (IFC) containing the salmon protein hydrolysate. The maximum enhancement of glucose uptake (40%) was demonstrated in cells treated with the FFC fraction at 1 ng/mL without insulin while a synergetic effect of FFC peptides at 1 ng/ mL with insulin was observed (enhancement of glucose uptake by 31%). Higher bioactivity demonstrated for FFC at pH 6 as compared with IFC raise two hypotheses: 1. The peptides demonstrating inhibiting effect on bioactive peptides are present in IFC and consequently EDFM is a convenient way to isolate those inhibitor peptides that may pass through the UF membrane amplifying therefore antidiabetic activity of FFC. 2. The antidiabetic enhancing peptides are present in IFC and their relative abundance was increased after EDFM treatment in FFC enhancing its bioactivity as compared with IFC.

Electrodialysis-Based Separation Technologies Chapter | 10

369

These results are of great interest since salmon frames, which are byproducts, could be a promising source and a natural alternative of bioactive peptide fractions to be incorporated in functional foods or used as nutraceuticals to help prevent type 2 diabetes (Pilon et al., 2010; Chevrier et al., 2015).

10.4.1.3 Simultaneous Hydrolysis and Separation of Peptides Very recently, Doyen et al. (2013) and Suwal et al. (2017) investigated the feasibility of using an electrodialytic membrane bioreactor for
LG hydrolysis. The hydrolysis of β-LG protein and the fractionation of generated peptides were performed in one step in an EDUF stacked. The enzymatic hydrolysis was started by the addition of trypsin solution and at the same time by the application of a constant electric field of 14 V/cm between EDUF electrodes. During hydrolysis, the reaction was maintained at pH 7.8, corresponding to the optimum pH value of trypsin. After 240 minutes of treatment, among the 31 peptides detected in the β-LG hydrolysate, 15 and 4 peptides were detected respectively, in the anionic (A2CR) and cationic (C1CR) peptide recovery compartments. Chromatograms of the control β-LG hydrolysate obtained after 240 minutes of trypsin hydrolysis (Fig. 10.6) were used to estimate peptide migration rates in A2CR and C1CR peptide recovery compartments. The authors reported that the peptide migration rates increased continuously as a function of time for both recovery compartments. Among the 15 peaks detected in the A2CR, peptide sequences IDALNENK (β-LG

FIGURE 10.6 Respective peptide profiles in each compartment at the end of the process. AEM, Anion exchange membrane; CEM, cation exchange membrane; E, trypsin enzyme; P1, cationic peptide; P2, anionic peptide; UFM, ultrafiltration membrane; β-LG, β-lactoglobulin protein.

370

Separation of Functional Molecules in Food by Membrane Technology

84
91) and VAGTWYSLAMAASDISLLDAQSAPLR (β-LG 15
40) were previously identified as antihypertensive (Chobert et al., 2005). Peptide sequence GLDIQK (β-LG 9
14) was identified as hypocholesterolemic (Nagaoka et al., 2001) and antihypertensive (Pihlanto-Leppala et al., 1998). Peptide sequence IIAEK (β-LG 71
75) was identified as a hypocholesterolemic peptide (Nagaoka et al., 2001). Finally, sequence VLVLDTDYKK (β-LG 92
101) was previously identified as an antimicrobial peptide (Pellegrini et al., 2001) and presented a migration rate of 17%. Among the 4 cationic peptides, the peptide sequence ALPMHIR (β-LG 142
148), identified as β-lactokinin and known to exert an important antihypertensive effect, was recovered with an estimated 66% migration rate. To our knowledge, it was the first time that hydrolysis was conducted under an electric field to simultaneously separate anionic and cationic peptides produced. To complete this work, tryptic hydrolysis of WPI was performed simultaneously during (in situ) and before (ex situ) fractionation by ED with UF membrane to obtain bioactive peptides (Suwal et al., 2017). Peptide migration to anionic (A2RC) and cationic (C1RC) peptide recovery compartments was strongly dependent on the digestion strategy used. Indeed, peptide migration to the A2RC was observed to be higher with in situ digestion while peptide migration to the C1RC was higher in an ex situ digestion: a final peptide concentration of 103.10 6 2.76 μg/mL was found in the C1RC (ex situ) while it was 49.65 6 6.13 μg/mL in the A2RC (in situ). HPLC-MS studies showed 23 major peaks that were generated by tryptic digestion of WPI. Seven of these peptides migrated to the A2RC while nine and eight peptides migrated to the C1RC for ex situ and in situ digestions respectively. Among the peptides recovered in anionic compartment one peptide (IDALNENK) is known to have hypocholesterolemic effect and one (VYVEELKPTPEGDLEILLQK) to have antihypertensive effect. The EDFM technique also separated cationic peptides such as VLVLDTDYKK and VAGTWY, which are peptides having antimicrobial activities. In addition, in cationic recovery compartments peptides with amino acid sequences ALPMHIR and TKIPAVFK are well known to possess antihypertensive and hypocholesterolemic effects respectively.

10.4.2 Separation of Protein Bovine lactoferrine (bLF) is a 703-amino acid glycoprotein and an 80 kDa MW protein originally isolated from milk (Guo et al., 2011; Ndiaye et al., 2010). In previous works, bLF has already demonstrated antimicrobial, antiinflammatory, immunostimulatory, and anticancer properties (Arnold et al., 2002; Mohan et al., 2007). Consequently, due to bLF large potential applications, Ndiaye et al. (2010) studied the feasibility of bLF electroseparation by EDFM to isolate and concentrate high purified fractions. The EDFM cell was composed of 1 AEM, 1 CEM, and a 1500-kDa polyethersulfone (PES) UFM MWCO. Two different solutions were fractionated. Indeed, the first

Electrodialysis-Based Separation Technologies Chapter | 10

371

solution was a model solution that contained 0.1% of bLF and the second solution was composed of a bLF-enriched whey solution. The authors reported that after 240 minutes of bLF model solution fractionation by EDFM under constant voltage difference of 20 V, a total bLF migration rate of 46% was obtained in the cationic bLF-enriched compartment. Concerning the bLF-enriched whey solution, the migration rate obtained after 240 minutes of EDFM fractionation reached a maximal value of 14.6% at pH 3.0, corresponding to a transport rate of 8.9 g/m2 h. This difference in migration rates was explained by the presence, in the whey solution, of other proteins competing with bLF for the migration and/or by a decrease of bLF electrophoretic mobility.

10.4.3 Separation of Chitosan Oligomers Chitosan oligomers are a succession of glucosamine molecules that are widely used as bioactive molecules to promote the formation and repairing of cartilage (Francis Suh and Matthew, 2000) and, to reduce the progression of osteoarthritis (Phoon and Manolios, 2002). Moreover, oligomers of higher MW had already demonstrated antitumoral, immunostimulating, antifungal, and antimicrobial activities (Shahidi et al., 1999). Enzymatic hydrolysis of the chitosan by an enzyme such as a chitosanase is generally performed in industry to produce oligomers of desirable range of polymerization. However, after hydrolysis, the mixture recovered is composed of oligomers with different MWs. Thus, to separate the different chitosan oligomer fractions, Aider et al. (2008) used EDFM to obtain more purified fractions. The experiments consisted of studying the effect of UFM MWCOs (500, 1000, 5000, 10,000, and 20,000 Da) stacked in the EDFM system for the electroseparation of a mixture of three chitosan oligomers (dimer, trimer, and tetramer) at pH 4.0, the maximal chitosan electrophoretic mobility (Aider et al., 2008). Indeed, at pH 4.0, chitosan oligomers are positively charged and consequently migrated toward the cathode. The process duration was fixed for 240 minutes under constant voltage difference of 5 V. Aider et al. (2008) reported that MWCO process duration and chitosan oligomers’ chain length had a significant impact on electromigration of chitosan oligomers. Indeed, a 500-Da UFM MWCO did not allow the migration of chitosan whatever the oligomers tested. The 1000-Da UFM MWCO allowed only the dimer and tetramer migration during the first 180-minute EDFM treatment, whereas the tetramer was only detected after 240 minutes. An UFM with MWCO of 5000 Da allowed the migration of dimer and tetramer as observed for the 1000 Da. However, in this case, the tetramer was detected after 120 minutes of EDFM treatment. UFM MWCO of 10,000 Da allowed the migration of dimer and trimer and after 120 minutes of treatment, and tetramer was detected as well as observed for the 5000 Da UFM MWCO except that the migration rates were more important. With UFM

372

Separation of Functional Molecules in Food by Membrane Technology

MWCO of 20,000 Da stacked in the system, the electromigration rate for all the chitosan oligomers increased as a function of time. Moreover, this EDFM configuration allowed the highest chitosan oligomer recoveries. In a further study, Aider et al. (2009) studied the effect of solution flow velocity and electric field strength on chitosan oligomers to improve the EDFM fractionation. Experiments were the same as those described previously except that only UFM MWCO of 10,000 Da was used. Moreover, three constant electric field strengths (2.5, 5, and 10 V/cm) and three solution flow velocities (2.77, 8.33, and 13.88 cm/s) were tested. The authors observed that the solution flow velocity had no impact on chitosan oligomers’ electroseparation. However, concerning the effect of electric field strength, the authors demonstrated that a value of 2.5 V/cm was sufficient to obtain a solution composed only of the dimer and trimer until an operating time of 120 minutes.

10.4.4 Concentration of Cranberry Juice in Phenolic Compounds Phenolic components contained in cranberry juice (proanthocyanidins and anthocyanidins) are known for their antioxidant and anticancer properties (Duthie et al., 2006; Yan et al., 2002). Moreover, cranberry juice is used in the prevention and treatment of urinary system infections (Jepson and Craig, 2008) and for the treatment of periodontitis (1) by inhibition of oral Streptococci colonization on tooth surfaces and (2) by its capacity to reduce Porphyromonas gingivalis to colonize periodontal sites (Yamanaka et al., 2004). Proanthocyanidins from cranberry juice offer protection against urinary tract infections by inhibiting the adherence of Escherichia coli to human body cells (Foo et al., 2000). Finally, proanthocyanidins inhibit adhesion of Helicobacter pylori, which is responsible of gastric ulcer development (Burger et al., 2002). Consequently, as proanthocyanidins and anthocyanins are charged molecules under specific pH values, Bazinet et al. (2009) used EDFM technology to produce a cranberry juice enriched with natural phenolic antioxidant compounds. Total concentrations of proanthocyanidins and anthocyanins were increased by 34.8% and 52.9% respectively in cranberry juice treated with EDFM system. Moreover, an 18% increase of the antioxidant capacity of the enriched cranberry juice was obtained by the EDFM treatments. However, in this first preliminary work on the subject, it was demonstrated that if used as is, this technology has the disadvantage to generate a raw juice impoverished in antioxidants with an unsatisfactory color and taste. To transpose this technology on an industrial scale for the production of an antioxidant-enriched cranberry juice, the authors proposed the integration of EDFM to the conventional process used for cranberry juice production (Fig. 10.7), in a way avoiding the generation of a source juice impoverished in polyphenols. This integration of EDFM was tested further and demonstrated to be successful (Bazinet et al., 2009).

Electrodialysis-Based Separation Technologies Chapter | 10

373

Cranberry juice bottling unit

Juice with very low variation in antioxidants

Circuit 1

Antioxidantenriched cranberry

Raw cranberry juice

Electrolyte (salt)

Circuit 2

FIGURE 10.7 Schematic flow diagram for the integrated production of antioxidant-enriched cranberry juice by EDFM.

To complete this study different FM with distinct constitutive materials and cutoffs were tested (Husson et al., 2013). The higher enrichment yields in anthocyanins (24%) without any significant depletion of polyphenols from the raw juice were obtained with two FMs: (1) polyvinylidene fluoride with a cutoff of 150 kDa and (2) PES with a cutoff of 500 kDa. It was also demonstrated that the enrichment in anthocyanins was influenced by the FM zeta potential (ZP). Indeed, for FMs presenting positive surface charge (ZP close to 2
3 mV), repulsions would be established with anthocyanins positively charged and would avoid their electromigration through the membrane. On the opposite, FMs with moderate negative charges (ZP close to 25 mV) seem to facilitate the electrotransfer of anthocyanins due to the attraction between opposite charges.

10.5 CONCLUSION Since 1960, ED has been a well-known technology in the food industry, with many plants around the world, but its main applications are still focused on the basic demineralization of food solution under a DC current. A recent

374

Separation of Functional Molecules in Food by Membrane Technology

upcoming application of conventional ED is the tartaric stabilization of wine with a growing market all around the world. With the development of bipolar membrane in the 1980s, new applications were developed over the next 15 years, but the majority of these applications for the food sector had not yet been exploited at the industrial scale and needed further experiments. In the 2000s, a new hybrid ED-based technology, EDMF (ED membrane fouling), was proposed and demonstrated promising potential applications for high added value applications and valorization of byproducts, such as production of peptides fractions having health benefits on human (antidiabetic, anticancer, antimicrobial, antihypertensive, etc.). However, for this last technology, no equipment is actually commercially available at a large scale, but works are currently underway in Canada with an equipment manufacturer. In addition to new membrane manufacturing or membrane stack configurations, the most promising research field that could leverage the development of ED-based technologies is the use of pulsed electric field as a new mode of current, to diminish the potential risks of fouling/scaling, to decrease the concentration polarization phenomenon, and to increase mass transfer or to work over the limiting current density, to improve the mass transfer or benefit from the formation of electroconvective vortices.

ACKNOWLEDGMENTS This project has been supported by the National Natural Science Foundation of China (Nos. 21490581, 91534203, 21476220), International Partnership Program of Chinese Academy of Sciences (No. 21134ky5b20170010), National High Technology Research and Development Program 863 (No. 2015AA021001), and K.C. Wong Education Foundation (2016-11). This work was also supported by the NSERC Industrial Research Chair on Electromembrane Processes, aiming toward the ecoefficiency improvement of biofood production lines (Grant IRCPJ 492889-15 to Laurent Bazinet( and the NSERC Discovery Grants Program (Grant SD RGPIN-201804128 to Laurent Bazinet).

REFERENCES Adhikary, S.K., Harkare, W.P., Gowindan, K.P., Nanjundaswamy, A.M., 1983. Deacidification of fruit juices by electrodialysis. Indian J. Technol. 21, 120
123. Aider, M., Brunet, S., Bazinet, L., 2008. Electroseparation of chitosan oligomers by electrodialysis with ultrafiltration membrane (EDUF) and impact on electrodialytic parameters. J. Memb. Sci. 309 (1
2), 222
232. Aider, M., Brunet, S., Bazinet, L., 2009. Effect of solution flow velocity and electric field strength on chitosan oligomer electromigration kinetics and their separation in an electrodialysis with ultrafiltration membrane (EDUF) system. Sep. Purif. Technol. 69 (1), 63
70. Akpinar-Bayizit, A., Ozcan, T., Yilmaz-Ersan, L., 2009. Membrane processes in production of functional whey components. Mljekarstvo 59 (4), 282
288. Ali, F., Ippersiel, D., Lamarche, F., Mondor, M., 2010. Characterization of low-phytate soy protein isolates produced by membrane technologies. Innov. Food Sci. Emerg. Technol. 11 (1), 162
168.

Electrodialysis-Based Separation Technologies Chapter | 10

375

Ali, F., Mondor, M., Ippersiel, D., Lamarche, F., 2011. Production of low-phytate soy protein isolate by membrane technologies: impact of salt addition to the extract on the purification process. Innov. Food Sci. Emerg. Technol. 12 (2), 171
177. Amundson, C.H., Watanawanichakorn, S., Hill, C.G., 1982. Production of enriched protein fractions of β-lactoglobulin and α-lactalbumin from cheese whey. J. Food Process. Preservat. 6, 55
71. Anderson, J., Blake, J., Turner, J., Smith, B., 1998. Effects of soy protein on renal function and proteinuria in patients with type 2 diabetes. Am. J. Clin. Nutr. 68 (6), 1347S
1353S. Andres, Z., Riera, F.A., Alvarez, R., 1995. Skimmed milk demineralization by electrodialysis: conventional versus selective membranes. J. Food Eng. 26, 57
66. Arnold, D., Di Biase, A.M., Marchetti, M., Pietrantoni, A., Valenti, P., Seganti, L., et al., 2002. Antiadenovirus activity of milk proteins: lactoferrin prevents viral infection. Antiviral. Res. 53 (2), 153
158. Balster, J., Pu¨nt, I., Stamatialis, D.F., Lammers, H., Verver, A.B., Wessling, M., 2007. Electrochemical acidification of milk by whey desalination. J. Memb. Sci. 303, 213
220. Bazinet, L., 2005. Electrodialytic phenomena and their applications in the dairy industry: a review. Crit. Rev. Food. Sci. Nutr. 44, 525
544. Bazinet, L., Firdaous, L., 2013. Separation of bioactive peptides by membrane processes: technologies and devices. Recent Patents Biotechnol. 7 (1), 9
27. Bazinet, L., Lamarche, F., Labrecque, R., Toupin, R., Boulet, M., Ippersiel, D., 1997a. Electroacidification of soybean proteins for production of isolate. Food Technol. 51 (9), 52
56. Bazinet, L., Lamarche, F., Labrecque, R., Ippersiel, D., 1997b. Effect of KCl and soy protein concentrations on the performance of bipolar membrane electroacidification. J. Agric. Food. Chem. 45 (7), 2419
2425. Bazinet, L., Lamarche, F., Ippersiel, D., 1998. Comparison of chemical and bipolar-membrane electrochemical acidification for precipitation of soybean proteins. J. Agric. Food. Chem. 46 (5), 2013
2019. Bazinet, L., Lamarche, F., Ippersiel, D., 1999a. Ionic balance: a closer look at the K 1 migrated and H 1 generated during bipolar membrane electro-acidification of soybean proteins. J. Memb. Sci. 154 (1), 61
71. Bazinet, L., Lamarche, F., Ippersiel, D., Amiot, J., 1999b. Bipolar membrane electroacidification to produce bovine milk casein isolate. J. Agric. Food. Chem. 47 (12), 5291
5296. Bazinet, L., Lamarche, F., Ippersiel, D., Gendron, C., Mahdavi, B., Amiot, J., 2000a. Comparison of electrochemical and chemical acidification of skim milk. J. Food Sci. 65 (8), 1303
1307. Bazinet, L., Ippersiel, D., Labrecque, R., Lamarche, F., 2000b. Effect of temperature on the separation of soybean 11S and 7S protein fractions during bipolar membrane electroacidification. Biotechnol. Prog. 16, 292
295. Bazinet, L., Gendron, C., Ippersiel, D., Rene´-Paradis, J., Te´treault, C., Beaudry, J., et al., 2002. Effect of type of added salt and ionic strength on physicochemical and functional properties of casein isolates produced by electroacidification. J. Agric. Food Chem. 50 (23), 6875
6881. Bazinet, L., Ippersiel, D., Mahdavi, B., 2004. Fractionation of whey protein by bipolar membrane electroacidification. Innovat. Food Sci. Emerg. Technol. 5, 17
25. Bazinet, L., Cossec, C.L., Gaudreau, Hl.N., Desjardins, Y., 2009. Production of a phenolic antioxidant enriched cranberry juice by electrodialysis with filtration membrane. J. Agric. Food. Chem. 57 (21), 10245
10251.

376

Separation of Functional Molecules in Food by Membrane Technology

Bazinet, L., Amiot, J., Poulin, J.F., Tremblay, A., Labbe´, D., 2013. Process and system for separation of organic charged compounds. Canadian Patent 2558,164. Benitez, R.M., Ortero, G.M., 2012. In: Benitez, R.M., Ortero, G.M. (Eds.), Whey: Types, Composition and Health Implications. Nova Science Publishers, Hauppauge, NY, p. c2012. Bezrukov, S.M., 2000. Functional consequences of lipid packing stress. Curr. Opin. Colloid Interface Sci. 5, 237
243. Bigelow, C.C., 1967. On the average hydrophobicity of proteins and the relation between it and protein structure. J. Theor. Biol. 16, 187
211. Bleha, M., Tishenko, G., Sumberova, V., Kudela, V., 1992. Characterization of the critical state f membranes in electrodialysis desalination of milk whey. Desalination 86, 173
186. Bukhovets, A.E., Savel’eva, A.M., Eliseeva, T.V., 2009. Separation of amino acids mixtures containing tyrosine in electromembrane system. Desalination 241 (1
3), 68
74. Burger, O., Weiss, E., Sharon, N., Tabak, M., Neeman, I., Ofek, I., 2002. Inhibition of Helicobacter pylori adhesion to human gastric mucus by a high-molecular-weight constituent of cranberry juice. Crit. Rev. Food. Sci. Nutr. 42 (Suppl. 3), 279
284. Calle, E.V., Ruales, J., Dornierb, M., Sandeauxc, J., Sandeauxc, R., Pourcelly, G., 2002. Deacidification of the clarified passion fruit juice. (P. edulis flavicarpa). Desalination 149, 357
361. Cheftel, J.C., Cuq, J.L., Lorient, D., 1985. Amino-acids, peptides and proteins. In: Fenema, O.R. (Ed.), Food Chemistry: 2nd Edition Revised and Expended. Marcel Dekker Inc., New York. Chen, D., Wang, S., Huang, T., 1995. Separation of phenylacetic acid, 6 aminopenicillanic acid and penicillin G with electrodialysis under constant current. J. Chem. Technol. Biotechnol. 64 (3), 284. Chen, G.Q., Eschbach, F.I.I., Weeks, M., Gras, S.L., Kentish, S., 2016. Removal of lactic acid from acid whey using electrodialysis. Sep. Purif. Technol. 158, 230
237. Chevrier, G., Mitchell, P.L., Rioux, L.E., Hasan, F., Jin, T., Roblet, C.R., et al., 2015. Prevention of obesity-linked glucose intolerance, inflammation and dyslipidemia in LDLR 2 / 2 ApoB100/100 mice by low molecular weight peptides isolated from salmon protein. J. Nutr. 145, 1415
1422. Chico´n, R., Belloque, J., Alonso, E., Lo´pez-Fandino, R., 2009. Antibody binding and functional properties of whey protein hydrolysates obtained under high pressure. Food Hydrocolloids. 23 (3), 593
599. Chindapan, N., Devahastin, S., Chiewchan, N., 2009. Electrodialysis desalination of fish sauce: electrodialysis performance and product quality. J. Food. Sci. 74, E363
E371. Chindapan, N., Devahastin, S., Chiewchan, N., Sablani, S.S., 2011. Desalination of fish sauce by electrodialysis: effect on selected aroma compounds and amino acid compositions. J. Food Sci 76, S451
S457. Chindapan, N., Devahastin, S., Chiewchan, N., Sablani, S.S., 2013. Modeling and optimization of electrodialytic desalination of fish sauce using artificial neural networks and genetic algorithm. Food Bioprocess Technol. 6, 2695
2707. Chobert, J.M., El-Zahar, K., Sitohy, M., Dalgalarrondo, M., Me´tro, F., Choiset, Y., et al., 2005. Angiotensin I-converting-enzyme (ACE)-inhibitory activity of tryptic peptides of ovine β-lactoglobulin and of milk yoghurts obtained by using different starters. Lait 85, 141
152. Couture, R., Rouseff, R., 1992. Debittering and deacidifying sour orange (Citrus aurantium) juice using neutral and anion exchange resins. J. Food. Sci. 57, 380
384. Daufin, G., Escudier, J.-P., Carre`re, S., Be´rot, S., Fillaudeau, L., Decloux, M., 2001. Recent and emerging applications of membrane processes in the food and dairy industry. Trans. IChemE 79 (Part C), 89
102.

Electrodialysis-Based Separation Technologies Chapter | 10

377

De Wit, J.N., Klarenbeek, G., 1978. Method for the clarification of liquids containing whey protein. Great Britain Patent 1519,897. Doyen, A., Beaulieu, L., Saucier, L., Pouliot, Y., Bazinet, L., 2011. Demonstration of in vitro anticancer properties of peptide fractions from a snow crab by-products hydrolysate after separation by electrodialysis with ultrafiltration membranes. Sep. Purif. Technol. 78, 321
329. Doyen, A., Husson, E., Bazinet, L., 2013. Use of an electrodialytic reactor for the simultaneous β-lactoglobulin enzymatic hydrolysis and fractionation of generated bioactive peptides. Food. Chem. 136, 1193
1202. Doyen, A., Udenigwe, C.C., Mitchell, P., Marette, A., Aluko, R.E., Bazinet, L., 2014. Antidiabetic and antihypertensive activities of two flaxseed protein hydrolysate fractions revealed following their simultaneous separation by electrodialysis with ultrafiltration membranes. Food. Chem. 145, 66
76. Dunsford, P., Boulton, R., 1981. The kinetics of potassium bitatrate crystallization from table wines. I. Effect of particle size, particle surface area and agitation. Am. J. Enol. Viticult. 32, 100
105. Duthie, S., Jenkinson, A., Crozier, A., Mullen, W., Pirie, L., Kyle, J., et al., 2006. The effects of cranberry juice consumption on antioxidant status and biomarkers relating to heart disease and cancer in healthy human volunteers. Eur. J. Nutr. 45 (2), 113
122. Fidaleo, M., Moresi, M., Cammaroto, A., Ladrange, N., Nardi, R., 2012. Soy sauce desalting by electrodialysis. J. Food Eng. 110, 175
181. Firdaous, L., Dhulster, P., Amiot, J., Gaudreau, A., Lecouturier, D., Kapel, R., et al., 2009. Concentration and selective separation of bioactive peptides from an alfalfa white protein hydrolysate by electrodialysis with ultrafiltration membranes. J. Memb. Sci. 329 (1
2), 60
67. Fisher, R.R., Glatz, C.E., Murphy, P.A., 1986. Effects of mixing during acid addition on fractionally precipitated protein. Biotech. Bioeng. 28, 1056
1063. Foo, L.Y., Lu, Y., Howell, A.B., Vorsa, N., 2000. The structure of cranberry proanthocyanidins which inhibit adherence of uropathogenic P-fimbriated Escherichia coli in vitro. Phytochemistry 54 (2), 173
181. Francis Suh, J.K., Matthew, H.W.T., 2000. Application of chitosan-based polysaccharide biomaterials in cartilage tissue engineering: a review. Biomaterials. 21 (24), 2589
2598. Gonc¸alvesa, F., Fernandes, C., Cameira dos Santos, P., Norberta de Pinho, M., 2003. Wine tartaric stabilization by electrodialysis and its assessment by the saturation temperature. J. Food Eng. 59, 229
235. Greiter, M., Novalin, S., Wendland, M., Kulbe, K.D., Fischer, J., 2002. Desalination of whey by electrodialysis and ion-exchange resins: analysis of both processes with regard to sustainability by calculating their cumulative energy demand. J. Membr. Sci. 210, 91
102. Guo, M., Zhang, L.Y., Lu¨, W.J., Cao, H.R., 2011. Analysis of the spectroscopic characteristics on the binding interaction between tosufloxacin and bovine lactoferrin. J. Luminescence 131 (4), 768
775. Guo, Z., Vikbjerg, A.F., Xu, X., 2005. Enzymatic modification of phospholipids for functional applications and human nutrition. Biotechnol. Adv. 23 (3), 203
259. Hannun, Y.A., Luberto, C., Argraves, K.M., 2001. Enzymes of sphingolipid metabolism: from modular to integrative signaling. Biochemistry 40, 4893
4903. Huang, C.H., Xu, T.W., Zhang, Y.P., Xue, Y.H., Chen, G.W., 2007. Application of electrodialysis to the production of organic acids: state-of-the-art and recent developments. J. Membr. Sci. 288, 1
12.

378

Separation of Functional Molecules in Food by Membrane Technology

Huang, C.H., Xu, T.W., 2006. Electrodialysis with bipolar membranes for sustainable development. Environ. Sci. Technol. 2006 (40), 5233
5243. Husson, E., Araya-Farias, M., Desjardins, Y., Bazinet, L., 2013. Selective anthocyanins enrichment of cranberry juice by electrodialysis with ultrafiltration membranes stacked. Innovat. Food Sci. Emerg. Technol. 17, 153
162. Irvine, R.F., Schell, M.J., 2001. Back in the water: the return of the inositol phosphates. Nat. Rev. Mol. Cell Biol. 2, 327
338. Ishihara, K., Oyaizu, S., Fukuchi, Y., Mizunoya, W., Segawa, K., Takahashi, M., et al., 2003. A soybean peptide isolate diet promotes postprandial carbohydrate oxidation and energy expenditure in type II diabetic mice. J. Nutr. 133 (3), 752
757. Jepson, R., Craig, J.C., 2008. Cranberries for preventing urinary tract infections. Cochrane Database Syst. Rev. Available from: http://dx.doi.org/10.1002/14651858. Kattan Readi, O.M., Girone`s, M., Nijmeijer, K., 2013a. Separation of complex mixtures of amino acids for biorefinery applications using electrodialysis. J. Membr. Sci. 429, 338
348. Kattan Readi, O.M., Girone`s, M., Wiratha, W., Nijmeijer, K., 2013b. On the isolation of single basic amino acids with electrodialysis for the production of biobased chemicals. Ind. Eng. Chem. Res. 52 (3), 1069
1078. Kilara, A., Sharkasi, T.Y., 1986. Effects of temperature on food proteins and its implications on functional properties. Crit. Rev. Food Sci. Nutr. 23, 323
395. Kinsella, J.E., Whitehead, D.M., 1989. Proteins in whey: chemical, physical, and functional properties. Adv. Food. Nutr. Res. 33, 343
438. Kokubu, T., Yamauchi, T., Miyagi, S., 1983. Application of electrodialysis in sugar industry. Food Ind. 9, 26
31. Lam Quoc, A., Lamarche, F., Makhlouf, J., 2000. Acceleration of pH variation in cloudy apple juice using electrodialysis with bipolar membranes. J. Agric. Food. Chem. 48, 2160
2166. Lasanta, C., Go´mez, J., 2012. Tartrate stabilization of wines. Trends Food Sci. Technol. 28, 52
59. Lau, A., McLaughlin, A., McLaughlin, S., 1981. The adsorption of divalent cations to phosphatidylglycerol bilayer membranes. Biochim. Biophys. Acta 645, 279
292. Lehmann, H., Wasen, D., 1990. Method of dephospholipidating whey. US Patent 4,897,279. Lin Teng Shee, F., Angers, P., Bazinet, L., 2005. Precipitation of cheddar cheese whey lipids by electrochemical acidification. J. Agric. Food Chem. 53, 5635
5639. Liu, Y.H., Chang, C.T., Hung, M.Y., et al., 2014. Probiotics-containing soybean oligosaccharide product and preparation thereof. US Patent 8,877.269 B2. Madureira, A.R., Pereira, C.I., Gomes, A.M.P., Pintado, M.E., Malcata, F.X., 2007. Bovine whey proteins —overview on their main biological properties. Food Res. Int. 40 (10), 1197
1211. McMurdo, M.E.T., Bissett, L.Y., Price, R.J.G., Phillips, G., Crombie, I.K., 2005. Age Ageing 34, 256
261. Mier, M.P., Iban˜ez, R., Ortiz, I., 2008. Influence of process variables on the production of bovine milk casein by electrodialysis with bipolar membranes. Biochem. Eng. J. 40, 304
311. Mikhaylin, S., Bazinet, L., 2015. Electrodialysis in food processing. Reference Module in Food Science. Elsevier, pp. 1
8. Available from: http://dx.doi.org/10.1016/B978-0-08-1005965.03116-4. Mikhaylin, S., Nikonenko, V., Pourcelly, G., Bazinet, L., 2016. Hybrid bipolar membrane electrodialysis-ultrafiltration technology assisted by pulsed electric field for casein production. Green Chem. 18, 307
314.

Electrodialysis-Based Separation Technologies Chapter | 10

379

Mikhaylin, S., Patouillard, L., Margni, M., Bazinet, L., 2018. Protein production by bipolar membrane electrodialysis coupled with ultrafiltration: a sustainable process responding to the environmental issues of the growing human population. Green Chem. 20, 449
456. Mohan, K.V.P.C., Gunasekaran, P., Varalakshmi, E., Hara, Y., Nagini, S., 2007. In vitro evaluation of the anticancer effect of lactoferrin and tea polyphenol combination on oral carcinoma cells. Cell. Biol. Int. 31 (6), 599
608. Mok, C., Park, Y.S., Lee, K.H., Chung, J.H., Woo, G.J., Lim, S., 2001. Desalination of soy sauce by electrodialysis, 2001 IFT Annual Meeting, New Orleans, LA. Montiel, V., Garcı´a-Garcı´a, V., Gonza´lez-Garcı´a, J., Carmona, F., Aldaz, A., 1998. Recovery by means of electrodialysis of an aromatic amino acid from a solution with a high concentration of sulphates and phosphates. J. Membr. Sci. 140, 243
250. Nagaoka, S., Futamura, Y., Miwa, K., Awano, T., Yamauchi, K., Kanamaru, Y., et al., 2001. Identification of novel hypocholesterolemic peptides derived from bovine milk β-lactoglobulin. Biochem. Biophys. Res. Commun. 281, 11
17. Nagarale, R.K., Gohil, G.S., Shahi, V.K., 2006. Recent developments on ion-exchange membranes and electro-membrane processes. Adv. Colloid. Interface. Sci. 119, 97
130. Nash, A.M., Wolf, W.J., 1967. Solubility and ultracentrifugal studies on soybean globulins. J. Cereal Chem. 44, 183
192. Ndiaye, N., Pouliot, Y., Saucier, L., Beaulieu, L., Bazinet, L., 2010. Electroseparation of bovine lactoferrin from model and whey solutions. Sep. Purif. Technol. 74 (1), 93
99. Okada, K., Tomita, M., Tamura, Y., 1975. Electrodialysis in the treatment of dairy products. Symposium Separation Processes by Membranes, Ion-Exchange and Freeze Concentration in Food Industry, Paris, March 13
14. Pellegrini, A., Dettling, C., Thomas, U., Hunziker, P., 2001. Isolation and characterization of four bactericidal domains in the bovine β-lactoglobulin. Biochim. Biophys. Acta. 1526, 131
140. Phoon, S., Manolios, N., 2002. Glucosamine. A nutraceutical in osteoarthritis. Aust. Fam. Physician 31 (6), 539
541. Pihlanto-Leppala, A., Rokka, T., Korhonen, H., 1998. Angiotensin I converting enzyme inhibitory peptides derived from bovine milk proteins. Int. Dairy J. 8, 325
331. Pilon, G., Ruzzin, J., Rioux, L.E., Lavinge, C., Frøyland, L., Jacques, H., et al., 2010. Differential effects of various fish proteins in altering body weight, adiposity, inflammatory status, and insulin sensitivity in high-fat-fed rats. Metabolism 12, 005. Poulin, J.F., Amiot, J., Bazinet, L., 2006. Simultaneous separation of acid and basic bioactive peptides by electrodialysis with ultrafiltration membrane. J. Biotechnol. 123 (3), 314
328. Ran, J., Wu, L., He, Y.B., Wang, Y.M., Jiang, C.X., Ge, L., et al., 2017. Ion exchange membranes: new developments and applications. J. Membr. Sci. 522, 267
291. Raz, R., Chazan, B., Dan, M., 2004. Cranberry juice and urinary tract infection. Clin. Infect. Dis. 38 (10), 1413
1419. Roblet, C., Amiot, J., Pilon, G., Marette, A., Bazinet, L., 2014. Enhancement of glucose uptake in muscular cell by soybean charged peptides isolated by EDUF: activation of the AMPK pathway. Food. Chem. 147, 124
130. Roblet, C., Akhtar, M.J., Mikhaylin, S., Pilon, G., Gill, T., Marette, A., et al., 2016. Enhancement of glucose uptake in muscular cell by peptide fractions separated by electrodialysis with filtration membrane from salmon frame protein hydrolysate. J. Funct. Food 22, 337
346. Romanov, A.M., Zelentsov, V.I., 2007. Use of electrodialysis for the production of grape-based soft and alcoholic drinks. Surface Eng. App. Electrochem. 43, 279
286.

380

Separation of Functional Molecules in Food by Membrane Technology

Schlegel, R.A., Halleck, M.S., Williamson, P., 2003. Phospholipid transporters in the brain. Nutrition and Biochemistry of Phospholipids. AOCS Press, Champaign, IL, pp. 1
13. Serre, E., Rozoy, E., Pedneault, K., Lacour, S., Bazinet, L., 2016a. Deacidification of cranberry juice by electrodialysis: impact of membrane types and configurations on acid migration and juice physicochemical characteristics. Sep. Purif. Technol. 163, 228
237. Serre, E., Boutin, Y., Langevin, M., Lutin, F., Pedneault, K., Lacour, S., et al., 2016b. Deacidification of cranberry juice protects against disruption of in-vitro intestinal cell barrier integrity. J. Funct. Foods. 26, 208
216. Shahidi, F., Arachchi, J.K.V., Jeon, Y.J., 1999. Food applications of chitin and chitosans. Trends Food Sci. Technol. 10 (2), 37
51. Lin Teng Shee, F., Angers, P., Bazinet, L., 2005. Precipitation of cheddar cheese whey lipids by electrochemical acidification. J. Agric. Food. Chem. 53, 5635
5639. Lin Teng Shee, F., Angers, P., Bazinet, L., 2007. Delipidation of a whey protein concentrate by electroacidification with bipolar membranes (BMEA). J. Agric. Food. Chem. 55, 3985
3989. Slack, A.W., Amundson, C.H., Hill, C.G., 1986. Production of enriched β-lactoglobulin and α-lactalbumin whey protein fractions. J. Food Process. Preservat. 10, 19
30. Smithers, G.W., 2008. Whey and whey proteins—from “gutter-to-gold”. Int. Dairy J. 18 (7), 695
704. Soares, P.A.M.H., Geraldes, V., Fernandes, C., dos Santos, P.C., de Pinho, M.N., 2009. Wine tartaric stabilization by electrodialysis: prediction of required deionization degree. Am. J. Enol. Vitic. 60, 183
188. Strathmann, H., 2010. Electrodialysis, a mature technology with a multitude of new applications. Desalination 264, 268
288. Suwal, S., Rozoy, E´., Seifu Manenda, M., Doyen, A., Bazinet, L., 2017. Comparative study of in-situ and ex-situ enzymatic hydrolysis of milk protein and separation of bioactive peptides in an electromembrane reactor. ACS Sustain. Chem. Eng. 5 (6), 5330
5340. The´odet, C., Gandemer, G., 1994. Devenir des lipides au cours de la clarification du lactose´rum. Lait 74, 281
295. Tronc, J.-S., 1996. Inhibition des re´actions de brunissement enzymatique du jus de pomme frais non clarifie´ par e´lectrodialyse’. The`se de Maıˆtrise n 14431. Universite´ Laval, Sainte-Foy, Que´bec. Tronc, J.-S., Lamarche, F., Makhlouf, J., 1997. Enzymatic browning inhibition in cloudy apple juice by electrodialysis. J. Food. Sci. 1, 75
78. Vasileiou, I., Katsargyris, A., Theocharis, S., Giaginis, C., 2013. Current clinical status on the preventive effects of cranberry consumption against urinary tract infections. Nutr. Res. 33 (8), 595
607. Vera, E., Ruales, J., Dornier, M., Sandeaux, J., Sandeaux, R., Pourcelly, G., 2003a. Deacidification of the clarified passion fruit juice using different configurations of electrodialysis. J. Chem. Technol. Biotechnol. 78, 918
925. Vera, E., Ruales, J., Dornier, M., Sandeaux, J., Persin, F., Pourcelly, G., et al., 2003b. Comparison of different methods for deacidification of clarified passion fruit juice. J. Food Eng. 59, 361
367. Vera, E., Sandeaux, J., Persin, F., Pourcelly, G., Dornier, M., Piombo, G., et al., 2007a. Deacidification of clarified tropical fruit juices by electrodialysis. Part II. Characteristics of the deacidified juices. J. Food Eng. 78, 1439
1445. Vera, E., Sandeaux, J., Persin, F., Pourcelly, G., Dornier, M., Ruales, J., 2007b. Deacidification of clarified tropical fruit juices by electrodialysis. Part I. Influence of operating conditions on the process performances. J. Food Eng. 78, 1427
1438.

Electrodialysis-Based Separation Technologies Chapter | 10

381

Wang, Y.M., Pan, S.D., Xu, T.W., 2015. Electrodialysis for production of xylitol from acid hydrolysis method. CIESC J. 66 (9), 3529
3534. Wang, Y.M., Zhang, Z., Xu, T.W., 2016. Recovery of gamma-aminobutyric acid (GABA) from reaction mixtures containing salt by electrodialysis. Sep. Purif. Technol. 170, 353
359. White, S.H., Ladokhin, A.S., Jayasinghe, S., Hristova, K., 2001. How membrane shape protein structure. J. Biol. Chem. 276, 32395
32398. Wing, D.A., Rumney, P.J., Preslicka, C.W., Chung, J.H., 2008. Daily cranberry juice for the prevention of asymptomatic bacteriuria in pregnancy: a randomized, controlled pilot study. J. Urol. 180, 1367
1372. van Woudenbergh, G.J., van Ballegooijen, A.J., Kuijsten, A., Sijbrands, E.J.G., van Rooij, F.J. A., et al., 2009. Eating fish and risk of type 2 diabetes a population based, prospective follow-up study. Diabetes. Care 32, 2021
2026. Xu, T.W., 2005. Ion exchange membranes: state of their development and perspective. J. Membr. Sci. 263, 1
29. Xu, T.W., Huang, C.H., 2008. Electrodialysis-based separation technologies: a critical review. AIChE J. 54, 3147
3159. Yamanaka, A., Kimizuka, R., Kato, T., Okuda, K., 2004. Inhibitory effects of cranberry juice on attachment of oral streptococci and biofilm formation. Oral Microbiol. Immunol. 19 (3), 150
154. Yan, X., Murphy, B.T., Hammond, G.B., Vinson, J.A., Neto, C.C., 2002. Antioxidant activities and antitumor screening of extracts from cranberry fruit (Vaccinium macrocarpon). J. Agric. Food. Chem. 50 (21), 5844
5849. Yang, L., Jiang, B., Feng, B., Jin, Z., 2006. Electrodialysis in crude inulin purification. Food Sci. (Beijing) 27 (7), 119
123. Youssef, E.L., Rayess, Y.E., Mietton-Peuchot, M., 2016. Membrane technologies in wine industry: an overview. Crit. Rev. Food. Sci. Nutr. 56 (12), 2005
2020. Yuan, F., Wang, Q., Yang, P., Cong, W., 2016. Transport properties of amino acid ions at isoelectric point in electrodialysis. Sep. Purif. Technol. 168, 257
264. Zhang, Y., Pinoy, L., Meesschaert, B., et al., 2011. Separation of small organic ions from salts by ion-exchange membrane in electrodialysis. AIChE J. 57 (8), 2070
2076.

FURTHER READING Doyen, A., Saucier, L., Beaulieu, L., Pouliot, Y., Bazinet, L., 2012. Electroseparation of an antibacterial peptide fraction from snow crab by-products hydrolysate by electrodialysis with ultrafiltration membranes. Food. Chem. 132 (3), 1177
1184. Vera, E., Sandeaux, J., Persin, F., Pourcelly, G., Dornier, M., Ruales, J., 2009. Modeling of clarified tropical fruit juice deacidification by electrodialysis deacidification of the clarified passion fruit juice (P-edulis f. flavicarpa). J. Memb. Sci. 326 (2009), 472
483.

This page intentionally left blank

Chapter 11

Osmotic Driven Membrane Processes for Separation of Special Food Compounds Katalin Belafi-Bako1, Irena Petrinic´ 2, Claus He´lix-Nielsen3, Guofei Sun4, Ye Wee Siew4, Simon Alvisse4, Nguyen Xuan Tung4, Andras Boor1 and Nandor Nemestothy1 1

University of Pannonia, Veszprem, Hungary, 2University of Maribor, Maribor, Slovenia, Technical University of Denmark, Lyngby, Denmark, 4Aquaporin Asia Pte. Ltd, Singapore, Singapore 3

Chapter Outline 11.1 Introduction 11.2 Forward Osmosis for Up-Concentrating Food Compounds 11.3 Membrane Osmotic Distillation 11.3.1 Preservation of Valuable Compounds

383

386 389

11.3.2 Sensory Evaluation 11.4 Draw Solutions 11.5 Conclusions Acknowledgments References Further Reading

394 396 398 398 398 401

392

11.1 INTRODUCTION The phenomenon called osmosis is the selective transport of a solvent, mostly water, through a semipermeable membrane from a solution containing solute of lower concentration into a solution containing solute of higher concentration (draw solution) (Haynie, 2001). The layer separating the two solutions is the membrane. Fundamentally there are two distinct ways of water transport through a membrane driven by osmosis: either (1) in liquid state or (2) in vapor phase. In case of (1) selective membranes should be used that are able to retain the salts, while water can be passed. In case of water vapor transport (2) hydrophobic porous (micro- or ultrafiltration) membranes are applied.

Separation of Functional Molecules in Food by Membrane Technology. DOI: https://doi.org/10.1016/B978-0-12-815056-6.00011-5 © 2019 Elsevier Inc. All rights reserved.

383

384

Separation of Functional Molecules in Food by Membrane Technology

Regarding the first case, the driving force of the transport can be defined by the relation of the hydrostatic pressure difference and the osmotic pressure difference between the two sides of the membrane. Jw 5 AðσΔπ 2 Δphydr Þ 2

ð11:1Þ

1

where Jw is the water flux (kg/m /s ), A is the membrane permeability constant (s/m), σ is the reflective constant (
), Δπ is the osmotic pressure difference (Pa), and Δphydr is the hydrostatic pressure difference (Pa) between the two sides of the membrane. Δπ osmotic pressure is caused by the difference between the feed and draw (osmotic agent) solutions. While Δphydr is the difference of the pressures applied in the feed side (P1) and the draw regime (P2). Based on the equation four distinct processes can be defined (Cath et al., 2006). In Table 11.1 these techniques are summarized. The methods are characterized by the relations of the pressures (P1, P2, and Δπ), which determine the flow direction as well. In the first case, forward osmosis (FO), the hydrostatic pressure is the same in both sides; “conventional” osmosis takes place, and water molecules leave from the solution of lower concentration into the draw solution. In case of pressure assisted osmosis (PAO) a mild pressure is applied on the solution of lower concentration (P1), enhancing the rate of osmosis. Applying pressure on the other side (the solution of higher concentration, that is, the draw solution, P2), as well, pressure retarded osmosis (PRO) and reverse osmosis (RO) processes can be set up. If the pressure used is lower

TABLE 11.1 Osmotic Processes: Water Transport in Liquid Phase Osmotic Process

Abbreviation

Pressures

Flow Direction

Source

Forward osmosis (direct osmosis)

FO (DO)

Δπ

Direct

Hough (1993)

P1 5 0

Kravath (1975)

P2 5 0 P1 . P2

Direct

Blandin (2015)

PRO

P2
P1 , Δπ

Direct

Loeb (1975)

RO

P2
P1 . Δπ

Reverse

Sourirajan (1970), Fritzmann et al. (2007)

Pressure assisted osmosis, pressure enhanced osmosis

PAO

Pressure retarded osmosis Reverse osmosis

PEO

Osmotic Driven Membrane Processes Chapter | 11

385

than the osmotic pressure (P2
P1 , Δπ) water will flow into the draw solution (diluting it), and the volume of this flow can be so high that it may be exploited for power generation. This is the principle of osmotic power (https://www.power-technology.com/projects/statkraft-osmotic/). Pattle first described this interesting phenomenon: when a river mixes with the sea, a source of power is formed in terms of the lost osmotic pressure (Pattle, 1954). However, a practical method of exploiting it by using selectively permeable membranes was only developed much later, in the mid-1970s. The generation of electric energy through PRO was developed by Sidney Loeb (1975). The next step was done by a Norwegian company, Statkcraft, that filed the first patent for osmotic power membranes in 2003. Then the world’s first osmotic power or salinity gradient power generation plant (a prototype) was built in Tofte (Norway) by the same company, which was opened in 2009. The prototype plant has a designed capacity to generate 10 kW of electricity. The plant generates renewable and emissions-free energy and thus contributes to eco-friendly power, sometimes called blue energy, as well, since no CO2 emission occurs. Finally, when high pressure is applied in the secondary side (P2), this is RO, one of the most widely used membrane techniques (Fritzmann et al., 2007). RO belongs to the “pressure driven membrane processes,” where the mechanism of the separation is a type of filtration. In RO the membrane retains (almost) all the dissolved solutes, even the monovalent ions, small molecules, and the pure solvent can pass through the membrane. Thus it is a suitable method for water purification. Desalination of seawater can be accomplished by RO and it is used commonly to produce potable water. To better picture the volume of desalination by RO nowadays, a few figures are presented (http://idadesal.org/) here: G G

G G

The total number of desalination plants worldwide (2015) is 18,426. The global capacity of commissioned desalination plants (2015) is more than 86.8 million cubic meters per day. The number of countries where desalination is practiced is 150. More than 300 million people around the world rely on desalinated water for some or all of their daily needs.

Beyond the osmotic membrane processes mentioned earlier, there are some more similar techniques that exploit osmotic power. Osmotic evaporation (OE) or osmotic distillation (OD) (Hogan et al., 1998; Alves and Coelhoso, 2002) is a membrane process where water transport occurs, as well, but in gaseous phase (2). Water vapor passes through the pores of the membrane. The driving force is again the osmotic power, since a draw solution (osmotic agent) is applied on the secondary side of the membrane. The effectiveness of OD can be enhanced when it is combined with membrane distillation (MD) (Be´lafi-Bako´ and Koroknai, 2006; Nagaraj et al., 2006; Koroknai et al., 2006; Wang et al., 2001). The coupled process

386

Separation of Functional Molecules in Food by Membrane Technology

called membrane osmotic distillation (MOD) exploits the thermal effects, that is, the two solutions are thermostated separately at different temperatures: the draw solution is the cold side (chilled), while the feed solution is the warm side (heated). Thus the driving forces are added and the water flux can increase. Both FO, (PAO) and MOD (OD) are considered as effective membrane processes for water removal, which can be applied for the concentration of various aqueous solutions (Cuperus, 1998; El-Abbassi et al., 2013, Jiao et al., 2004). Moreover the methods are operated under mild conditions, additives are not needed, and hazardous materials are not produced, thus the techniques are regarded as environmentally safe, attractive solutions for numerous kinds of separations.

11.2 FORWARD OSMOSIS FOR UP-CONCENTRATING FOOD COMPOUNDS The use of FO in the form of concentrating freshly squeezed fruit juices using a semipermeable cloth bag containing the juice immersed in a concentrated brine solution probably has a long agricultural history (Cussler, 1984). One of the first scientific investigations on the application of FO for food concentration was conducted by Leonard Wickenden, an American chemist and organic gardening advocate. On January 11, 1934, Wickenden filed a patent application (which was granted in 1938) in which he discloses the use of a semipermeable membrane made of parchment or cellophane for upconcentration of fruit juices described as using both batch and continuous mode embodiments of the process (Wickenden, 1938). He specifically described the osmotic concentration of orange juice with 13 Brix as feed solution using a cellophane membrane and 80 Brix syrup as draw solution. He also describes an up-concentration of coffee from 15% solids to 75% solids producing a concentrate with an excellent flavor, and a 7:1 volume ratio up-concentration of milk. The formed milk concentrate had no cooked flavor and its color was not altered, except from the natural color that became darker by the concentration. The citations not only point to the potential of FO in liquid food up-concentration, but also to two major technological challenges: (1) how to minimize undesired flux of solutes from the feed to the draw solution; and (2) how to minimize undesired reverse solute flux from the draw to the feed solution. Both effects can have detrimental impact on the up-concentrated product such as the loss of essential flavors/ fragrances and/or contamination by draw solution agents (e.g., salts). With the advent of cellulose acetate membranes, the concept gained momentum (Popper et al. 1966). However, the problem of reverse solute flux remained a serious issue thus impairing the sensorial quality of the final up-concentrated product. In the 1990s, membranes with improved selectivity combined with lower internal concentration polarization (which will diminish

Osmotic Driven Membrane Processes Chapter | 11

387

the effective osmotic driving force) in the membrane material became available. This has spurred a new interest in FO for achieving high levels of concentration for a variety of liquid foods and food ingredients, and the research area has been described in several reviews (Petrotos & Lazarides, 2001; Rastogi, 2016; Terefe et al. 2016). An attractive liquid food to up-concentrate is coconut milk. Coconut milk and cream are found in many traditional Indian and Southeast Asian cuisines, in sweet and savory dishes. Today Indonesia and Thailand are some of the world’s largest exporters and consumers of coconut milk. Coconut milk beverages, containing less than 1%
2% fat, are increasingly being recognized as competitors to soya and almond milk products in the United States and Europe (Tetra Pak, 2016). Also, coconut milk is gaining interest as a dietary substitute for lactose intolerant consumers, and many brands are already diversifying their beverage offerings to include coconut milk. Recent studies have also pointed the beneficial effects of coconut milk oil medium chain triglycerides for treating diseases such as dementia (Alzheimer’s disease) (Henderson, 2004). However, coconut milk is a challenging feed solution. It is a thixotropic fluid (similar to yogurt), a type of non-Newtonian fluid with time-dependent viscosity where viscosity diminishes with time when the fluid is subjected to a constant speed gradient. As dewatering will lead to an increase in viscosity the FO membrane module must be designed to accommodate this increase. In a recent feasibility study, up-concentration of a coconut milk solution using Aquaporin Inside Tubular FO membrane module was attempted (Fig. 11.1). The (feed) was operated in a batch mode to attain a higher recovery. The draw solution was operated in a continuous mode to maintain a sufficient driving force. The experiment was performed in FO mode where active layer of the membrane is facing the feed side. Experimental details are summarized in Fig. 11.1.

FIGURE 11.1 FO up-concentration of coconut milk with batch feed and continuous draw in a cocurrent flow arrangement. Feed solution: 10 kg coconut milk (24% fat content). Draw solution: 1 M NaCl solution. Membrane: two Aquaporin Inside D50 Tubular forward osmosis (FO) modules in parallel (0.7 m2 total membrane area). Feed flow: 0.7 LPM (peristaltic pump) per D50 (lumen flow). Draw flow: 0.5 LPM (gearing pump) per D50 (shell flow). Feed and draw inlet pressures maintained at ,1 and ,0.1 bar respectively. Experiments were conducted at an ambient temperature of 17 C.

388

Separation of Functional Molecules in Food by Membrane Technology

With batch concentration operations and 1 M NaCl draw solution, it was possible to up-concentrate the 10 kg coconut milk 2 3 corresponding to approximately 50% recovery. Water is recovered from the coconut milk to the draw solution with time (Fig. 11.2A). The initially measured viscosity increased from 700 to 5000 cP whereas the equilibrium viscosity increased from 500 to 800 cP reflecting the thixotropic nature of the coconut milk (Fig. 11.2B). For sanitary reasons it is crucial that the membrane modules (and any food contact material) can be cleaned. This was tested using a standard cleaning protocol, see Fig. 11.2C, and the results are shown in Fig. 11.2D. As can be seen, both the water flux Js and reverse salt flux Js were similar before and after cleaning demonstrating the feasibility to use tubular FO modules for coconut milk up-concentration. Besides coconut milk also coconut water is of interest as feed in FO upconcentration. Coconut water can be used directly as a nutritional drink and its sodium and potassium content makes it an ideal drink for rehydration. However, it is not commercially viable to transport large amounts of single strength coconut water (92%
95% water content) in bulk to markets where coconuts are not readily available (Tetra Pak, 2016). Thus, it is desirable to be able to concentrate coconut water to higher soluble solid levels of 60
65 Brix (or 35%
40% water content), saving transportation costs. (A)

(B) 6000

60%

5000 5000 Viscosity (cP)

Recovery (%)

50% 40% 30% 20%

Before Coconut Milk Application Test

3000

After Coconut Milk Application Test

2000 1000

10% 0%

4000

0

60

120 180 Time (mins)

240

300

First Viscosity Reading (cP) Stabilized Viscosity Reading (cP)

(D)

800 700

Duration

10% citric acid recirculation rinse

30 mins

10% ethanol recirculation rinse

30 mins

RO water recirculation rinse

30 mins

25L RO water flush

As required

688

698

600 Jw (LMH) and js (gMH)

Step

500

0

360

(C)

800

700

500 400 300 200 100

056

055

000 LS-D50-20 LS-D50-20 Before After

FIGURE 11.2 (A) Forward osmosis (FO) up-concentration of coconut milk. Recovery as a function of time. (B) Viscosity measurements before after and FO up-concentration. (C) Cleaning protocol used. (D) Membrane integrity tests performed with De-ionised water (DI) water as feed and 1 M NaCl as draw.

389

Brix (%),viscpsity (cP) and recovery (%)

Osmotic Driven Membrane Processes Chapter | 11

100 Brix (%)

90

Viscosity (cP) Recovery (%)

80 70 60 50 40 30 20 10 0 0

50

100

150 Time (mins)

200

0

FIGURE 11.3 FO up-concentration of coconut water. Batch test for feed and single pass for draw. (Draw solution: 2.5 M MgCl2).

In the current state of the art for producing coconut water concentrate, fresh coconut water is first passed through a preconcentration stage of RO to increase the total solids. Then, it goes into a multiple effect evaporation stage to achieve 60
65 Brix, at which concentration the coconut water is to some degree self-preserving (Tetra Pak, 2016). A proof-of-concept experiment was conducted where a concentration .60 Brix was achieved using Aquaporin Inside FO membranes (Fig. 11.3). Both results obtained with coconut milk and water demonstrated the promising application of FO in liquid food up-concentrations.

11.3 MEMBRANE OSMOTIC DISTILLATION MOD, as mentioned earlier, could be applied for concentration of various aqueous solutions containing heat-sensitive compounds, which can be easily found in agro-food processes. One of the typical applications is the concentration of fresh juices from various fruits and vegetables (Kujawa et al., 2015), e.g., juices from camu-camu (Rodrigues et al., 2004; Souza et al., 2013), acerola (Pagani et al., 2011), sour cherry (Ra´cz et al., 2014), passion fruit (Vaillant et al., 2001), cranberry (Zambra et al., 2015), tomato (Petrotos et al., 1999), etc. (Table 11.2). In our laboratory, concentration of juices from colorful fruits by MOD was studied previously (Koroknai et al., 2008). These fruits, belonging to berry types: black and red currant, raspberry, and blackberry, have considerable antioxidant capacity due to their polyphenolic, anthocyanin, and flavonoid content and other special, valuable oxidant compounds (Koroknai et al., 2008). The antioxidant capacity (scavenging free radicals) can be measured by ferric reducing ability of plasma (Benzie and Strain, 1996), while total

390

Separation of Functional Molecules in Food by Membrane Technology

TABLE 11.2 Membrane Osmotic Distillation (MOD) for Concentration of Various Juices of Fruits and Vegetables Fruit/Vegetable

Source

Camu-camu

Rodrigues et al. (2004) Souza et al. (2013)

Acerola

Pagani et al. (2011)

Cranberry

Zambra et al. (2015)

Tomato

Petrotos et al. (1999)

Sour cherry

Ra´cz et al. (2014)

Passion fruit

Vaillant et al. (2001)

Grape

Kujawa et al. (2015)

Red currant

Koroknai et al. (2008)

Raspberry

Koroknai et al. (2008)

Black currant

Ba´nvo¨lgyi et al. (2009)

phenol and anthocyanin content (having antioxidant capacity) can be determined by Folin-Ciocalteu reagent (Singleton et al., 1999) and an ultraviolet (UV) spectrometric method (Giusti and Wrolstad, 2001), respectively. Our experiments have proven that the mild membrane technique, MOD, is able to preserve almost all of these valuable compounds during the concentration process. The range of these colorful berry fruits has been recently extended with similar, but wild-grown species (Belafi-Bako and Boor, 2011; Boo´r et al., 2016), like elderberry (Sambucus nigra), Cornelian cherry (Cornus mas), common whitebeam (Sorbus aria), and blackthorn (Prunus spinosa). These fruits are rich in coloring compounds and have antioxidant capacities as well (Dawidowicz et al., 2006; Serpil and Ilkay, 2008; Sidor and GramzaMichałowska, 2014). These fruit plants are native to Hungary and have grown there for a long time. They have been used traditionally for homemade manufacturing of jam, stewed fruit, marmalade, syrup, cookies, pastries, and soft drinks. The fresh juices were collected in the ripening season in the Transdaubian region (nearby) and, after removing the seeds, were pressed in a laboratory scale mechanical squeeze to get the fresh juices. Then the juices were stored in the deep-freezer until the measurements. The juices were concentrated by MOD in laboratory scale experiments and the preservation of antioxidant capacity was measured.

Osmotic Driven Membrane Processes Chapter | 11

391

The measurements were carried out (Belafi-Bako and Boor, 2011) in a hollow fiber membrane module (Microdyn), using hydrophobic polypropylene membrane, with a membrane surface area of 70 cm2. The 6 M CaCl2 draw solution was circulated in the secondary side of the module (shell side) at 18 C, while the fruit juices to be concentrated were circulated in the primary side (in the tubes) at 35 C temperature. Both solutions were circulated by peristaltic pumps on a countercurrent mode of operation with a flow rate of 10 L/h. To follow the water transport the weight of the fruit juice was inline measured and the total solid substance (TSS) of the samples was determined. In Fig. 11.4 the changes of volume and TSS of the elderberry juice are presented as a function of time as an example. As can be seen elderberry juice was managed to concentrate considerably up to 54% TTS. The time courses of the concentration processes of the other juices were similar. The results of the concentration measurements of all four fruit juices are summarized in Table 11.3 including the initial and final TSS of the juices, the process time, and the average flux values calculated. As can be seen from the table the flux values obtained varied between 400 and 570 mL/m2h.

FIGURE 11.4 Changes of volume and total solid substance (TSS) of elderberry juice during the concentration process.

392

Separation of Functional Molecules in Food by Membrane Technology

TABLE 11.3 Experimental Results of the Concentration Processes by MOD Fruit Juices

Cornelian Cherry

Initial conc. TSS (%)

Blackthorn

Elderberry

22.5

22.0

57

64

53

54

Process time (h)

140

280

200

120

Average flux values (mL/m2h)

570

400

500

550

Final conc. TSS (%)

7.0

Whitebeam

8.0

Antioxidant capacity (mg AS/L)

400 350 300 250 200 150 100 50 0

Elderberry juice (rediluted)

Whitebeam juice Cornelian cherry juice (rediluted) (rediluted)

Blackthorn juice (rediluted)

Elderberry juice (initial)

Whitebeam juice (initial)

Blackthorn juice (initial)

Cornelian cherry juice (initial)

FIGURE 11.5 Antioxidant capacity of the fresh and rediluted juices.

11.3.1 Preservation of Valuable Compounds The antioxidant capacity, total phenol, and anthocyanin contents of the fresh and rediluted samples were determined and compared. (The concentrated syrups were rediluted by water exactly to the initial TSS level.) The results of the measurements are presented in Figs. 11.5
11.7. The antioxidant capacity of elderberry, whitebeam and blackthorn juices were preserved almost entirely. In the case of Cornelian cherry a bit more loss was observed during the process. The anthocyanin contents of whitebeam, Cornelian cherry, and blackthorn juices were well preserved, and the differences between the initial and the rediluted juices were less than 10%. While in elderberry juice only 80% of the initial anthocyanin content was preserved.

Osmotic Driven Membrane Processes Chapter | 11

393

1000

Anthocyanin content (mg/L)

900 800 700 600 500 400 300 200 100 0

Elderberry juice (rediluted)

Whitebeam juice Cornelian cherry juice (rediluted) (rediluted)

Blackthorn juice (rediluted)

Elderberry juice (initial)

Whitebeam juice (initial)

Blackthorn juice (initial)

Cornelian cherry juice (initial)

FIGURE 11.6 Anthocyanin content of the fresh and rediluted juices. 70

Polyphenol content (mg/cm3)

60 50 40 30 20 10 0

Elderberry juice (rediluted)

Whitebeam juice Cornelian cherry juice (rediluted) (rediluted)

Blackthorn juice (rediluted)

Elderberry juice (initial)

Whitebeam juice (initial)

Blackthorn juice (initial)

Cornelian cherry juice (initial)

FIGURE 11.7 Total polyphenol content of the fresh and rediluted juices.

The loss of polyphenol contents of elderberry, Cornelian cherry, and blackthorn juices was within 5%, while in whitebeam juice it was somewhat higher (approximately 10%). Thus preservation of the polyphenol content can be regarded quite well during the up-concentration process. As a summary the results in Figs. 11.5
11.7 confirmed that both the antioxidant capacity, the anthocyanin, and total polyphenol content of the juices were managed to be preserved almost entirely. Hence it was shown that the MOD process is a mild membrane technique, suitable for concentration of fruit juices.

394

Separation of Functional Molecules in Food by Membrane Technology

11.3.2 Sensory Evaluation To assess a successful technology in the food industry entirely, it is not enough to check the conditions of the up-concentration process and preservation of valuable contents. Evaluation of other features of the juices (e.g., taste, flavor, odor, color, consistency) is needed as those characteristics are important for the consumers. These characters can be judged by organoleptic (sensory) evaluations, where the products, after careful preparations, are investigated and compared by numerous assessors regarding given aspects. To compare the sensory features of the initial, fresh, and rediluted juices, a special organoleptic method was used, called the paired-comparison test (UNI EN ISO 5495) (http://www.gustosalutequalita.it), where the differences of two samples were studied. The assessors evaluated comparatively the two juices of each fruit, studying the consistency, odor, color, and flavor and scored the difference from 1 to 5. The smaller the difference between the samples, the higher score (5) can be given. Finally all the grades were averaged and summarized for the four properties (Table 11.4.). To perform the sensory evaluation test in the organoleptic laboratory two differently coded samples were presented to each panelist simultaneously and the panelist’s task was to score the difference. The two samples, A and B, were presented randomized serving sequences (AB and BA). In the organoleptic evaluation 10 panelists participated. The consistency, odor, color, and flavor of the fresh and rediluted juices from Cornelian cherry, blackthorn, white beam, and elderberry were evaluated according to the paired-comparison test. The assessors scored the differences between the sample pairs from 1 to 5. The results of the sensory evaluation for Cornelian cherry, as an example, are summarized in Table 11.5. The average results of the organoleptic evaluations for all the fruit juices are presented in Table 11.6. As can be seen the highest average score was achieved for the consistency, but the differences for other properties were

TABLE 11.4 Grades of the Sensory Evaluation Grades

Qualification

Quality Class

Total scores

5

Excellent

Excellent

17.6
20.0

4

Good

Good

15.2
17.5

3

Satisfactory

Satisfactory

13.2
15.1

2

Sufficient

Sufficient

11.2
13.1

1

Insufficient

Insufficient

, 11.2

Osmotic Driven Membrane Processes Chapter | 11

395

TABLE 11.5 Grades of the Sensory Evaluation for the Juices of Cornelian Cherry Assessor

Difference Between the Sample Pair Consistency

Odor

Color

Flavor

Total

1

5

4

4

5

18

2

5

5

5

5

20

3

5

4

4

5

18

4

5

3

4

5

17

5

5

4

4

5

18

6

4

5

5

4

18

7

4

4

4

4

16

8

5

5

4

5

19

9

5

4

4

4

17

10

4

5

4

4

17

Average

4.7

4.3

4.2

4.6

17.8

TABLE 11.6 Average Results of the Organoleptic Tests Fruit

Consistency

Odor

Color

Flavor

Total

Cornelian cherry

4.7

4.3

4.2

4.6

17.8

Blackthorn

4.5

4.3

4.6

4.2

17.6

Whitebeam

4.3

4.5

4.1

4.5

17.4

Elderberry

4.8

4.2

4.5

4.2

17.7

really small, as well, which means that the quality of the concentrated and rediluted juices were similar to the initial, fresh juices. It was found during the sensory evaluations that no heat damage occurred during the concentration process; no burnt sugar taste (caramel) could be recognized in the juices. The average total values for the evaluation of the juice pairs fall into the “excellent” quality class (range 17.6
20), that is, there was only a minimal difference found between the fresh and rediluted juices. Hence these tests confirmed that the MOD technique for the concentration of fruit juices is really a mild method and able to preserve the valuable compounds.

396

Separation of Functional Molecules in Food by Membrane Technology

TABLE 11.7 Classification of Draw Solutes Types of Draw Solute

Examples

Features

Ordinary

Inorganic salt

NaCl, Na2SO4, MgCl2, K acetate, Al2(SO4)3, CaCl2

High osmotic pressure, low cost

Organic salt

Zwitterions, glycine, urea

Low reverse diffusion

Nutrient

Sucrose, fructose, glucose

Suitable for food applications

Thermolytic salt (volatile)

NH4HCO3

High osmotic pressure

Ionic liquid

Betaine bis (trifluoromethylsulfonyl) imide ((Hbet)(Tf2N));

High cost, synthetic

Unconventional

Sodium tetraethylenepentamine heptaacetate (STPH) Surfactant

Triton-X 114

Nanoparticle

Hydrophilic magnetic nanoparticles

Easy regeneration

Gas

SO2 (with KNO3), NH3, CO2

Easy regeneration

Polymer

Polymer hydrogels

Poor water flux

Polyelectrolyte

Carboxylate polyelectrolyte

High viscosity

11.4 DRAW SOLUTIONS For both FO and MOD usage of a draw solution is essential and crucial. Ideally a draw solution has the following features (Ge et al., 2013; Linares et al., 2017; Shon et al., 2015): G G

G G G G G G

ability to generate high osmotic pressure (driving force) the reverse flux of the draw solute is minimal (concentration polarization) easy to regenerate the draw solute has a small molecular weight and low viscosity in water solid state at ambient temperature (easier to handle) compatibility with the membrane low toxicity low cost

Osmotic Driven Membrane Processes Chapter | 11

397

The types of draw solutes can be classified by various aspects. Table 11.7 presents a classification system, where “ordinary” (inorganic and organic salts, nutrients, volatile compounds) and “unconventional” solutes (nanoparticles, ionic liquids, gases, polymers, etc.) are listed (www.forwardosmosistech.com; Liu et al., 2015; Hoover et al., 2011; Long and Wang, 2015). Among these “unconventional” solutes much literature is available on ionic liquids (Gale and Lovering, 1982; Inman and Lovering, 1981; Kirchner, 2009; Marcus, 2016). Recently the range of unconventional draw solutes were extended (Li and Wang, 2013; Knoerzer et al., 2016) with the so-called smart draw agents possessing responsive properties. These interesting smart draw agents (“advanced materials”) include, e.g., functionalized magnetic nanoparticles, thermoresponsive polyelectrolytes, and stimuli-responsive polymer hydrogels. During the operation of both methods (FO and MDO) the draw solution is being continuously diluted, thus the driving force is gradually decreasing, the process is slower and slower, and finally the process should be stopped. To apply the draw solution again, regeneration is extremely important. Various methods are known for regeneration of the draw solutions (www. forwardosmosistech.com; Singh and Hankins, 2016), they are categorized and presented in Table 11.8. One has to note, however, that in certain cases (direct use) it is possible to avoid regeneration: when the diluted draw solution can be used itself, e.g., in the case of the hydration pack, where a sugar-and-nutrient draw solution is applied to provide an energy drink from any kind of (polluted) water. As can be seen from the table, the most suitable regeneration process should be chosen by considering the features of the draw solution. For

TABLE 11.8 Classification of Regeneration Methods for Draw Solution Regeneration Methods Thermal processes

Membrane based processes

Other processes

For Draw Solution Multivapor compression

Salts

Multistage flash distillation

Salts

(Simple) heating

Gas (SO2)

Nanofiltration

Multivalent salts

Pervaporation

Volatile compounds

Reverse osmosis

Monovalent salts

Membrane distillation

Salts

Precipitation

e.g., Al2(SO4)3 with Ca (OH)2

398

Separation of Functional Molecules in Food by Membrane Technology

regeneration of the most widely used salts, however, the majority of the processes available are quite energy-intensive. If we want to elaborate a sustainable and economical FO process for industrial use, a balance should be found between water recovery by FO and the cost of the draw recovery process.

11.5 CONCLUSIONS Both FO and MOD can be considered as effective, environmentally safe, modern concentration processes, which are operated under mild conditions, thus the valuable heat-sensitive compounds can be preserved during the procedures. A case study was reported in this chapter. FO was applied for up-concentration of coconut milk, while MOD was used for production of syrups from juices made of various colorful wild-grown fruits (Cornelian cherry, blackthorn, white beam, and elderberry). The advantageous features of both osmotic driven membrane processes were proven experimentally.

ACKNOWLEDGMENTS The research work was partly supported by the projects EFOP 3.6.1-16-2016-00015 entitled “University of Pannonia’s comprehensive institutional development program to promote Smart Specialization Strategy” and GINOP-2.3.2-15 entitled “Excellence of strategic R 1 D workshops; Development of modular, mobile water treatment systems and waste water treatment technologies based on University of Pannonia to enhance growing dynamic export of Hungary (2016-2020)”.

REFERENCES Alves, V.D., Coelhoso, I.M., 2002. Mass transfer in osmotic evaporation: effect of process parameters. J. Membr. Sci. 208, 171
179. Ba´nvo¨lgyi, S., Horva´th, S., Stefanovits-Ba´nyai, E´., Be´ka´ssy-Molna´r, E., Vatai, G., 2009. Integrated membrane process for blackcurrant (Ribes nigrum L.) juice concentration. Desalination 241 (1-3), 281
287. Be´lafi-Bako´, K., Koroknai, B., 2006. Enhanced flux in fruit juice concentration: coupled operation of osmotic evaporation and membrane distillation. J. Membr. Sci. 269, 187
193. Belafi-Bako, K., Boor, A., 2011. Concentration of Cornelian cherry fruit juice by membrane osmotic distillation, Desal. Water Treat. 35, 271274. Benzie, I.F.F., Strain, J.J., 1996. The ferric reducing ability of plasma (FRAP) as a measure of “antioxidant power”: the FRAP assay. Anal. Biochem. 239, 70
76. Blandin, G., (2015): Fundamentals and Challenges of Pressure Assisted Osmosis to Combine Desalination and Water Reuse, (Dissertation). Ghent University. Faculty of Bioscience Engineering; University of New South Wales. Faculty of Chemical Engineering. Boo´r, A., Be´lafi-Bako´, K., Nemesto´thy, N., 2016. Concentration of colourful wild berry fruit juices by membrane osmotic distillation via cascade model systems. J. Membr. Sci Res. 2, 201
206.

Osmotic Driven Membrane Processes Chapter | 11

399

Cath, T.Y., Childress, A.E., Elimelech, M., 2006. Forward osmosis: principles, applications and recent developments. J. Membr. Sci. 281, 70
87. Cuperus, F.P., 1998. Membrane processes in agro-food: state of the art and new opportunities. Sep. Purif. Technol. 14 (1
3), 233
239. Cussler, E.L., 1984. Diffusion: Mass Transfer in Fluid Systems. Cambridge University Press, Cambridge, England. Dawidowicz, A.L., Wianowska, D., Baraniak, B., 2006. The antioxidant properties of alcoholic extracts from Sambucus nigra L. (antioxidant properties of extracts). LWT - Food Sci. Technol. 39, 308
315. El-Abbassi, A., Khayet, M., Kiai, H., Hafidi, A., Garcı´a-Payo, M.C., 2013. Treatment of crude olive mill wastewaters by osmotic distillation and osmotic membrane distillation. Sep. Purif. Technol. 4, 327
332. Fritzmann, C., Lo¨wenberg, J., Wintgens, T., Melin, T., 2007. State-of-the-art of reverse osmosis desalination. Desalination 216, 1
76. Gale, R.J., Lovering, D.L. (Eds.), 1982. Molten Salt Techniques. Plenum Press, New York. Ge, Q., Ling, M., Chung, T.S., 2013. Draw solutions for forward osmosis processes: developments, challenges and prospects for the future. J. Membr. Sci. 442, 225
237. Giusti, M.M., Wrolstad, R.E., 2001. Anthocyanins: characterization and measurement with UVvisible spectroscopy. In: Wrolstad, R.E. (Ed.), Current Protocols in Food Analytical Chemistry. Wiley, New York, pp. F1.2.1
1.2.13. Haynie, D.T., 2001. Biological Thermodynamics. Cambridge University Press, Cambridge. Henderson, S., 2004. Use of Medium Chain Triglycerides for the Treatment and Prevention of Alzheimer’s Disease and Other Diseases Resulting From Reduced Neuronal Metabolism II. US Patent US 6,835,750 B1. Hogan, P.A., Canning, R.P., Peterson, P., Johnson, R.A., Michaels, A.S., 1998. A new option: osmotic distillation. Chem. Eng. Progress 7, 49
61. Hoover, L.A., Phillip, W.A., Tiraferri, A., Yip, N.Y., Elimelech, M., 2011. Forward with osmosis: emerging applications for greater sustainability. Environ. Sci. Technol. 45 (23), 9824
9830. Hough, G., 1993. A simple model for osmotic dehydration of Apples. Lebensm.- Wiss. U.- Technol. 26, 151
156. Inman, D., Lovering, D.L. (Eds.), 1981. Ionic Liquids. Springer International Publishing, New York. Jiao, B., Cassano, A., Drioli, E., 2004. Recent advances on membrane processes for the concentration of fruit juices: a review. J. Food Eng. 63, 303
324. Kirchner, B., 2009. Ionic Liquids. Springer International Publishing. Knoerzer, K., Juliano, P., Smithers, G.W. (Eds.), 2016. Innovative Food Processing Technologies: Extraction, Separation, Component Modification and Process Intensification. Woodhead Publishing, Duxford, Cambridge, Kidlington, United Kingdom. Koroknai, B., Kiss, K., Gubicza, L., Be´lafi-Bako´, K., 2006. Coupled operation of membrane distillation and osmotic evaporation in fruit juice concentration. Desalination 200, 526
527. Koroknai, B., Csana´di, Z., Gubicza, L., Be´lafi-Bako´, K., 2008. Preservation of antioxidant capacity and flux enhancement in concentration of red fruit juices by membrane processes. Desalination 228, 295
301. Kravath, R.E., Davis, J.A., 1975. Desalination of sea water by direct osmosis. Desalination 16 (2), 151
155. Kujawa, J., Guillen-Burrieza, E., Arafat, H.A., Kurzawa, M., Wolan, A., Kujawski, W., 2015. Raw juice concentration by osmotic membrane distillation process with hydrophobic polymeric membranes. Food Bioprocess Technol. 8 (10), 2146
2158.

400

Separation of Functional Molecules in Food by Membrane Technology

Li, D., Wang, H., 2013. Smart draw agents for emerging forward osmosis application. J. Mater. Chem. A 45 (1), 14049
14060. Linares, R.V., Li, Z.U., Elimelech, M., Amy, G., Vrouwenvelder, H. (Eds.), 2017. Recent Developments in Forward Osmosis Processes. IWA Publishing, London. Liu, B., Yu, J., Li, D., Lu, D., Zhang, J., 2015. Research advance on draw solution in forward osmosis process. Proc. Int. Conf. Adv. Energy Env. Sci (ICAEES). Atlantis Press, pp. 898
902. Loeb, S., 1975. Osmotic power plants. Science 189 (4203), 654
655. Long, Q.W., Wang, Y., 2015. Sodium tetraethylenepentamine heptaacetate as novel draw solute for forward osmosis—synthesis, application and recovery. Energies 8, 12917
12928. Marcus, Y., 2016. Ionic Liquid Properties: From Molten Salts to RTILs. Springer International Publishing, New York. Nagaraj, N., Patil, B.S., Biradar, P.M., 2006. Osmotic membrane distillation - A brief review. Int. J. Food Eng. 2 (2), 2
22. Pagani, M.M., Rocha-Lea˜o, M.H., Couto, A.B.B., Pinto, J.P., Ribeiro, A.O., dos Santos Gomes, F., et al., 2011. Concentration of acerola (Malpighia emarginata DC.) juice by integrated membrane separation process. Desalin. Water Treat. 27 (1
3), 130
134. Pattle, R.E., 1954. Production of electric power by mixing fresh and salt water in the hydroelectric pile. Nature 174 (4431), 660. Petrotos, K.B., Lazarides, H.N., 2001. Osmotic concentration of liquid foods. J. Food Eng. 49 (2
3), 201
206. Petrotos, K.B., Quantick, P.C., Petropakis, H., 1999. Direct osmotic concentration of tomato juice in tubular membrane module configuration. II. The effect of using clarified tomato juice on the process performance. J. Membr. Sci. 160 (2), 171
177. Popper, K., Carmirand, W.M., Stanley, W.L., Nury, F.S., 1966. Osmotic emergency purifier of seawater. Nature 211, 297
298. Ra´cz, G., Alam, M.R., Arekatte, Ch. K., Alberta, K., Papp, N., Stefanovits-Ba´nyai, E´., et al., 2014. Potassium acetate solution as a promising option to osmotic distillation for sour cherry (Prunus cerasus L.) juice concentration. Acta Alimentaria 43, 114
206. Rastogi, N.K., 2016. Opportunities and challenges in application of forward osmosis in food processing. Crit. Rev. Food Sci. Nutr. 56, 266
291. Rodrigues, R.B., Menezes, H.C., Cabral, L.M.C., Dornier, M., Rios, G.M., Reynes, M., 2004. Evaluation of reverse osmosis and osmotic evaporation to concentrate camu-camu juice (Myrciaria dubia). J. Food Eng. 63, 97
102. Serpil, T., Ilkay, K., 2008. Physico-chemical and antioxidant properties of Cornelian cherry fruits (Cornus mas L.) grown in Turkey. Sci. Hortic. (Amsterdam) 116, 362
366. Shon, H.K., Phuntsho, S., Zhang, T.C., Surampalli, R.Y., 2015. Forward Osmosis Fundamentals and Applications. American Society of Civil Engineers. Sidor, A., Gramza-Michałowska, A., 2014. Advanced research on the antioxidant and health benefit of elderberry (Sambucus nigra) in food - a review. J. Funct. Foods . Available from: https://doi.org/10.1016/j.jff.2014.07.012. Singh, R., Hankins, N., 2016. Emerging Membrane Technology for Sustainable Water Treatment. Elsevier, Amsterdam. Singleton, V., Orthofer, L.R., Lamuela-Raventos, R.M., 1999. Analysis of total phenols and other oxidation substrates and antioxidants by means of Folin-Ciocalteu reagent. Method Enzymol. 299, 152
178. Sourirajan, S., 1970. Reverse Osmosis. Academic Press, Toronto, Canada.

Osmotic Driven Membrane Processes Chapter | 11

401

Souza, A.L.R., Pagani, M.M., Dornier, M., Gomes, F.S., Tonon, R.V., Cabral, L.M.C., 2013. Concentration of camu-camu juice by the coupling of reverse osmosis and osmotic evaporation processes. J. Food Eng. 119 (1), 7
12. Terefe, N.S., Janakievski, F., Glagovskaia, O., De Silva, K., Horne, M., Stockmann, R., 2016. Forward osmosis: a novel membrane separation technology of relevance to food and related industries. In: Knoerzer, K., Juliano, P., Smithers, G. (Eds.), Innovative Food Processing Technologies. Woodhead Publishing, Duxford, Cambridge, Kidlington, United Kingdom, pp. 177
205. Tetra Pak, 2016. The Coconut Handbook: Technology, Engineering, Agriculture. Tetra Pak International S.A, 9789176115978 pp. 1
183. Vaillant, F., Jeanton, E., Dornier, M., O’brien, G.M., Reynes, Decloux, M., 2001. Concentration of passion fruit juice on an industrial pilot scale using osmotic evaporation. J. Food Eng. 47, 195
202. Wang, Z., Zheng, F., Wu, Y., Wang, S., 2001. Membrane osmotic distillation and its mathematical simulation. Desalination 139, 423
428. Wickenden, L., 1938. Concentration of Liquid Food Material. US Patent 2,116,920. Zambra, C., Romero, J., Pino, L., Saavedra, A., Sanchez, J., 2015. Concentration of cranberry juice by osmotic distillation process. J. Food Eng. 144, 58
65.

FURTHER READING Alves, V.D., Koroknai, B., Be´lafi-Bako´, K., Coelhoso, I.M., 2004. Using membrane contactors for fruit juice concentration. Desalination 162, 263
270. Crespo, J.G., Bo¨ddeker, K.W., 1994. Membrane Processes in Separation and Purification. Springer Verlag, New York. Field, R.W., Bekassy-Molnar, E., Lipnizki, F., Vatai, G., 2017. Engineering Aspects of Membrane Separation and Application in Food Processing. CRC Press, Boca Raton. Herron, J.R., Beaudry, E.G., Jochums, C.E., Medina, L.E., 1994. Osmotic Concentration Apparatus and Method for Direct Osmotic Concentration of Fruit Juice. US Patent 5,281,430. January 25. Li, N.N., Fane, A.G., Winston, W.S., Ho M.S., B.S., Matsuura, T., 2008. Advanced Membrane Technology and Applications. John Wiley & Sons, Inc, Hoboken, NJ. Merson, R.L., Morgan, A.I., 1968. Juice concentration by reverse osmosis. Food Technol. 22 (5), 631
634. Mulder, M., 1991. Basic Principles of Membrane Technology. Kluwer Academic Publishers, London, UK. Nayak, C.A., Rastogi, N.K., 2010. Comparison of osmotic membrane distillation and forward osmosis membrane processes for concentration of anthocyanin. Desalin. Water Treat. 16, 134
145. Pepper, D., Orahard, A.C.J., Merry, A.J., 1985. Concentration of tomato juice and other fruit juice by reverse osmosis. Desalination 53 (1
3), 157
166. Veberic, R., Jakopic, J., Stampar, F., Schmitzer, V., 2009. European elderberry (Sambucus nigra L.) rich in sugars, organic acids, anthocyanins and selected polyphenols. Food. Chem. 114, 511
515.

This page intentionally left blank

Index Note: Page numbers followed by “f ” and “t” refer to figures and tables, respectively.

A Absolute temperature, 124 ABTS. See 2,20-Azino-bis(3ethylbenzothiazoline-6-sulphonic acid) (ABTS) ACE. See Angiotensin-converting enzyme (ACE) Acid juices, deacidification of, 360
361, 360f Acid pH, 364
365 Acid whey, 284, 351
352 Acinetobacter species, 80 Acoustic cavitation, 289 Acoustic streaming, 291 Activated charcoal adsorption, 142
143 Active peptides, 34
42, 35t Adsorption, 204 AEM. See Anion exchange membranes (AEM) Agricultural byproducts, 133 Agro-food byproducts, 342 processes, 389 wastewaters, 196 Artichoke wastewaters, 205
206 food wastewaters and by-products, 209
214 olive mill wastewaters, 199
205 phenolic compounds recovery from, 199
214 wastewaters from winemaking industry, 206
209 AIS. See Alcohol insoluble solids (AIS) Albumin, 34
41 Alcohol insoluble solids (AIS), 169 Aliphatic polymeric material, 90
91 Alkali agents, 234
235 Alkaline, 242 lignin, 248
249 treatments, 234
235

α-lactalbumin (α-LA), 52, 272, 281
282, 351
352 ALPMHIR (peptide sequence), 369
370 Aluminum oxide (Al2O3), 165 Amino acids, 357
358, 358f Angiotensin-converting enzyme (ACE), 51 Animal blood, 50 Animal origin, bioactive compounds of, 49
52 Animal-derived food waste, 51 Anion exchange membranes (AEM), 350
351 Anthocyanidins, 42
43 Anthocyanins, 175
176, 392, 393f Antidiabetic enhancing peptides, 368 Antimicrobial effects, lignin, 238
239 Antioxidant(s), 53 activity, 32, 41 capacity, 392 compounds, 206 Apparent rejection coefficient, 122
123 Apple (Malus pumila), 172 juice, 157
161, 172
174 pomace, 46 Arabinose, 134f Arabinoxylans, 48 Artichoke byproducts, 339 wastewaters, 205
206 Aspergillus niger, 169
170, 178 Aspergillus oryzae, 119
120 Asymmetric membranes, 5
6, 164, 164f. See also Ceramic membrane(s) composite, 6 integral, 6 Autochthonous human gastrointestinal lactic acid bacteria, 219 Autohydrolysis, 133 2,20-Azino-bis(3-ethylbenzothiazoline-6sulphonic acid) (ABTS), 342

403

404

Index

B Bacillus strains, 92 Bacillus subtilis, 92
93 Bacterial food-borne pathogens, 274 Banana (Musa acuminata), 177
178 juice, 177
179 Batch concentration, 308 Batch operation, 17 BDE. See Bond dissociation enthalpy (BDE) Benzene ring
benzene ring interactions, 205 Bergamot, 212
213 β-galactosidase, 120
121, 140
141 β-Glucans, 45
46 β-lactoglobulin (β-LG), 272, 351
352 Bifidogenic effect, 139 Bioactive compounds, 32, 196 of animal origin, 49
52 from plant and animal-derived products or waste, 35t recovery from conventional and nonconventional sources, 33
52 lipids, 48
49 polyphenols, 42
45 polysaccharides, 45
48 proteins and active peptides, 34
42 Bioactive peptide fraction separation, 367
368, 367f Bioemulsifiers, 80
84 Biological activities, 41 Biological oxygen demand (BOD), 319 Biopolymers, 230 Biosurfactants, 79
84, 81t downstream processing, 87
91 membrane choice, 100
101 cleaning, 101
102 production, 84
87, 85t UF of lipopeptide biosurfactants, 92
96 of MEL, 98
100 of Rhamnolipid biosurfactants, 97
98 Biotechnological applications, 199 Biotechnological processes, 114 Black Kraft liquor fractionation, 245
246 Blanching, 159t bLF. See Bovine lactoferrine (bLF) Blood orange juice, 175 BOD. See Biological oxygen demand (BOD) Bond dissociation enthalpy (BDE), 230
231 Bovine lactoferrine (bLF), 370
371 Bovine serum albumin (BSA), 272

Brewer’s spent grain, 48 BSA. See Bovine serum albumin (BSA) By-products, 209
214

C CA. See Cellulose acetate (CA) Calcium, 268
269 Calcium lignosulfonate, 240 Calcium phosphate, 288, 322 Calcium tartrate (CaT), 354 CAP. See Cellulose acetate phthalate (CAP) Carboxylic acids, 80 Carotenoids, 35t, 240 Cascade fractionation process, 248 Cascade NF, 308 Casein(s) micelles, 272
273 proteins, 272 Cassava wastewater, 95
96 CaT. See Calcium tartrate (CaT) Cation exchange membranes (CEM), 350
351 Cellophane membrane, 386 Cellulose acetate (CA), 166 asymmetric membranes, 1
3 membranes, 386
387 Cellulose acetate phthalate (CAP), 166 Cellulose nitrate (CN), 166 Cellulose triacetate (CTA), 166 Cellulosic pulp application, 248 Cellulosics, 90
91 CEM. See Cation exchange membranes (CEM) Centrifugation, 361
362 Ceramic membrane(s), 145, 244. See also Asymmetric membranes cleaning, 252 Cheese, 267
268 brine purification, 284
285 milk standardization, 279
280 processing, 278
280 whey, 364 Chemical oxygen demand (COD), 196, 286, 352
353 Chitosan oligomer separation, 371
372 Chitosanase, 371 Chlorogenic acid, 209 Chromatographic methods, 88, 198 Chronic diseases, 43
44 CIP. See Clean-in-place (CIP) Citrus fruits, 179
181 peels, 46

Index Clean-in-place (CIP), 285
286 Cleaning agents, 101 Clostridium spp., 274 Cloudy apple juice, enzymatic browning inhibition in, 359
360 CMC. See Critical micelle concentration (CMC) CN. See Cellulose nitrate (CN) Coconut milk, 387, 387f, 388f COD. See Chemical oxygen demand (COD) Coions, 307 Cold sterilization technique, 168, 176 Collagen, 50 Colloidal fouling in NF, 129 Column chromatography, 51 Commercialization, 84
86 Compact membrane modules, 1
2 Composite asymmetric membranes, 6 Concentration factor (FC), 129
130 Concentration polarization, 2
3, 15
17, 15f, 54, 100
101, 122
123, 250 Concentration-driven process, 3 Constant volume diafiltration (CVD), 328 Contamination, 157
161 Continuous operation, MF/UF/NF/RO, 18 Conventional extraction techniques, 197
198 Conventional methods, 87 Conventional nonthermal fruit juice processing techniques, 157 Conventional process, 161 Conventional separation technologies, 198
199 Conventional sources, bioactive compounds recovery from, 33
52 Coupled depectinization, 179, 180t Cranberry juice concentration integrated production of antioxidantenriched cranberry juice, 373f in phenolic compounds, 372
373 Critical micelle concentration (CMC), 79
80, 88 Critical transmembrane pressure (TMPC), 128
129, 130f CTA. See Cellulose triacetate (CTA) CVD. See Constant volume diafiltration (CVD) Cyclopeptides, 80

D Dairy industry, 319
325 concentration and demineralization of UF whey permeate, 320
322

405

concentration and demineralization of whey, 319
320 recovery of lactic acid, 322
325 Dairy proteins, 361 DCMD. See Direct contact membrane distillation (DCMD) Deacidification, 353 of acid juices, 360
361, 360f Dead-end filtration cell, 138 Degree of polarization, 130
131 Degrees of polymerization (DP), 117 Dementia, 387 Demineralization of UF whey permeate, 320
322 of whey, 319
320 Depectinization, 168
182, 171t of fruit juices, 163 process, 179 Desalting technologies, 355
356 DF. See Diafiltration (DF) DHA. See Docosahexaenoic acid (DHA) Diafiltration (DF), 246, 308 Dialysis, 7t Diananofiltration, 210
211 Dietary fiber, lignin as part of, 237
238 Digesters, 232
233 Direct contact membrane distillation (DCMD), 285 “Dissolving degree” cellulose, 249
250 DLS. See Dynamic light scattering (DLS) Docosahexaenoic acid (DHA), 49
51 Donnan exclusion, 307 Downstream processing of biosurfactants, 87
91, 89t assessment of separation performance, 91 UF process and equipment, 90
91 wastes, 196 DP. See Degrees of polymerization (DP) Driving force, 24 Dynamic light scattering (DLS), 92
93 “Dynamic membrane”, 276

E Eating habits, 113
114 “Eco-friendly” extraction processes, 44
45 ED. See Electrodialysis (ED) ED with bipolar membrane (EDBM), 350
351, 350f application in food processing, 359
366 to fruit juices, 359
361

406

Index

ED with bipolar membrane (EDBM) (Continued) to phospholipids, 364
366 to proteins, 361
364 ED with filtration membrane (EDFM), 366 applications in food processing, 366
373 to protein hydrolysates, 367
370 concentration of cranberry juice in phenolic compounds, 372
373 separation of chitosan oligomers, 371
372 of protein, 370
371 ED with ultrafiltration membrane (EDUF), 366 EDBM. See ED with bipolar membrane (EDBM) EDBM coupled with ultrafiltration (EDBMUF), 362
363 EDFM. See ED with filtration membrane (EDFM) EDUF. See ED with ultrafiltration membrane (EDUF) Effective transmembrane pressure (TMPe), 130
131 Eicosapentaenoic acid (EPA), 49
51 Electrodialysis (ED), 3, 7t, 21
24, 53, 269, 306, 350
351, 350f. See also Hydrolysis characterization of ion-exchange membranes, 22 in food processing, 351
358 plant, 23f process operation, 23
24 stack operation, 22f Electrodialysis-based separation technologies ED and EDBM, 350
351, 350f ED in food processing, 351
358 EDBM applications in food processing, 359
366 EDFM applications in food processing, 366
373 Electromembrane process, 362
363 Enzymatic/enzymes, 114, 169 browning inhibition in cloudy apple juice, 359
360 depectinization of juices, 168
171 hydrolysis, 117, 369
370 transglycosylation, 117 EPA. See Eicosapentaenoic acid (EPA) Epigallocatechin gallate, 209 Essential oils, 35t

Ethanol, 48
49, 101
102 3-Ethoxy-4-hydroxybenzaldehyde. See Ethyl vanillin molecule Ethyl vanillin molecule, 255, 255f Extended Nernst
Planck equation, 122
123 Extractable polyphenols, 43 Extraction, 207, 208f

F Fatty acids, 35t FC. See Concentration factor (FC) Feed additive, lignin as, 238 Feed composition, 131
132 Feed-and-bleed configuration, 308 FFC. See Final feed compartment (FFC) Film theory, 16, 326
327 Filtration, 136 performance, 204 process, 163
164 Filtration membrane (FM), 366 Final feed compartment (FFC), 368
369 Fish protein, 51 sauce, 356 5-Stages Universal Recovery Process, 52
53, 329
338 Flat-sheet NF membranes, 313
314 UF membranes, 201 Flavones, 42
43 Flavonoids, 42
43 Flavonols, 42
43, 201 Flux decline, 183
184, 204, 250 FM. See Filtration membrane (FM) FO. See Forward osmosis (FO) Food consumption, 239 functional compounds recovery from food processing byproducts, 329
342 industry, 113
114, 210
211, 327, 361 antimicrobial effects, 238
239 lignin application in, 236
241 lignin as feed additive, 238 lignin as part of dietary fiber, 237
238 packaging and films, 240
241 prebiotic effects of lignin and weight gain, 239
240 processing ED in, 351
358 EDBM applications in, 359
366 EDFM applications in, 366
373

Index industry, 156, 196, 198 wastewaters, 209
214 phenolic-based compounds extraction, 214
220 valorization, 197
199 Forward osmosis (FO), 269, 283, 384
385, 389f for up-concentrating food compounds, 386
389 FOS. See Fructooligosaccharides (FOS) Fouling process, 54, 287
289 Fractionation processes, 133
134 Fructooligosaccharides (FOS), 121, 143
144, 327 Fructooligosteviol, 144 Fruit juice, 156
163, 353
354 applications to deacidification of acid juices, 360
361 enzymatic browning inhibition in cloudy apple juice, 359
360 clarification, 164 processing, 310
314 analyses of biologically active compounds, 312t industries, 155, 178 NF membranes rejection toward analyzed compounds, 314f processing or preservation techniques for, 157
163, 159t treatment methods for fruit juice clarification/depectinization, 163
168 membrane, 164
166 Fruit(s), 155. See also Vegetable(s) byproducts, 33
34 Functional components for membrane separations bioactive compounds recovery, 33
52 challenges in functional food development, 32
33 separation and recovery of macro-and micromolecules, 52
67 Functional compound recovery from food processing byproducts, 329
342, 340f Functional foods, 31
32 challenges in functional food development, 32
33 Functional groups, 80 Functional ingredients, 31
32

G GABA. See Gamma-aminobutyric acid (GABA) Galactooligosaccharides (GOS), 119
120, 125f, 139
143, 139f, 327

407

Gamma-aminobutyric acid (GABA), 219, 358 Gas permeation (GP), 7t Gel filtration chromatography (GF chromatography), 144
145 Gelatin, 50 GLDIQK (peptide sequence), 369
370 Globulin, 34
41 Glycolipids, 80, 85t Glycosyl hydrolases, 119 Glycosyl residues, 131 GOS. See Galactooligosaccharides (GOS) GP. See Gas permeation (GP) Graded permeability membranes (GP membranes), 274
275 Gram-positive bacterial species, 92 Grape, 206 processing residues, 340
341 Green food processing, 306 Green solvents, 34 Guaiacyl units (G units), 230
231

H Hairy layer, 272 Helicobacter pylori, 372 Hemicelluloses, 177
178, 242 Hepatic encephalopathy, 142 HFLP ultrasound. See High-frequency lowpower ultrasound (HFLP ultrasound) High hydrostatic pressure, 159t High molecular weight compounds (HMW), 315
316 High-frequency low-power ultrasound (HFLP ultrasound), 289 Hindrance factor, 125
128 HMW. See High molecular weight compounds (HMW) Homogeneous membranes, 4, 5t Human digestive tract, 114 Hybrid recovery processes using UF, 94
95 Hydraulic permeability, 2
3 Hydroalcoholic stream, 207
208 Hydrolysis, 114, 369
370. See also Hydrolysis lignin recovery, 249
250 Hydrophobic porous membranes, 383 Hydrophobic substrates, 87
88 Hydrostatic pressure, 383
384 Hydroxycinnamic acid, 54, 201

408

Index

I IE. See Ion exchange (IE) IFC. See Initial feed compartment (IFC) IFT. See Interfacial tension (IFT) IG. See Immunoglobulins (IG) IIAEK (peptide sequence), 369
370 Immunoglobulins (IG), 52, 272 IMOS. See Isomaltooligosaccharides (IMOS) In situ ultrasound, 292
293, 294f Incentives, 33 Industrial membrane modules, 8 Initial feed compartment (IFC), 368
369 Inorganic components, 356
357 Inorganic materials, 4t Integral asymmetric membranes, 6 Integrated membrane process, 203f, 205, 210f, 212, 214f Interfacial tension (IFT), 79
80 Inuline, 114 Ion exchange (IE), 350
351 membranes, 21, 350
351 characterization, 22 resins, 352
353 Isoelectric solubilization/precipitation (ISP), 51 Isolates production, 361
363, 362f Isomaltooligosaccharides (IMOS), 137 purification, 137
138 ISP. See Isoelectric solubilization/precipitation (ISP)

J Juice(s), 179
182 enzymatic depectinization, 168
171 processing industry, 156

K KHT. See Potassium hydrogen tartrate (KHT) Kinetically controlled reaction, 119 Kluyveromyces lactis, 119
120 Kraft black liquor, 245
247 Kraft lignin, 234, 247 purification, 245
247 Kraft process, 232
234, 241
242, 248
250 Kraft pulping process, 233

Lactosucrose, 142
143 Lactulose, 142 Laminar regimen, 16 LC-PUFAs. See Long-chain polyunsaturated fatty acids (LC-PUFAs) LFHP ultrasound. See Low-frequency highpower ultrasound (LFHP ultrasound) Lignans, 43 Lignin, 230
232, 237, 242 application in food industry, 236
241 isolation, 232
236 alkaline treatments, 234
235 kraft process, 233
234 Organosolv process, 235
236 sulfite process, 232
233 linkages proportions and BDE in, 231t separation by UF, 241
257 LignoBoost, 245 Lignocellulosic biomass, 229
230 Lignocellulosic material, 114, 133 Lignosulfonate(s), 232
233 purification, 247
248 Limiting TMP (TMPL), 128
129 Linear correlation, 246
247 Lipids, 35t, 48
49 lipid-soluble minor components, 48 Lipopeptides, 80, 81t, 85t, 87
88. See also Peptide(s) mixtures separation, 95 UF of lipopeptide biosurfactants, 92
96 hybrid recovery processes using UF, 94
95 lipopeptides recovery from complex culture medium, 95
96 separation of lipopeptide mixtures, 95 surfactin separation by two-step UF method, 92
94, 92f LMW. See Low molecular weight (LMW) Long-chain arabino-XOS, 135 Long-chain polyunsaturated fatty acids (LCPUFAs), 49 Low molecular mass biosurfactants, 80 Low molecular weight (LMW), 308 Low water activity gel, 162 Low-frequency high-power ultrasound (LFHP ultrasound), 289

L

M

Lactic acid recovery, 322
325, 324f, 324t Lactose, 271, 283
284 hydrolysis, 141 lactose-derived prebiotics, 139
143

Macromolecular polyphenol, 43
44 Macromolecules, separation and recovery of, 52
67, 55t, 59t, 63t Macroscopic pretreatment (MF), 197

Index Magnesium, 268
269 Malus pumila. See Apple (Malus pumila) Mannosylerythritol lipids (MEL), 81t, 85t, 98
99 UF, 98
100 Marine microalgae strains, 49 Mass transfer, 119 coefficient, 122
123 at fluid phase circulating tangentially to membrane, 15
17 MBRs. See Membrane bioreactors (MBRs) MC. See Must concentrate (MC) MD. See Membrane distillation (MD) ME. See Membrane emulsification (ME) Meat byproducts, 50 MEL. See Mannosylerythritol lipids (MEL) Membrane based separation processes (MSP), 168 Membrane bioreactors (MBRs), 118
119, 119f, 137f oligosaccharides synthesis in, 118
122 Membrane distillation (MD), 166, 269
270, 283, 385
386 Membrane emulsification (ME), 204 Membrane osmotic distillation (MOD), 385
386, 389
395, 390t, 392t preservation of valuable compounds, 392
393 sensory evaluation, 394
395 Membrane technology, 129
130, 242
244, 247
248, 286, 310, 325 classification, 6
7 membrane separation processes, 7t electrodialysis, 21
24 materials, 3 in manufacturing of synthetic membranes, 4t membrane modules, 8
12 pervaporation, 24
27 pressure-driven membrane processes, 12
21 purification of specific oligosaccharides by, 133
145 separation and recovery of macro-and micromolecules, 52
67, 55t, 59t, 63t structures, 3
6 asymmetric membranes, 5
6 homogeneous membranes, 4, 5t microporous membranes, 4, 5t Membrane(s), 164
166, 167t based clarification of fruit and vegetable juices, 168
182

409

apple juice, 172
174 banana juice, 177
179 mosambi juice, 179 orange juice, 174
176 other juices, 179
182 pineapple juice, 176
177 based processes, 161, 164
166 choice, 100
101 classification, 166 cleaning, 101
102, 288 ultrasound application to, 292
295 filtration, 44
45, 101, 287 of XOS, 136
137 flux decay in lignin separation process, 250
253 example of fouling influence on permeate flux, 251f general influence of membrane cutoff, feed concentration, and TMP, 253f polarization concentration and gel layer effect, 251f fouling, 101
102, 250, 290, 308, 353
354 membrane-based technologies, 197 as emerging tools for food wastewaters valorization, 197
199 in phenolic recovery, 220
222 modules, 8
12 comparison in reverse osmosis, 12t hollow fiber, 9
10, 10f plate-and-frame, 8, 9f properties and applications, 11t spiral wound, 10
12, 10f tubular, 8
9 modules, 166
168 operations, 198
199, 211f processes, 1
2, 44
45, 212
213, 241, 268f, 306, 340
341 separation, 7t, 51, 117
118, 323, 339 application of sonication, 289
295 cheese processing, 278
280 commercial applications of membrane filtration processes, 269f in dairy industry, 267
268 fouling, 287
289 milk and whey, 270
273 milk processing, 273
278 waste treatment, 285
287 whey processing, 280
285 Methanol, 101 3-Methoxy-4-hydroxybenzaldehyde. See Methyl vanillin molecule Methyl vanillin molecule, 255, 255f

410

Index

MF. See Macroscopic pretreatment (MF); Microfiltration (MF) Microalgae, 49 Microbial growth control, 273
275 Microfiltration (MF), 2, 7t, 47, 53, 145, 166, 169, 197, 244
246, 284
287, 306. See also Ultrafiltration (UF) permeate, 341 Micromolecules, separation and recovery of, 52
67, 55t, 59t, 63t Microporous collagen films, 50 Microporous membranes, 4, 5t Microwave extraction, 47 Milk, 270
273 average composition of bovine milk, 271t casein micelles, 272
273 components and composition, 270
271 processing control of microbial growth, 273
275 on farm concentration, 273 milk fat fractionation, 278 milk protein fractionation, 275
277 proteins, 272 MOD. See Membrane osmotic distillation (MOD) Modern extraction techniques, 34 Molar mass, phenolic compounds, 199, 200f Molecular weight (MW), 14
15, 34
41, 90
91, 358 Molecular weight cutoff (MWCO), 14
15, 42, 53, 88, 90
91, 97, 198, 245
246, 279
280, 307, 367
368 Molecular weight distribution (MWD), 246
247 Monomer phenolic compounds, 249 Monosaccharides, 48, 117 Moroccan Valencia orange juice, 175 Mosambi (Citrus limetta), 179 juice, 179 MSP. See Membrane based separation processes (MSP) Multistage NF, 308 Multistage pressure-driven membrane processes, 197 Musa acuminata. See Banana (Musa acuminata) Must concentrate (MC), 318 Must processing, 314
318 Mutagenesis, 86 MW. See Molecular weight (MW) MWCO. See Molecular weight cutoff (MWCO) MWD. See Molecular weight distribution (MWD)

N Nanofiltration (NF), 1
2, 7t, 53, 94
95, 164, 166, 197, 244
246, 268, 284
286, 306. See also Microfiltration (MF); Ultrafiltration (UF) application in food processing products and byproducts, 330t aspects of NF membranes, 307
309 dairy industry, 319
325 in diafiltration mode, 144 fruit juice processing, 310
314 membranes, 120
121, 126t, 318, 338 NF270 membrane, 311 of oligosaccharides retention to concentration polarization, 122
124 effect of solute concentration and feed composition, 131
132 effect of temperature, 124
128 effect of transmembrane pressure, 128
131 permeate, 323, 338 recovery of functional compounds from food processing byproducts, 329
342 sugar industry, 325
328 wine and must processing, 314
318 Natural based polymers, 4t Nejayote. See Nixtamalization wastewaters Neocontaminants, 128 Nernst
Planck equation, 326
327 NF. See Nanofiltration (NF) Nixtamalization wastewaters, 213
214 Nominal molecular weight cutoff (NMWCO), 90
91 Nonconventional sources, bioactive compounds recovery from, 33
52 Nondestructive technologies, 53 Nondigestible polysaccharides, 114 Nonextractable polyphenols, 43 Nonstarch polysaccharides (NSPs), 45 Nonthermal technology, 157, 159t NSPs. See Nonstarch polysaccharides (NSPs) Nutraceuticals, 196 Nutritional parameters, 176

O Oat, 34
41 OD. See Osmotic distillation (OD) OE method. See Osmotic evaporation method (OE method) Oil-pumpkin biomass, 46

Index Oligosaccharides, 114, 118f. See also Xylooligosaccharides (XOS) nanofiltration, 122
132 prebiotics, 113
118 purification, 144
145 synthesis in membrane bioreactors, 118
122 Olive mill wastewaters (OMWs), 199
205, 338 Olive phenols, 219 Omega-3 polyunsaturated fatty acids (ω-3 PUFAs), 50
51 OMWs. See Olive mill wastewaters (OMWs) On farm concentration, 273 Orange (Citrus sinensis), 174 juice, 174
176 “Ordinary” solutes, 397 Organic membranes, 165 Organic molecule adsorption, 101 Organic nonsugar components, 356
357 Organic solvents, 87, 101 Organoleptic evaluations, 394
395 Organosolv lignin, 248
249 liquors, 242 process, 235
236 Osmosis, 383 Osmotic dehydration, 166 Osmotic distillation (OD), 172, 202
203, 385 Osmotic driven membrane processes draw solutions, 396
398, 396t, 397t forward osmosis for up-concentrating food compounds, 386
389 MOD, 389
395 osmotic processes, 384t Osmotic evaporation method (OE method), 173, 385 Osmotic membrane processes, 385 Osmotic pressure, 383
384 Ozone processing, 159t

P p-hydroxyphenyl units (H units), 230
231 PA. See Polyamide (PA) PAA-Na salt. See Poly(acrylic acid) sodium salt (PAA-Na salt) Palm pressed fibers, 49 PAN. See Polyacrylonitrile (PAN) PAO. See Pressure assisted osmosis (PAO) Passionfruit, 179
181 Pasteurization, 159t

411

PC. See Polycarbonate (PC) PE. See Polyethylene (PE) Pectates, 162 Pectic acid, 162 Pectic polysaccharides, 45
46 Pectin-derived oligosaccharides (POS), 144
145 Pectin(s), 45
46, 161
162, 173
174, 177
178 heterogeneity of pectin structure, 46 removal and clarification of juices depectinization and membrane based clarification, 168
182 membrane based separation processes, 168 processing or preservation techniques for fruit and vegetable juice, 157
163 treatment methods for fruit juice clarification/depectinization, 163
168 Pectinase treatment, 170 Pectinates, 162 Peptide(s), 50, 368. See also Lipopeptides separation, 369
370, 369f Pervaporation (PV), 7t, 24
27, 24f, 166 PES. See Polyethersulfone (PES) Phase inversion process, 90
91 Phenolic acids, 43 Phenolic compounds, 196, 215t chemical structure, 199, 200f cranberry juice concentration in, 372
373 economic overview of membrane-based technologies, 220
222 extracted from food wastewaters, 214
220 membrane-based technologies as emerging tools, 197
199 recovery from agro-food wastewaters, 199
214 recovery of phenolic compounds from agrofood wastewaters, 199
214 Phenolic fragments, 238
239 Phenolic recovery, membrane-based technologies in, 220
222 Phenylpropanoid units, 230
231 Phospholipid(s), 48 application to, 364
366 phospholipid-enriched fraction production, 365
366 pH
temperature flocculation, 205 Phytochemicals, 32, 49 Pineapple (Ananas comosus), 156, 176 juice, 176
177 Plasma protein, 50

412

Index

Plate polymeric membranes, 248 Polar interactions, 202 Polar lipids, 364 Polar micromolecules, 54 “Polarity resistance” phenomenon, 202 Polarization concentration, 251
252 retention to concentration, 122
124 Poly(acrylic acid) sodium salt (PAA-Na salt), 283 Polyacrylonitrile (PAN), 90
91, 165 Polyamide (PA), 166 Polyamide TFC NF membranes, 311 Polycarbonate (PC), 166 Polyethersulfone (PES), 90
91, 166, 313, 370
371 Polyethylene (PE), 165 Polyimide, 90
91 Polymeric biosurfactants, 80 Polymeric materials, 90
91 Polymeric membranes, 145
146, 200
201, 248
249, 308 Polymeric NF membranes, 307 Polyphenoloxidases (PPO), 177
178, 359
360 Polyphenols, 35t, 42
45, 199
200, 209, 338 Polypropylene (PP), 166 Polysaccharides, 35t, 45
48, 209, 230 Polysulfone (PSU), 54, 90
91, 101
102, 165, 247
248 Polytetrafluoroethylene (PTFE), 283 Polyunsaturated fatty acids (PUFAs), 48 Polyvinyl alcohol (PVA), 166 Polyvinyl chloride (PVC), 166 Polyvinylidene difluoride (PVDF), 90
91, 166, 310
311 Pore size distribution (PSD), 275 Porous membranes, 164 POS. See Pectin-derived oligosaccharides (POS) Potassium hydrogen tartrate (KHT), 354 PP. See Polypropylene (PP) PPO. See Polyphenoloxidases (PPO) Prebiotic(s), 114 effects of lignin and weight gain, 239
240 oligosaccharides, 114, 115t “Precheese”, 280 Prehydrolyzed liquors, 249
250 Preservation of fruits and vegetables dates, 155 processing or preservation techniques for fruit and vegetable juice, 157
163 of valuable compounds, 392
393

Pressure assisted osmosis (PAO), 384
385 Pressure retarded osmosis (PRO), 384
385 Pressure-driven membrane processes, 12
21, 13f, 198, 220
221, 306
307, 310, 338, 385 case study applications, 19
21 batch sheep milk concentration, 19f fractionation of sheep cheese whey, 20f technoeconomical analysis, 19
21, 20t mass transfer at fluid phase circulating tangentially to membrane, 15
17 membrane characterization, 12
15 operation modes, 17
18 batch operation, 17 continuous operation, 18 PRO. See Pressure retarded osmosis (PRO) Proanthocyanidins, 372 Product yield, 87
88 Prolamin, 34
41 Protein(s), 34
42, 35t, 96
98 applications to protein hydrolysates, 367
370 improvement of protein hydrolysate bioactivity, 368
369 separation of bioactive peptide fractions, 367
368 simultaneous hydrolysis and separation of peptides, 369
370 denaturation, 273
274 fractionation, 363
364 production of isolates, 361
363 separation, 370
371 PS. See Polysulfone (PSU) PSD. See Pore size distribution (PSD) Pseudomonas aeruginosa YPJ-80, 97 Pseudozyma species, 98
99 P. tsukubaensis, 99 PSF. See Polysulfone (PSU) PSU. See Polysulfone (PSU) PSU surface modified by poly-ethylene glycol (PSU-g-PEG), 101
102 PTFE. See Polytetrafluoroethylene (PTFE) PUFAs. See Polyunsaturated fatty acids (PUFAs) Pulp and paper industry, 241 Pulping method, 245 Pulsed electric field, 159t Purification, 207, 208f of fructooligosaccharides, 143
144 of galactooligosaccharide and lactosederived prebiotics, 139
143 of IMOS, 137
138

Index purification of oligosaccharides, 144
145 of specific oligosaccharides by membrane technology, 133
145 of xylooligosaccharides, 133
137 PV. See Pervaporation (PV) PVA. See Polyvinyl alcohol (PVA) PVC. See Polyvinyl chloride (PVC) PVDF. See Polyvinylidene difluoride (PVDF)

R Raw fruit juice, 353 RC. See Regenerated cellulose (RC) Recovered proteins, 41 Recovery factor (Δ), 309 Rectified must concentrate (RMC), 318 Regenerated cellulose (RC), 94 membranes, 202 Rejection coefficient (R), 91 Renewable sources, 33 Response surface methodology (RSM), 169
170 Retention to concentration polarization, 122
124 Reverse osmosis (RO), 1
2, 7t, 164, 166, 172, 197, 268, 282
283, 306, 384
385 Reversible osmosis, 244 Rhamnogalacturonan I pectins (RG-I pectins), 161
162 Rhamnogalacturonan II pectins (RG-II pectins), 161
162 Rhamnolipids, 81t, 85t UF of rhamnolipid biosurfactants, 97
98 Rice bran, 48
49 byproducts, 41
42 RMC. See Rectified must concentrate (RMC) RO. See Reverse osmosis (RO) RSM. See Response surface methodology (RSM)

S Salting/brining, 284
285 Salty whey, 286
287 Sauce, 355
356 SCC. See Spinning cone column (SCC) SDS. See Sodium dodecyl sulfate (SDS) Selective ED (SED), 352
353 Selective fermentation, 142
143 Selective membranes, 383 SelRO Caustic Recovery System, 286

413

Semipermeable membrane, 383 Sensory evaluation, 394
395, 394t, 395t Separation, 164, 196, 385 of phenolic micromolecules, 202 of raffinose, 141
142 Serum protein concentrate (SPC), 276 Serum protein isolate (SPI), 276 Serum proteins. See Whey—proteins Shelf life of processed juice, 157 of typical natural juices, 158t Short-chain fatty acids, 114 Simulating moving bed chromatography, 142
143 Skim milk, 275
276 Sodium dodecyl sulfate (SDS), 84
86 Sodium hydroxide (NaOH), 233, 235 Sodium sulfide (Na2S), 233 Sol
gel process, 145 Soluble dietary fibers, 45 Solute concentration effect, 131
132 Solution
diffusion model, 25 Solvent(s), 43 extraction, 117, 199
200 precipitation adsorption, 117 Sonication application, 289
295 application of ultrasound to membrane cleaning, 292
295 mechanisms of ultrasonic membrane cleaning in dairy applications, 291
292 physical and chemical effects of ultrasound, 289
290 ultrasonic reduction of fouling buildup, 290 Sophorolipids, 81t, 85t Soy proteins, 361 sauce, 355
356 Soybean oligosaccharides, 114, 144
145, 357 SPC. See Serum protein concentrate (SPC) SPI. See Serum protein isolate (SPI) Spinning cone column (SCC), 315 SPs. See Starche polysaccharides (SPs) ST. See Surface tension (ST) Starche polysaccharides (SPs), 45 Sterilization, 159t Stilbens, 43 Stirred cell device, 88
90 Sugar(s), 356
357 acid pectin gel, 162 industry, 325
328, 329f

414

Index

Sulfite liquors, 247 process, 232
233 Sulfur dioxide (SO2), 232
233 Sulfur-free lignin fractionation, 248
249 Sulfuric acid (H2SO4), 232
233 Supramolecular structures, 87 Surface tension (ST), 79
80 Surfactin separation by two-step UF method, 92
94, 92f Sustainable extractive technology, 206 Sustainable flux, 128
129 Sustainable process, 210 Sustainable water management, 143, 196 Sweet whey, 351
352 Symmetric membranes, 164, 164f Synthetic membranes, 1
2 Synthetic polymers, 4t Syringyl units (S units), 230
231

T TAA. See Total antioxidant activity (TAA) TDS concentration. See Total dissolved solids concentration (TDS concentration) Technoeconomical analysis, 19
21 Temperature effect, 124
128 TFC. See Thinfilm composite (TFC) TFC NF spiral-wound membrane, 322 Thermal drying process, 159t Thermal technology, 157
161, 159t, 198 Thinfilm composite (TFC), 308 Threshold flux, 128
129 theories, 204 TMP. See Transmembrane pressure (TMP) Tocotrienols, 48
49 Total antioxidant activity (TAA), 312 Total dissolved solids concentration (TDS concentration), 286
287 Total recycle configuration, 308 Total solid substance (TSS), 391, 391f Total soluble sugar (TSS), 172 Traditional cold tartaric stabilization method, 23
24 Transgalactosylation products, 139 Transglycosylation, 114, 117 reactions, 114 Transmembrane curves, 132 Transmembrane pressure (TMP), 128
129, 172, 252
253, 311 effect, 128
131 gradient, 269
270

Trehalose lipids, 87
88 TSS. See Total solid substance (TSS); Total soluble sugar (TSS) Turbulent regimen, 16 Two-step UF method, surfactin separation by, 92
94, 92f

U UF. See Ultrafiltration (UF) UHT. See Ultrahigh temperature (UHT) Ultra-Osmosis process, 286
287 Ultrafiltration (UF), 1
2, 7t, 42, 51
53, 88, 145, 164, 166, 169, 197, 244
246, 268, 279, 281t, 285
286, 306, 361
362. See also Nanofiltration (NF) concentration and demineralization of UF whey permeate, 320
322 hybrid recovery processes using, 94
95 lignin separation, 241
257 hydrolysis lignin recovery, 249
250 Kraft lignin purification, 245
247 lignin characteristics in function of pulping process, 243t lignosulfonates purification, 247
248 membrane flux decay in, 250
253 sulfur-free lignin fractionation, 248
249 of lipopeptide biosurfactants, 92
96 membranes, 207, 287, 338 process and equipment, 90
91 purification as source for food industry, 253
257 characteristics and operating conditions of UF processes, 254t vanillin purification by UF, 255
257 of Rhamnolipid biosurfactants, 97
98 technology, 47 Ultrahigh temperature (UHT), 273
274 Ultrahigh-pressure extraction, 47 Ultrasonic bath, membrane modules submerging in, 292 Ultrasonic membrane cleaning mechanism in dairy applications, 291
292 Ultrasonic reduction of fouling buildup, 290 Ultrasound, 289 application to membrane cleaning, 292
295 membrane modules submerged in ultrasonic bath, 292 preventing membrane damage from ultrasound, 292 in situ ultrasound, 292
293, 294f

Index applied to recirculating cleaning solution or as pretreatment, 293
295 physical and chemical effects, 289
290 Ultraviolet (UV), 389
390 irradiation, 159t spectrometric method, 389
390 “Unconventional” solutes, 397 Uniform transmembrane pressure system (UTMP system), 274
275 Up-concentrating food compounds, 386
389 UTMP system. See Uniform transmembrane pressure system (UTMP system) UV. See Ultraviolet (UV)

V Vacuum method, 24 Value-added applications, 241
242 Vanillin, 247
248 purification by UF, 255
257 Variable volume diafiltration (VVD), 328 VCFs. See Volume concentration factors (VCFs) VDF. See Volume dilution factor (VDF) Vegetable(s), 155. See also Fruit(s) byproducts, 33
34 juice, 157
163 processing or preservation techniques for, 157
163 wastes, 46 Vivinal GOS, 139
140 VLVLDTDYKK sequence, 369
370 Volume concentration factors (VCFs), 277 Volume dilution factor (VDF), 322 Volume reduction factor (VRF), 280, 309 Volume reduction ratio (VRR), 274
275 VRF. See Volume reduction factor (VRF) VRR. See Volume reduction ratio (VRR) VVD. See Variable volume diafiltration (VVD)

W Waste treatment, 285
287 Wastewaters from winemaking industry, 206
209 Water transport in liquid phase, 384t Whey, 270
273, 280, 351
353 cheese brine purification, 284
285 concentration, 282
283

415

concentration and demineralization, 319
320, 321t demineralization, 283
284 fractionation process, 51
52 lipid precipitation, 364
365 phospholipids, 364 processing, 280
285 protein fractionation, 281
282 proteins, 272, 276, 281
282 WPC, 280
281 Whey protein concentration/concentrate (WPC), 276, 280
281, 352
353 electroacidification, 364
365 Whey protein isolate (WPI), 363
364 WHO. See World Health Organization (WHO) Wine, 354
355 processing, 314
318 production of low sugar content fruit juices, 317f two-stage NF process for production, 316f wastewaters from winemaking industry, 206
209 Winery sludge, 207 Wood pulp, 232
233 World Health Organization (WHO), 48 WPC. See Whey protein concentration/ concentrate (WPC) WPI. See Whey protein isolate (WPI)

X XOS. See Xylooligosaccharides (XOS) Xylan, 134f Xylitol, 357 Xylooligosaccharides (XOS), 121
122, 326
327. See also Oligosaccharides purification, 133
137, 134f Xylose, 134f

Y Yield of recovery, 123
124 Yogurt, 267
268

Z Zero fouling, 128
129 Zeta potential (ZP), 373 Zirconium oxide (ZrO2), 165