Functional Food Ingredients and Nutraceuticals Processing Technologies

Citation preview

2441_title 7/18/06 11:28 AM Page 1

Functional Food Ingredients and Nutraceuticals Processing Technologies Edited by

John Shi, Ph.D.

Boca Raton London New York

CRC is an imprint of the Taylor & Francis Group, an informa business

© 2007 by Taylor & Francis Group, LLC

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487‑2742 © 2007 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Printed in the United States of America on acid‑free paper 10 9 8 7 6 5 4 3 2 1 International Standard Book Number‑10: 0‑8493‑2441‑6 (Hardcover) International Standard Book Number‑13: 978‑0‑8493‑2441‑3 (Hardcover) This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the conse‑ quences of their use. No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www. copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978‑750‑8400. CCC is a not‑for‑profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging‑in‑Publication Data Functional food ingredients and nutraceuticals : processing technologies / edited by John Shi. p. cm. ‑‑ (Functional foods & nutraceuticals series ; 9) Includes bibliographical references and index. ISBN 0‑8493‑2441‑6 (alk. paper) 1. Functional foods. I. Shi, John. II. Series. QP144.F85F85 2006 613.2‑‑dc22 Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

© 2007 by Taylor & Francis Group, LLC

2006043862

2441_C000.fm Page v Thursday, July 20, 2006 9:07 AM

Foreword The linkages between diet and health, and their economic implications, continue to be elucidated and to gain public acceptance as credible science in this fascinating field of study. Knowledge of the specific contributions that functional foods, bioactive compounds, and nutraceuticals make to our health has grown immensely over the past decade. So, too, has the desire of consumers to purchase and integrate these materials into healthful diets, and for mainstream retail chains to give these products lucrative shelf space increased. To effectively disseminate this knowledge and achieve broad public benefit from this work, it is essential that emerging factual information in the scientific literature be distilled, assembled, and consolidated into more concise communications, as presented here, that can be put into action throughout the food continuum. This ninth volume in the CRC Press Functional Foods and Nutraceuticals Series, edited by Dr. G. Mazza (AAFC-PARC, Summerland, BC), continues the work of the earlier volumes and moves farther along the continuum to address the most recent advances in food engineering and processing. These advances hold the greatest promise for the immediate future to achieve stable, high-volume production of functional food and nutraceutical products of defined and reliable composition, thereby fostering the health and economic benefits that could ensue. Dr. John Shi, an internationally wellknown scientist in this field, and coeditor of an earlier volume of the series, steps out independently to take the role of editor as well as contributing author on this volume. The team of contributing authors, representing five continents, reflects the international distribution of interest and expertise on this topic, and the essential widespread collaboration that is required to generate advances most effectively. In 16 chapters, this volume provides a sequential study of key factors in the preparation of functional foods and nutraceuticals, from the selection of sources through the extraction, purification, decontamination, packaging, and preservation of a variety of products in these categories. Where the earlier volumes focused on an array of products and their biochemical constituents, this volume is process oriented, although it touches on a variety of products as well as their characteristics and how individual properties influence the suitability of the processes described. The authors and editor have done a masterful job and are to be congratulated for their success in selecting the most significant issues and the most promising opportunities for presentation in this enlightening publication. The dissemination of the information this volume contains will accelerate the integration of these approaches and technologies into the production processes of tomorrow. This book will be an excellent resource for students of food science, life sciences, as well as food industry professionals interested in functional foods. Yvon Martel, Ph.D. A/Assistant Deputy Minister Agriculture and Agri-Food Canada

© 2007 by Taylor & Francis Group, LLC

2441_C000.fm Page vii Thursday, July 20, 2006 9:07 AM

Preface People from North America and the European community are committing more time and effort to have greater control over their health by exploring alternative or herbal medicines and natural health products for disease prevention and better health. There has been a growing interest in the role of special micronutrients (phytochemicals) on human health and well-being. The relationships among special food components, their physiological functionality, and health benefits have been revealed progressively in the recent years. Functional foods contain significant levels of natural extracts, concentrates, or natural ingredients that are extracted from natural sources. These can reduce current healthcare costs by health improvement and disease prevention. During the past decade, the trend of functional food consumption was consumer driven. It also serves the needs of the aging population that wants greater control over their health and well-being. This trend is expected to continue, and as a result, scientific information on all aspects of functional foods is vital to the advancement of this emerging sector. The increase in consumer demand for functional foods has prompted international health organizations and government agencies to develop specific guidelines for their production and use. The scientific community must, therefore, utilize modern technologies to ensure efficacy and safety in the manufacturing of functional foods. Manufacturers are always eager to fulfill the consumer’s desire for functional food products that could be used to promote health. In order to provide a better understanding and to disseminate the latest developments in this rapidly expanding field, this ninth volume of the CRC Press Functional Foods and Nutraceuticals Book Series, Functional Foods and Nutraceuticals: Processing Technologies, was developed. Sixteen chapters in this book cover a broad spectrum of functional foods from biological material, applications of engineering techniques in functional food production, process engineering and modeling, functional food bioavailability, to product quality. The emphasis is on (1) applications of engineering techniques such as high pressure, supercritical fluid, membrane, microencapsulation, and molecular distillation in the processing of functional foods; (2) stability of bioactive components and antioxidative properties during processing and shelf life; (3) improvement in bioavailability of bioactive components by physical and chemical methods; and (4) mechanisms of antioxidant action and clinical and epidemiological evidence of functionality. The contributing authors are international experts in their respective fields, and I am grateful to each and every one of them for their thoughtfulness in contributions to this book. This book will be of interest to a wide spectrum of professionals from food scientists and technologists, nutritionists, biochemists, and engineers to entrepreneurs worldwide. It will also serve as a unique reference for food scientists for the R&D departments of food companies that are working with functional foods

© 2007 by Taylor & Francis Group, LLC

2441_C000.fm Page viii Thursday, July 20, 2006 9:07 AM

and ingredients. Additionally, it will serve as a source of basic information for college and university students majoring in food science and technology, food processing, and engineering. Readers will obtain sound scientific knowledge of engineering techniques and the quality of functional foods and nutraceuticals. John Shi, Ph.D. Editor

© 2007 by Taylor & Francis Group, LLC

2441_C000.fm Page ix Thursday, July 20, 2006 9:07 AM

Series Editor’s Preface The Functional Foods and Nutraceuticals Book Series, launched in 1998, was developed to provide a timely and comprehensive treatment of the emerging science and technology of functional foods and nutraceuticals which are shown to play a role in preventing or delaying the onset of diseases, especially chronic diseases. The first eight volumes in the series, Functional Foods: Biochemical and Processing Aspects, Volumes 1 and 2; Herbs, Botanicals and Teas; Methods of Analysis for Functional Foods and Nutraceuticals; Handbook of Fermented Functional Foods; Handbook of Functional Dairy Products; Handbook of Functional Lipids; and Dictionary of Functional Foods and Nutraceuticals, have received broad acceptance by food, nutrition, and health professionals. The latest volume, Functional Foods and Nutraceuticals: Processing Technologies, edited by Dr. John Shi, addresses the most recent advances in processing technologies for functional food ingredients and nutraceuticals. Distinctive features of this book include in-depth treatments of the peer-reviewed literature on supercritical fluid extraction, pressurized low polarity water extraction, membrane separation, distillation, dehydration, food pasteurization, and sterilization with high pressure, microencapsulation of omega-3 fatty acids, and bioprocessing. Other topics addressed include microbial modeling for bioreactor design, biochemical reactions in supercritical fluids, modeling of supercritical fluid extraction of plant material, stability of lycopene during processing, functional foods packaging, fruits with high antioxidant activity, and biological antioxidation mechanisms quenching of peroxynitrite. The book contains 16 excellent chapters written by 32 international experts at the forefront of functional food and nutraceutical science and technology. It is hoped that the effort will be beneficial to process engineers; food, nutrition, and health practitioners; and students, researchers, and entrepreneurs in industry, government, and university laboratories. Earlier volumes in the series addressed a range of topics and include: Functional Foods: Biochemical and Processing Aspects, Volume 1, the first volume of the series, edited by G. Mazza, is a bestseller, and is devoted to functional food products from oats, wheat, rice, flaxseed, mustard, fruits, vegetables, fish, and dairy products. In Volume 2, edited by Drs. John Shi, G. Mazza, and Marc Le Maguer, the focus is on the latest developments in the chemistry, biochemistry, pharmacology, epidemiology, and engineering of tocopherols and tocotrienols from oil and cereal grain, isoflavones from soybeans and soy foods, flavonoids from berries and grapes, lycopene from tomatoes, limonene from citrus, phenolic diterpenes from rosemary and sage, organosulfur constitutes from garlic, phytochemicals from echinacea, pectin from fruit, and omega-3 fatty acids and docosahexaenoic acid from flaxseed and fish products.

© 2007 by Taylor & Francis Group, LLC

2441_C000.fm Page x Thursday, July 20, 2006 9:07 AM

The book Herbs, Botanicals and Teas, edited by Drs. G. Mazza and Dave Oomah, provides an in-depth literature review of the scientific and technical information on the chemical, pharmacological, epidemiological, and clinical aspects of garlic, ginseng, Echinacea, ginger, fenugreek, St. John’s wort, ginkgo biloba, kava kava, goldenseal, saw palmetto, valerian, evening primrose, liquorice, bilberries and blueberries, and green and black teas. The book, which is superbly referenced, also contains chapters on international regulations and quality assurance and control for the herbal and tea industry. The volume Methods of Analysis for Functional Foods and Nutraceuticals, edited by Dr. W. Jeffrey Hurst, presents advanced methods of analysis for carotenoids, phytoestrogens, chlorophylls, anthocyanins, amino acids, fatty acids, flavonoids, water-soluble vitamins, and carbohydrates. The fifth volume of the series, Handbook of Fermented Functional Foods, edited by Dr. Edward R. Farnworth, provides a comprehensive, state-of-the-art treatment of the scientific and technological information on the production of fermented foods, the microorganisms involved, the changes in composition that occur during fermentation and, most importantly, the effects of these foods and their active ingredients on human health. The Handbook of Functional Dairy Products, edited by Drs. Colette Shortt and John O’Brien, contains outstanding chapters dealing with probiotic lactobacilli and bifidobacteria, lactose hydrolyzed products, trans-galactooligosaccharides as prebiotics, conjugated linoleic acid (CLA) and its antiatherogenic potential and inhibitory effects on chemically induced tumors, immuno-enhancing properties of milk components and probiotics, and calcium and iron fortification of dairy products. On lipids, we have the volume, Handbook of Functional Lipids, edited by Professor Casimir C. Akoh, which presents up-to-date information on all major scientific and technological aspects of functional lipids, including isolation, production, and concentration of functional lipids; lipids for food functionality; lipids with health and nutritional functionality; and the role of biotechnology for functional lipids. Finally, the Dictionary of Functional Foods and Nutraceuticals, written by N. A. Michael Eskin and Snait Tamir, is essentially a mini-encyclopedia that provides the reader with valuable and up-to-date information on the occurrence, chemistry, and biological activity/efficacy of 480 functional foods and nutraceuticals. The information is based solely on peer-reviewed literature, and it is presented alphabetically in a clear and concise manner. G. Mazza, Ph.D., FCIFST Series Editor

© 2007 by Taylor & Francis Group, LLC

2441_C000.fm Page xi Thursday, July 20, 2006 9:07 AM

Acknowledgments This book is a product of collaborative efforts, and a number of scientists gave their valuable time to review the manuscripts. The editor is indebted to their contributions and support that improved this book significantly. The reviewers include Dr. Ireneo Kikic (University of Trieste, Italy); Dr. Thomas Gamse (Graz University of Technology, Austria); Drs. Yukio Kakuda, Gauri S. Mittal, and Warrer Stiver (University of Guelph, Canada); Drs. Selma Guigard and Marleny Saldana (University of Alberta, Canada); Dr. Paulyn Appah (Food Development Centre, Manitoba, Canada); and Drs. Giuseppe Mazza, Christopher Young, and Lamin S. Kassama (Agriculture and Agri-Food Canada). The editor also would like to express appreciation to Dr. Yvon Martel (A/Assistant Deputy Minister, Agriculture and Agri-Food Canada) for his kind preparation of the foreword for this book; and to Drs. John Lynch (Science Director) and Maria Nazarowec-White (Program Coordinator) for their help and support. The editor also wishes to acknowledge the encouragement and help from Dr. Jerry King (Los Alamos National Laboratory, Chemistry Division, United States); Dr. Asbjørn Gildberg (Norwegian Institute of Fishery and Aquaculture, Norway); Drs. Amparo Chiralt and Pedro Fito (Polytechnic University of Valencia, Spain); Drs. Albert Ibarz and Joaquin Giner Segui (University of Lleida, Spain); Dr. Sam K. C. Chang (North Dakota State University, United States); Dr. James H. May (University of Hawaii at Manoa, United States); Dr. Yueming Jiang (South China Institute of Botany, Chinese Academy of Sciences, China); Dr. Eleanor Riemer, Ms. Susan Lee, and Ms. Patricia Roberson (CRC/Taylor & Francis) for their support and encouragement in the preparation of the book proposal and manuscript.

© 2007 by Taylor & Francis Group, LLC

2441_C000.fm Page xiii Thursday, July 20, 2006 9:07 AM

Editor Dr. John Shi is a senior research scientist in the Federal Department of Agriculture and Agri-Food Canada, and is also an adjunct professor of food engineering, at the School of Engineering, University of Guelph. He is co-editor of two books, Functional Foods II – Biochemical and Processing Aspects, and Asian Functional Foods, published by CRC Press, now Taylor & Francis. He graduated from Zhejiang University, China, earned a Masters degree in 1985, and Ph.D. in 1994 from Polytechnic University of Valencia, Spain. Dr. Shi is an international editor of the Journal of Food Science and Nutrition and Nutraceuticals and Foods, and also a member of the editorial boards of the Journal of Medicinal Foods and Journal of Agriculture, Food and Environment. As a post-doctoral fellow, he conducted research at North Dakota State University, USA; and as visiting professor he conducted international collaborative research at the Norwegian Institute of Fishery and Aquaculture, Norway, and at Lleida University, Spain. He was keynote speaker at a number of international conferences in the United States, Canada, Japan, China, Korea, Thailand, Spain, and Colombia. He has published more than ninety research papers in international scientific journals and twenty book chapters. His current research interests focus on processing technologies to separate health-promoting components from natural products and to develop functional foods.

© 2007 by Taylor & Francis Group, LLC

2441_C000.fm Page xv Thursday, July 20, 2006 9:07 AM

Contributors Mary Ann Augustin, Ph.D. Science Manager Food Chemistry and Formulation Science Food Science Australia Victoria, Australia

Caye M. Drapcho, Ph.D. Associate Professor Department of Biosystems Engineering Clemson University Clemson, SC, United States

Alberto Bertucco, Ph.D. Professor Institute of Chemical Engineering University of Padova Padova, Italy

Hideo Etoh, Ph.D. Professor Faculty of Agriculture Shizuoka University Shizuoka, Japan

Mércia de Fátima M. Bettini, Ph.D. Technical Director of Flavor Tec - Aromas de Frutas Ltda Pindorama, Brazil

Maja Habulin, Ph.D. Associate Professor Faculty of Chemistry and Chemical Engineering University of Maribor Maribor, Slovenia

Juan Eduardo Cacace, M.S. Visiting Research Engineer Pacific Agri-Food Research Center Agriculture and Agri-Food Canada Summerland, BC, Canada Feng Chen, Ph.D. Associate Professor Department of Food Science and Human Nutrition Clemson University Clemson, SC, United States Louise Deschênes, Ph.D. Research Scientist Food Research and Development Centre Agriculture and Agri-Food Canada St. Hyacinthe, QC, Canada

© 2007 by Taylor & Francis Group, LLC

Yueming Jiang, Ph.D. Professor South China Institute of Botany Chinese Academy of Sciences Guangzhou, China Yukio Kakuda, Ph.D. Professor Department of Food Science University of Guelph Guelph, ON, Canada Lamin S. Kassama, Ph.D. Post-doctoral Research Fellow Guelph Food Research Centre Agriculture and Agri-Food Canada Guelph, ON, Canada

2441_C000.fm Page xvi Thursday, July 20, 2006 9:07 AM

ˇ Zeljko Knez, Ph.D. Professor Faculty of Chemistry and Chemical Engineering University of Maribor Maribor, Slovenia Aditya Kulkarni, M.S. Research Assistant Faculty of Agriculture Shizuoka University Shizuoka, Japan Ashwani Kumar, Ph.D. Research Scientist Institute for Chemical Process and Environmental Technology National Research Council of Canada Ottawa, ON, Canada Chiara G. Laudani, Ph.D. Associate Professor Faculty of Chemistry and Chemical Engineering University of Maribor Maribor, Slovenia Giuseppe Mazza, Ph.D., FCIFST Principal Research Scientist Pacific Agri-Food Research Center Agriculture and Agri-Food Canada Summerland, BC, Canada Gauri S. Mittal, Ph.D. Professor School of Engineering University of Guelph Guelph, ON, Canada Valérie Orsat, Ph.D. Research Engineer Bioresource Engineering Department Macdonald Campus of McGill University Ste-Anne de Bellevue, QC, Canada

© 2007 by Taylor & Francis Group, LLC

ˇ ˇ Ph.D. Mateja Primozic, Associate Professor Faculty of Chemistry and Chemical Engineering Laboratory for Separation Processes University of Maribor Maribor, Slovenia G. S. Vijaya Raghavan, Ph.D. Professor Bioresource Engineering Department Macdonald Campus of McGill University Ste-Anne de Bellevue, QC, Canada Luz Sanguansri, Ph.D. Research Scientist Food Science Australia Victoria, Australia John Shi, Ph.D. Research Scientist Guelph Food Research Centre Agriculture and Agri-Food Canada Guelph, ON, Canada Helena Sovová, Ph.D. Professor Institute of Chemical Process Fundamentals Academy of Sciences of the Czech Republic Prague, Czech Republic Sara Spilimbergo, Ph.D. Professor Department of Materials Engineering and Industrial Technologies Faculty of Engineering University of Trento Trento, Italy

2441_C000.fm Page xvii Thursday, July 20, 2006 9:07 AM

Terry H. Walker, Ph.D. Associate Professor Department of Biosystems Engineering Clemson University Clemson, SC, United States Shiow Y. Wang, Ph.D. Plant Physiologist/Biochemist Fruit Laboratory ARS-USDA Beltsville, MD, United States

© 2007 by Taylor & Francis Group, LLC

Xiaoqin Zhou, M.S. Professional Engineer Department of Chemical Engineering University of Waterloo Waterloo, ON, Canada

2441_C000.fm Page xix Thursday, July 20, 2006 9:07 AM

Table of Contents PART I

Supercritical Fluid Extraction Technology

Chapter 1 Supercritical Fluid Technology for Extraction of Bioactive Components ...............3 John Shi, Lamin S. Kassama, and Yukio Kakuda Chapter 2 Solubility of Food Components and Product Recovery in the Supercritical Fluid Separation Process....................................................................45 John Shi and Xiaoqin Zhou Chapter 3 Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials .........................................................................................................75 Helena Sovová Chapter 4 Biochemical Reactions in Supercritical Fluids .....................................................111 ˇ ˇ ˇ Zeljko Knez, Chiara G. Laudani, Maja Habulin, Mateja Primozic

PART II

Pressurized Low Polarity Water Extraction, Membrane Separation, Distillation, and Dehydration Technologies

Chapter 5 Pressurized Low Polarity Water Extraction of Biologically Active Compounds from Plant Products ..........................................................................135 Juan Eduardo Cacace and Giuseppe Mazza Chapter 6 Purification of Orange Peel Oil and Oil Phase by Vacuum Distillation ..............157 Mércia de Fátima M. Bettini

© 2007 by Taylor & Francis Group, LLC

2441_C000.fm Page xx Thursday, July 20, 2006 9:07 AM

Chapter 7 Dehydration Technologies to Retain Bioactive Components ...............................173 Valérie Orsat and G. S. Vijaya Raghavan Chapter 8 Membrane Separation Technology in Processing Bioactive Components ...........193 Ashwani Kumar

PART III

Bioprocessing Technology

Chapter 9 Bioprocessing Technology for Production of Nutraceutical Compounds ............211 Terry H. Walker, Caye M. Drapcho, and Feng Chen Chapter 10 Microbial Modeling as Basis for Bioreactor Design for Nutraceutical Production .......................................................................................237 Caye M. Drapcho

PART IV

Preservation and Packaging Technologies

Chapter 11 Food Pasteurization and Sterilization with High Pressure ...................................269 Alberto Bertucco and Sara Spilimbergo Chapter 12 Microencapsulation and Delivery of Omega-3 Fatty Acids .................................297 Luz Sanguansri and Mary Ann Augustin Chapter 13 Packaging Technologies of Functional Foods.......................................................329 Louise Deschênes

PART V

Antioxidant Properties and Material

Chapter 14 Biological Antioxidation Mechanisms: Quenching of Peroxynitrite ...................341 Aditya Kulkarni and Hideo Etoh

© 2007 by Taylor & Francis Group, LLC

Chapter 15 Stability of Lycopene During Food Processing ....................................................353 John Shi, Yukio Kakuda, Yueming Jiang, and Gauri S. Mittal Chapter 16 Fruits with High Antioxidant Activity as Functional Foods ................................371 Shiow Y. Wang

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 3 Thursday, July 20, 2006 9:35 AM

1

Supercritical Fluid Technology for Extraction of Bioactive Components John Shi and Lamin S. Kassama Agriculture and Agri-Food Canada

Yukio Kakuda University of Guelph

CONTENTS 1.1 1.2

Introduction ......................................................................................................4 Process Concept Schemes and Systems ..........................................................6 1.2.1 Process Scheme and System................................................................7 1.2.1.1 Single-Stage Extraction Process...........................................7 1.2.1.2 Multistage Extraction Process ..............................................9 1.2.2 Physicochemical Properties of Supercritical (CO2) Fluids ...............10 1.2.2.1 Phase Diagram ....................................................................10 1.2.2.2 Physical Properties..............................................................10 1.3 Applications in the Food Industry .................................................................12 1.3.1 Extraction of Bioactive Compounds..................................................12 1.3.2 Fractionation of Flavors and Fragrances ...........................................18 1.3.3 Cholesterol-Free Food Products ........................................................20 1.3.4 Separation of Spices and Essential Oils............................................21 1.3.5 Decaffeination of Coffee and Tea......................................................23 1.3.6 Fish Oil Concentration.......................................................................26 1.4. Factors Affecting Extraction Yield ................................................................27 1.4.1 Pressure ..............................................................................................27 1.4.2 Temperature........................................................................................28 1.4.3 Moisture Content of Raw Materials ..................................................29 1.4.4 Cosolvent............................................................................................30

3

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 4 Thursday, July 20, 2006 9:35 AM

4

Functional Food Ingredients and Nutraceuticals

1.4.5 Particle Size........................................................................................32 1.4.6 Flow Rate ...........................................................................................32 1.4.7 Effect of Time on Yield .....................................................................34 1.5 Summary ........................................................................................................36 References................................................................................................................37

1.1 INTRODUCTION Some chemicals are toxic and if consumed can lead to cancer and other ailments of public concern. As a result, the conventional solvent extraction methods are viewed with suspicion with regards to their role in the manufacturing of functional foods. Extracts from natural sources are key elements in the manufacturing of functional foods. These are any food fortified with extracts from natural sources or nutraceutical. A new paradigm, supercritical extraction technology, was developed to extract bioactive components for use as supplements for functional foods. This technology uses supercritical fluids as the extracting solvent which has a significant attribute paramount to the food and pharmaceutical industry because it leaves no residue in the extract and is gaseous at ambient temperature. Extraction with organic solvents is a well-established technique for selective separation of specific constituents from plant products. Organic solvents with low boiling point such as ethyl acetate, methanol, dichloromethane, and so forth, have been successfully used to isolate bioactive components from hops, spices, oil seeds, and other plant products. The procedure is used to decaffeinate coffee and tea, and to remove nicotine from tobacco. The solvents used for this purpose must meet the legal requirements put in place to ensure food quality and safety, and these regulations vary from country to country. These requirements are: high degree of purity, chemical stability, inert (no reaction with food constituent), low boiling point, and no toxic effects. Criteria for these regulations are set by national and international bodies such as the US Food and Drug Administration (FDA); European Economic Commission (EEC) Codex committee; the Canadian Food Inspection Agency (CFIA); and FAO/WHO Codex Alimentarius Commission. Most countries have regulations stating which extraction solvents are generally regarded as safe (GRAS). For example, the Canadian food additive regulation (FDAR: B.10.045) for food chemical codex specification and the EEC directive (88/344/EEC) state the regulations regarding the use of extraction solvents in the production of foodstuffs and food ingredients within Canada and the EEC. Although only mandatory within Canada and the EEC, in general they are comparable to those in other countries. Public health, environmental, and safety issues are the major concerns in the use of organic solvents in food processing. The possibility of solvent residues remaining in the final product has been a growing concern to consumers, thus warranting stringent environmental regulations. The demand for ultrapure and highadded-value products is redirecting the focus of the food and pharmaceutical industries into seeking the development of new and clean technologies for their products. The supercritical fluid extraction (SFE) technology has provided an excellent alternative to the conventional organic solvent extraction methods. Although the

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 5 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

5

technology was known for more than 100 years, its application in the food and pharmaceutical industry began only three decades ago.1,2 Since that time, over 100 plants of different capacities have been built globally for extraction of desired solutes from solid materials.3 Supercritical extraction is a novel separation technique that utilizes the solvent properties of fluids near their thermodynamic critical points.4 A variety of processes involving extractions with supercritical fluids (SCF) have been developed and are regarded as a viable extraction technology that meets the food quality and safety requirements. The physicochemical properties of supercritical fluids, such as the density, diffusivity, viscosity, and dielectric constant can be controlled by varying the operating conditions of pressure and temperature or both in combination.1,3,5 Many supercritical fluids (carbon dioxide, ethane, propane, butane, pentane, ethylene, ammonia, sulfur dioxide, water, chlorodifluoromethane, etc.) are used in supercritical extraction processes. Sihvonen et al.,1 Brunner,3 and Rozzi and Singh6 recommended carbon dioxide (CO2) in their reports because of its favorable properties and the ease of changing selectivity by the addition of agents such as ethanol and other polar solvents. Supercritical carbon dioxide (SC-CO2) is the most desirable supercritical fluid for extracting natural products for foods and medicinal uses. Its characteristic traits are inert, nonflammable, noncorrosive, inexpensive, availability, odorless, tasteless, environmentally friendly, and GRAS status.7 Its near-ambient critical temperature makes it ideal for thermolabile natural products.8,9 Supercritical CO2 has the following advantageous attributes over other solvents: (a) it has a solvating power similar to organic liquid solvents and higher diffusivities, lower surface tension, and viscosity; (b) separation can be affected by simply changing the operating pressure or temperature to alter the solvating power of the solvent; and (c) modifying CO2 with a cosolvent can significantly augment the selective and separation power and in some cases extend its solvating powers to polar components. Supercritical carbon dioxide is being given a great deal more attention as an alternative to industrial solvents as a result of (a) increased governmental scrutiny and new regulations restricting the use of common industrial solvents such as chlorinated hydrocarbons; (b) its nontoxic and environmentally friendly attributes, given that it leaves no traces of solvent residue in food; (c) sharp increase in energy cost, which increased the cost of traditional energy-intensive separation technique, such as distillation; (d) carbon dioxide is cheap, safe to use, recyclable, and has minimum disposal cost required; (e) stringent pollution-control legislation prompting industries to seek alternative means of waste treatment and utilization; and (f) increased performance demands on materials, which traditional processing techniques cannot meet. It must, however, be stated that commercial applications of the SFE technology remain limited to a few high-value products due to high capital investment,10,11 its novelty, and complex operating system. Adoption of the technology is on the rise as a result of advances in processing, equipment, and the realization of producing high-value products with high profitability.1,10 Supercritical fluid extraction is one of the most promising technologies being adopted in the chemical, food, pharmaceutical, and neutraceutical industries. Analytical techniques are superior to the conventional organic solvents when applied for qualitative and quantitative extraction

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 6 Thursday, July 20, 2006 9:35 AM

6

Functional Food Ingredients and Nutraceuticals

of natural thermolabile products.12,13 Research efforts have proven that the SFE technologies could be successfully employed in the extraction and fractionation of fats and oils; purification of solid matrix; separation of tocopherols and antioxidants; removal of pesticide residue from herbs, medicines, and food products; detoxification of shellfish; and concentration of fermented broth, fruit juices, essential oils, spices, and coffee; and separation of caffeine, and so forth.3,14–16

1.2 PROCESS CONCEPT SCHEMES AND SYSTEMS The supercritical fluid extraction technology was conceptualized on the basis of obtaining pure extract without solvent residues which could be detrimental to consumers of food and pharmaceutical products. Extraction is an analytical process used to separate and isolate a targeted component from other substances. The success of the process is dependent on the distribution of the analyte between two phases, the separation and stationary phases.17 In the phase equilibrium between liquid and gas, the partition of the liquid phase increases with increasing pressure and decreases with increasing temperature. If the temperature and pressure are simultaneously increased, the transport properties of both liquid and gas increases and thus convergence occurs. When used under practical pressure and temperature conditions of 5 to 50 MPa and ambient to 300°C, respectively, the solubility properties of the supercritical fluid are greatly influenced by its density, diffusivity, and viscosity. The SC-CO2 is liquid-like and has a higher extraction flux than those obtained with organic liquids solvents. King et al.17 stated that at high CO2 densities, its solvent properties were similar to organic solvents like chloroform and acetone, and if intermediate compression were applied, it behaved like a nonpolar hydrocarbon such as n-pentane or diethyl ether. The separation phase occurs during the dynamic extraction period, while the stationary phase is the sample material loaded as a fixed-bed in the extraction column. Supercritical extraction (SCE) involves the use of compressed gases at or above their critical temperature (Tc) and pressure (Pc). It utilizes the ability of these special fluids to become excellent solvents to solvate certain solutes (bioactive components) from a solid matrix.6 The solute extraction stream from the sample matrix is directly proportional to the product of solubility and diffusivity in the supercritical medium. Hence the solute’s solubility increases with pressure, while its corresponding diffusivity is expected to decrease by two orders of magnitude. The solvent capacity is mainly the function of density and can be improved with the addition of a cosolvent, which modifies the density and polarity of the supercritical fluid, thus significantly increasing the yield. This technology has been successfully applied in the extraction of bioactive components (antioxidants, flavonoids, lycopene, essential oils, lectins, carotenoids, etc.) from a variety of biological materials such as hops, spices, tomato skins, and other raw or waste agricultural materials. The process requires intimate contact between the packed bed formed by a ground solid substratum (fixed-bed of extractable material) with a supercritical fluid.18 During the supercritical extraction process, the solid phase comprised of the solute and the insoluble residuum (matrix) is brought

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 7 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

Back pressure valve

Pressure valve

CO2 outlet Separator

Extractor column

Heat exchanger

Heat exchanger

CO2 Pump

Cooling bath

7

Extracts outlet

CO2 Source

FIGURE 1.1 Schematic diagram of a typical single-stage supercritical fluid extraction system with CO2.

into contact with the fluid phase which is the solution of the solute in the supercritical fluid (solvent). The extracted material is then conveyed to a separation unit. The power of supercritical fluid extraction is linked to the solubility and phase equilibrium of substances in the compressed gas. The targeted bioactive component being extracted must be soluble in the supercritical fluid. Controlling the pressure and temperature of SCF varies the solubility and phase equilibria. The extraction of pure and high-value extract is accomplished without risk of environmental pollution or residual solvent contamination in the final product.

1.2.1 PROCESS SCHEME

AND

SYSTEM

The extraction with supercritical fluids is comparable to liquid-liquid solvent extraction even though, with supercritical extraction, compressed gas is used instead of organic solvents and the applied pressure is crucial. The supercritical fluid extraction process is governed by four key steps: extraction, expansion, separation, and solvent conditioning. The steps are accompanied by four generic primary components: extractor column (high-pressure vessel), pressure control valves, separator column, and pressure intensifier (pump) (Figure 1.1) for the recyclable solvent.3,5,17,19,20 The system has other built-in accessories, such as heat exchangers for providing a source of heating; condensers for condensing supercritical fluids to liquid; storage vessels; and a supercritical fluid source. Raw materials are usually ground and charged into a temperature-controlled extractor column forming a fixed bed, which is usually the case for batch and single-stage mode. 1.2.1.1 Single-Stage Extraction Process The supercritical fluid is fed at high pressure by means of a pump, which pressurizes the extraction tank and also circulates the supercritical medium throughout the system. Figure 1.2 shows an example of a typical single-stage supercritical extraction system. Once the SCF and the feed reach equilibrium in the extraction vessel, through the manipulation of pressure and temperature to achieve the ideal operating conditions, the extraction process proceeds. The mobile phase consisting of the SCCO2 fluid and the solubilized components are transferred to the separator where the

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 8 Thursday, July 20, 2006 9:35 AM

Functional Food Ingredients and Nutraceuticals

Three stage separator columns

CO2 inlet

Extractor column

Pressure regulator

Dry test meter

Extracts

8

Ice water trap

CO2 pump

Co-solvent inlet

Mixer

CO-solvent pump

FIGURE 1.2 Schematic diagram of supercritical fluid extraction system used to fractionate bioactive components from plant matrix using supercritical carbon dioxide.

solvating power of the fluid is decreased by increasing the temperature or decreasing the pressure of the system. The extract precipitates in the separator while the supercritical fluid is either released or recycled back to the extractor. In the case where highly volatile components are being extracted, a multistage configuration may have to be employed as shown in Figure 1.2. As the solutions leave the extractor and flow to the first separation vessel via the pressure regulator, the pasty oleoresins settle to the bottom as they separate and can be collected, while the remaining solution goes to the second-stage separator where the fractionation of the volatile components occurs. For more sensitive products, the third stage of separation would be required for the complete isolation of pure volatile components. Saltzman et al.21 presented a design (Figure 1.2) where the solution flows through a heated valve and precipitates into a preweight U-tube in an ice-water bath. The glass wool at the U-tube exit traps the entrained solutes in the gas, while the gas flows through a dry-test meter which monitors the flow rate. Oszagyan et al.22 used a similar system as illustrated in Figure 1.2 to extract essential oil from Lavandula intermedia Emeric ex aloisel and herb of Thymus vulgaris L., and further fractionated volatile components (-Cymene, -Terpinene, Thymol, and Carvacrol) while Ozcan23 used it to fractionate volatile components from Turkish herbal tea (Salvia aucheri Bentham Var. canaescen Boiss and Heldr.). Duquesnoy et al.24 and Boutekedjiret et al.25 extracted and fractionated volatile compounds from plant materials using SCE with a similar multistage fractionation method. The processes described above are semibatch continuous, where the SCF flows in a continuous mode while the extractable solid feed is charged into the extraction vessel in batches. In a commercial processing plant, multiple extraction vessels are sequentially used to enhance process performance and output. Although the system is interrupted at the end of the extraction period, when the process is switched to a

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 9 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

9

2 Stage separation vessels

Storage tank

3 Stage extraction columns

Co-solvent inlet

CO2 inlet Pu mps

FIGURE 1.3 Schematic diagram of commercial scale multistage supercritical fluid extraction system used to fractionate bioactive components.

prepared vessel for extraction, the unloading and/or loading of the spent vessels can be carried out while extraction is in progress, reducing the downtime and improving the production efficiency. 1.2.1.2 Multistage Extraction Process A semicontinuous approach on a commercial scale uses a multiple-stage extraction process which involves running the system concurrently by harnessing a series of extraction vessels in tandem as shown in Figure 1.3. In this system the process is not interrupted at the end of extraction period for each vessel, because the process is switched to the next prepared vessel by control valves for extraction while unloading or loading the spent vessels; although imperfect, continuity is attained. The primary extraction stages operate in a similar mode to the ones depicted in Figure 1.2 and Figure 1.3. The raffinate from the premier stage enters the first separation vessel while separation and fractionation of different compounds occurs based on their relative solubility. The options of cosolvent are available to enhance the solvent power of separation of specific components. This is effective for cases where more than one targeted component is to be extracted, giving the flexibility to vary the extraction parameters such as pressure and temperature to achieve different solubilities for different components being extracted at each stage of operation. Gamse26 suggested that highly soluble substances could be extracted at the initial stages at low SCF density, and by increasing the density in the subsequent stages remove the less soluble substances. The supercritical pressure, temperature, and flow rate at each stage could be controlled independently.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 10 Thursday, July 20, 2006 9:35 AM

10

Functional Food Ingredients and Nutraceuticals

1.2.2 PHYSICOCHEMICAL PROPERTIES FLUIDS

OF

SUPERCRITICAL (CO2)

The physicochemical properties of the supercritical fluids are crucial to the understanding of the process design calculation and modeling of the extraction process. Therefore, selectivity of solvents to discriminate solutes is a key property for the process engineer. Physical characteristics such as density and interfacial tension are important for separation to proceed; the density of the extract phase must be different from that of the raffinate phase and the interfacial properties influence coalescence, a step that must occur if the extract and raffinate phase are to separate. 1.2.2.1 Phase Diagram Supercritical state of fluid is influenced by temperature and pressure above the critical point. The critical point is the end of the vapor-liquid coexistence curve as shown on the pressure-temperature curve in Figure 1.4 where a single gaseous phase is generated. When pressure and temperature are further increased beyond this critical point, the fluid enters a supercritical state. At this state no phase transition will occur regardless of any increase in pressure or temperature nor will it transit to a liquid phase. Hence, diffusion and mass transfer during supercritical extraction are about two orders of magnitude greater than in the liquid state. 1.2.2.2 Physical Properties Substances that have similar polarities will be soluble in each other, but increasing deviation in polarity will make solubility increasingly difficult. Intermolecular polarities exist as a result of van der Waals forces, and, although solubility behaviors

Pressure (MPa)

Supercritical region Liquid phase PC Solid phase

Triple point

Gas phase Tc = 31.1°C Pc = 7.38 MPa TC

Temperature (°C)

FIGURE 1.4 Supercritical pressure-temperature diagram for carbon dioxide.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 11 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

11

TABLE 1.1 Comparison of Physical Properties of Supercritical CO2 at 20 MPa and 55°C with Some Selected Liquid Solvent at 25°C Properties

CO2

n-Hexane

Methylene Chloride

Methanol

Density (g/mL) Kinematic viscosity (m2/s) Diffusivity (m2/s) Cohesive energy density (δ(cal/cm∏))

0.75 1.0 6.0 × 109 10.8

0.66 4.45 4.0 × 109 7.24

1.33 3.09 2.9 × 109 9.93

0.79 6.91 1.8 × 109 14.28

Source: Modified from King et al.17

depend on the degree of intermolecular attraction between molecules, the discrimination between different types of polarities is also important. Substances dissolve in each other if their intermolecular forces are similar, or if the composite forces are made up in the same way. Properties such as the density, diffusivity, dielectrical constant, viscosity, and solubility are paramount to supercritical extraction process design. The dissolving power of SCF depends on its density and the mass transfer characteristic, and is superior due to its high diffusivity, low viscosity, and interfacial tension to liquid solvents. Although many different types of supercritical fluids are in existence and have many industrial applications, CO2 is the most desired for SCE of bioactive components. Table 1.1 shows some physical properties of compressed (20 MPa) supercritical CO2 at 55°C compared to condensed liquids commonly used as extraction solvents at 25°C. It should be noticed that supercritical CO2 exhibited similar density as those of the liquid solvents, while less viscous and highly diffusive. This fluidlike attribute of CO2 coupled with its ideal transport properties and other quality attributes outlined above make it a better choice over other solvents. The specific heat capacity (Cp) of CO2 rapidly increases as the critical point (31.1°C temperature, 7.37 MPa pressure, and 467.7 g/L flow rate) is approached. Like enthalpy and entropy, the heat capacity is a function of temperature, pressure, and density.5 Under constant temperature both the enthalpy and entropy of supercritical CO2 decreases with increased pressure and increases with temperature at constant pressure. The change in the specific heat as a result of varying the pressure and temperature is also dependent on density. For example, under constant temperature, specific heat increases with increased density up to a certain critical level. Above this critical level, any further increase of density decreases the specific heat. Sample matrix is an important parameter that influences solubility and mass transfer process during SCE. Properties such as particle shape and size distribution, porosity and pore size distributions, surface area, and moisture content influence solubility and mass transfer. The presence of water (moisture content) in the sample matrix during supercritical extraction also has an effect on the extraction outcome.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 12 Thursday, July 20, 2006 9:35 AM

12

Functional Food Ingredients and Nutraceuticals

1.3 APPLICATIONS IN THE FOOD INDUSTRY One of the most important trends in the food industry today is the demand for allnatural food ingredients that are free of chemical additives. Natural food antioxidants are derivatives of plant by-products. A quantum leap in supercritical fluid extraction technology was made by its applications in decaffeinating coffee, tea, and other bioactive (essential oils from spices) components used as ingredients in foods. Likewise, SCF extraction is used to extract flavor and fragrance components, highvalue compounds used in the food, pharmaceuticals, and neutraceutical industries. With this technology, extracts of natural nutrients could be utilized without the fear of organic solvent residues. A compendium of process parameters used for different product applications is listed in Table 1.2. Various processes of extracting bioactive components from agricultural materials have been used commercially. The major hurdles to overcome in SCE are to obtain extracted compounds with similar bioactivity as the synthetic compounds and also to maintain the required flavor, odor, and color components that may be detectable in the treated food product. Many extraction processes employ the following methods to extract bioactive components from plants (Labiatae family): solvent (polar and nonpolar) extraction and aqueous alkaline58; extraction of vegetable oils and mono/diglycerides59 by steam distillation and molecular distillation.23 These processes are limited by numerous disadvantages, because the solvents used are not very selective for the bioactive components. Consequently, the resulting extracts are not as pure as synthetic chemical compounds. The solvents used include hydrocarbons such as hexane, acetone, and methyl chloride which leave unwanted residues in the food products, which are prohibited in some instances by regulatory bodies for use in food as discussed earlier. Many researchers applied SCE technology for the extraction of bioactive components.3,56,60–62

1.3.1 EXTRACTION

OF

BIOACTIVE COMPOUNDS

Some bioactive components are phenolic or polyphenolic compounds of plant origins which interfere with the formation of free radicals, thus preventing the formation of hydroperoxides. However, during food processing, especially conventional thermal processing of food products, these bioactive components are lost. To restore these components necessitates fortification. Some of the more common bioactive compounds include lycopene, flavonoids, tocopherol, lecithin, ascorbic acid, citric acid, polyphenols, and so forth. Most of the separation procedures involve physical and chemical processes such as centrifugation, filtration, membrane separation, precipitation, chromatography, solvent extraction, crystallization, evaporation, molecular distillation, and SC-CO2. To overcome the separation problems encountered when producing soluble materials via such processing procedures requires conditions that may have detrimental effects on the nature of the extraction technique or the product stability. In order to achieve extract labile compounds, a rapid separation process may be needed to avoid any significant product loss. Most bioactive components used as food additives are in a concentrated form. Appropriate extraction procedures are consequently required

© 2007 by Taylor & Francis Group, LLC

Product Extracted

Raw Material

Sunflower oil Soybean oil

Sunflower seed Soybean

Corm germ oil Seed oil

Corn germ Avocado

Raw Material Pretreatment

Component Concentration (%)

Temp (°C)

Pressure (MPa)

Ground

40−50 32–35

Flaked, 20.1 0.38–5.1mm Dried, milled 23.4 Dried, ground (< 0.25 mm)

50

7.58–8.27

50 70

55.2 75.8

50 50 70–75 50

Cottonseed Paprika Peanuts Peanut hearts Rice bran Sorghum bran Sorghum germ Soybean Wheat bran Wheat germ

© 2007 by Taylor & Francis Group, LLC

CO2 Flow Rate Time (L/h) (h)

MC (%)

2.5

Recovery (%)

Cosolvent

Source

Ethanol

36

Cocero & Calvo27 Friedich et al.28

Ethanol

50 58.2

Ronyai et al.29 Friedich & Pryde30

900–1080

9.8

1140

24

3.5 3.4

55.2 55.2 68.9 68.9

8 7 8 11

8.6 5 9.0 2.8

30.8 7.2 48.0 42.6

70 78

65.5 66.2

6 11

6.3 6.1

19.2 5.0

70

68.9

7

9.4

16.8

50 50 50

55.2 55.2 55.2

8 5 6

11.4 11.4 10.1

19.4 4.0 7.0

2441_C001.fm Page 13 Thursday, July 20, 2006 9:35 AM

TABLE 1.2 Supercritical Extraction Process Parameters for Some Selected Bioactive Components from Agricultural Material by SC-CO2

Product Extracted (Essential oil) Limonene Carvone Anethole Eugenol and Caryophyllene Yaleranone Wheat germ oil

Licochalcone A Licochalcone B Defatted mustard Canola oil

Bergaptene

Raw Material

Raw Material Pretreatment

Component Concentration (%)

Sweet Dried ground orange Sieved (1, 0.7, Caraway 0.4, 0.2, 0.08 seed mm) Anise seed Cloves Spikenard Wheat germ Milled, 10.2 powder (20 µm diameter) Licorile Cut root Mustard Dried, fully 16 seed pressed (1mm) Canola seed Flaking (0.2–0.5 mm) cooking (90°C) Bergamot Peel dried 0.15 Citrus ground

© 2007 by Taylor & Francis Group, LLC

Temp (°C)

Pressure (MPa)

CO2 Flow Rate Time (L/h) (h)

23–40 9–10

60

35–50 13–41

0.12

2

35

12.7

360

0.5

40

30.4

300–400

3

45–70 41–62

180

3

40–60 8–10

2.6

2

MC (%)

Recovery (%)

Cosolvent

Source

6.7

60

Sovava et al.31

4.3–11.5

98.7

Ge et al.32

5% Ethanol 25–30

Ethanol

17–19

80

Taniguchi et al.33

44

Dunford and Temelli34

85

Poiana et al.35

2441_C001.fm Page 14 Thursday, July 20, 2006 9:35 AM

TABLE 1.2 (CONTINUED) Supercritical Extraction Process Parameters for Some Selected Bioactive Components from Agricultural Material by SC-CO2

Defatted & decholesterol of muscle Oil Bixin pigment Defatted & decholesterol beef Bixin pigment

Evening primrose oil Cardamom oil

Alfalfa

Leaf protein concentration Dehydrated Spray dried beef powder + air dried chunk Antarctic Freeze-dried 16.2 krill meal & ground Annatto Seed, 3.92 seed mm Fresh 19 frozen Ground beef Annatto Whole seed 1.3 seed (11.6 oil)

40

Evening primrose seed Cardamom seeds

Decholesterol beef tallow β-carotene

Crude beef tallow Sweet potato

Lanolin

Woolgrease

Dried, ground, 0.355 mm Freshly ground seeds, mesh

Freeze dried ground ( 0.25 mm)

© 2007 by Taylor & Francis Group, LLC

30– 70

300–360

90 70

Favati et al.36

1.4–3.2

87

Wehling et al.37

7.83

99

Yamaguchi et al.38 Chao et al.39

56-lipid 26-cholesterol

Chao et al.40

3

45–55 23–3927.6 300–600

80

24.5

51.4

3–4

50 60

34.5 31

1–2

35

31.

117.2 – 144.8 117.2 – 144.8

50

29.6

1.2

21.9

60

70

64.8

40–60

40

10

0.15–0.2

40

34.5

226.7250

94

48

41.4

840–1080

80

38

396–2880

4–8

Degnan et al.41

Soybean oil 20% v 1.7–5 < 8

97

Favati et al.42

3

85–95

Gopalakrishnan & Narayanan43

60–70

Chao et al.44

80

Spanos et al.45

4–6

15.5

10

CygnarowiczProvost et al.46

2441_C001.fm Page 15 Thursday, July 20, 2006 9:35 AM

Carotene lutein

Product Extracted

Raw Material

Phospholipid mixture

Soybean flakes

Tocopherol enrichment

Soybean flakes Rice bran Carrot

Carotene Essential oil

Star anise fruit

Protein

Spring mackerel Black pepper Eucalyptus leaves Ginger rhizomes Palm oil fiber

Essential oil Piperine Essential oil Essential oils gingeroles Triglycerides carotenoid

Raw Material Pretreatment

Component Concentration (%)

Ground, 5.134 mg/g thinly flakes, (0.1–0.25 mm)

Frozen puree (0.93mm) Dried, ground, particle size (0.57 mm) Freeze-dried, chopped Dried ground

Oven dried

© 2007 by Taylor & Francis Group, LLC

Pressure (MPa)

CO2 Flow Rate Time (L/h) (h)

MC (%)

Source

68.2

43.6–109

80

25

300

500 ppm

55

20.7

90

8–9

26.5

6.5

5.4

1.5

12.7

97

35

34.5

1080

15

1.6

87

40

9–15

2.55

9

18

Perakis et al.49

50

9

1.1

2.5

9.5

2.4

Porta et al.50

138

2

8.8

8.4

Catchpole et al.51

1.45

2.25

5

7

Franca and Meirele52

1.5 5.7

30

40 45– 55

30 20–30

10.8

Recovery (%)

Cosolvent

80

Air drying Dried ground

Temp (°C)

81

Ethanol (5% v)

Ethanol 10%

60-Soybean King et al.47 70-Rice bran 80-Wheat germ Vega et al.48

2441_C001.fm Page 16 Thursday, July 20, 2006 9:35 AM

TABLE 1.2 (CONTINUED) Supercritical Extraction Process Parameters for Some Selected Bioactive Components from Agricultural Material by SC-CO2

Essential oil vanillin Essential oil Capsaicine alkaloids Essential oil Antioxidants

Essential oil

Tomatoes Peppermint & Spearmint Vanilla pod Chili pepper

Dried skin ground Cut leaves

65 35

Dried ground

36

36

11

Dried ground

2

40

30

50

15

Coriander

Dried grounds 96 (0.4 mm) Sweet Thai Dried ground Tamarinds (mesh 40–70) Grape seed Wash dry 80 ground

© 2007 by Taylor & Francis Group, LLC

40–80 17–28

30

0.5

24–43 6–18

3.42

4–9

35–80 10–30

35–75 20–47

Cadoni et al.53

Chloroform 14–82

76

Barton et al.54

97 2

3

3

0.3

5

0.75

Catchpole et al.51

8.8

0.61

10

Anitescu et al.55

Ethanol

Luengthanaphol et al.56

2% Ethanol 77

Lee et al.57

2441_C001.fm Page 17 Thursday, July 20, 2006 9:35 AM

Lycopene -Carotene Essential oil Menthol

2441_C001.fm Page 18 Thursday, July 20, 2006 9:35 AM

18

Functional Food Ingredients and Nutraceuticals

when preparing them from their original matrices. Some compounds in the concentrated form are thermolabile, volatile, and prone to degradation when subject to intensive heat. Supercritical extraction with CO2 is the most viable method for food applications. Baysal et al.63 extracted lycopene and -carotene from tomato using SC-CO2. The processing conditions used were extraction pressure of 20, 25, and 30 MPa; temperatures of 35°C, 45°C, 55°C, and 65°C; resident time of 1, 2, and 3 h; and CO2 flow rate of 2, 4, and 8 kg/h. The best conditions for lycopene extraction were 2 h at a flow rate of 4 kg/h, pressure of 30 MPa, temperature of 55°C, and with the addition of 5% cosolvent (ethanol). They noted that if too much ethanol is used, it decreased the homogeneity of the extraction mixture, and reduced the separation efficiency. Brunner and Peter64 and Vega et al.48 corroborated this finding. Luengthanaphol et al.56 compared the SC-CO2 extraction to other methods and their effects on bioactivity. Their results compared to Tsuda et al.65 indicated the superiority of SC-CO2 coupled with cosolvent as shown in Table 1.3. Both studies show the superiority of antioxidant extraction with SC-CO2 and modifiers, although some disparity occurred which could have been caused by the variety used. Macias-Sanchez et al.61 extracted carotenoids and chlorophyll from Nannochloropsis gaditana and achieved the highest yield at 20 MPa and 60°C, the optimal pressure and temperature, respectively. Wang et al.66 also reported that the antioxidant activity of Bupleurum kaoi Liu fractionated with SC-CO2 gave the highest yield of phenol and the strongest antioxidant capacities.

1.3.2 FRACTIONATION

OF

FLAVORS

AND

FRAGRANCES

The extraction of compounds, for flavor and fragrances, by supercritical CO2 is paramount in the food industry. Mother Nature is a splendid synthesizer of flavors and fragrances in natural products. The cleaner and safer attributes make supercritical

TABLE 1.3 Yields and Activities of Antioxidants Extracted by Different Methods from Sweet Thai Tamarind Seed. SC-CO2 Extraction at 30 MPa and 80°C; Organic Solvent Extraction at Room Temperature PV after 24 h (meq/kg dry weight)

(-) Epicatechin Yield (mg/100 g dry weight ) Extraction SC-CO2 SC-CO2 + Cosolvent (10% Ethanol) Ethanol α-tocopherol

Luengthanaphol et al.56 Tsuda et al.65

Luengthanaphol et al.56

Tsuda et al.65

0.022 13

0.336 26

– 231

– ≈26

25

32

– 157

≈25

56

65

Source: Modified from Luengthanaphol et al. and Tsuda et al.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 19 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

19

TABLE 1.4 The Results of Gas Chromatography and Mass Spectrophotometer Analysis of Flavor Compounds Extracted by the Likens-Nickerson Extraction and the Supercritical CO2 Extraction Methods of Basmati Rice Likens-Nickerson Extraction Peak No.

Retention Time (min)

2 3 5 8 13 14 19 21 22 24

7.51 8.11 9.51 14.52 21.98 22.68 27.71 28.53 30.45 32.14

Hexamethyl disiloxane* Silicate anion tetramer* Dibutyl phthalate* Hexane cyclotrisiloxane* Pentanal Butan-2-one-3-Me Heptanol 2-Heptanone Octanol Octanal

28 30 31 34 35 37

35.14 36.12 37.34 40.22 42.77 43.82

2-Octanone Nonanal Decanol 2-Decenal Undecane Benzoic acid 2,5-bis (trimethylsiloxy) benzene*

39

46.28

Compound

SC-CO2 Peak No.

Retention Time (min)

Compound

12 13 19 20 21 23 29 30 31 32 34 37 40

22.11 22.56 28.32 29.27 30.92 32.26 35.19 35.44 36.13 37.42 40.31 42.52 45.11

Pentanal Butan-2-one-3-Me Heptanol 2-Heptanone Octanol Octanal 2-Octenal 2-Octanone Nonanal Decanol 2-Decenal Undecane Dodecane

44

45.24

Tetradecane

Compounds identified as artifacts. Source: Modified from Bhattacharjee et al.62

technology an ideal candidate for extracting such valuable and heat-sensitive products when compared to toxic organic solvents. The high-value-added natural products are good for use in soft drinks. An example is ginger extract which gives the pungency and flavor in ginger ale drinks. Bhattacharjee et al.62 compared the Likens-Nickerson extraction and the SCCO2 extraction methods on Basmati rice. The results showed that the SCE technique was superior and extracted the most flavor components that bore the closest resemblance to the original Basmati flavor (Table 1.4). The SC-CO2 technique extracted more flavor compounds and produced purer (resembles original basmati flavor) extracts than its counterpart (Liken-Nickerson method). Desired fragrances are isolated from concentrates extracted from flowers. The process consists of initial solvent extraction, usually with hexane, which yields an intermediate product called concrete. This product contains fragrances and other components like paraffin, fatty acids, fatty acids methyl ester, di-, and tri-terpenic

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 20 Thursday, July 20, 2006 9:35 AM

20

Functional Food Ingredients and Nutraceuticals

compounds, pigments, and so forth. Postprocessing of the concrete can be done using SC-CO2 extraction. Reverchon67 used single-step SCE at a pressure of 8 MPa and temperature of 40°C followed by a two-stage separation procedure at a pressure of 8 MPa and temperature of –16°C at the first stage and a pressure of 1.5 MPa and a temperature of 0°C at the second stage. Under these optimum conditions the extracted volatile rose oil contains 50% 2-phenylethanol. When a cosolvent (ethanol) is mixed with SC-CO2, a yield of 50% to 60% was observed.68 Jasmine fragrance extracted at 12 MPa and 40°C gave results superior to those of other solvents. Sastry and Mukhopadhyay68 experienced increased yield of 45% to 53% with the use of cosolvents. Similarly, SC-CO2 has been used effectively to extract fragrances from orange, marigold, sandalwood, vetiver, and so forth.

1.3.3 CHOLESTEROL- FREE FOOD PRODUCTS Cholesterol is an unavoidable substance we require for the daily maintenance of our body. Lipoproteins are vehicles that transport cholesterol to various body tissues to be used, stored, or excreted, but if cholesterol is in excess it can lead to coronary heart problems (atherosclerosis). Low-density lipoprotein (LDL) termed “bad” cholesterol causes fat buildup in the arteries increasing the risk of heart disease. Highdensity lipoprotein (HDL) termed “good” cholesterol transports cholesterol back to the liver, where endogenous metabolism prevents cholesterol buildup and reduces the risk of heart disease. The indiscriminate consumption of saturated fats in our diet may raise the total LDL (> 100 mg/dL) level and decrease HDL (< 35 mg/dL) level, thus increasing the risk of heart disease.69 The recommended daily intake of cholesterol is about 300 mg.70 The correlation between serum cholesterol level and mortality rate of cardiovascular disease has been reported in many studies.69,71 Pork has cholesterol content of 30 to 450 mg/100 g, poultry 70 mg, fish 35 to 70 mg/100 g, and beef 65 to 331mg. One common source of cholesterol is from the consumption of fried fast-food products. The fast-food industry uses hydrogenated fats for their deep-fat frying processes because of its stability and high economic turnover. The hydrogenated fat is the potential source of trans fatty acids which are taken up by the fried foods during cooking (French fries, onion rings, chicken nuggets, etc.) and ultimately ingested by the consumer. Trans fats have been shown to increase LDL cholesterol levels and reduce HDL cholesterol levels, thus raising the risk of heart disease. Public health initiatives such as the National Cholesterol Education programs have raised consumer awareness, resulting in the advocation for healthy foods with low cholesterol. Thus, the food industry is under tremendous pressure to address this consumer concern. Supercritical fluid extraction is an emerging technology with great potential to revolutionize the oil/fat industry. Many researchers (Dunford and Temelli34; Temelli et al.34,72) reported the feasibility of supercritical fluid extraction of lipids from food without compromising their organoleptic quality. Chao et al.44 used a similar extractor configuration (Figure 1.3) as discussed earlier with three stage separations, to remove cholesterol. By decreasing the pressure at each stage sequentially from 17 to 11 to 4 MPa, they were able to achieve higher selectivity for cholesterol at the lower pressures. The results also showed that the fraction collected from the third

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 21 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

21

TABLE 1.5 Solubility of Cholesterol in Supercritical CO2 under Different Operating Conditions Run No.

Pressure (MPa)

Temperature (°C)

Density of CO2 (g/L)

Solubility of Cholesterol (mg/L)

1 2 3 4 5 6 7 8

10 10 10 10 13 13 13 13

40 45 50 55 40 45 50 55

602.4 494.4 408.1 342.3 723.4 655.8 588.5 527.7

146.9 82.6 47.2 28.5 289.2 234.8 182.7 141.1

Source: Modified from Yeh et al.74

separator at 4 MPa contained concentrated cholesterol ranging from 272 to 433 mg/100 g lipid. Chao et al.40 used SC-CO2 at an operating pressure of 10 to 30 MPa and temperature range of 30°C to 50°C to reduce the cholesterol level in ground beef. Similarly, Hardardottir and Kinsella73 explored the removal of lipids and cholesterol from fish muscle with SC-CO2. They removed over 80% to 99% cholesterol using pressures of 14 to 35 MPa and temperatures of 40°C to 50°C. Although the authors noted limited effect on lipid/cholesterol yield with increased extraction pressure and temperature, increased extraction time from 3 to 9 h significantly increased the yield. Yeh et al.74 used eight operating conditions shown in Table 1.5 to optimize their process and observed that at 10.3 MPa and 55°C operating pressure and temperature, respectively, cholesterol level was reduced from 2867 mg/100 g to 14.1 mg/100 g. Supercritical-CO2 technology was used to fractionate milk fat, which is an excellent raw material with specific functionalities used in many products.75 Extracting cholesterol from anhydrous milk fat with SC-CO2 used in conjunction with adsorbents (silica gel) to maximize yield was demonstrated by Huber et al.76

1.3.4 SEPARATION

OF

SPICES

AND

ESSENTIAL OILS

Spices have strongly flavored or aromatic components that can be used in small quantities in food as a preservative or flavoring ingredient. Chili (capsicium species), ginger (Zingiber officinalis), and pepper (piper nigrum L) are classic pungent flavorines while ginger and chili have additional nutraceutical values. These products have high economic value in their concentrated form. The extraction of spices is usually carried out in two stages: stage one separates pungent oleoresins and the second stage the essential oil fractions. Essential oils are typically volatile terpenes and esters. Essential oils are concentrated pure plant extracts that have long been revered for their therapeutic applications and are derivatives from flowers, leaves,

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 22 Thursday, July 20, 2006 9:35 AM

22

Functional Food Ingredients and Nutraceuticals

stems, berries, rinds, resins, or roots of plants. These are very important ingredients and food additives of high value. Catchpole et al.51 performed a detailed study on the extraction of spices and essential oils using SC-CO2, propane, and dimethyl ether. They reported ginger to be the easiest of all spices in terms of optimized yield relative to pressure and temperature, while capsaicin in chili could be extracted at moderate pressure and temperature, especially with the use of modifiers. Chili oil fraction contains fatty oil and carotenoids, and is speculated that the fatty oil acts as modifier for the capsaicins.77 The author extracted black pepper with much duress, because of its viscous characteristics, thus requiring higher pressure and moderate to high temperatures. The use of supercritical propane for extracting spices was reported by Illes et al.78 They found propane to adequately extract fatty oils, tocopherols, and carotenoids, but inadequate for capsaicins, while CO2 was adequate for capsaicinoids, fatty oils, and tocopherols, but not for carotenoids. Figure 1.5 shows the supercritical extraction of ginger with three extraction fluids (CO2, propane, and dimethyl ether).51 Propane gave the lowest yields while dimethyl ether gave the highest yield. They reported dimethyl ether to have mutual solubility with water. Ginger contains high amounts of volatiles, and CO2 extraction offers the advantage of dividing the extract into oleoresins and essential oil fractions by using a two-stage separation procedure with sequential pressure reduction. Similarly, if propane or dimethyl ether is used, considerable heating is required which ultimately results in thermal degradation, and a larger energy requirement in the form of cooling, depressurization, and boiling to recover the essential oils. In the case of ginger, the oxygenated fraction was much greater than the steam distilled oils and the gingerols in the oleoresin were extracted without decomposition. Oleoresins and piperine from pepper were extracted with insignificant losses although a longer processing time was required. Similar trends were observed for chili and pepper (Figure 1.6 and Figure 1.7). The extracts contained carotenoid pigments, and those obtained with SC-CO2 were bright red with pink residues, while those from propane and dimethyl ether were dark red. The extract obtained from chili with SCCO2 corroborated the results of other researchers5,79 in being viscous, pastry yellow, and semisolid, while those extracted with dimethyl ether were yellow/black and liquid at room temperature with a high quantity of water resulting in the dilution of the essential oil and piperine content. Nguyen et al.80 described the extraction of antioxidants from Labiatae herbs (rosemary, sage, oregano, and thyme) with SCCO2 at pressures in the vicinity of 50 MPa and temperatures ranging from 80°C to 100°C. The extracted oleoresin was precipitated into two fractions at various pressures and temperatures. The first fraction consisted of a green-brown, oil-soluble, heat-stable, resin containing less than 2% essential oil and exhibiting remarkable antioxidant properties. The second fraction was the essential oil containing more than 95 mL steam distilled oil/100 g. The use of SC-CO2 for the production of essential oils or oleoresins from spices is possible by selecting suitable pressures and temperatures. The oils extracted with supercritical technology were found to be superior in terms of their chemical composition and higher percentage of sesquiterpene compounds. Supercritical extraction with CO2 and hydrodistillation extraction methods were used to extract essential oil

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 23 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

23

160

Yield (g/kg)

120 80 40 0 0

5

10 15 Solvent (kg/kg ginger )

CO2, 313 K

Propane, 323 K

20

25

Dimethylether, 313 K

FIGURE 1.5 Supercritical extraction yield of ginger using CO2, propane, and dimethyl ether (modified from Catchpole et al.51).

160

Yield (g/kg)

120 80 40 0 10

5

0

15

20

25

Solvent (kg/kg chili) CO2, 313 K

Propane, 323 K

Dimethylether, 313 K

FIGURE 1.6 Supercritical extraction yield of chili using CO2, propane, and dimethyl ether (modified from Catchpole et al.51).

(Juniperus communis L.)9 (Table 1.6). Oils obtained by SC-CO2 and hydrodistillation showed significant differences; the former was more selective and particularly efficient for the isolation of -thoujone and limonene. Anitescu et al.55 did a comparative analysis of coriander oil with supercritical CO2 and stream distillation (Table 1.7). They concluded that oils obtained by supercritical extraction gave a superior aroma compared to both the commercial and hydrodistillation extracted oils.

1.3.5 DECAFFEINATION

OF

COFFEE

AND

TEA

Caffeine (1,3,7-trimethylxanthine) is a bioactive plant component commonly found (since 1820s) in popular beverages such as teas (Camellia sinensis), coffees (Coffee Arabica, canephora, liberica), and soft drinks.81 Caffeine is a secondary metabolite, the product of nucleic acid catabolism, and belongs to the group of compounds

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 24 Thursday, July 20, 2006 9:35 AM

24

Functional Food Ingredients and Nutraceuticals

200

Yiled (g/kg)

150 100 50 0 0

5

10

CO2, 313 K

15 20 25 Solvent (kg/kg pepper) Propane, 323 K

30

35

Dimethylether, 313 K

FIGURE 1.7 Supercritical extraction yield of pepper using CO2, propane, and dimethyl ether (modified from Catchpole et al.51).

TABLE 1.6 Essential Oil (Juniperus communis) Oil Extracted by Comparing Supercritical Extraction and Hydrodistillation (HD) Methods Pressure Temperature Dynamic time Modifier No. Compound 1 α-Thoujone 2 α-Pinene 3 Sabinene 4 Myrcene 5 3-Carene 6 Limonene 7 Terpinolene

(MPa) (°C) (min) (µL) RI* 928 943 972 987 1007 1020 1071

20 45 20 – 1 25.1 1.3 2.3 2.8 35.0 22.0 2.6

20 45 30 – 2 26.2 1.4 2.1 3.8 36.8 24.2 –

20 55 30 – 3 26.9 1.4 2.3 3.8 37.1 22.9 2.6

35 45 30 – 4 17.0 1.8 2.1 1.3 19.9 9.0 0.8

35 55 30 – 5 13.0 – 1.8 1.2 17.1 8.6 2.1

35. 55 30 80 6 22.0 1.9 1.9 2.4 29.1 15.6 2.7

35 55 30 400 7 4.0 22.4 34.6 4.2 2.6 6.1 2.7

HD

– 24.5 0.4 3.4 39.4 – 3.1

* Retention Index; Source: Modified from Pourmortazavi et al.9

known as purine alkaloids. Excessive ingestion of caffeine may cause certain health problems such as palpitations, gastrointestinal disturbance, anxiety, tremor, increased blood pressure, dizziness, and insomnia.82–85 The aroma and flavor coupled with the stimulant effects comes from caffeine. Coffee beans have about 2% to 3% while tea leaves have about 5% caffeine,82 depending on the variety and species. Decaffeinated coffee must contain less than 0.1% caffeine on a dry weight basis, as specified by EEC regulations. Therefore, decaffeination of coffee and tea poses significant challenges to both the producers and processors. The demand for decaffeinated coffee is high on the world market. It accounts for more than 20% of all coffee sales in the USA, with a 50% growing demand among the adult population.82 Research in

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 25 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

25

TABLE 1.7 Results of Comparative Analysis of Essential (Coriander) Oil between Supercritical CO2 and Steam Distillation Processes

No.

Compounds

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

α-Thujene α-Pinene Camphene Sabinene β-Pinene Ββ-Myrcene ∆3-Carene Limonene 1,8-Cineole Linalol Camphor Menthol ρ-Cymen-8-ol Cis-Hex-3-enyl butyrate α-Terpineol β-Citronellol Neral Carvone Anethole Carvacrol Neryl acetate Geranyl acetate β-Caryophylene α-Humulene Eugenyl acetate β-Caryophylene Unidentified Compounds

Kovats RI

Comm Oil

Steam Distillation Oil

SC-CO2 Oil

928 936 951 975 980 990 1006 1030 1033 1103 1147 1174 1184 1186 1192 1226 1241 1245 1287 1299 1363 1382 1428 1463 1526 1594

Trace (Tr) 3.3 0.6 0.1 1.0 1.2 1.1 2.4 Tr 63.8 5.5 Tr Tr Tr 1.0 0.1 0.1 0.5 0.7 Tr 0.1 1.0 Tr Tr – Tr 3.8

Tr 2.3 0.4 0.3 0.3 0.8 0.3 2.3 0.1 62.8 5.6 0.1 0.1 0.1 0.9 0.3 0.1 1.0 0.4 0.1 0.1 1.8 2.1 0.3 Tr Tr 2.9

0.1 2.8 1.5 0.9 0.9 1.0 0.3 2.7 0.1 61.9 5.6 0.1 0.3 0.2 0.6 0.2 0.2 1.0 0.4 0.2 0.2 2.4 0.8 0.2 0.2 0.2 1.7

Source: From Anitescu et al.55

genetics engineering to produce transgenic tea and coffee plants deficient in caffeine is in progress.82,83,86 However, the consumption of genetically modified products is still contentious globally, and SCE technology gives the best option in combating these critical issues. The decaffeination of coffee and tea using SCE technology is among the first known commercial operation in the food industry. In the past, methylene chloride was used for decaffeination of coffee with one cycle of production lasting from 24 to 36 h and the end products usually contained toxic residues, thus posing more harm than the caffeine. Due to its suspected carcinogenic effect, the FDA placed regulations against methylene chloride use. However, decaffeination process with SC-CO2 fluid can be accomplished on green

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 26 Thursday, July 20, 2006 9:35 AM

26

Functional Food Ingredients and Nutraceuticals

coffee or roasted coffee beans or tea leaves without deleterious effects on the flavor even after 10 h of processing, and many patents exist for such processes. The process requires charging the extraction vessel containing the coffee beans with CO2 at a pressure of 7 to 22 MPa and temperature of 31°C. The caffeine is dissolved in the SC-CO2 stream, which subsequently enters a washing tower or alternatively activated carbon scrubbers, distillation, recrystallization, or reverse osmosis are used in some instances to entrain caffeine. The method can strip coffee of its caffeine content (0.7% to 3%) by 71% to 97%.87 The caffeine recovered is sold for medicinal purposes and or for use in soft drinks. Peker et al.88 reported that soaking raw coffee beans in water prior to processing enhances the rate of decaffeination.

1.3.6 FISH OIL CONCENTRATION Fish oils are characterized by a high percentage of unsaturated straight-chain fatty acids ranging from C14 to C22 with one to six double bonds. They contain essential fatty acids (EFA) and polyunsaturated acids, grouped into omega-6 and -3 EFAs. The main sources of omega-3 (-3) are flaxseed, walnut, and marine plankton and fish. This review would focus on -3 oils derived from fish. Eicosapentaenoic acid (EPA) and decosahexaenoic acid (DHA) are predominant in fish oil, and have been reported to contribute to the prevention of atherosclerosis, heart attack, depression, and cancer if consumed in sufficient quantities.89 Fish oil derivatives in the form of -3 oils are in high demands as food additives. For example, Ocean Nutrition’s ME3TM Omega-3 powder is currently used in several breads. In the US, Wegman’s Food Markets, Rochester, NY, launched breads fortified with MEG omega-3 fats, two slices of which offers 80 to 90 mg of omega-3s.90 Encapsulated omega-3 fatty acids forms are available for fortified bakery products. Fish oils are processed as fatty acids or as methyl or ethyl esters which are more stable than the free acids form.91 Fatty acids are highly soluble in CO2 and as a result SC-CO2 extraction is a preferred method of fractionation. With this technology, it is possible to separate heat-sensitive compounds (ω-3 fatty acids) and avoid toxic solvent residues in the final product. The isolation and fractionation of ω-3 PUFA (polyunsaturated fatty acid) from fish, fish oil, and esters using SC-CO2 have been studied by several researchers.72,92,93 Eisenbach93 fractionated the ethyl esters from cold fish oil using SC-CO2 at a pressure of 15 MPa and an extracting temperature of 50°C. Alkio et al.94 produced EPA and DHA with 50% and 90% purity, respectively, from trans-esterified tuna oil using carbon dioxide. Temelli et al.72 obtained the highest yield of -3 fatty acids at 35 MPa and 35°C without denaturing the protein during SCE. They also compared solvent (hexane) extraction to SC-CO2 as shown in Table 1.8. A higher concentration of -3 was achieved with SC-CO2 (Table 1.8). At 25 MPa pressure and temperature from 40°C to 80°C, no significant effect on yield was observed in oil extraction from krill.38 Hardardotti and Kinsella73 did not see any yield change on the recovery of rainbow trout at operating pressure ranging from 13 to 35 MPa and temperature range of 40°C to 50°C.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 27 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

27

TABLE 1.8 Fatty Acid Composition of Fall Atlantic Mackerel Oil Concentrate Extracted with Hexane and SC-CO2 at Pressure 34.5 MPa and Temperature of 35°C Fractionated by Gas Chromatograph Fatty Acid

GC Retention Time (min)

Hexane Extract

SC-CO2 Extract

C16:0 C16:1 C18:0 C18:1 C18:3 C20:0 C20:1 C20:5 (EPA,ω-3) C22:5 C20:5 (DHA,ω-3) EPA+DHA (ω-3)

8.07 8.53 12.49 13.09 15.58 18.05 18.14 22.79 30.08 31.43

15.86 6.65 3.53 3.59 0.88 11.22 1.20 6.07 1.34 8.58 14.65

16.66 7.94 3.16 3.16 1.04 7.84 1.20 8.76 1.38 8.97 17.73

Source: Modified from Tamelli et al.72

1.4. FACTORS AFFECTING EXTRACTION YIELD Several bioactive components were extracted successfully by the SC-CO2 extraction method as outlined in the preceding sections. Optimization of yield is a function of various independent parameters. Process parameters such as solvent flow rate, resident time, moisture content, particle sizes, and particle size distribution in conjunction with supercritical pressures and temperatures are key parameters for achieving optimum results. Most of these parameters can have individual or combined effects on the extraction rate of a process; for example, the resident time can have an immense influence on the composition of the extracted compound.

1.4.1 PRESSURE Figure 1.8 is a typical extraction rate curve, and it is apparent that pressure significantly influences the rate of extraction, and likewise the extraction time. Extrapolating the normalized yield at the point where the yield curve become asymptotic gives significant different normalized yields of 15, 11, and 4% for pressures of 10, 9, and 8 MPa, respectively.8 Macias-Sanchez et al.61 observed similar trends in the SC-CO2 extraction of carotenoids and chlorophyll a from Nannochloropsis gaditana, although as pressure increased beyond a critical point the yields dropped as a result of increased density. Higher density causes a double effect: increases the solvation power and reduces the interaction between the fluids and matrix, thus decreasing the diffusion coefficient. Excessive pressure also increases the compactness of the sample matrix, thus reducing the pore sizes and apparently reducing the mass transport, which eventually diminishes the yield.95

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 28 Thursday, July 20, 2006 9:35 AM

28

Functional Food Ingredients and Nutraceuticals

16

Yield (%)

12 8 4 0 0

100

200 8 MPa

300 400 Time (min) 9 MPa

500

600

700

10 MPa

FIGURE 1.8 The effect of pressure change on bioactive compound yield during supercritical extraction of essential oil from Juniperus oxycedrus on extraction rate at flow rate 1.5 kg/h and temperature 50°C (modified from Marongiu et al.8).

The selectivity of solute extraction is a function of extraction pressure, and if it increases, different solutes are extracted. D’Andrea et al.96 also found that with a working pressure of 25 MPa and temperature of 55°C, they attained optimum yield for the extraction of azadirachtin and 3-tigloylazadirachtol from neem seeds. Similarly, Tonthubthimthong et al.95 reported optimum yield at a pressure of 23 MPa and a temperature of 55°C.

1.4.2 TEMPERATURE Temperature is a parameter with significant influence on SFE; therefore manipulating it could have adverse implication on the process and yield. Figure 1.9 shows the general trend of increased extraction yield as temperature increases relative to the pressure. Tonthubthimthong et al.95 reported similar trends for extracting nimbin from neem seeds at 20 MPa and a CO2 flow rate of 0.62 mL/min, and found 35°C the optimum temperature for their process. Ge et al.32 indicated a temperature of 35°C gave the highest yield in the first 45 min of extraction, but during prolonged extraction from 45 to 120 min, the highest temperature condition was shown to produce the highest yield (Figure 1.9). Although many literature reports correlated increased temperature to yield,45 many others showed no particular trends as far as temperature was concerned.97 Some researchers reported yield was inversely proportional to temperature under 15 MPa. The combined effect of pressure and temperature on cholesterol extraction was studied by Chao et al.44 At a pressure-temperature setting of 34 MPa and 50°C, cholesterol yield of 160 mg/100 g was realized compared to 430 mg/100 g when the temperature drops to 40°C and 2.5 kg of CO2 was used. As the mass of CO2 increased, the yields decreased, but the lowest temperature still maintains the highest yield as shown in Figure 1.10. Also, under constant temperature, increases were achieved when pressure decreased. The results demonstrated that higher selectivity is possible at lower pressures and higher temperatures. Froning et al.98 corroborated this fact based on their experiment with lipid and cholesterol extraction from

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 29 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

29

Yield (mg/100 g)

2600

2200

1800

1400 10

15

20

25 Pressure (MPa)

35°C

40°C

30

45°C

35

40

50°C

FIGURE 1.9 The effect of temperature change on bioactive compound yield during supercritical extraction of wheat germ on extraction, time 120 min, rate at flow rate 2.0 mL/min and sample size 5g (modified from Ge et al.32).

Cholesterol (mg/100 g)

500 400 300 200 100 0 0

5

10

15

20

25

Carbon dioxide (kg)

34.5 MPa/50°C 34.5 MPa/40°C

13.8 MPa/50°C 24.1 MPa/40°C

FIGURE 1.10 Supercritical carbon dioxide extraction of cholesterol in beef tallow, different pressures and temperatures (modified from Chao et al.44).

dehydrated chicken meat. The combination of pressure and temperature of 38.6 MPa and 55°C, respectively, yielded lipid (89%) and cholesterol (90%), while a pressure and temperature combination of 30.3 MPa and 45°C, respectively, produced much lower yield.

1.4.3 MOISTURE CONTENT

OF

RAW MATERIALS

Moisture content is a factor that influences extraction yield of bioactive compounds as shown in Table 1.9. A maxima yield of 1678 mg/100 g was achieved at 5% moisture content, and any further increase or decrease in moisture reduced that yield. Therefore, it is important to establish this magic number in order to maximize yield in the SCE process. The effect of sample pretreatment is crucial in attaining this

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 30 Thursday, July 20, 2006 9:35 AM

30

Functional Food Ingredients and Nutraceuticals

TABLE 1.9 The Effect of Moisture Variation on Bioactive Compound Yield during Supercritical CO2 Extraction of Wheat Germ Water Content (% wet basis)

Yield (mg/100 g)

4 5 8 12

1470 1678 1352 1290

Source: Modified from Ge et al.32

objective. Water removal in most cases frees the internal pores and thus increases the mass transport intensity, because water in the sample matrix would inhibit the flow of SCF by changing surface tension and contact angles as a result of phase interaction between the three components (water, sample matrix, and SCF). For example, the higher the moisture content, the higher the probability for the formation of a thin film of water between the sample matrix and the SCF phase. Water has a small but finite solubility in SC-CO2, and as a result it can also be extracted with the targeted components, and its separation can then only be done at the end of the process.

1.4.4 COSOLVENT The use of cosolvent (entrainers) during SCF extraction is key to enhancing the extraction efficiency and cost effectiveness of the processes. Joslin et al.99 indicated two significant attributes of cosolvents: the interaction between the cosolvent and the solute (direct effect) and the cosolvent-solvent interactions (indirect effect). Cosolvents used in small doses (1% to 5% mol) in SCF can change the overall characteristics of the extraction fluid such as polarity, solvent strength, and specific interactions. These changes in turn can significantly alter the density and compressibility of the original SCF.5 Additionally, they can improve selectivity for desired components and facilitate selective fractional separations. Table 1.10 summarizes the results of different cosolvents (ethanol, methylene chloride, and methanol). Ethanol was found to have the greatest enhancement factor while methanol had the lowest.100 The solubility enhancement is the result of complex interaction between the -carotene, SC-CO2 , and cosolvent as indicated by Joslin et al.99 Temelli et al.101 observed an enhancement factor of 64, 63, and 29 by using ethanol as cosolvent to extract palmitic acid, stearic and behenic fatty acids, respectively. Baysal et al.63 used ethanol at different concentrations (5%, 10%, and 15%) to recover -carotene and lycopene from tomato paste. Although they observed that, with a high ethanol

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 31 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

31

TABLE 1.10 The Effect of Fractionation of β-Carotene in Cosolvent in Supercritical Carbon Dioxide Mixtures at the Temperature of 70°C. Enhancement Factors were Determined Based on Density (17 mol–1) Pressure (MPa)

21.2 24.9 28.7 32.8 35.8 40.0 43.9

CO2 Density (mol–1) β-Carotene CO2 15.71 16.73 17.65 18.43 18.91 19.49 19.95

Yields (× 107)

1.95 3.33 6.23 10.00 12.50 19.10 25.4

β-Carotene CO2 + 1 wt % Ethanol (Enhancement factor = 4.7) 22.3 15.92 9.6 24.9 16.73 19.5 31.6 18.22 25.2 37.4 19.14 37.5 β-Carotene CO2 + 1 wt % methylene chloride (Enhancement factor = 3.5) 23.4 16.28 12.8 24.7 16.67 13.3 31.2 18.16 21.7 37.0 19.08 27.7 β-Carotene CO2 + 1 wt % Methanol (Enhancement factor = 2.1) 18.0 13.92 3.98 26.8 17.26 9.62 33.0 18.47 15.60 37.3 19.12 30.60 Source: Modified from Cygnarowicz et al.100

concentration, the extraction was hindered due to a decrease in the homogeneity of the extraction mixture, no statistically significant differences were found between the 10% and 15% concentrations. Quancheng et al.102 also investigated the effect of cosolvents on the extraction of tocopherols. They extracted tocopherols from rapeseed deodorizer distillate at 2%, 4%, and 6% with methanol, ethanol, and mixed

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 32 Thursday, July 20, 2006 9:35 AM

32

Functional Food Ingredients and Nutraceuticals

ethanol as cosolvents, under operating conditions of 12 MPa and 60°C. The use of 4% cosolvent gave the best tocopherol yield followed by 2% and 6%, and the 2% and 6% methanol concentration gave the best yield of the three and the mixed ethanol the worst.

1.4.5 PARTICLE SIZE Particle size may have a significant impact on the flow behavior of SCF in the sample matrix. The mechanism of sample pretreatment, for example the methods of drying (air, oven, vacuum, or freeze drying), would influence particle sizes when subjected to attrition or size reduction. The sizes of particles, shapes, and their random layout (size distribution) would determine what goes through the medium and how fast. The layout would influence the type of pore, either open or blind pores, or their degree of interconnectedness. Process parameters such as pressure influence particle size distributions. Pressure tends to create compactedness,95 thus decreases the intergranular porosity resulting in increased solid density. The smaller the particle sizes, the larger is the surface area, and as a result bioactive components are released easily. However, Coelho et al.103 observed no significant effect of particle sizes on the extraction yield as a function of extraction time at a fixed flow rate. The oxygenated compounds increased from 81% to 85% as particle size decreased, as shown in Table 1.11. However, the findings of Ge et al.32 were contrary to Coelho et al.103 in their study on the effect of particle sizes on wheat germ (Table 1.12). Papamichail et al.104 extracted essential oil from celery with SC-CO2. They experienced increased yield (more oil released) as the particle sizes of the seed decreased and attributed that to the pretreatment milling and sieving. Maximum yield of 1838 mg/100 g was obtained with optimum particle size of 0.505 mm. Also, they observed that very fine and big particle sizes have low extraction yield probably due to higher resistances to mass transfer because of the compact tendency reflecting reduced pore sizes in finer particle sizes while less interactions with the supercritical fluids in the case for the latter. Likewise larger particles contain undamaged cell walls rendering then impervious,60 although some believed that higher SC-CO2 flow rate is capable of degrading protein structure to release the targeted bioactive components.105 Nelson106 revealed the need to break the lipoprotein matrix in order to release the embedded lipids in fish during SCE. Papamichail et al.104 reported that it was possible to extract more essential oil per kg of CO2 at a lower flow rate due to the intraparticle diffusion resistance. Therefore, a threshold level for each product has to be experimentally determined in pursuance to pilot plant or industrial processing.

1.4.6 FLOW RATE Coelho et al.103 presented three scenarios of 2.3, 1.5, and 0.85 kg/h flow rates as a function of extraction time and concluded the highest flow rate of 2.3 kg/h gave the higher rate of extraction of Foeniculum vulgare volatile oil with SC-CO2 at 9 MPa and 40°C. A similar trend was observed by Ge et al.32 who reported a flow rate of 3 mL/min yielded 1927 mg/100 g of bioactive component from wheat germ. Some

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 33 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

TABLE 1.11 Comparative Supercritical CO2 and Hydrodistillation Extraction of Volatile Components from Fennel (Foeniculum vulgare) Fruits of Different Particle Sizes SCE CO2 (%)

HD (%) Volatile Components Canfene Sebinene Myrcene -Phellandrene Limonene -Terpinene Terpinolene Fenchone Estragol (E)-Anethole Piperitenone oxide Unknowns Waxes

Stalks

Fruits (0.5 mm)

Fruits (0.55 mm)

Fruits (0.35 mm)

0.2 0.1 1.2 2.0 2.1 Tr 0.5 15.8 18.9 42.5 0.2 4.8

0.2 0.2 1.4 2.2 3.6 0.1 0.6 16.8 20.9 42.2 0.2 3.4

0.2 0.2 1.4 2.2 3.5 Tr 0.6 16.2 21.0 42.5 0.3 5.5 0.6

0.1 0.2 1.3 1.9 3.1 Tr 0.6 17.1 21.9 44.6 0.3 0.3

Tr (Trace < 0.05)). Source: Modified from Coelho et al.103

TABLE 1.12 The Effect of Particle Sizes on Bioactive Compound Yield during Supercritical CO2 Extraction of Wheat Germ Sieving (mesh)

Particle Size (mm)

Yield (mg/100 g)

No grinding 20 30 40 60 80 100

2.1 0.86 0.51 0.40 0.22 0.18 0.13

1610 1710 1838 1550 1070 890 742

Source: Modified from Ge et al.32

© 2007 by Taylor & Francis Group, LLC

33

2441_C001.fm Page 34 Thursday, July 20, 2006 9:35 AM

34

Functional Food Ingredients and Nutraceuticals

Yield (kg/kg feed)

0.18

0.12

0.06

0 0

50

100 150 Time (min) 1.1 kg/h

200

250

3 kg/h

FIGURE 1.11 The effect of flow rate on supercritical extraction of essential oil from celery at pressure 15 MPa and temperature 45°C (modified from Baysal et al.63).

of these facts were corroborated by Peker et al.88 in their experimental study on extraction rates of coffee beans with SC-CO2. They indicated the need for long extraction time in conditions where low flow rate is used. Figure 1.11 shows an apparent yield at high flow rate for extracting celery oil104 and Juniperus oxycedrus essential oil.8 Summarized in Table 1.13 are the results of Baysal et al.63 where a flow rate of 4 kg/h was identified as the optimum condition for attaining the highest yield. Similar trend was observed by Ferreira et al.18 for extracting essential oil from black pepper. They observed larger yield at 30 MPa using the upper-level flow rate (10.54 kg/s). When maximum solubility is attained, the highest CO2 flow rate would offer the highest recovery in extracting lipids from fish.72

1.4.7 EFFECT

OF

TIME

ON

YIELD

Several factors have direct or indirect implication on yield during SCE. Resident time is an important factor that influences yield and the economic viability of the process. Other factors such as temperature and pressure could have individual or

TABLE 1.13 Effect of Flow Rate on the Supercritical Extraction of Lycopene and -Carotene from Tomato Paste Flow Rate (kg/h)

Extraction Time (h)

Lycopene (%)

β-carotene (%)

2 4 8

4 2 1

14 22 20

30 43 34

Source: Baysal et al.63

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 35 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

35

60

Content (%)

50 40 30 20 10 0 0

50

100 HM

150 200 Time (min) OM

250 HS

300

350

OS

Cumulative quantity (g)

(a)

1.5

1

0.5

0 0

50

100

150

200

250

300

350

Time (min) HM

OM

HS (b)

OS

Overall

FIGURE 1.12 Families of flavor compounds extracted from Santolina Insularis at different supercritical extraction times. HM: hydrocarbon monoterpenes; OM: Oxygenated monoterpenes; HS: Hydrocarbon sesquiterpenes; and OS: Oxygenated sesquiterpenes (modified from Cherchi et al.107).

combined effects. Cherchi et al.107 performed detailed analysis of flavor compounds in essential oil extracted from Santolina insularis by SC-CO2 extraction. They reported the change in concentration exhibited by the monoterpenes from 50% in the fraction collected after 30 min to 10% in the fraction collected after 240 min (Figure 1.12a) under optimum conditions of 9 MPa and 50°C and a two-stage separation. The first-stage separation was accomplished under 9 MPa and −12°C, while the final stage used 2 MPa and 15°C. The yield becomes asymptotic at 1.75% with increased extraction time while the rate decreases (Figure 1.12b). Hawthorne et al.108 studied the extraction rate of basil conducted at 30 MPa and 45°C for 10 min and identified 1,8-cineole, estragole, eugenol, and selinene. Yield in most cases

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 36 Thursday, July 20, 2006 9:35 AM

36

Functional Food Ingredients and Nutraceuticals

was dependent on time. Tamelli et al.72 considered 3 to 4 h sufficient time to extract all extractable lipids from freeze-dried krill and 6 h for rainbow trout regardless of the conditions.

1.5 SUMMARY Supercritical carbon dioxide extraction has been shown to be a viable alternative to the conventional solvent extraction technique to extract bioactive components from agricultural materials. It offers a unique advantage of adding value to agricultural waste by extracting antioxidants and flavonoids (lycopene from tomato skin, essential oils, and flavor and fragrances, etc.), which are then used for the fortification of foods and other applications. Its drawbacks are the difficulties in extracting polar compounds and its susceptibility of extracting compounds from a complex matrix where the phase interaction with the intrinsic properties of the product inhibits its effectiveness. Many of these drawbacks can be ameliorated by using cosolvents. However, much investigation is required to understand the solvation effects on targeted bioactive components being extracted. Supercritical fluid extraction technology can be utilized to provide healthy snack foods, a problem that faces the snack food industry, with regard to fat/oil contents, which is becoming a greater public health concern. A lot of cost is incurred reformulating products such as French fries, onion rings, and other snack foods to eliminate trans fats and high cholesterol content. Defatting and decholesterol with SCF has been demonstrated as applicable to food products. Although most of the tests were conducted on dehydrated products, research has shown successful application of SC-CO2 on high moisture products where extraction could be accomplished without compromising the organoleptic characteristics. The SCF technology is available in the form of single-stage batch and could be upgraded to multistage semicontinuous batch coupled with a multiseparation process. Although much work was accomplished, it would not warrant complacency, since batch modes render the SCF technology cumbersome for certain industrial applications, and has been the drawback for broader commercial adaptation. However, the needs to improve the design into continuous modes are growing. One possibility is to integrate membrane technology into a supercritical process. Study on membrane and SC-CO2 has been attempted and needs an aggressive pursuance. This concept would make feasible the extraction of bioactive components from aqueous, less viscous, and pumpable substances in a continuous mode. Supercritical carbon dioxide extraction could only be cost effective under largescale production, which made it ideal for decaffeination of coffee, tea, and hops. Although it is expensive, it is much more economical when compared to conventional solvent extraction at high production scales. With improved processing conditions and reduced cost, SCE will become even more economical at low throughput.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 37 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

37

REFERENCES 1. Sihvonen, M., Jarvenpaa, E., Hietaniemi, V., and Huopalahti, R., Advances in supercritical carbon dioxide technologies, Trends in Food Science & Technology, 10, 217–222, 1999. 2. Steinhagen, V., Bork, M., and Koerner, J.P., Application of supercritical fluids in pharmaceutical industry, (unknown); 3. Brunner, G., Supercritical fluids: technology and application to food processing, Journal of Food Engineering, 67, 21–33, 2005. 4. Uquiche, E., del Valle, J.M., and Ortiz, J., Supercritical carbon dioxide extraction of red pepper (Capsicum annuum L.) oleorisin, Journal of Food Engineering, 65, 55–66, 2004. 5. Mukhopadhyay, M., Natural Extracts Using Supercritical Carbon Dioxide, CRC Press, Boca Raton, FL, 2000. 6. Rozzi, N.L. and Singh, R.K., Supercritical fluids and the food industry: comprehensive review, Food Science and Food Safety, 1, 33–34, 2002. 7. Rizvi, S.S.H., Yu, Z.R., Bhaskar, A.R., and Chidambara, R.C.B., Fundamentals of processing with supercritical fluids, in Supercritical Fluids Processing of Food and Bio Materials, anon. eds., Chapman and Hall, Glasgow, 1994, 1–26. 8. Marongiu, B., Porcedda, S., Caredda, A., De Gioannis, B., Vargiu, L., and La Colla, P., Extraction of Juniperus Oxycedrus Ssp Oxycedrus essential oil by supercritical carbon dioxide: influence of some process parameters and biological activity, Flavour and Fragrance Journal, 18, 390–397, 2003. 9. Pourmortazavi, S.M., Baghaee, P., and Mirhosseini, M.A., Extraction of volatile compounds from Juniperus communis L. leaves with supercritical fluid carbon dioxide: comparison with hydrodistillation, Flavour and Fragrance Journal, 19, 417–420, 2004. 10. Rosa, P.T.V. and Meireless, A.A.A., Rapid estimation of the manufacturing cost of extracts obtained by supercritical fluid extraction, Journal of Food Engineering, 67, 235–240, 2005. 11. Perrut, M., Supercritical fluid applications: industrial development and economic issues, Industrial and Engineering Chemistry Research, 39, 4531–4535, 2000. 12. Ibanez, E., Palacios, J., Senorans, F.J., Santa-Maria, G., Tabera, J., and Reglero, G., Isolation and separation of tocopherols from olive by-products with supercritical fluids, Journal of the American Oil Chemists’ Society, 77, 187–190, 2000. 13. Moret, S. and Conte, L.S., Polycyclic aromatic hydrocarbons in edible fats and oils: occurrence and analytical methods, Journal of Chromatography A, 882, 245–253, 2000. 14. Mohamed, R.S. and Mansoori, G.A., The use of supercritical fluid technology in food processing, Featured article, Journal of Food Technology Magazine, 2002. 15. Eggers, R., Ambrogi, A., and von Schnitzler, J., Special features of SCF solid extraction of natural products: deoiling of wheat gluten and extraction of rose hip oil, Brazilian Journal of Chemical Engineering, 17, 329–334, 2000. 16. Gonzalez, J., Fontal, C.O.I., Vieytes, M.R., Vietyes, J.M., and Botana, L.M., Basis for a new procedure to eliminate diarrheic shellfish toxins from a contaminated matrix, Journal of Agricultural and Food Chemistry, 50, 400–405, 2002. 17. King, J.W., Hill, H.H., Jr., and Lee, M.L., Analytical supercritical fluid chromatography and extraction. in Supplement and Cumulative Index, anon. eds., John Wiley & Sons, New York, 1993, 1–83.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 38 Thursday, July 20, 2006 9:35 AM

38

Functional Food Ingredients and Nutraceuticals 18. Ferreira, S.R.S. and Meireles, M.A.A., Modeling the supercritical fluid extraction of black pepper (Piper Nigrum L.) essential oil, Journal of Food Engineering, 54, 263–269, 2002. 19. Rezzoug, S.A., Baghdadi, M.W., Louka, N., Boutekedjiret, C., and Allaf, K., Study of a new extraction process: controlled instantaneous decompression, application to the extraction of essential oil from rosemary leaves, Flavour and Fragrance Journal, 13, 251–258, 1998. 20. Reverchon, E., Supercritical fluid extraction and fractionation of essential oils and related products, Journal of Supercritical Fluids, 10, 1–37, 1997. 21. Saltzman, M.W., Sheppard, N.F., Jr., McHugh, M.A., and Dause, R.B., Controlled antibody release from matrix of poly (Ethylene-Covinyl Acetate) fractionated with a supercritical fluid, Journal of Applied Polymer Science, 48, 1793–1800, 1993. 22. Oszagyan, M., Simandi, B., and Sawinsky, J., Supercritical fluid extraction of volatile compounds from Lavandin and Thyme, Flavour and Fragrance Journal, 11, 157–165, 1996. 23. Ozcan, M., Tzakou, O., and Couladis, M., Essential oil composition of Turkish herbal tea (Salvia aucheri Bentham var. canescens Boiss & Heldr.), Flavour and Fragrance Journal, 18, 213–226, 2003. 24. Duquesnoy, E., Marongiu, B., Castola, V., Piras, A., Porcedda, S., and Casanova, J., Analysis of the volatile concentrate of the leaves of Pinus nigra subsp., Chimie et Biomas, Univeriste de Corse, France, 2004. 25. Boutekedjiret, C., Bentahar, F., Belabbes, R., and Bessiere, J.M., Extraction of rosemary essential oil by steam distillation and hydrodistillation, Flavour and Fragrance Journal, 18, 481–484, 2003. 26. Gamse, T., Supercritical extraction of liquid mixtures, Lecture Series Socrates Intensive Course: High pressure chemical engineering process: basic applications, Institut für Thermische Verfahrenstechnik und Umwelttechnik, Graz, Austria, 2003. 27. Cocero, M.J. and Calvo, L., Supercritical fluid extraction of sunflower seed oil with CO2 ethanol mixtures, Journal of the American Oil Chemists’ Society, 73, 1573, 1996. 28. Friedrich, J.P., List, G.R., and Heakin, A.J., Petroleum-free extraction of oil from soybeans with supercritical CO2, Journal of the American Oil Chemists’ Society, 59, 288, 1982. 29. Ronyai, E., Simandi, B., Tomoskozi, S., Deak, A., Vigh, L., and Weinbrenner, Z., Supercritical fluid extraction of corn germ with carbon dioxide-ethyl alcohol mixture. in Proc. 4th Int. Symp., Supercritical Fluid Extraction, Sendai, Japan, 1997. 30. Friedrich, J.P. and Pryde, E.H., Supercritical CO2 extraction of lipid-bearing materials and characterization of products, Journal of the American Oil Chemists’ Society, 61, 223, 1984. 31. Sovova, H., Komers, R., Kucera, J., and Jez, J., Supercritical carbon dioxide extraction of caraway essential oil, Chemical Engineering Science, 49, 2499–2505, 1994. 32. Ge, Y., Yan, H., Hui, B., Ni, Y., Wang, S., and Cai, T., Extraction of natural vitamin E from wheat germ by supercritical carbon dioxide, Journal of Agricultural and Food Chemistry, 50, 686–689, 2002. 33. Taniguchi, M., Nomura, R., Kijima, I., and Kobayashi, T., Preparation of defatted mustard by extraction with supercritical carbon dioxide, Agricultural and Biological Chemistry, 51, 413–417, 1987. 34. Dunford, N.T. and Temelli, F., Extraction and fractionation of canola phospholipids with supercritical carbon dioxide and ethanol mixture, Journal of the American Oil Chemists’ Society, 70, 1009–1015, 1995.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 39 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

39

35. Poiana, M., Fresa, R., and Mincione, B., Supercritical carbon dioxide extraction of bergamot peels: extraction kinetics of oil and its components, Flavour and Fragrance Journal, 14, 358–366, 1999. 36. Favati, F., King, J.W., Friedrich, J.P., and Eskins, K., Supercritical CO2 extraction of carotene and lutein from leaf protein concentrate, Journal of Food Science, 1532–1536, 1988. 37. Wehling, R., Froning, G.W., Cuppett, S.L., and Niemann, L., Extraction of cholesterol and other lipids from dehydrated beef using supercritical carbon dioxide, Journal of Agricultural and Food Chemistry, 40, 1204–1206, 1992. 38. Yamaguchi, K., Murakami, M., Nakano, H., Konosu, S., Kokura, T., Yamamoto, H., Kosaka, M., and Hata, K., Supercritical carbon dioxide extraction of oil from Antarctic krill, Journal of Agricultural and Food Chemistry, 34, 904–907, 1986. 39. Chao, R.R., Mulvaney, S.J., Swanson, D.R., Hsieh, F., and Tempesta, M.S., Supercritical CO2 extraction of annatto (Bixa orellana) pigments and some characteristics of the color extracts, Journal of Food Science, 56, 80–83, 1991. 40. Chao, R.R., Mulvaney, S.J., Bailey, M.E., and Fernando, L.N., Supercritical CO2 conditions affecting extraction of lipid and cholesterol from ground beef, Journal of Food Science, 56, 183–187, 1991. 41. Degnan, A.J., von Elbe, J.H., and Hartel, R.W., Extraction of annatto seed pigment by supercritical carbon dioxide, Journal of Food Science, 56, 1656–1659, 1991. 42. Favati, F., King, J.W., and Mazzati, M., Supercritical carbon dioxide extraction of evening primrose oil, Journal of the American Oil Chemists’ Society, 68, 422, 1991. 43. Gopalakrishnan, N. and Narayanan, C.S., Supercritical carbon dioxide extraction of cardamom, Journal of Agricultural and Food Chemistry, 39, 1976, 1991. 44. Chao, R.R., Mulvaney, S.J., and Hanah, H., Effects of extraction and fractionation pressures on supercritical extraction of cholesterol from beef tallow, Journal of the American Oil Chemists’ Society, 70, 139–143, 1993. 45. Spanos, G., Chen, H., and Schwartz, S., Supercritical CO2 extraction of -carotene from sweet potatoes, Journal of Food Science, 58, 817–820, 1993. 46. Cygnarowicz-Provost, M., King, J.W., Marmer, W.N., and Magidman, P., Extraction of woolgrease with supercritical carbon dioxide, Journal of the American Oil Chemists’ Society, 71, 223–225, 1994. 47. King, J.W., Favati, F., and Taylor, S.L., Production of tocopherol concentrates by supercritical fluid extraction and chromatography, Separation Science Technology, 31, 1843–1857, 1996. 48. Vega, P.J., Balaban, M., Sims, C.A., O'Keefe, S.F., and Cornell, J.A., Supercritical carbon dioxide extraction efficiency for carotenes from carrots by RSM, Journal of Food Science, 61, 757–759, 1996. 49. Perakis, C., Louli, V., and Magoulas, K., Supercritical fluid extraction of black pepper oil, Journal of Food Engineering, 2004. 71(4), 386–393, 2004. 50. Porta, G.D., Porcedda, S., Morongiu, B., and Reverchon, E., Isolation of eucalyptus oil by supercritical fluid extraction, Flavour and Fragrance Journal, 14, 214–218, 1999. 51. Catchpole, O.J., Grey, J.B., Perry, N.B., Burgess, E.J., Redmond, W.A., and Porter, N.G., Extraction of chili, pepper and ginger with near critical CO2, propane and dimethyl ether: analysis of the extracts by quantitative nuclear magnetic resonance, Journal of Agriculture and Chemistry, 51, 4853–4860, 2003. 52. Franca, L.F. and Meireles, M.A.A., Modeling the extraction of carotene and lipids from pressed palm oil (Elaes guineensis) fiber using supercritical CO2, The Journal of Supercritical Fluids, 18, 35–47, 2000.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 40 Thursday, July 20, 2006 9:35 AM

40

Functional Food Ingredients and Nutraceuticals 53. Cadoni, E., De Giorgi, M.R., Medda, E., and Poma, G., Supercritical CO2 extraction of lycopene and β−carotene from ripe tomatoes, Dyes and Pigments, 44, 2732, 2000. 54. Barton, P., Hughes, R.E., and Hussein, M.M., Supercritical carbon dioxide extraction of peppermint and spearmint, Journal of Supercritical Fluids, 5, 157–162, 1992. 55. Anitescu, G., Doneanu, C., and Radulescu, V., Isolation of coriander oil: comparison between steam distillation and supercritical CO2 extraction, Flavour and Fragrance Journal, 12, 173–176, 1997. 56. Luengthanaphol, S., Mongkholkhajornsilp, D., Douglas, S., Douglas, P.L., Pengsopa, L.I., and Pongamphai, S., Extraction of antioxidants from sweet Thai tamarind seed coat: preliminary experiments, Journal of Food Engineering, 63, 247–252, 2004. 57. Lee, W.Y., Cho, Y.J., Oh, S.L., Park, J.H., Cha, W.S., Jung, J.Y., and Choi, Y.H., Extraction of grape seed oil by supercritical CO2 and ethanol modifier, Food Science and Biotechnology, 9, 174–178, 2000. 58. Konig, W.A. and Hochmuth, D.H., Enantioselective gas chromatography in flavor and fragrance analysis: strategies for the identification of known and unknown plant volatiles, Journal of Chromatography, 42, 1–17, 2004. 59. Franca, L.F. and Meireles, M.A.A., Extraction of oil from pressed palm oil (Elaes guineensis) fibers using supercritical CO2, Ciencia e Technologia de Alimentos, 17, 1–12, 1997. 60. del Valle, J.M., de la Fuente, J.C., and Cardarelli, D.A., Contribution to supercritical extraction of vegetable substrates in Latin America, Journal of Food Engineering, 67, 35–57, 2005. 61. Macias-Sanchez, M.D., Mantell, C., Rodriguez, M., Martinez de la Ossa, E., Lubian, L.M., and Montero, O., Supercritical fluid extraction of carotenoids and chlorophyll from Nannochloropsis gaditana, Journal of Food Engineering, 66, 247–251, 2005. 62. Bhattacharjee, P., Ranganathan, T.V., Singhal, R.S., and Kulkarni, P.R., Comparative aroma profiles using supercritical carbon dioxide and likens-nickerson extraction from a commercial brand of basmati rice, Journal of the Science of Food and Agriculture, 83, 880–883, 2003. 63. Baysal, T., Ersus, S., and Starmans, D.A.J., Supercritical CO2 extraction of betacarotene and lycopene from tomato paste waste, Journal of Agricultural and Food Chemistry, 48, 5507–5511, 2000. 64. Brunner, G. and Peter, S., On the solubility of glycerides and fatty acids in compressed gases in the presence of an entrainer, Separation Science Technology, 17, 199–214, 1982. 65. Tsuda, T., Mizuno, K., Oshima, K., Kawakishi, S., and Osawa, T., Supercritical carbon dioxide extraction of antioxidative components from tamarind (Tamarindus indica L.) seed coat, Journal of Agricultural and Food Chemistry, 42, 2671–2674, 1995. 66. Wang, B.J., Liu, C.-T., Tseng, C.-Y., and Yu, Z.-R., Antioxidant activity of Bulpeurum kaoi (Chao et Chuang) fractions fractionated by supercritical CO2, Lebensmitted Wissenschaft und Technologic, 38, 281–287, 2005. 67. Reverchon, E., Rose concentrate fractionation by supercritical CO2, The Journal of Supercritical Fluids, 9, 199–204, 1996. 68. Sastry, S.V.G. and Mukhopadhyay, M., Substrate Hindrance in Supercritical Extraction of Fragrance from Jasmine Flowers, 3rd International Symposium Supercritical Fluids, France, 1994. 69. NHLBI (National Heart, Lung, and Blood Institute), retrieved from http://nhlbisupport.com/chd1/ehdexp.htm., 1999.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 41 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

41

70. James, W.P.T. and Ralph, A., Policy and a prudent diet, in Human Nutrition and Dietetics, Garrow, J.S., James, W.T.P., and Ralph, A., Eds., Churchill Livingstone, Edinburgh, 2000. 71. Griffin, B., Lipoprotein atherogenicity: an overview of current mechanism, Proc. Nutr. Soc. Proceedings of Nutrition Society, 58, 163–169, 1999. 72. Tamelli, F., LeBlanc, E., and Fu, L., Supercritical CO2 extraction of oil from Atlantic mackerel (Scomber scombrus) and protein functionality, Journal of Food Science, 60, 703–06, 1995. 73. Hardardottir, I. and Kinsella, J.E., Extraction of lipid and cholesterol from fish muscle with supercritical fluids, Journal of Food Science, 53, 1656–1661, 1988. 74. Yeh, A., Liang, J.H., and Hwang, L.S., Separation of fatty acids esters from cholesterol in esterified natural and synthetic mixtures by supercritical carbon dioxide, Journal of the American Oil Chemists’ Society, 68, 224–229, 1991. 75. Rizvi, S.S.H. and Bhaskar, A.R., Supercritical fluid processing of milk fat: fractionation, scale-up, and economics, Food Technology, February 1995. 76. Huber, W., Molero, A., Pereyra, C., and de la Ossa, E.M., Dynamic supercritical CO2 extraction of cholesterol from anhydrous milk fat, International Journal of Food Science, 31, 143–151, 1996. 77. Peusch, M., Mueller, S.E., Petz, M., Mueller, A., and Anklam, E., Extraction of capsaicinoids from chillies (Capsicum frutescens L.) and paprika (Capsicum annuum L.) using supercritical fluids and organic solvents, Z. Lebensm. Unters, Forsch. A, 17, 177–186, 1997. 78. Illes, V., Daood, H.G., Perneczki, S., Szokonya, L., and Then, M., Extraction of coriander seed oil by CO2 and propane at super- and subcritical conditions, Journal of Supercritical Fluids, 17, 177–186, 2000. 79. Stahl, E., Quirin, K.W., and Gerard, D., Dense Gases for Extraction and Refining, Springer-Verlag, Berlin, 81–112, 1988. 80. Nguyen, U., Process for extracting antioxidants from labiatae herb, U.S. Patent 5,017,397, 1991. 81. Ashihara, H. and Crozier, A., Caffeine: a well known but little mentioned compound in plant science, Trends in Plant Science, 6, 407–413, 2001. 82. Jameel, S., Genetically decaffeinated coffee, Journal of Bioscience, 28, 529–531, 2003. 83. Ogita, S., Uefuji, H., Choi, Y.E., Hatanaka, T., Ogawa, M., Yamaguchi, Y., Koizumi, N., and Sana, H., Genetic modification of coffee plants, Journal of Plant Biotechnology, 4, 91–94, 2002. 84. Chou, T.M. and Benowitz, N.L., Caffeine and coffee: effect on health and cardiovascular disease, Comprehensive Biochemistry and Physiology, 109C, 173–189, 1994. 85. Nurminen, M.L., Coffee, caffeine and blood pressure, European Journal of Clinical Nutrition, 53, 831–839, 1999. 86. Uefuji, H., Ogita, S., Yamaguchi, Y., Koizumi, N., and Sano, H., Molecular cloning and functional characterization of three distinct N-Methyltransferases involved in the caffeine biosynthetic pathway in coffee plants, Plant Physiology, 132, 372–380, 2003. 87. Caragay, A.B. and Little, A.D., Supercritical fluids for extraction of flavours and fragrances from natural products, Perfumer & Flavorsito, 43–54, 1981. 88. Peker, H., Srinivasan, M.P., Smith, J.M., and McCoy, B.J., Caffeine extraction rates from coffee beans with supercritical carbon dioxide, AIChE Journal, 38, 761–770, 1992. 89. Simopoulos, A., Omega-3 fatty acids in health and disease and growth and development, American Journal of Clinical Nutrition, 54, 438–463, 1991.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 42 Thursday, July 20, 2006 9:35 AM

42

Functional Food Ingredients and Nutraceuticals

90. Ohr, L.M., Nutraceuticals and functional foods, Food Technology, 59, 63–65, 2005. 91. Espinosa, S., Diaz, S., and Brignole, E.A., Thermodynamic modeling and process optimization of supercritical fluid fractionation of fish oil fatty acid ethyl esters, Industrial and Engineering Research, 41, 1516–1527, 2002. 92. Dunford, N.T., Temelli, F., and LeBlanc, E., Supercritical CO2 extraction of oil and residual proteins from Atlantic Mackerel (Scomber Scombrus) as affected by moisture content, Journal of Food Science, 62, 289–294, 1997. 93. Eisenbach, W., Supercritical extraction, Physical Chemistry, 88, 882, 1984. 94. Alkio, M., Gonzalez, C., Jantii, M., and Altonen, O., Purification of polyunsaturated fatty acid ester from tuna oil with supercritical fluid chromatography, Journal of American Oil Chemists’ Society, 77, 315, 2000. 95. Tonthubthimthong, P., Chuapraset, S., Douglas, P., and Luewisutthicat, W., Supercritical CO 2 extraction of nimbin from neem seeds: an experimental study, Journal of Food Engineering, 47, 289–293, 2001. 96. D'Andrea, A., Ferri, D., Maccioni, O., van der Esch, S.A., and Vitali, F., Application of SCF technology to the extraction of active substances from Azadirachta indica (A. Juss) seeds, in Practice Oriented Results on the Use and Production of Neem Ingredients and Pheromones, anon. eds., Giessen, Druck and Graphic, 1994. 97. Gomez, A.M., Lopez, C.P., and de la Ossa, E.M., Recovery of grape seed oil by liquid and supercritical carbon dioxide extraction: a comparison with conventional solvent extraction, The Chemical Engineering Journal, 61, 227–231, 1996. 98. Froning, G.W., Fieman, F., Wehling, R.L., Cuppett, S.L., and Niemann, L., Supercritical carbon dioxide extraction of lipids and cholesterol from dehydrated chicken meat, Poultry Science, 73, 571–575, 1994. 99. Joslin, C.G., Gray, C.G., and Goldman, S., Solubility in supercritical fluids from the virial equation of state, Molecular Physics, 89, 489–503, 1996. 100. Cygnarowicz, M.L, Maxwell, R.J., and Seider, W.D., Equilibrium solubilities of -carotene in supercritical carbon dioxide, Fluid Phase Equilibria, 59, 57–71, 1990. 101. Tamelli, F. and Guclu-Ustundag, O., Solubility behaviour of ternary systems of lipids, cosolvents and supercritical carbon dioxide, unknown. 102. Quancheng, Z., Guihua, S., and Hong, J., Concentration of tocopherols by supercritical carbon dioxide with cosolvents, European Food Research and Technology, 219, 398–402, 2004. 103. Coelho, J.A.P., Pereira, A.P., Mendes, R.L., and Palavra, A.M.F., Supercritical carbon dioxide extraction of foeniculum vulgare volatile oil, Flavour and Fragrance Journal, 18, 316–319, 2003. 104. Papamichail, I., Loili, V., and Magoulas, K., Supercritical fluid extraction of celery seed oil, Journal of Supercritical Fluids, 18, 213–226, 2000. 105. Femenia, A., Garcia-Marin, M., Simal, S., Rossello, C., and Blasco, M., Effects of supercritical carbon dioxide (SC-CO2) oil extraction on the cell walls composition of Almond fruits, Journal of Agricultural and Food Chemistry, 49, 5828–5834, 2001. 106. Nelson, G.J., Isolation and purification of lipids from biological matrix, in Analysis of Fat, Oils and Lipoprotein, anon. ed., American Oil Chemists’ Society, Champaign, IL, 1991, 20–59. 107. Cherchi, G., Deidda, D., De Gioannis, B., Marongiu, B., Pompei, R., and Porcedda, S., Extraction of Santolina Insularis essential oil by supercritical carbon dioxide: influence of some process parameters and biological activity, Flavour and Fragrance Journal, 16, 35–43, 2001.

© 2007 by Taylor & Francis Group, LLC

2441_C001.fm Page 43 Thursday, July 20, 2006 9:35 AM

Supercritical Fluid Technology for Extraction of Bioactive Components

43

108. Hawthorne, S.B., Riekkola, M.L., Serenius, K., Holm, Y., Hiltunen, R., and Hartonen, K., Comparison of hydrodistillation and supercritical-fluid extraction for the determination of essential oils in aromatic plants. Journal of Chromatography, 634, 297–308, 1993.

© 2007 by Taylor & Francis Group, LLC

2441_C002.fm Page 45 Tuesday, July 18, 2006 8:12 AM

2

Solubility of Food Components and Product Recovery in the Supercritical Fluid Separation Process John Shi1 and Xiaoqin Zhou2 1 2

Agriculture and Agri-Food Canada University of Waterloo

CONTENTS 2.1 2.2 2.3

Introduction ....................................................................................................45 Solubility of Food Components in SFE ........................................................46 Factors Affecting Solubility in SFE-CO2 ......................................................48 2.3.1 Solvent Selectivity .............................................................................51 2.3.2 Polar and Nonpolar Solvents .............................................................52 2.3.3 Cosolvents-Modification of SFE Fluid Phase Behavior ...................53 2.3.4 Pressure Effects..................................................................................57 2.3.5 Effect of Water Content in Material ..................................................58 2.4 Solubility Prediction ......................................................................................60 2.4.1 Equations of State (EOS)...................................................................65 2.5 Summary ........................................................................................................67 References................................................................................................................68

2.1 INTRODUCTION Supercritical fluid extraction (SFE) has been used in the food industry since 1979,1 and has accrued 23 years of experience in the extraction of oils from plant materials. Initially, it was mainly used in the decaffeination of coffee and tea and the extraction of spices and hops. Those had to be large-scale processing operations in order to make them economically feasible and cost effective. However, with the development

45

© 2007 by Taylor & Francis Group, LLC

2441_C002.fm Page 46 Tuesday, July 18, 2006 8:12 AM

46

Functional Food Ingredients and Nutraceuticals

and advancements in processing conditions and equipments, the cost effectiveness improved allowing the small-scale productions of ultrahigh quality or high margin product using SFE for commercial distribution. The hottest area for supercritical extraction application is with nutraceuticals, as supercritical fluid extraction can be operated at mild conditions and will retain the bioactivity of components. Solubility is the amount of a substance dissolved by the solvent at thermodynamic equilibrium. For a multicomponent system (different components in the analyzed matrix and a mixture of solvents), the solubility of various components determines the selectivity and gives the opportunity of solving a separation problem. To design a supercritical fluid extraction process, the solubility of solutes in the supercritical fluid is fundamental. It influences the ideal operating conditions of the extractor and recovery unit. Also, it controls the minimum amount of supercritical fluid required to complete the extraction. In addition, the primary characteristics of a supercritical fluid which make it so attractive is the continuously changing solvating power. Since the solvating power is sensitive to temperature and pressure in the supercritical region, it can be finely adjusted by varying temperature and pressure. The goal of optimizing temperature and pressure of an SFE system is to maximize the solubility of solutes and increase their upper limit of yield. In past years, study has focused on the experimental solubility data in the supercritical region, solubility prediction, and improvement (modifying solvent). As the molecular structures of the food component in nutraceuticals become more complex, the interaction between solute, solvent, and the solubility of biocomponents in SCF become more complex. In the following sections, recent developments of the research of solubility in SFE related with food science are reviewed.

2.2 SOLUBILITY OF FOOD COMPONENTS IN SFE Because solubility is important to the process design of SFE, a lot of experimental work has been done on the measurements of solubilities of food component in SCF. Typical solubility behavior of a solid solute in SCF solvent is shown at Figure 2.1. Two convergence points are shown at PL and PU, and one minima (Pmin). At low pressure region (< Pmin), the solubility decreases as pressure increases, while in the region of pressure Pmin< P < PU, the solubility increases sharply with the increase of pressure. This region is usually observed in the near-critical and high compressed region of an SCF solvent. It indicates that in the supercritical fluid extraction, the solubility can be controlled by pressure. In other words, it is related to the state of the supercritical fluid. Hence, the prediction of solubility is usually based on the equation of state (EOS) of the solvent. The crossover points PL and PU are attracting some interest as a method of separating components with small difference in selectivity, such as isomers. For a multicomponent system, the crossover point of each component may not overlap, so there is a “crossover region” where most of the crossover points are located. At this point or in this region, the solubility of components are similar. Johnston et al.2 pointed out that the crossover point in fact is the turning point of the isotherm line of solubility, where

© 2007 by Taylor & Francis Group, LLC

2441_C002.fm Page 47 Tuesday, July 18, 2006 8:12 AM

Solubility of Food Components and Product Recovery

T3

47

T2 T1

Mole fraction

T3 > T2 > T1

Pmin

PL

Pressure

PU

FIGURE 2.1 Solubility behavior of a solid solute in an SCF solvent.

∂ ( ln y2 ) =0 ∂T where y2 is the mole fraction of solute in gas phase and T is temperature. For one component, a slight increase in temperature at PU will cause the solubility to increase above the crossover pressure and decrease below the crossover pressure. The results are reversed at PL. Under these conditions a retrograde region is formed. For multicomponent systems, at the crossover point of one component, its solubility will not change with temperature, while others do, so the selectivity of those components increases. Through several cyclings of retrograde crystallization or solvation, the components are separated. The operation becomes similar to distillation and requires a temperature gradient. Chimowitz and Pennisi,3 and Foster et al.4 gave a detailed description of the operation in the crossover pressure region. Because fewer data are available in the critical region, this “distillation” operation runs more often in the upper crossover pressure than in lower crossover pressure. The other interesting observation is the sharp change of solubility with pressure. In view of the macroscopic thermodynamics, the influence of the pressure to the solubility can be explained by partial molar volume V2 directly ln φ2 =

1 RT

∫

P

V2 dp

0

where V2s is the saturate volume, f2 is the fugacity coefficient, and R is the gas constant. The partial molar volume is a differential quantity that describes the solution behavior at a particular pressure. Here, fugacity coefficient f2 is the pressure integral of partial molar volume (Equation 2.3). Using the equation above, the solubility behavior vs. pressure can be easily explained.

© 2007 by Taylor & Francis Group, LLC

2441_C002.fm Page 48 Tuesday, July 18, 2006 8:12 AM

48

Functional Food Ingredients and Nutraceuticals

When the pressure is much lower than the critical point, the partial molar volume is not a function of composition, so the equation is simplified to ∂ ( ln y2 ) V2s − V2 = ∂P RT At low pressure V2s 0 , Y j+ = 0 for Xbj = 0

3.4.2 MODEL

FOR

(3.21)

VEGETABLE OIL EXTRACTION

To illustrate the dependence of concentration profiles and the extraction curve on the type of phase equilibrium, we have chosen a basic set of model parameters, r = 0.6, Γ = 0.167, Θf = 0.25, and Θc = 12 (see Table 3.1). To simulate SFE of vegetable oils with plug flow, model Equation 3.5b through Equation 3.9b plus Equation 3.12b and Equation 3.14b were numerically integrated together with Equation 3.21 using the basic set of parameters and n = 40; the results are shown in Figure 3.3. The depicted solid phase concentration is calculated as mean concentration in particle X = rXb + (1 - r)Xc. The concentration profiles in both phases gradually shift to the right-hand side of the figure, in the direction of solvent flow. As long as the fluid phase concentration at solvent outlet remains constant, the extraction curve e/xu versus τ is a straight line. When the extraction of solute from broken cells is complete, mass transfer from particles continues much slower and concentration profiles flatten; the slope of the extraction curve decreases proportionally with the decrease in

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 88 Tuesday, July 18, 2006 8:19 AM

88

Functional Food Ingredients and Nutraceuticals 1 0.8

Y

0.6 0.4 0.2 0

0

0.2

0.4

0.6

0.8

1

0.6

0.8

1

z 1 0.8

X

0.6 0.4 0.2 0

0

0.2

0.4

z 1 0.8

e/xu

0.6 0.4 0.2 0

0

10

20 t

30

40

FIGURE 3.3 Simulated concentration profiles and extraction curve for plug flow and constant solubility. Model parameters: r = 0.6, Γ = 0.167, Θf = 0.25, Θc = 12, n = 40. Fluid phase concentration profiles Y(z) and solid phase concentration profiles X(z) shifting to the right and decreasing: (_______) solution of complete model equations for τ = 1, 2, 4, 6, 8, 10, 20, 40. Extraction yield e/xu: (_______) solution of complete model equations; (+ + + +) approximate sections of extraction curve according to Equation 3.24 and Equation 3.25.

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 89 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

89

outlet fluid phase concentration. The transition from equilibrium-controlled period to diffusion-controlled period occurs at extraction yield higher than the yield of easily accessible solute rxu . The reason for the increase is that a certain amount of solute from the intact cells is transferred into the broken cells and becomes easily accessible while the solute deposited initially in broken cells is transferred into the fluid phase.

3.4.3 EQUILIBRIUM-CONTROLLED PERIOD Several researchers assume constant solubility and plug flow in order to model the first period of oil extraction with SC-CO2. Lee et al.9 solved the mass balance equations for oil extraction from rape seed flakes numerically and adjusted external mass transfer coefficient kfa0 to experimental data with the result that kfa0 was proportional to u0.54. Brunner10 pointed out that oil extraction is a steady-state process as long as any easily accessible oil remains at the solvent inlet, where it is depleted first. (It can be shown that the steady-state extraction lasts until τ = τ1 = Θf(1 – Γ)/Γ. Thus, in our example depicted in Figure 3.3, τ1 = 1.25.) During the steady-state extraction, both fluid phase concentration profile and extraction rate are constant and the mean driving force in the extraction bed ys – ymean is related to measurable quantities ys and y(h = H) according to ys − ymean =

y (h = H ) ys ln ys − y ( h = H )

(3.22)

Taking into account that in a steady-state period ∂y/∂t = 0, volumetric mass transfer · coefficient kfa0 can be determined from mass flux Vkfa0ρf (ys – ymean) = Qy(h = H). Ferreira et al.11 applied this procedure to evaluate the external mass transfer coefficient with experimental data for black pepper with SC-CO2 at 15–30 MPa and 30C –50C. The botanic material was ground to a mean size of approximately 0.1 mm, and the interstitial solvent velocity, u, was in the range of 0.2–0.9 mm/s. Mass transfer coefficient kf , increasing with increasing interstitial velocity, was found to be independent of pressure and temperature and its value was between 3 × 10–4 and 9 × 10–4 m/s. In terms of the present model, Equation 3.22 corresponds to a steady-state fluid phase concentration profile  z  Y = 1 − exp  −  Θ f 

(3.23)

(depicted in Figure 3.3a) for τ = 1, and the dimensionless extraction yield is based on the following relationship

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 90 Tuesday, July 18, 2006 8:19 AM

90

Functional Food Ingredients and Nutraceuticals

  1 e = Γr τ 1 − exp  − xu  Θf 

   

(3.24)

The exponential term for Θf < 0.22 is lower than 0.01. For these values of Θf , the solvent flowing out from the extraction bed is practically saturated with solute. After exhaustion of the easily accessible solute from particles at the solvent inlet, that is, when xb(z = 0) becomes zero, the fluid phase concentration profile starts shifting toward the solvent outlet and the overall extraction rate gradually slows down.

3.4.4 DIFFUSION-CONTROLLED PERIOD In the second part of the extraction curve the fluid phase concentration is much lower than the solute solubility in the solvent. If the assumption on equilibrium at a particle surface and constant solubility holds, the solute that arrives from the particle core to its surface dissolves immediately in the solvent and the concentration in broken cells becomes practically zero. The driving force for internal diffusion according to Equation 3.12 is then equal to xc and the extraction yield in the second period is approximated by the following equation e  τ − τc  = 1 − (1 − r ) exp  −  xu Θ c 

(3.25)

where the shift in time, τc > 0, is related to the fact that the driving force for solute transfer from the intact to broken cells was decreased in the first extraction period by the nonzero concentration in the broken cells. As shown in Figure 3.3, the extraction yield calculated using the complete model equations and the yield according to Equation 3.25 with τc adjusted to 3.75 overlap in the second extraction period.

3.4.5 APPROXIMATE SOLUTION

FOR

BOTH PERIODS

An approximate model with plug flow based on the concept of broken and intact cells was derived by Sovová2 as an extension of Lack’s model. The model simulates both concentration profiles and extraction yield, which is given as   1 e = Γr τ 1 − exp  − xu  Θf 

 Θf   for τ ≤ τ m = Γ 

(

)

   exp  τ − τ m Θc  − 1 + r e  = r Γτ − Θ f exp  r ΓΘc ln − 1  xu r     for

© 2007 by Taylor & Francis Group, LLC

  Θ f    

2441_C003.fm Page 91 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

τ < τn =

91

 Θf  1  + Θ c ln  1 − r + r exp  ,  r ΓΘ c   Γ 

 e  1    τ − τ m    − 1 expp  − = 1 − r ΓΘ c ln 1 + (1 − r )  exp     r ΓΘ c    Θ c   xu   for τ ≥ τn

(3.26)

The steady-state extraction period according to Equation 3.26 finishes when τ = τm. Then the fluid phase concentration profile starts shifting through the extraction bed. At time τ = τn the last easily accessible solute is transferred from the broken cells to fluid phase and the extraction is further controlled solely by intraparticle diffusion. A model2 was derived for γ 0 where the solute accumulation in the fluid phase was not taken into account. Therefore a discrepancy exists between extraction curves calculated according to Equation 3.26 and those obtained by numerical solution to the complete model equations, as illustrated in Figure 3.4 for the basic set of parameters r = 0.6, Γ = 0.167, Θf = 0.25, and Θc = 12. To simulate precisely the extraction curve calculated with the complete model, the approximate parameters would have to be adjusted to r = 0.697, Γ = 0.211, Θf = 0.917, and Θc = 11.6. Thus, the approximate model is useful to smooth the experimental data and obtain the first estimation of model parameters, which, however, should be refined by further calculations. As the approximate model does not involve solute-matrix interaction, it should not be applied for the system where solute interacts with matrix. 1 0.8

e/xu

0.6 0.4 0.2 0

0

10

20 τ

30

40

FIGURE 3.4 Comparison of extraction curves calculated for constant solubility, plug flow, and model parameters r = 0.6, Γ = 0.167, Θf = 0.25, Θc= 12, and n = 40. (_______) solution of complete model equations; (+ + + +) approximate model2 with analytical solution.

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 92 Tuesday, July 18, 2006 8:19 AM

92

Functional Food Ingredients and Nutraceuticals

3.5 SFE OF ESSENTIAL OILS 3.5.1 SOLUTE-MATRIX INTERACTION Essential oils as mixtures of volatile and predominantly low-polar compounds are more soluble in SC-CO2 than vegetable oils. Their main components, terpenes and oxygenated terpenes, are fully miscible with SC-CO2 at temperatures of 40°C –50°C and pressures of 10–20 MPa, as Reverchon states in a comprehensive review on SFE of essential oils.12 Nevertheless, when essential oils are extracted with SC-CO2 from leaves or other botanic materials where their concentrations are only a few percent or less, their equilibrium fluid phase concentration is much lower than that measured in absence of tissue matrix. For example, Goto et al.13 observed that the equilibrium fluid phase concentration of menthol, a major component in peppermint essential oil, during SFE from peppermint leaves, was two orders of magnitude lower than the solubility of pure menthol in SC-CO2. To explain how the matrix affects the essential oil equilibrium, the authors assumed that the essential oil is adsorbed on the lipids present in the leaves and thus the experimental equilibrium concentration is represented by an adsorption isotherm. To describe the solute-matrix interaction they applied the simplest and the most frequently used relationship for SFE, linear equilibrium with partition coefficient K y + = Kxb

(3.27)

Its dimensionless form for j-th mixer is Y j+ = KXbj , K = 1 (1 − Γ )

3.5.2 SIMULATION

OF

(3.28)

ESSENTIAL OIL EXTRACTION

Model Equation 3.5b through Equation 3.9b plus Equation 3.12b and Equation 3.14b together with Equation 3.28 were numerically integrated for the basic set of model parameters and n = 40. Due to the linear equilibrium, completely different concentration profiles from vegetable oil extraction were obtained, as shown in Figure 3.5. As the solid phase concentration gradually decreases during the extraction process, the equilibrium concentration in fluid phase at a particle surface also decreases likewise the driving force y+ - y. As a result, the concentration profiles corresponding to solute-matrix interaction are flatter and the extraction becomes slower than in the case without solute-matrix interaction.

3.5.3 TWO EXTRACTION PERIODS The first and second extraction periods cannot be completely separated as it was in the case without solute-matrix interaction. On one hand, the solute diffusing from the intact to broken cells during the first period increases the concentration in broken cells and therefore also the equilibrium fluid phase concentration. On the other hand, as the concentration in the broken cells is directly proportional to the fluid phase

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 93 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

93

1 0.8

Y

0.6 0.4 0.2 0

0

0.2

0.4

0.6

0.8

1

0.6

0.8

1

z 1 0.8

X

0.6 0.4 0.2 0

0

0.2

0.4 z

1 0.8

e/xu

0.6 0.4 0.2 0

0

10

20 τ

30

40

FIGURE 3.5 Simulated concentration profiles and extraction curve for linear equilibrium and plug flow. Model parameters: r = 0.6, Γ = 0.167, Θf = 0.25, Θc = 12, and n = 40. a) Decreasing fluid phase concentration profiles Y(z) and solid phase concentration profiles X(z): (_______) solution of complete model equations for τ = 1, 2, 4, 6, 8, 10, 20, 40. Extraction yield e/xu: (_______) solution of complete model equations; (+ + + +) approximate sections of extraction curve according to Equation 3.24 and Equation 3.29.

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 94 Tuesday, July 18, 2006 8:19 AM

94

Functional Food Ingredients and Nutraceuticals

concentration at a particle surface, it is neither zero nor constant in the second extraction period. Nevertheless, at least the initial slope of the extraction curve can be estimated according to Equation 3.24 as in the case of constant solubility (Figure 3.5). The extraction in the second extraction period is slower than that derived for constant solubility with Equation 3.25, because the nonzero concentration xb reduces the driving force. An acceptable estimate for extraction yield in the second period is given in the following equation1  τ − τc e = 1 − (1 − r ) exp  − xu  Θ c 1 + (1 − r ) K ΓΘ c r 

(

)

    

(3.29)

The agreement of Equation 3.29 with the results obtained by solution of complete model equations for the basic set of model parameters is shown in Figure 3.5.

3.6 SFE FOR COMBINED EQUILIBRIUM 3.6.1 EQUILIBRIUM ACCORDING

TO

PERRUT

ET AL.

A versatile SFE model would combine the constant solubility and linear equilibrium in one expression. Two papers independently solving this task were published almost simultaneously. Goto et al.14 suggested the BET adsorption isotherm which is often used to describe adsorption equilibria. The BET isotherm simulates a smooth increase in equilibrium fluid phase concentration with increasing solid phase concentration, from y+ determined solely by solute-matrix interaction at lowest solid phase concentrations to y+ asymptotically approaching ys at high solid phase concentrations. The other equation, proposed by Perrut et al.15 as a simplified description of phase equilibrium during SFE from plant materials, defines the equilibrium fluid phase concentration as a discontinuous function of solid phase concentration as depicted in Figure 3.6 y ys

b

a

c

KXf d Xf

X1

FIGURE 3.6 Combined equilibrium relationship with constant solubility for high solid phase concentrations and with linear equilibrium for low solid phase concentrations according to Perrut et al.15

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 95 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

y + = ys for xb > xt

y + = Kxb for xb ≤ xt ,

95

(3.30)

where ys > Kxb. The sudden fall in equilibrium concentration y+ at xb = xt can be explained as follows. Matrix has certain capacity of interacting with solute, as expressed by transition concentration xt. As long as the solid phase concentration is less than or equal to xt, all solute in solid phase interacts with matrix and the equilibrium fluid phase concentration is determined by the distribution coefficient K. When the solid phase concentration is higher, part of solute remains free, unbound to the matrix, and its equilibrium fluid phase concentration equals solute solubility. According to this simplified model, when SFE starts at xb0 > xt, solid phase concentration gradually decreases to xt and when it reaches this value, the equilibrium fluid phase concentration at the particle surface falls from ys to Kxt, below the fluid phase concentration in the diffusion layer close to the surface. Mass transfer from the broken cells to fluid phase is interrupted until the fluid phase concentration drops below y = Kxt. In the meantime, the solute is only washed out

(

j f = k f a0 ρ f y + − y

)

for xb ≠ xt or xb = xt , y < Kxt , j f = 0 otherwise. (3.31)

Equation 3.30 includes both equilibrium relationships mentioned above: for xt = 0 it is identical with Equation 3.20 for constant solubility, and for xt > xb0 it describes linear equilibrium according to Equation 3.27. It seems that any solute can occur both in free form and bound to the matrix, depending on its solid phase concentration and on matrix and solvent properties. Indeed, Perrut et al.15 observed the combined equilibrium during SFE of vegetable oil from seed, where constant equilibrium would be expected. When the extraction with plug flow and combined equilibrium (Equation 3.30) starts in section “a” of the equilibrium curve (Figure 3.6), the whole period of solute extraction from the broken cells is governed by constant solubility and the solutematrix-interaction does not influence the course of the extraction. When the extraction starts in section “b,” the outlet concentration firstly approaches the solubility, then the equilibrium value at transition concentration xt, and finally the intraparticle diffusion-controlled extraction follows. When the extraction starts in section “c,” the amount of free solute present in the extraction bed at t = 0 is not sufficient to saturate the solvent and thus the initial outlet concentration is lower than solubility; as soon as the initial solution is washed out of the extractor, only the solute interacting with matrix is extracted. Finally, when the extraction starts in section “d,” all solute interacts with matrix, as described in Section 3.5. In the dimensionless model for a series of mixers, Equation 3.30 reads as follows Y j+ = 1 for Xbj > Xt , Y j+ = KXbj for Xbj ≤ Xt

© 2007 by Taylor & Francis Group, LLC

(3.32)

2441_C003.fm Page 96 Tuesday, July 18, 2006 8:19 AM

96

Functional Food Ingredients and Nutraceuticals

– – where Xt = xt/xu and K = Kxu/y0. Section “a” is characterized by 1 – Γ ≥ 1/K, section “b” by Xt < 1 – Γ < 1/ K , section “c” by Xt = 1 – Γ and K < 1 (1 − Γ ) , and section “d” by Xt ≥ 1 – Γ and K = 1 (1 − Γ ) .

3.6.2 SIMULATION

OF

SFE

WITH

COMBINED EQUILIBRIUM

Model Equation 3.5b through Equation 3.9b plus Equation 3.12b and Equation 3.14b, together with Equation 3.32 were numerically integrated with the basic set of the same parameter values and number of mixers as in the case of constant equilibrium (r = 0.6, Γ = 0.167, Θf = 0.25, Θc = 12, and n = 40), using two more – equilibrium parameters Xt = 0.48 and K = 0.5. Thus, the extraction starts in section “b.” The results shown in Figure 3.7 apparently indicate an extraction with constant equilibrium where the second, diffusion-controlled period starts as soon as the dimensionless solid phase concentration in the broken cells drops to Xt. Further analysis, however, will show that the shape of the extraction curve starting in section “b” is different from that of constant equilibrium.

3.6.3 TWO EXTRACTION PERIODS The first extraction period consists of two sections of different slopes. If the external mass transfer resistance is neglected for the sake of simplicity, the shape of the extraction curve in the first extraction period is calculated as1 e 1 − Γ − Xt = Γr τ for τ ≤ τ1 = 1 + xu Γ 1 − KXt

(

e = Γr  τ1 + KXt ( τ − τ1 )  for τ >τ1 xu

) (3.33)

In the second, diffusion-controlled period, the solute interacting with matrix in broken cells is practically in equilibrium with the solute in fluid phase, and the extraction curve is approximately described by Equation 3.29 (unless Xt = 0, when Equation 3.25 should be applied). As Figure 3.7 shows, the agreement of estimated extraction curve sections with complete extraction curve is good.

3.7 FLOW PATTERNS 3.7.1 AXIAL DISPERSION In the previous sections we have shown how phase equilibrium and diffusion resistance affect the shapes of concentration profiles and the extraction curve. Another aspect that should be taken into account and which is often omitted is the flow pattern in the extraction bed. To express deviations from plug flow, Goodarznia and Eikani16 inserted an axial diffusion term into the differential mass balance equation

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 97 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

97

1 0.8

Y

0.6 0.4 0.2 0

0

0.2

0.4

0.6

0.8

1

0.6

0.8

1

z

1 0.8

X

0.6 0.4 0.2 0

0

0.2

0.4 z

1 0.8

e/xu

0.6 0.4 0.2 0

0

20

40 τ

60

80

FIGURE 3.7 Simulated concentration profiles and extraction curve for plug flow and com– bined equilibrium. Model parameters: r = 0.6, Γ = 0.167, Θf = 0.25, Θc = 12, Xt = 0.48, K = _______ 0.5 concentration profiles X(z) shifting to the right and decreasing: ( ) solution of complete model equations for τ = 1, 2, 4, 6, 8, 10, 20, 40. Extraction yield e/xu: (_______) solution of complete model equations; (- - - -) approximate equilibrium-controlled section according to Equation 3.33; (+ + + +) approximate diffusion-controlled section according to Equation 3.29.

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 98 Tuesday, July 18, 2006 8:19 AM

98

Functional Food Ingredients and Nutraceuticals

ρs (1 − ε ) (1 − r )

 ∂y ∂xc ∂x ∂2 y  ∂y − Dax 2  = 0 + ρs (1 − ε ) r b + ρ f ε  + u ∂h ∂t ∂t ∂h   ∂t (3.34)

where Dax is the axial dispersion coefficient. Reis-Vasco et al.6 applied Equation 3.34 on SC-CO2 extraction of mint oil from leaves and found out that Equation 3.34 yields a better representation of experimental extraction curves than the plug flow model. The axial dispersion values adjusted to fit experimental data were decreasing with increasing flow rate and were about one order of magnitude larger than those obtained from the correlations found in literature. Nevertheless, the authors regarded the introduction of an additional adjustable parameter into the model as too high a cost for slightly better accuracy and therefore selected the plug flow model as the most useful one. Another possibility to express axial dispersion in extraction bed is decreasing the number of mixers in the model Equation 3.5b through Equation 3.14b. Though the changes in the degree of axial dispersion are not continuous as according to Equation 3.34, it is possible to cover a wide range of flow patterns from plug flow for n ∞ up to ideal mixer for n = 1, which is discussed in the next section.

3.7.2 MODELS

FOR

DIFFERENTIAL EXTRACTION BED (IDEAL MIXER)

The models published for SFE from plant materials by Goto et al.13,14 consider a differential extraction bed where no concentration gradient is assumed. Thus, the differential extractor is described by ordinary differential equations for an ideal mixer (Equation 3.6b through Equation 3.8b for n = 1). Such models are simpler than the plug flow models and some of them have analytical solutions. They would be ideal for SFE application in short extraction beds or when strong axial dispersion exists in fluid phase. Compared to plug flow, the extraction in an ideal mixer is slower in the first extraction period due to the lower driving force (Figure 3.8). The following equation

y=

y+ 1+ Θf

and

Γr τ e = xu 1 + Θ f

(3.35)

holds for the initial steady-state period. The effect of external mass transfer resistance on outlet concentration and the extraction rate is more pronounced than in the case of plug flow (Equation 3.24).

3.7.3 NATURAL CONVECTION The best extractor performance is achieved with a plug flow. Unfortunately, supercritical fluids are extremely prone to natural convection, due to their low kinetic viscosity. Such convection can be induced by solvent density differences caused by the differences in local solute concentrations. The occurrence of natural convection

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 99 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

99

1 0.8

e/xu

0.6 y + = ys

0.4 0.2 0

0

10

20 τ

30

40

1 0.8

e/xu

0.6 y+ = Kxb

0.4 0.2 0

0

10

20

30 τ

40

50

60

1 0.8

e/xu

0.6 y+ = ys for xb > xt y+ = Kxb for xb ≤ xt

0.4 0.2 0

0

20

40 τ

60

80

FIGURE 3.8 Effect of flow pattern on extraction curves for three types of equilibrium. Model parameters: r = 0.6, Γ = 0.167, Θf = 0.25, Θc = 12, n = 40, and for combined equilibrium Xt – = 0.48 and K = 0.5. (_______) plug flow; (– – – –) ideal mixer; (- - - - -) two parallel flows with unc = 0.75u.

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 100 Tuesday, July 18, 2006 8:19 AM

100

Functional Food Ingredients and Nutraceuticals

depends also on the direction of solvent flow through the extraction bed, as Beutler et al.17,18 discovered. They observed this phenomenon in SC-CO2 extraction of different solutes, such as black pepper oleoresin or fat, from botanical matrix. When the flow was from the top to the bottom, the fat extraction yield was twice as high as for the opposite flow direction. The authors explained this phenomenon by higher density of solute-laden CO2 than that of pure solvent; the solute-laden solvent can be more easily withdrawn from the bottom of the extractor than from its top. The better withdrawal of solute leads to higher concentration gradients in the solid phase and thus to a higher mass transfer driving force. Barton et al.19 observed the effect of flow direction when they extracted peppermint oil with dense CO2. Though the extraction yields by downflow CO2 reached 90 to 98% of those extracted by steam distillation, the yield of peppermint oil in the CO2 upflow mode was only 76%, even at higher solvent treatments and extraction times. Also, Sovová et al.20 who extracted grape seed oil with SC-CO2 observed retardation in extraction when the flow direction was switched to the upflow mode; the effect was more pronounced for lower interstitial velocities. Equation 3.26 was used to simulate experimental data. While a simple plug flow model was adequate for SFE in downflow mode, the natural convection in upflow mode was simulated assuming parallel flows of different velocities in the extraction bed. The simplest application of this approach is splitting the bed into halves with interstitial velocities u + unc and u – unc. The resulting extraction curves calculated for the basic set of model parameters with unc = 0.75u are compared with the results of the simple plug flow in Figure 3.8; the effect of different velocities is most pronounced for constant solubility. The existence of natural convection in a fixed bed where supercritical solvent flows slowly in the direction to the top was confirmed by Dams21 and Stüber et al.22, who dissolved in dense CO2 solid substances from porous particles. Dams observed the natural convection indicated by flattening of solid phase concentration profiles. Stüber et al. measured extraction curves for a wide range of extraction conditions; to simulate the results they applied the model for a differential bed and included the effects of gravity-opposing flow (upflow) and gravity-assisting flow (downflow) on extraction rate into correlations for external mass transfer coefficient.

3.7.4 CHANNELING

AND

SCALE UP

OF

EXTRACTOR

Another unfavorable type of flow pattern occurring in extraction beds is channeling. When the extraction bed is inhomogeneous, the solvent flows preferably through the paths of minimum hydraulic resistance and extracts the solute from their vicinity. Most of the solute in the extraction bed has very limited contact with the flowing solvent and as a result the extraction may even be retarded by orders of magnitude. With an increase in the cross section of the extractor the probability of inhomogeneous particle distribution and the tendency of channeling increase. This tendency depends also on the properties of extracted material; it is large for example when extracted particles are sticky or when they are too small and the extraction bed has therefore large hydraulic resistance. Berna et al.23 published experimental results on SC-CO2 extraction of orange peel essential oil from different cultivars and on

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 101 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

101

different scales, showing a large extraction retardation on production scale compared to the pilot scale. When a bed of inert particles of diatomaceous earth was placed in the extractor as solvent flow distributor, the difference between extraction yields measured in a small and a large extractor was significantly reduced. We have shown1 that the model for a differential extraction bed can simulate SFE in channeling extractor, when the external volumetric mass transfer coefficient kfa0 is adjusted to unrealistically low values, corresponding to the difficult solute approach from distant particles to flow channels. Prediction of scale-up effects is, however, difficult because there is no rule for predicting the extent of channeling in a larger-scale extraction bed. On the other hand, if the channeling is suppressed by proper measures, and flow pattern in the larger extractor is similar to that in the smaller extractor, the model with equilibrium and mass transfer parameters adjusted to the data measured in the smaller extractor is able to predict well the large-scale extraction.

3.8 PRELIMINARY EVALUATION OF EXTRACTION CURVES 3.8.1 SPLITTING EXTRACTION CURVE

INTO

TWO SECTIONS

Before the SFE is simulated using the complete model equations, model parameters should be roughly estimated from the shapes of extraction curves. The first condition for measurement of “readable” extraction curves is to suppress deviations from plug flow and particularly the channeling. This can be achieved relatively easily in a small-scale extractor, especially when the gravity-assisting solvent flow direction is chosen. In the case without solute-matrix interaction, the equilibrium- and diffusioncontrolled sections of extraction curves are obvious at first sight (Figure 3.3). In the case of solute-matrix interaction the differentiation is more difficult but possible especially when at least two extraction curves measured at identical conditions except for residence times are available. The extraction of easily accessible solute should be close to equilibrium extraction, which is possible due to excellent transport properties of supercritical solvents. Let us assume that the second experiment will be performed at identical solvent flow rate and with twofold solid charge in the extractor. (It is better to maintain the flow rate constant, as its change would affect not only the residence time but also the external mass transfer coefficient.) In our example, extraction yields e were calculated first for model parameters given in Table 3.1 and then for the same parameters except for a doubled residence time. The obtained extraction curves e(q) and e(t) are compared in Figure 3.9. The slopes of e(q) are in the initial, equilibrium-controlled extraction period practically equal to the equilibrium fluid phase concentration until about 11 g/(kg matrix) is extracted. The curves stop overlapping in the following, transition, period controlled by both equilibrium and intraparticle diffusion. Finally, when approximately 21 g/(kg matrix) has been extracted and the extraction is controlled almost solely by intraparticle diffusion, the shapes of both curves e(t) are almost identical, as can be proved by shifting the first extraction curve by a proper time interval as indicated in Figure 3.9. After identification of equilibrium-

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 102 Tuesday, July 18, 2006 8:19 AM

102

Functional Food Ingredients and Nutraceuticals 30

e, g/kg

20

10

0

0

10

20 q, kg/kg

30

40

30

e, g/kg

20

10

0

0

50

100

150 t, min

200

250

300

FIGURE 3.9 Estimation of model parameters from extraction curves e(q) and e(t) measured close to plug flow and with almost saturated solvent in the first extraction period. Solute matrix interaction (linear equilibrium) is assumed. (_______) simulated extraction curve for residence time tr = 4 min, model parameters according to Table 3.1; (- - - -) the same curve shifted by 36 min to the right; (– – – –) simulated extraction curve for residence time tr = 8 min (q· = 0.0625 min-1) and other physico-chemical parameters unchanged; (+ + + +) equilibrium model Equation 3.36.

and diffusion-controlled extraction curve sections, each section can be evaluated independently according to different relationships. It should be taken into account that the models simulate the extraction of one compound or one pseudocompound. When two or more solutes with completely different solubilities and comparable contents in botanic material are extracted and when the only experimental data available is the extraction curve measured for total extract, the curve should be simulated as a superposition of extraction curves for individual components and not as an extraction curve for a single compound.

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 103 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

103

3.8.2 EQUILIBRIUM-CONTROLLED PERIOD The equilibrium parameters are estimated by fitting approximate model equations to the first, equilibrium-controlled part, of the extraction curve. The equations given in Sections 3.4.3 and 3.6.3 as functions of dimensionless parameters can be rearranged for Θf → 0 to
e = qys = ys qt

(3.36)

for the case without solute-matrix interaction, and to
, y0 = e = qy0 = y0 qt

Kxu 1+ Kγ r

(3.37)

for solute-matrix interaction described by linear equilibrium. For combined equilibrium according to Equation 3.30, depicted in Figure 3.6, the shape of the first section of extraction curve depends on where on the equilibrium curve the extraction starts, as we discussed in Section 3.6.1. Equation 3.36 should be applied for the extraction starting in section “a,” Equation 3.37 for the extraction starting in “d,” and the following equation
for t ≤ t1 = e = qys = ys qt

r ( xu − xt ) − γKxt , q
( ys − Kxt )

e = q
 ys t1 + Kxt ( t − t1 )  for t > t1.

(3.38)

for the extraction starting in section “b.” From the slopes of the curve we can read equilibrium characteristics ys and Kxt, and from t1 we can estimate xt, after r is estimated from the diffusion-controlled part of the extraction curve. The equation for the extraction starting in section “c” is similar to Equation 3.38 but its initial slope e/q is lower than ys. More details on modeling the extraction with combined equilibrium are given elsewhere.1

3.8.3 DIFFUSION-CONTROLLED PERIOD The intraparticle diffusion time is much larger than the time constants for other phenomena affecting SFE and thus the diffusion characteristics can be obtained easily from the second extraction period data. Extraction yield in this period is according to Equation 3.25 and Equation 3.29 approximated by  1− r  t  for y + = Kxb1 . e = xu 1 − C1 exp  −   , C2 = t c for y + = ys , C2 = t c +
  C Kq 2   (3.39)

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 104 Tuesday, July 18, 2006 8:19 AM

104

Functional Food Ingredients and Nutraceuticals

The content of extractable solute in botanic material, xu, and constants C1, C2 are determined by fitting Equation 3.37 to the experimental extraction curve. The fact that the shape of the extraction curve in the second period is known has practical impact in analytical SFE as it enables the prediction of analyte content in extracted sample, Nxu, from several (at least three) measurements in the second period. If the amount of extract is determined in this period at three time points with the same time intervals, the initial amount of solute in the extraction bed is calculated after rearrangement of Equation 3.39 as

E∞ = Nxu = m1 +

 ( e2 − e1 )2  m22 = N  e1 +  2 e2 − e1 − e3  m2 − m 3 

(3.40)

where m1 = E(t1) is the amount of extract in the first measurement, m2 = E(t2) - E(t1) is the increase in extract in the first time interval and m3 = E(t3) - E(t2) is the increase in extract in the second interval. Nowadays, with computers, a direct fitting of Equation 3.39 to more than three experimental points is recommended instead of the three-point extrapolation, which is extremely sensitive to experimental error. To express the relation between intraparticle diffusion time (tc), effective intraparticle diffusivity (De), and particle size and shape, Reverchon24 used the results of Villermaux25 who examined approximate relationships between diffusion times and particles of different shapes and found the following relationship tc = µ

l2 De

(3.41)

where µ is equal to 3/5, 1/2, and 1/3 for spheres, cylinders, and slabs, respectively, and l is the characteristic particle dimension. Though the model applied by Reverchon and Villermaux does not take into account the existence of broken cells and easily accessible solute in particles, it can be easily applied to the second extraction period when the characteristic particle core dimension, lc, is substituted for l: tc = µ

lc2 De

(3.42)

As Equation 3.16 holds simultaneously, the solid phase mass transfer coefficient is approximately kc =

© 2007 by Taylor & Francis Group, LLC

De µlc

(3.43)

2441_C003.fm Page 105 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

105

Particles are usually assumed to be of spherical shape. Characteristic dimension for spherical particles of diameter d is equal to d/6 and their core characteristic dimension is d (1 − r ) 6

13

lc =

(3.44)

Inserting Equation 3.44 in Equation 3.42 we obtain the intraparticle diffusion time for spherical particles

tc =

d 2 (1 − r ) 60 De

2 3

(3.45)

Equation 3.45 written for r = 0 and modified for simultaneous external mass transfer resistance was applied in SFE from plant materials several times, for example, by del Valle et al.26 Reverchon24 demonstrates with experimental data for essential oil extraction from ground leaves of different particle sizes the importance of correct modeling of particle shape. The internal diffusion time was less dependent on particle size than according to Equation 3.45 derived for spherical particles. When it was taken into account that the extracted particles had a form of slabs whose characteristic dimension is less dependent on particle size and more on the slab thickness, which is not affected by grinding, a good agreement of model calculations with experimental extraction curves was achieved. Other published models are based on the assumption that the intraparticle mass transfer coefficient is related to solute diffusion in particle pores; mass balance equations then include the rate of solute transfer from solid phase to the pores or solid-fluid equilibrium at pore walls, and the diffusion in pores. For example, Goto et al.13 assumed adsorption equilibrium according to Equation 3.30 on the pore surface. Diffusion in the pores filled with solvent is also involved in shrinking core models (Roy et al.27 and Akgün et al.28). These models are based on the assumption that as the solute in spherical particle is depleted, its concentration in solid phase remains unchanged but the volume of the solid phase containing the solute, which is regarded as a particle core, is diminishing; the intraparticle mass transfer resistance increases as the solute has to diffuse through increasing length of pores between the shrinking core and particle surface. Shrinking core models do not require any knowledge of equilibrium equation y+ = y+(x) as the equilibrium solid phase concentration at a particle core surface remains constant. Whatever models are used to simulate intraparticle diffusion, all of them, after some simplification, lead to the equation for extraction yield in the form of Equation 3.39.

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 106 Tuesday, July 18, 2006 8:19 AM

106

Functional Food Ingredients and Nutraceuticals

3.9 CONCLUSIONS Supercritical extraction of botanic materials is characterized, in contrast to other extraction processes, by the combination of special properties of supercritical fluids (liquid-like density, very low kinematic viscosity, and good transport processes) and extracted plant materials (low permeability of intact dry plant tissue in contrast to the easily accessible solute in broken cells close to the particle surface). The process of supercritical fluid extraction can be split into two periods, the first one governed by phase equilibrium where solute is extracted from broken cells, and the second one controlled by intraparticle diffusion where the solute is extracted from intact plant tissue in the particle core. Using mathematical model with broken and intact cells, we have analyzed the effects of three phenomena — phase equilibrium, mass transfer, and flow pattern — on the extraction process. A simple experimental design enables us to distinguish the extraction periods and estimate phase equilibrium and mass transfer parameters separately. Solvent flow pattern significantly affects extraction yields. Scale up of the extraction process is usually connected with unfavorable changes in flow pattern, accompanied by a decrease in extractor performance. These changes are related to the inhomogeneous structure of the extraction bed formed by solid particles. Though the changes in flow pattern are not fully predictable, they can be largely suppressed when appropriate measures including the selection of solvent flow direction are taken.

ACKNOWLEDGMENT Financial support of project AS CR No. K4040110 is gratefully acknowledged.

GLOSSARY a0 c d Dax De e E h H j k K l M n N q q


Specific particle surface, m2/m3 Fluid phase concentration, kg/m3 Particle size, m Axial dispersion coefficient, m2/s Effective diffusivity in particle core, m2/s (=E/N) extraction yield, kg/(kg matrix) Extract, kg Axial coordinate, m Extraction bed height, m Flux across interface, kg/m3 s Mass transfer coefficient, m/s Distribution coefficient (equilibrium parameter) defined by Equation 3.27 Characteristic particle dimension, m Solvent in extraction bed, kg Number of mixers Insoluble solid phase (matrix), kg (=Q/N) specific solvent amount, kg/kg (= Q
/N) specific solvent flow rate, kg/kg s

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 107 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

Q Q
r t te tr T u V xb xc xt

107

Solvent passed through extraction bed, kg Solvent flow rate, kg/s Grinding efficiency (fraction of broken cells in plant tissue) Extraction time, s (=tr/Γ) equilibrium extraction time, s Mean residence time, s Temperature, K Interstitial velocity, m/s Extraction bed volume, m3 Concentration in broken cells, kg/(kg matrix) Mean concentration in intact cells (particle core), kg/(kg matrix) Transition concentration (equilibrium parameter) defined by Equation 3.30, kg/(kg matrix) Solute in untreated material, kg/(kg matrix) (=x/xu) dimensionless solid phase concentration Fluid phase concentration, kg/(kg solvent) Solubility (equilibrium parameter), kg/(kg solvent) Equilibrium fluid phase concentration, kg/(kg solvent) (y/y0) dimensionless fluid phase concentration Dimensionless axial coordinate

xu X y ys y+ Y z

Greek letters

γ Γ ε Θ ρ τ

Solvent-to-matrix ratio in extraction bed, kg/kg (=γy0/(rxu)) initial free solute ratio, kg/kg Void fraction Dimensionless mass transfer resistance defined by Equation 3.18 Density, (kg solvent or matrix)/(m3 fluid or solid phase) (= t/tr) dimensionless time

Subscripts 0 c f s

Initial conditions Particle core Fluid phase Solid phase

REFERENCES 1. Sovová, H., Mathematical model for supercritical fluid extraction of natural products and extraction curve evaluation, J. Supercrit. Fluids, 33, 35, 2005. 2. Sovová, H., Rate of the vegetable oil extraction with supercritical CO2: I. modelling of extraction curves, Chem. Eng. Sci., 49, 409, 1994. 3. Marrone, C. et al., Almond oil extraction by supercritical CO2: experiments and modelling, Chem. Eng. Sci., 53, 3711, 1998.

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 108 Tuesday, July 18, 2006 8:19 AM

108

Functional Food Ingredients and Nutraceuticals

4. Reverchon, E. et al., Supercritical fractional extraction of fennel seed oil and essential oil: experiments and mathematical modeling, Ind. Eng. Chem. Res., 38, 3069, 1999. 5. Reverchon, E. and Marrone, C., Modeling and simulation of the supercritical CO2 extraction of vegetable oils, J. Supercrit. Fluids, 19, 161, 2001. 6. Reis-Vasco, E.M.C. et al., Mathematical modelling and simulation of pennyroyal essential oil supercritical extraction, Chem. Eng. Sci., 55, 2917, 2000. 7. del Valle, J.M. and Aguilera, J.M., An improved equation for predicting the solubility of vegetable oils in supercritical CO2, Ind. Eng. Chem. Res., 27, 1551, 1988. 8. Angus, S., Armstrong, B., and de Reuck, K.M., International Thermodynamic Tables of Fluid State, CO2, Pergamon Press, Oxford, 1976, p. 37. 9. Lee, A.K.K., Bulley, N.R., and Meisen, A., Modelling of supercritical CO2 extraction of canola oilseed in fixed beds, J. Am. Oil Chem. Soc., 63, 921, 1986. 10. Brunner, G., Mass transfer from solid material in gas extraction, Ber. Bunsenges. Phys. Chem., 88, 887, 1984. 11. Ferreira, S.R.S. et al., Supercritical fluid extraction of black pepper (Piper nigrum L.) essential oil, J. Supercrit. Fluids, 14, 235, 1999. 12. Reverchon, E., Supercritical fluid extraction and fractionation of essential oils and related products, J. Supercrit. Fluids, 10, 1, 1997. 13. Goto, M., Sato, M., and Hirose, T., Extraction of peppermint oil by supercritical carbon dioxide, J. Chem. Eng. Japan, 26, 401, 1993. 14. Goto, M. et al., Modeling supercritical fluid extraction process involving solute-solid interaction, J. Chem. Eng. Japan, 31, 171, 1998. 15. Perrut, M., Clavier, J.Y., Poletto, M., and Reverchon, E., Mathematical Modeling of sunflower seed extraction by supercritical CO2, Ind. Eng. Chem. Res., 36, 430, 1997. 16. Goodarznia, I. and Eikani, M.H., Supercritical carbon dioxide extraction of essential oils: modeling and simulation, Chem. Eng. Sci., 53, 1387, 1998. 17. Beutler, H.J. et al., Einflu der Lösungsmittelführung auf den Hochdruck-ExtraktionsProze, Chem. Ing. Tech., 60, 773, 1988. 18. Beutler, H.J., Lenhard, U., and Lürken, F., Erfahrungen mit der CO2-Hochdruckextraktion auf dem Gebiet der Fettextraktion, Fat Sci. Technol., 90, 550, 1988. 19. Barton, P., Hughes, R.E., Jr., Hussein, M.M., Supercritical carbon dioxide extraction of peppermint and spearmint, J. Supercrit. Fluids, 5, 157, 1992. 20. Sovová, H., Kuera, J., and Jeû, J., Rate of vegetable oil extraction with supercritical CO2 : II. extraction of grape oil, Chem. Eng. Sci., 49, 415, 1994. 21. Dams, A., Stoffübergang bei der überkritischen Extraktion im Festbett, Chem. Ing. Tech., 61, 712, 1989. 22. Stüber, F. et al., Supercritical fluid extraction of packed beds: external mass transfer in upflow and downflow operation, Ind. Eng. Chem. Res., 35, 3618, 1996. 23. Berna, A. et al., Supercritical CO2 extraction of essential oil from orange peel: effect of the height of the bed, J. Supercrit. Fluids, 18, 224, 2000. 24. Reverchon E., Mathematical modeling of supercritical extraction of sage oil, AICHE J., 42, 1765, 1996. 25. Villermaux, J., Chemical engineering approach to dynamic modelling of linear chromatography, J. Chromatogr., 406, 11, 1987. 26. del Valle, J.M., Napolitano, P., and Fuentes, N., Estimation of relevant mass transfer parameters for the extraction of packed substrate beds using supercritical fluids, Ind. Eng. Chem. Res., 39, 4720, 2000. 27. Roy, B.C., Goto, M., and Hirose, T., Extraction of ginger oil with supercritical carbon dioxide: experiments and modeling, Ind. Eng. Chem. Res., 35, 607, 1996.

© 2007 by Taylor & Francis Group, LLC

2441_C003.fm Page 109 Tuesday, July 18, 2006 8:19 AM

Modeling of Supercritical Fluid Extraction of Bioactives from Plant Materials

109

28. Akgün, M., Akgün, N.A., and Dincer, S., Extraction and modeling of lavender flower essential oil using supercritical carbon dioxide, Ind. Eng. Chem. Res., 39, 473, 2000.

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 111 Tuesday, July 18, 2006 8:28 AM

4

Biochemical Reactions in Supercritical Fluids ˇ Zeljko Knez, Chiara G. Laudani, Maja Habulin, ˇ ˇ Mateja Primozic University of Maribor

CONTENTS 4.1 4.2 4.3

Introduction ..................................................................................................111 Enzyme Catalysis in Nonconventional Media ............................................114 Enzyme Catalysis in Supercritical Fluids (SCFs) .......................................117 4.3.1 Enzyme Stability in Supercritical Fluids.........................................118 4.3.2 Effect of Pressure.............................................................................119 4.3.3 Number of Pressurization-Depressurization Steps ..........................121 4.3.4 Effect of Temperature ......................................................................122 4.3.5 Effect of Water Activity ...................................................................122 4.3.6 Enzyme Reactors..............................................................................124 4.3.6.1 Process Schemes and Downstream Processing Schemes..........................................................124 4.3.7 Processing Costs ..............................................................................126 4.4 Conclusion....................................................................................................126 References..............................................................................................................127

4.1 INTRODUCTION Enzymatic catalysis has gained considerable attention in recent years as an efficient tool for synthesis of natural products, pharmaceuticals, fine chemicals, and food ingredients. The production of fine chemicals results in the generation of considerable volumes of waste, as the syntheses generally include a number of steps. The yield of each of these steps is usually 60% to 90%, but 10% is not unusual. Based on these data we can conclude that typically 1 kg of end product leads to the generation of 15 kg of wastes or more. Most generated wastes are solvents and by-products from solvents and intermediates. Therefore, ideally several reactions should be performed in water or in supercritical fluids.

111

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 112 Tuesday, July 18, 2006 8:28 AM

112

Functional Food Ingredients and Nutraceuticals

The high selectivity and mild reaction conditions of enzymatic transformations make them an alternative to the synthesis of complex bioactive compounds, which are often difficult to obtain by standard chemical routes. However, the majority of organic compounds are not highly soluble in water, which was traditionally perceived as the only suitable reaction medium for the application of biocatalysts. The realization that most enzymes can function perfectly well under nearly anhydrous conditions and, in addition, display a number of useful properties, for example, highly enhanced stability and different selectivity, has dramatically widened the scope of their application to organic synthesis. Another great attraction of using organic solvents rather than water as a reaction solvent is the ability to perform synthetic transformations with relatively inexpensive hydrolytic enzymes. Generally, in vivo, the synthetic and hydrolytic pathways are catalyzed by different enzymes. However, elimination of water from the reaction mixture enables the “reversal” of hydrolytic enzymes and thus avoids the use of the expensive cofactors or activated substrates that are required for their synthetic counterparts. Water is the most common solvent for biochemical reactions but in a biotechnological perspective, there are a lot of advantages of conducting enzymatic conversions in monophasic organic solvents as opposed to water.1 Some are listed below: • • • •

High solubility of most organic (nonpolar) compounds in nonaqueous media Ability to carry out new reactions impossible in water because of kinetic or thermodynamic restrictions Reduction of water-dependent side reactions Insolubility of enzymes in organic media, which allows their easy recovery and reuse

However, the use of organic solvents can be problematic because of their toxicity and flammability, and also there are increasing environmental concerns. As a result, supercritical fluids (SCFs) have attracted much attention in recent years as an alternative to organic solvents for enzymatic reactions. The present research activity will be focused on the development of selective methods for the production of polyfunctional molecules by enzymatic reactions in supercritical carbon dioxide (SC-CO2). Among all the possible SCFs, carbon dioxide is largely used. The use of SC-CO2 instead of organic solvents in biocatalysis presents several additional advantages. It performs mainly as a lipophilic solvent, nontoxic, nonflammable, and cheap. Supercritical CO2 has been most frequently used as a medium for biotrasformations: its critical pressure (73.8 MPa) is “acceptable” and temperature (31.1°C) is consistent with the use of enzymes and labile solutes. It has the GRAS (Generally Regarded As Safe) status. In addition, its “naturalness” is greatly appreciated by the food and healthcare related industries. The use of SCFs as solvents for enzymatic transformations is a relatively new area of research, which is expected to expand in the future. Close to the critical point, small changes in temperature or pressure can produce large changes in density and solvation ability of SCFs. This attribute of SCFs can be fruitfully exploited and integrated in biotransformation and downstream

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 113 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

113

processing steps in a single bioreactor. Beyond the critical point, both phases are indistinguishable, and the fluid is monophasic and occupies all the vessel volume. It can be described as a dense gas or an expanded liquid. Generally, SCFs exhibit liquid-like density and therefore have good solvating power, but they retain gas-like compressibility. Consequently, it is possible to control their solvating power by changing the pressure and/or temperature, with a continuous transition from a good to poor solvent. Moreover, low viscosity and high diffusion coefficients of these fluids enhance mass transport and reaction kinetics. These unique properties of SCFs enable one to design efficient integrated processes by coupling an enzymatic reaction with subsequent fractionation and product recovery steps. Molecules in the supercritical (SC) phase are not uniformly distributed in space, but the solvent molecules aggregate around the solute through solvent-solute intermolecular interactions forming clusters, where the aggregated molecules are in dynamic equilibrium with free solvent molecules. Thus, the solvation depends strongly on the density of the SCFs and differs from that in the liquid solution or gas phase.2 When catalytic reactions are performed in SCFs, the outcome of the reactions can be affected in a number of ways. In general, replacement of conventional liquid solvents by SCFs can increase the rate and tune the selectivity of reactions for the following reasons:3 •

•

•

Rapid diffusion of solutes or weakening of the solvation around reacting species facilitates the reactions and sometimes changes the reaction pathway Local clustering of solutes or solvents resulting in an appreciable increase in the local concentration of substrate (and catalyst) causes acceleration of the reaction Reduction or increase in the cage effect often affects the reaction performance of rapid chemical transformations such as radical reactions

In addition to these benefits, SCF chemistry, as a reaction media it poses economical, technical, environmental, and health advantages. The high volatility of CO2 allows it to be completely and easily removed from the product, resulting in an overall “solvent-free” reaction. By using SC-CO2 , an integrated production process can be performed, because it can act as solvent for the reaction and also as a separation medium. The variable solvating power of SC-CO2 and other SCFs facilitates the integration of biocatalytic and downstream processing steps in a single robust bioreactor. The main drawback of SC-CO2 is that it has limited solvating power with respect to polar compounds. This is a serious limitation for biotechnological applications where most natural molecules of interest (e.g., alkaloids, carotenoids, phenols, proteins, and sugars) are only sparingly soluble in SC-CO2. In this case, a polar cosolvent or so-called entrainer, such as acetone, ethanol, methanol, or water, is added in order to increase the polarity of the medium and to solubilize the target solute via the formation of hydrogen bonds. Typically, cosolvents are added to the SCF at moderate concentrations of less than 10 mol %.4 In a batch reactor, the cosolvent can be added directly into the reactor prior to pressurization, whereas in a continuous process, the addition should be made to the CO2 inflow via a liquid pump to deliver a constant flow rate at the operating pressure. However, the use of

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 114 Tuesday, July 18, 2006 8:28 AM

114

Functional Food Ingredients and Nutraceuticals

another component in the system further increases the complexity and may also complicate downstream processing. Moreover, the solubility enhancement effect of a cosolvent is usually limited in the case of very polar compounds. Two alternative methods have been developed for some specific cases. To solubilize polyols (e.g., glycerol and sugars) has been proposed5 to form a hydrophobic complex between the polyol and phenylboronic acid (PBAC), which is much more soluble in the SC phase. This method was used to perform the esterification of glycerol and sugar with oleic acid in SC-CO2. Another method involves the adsorption of polar substrates onto a solid hydrophilic support such as silica gel. Compared with the former, this approach is more general because it is not necessary to have two vicinal hydroxyl groups in the substrate molecule.5 In addition, recent advances in the understanding of the chemical properties of materials that are soluble in CO2 have permitted the development of novel surfactants that allow dissolution of both hydrophilic and hydrophobic materials in CO2. The use of SCFs decreases the mass transfer limitations because of the high diffusivity of reactants in the SC medium, the low surface tension, and because of the relatively low viscosity of the mixture. The Schmidt number, Sc = η/ρ⋅D (where η is dynamic viscosity, D is diffusivity, and ρ is density), for CO2 at 200 bars is 45 times lower than for water at 1 bar and 20°C. High diffusivity of SCF and low surface tension lead to reduced internal mass transfer limitations for heterogeneous chemical or biochemical catalysis. One of the main advantages of the use of dense gases as a solvent for enzymecatalyzed reactions is the simple downstream processing. The physico-chemical properties of dense gases are determined by their pressure and temperature, and are especially sensitive near their critical point. By reducing the solvent-power of a dense gas in several stages, fractionation of the product and unreacted reactants is possible. Fractionation is also possible by extracting the mixture, usually with the same dense gas as used in the reaction, but under different process conditions. In all downstream processing schemes, various particle formation techniques or chromatographic techniques can be integrated.

4.2 ENZYME CATALYSIS IN NONCONVENTIONAL MEDIA The breakthrough of biocatalysis in nonaqueous media started in the early 1980s.6–10 Nowadays, it is well established that many enzymes, such as lipases, can remain active and stable in pure organic solvents. Changing the hydrophobicity of predominantly aqueous media through addition of organic solvents causes the hydrophobic effects, which keep the hydrophobic residues buried in the core of the proteins and folded in an aqueous environment where enzymes are kinetically trapped in their native structure in organic solvents.11–14 In addition, in the media of low water content, enzyme inactivation, caused by hydrolysis of peptide bonds and deamidation of asparagine and glutamine residues, is reduced.11 Indeed, many times enzymes are more stable in organic solvents than in water. For instance, increased thermal stability in dry organic solvents with substantial increase in enzyme half-life at 100°C

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 115 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

115

TABLE 4.1 Advantages of Biocatalysis in Nonaqueous Media Increased substrate solubility Simplified recovery of biocatalyst Shift to synthetic reactions Mild reaction conditions and minimization of side-reactions Environmentally benign catalyst High selectivity Simplified work-up of products Avoids microbial contaminations

compared to water has been observed.9,15 There are a number of advantages of using enzymes in organic solvents. These are listed in Table 4.1. Many substrates that are insoluble in water can be dissolved by organic solvents. Enzymes are often insoluble, an advantage that simplifies their recovery and reuse many times and is economically very important. A change from an aqueous environment can favor a shift in the equilibrium, enabling synthetic reactions to be achieved with hydrolytic enzymes. Moreover, the unique selectivity and activity of enzymatic reactions is achieved under mild reaction conditions. The enzyme instability in harsher reaction conditions is common in industrial processes, which can, of course, be a disadvantage. The search for enzymes from various extremophilic organisms may, however, result in biocatalysts that can withstand more extreme conditions.16,17 Furthermore, enzymes are environmentally benign and are, unlike metal catalysts, completely degraded in nature. Although some enzymes do display perfect selectivity, many others can receive a broad range of unnatural substrates still with high chemo-, regio- or enantioselectivity. Additionally, the use of an organic solvent often simplifies work-up procedures and avoids microbial contamination of the reaction medium.11,18 The conventional notion that enzymes are only active in aqueous media has long been discarded, thanks to the numerous studies documenting enzyme activities in nonaqueous media, including pure organic solvents and supercritical fluids (SCFs). Enzymatic reactions in nonaqueous solvents offer new possibilities for producing useful chemicals (emulsifiers, surfactants, wax esters, chiral drug molecules, biopolymers, peptides and proteins, modified fats and oils, structured lipids, and flavor esters). According to conventional notion, enzymes are active only in water. Historically, enzymatic catalysis has been carried out primarily in aqueous systems. Although water is a poor solvent for preparative organic chemistry, it is the unique specificity of enzymes that drew the interest of chemists who were seeking highly selective catalytic agents. Experiments to place enzymes in systems other than aqueous media date back to the end of the nineteenth century.19–23 Initial studies considered the addition of small quantities of water-miscible organic solvents like ethanol or acetone to aqueous enzyme solutions ensuring high availability of water to retain the catalytic activity of enzymes. Then, the biphasic mixtures (aqueous

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 116 Tuesday, July 18, 2006 8:28 AM

116

Functional Food Ingredients and Nutraceuticals

enzyme solution emulsified in a water-immiscible solvent such as isooctane or heptane) were used, in which the substrates from the organic phase diffuse into the aqueous phase, that undergoes enzymatic reaction and the products diffuse back. The size of water droplets may be reduced to facilitate mass transfer resulting in the formation of microemulsions or reverse micelles, whose stabilization is achieved by adding surfactants.24,25 Developments in using enzymes in nearly nonaqueous solvents containing traces of water have stimulated research in achieving various kinds of enzymatic transformations.26–34 Enzymatic reactions in nonaqueous solvents offer numerous possibilities for the biotechnological production of useful chemicals using reactions that are not feasible in aqueous media. These reactions include chiral synthesis or resolution;35–38 production of high-value pharmaceutical substances;39–43 modification of fats and oils;44 synthesis of flavor esters and food additives;45–51 production of biodegradable polymers,52 peptides, proteins, and sugar-based polymers.53 In nonaqueous solvents, hydrolytic enzymes could undergo synthetic reactions while they also exhibit altered selectivities,11 pH memory,10,15,54 increased activity and stability at elevated temperatures,9,55 regio-, enantio- and stereoselectivity,44,56 and may also be affected by their water activity.57 Currently, there is a considerable interest in the use of enzymes (particularly lipases, esterases, and proteinases) as catalysts in organic synthesis.18,44,56,58–62 Five major technological advances are believed to have significantly influenced the industry for adopting enzymatic biotransformations:63 (1) the development of large-scale downstream processing techniques for the release of intracellular enzymes from microorganisms; (2) improved screening methods for novel biocatalysts;64–68 (3) the development of immobilized enzymes; (4) biocatalysis in organic media; and most recently, (5) recombinant-DNA (r-DNA) technology to produce enzymes at a reasonable cost. There seems to be no agreement as to why the biocatalysis in organic media could not have taken off earlier.69–71 Perhaps the traditional belief that most enzymes are incompatible with most organic syntheses in nonaqueous media, which also poses a psychological hurdle. Also, until recently, there was no demand for enantiopure compounds, and hence, no enzymes needed to be used. The establishment of industrial processes,72,73 and the realization that most enzymes can function well in organic solvents,9,10,74,75 have heightened interest in the use of enzymes. Also, the need for enantiomerically pure drugs is driving the demand for enzymatic processes. This combined with the discovery of strikingly new properties of enzymes in organic solvents has led to the establishment of organic phase enzyme processes in industry.56,60 Enzymes occupy a unique position in synthetic chemistry due to their high selectivities and rapid catalysis under ambient reaction conditions. Nevertheless, synthetic chemists have been reluctant to employ enzymes as catalysts, because most organic compounds are water insoluble, and the removal of water is tedious and expensive. The fact that enzymes are stable, and in some cases, improve their high specificity in near anhydrous media, has dramatically changed the prospects of employing enzymes in synthetic organic chemistry. The problems that arise for most biotransformations are low solubility of reactants and products, and limited stability of biocatalysts. Carrying out reactions in an aqueous-organic two-phase system

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 117 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

117

would be a solution to overcome the first problem. This is not always possible due to the limited stability of enzymes at liquid-liquid interface or in organic solvents. Hence, other approaches are necessary. These include addition of complexing agents such as dimethylated cyclodextrins or adsorbing materials like Amberlite™ XAD7 resins (Eli Lilly, Indianapolis, IN, USA), use of membrane-stabilized interface (Sepracor™, Marlborough, MA, USA) and continuous extraction of reaction products (Forshungs-zentrum-Julich, Julich, Germany). The catalyst’s stability can be increased using a variety of methods including the addition of antioxidants (e.g., dithiothreitol), immobilization, cross-linking, separation from deactivating reagents, variation of reaction conditions, and by genetic engineering.76 Solvent systems used as the reaction media for enzymatic catalysis may be categorized as: (1) aqueous; (2) water: water-miscible (monophasic aqueous-organic system); (3) water: water-immiscible (biphasic aqueous-organic system); (4) nonaqueous (monophasic organic system); (5) anhydrous; (6) supercritical fluids; (7) reversed micelles; (8) solvent-free systems; (9) gas phase; and (10) ionic liquids.

4.3 ENZYME CATALYSIS IN SUPERCRITICAL FLUIDS (SCFs) During the last decade, the tremendous potential of enzymes as a practical catalyst for chemical processes in nonaqueous environments has been well recognized. The use of biocatalysts in organic solvents offers many advantages over using pure water. Among these media, SCFs, such as carbon dioxide (SC-CO2), exhibit properties similar to organic solvents, but with the additional capacity of enhancing transport phenomena (due to their high diffusivities) and facilitate reaction products separation by tuning solvent power, which makes them more attractive to be used as “greendesigner” solvents.77,78 The interest of using biocatalysts in SC-CO2 has been growing rapidly in recent years, mainly in industrial and pilot plant applications.79 The growing interest in SCF technology results from the attractive possibilities offered by this technique: processing at moderate, usually ambient temperatures, use of nontoxic, nonflammable, and environmentally acceptable solvents (usually pressurized CO2) and easy-to-change solvent power (not possible with conventional organics). For example, SCFs at temperatures and pressures slightly above the critical points (e.g., 31°C and 73.8 MPa for CO2), exhibit unique combined properties: liquid-like density (and hence solvent power) and high compressibility, very low viscosity, and high diffusivity. The first two properties make the solvent power controllable by changing pressure and/or temperature, while low viscosity and high diffusivity markedly enhance mass transport phenomena and hence kinetics of a process. However, enzymes are not soluble in SCFs, therefore enzymatic catalysis in SCFs would always be heterogeneous. The use of enzymes as catalysts in nonaqueous media has been described in scientific literature since the middle of the 1980s.10 Organic media offer certain advantages over aqueous media: stabilization of enzymes, dissolution of hydrophobic compounds, and the feasibility of shifting thermodynamic equilibrium toward the synthesis of esters and amides (e.g., in the case of hydrolytic enzymes). Certain

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 118 Tuesday, July 18, 2006 8:28 AM

118

Functional Food Ingredients and Nutraceuticals

SCFs, such as CO2, may prove an interesting alternative inasmuch as they exhibit properties similar to organic solvents and their solvent power, which is dependent on the specific density (and hence easily controllable by regulation of pressure and temperature), may be advantageously used in the recovery of products. This opens the way for an integrated production of product recovery processes. The use of SCFs as media for enzymatic reaction was first proposed in the middle of the 1980s by Randolph et al.,80 Hammond et al.,81 and Nakamura et al.82 The aim of the initial studies was to demonstrate that enzymes are active in SCCO2. Randolph et al.80 showed that alkaline phosphatase and cholesterol oxidase are active in SC-CO2; Hammond et al.81 demonstrated the same for polyphenyl oxidase; and Nakamura et al.82 carried out successful transesterification reactions using lipase. Successively, Randolph et al.83 examined the effect of aggregation of cholesterol on cholesterol oxidase activity and found that the addition of cosolvents (entrainers) promotes aggregation and thus resulted in an increase in reaction rate proportional to the degree of aggregation. Recently, the benefit of using SCFs for enzymatic reactions has been demonstrated by Mori and Okahata84 and Kamat et al.85,86 and Chaudhary et al.,87 for example, for improved reaction rates, control of selectivities by pressure, and so forth. Some examples of enzymatic reactions are shown in Figure 4.1 and Figure 4.2. Most of the research published to date dealt with two problems: (1) conformation and stability in the supercritical environment (mainly CO2) and the effect of pressure on reaction rate; (2) the effect of water/moisture content on the activity of enzymes.

4.3.1 ENZYME STABILITY

IN

SUPERCRITICAL FLUIDS

The use of enzymes in supercritical fluids is full of potential problems. For example, the number of parameters that influences the stability of the enzymes increases dramatically when using supercritical fluid. This is the reason why, up to now, no prediction could be made on whether enzymes are stable under supercritical conditions or if the equivalent of even higher activity and selectivity is available compared OH O

.

OH O

OH O OEt

Pseudomonas SP

OH

OEt +

SC CO2

FIGURE 4.1 Reaction scheme of the enantioselective hydrolysis of HPAE in SC-CO2.88

R R1

+ H

O

O

OH C

CH3 C

O

C

CH3

Pseudomonas SP SC CO 2

+

C R

1

H

OH

R

OH

R

C R

1

+ CH3

O C

OH

H

FIGURE 4.2 Reaction scheme of the enzymatic esterification of secondary alcohols in SC-CO2.89

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 119 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

119

with the reactions in organic solvents. In the following sections the influence of parameters on the enzyme stability will be discussed. Early investigations80–82 demonstrated that certain enzymes are active in SCCO2. Randolph et al.83 first studied the conformation of several spin-labeled variants of cholesterol oxidase in SC-CO2 and concluded that these proteins were not influenced by the SC-CO2 environment; a similar result has recently been obtained for lipase.90 In contrast, Kasche et al.91 reported that α-chymotrypsin, trypsin, and penicillin amidase were partially denaturated by SC-CO2 and suggested that the decompression process led to their denaturation, but no in situ measurements were conducted to substantiate this suggestion. Most recently Zagrobelny and Bright92 carried out a more detailed examination of the same problem. The conformation of trypsin in SC-CO2 was studied at the pressure range of 5 to 25 MPa and monitored the conformational changes of trypsin in situ as a function of pressure. Their results clearly demonstrate that: (1) significant changes in protein conformation can be induced by SC solvents; (2) most of the conformational changes occur during compression; (3) the native trypsin conformation is only slightly more stable than the unfolded form. Performance characteristics, specific rates of conversion, and yield factors are essential for rating any technological process. On the whole, enzymatic reactions in SC-CO2 proceed at rates similar to those of organic solvents such as n-hexane93,94 and cyclohexane.90 As well as ensuring similar rates of processing and enzyme stability, the SC technology offers important advantages over organic solvent technology, such as ecological friendliness and product fractionation, which can easily be linked with direct micronization and crystallization from SC-CO2 by fluid expansion. In addition, CO2 does not usually oxidize substrates and products, allowing the process to be operated at a temperature of 40°C. Although many enzymes are stable in SCFs, one should pay considerable attention to finding the correct reaction conditions for each substrate/enzyme/SCF system. Although successful reactions have been reported with Subtilisin Carlsberg protease and Candida lipases in SCCO2, there is also evidence for their instability85,95,96 or the existence of a narrow pressure range of activity.97,98 These enzymes are fairly stable in other SCFs such as fluoroform, ethylene, ethane, propane, and sulphur hexafluoride.85 Immobilized Mucor miehei lipase appears to be very stable in SC-CO2; it is a monomeric enzyme with three stabilizing disulfide bonds,99 which may play a role in maintaining its activity in SC-CO2. Nevertheless, one of the most important advantages of using the SCFs as reaction media is that upstream processing after reaction can be greatly simplified as the technique is easily combined with other unit operations.

4.3.2 EFFECT

OF

PRESSURE

The influence of the system pressure on the stability of enzymes is not so significant up to 30 MPa. This is of great advantage because it means that the solvent-power of the supercritical fluid can be adjusted for reaction performance. On the one hand, the solubility of substances increases with higher pressures because of a higher fluid density and this is essential to bring the initial products in the reactor and remove the end products from the reactor. On the other hand, higher pressure normally

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 120 Tuesday, July 18, 2006 8:28 AM

120

Functional Food Ingredients and Nutraceuticals

results in higher reaction rates. Therefore a pressure increase is in most cases positive for enzymatic reactions. An isothermal change of pressure of SCFs may change the reaction rate by changing either solubility or the reaction rate constant. A pressure increase improves solubility, thereby increasing production rates, and this effect is most pronounced in the near-critical region. Certain enzymes show considerable apparent pressure activation.100 Pressure can modify the catalytic behavior of enzymes by changing, for example, the rate-limiting step101 or modulating the selectivity of the enzyme.102 If an enzyme is stable in a supercritical fluid, its stability is usually not influenced by pressure ranges of up to 30 MPa. Conversely, reaction rates may be influenced by pressure. In most cases a pressure increase acts positively for enzymatic reactions or there are no changes in the reaction rates. Pressure-induced deactivation of enzymes takes place mostly at pressure exceeding 150 MPa. Reversible pressure denaturation mostly occurs at pressures below 300 MPa, and higher pressures are required to cause irreversible denaturation.103 The effect of pressure on the reaction rate constant has not yet been determined, but the effect on the overall production rate has been examined in several papers. Erickson et al.104 carried out transesterification of triglycerides, using lipase from Rhizopus arrhizius. The reactants were trilaurin and palmitic acid and the pressure ranged from 10 to 30 MPa. A strong negative effect of pressure increased the rate of palmitic acid incorporation into triglyceride was detected, especially in the nearcritical region. The interesterification of trilaurin and myristic acid, catalyzed by lipase, was investigated by Miller et al.90 in the pressure range of 6 to 11 MPa. The interesterification rate and the overall rate (based on total trilaurin conversion) increased with increase in pressure; however, the interesterification rate increased much more rapidly than the overall rate, indicating that the selectivity of the reaction for interesterification over hydrolysis improved at higher pressures. The operational stability of enzymes in SC-CO2 is of crucial importance from the point of view of application. Miller et al.90 measured the interesterification rate over 80 h of continuous operation and observed no loss of activity of lipase. Cholesterol oxidase is stable at 10 MPa and 35°C for at least 50 h.83 Pressure has also been found to have little effect on the stability of lipase from Mucor miehei in the range from 13 to 18 MPa, causing only 10% loss of activity93 after six days at 40°C, unlike temperature, which contributed to a 20% loss at 60°C. Additionally, in some cases a negative effect of pressure on the catalytic activity of biocatalysts in compressed gases may be observed. The catalytic efficiency of subtilisin Carlsberg suspended in compressed propane, near-critical ethane, near-critical carbon dioxide, and tert-amyl alcohol at constant temperature and pressure up to 30 MPa and fixed enzyme hydration was lowered.105 In near-critical fluids an increase in pressure of only 20 MPa caused a sixfold decrease in the catalytic efficiency of subtilisin in CO2. In SC-CO2 the formation of carbamates are responsible for lower enzymatic activity in this medium. Carbamates are the product of the reaction between basic free amino groups in the enzymes and acidic SC-CO2.95 On the other hand, lysozyme lipase unfolded and partially oligomerized in moist SC-CO2 at 80°C and its denaturation was not caused by interaction with SC-CO2 but by heating the protein in the presence of water, as found by Weder.106

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 121 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

121

One of the advantages of using SCFs as enzymatic reaction media is separation of products from the reaction mixture by changing the pressure of the supercritical fluid. With respect to the facts mentioned previously, the solvent power of the SCF can be adjusted for running reactions, and products can be easily removed from the reactor. When the lipid-coated lipase was employed in SC-CHF3 the enzyme activity107 could be switched on and off by adjusting the pressure or temperature of the SC-CHF3. The effect of pressure on the extent of conversion and product composition of the enzyme-catalyzed hydrolysis of canola oil in SC-CO2 was investigated using lipase from Mucor miehei immobilized on macroporous anionic resin.108 A conversion of 63% to 67% (triglyceride disappearance) was obtained at 24 to 38 MPa. Monoglyceride production was favored at 24 MPa. The amount of product obtained was higher at 24 to 38 MPa due to its enhanced solubility in SC-CO2.

4.3.3 NUMBER

OF

PRESSURIZATION-DEPRESSURIZATION STEPS

The influence of pressurization-depressurization steps in batch reactors on the enzyme activity is of importance to many researchers.109 Pressurizing an enzyme does not play an important role, but depressurization is usually the step that influences residual enzyme activity. The supercritical fluid permeates through the enzymes by diffusion, a process that is relatively slow. After a certain time the enzymes are saturated with the SCFs. If the expansion is too fast and the fluid diffuses out of the enzyme slowly, this causes a higher fluid pressure in the enzyme than in the system. The pressure difference results in cell-cracking, where the cell membranes are broken by the resulting pressure inside the cells. This causes the unfolding of the enzymes and the loss in tertiary structure, which is required for their activity and selectivity. Experiments have shown that depressurization from supercritical CO2 conditions is much smoother than entering the two-phase region and expanding the gas of the fluid.110 This is caused by a change in density which is continuous in supercritical conditions. The expanding liquid CO2 causes evaporation of the fluid accompanied by a large change in density, and this volume expansion causes the unfolding of the enzyme. Depressurization is important when using the benefit of SCFs for simple downstream processing. In this case, by operating a cascade of depressurizations (with a possible change in temperature), product fractionation can be achieved.111 In order to have successful industrial applications, enzymes as biocatalysts must retain their activity for a considerable period of time. The activity of the lipase from Candida antarctica for the production of isoamyl acetate in SC-CO2 was studied by Romero et al.111 in a tubular reactor. The yield of isoamyl acetate was 100% for 30 days, and then slowly decreased. Habulin et al.112 found similar results with immobilized Rhizomucor miehie lipase, reporting a 4% decrease in conversion after one month of treatment. Cholesterol oxidase from different sources can exhibit different stabilities in SCCO2.83 The cholesterol oxidase from Gloecysticum retained its activity for three days and the one from Streptomyces sp. for only one hour. In some cases the half-life of

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 122 Tuesday, July 18, 2006 8:28 AM

122

Functional Food Ingredients and Nutraceuticals

the biocatalysts can be increased by adjusting pressure.113 The half-life of α-chymotrypsin increased when the pressure increased from 8 to 15 MPa.

4.3.4 EFFECT

OF

TEMPERATURE

The effect of temperature is much more significant than the pressure effect. For enzyme stability, a temperature increase above certain levels, depending on the enzyme sources, results in the deactivation of the enzyme. This limitation places the temperature limits on the extraction process. Normally, at higher pressure levels an increase in temperature results in higher solubilities of substances in SCFs because the increase in the vapor pressure of the compounds to be dissolved overcomes the reduction in density. Reaction rate also increases at higher temperatures, although enzyme deactivation at higher temperatures may occur. At the moment no correlation between temperature and the stability of the different types of enzymes is available. Lipase from Aspergillus niger was incubated in SC-CO2 at 30 MPa and different temperatures.114 Its residual activity was optimal at 323K. At higher temperatures a rapid decrease in activity was observed. This thermal deactivation was connected to changes in the water distribution in the system. In microaqueous media, including SCFs, thermal stability of biocatalysts can also be improved. Reaction rate for subtilisin protease-catalyzed reactions increased by 80% in SC-CO2.115 The optimal temperature for esterification between n-butyric acid and isoamyl alcohol, catalyzed by porcine pancreas lipase, increased from 313K at atmospheric pressure to 323K in near-critical propane.116 The improved stability of the lipase in the low-water-content environment is a consequence of a well-known fact that reactions, which may be responsible for the denaturation of enzymes, are hydrolytic and therefore require water.117 Thermal activation and deactivation (energy of activation and deactivation enthalpy, respectively) may be determined from the Arrhenius diagram. The ratio between the amounts of inactive and active forms of the enzyme at a temperature, where the greatest initial reaction rate is observed, is expressed with a deactivation constant. If the activation and deactivation enthalpy values are high this indicates that enzyme activity is very dependent on temperature.

4.3.5 EFFECT

OF

WATER ACTIVITY

Water concentration in the reaction system is one of the most important factors that influence activity of an enzyme. Water is crucial to enzymes and affects enzyme action in various ways: by influencing enzyme structure via noncovalent binding and disruption of the hydrogen bonds; facilitating reagent diffusion; and influencing the reaction equilibrium.118 The subtilisin Carlsberg catalyzed transesterification of N-acetyl phenylalanine methyl ester,119 N-acetyl phenylalanine ethyl ester,120 N-trifluoroacetyl phenylalanine methyl ester,121 and N-trifluoroacetyl phenylalanine ethyl ester122 was studied in SC-CO2. The water content of the reaction affected the reactivity of the system; for

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 123 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

123

the transesterification of the methyl esters with ethanol the optimum concentration of water was determined to be about 0.74 M, while during the transesterification of the ethyl esters123 with methanol it was about 1.3 M. Use of an enzyme in pure SC-CO2 may lead to removal of the water, which is included or bonded to the enzyme. The quantity of the removed water is temperature and pressure depended and if too high may lead to enzyme denaturation and loss of activity. The solubility of water in SC-CO2 can be calculated by Chrastil equation124 a  c = ρk ⋅ exp  + b  T  where c is the solubility (g/L), ρ is the CO2 density (g/L), and T the temperature (K). The calculated water parameters are k = 1.549, a = −2826.4, and b = –0.807. Water is important in the supercritical fluid because water-saturated CO2 causes the inhibition of enzymes and consequent loss of activity. The optimal water concentration has to be determined for each enzyme separately. Enzymes require a specific amount of water to maintain their active conformation. Enzyme stability generally decreases with increasing water concentration, whereas their activities require some water to be present. Therefore, the water content has to be optimized to find the best balance between enzyme life and activity. If water acts as a substrate for an enzymatic reaction, the optimal parameters for continuous reaction require among other things enough moisture to compensate for complete reaction and sufficient enzyme moisture for losses due to water solubility in SC-CO2.125,126 However, if the water concentration in the supercritical medium is too high or if it is a product of the reaction, the increase in moisture may cause enzyme deactivation. Not surprisingly, all studies published so far have pointed out the strong influence of moisture on enzymatic activity and reaction rates. Optimum water content in the support was estimated at 10 wt %, irrespective of the operating conditions,94 but this value may be contested in view of other reports.127 To prevent dehydration of the enzyme, the fluid in contact with the protein must contain water. The most hydrophilic hydrocarbons (e.g., hexane) dissolve 0.01% water, but SC-CO2 may dissolve as much as 0.3% to 0.35% water. However, it is not the solubility of water itself but the partition of water between the enzymatic support and the solvent (SC-CO2), which matters. Marty et al.94 carried out an extensive analysis of the partition of water between the enzymatic support and SC-CO2 as a function of pressure and temperature. They found that increasing temperature and pressure had a negative effect on the adsorption of water by the support. This is opposite to the results obtained by gas adsorption, which suggests that the solvation effects predominate over the vapor pressure effects. The same authors also extensively studied the influence of ethanol (entrainer) content in SC-CO2 and found that ethanol has a strong “drying” effect on the enzyme support; indeed, the more hydrophilic the fluid, the more pronounced is the dehydration of the enzymes. Increasing water content, above the optimum level, adversely affects the overall performance. This appears to be related to hydrophilic hindrance of the hydrophobic substrate on its way to the

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 124 Tuesday, July 18, 2006 8:28 AM

124

Functional Food Ingredients and Nutraceuticals

active sites on the enzyme, and eventually makes the thermodynamic equilibrium less favorable.128 Chulalaksananukul et al.129 measured the residual activity of lipase from Mucor miehei after a day in SC-CO2 at a temperature range of 40°C to 100°C and at various water concentrations. As the temperature increased, the enzyme molecule at first unfolded reversibly and then underwent one or more of the following reactions: formation of incorrect or scrambled structures, cleavage of disulfide bonds, deamination of trypsine residues, and hydrolysis of peptide bonds. Each process requires water and is therefore accelerated with increasing water concentration.

4.3.6 ENZYME REACTORS 4.3.6.1 Process Schemes and Downstream Processing Schemes Batch-stirred tank, extractive semibatch, recirculating batch, semicontinuous flow, continuous packed-bed, and continuous-membrane reactors have used dense gases as solvents. 4.3.6.1.1 Batch-Stirred-Tank Reactor Batch-stirred-tank reactors are usually used for screening enzymatic reactions in dense gases.130,131 The design of the system is shown in Figure 4.3. Initially, the substrates are pumped into the reactor and then the enzyme-preparation is added. Finally, dry gas is pumped into the reactor, up to the desired pressure. The initial concentration of the reactant must never exceed its solubility limit in the gas.

T = const.

PI

P

2 Gas 1

FIGURE 4.3 Design of experimental batch-stirred-tank apparatus for synthesis under high pressure. 1 magnetic stirrer and heater; 2 reactor; P high-pressure pump; PI pressure indicator.

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 125 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

125

PI

PI

HPP

PI

TIR

TIR H

Gas

3

FI

HPP HPP 4

1

2

FIGURE 4.4 Design of experimental continuous packed-bed apparatus for synthesis under supercritical conditions. 1, 2 substrates; 3 reactor; 4 separation column; HPP high-pressure pump; PI pressure indicator; TIR temperature indicator and regulator; H heat exchanger; FI flow indicator.

4.3.6.1.2 Continuous Packed-Bed Reactor A continuous packed-bed reactor is presented in Figure 4.4. The pump delivers a high-pressure gas into the system. The substrates are pumped into the system, using a liquid pump. The initial concentration of the reactant must never exceed its solubility limit in the gas. Supercritical CO2 is depressurized through the expansion valves into separator column 4, where the product and the unreacted substrates are recovered. The gas phase is finally vented into the atmosphere after measuring the flow rate with a rotameter. The gas can be condensed and recycled when used on a pilot or industrial scale. 4.3.6.1.3 Continuous High-Pressure Enzyme Membrane Reactor In a continuous high-pressure enzyme membrane reactor (Figure 4.5), a 35-mmdiameter membrane is placed between two sintered plates and fitted in the reactor. A measured amount of the catalyst (hydrated enzyme preparation) is put in the reactor, which is electrically heated, with a heating jacket, to constant temperature. The substrates and the gas are pumped into the membrane reactor with the highpressure pump. The products and unreacted reactants are collected in the separator. The catalyst remains in the reactor (behind the membrane). Continuous processes in particular are favored in the industry because they are cost efficient and the reactors can be kept relatively smaller in size.132,133 This reduction in size reduces both costs and safety problems of the high-pressure equipment needed for SC reactions.

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 126 Tuesday, July 18, 2006 8:28 AM

126

Functional Food Ingredients and Nutraceuticals

TIR PI

P

PI

H

4

FI

TIR

Gas P

P

3 2

1

1

FIGURE 4.5 High-pressure continuous stirred-tank membrane reactor. 1 substrates; 2 magnetic stirrer and heater; 3 reactor with membrane; 4 separation column; P high-pressure pump; TIR temperature indicator and regulator; PI pressure indicator; FI flow indicator.

4.3.7 PROCESSING COSTS Economic evaluations were made for the enzyme-catalyzed production of oleyl oleate in a high-pressure batch-stirred-tank reactor (HP BSTR) and in a high-pressure continuously operated reactor (HP COR). It was assumed that the reaction mixture was completely precipitated from CO2. The production costs are strongly dependent on the solubility of the substrates in dense gas, the enzyme lifetime, and the productivity of the biocatalyst. The study shows that the return on investment would take less than one year.109

4.4 CONCLUSION The application of SCFs as reaction media for enzymatic synthesis has several advantages, such as the higher initial reaction rates, higher conversion, possible separation of products from unreacted substrates, over solvent-free, or solvent systems (where water or organic solvents are used). Owing to the lower mass transfer limitations and mild (temperature) reaction conditions, at first the reactions which were performed in nonaqueous systems will be transposed to supercritical media. An additional benefit of using SCFs as reaction media is that they give simple and ecologically safe (no heat and solvent pollution) recovery of products. However, for some specific reactions, solvent-free systems are preferred because of their higher yields. The main area of development should be related to the hydrolysis of glycerides, transesterification, esterification, and interesterification reactions. As lipases have high and stable activity in SC-CO2 (even at high temperatures) even greater and more intensive developments are expected in the future. Only a few papers on the separation and synthesis of chiral compounds have been published so far. Because

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 127 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

127

enzymes have high selectivity, and owing to the great importance of enantioselective synthesis or enantiomeric resolution in the pharmaceutical industry, this area of research is expected to have substantial growth.

REFERENCES 1. Dordick, J.S., Enzymatic catalysis in monophasic organic-solvents, Enzyme Microb. Technol., 11, 194, 1989. 2. Kajimoto, O., Solvation in supercritical fluids: its effects on energy transfer and chemical reactions, Chem. Rev., 99, 355, 1999. 3. Ikariya, T. and Kayaki, Y., Supercritical fluids as reaction media for molecular catalysis, Catal. Surv. Jpn., 4, 39, 2000. 4. Wong, J.M. and Johnston, K.P., Solubilization of biomolecules in carbon dioxidebased supercritical solutions, Biotechnol. Prog., 2, 29, 1986. 5. Castillo, E. et al., Polar substrates for enzymatic-reactions in supercritical CO2: how to overcome the solubility limitation, Biotechnol. Lett., 16, 167, 1994. 6. Antonini, E., Carrea, G., and Cremonesi, P., Enzyme catalyzed reactions in waterorganic solvent two-phase systems, Enzyme Microb. Technol., 3, 291, 1981. 7. Martinek, K., Semenov, A.N., and Berezin, I.V., Enzymatic synthesis in biphasic aqueous-organic systems, I. Chemical equilibrium shift, Biochim. Biophys. Acta, 658, 76, 1981. 8. Martinek, K. et al., Colloidal solution of water in organic solvents: a microheterogeneous medium for enzymatic reactions, Science, 218, 889, 1982. 9. Zaks, A. and Klibanov, A.M., Enzymic catalysis in organic media at 100°C, Science, 224, 1249, 1984. 10. Zaks, A. and Klibanov, A.M., Enzyme-catalyzed processes in organic solvents, Proc. Natl. Acad. Sci. U.S.A., 82, 3192, 1985. 11. Klibanov, A.M., Improving enzymes by using them in organic solvents, Nature, 409, 241, 2001. 12. Partridge, J., Moore, B.D., and Hailing, P.J., Alpha-chymotrypsin stability in aqueousacetonitrile mixtures: is the native enzyme thermodynamically or kinetically stable under low water conditions? J. Mol. Catal. B: Enzym., 6, 11, 1999. 13. Griebenow, K. and Klibanov, A.M., On protein denaturation in aqueous-organic but not in pure organic solvents, J. Am. Chem. Soc., 118, 11695, 1996. 14. Wescott, C.R. and Klibanov, A.M., The solvent dependence of enzyme specificity, Biochim. Biophys. Acta, 1206, 1, 1994. 15. Zaks, A. and Klibanov, A.M., Enzymatic catalysis in nonaqueous solvents, J. Biol. Chem., 263, 3194, 1988. 16. Persidis, A., Extremophiles, Nat. Biotechnol., 16, 593, 1998. 17. Adams, M.W.W. and Kelly, R.M., Enzymes from microorganisms in extreme environments, Chem. Eng. News, 73, 32, 1995. 18. Faber, K., Biotransformations in Organic Chemistry, 4th ed., Springer, Berlin, 2000. 19. Hill, A.C., Reversible zymohydrolysis, J. Chem. Soc., 73, 634, 1898. 20. Kastle, J.H. and Loevenhart, A.S., Concerning lipase, the fat-splitting enzyme, and the reversibility of its action, J. Am. Chem. Soc., 24, 491, 1900. 21. Bourquelot, E. and Bridel, M., Synthtse des glucosides d’alcools, B. l’aide de l’tmulsine et reversibilitt des actions fermentaires, Ann. Chim. Phys., 29, 145, 1913.

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 128 Tuesday, July 18, 2006 8:28 AM

128

Functional Food Ingredients and Nutraceuticals

22. Bourquelot, E. and Bridel, M.J., Travaux originaux sur une reaction synthetisante de l'emulsine, J. Pharm. Chim., 7, 569, 1912. 23. Dastoli, F.R. and Price, S., Further studies on xanthine oxidase in nonpolar media, Arch. Biochem. Biophys., 122, 289, 1967. 24. Martinek, K. et al., Micellar enzymology, Eur. J. Biochem., 155, 453, 1986. 25. Krishna, H.S. et al., Reverse micellar extraction for downstream processing of proteins/enzymes, Adv. Biochem. Eng. Biotechnol., 75, 119, 2002. 26. Klibanov, A.M., Enzymatic reactions in liquid and solid paraffins: application for enzyme-based temperature abuse sensors, Chemtech., 16, 354, 1986. 27. Klibanov, A.M., Enzymatic catalysis in anhydrous organic solvents, Trends Biochem. Sci., 14, 141, 1989. 28. Schoffers, E., Golebiowski, A., and Johnson, C.R., Enantioselective synthesis through enzymatic asymmetrization, Tetrahedron, 52, 3769, 1996. 29. Bornscheuer, U.T. and Kazlauskas, R.J., Hydrolases in Organic Synthesis—RegioandStereoselective Biotransformations, Wiley-VCH, Weinheim, 1999. 30. Gandhi, N.N. et al., Lipase-catalyzed esterification, Catal. Rev., 42, 439, 2000. 31. Giri, A. et al., Biotransformations using plant cells, organ cultures and enzyme systems: current trends and future prospects, Biotechnol. Adv., 19, 175, 2001. 32. Krishna, H.S. and Karanth, N.G., Lipases and lipase-catalyzed esterification reactions in nonaqueous media, Catal. Rev., 44, 499, 2002. 33. Panke, S. and Wubbolts, M.G., Enzyme technology and bioprocess engineering, Curr. Opin. Biotechnol., 13, 111, 2002. 34. Thomas, S.M., Di Cosimo, R., and Nagarajan, A., Biocatalysis: applications and potentials for the chemical industry, Trends Biotechnol., 20, 238, 2002. 35. Klibanov, A.M., Asymmetric transformations catalyzed by enzymes in organic-solvents, Acc. Chem. Res., 23, 114, 1990. 36. Collins, A.N., Sheldrake, G.N., and Crosby, J., Chirality in Industry, Wiley, Chichester, 1992. 37. Stinson, S.C., Chiral drugs, Chem. Eng. News, 78, 55, 2000. 38. Zaks, A., Industrial biocatalysis, Curr. Opin. Chem. Biol., 5, 130, 2001. 39. Zaks, A. and Dodds, D.R., Application of biocatalysis and biotransformations to the synthesis of pharmaceuticals, Drug Discov. Today, 2, 513, 1997. 40. Schulze, B. et al., Review of biocatalysis in the production of chiral fine chemicals, Spec. Chem., 18, 244, 1998. 41. McCoy, M., Biocatalysis grows for drug synthesis, Chem. Eng. News, 77, 10, 1999. 42. Patel, R.N., Biocatalytic synthesis of intermediates for the synthesis of chiral drug substances, Curr. Opin. Biotechnol., 12, 587, 2001. 43. Rasor, J.P. and Voss, E., Enzyme-catalyzed processes in pharmaceutical industry, Appl. Catal. A: Gen., 221, 145, 2001. 44. Bornscheuer, U.T., Enzymes in Lipid Modification, Wiley-VCH, Weinheim, 2000. 45. Krishna, H.S. and Karanth, N.G., Lipase-catalyzed synthesis of isoamyl butyrate: A kinetic study, Biochim Biophys Acta, 1547, 262, 2001. 46. Krishna, H.S. and Karanth, N.G., Response surface modelling of lipase-catalyzed isoamyl propionate synthesis, J. Food Sci., 67, 32, 2002. 47. Krishna, H.S. et al., Lipase-catalyzed synthesis of isoamyl butyrate: optimization by response surface methodology, J. Am. Oil. Chem. Soc., 76, 1483, 1999. 48. Krishna, H.S. et al., Optimization of isoamyl acetate production by using immobilized lipase from Mucor miehei by response surface methodology, Enzyme Microb. Technol., 26, 131, 2000.

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 129 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

129

49. Krishna, H.S., Prapulla, S.G., and Karanth, N.G., Enzymatic synthesis of isoamyl butyrate using immobilized Rhizomucor miehei lipase in non-aqueous media, J. Ind. Microbiol. Biotechnol., 25, 147, 2000. 50. Krishna, H.S. et al., Enzymatic synthesis of isoamyl acetate using immobilized lipase from Rhizomucor miehei, J. Biotechnol., 87, 193, 2001. 51. Krishna, H.S., Sattur, A.P., and Karanth, N.G., Lipase-catalyzed synthesis of isoamyl isobutyrate: optimization using a central composite rotatable design, Process Biochem., 37, 9, 2001. 52. Kobayashi, S., Enzymatic polymerization: a new method of polymer synthesis, J. Polym. Sci. A. Polym. Chem., 37, 3041, 1999. 53. Vulfson, E.N., Novel Surfactants: Preparation, Applications and Biodegradability, Surfactant Science Series, Marcel Dekker, New York, 1998, 279. 54. Klibanov, A.M., Enzyme memory: what is remembered and why? Nature, 374, 596, 1995. 55. Ahern, T.J. and Klibanov, A.M., The mechanism of irreversible enzyme inactivation at 100-degrees-C, Science, 228, 1280, 1985. 56. Bornscheuer, U.T., Biotechnology, Wiley-VCH, Weinheim, 2000, 277. 57. Halling, P.J., Biocatalysis in low-water media: understanding effects of reaction conditions, Curr. Opin. Chem. Biol., 4, 74, 2000. 58. Schmid, R.D. and Verger, R., Lipases: interfacial enzymes with attractive applications, Angew. Chem. Int. Ed. Engl., 37, 1609, 1998. 59. Carrea, G. and Riva, S., Properties and synthetic applications of enzymes in organic solvents, Angew. Chem. Int. Ed. Engl., 39, 2226, 2000. 60. Liese, A., Seelbach, K., and Wandrey, C., Industrial Biotransformations, Wiley-VCH, Weinheim, 2000. 61. Patel, R.N., Stereoselective Biocatalysis, Marcel Dekker, New York, 2000. 62. Koeller, K.M. and Wong, C.H., Enzymes for chemical synthesis, Nature, 409, 232, 2001. 63. Lilly, M.D., Eighth P.V. Danckwerts memorial lecture presented at Glaziers´ Hall, London, 13 May 1993, Advances in biotransformation processes, Chem. Eng. Sci., 49, 151, 1994. 64. Kieslich, K. et al., New Frontiers in Screening for Microbial Biocatalysts, Elsevier, Amsterdam, 1998. 65. Demirjan, D.C., Shah, P.C., and Moris-Varas, F., Screening for novel enzymes, Topics Curr. Chem., 200, 1, 1999. 66. Wahler, D. and Reymond, J.L, Novel methods for biocatalyst screening, Curr. Opin. Chem. Biol., 5, 152, 2001. 67. Asano, Y., Overview of screening for new microbial catalysts and their uses in organic synthesis: selection and optimization of biocatalysts, J. Biotechnol., 94, 65, 2002. 68. Ornstein, R.L., Improving Enzyme Catalysis: Screening, Evolution and Rational Design, Marcel Dekker, New York, 2002. 69. Halling, P.J. and Kvittingen, L., Why did biocatalysis in organic media not take off in the 1930s? Trends Biotechnol., 17, 343, 1999. 70. Klibanov, A.M., Answering the question: “Why did biocatalysis in organic media not take off in 1930s?” Trends Biotechnol., 18, 85, 2000. 71. Kvittingen, L., Response from Kvittingen, refers to Answering the Question: “Why did biocatalysis in organic media not take off in 1930s?” Trends Biotechnol., 18, 86, 2000. 72. Coleman, M.H. and Macrae, A.R., German Fat Process and Composition, Patent DE 27 05 608, 1977 [Unilever].

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 130 Tuesday, July 18, 2006 8:28 AM

130

Functional Food Ingredients and Nutraceuticals

73. Matsuo, T. et al., European Patent EP 00 35 883, 1981 [Fuji Oil]. 74. Zaks, A. and Klibanov, A.M., Substrate-specificity of enzymes in organic-solvents vs. water is reversed, J. Am. Chem. Soc., 108, 2767, 1986. 75. Zaks, A. and Klibanov, A.M., The effect of water on enzyme action in organic media, J. Biol. Chem., 263, 8017, 1988. 76. Burton S.G., Cowan, D.A., and Woodley, J.M., The search for the ideal biocatalyst, Nat. Biotechnol., 20, 37, 2002. 77. Jarzebski, A.B. and Malinowski, J.J., Potentials and prospects for application of supercritical fluid technology in bioprocessing, Process. Biochem., 30, 343, 1995. 78. Lozano, P. et al., Continuous green biocatalytic processes using ionic liquids and supercritical carbon dioxide, Chem. Commun., 7, 692, 2002. 79. Perrut, M., Supercritical fluid applications: industrial developments and economic issues, Ind. Eng. Chem. Res., 39, 4531, 2000. 80. Randolph, T.W. et al., Enzymatic catalysis in a supercritical fluid, Biotechnol. Lett., 7, 325, 1985. 81. Hammond, D.A. et al., Enzymatic-reactions in supercritical gases, Appl. Biochem. Biotech., 11, 393, 1985. 82. Nakamura, K. et al., Lipase activity in supercritical carbon-dioxide, Chem. Eng. Commun., 45, 207, 1986. 83. Randolph, T.W., Blanch, H.W., and Prausnitz, J.M., Enzyme-catalyzed oxidation of cholesterol in supercritical carbon-dioxide, AIChE J., 34, 1354, 1988. 84. Mori, T. and Okahata, Y., Effective biocatalytic transgalactosylation in a supercritical fluid using a lipid-coated enzyme, Chem. Commun., 20, 2215, 1998. 85. Kamat, S. et al., Biocatalytic synthesis of acrylates in organic-solvents and supercritical fluids, 1. Optimization of enzyme environment, Biotechnol. Bioeng., 40, 158, 1992. 86. Kamat, S.V. et al., Biocatalytic synthesis of acrylates in supercritical fluids: tuning enzyme-activity by changing pressure, P. Natl. Acad. Sci. USA, 90, 2940, 1993. 87. Chaudhary, A.K. et al., Polyester synthesis and insitu processing using supercritical fluids, Abstr. Pap. Am. Chem. S., 211, 240, 1996. 88. Hartmann, T., Meyer, H.H., and Scheper, T., The enantioselective hydrolysis of 3hydroxy-5-phenyl-4-pentenoicacidethylester in supercritical carbon dioxide using lipases, Enzyme Microb. Technol., 28, 653, 2001. 89. Cernia, E., Palocci, C., and Soro, S., The role of the reaction medium in lipasecatalyzed esterifications and transesterifications, Chem. Phys. Lipids, 93, 157, 1998. 90. Miller, D.A., Blanch, H.W., and Prausnitz, J.M., Enzyme-catalyzed interesterification of triglycerides in supercritical carbon-dioxide, Ind. Eng. Chem. Res., 30, 939, 1991. 91. Kasche, V., Schlothauer, R., and Brunner, G., Enzyme denaturation in supercritical CO2: stabilizing effect of s-s bonds during the depressurization step, Biotechnol. Lett., 10, 569, 1988. 92. Zagrobelny, J. and Bright, F.V., In situ studies of protein conformation in supercritical fluids: trypsin in carbon-dioxide, Biotechnol. Progr., 8, 421, 1992. 93. Marty, A. et al., Comparison of lipase-catalyzed esterification in supercritical carbondioxide and in normal-hexane, Biotechnol. Lett., 12, 11, 1990. 94. Marty, A. et al., Kinetics of lipase-catalyzed esterification in supercritical CO2, Biotechnol. Bioeng., 39, 273, 1992. 95. Kamat, S. et al., Biocatalytic synthesis of acrylates in organic-solvents and supercritical fluids, 3. does carbon-dioxide covalently modify enzymes, Biotechnol. Bioeng., 46, 610, 1995.

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 131 Tuesday, July 18, 2006 8:28 AM

Biochemical Reactions in Supercritical Fluids

131

96. De Carvalho, I.B., De Sampaio, T.C., and Barreiros, S., Subtilisin Hydration and Activity in Supercritical and Near-Critical Fluids, presented at Proc. 3rd Symp. Supercrit. Fluids, Strasbourg, October 17−19, 1994, 155. 97. Ikushima, Y. et al., Activation of a lipase triggered by interactions with supercritical carbon-dioxide in the near-critical region, J. Phys. Chem., 99, 8941, 1995. 98. Ikushima, Y. et al., Promotion of a lipase-catalyzed esterification in supercritical carbon dioxide in near-critical region, Chem. Eng. Sci., 51, 2817, 1996. 99. Jensen, B.H., Galluzzo, D.R., and Jensen, R.G., Partial-purification and characterization of free and immobilized lipases from mucor-miehei, Lipids, 22, 559, 1987. 100. Dufour, E., Hervé, G., and Halrtle, T., Hydrolysis of beta-lactoglobulin by thermolysin and pepsin under high hydrostatic-pressure, Biopolymers, 35, 475, 1995. 101. Gross, M., Auerbach, G., and Jeanicke, R., The catalytic activities of monomeric enzymes show complex pressure-dependence, FEBS Lett., 321, 256, 1993. 102. Okamoto, M. et al., High-pressure proteolytic digestion of food proteins: selective elimination of beta-lactoglobulin in bovine-milk whey concentrate, Agric. Biol. Chem., 55, 1253, 1991. 103. Cheftel, J.C., Applications des hautes pressions en technologie alimentarie, Ind. Alim. Agric., 108, 141, 1991. 104. Erickson, J.C., Schyns, P., and Cooney, C.L., Effect of pressure on an enzymaticreaction in a supercritical fluid, AIChE J., 36, 299, 1990. 105. Fontes, N. et al., Effect of pressure on the catalytic activity of subtilisin Carlsberg suspended in compressed gases, Biochim. Biophy. Acta, 1383, 165, 1998. 106. Weder, J.K., Studies on proteins and amino-acids exposed to supercritical carbondioxide extraction conditions, Food Chem., 175, 1984. 107. Mori, T. et al., Reversible activity control of enzymatic reactions in supercritical fluid: enantioselective esterification catalyzed by a lipid-coated lipase in supercritical fluoroform, Kobunshi Ronbunshu, 58, 564, 2001. 108. Rezaei, K. and Temelli, F., Lipase-catalyzed hydrolysis of canola oil in supercritical carbon dioxide, J. Am. Oil. Chem. Soc., 77, 903, 2000. ˇ Gamse, T., and Marr, R., High pressure process technology: fundamentals 109. Knez, Z., and applications, in Enzymatic Reactions (Industrial chemistry library, Vol. 9), Bertucco, A. and Vetter, G., Eds., Elsevier, Amsterdam, 2001. ˇ Stability of proteinase form Carica papaya ˇ ˇ M., and Knez, Z., 110. Habulin, M., Primozic, latex in dense gases, J. Supercrit. Fluid, 33, 27, 2005. 111. Romero, M.D. et al., Enzymatic synthesis of isoamyl acetate with immobilized Candida antarctica lipase in supercritical carbon dioxide, J. Supercrit. Fluid, 33, 77, 2005. ˇ Supercritical carbon dioxide as a medium for 112. Habulin, M., Krmelj, V., and Knez, Z., enzymatically catalyzed reaction, in Proc. High Pressure Chemical Engineering, von Rohr, R. and Trepp, Ch., Eds., Elsevier, Amsterdam, 1996, 85. 113. Lozano, P. et al., Stability of immobilized alpha-chymotrypsin in supercritical carbon dioxide, Biotechnol. Lett., 18, 1345, 1996. ˇ Habulin, M., and Primozic, ˇ ˇ M., Hydrolases in supercritical CO2 and their 114. Knez, Z., use in a high-pressure membrane reactor, Bioprocess. Biosyst. Eng., 25, 279, 2003. 115. Pasta, P. et al., Subtilisin-catalyzed trans-esterification in supercritical carbon-dioxide, Biotechnol. Lett., l1, 643, 1989. ˇ Activity and stability of lipases from different sources in 116. Habulin, M. and Knez, Z., supercritical carbon dioxide and near-critical propane, J. Chem. Technol. Biotechnol. 76, 1260, 2001.

© 2007 by Taylor & Francis Group, LLC

2441_C004.fm Page 132 Tuesday, July 18, 2006 8:28 AM

132

Functional Food Ingredients and Nutraceuticals

117. Mattiasson, B. and Aldercreutz, P., Tailoring the microenvironment of enzymes in water-poor systems, Trends in Biotechnol., 9, 394, 1991. 118. Krishna, H.S., Developments and trends in enzyme catalysis in nonconventional media, Biotechnol. Adv., 20, 239, 2002. 119. Aaltonen, O., Enzymatic catalysis, in Chemical Synthesis using Supercritical Fluids Jessop, P.G. and Leiner, W., Eds., Wiley-VCH, Weinheim, 1999. 120. Turner, C. et al., Lipase-catalyzed reactions in organic and supercritical solvents: application to fat-soluble vitamin determination in milk powder and infant formula, Enzyme Microb. Tech., 29, 111, 2001. 121. Rezaei, K. and Temelli, F., On-line extraction-reaction of canola oil using immobilized lipase in supercritical CO2, J. Supercrit. Fluid, 19, 263, 2001. 122. Schmitt-Rozieres, M., Deyris, V., and Comeau, L.C., Enrichment of polyunsaturated fatty acids from sardine cannery effluents by enzymatic selective esterification, J. Am. Oil. Chem. Soc., 77, 329, 2000. 123. Smallridge, A.J., Trewhella, M.A., and Wang, Z., The enzyme-catalysed stereoselective transesterification of phenylalanine derivatives in supercritical carbon dioxide, Aust. J. Chem., 55, 259, 2002. 124. Chrastil, J., Solubility of solids and liquids in supercritical gases, J. Phys. Chem., 86, 3016, 1982. 125. Hampson, J.W. and Foglia, T.A., Effect of moisture content on immobilized lipasecatalyzed triacylglycerol hydrolysis under supercritical carbon dioxide flow in a tubular fixed-bed reactor, J. Am. Oil. Chem. Soc., 76, 777, 1999. 126. Martinez, J.L., Rezaei, K., and Temelli, F., Effect of water on canola oil hydrolysis in an online extraction-reaction system using supercritical CO2, Ind. Eng. Chem. Res., 41, 6475, 2002. ˇ The influence of water on the synthesis of n-butyl oleate 127. Leitgeb, M. and Knez, Z., by immobized Mucor miehei lipase, J. Am. Oil. Chem. Soc., 67, 775, 1990. 128. Basheer, S., Mogi, K., and Nakajima, M., Surfactant-modified lipase for the catalysis of the interesterification of triglycerides and fatty acids, Biotechnol. Bioeng., 45, 187, 1995. 129. Chulalaksananukul W., Condoret J.S., and Combes D., Geranyl acetate synthesis by lipase-catalyzed transesterification in supercritical carbon dioxide, Enzyme Microb. Technol., 15, 691, 1993. ˇ Habulin, M., and Krmelj, V., Enzyme catalyzed reactions in dense gases, 130. Knez, Z., J. Supercrit. Fluid, 14, 17, 1998. ˇ and Habulin, M., Compressed gases as alternative enzymatic-reaction sol131. Knez, Z. vents: a short review, J. Supercrit. Fluid, 23, 29, 2002. 132. Hitzler, M.G. et al., Selective catalytic hydrogenation of organic compounds in supercritical fluids as a continuous process, Organic Process Res. Dev., 2, 137, 1998. 133. Tundo, P., Continuous Flow Methods in Organic Synthesis, Ellis Horwood, Chichester, 1991.

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 135 Tuesday, July 18, 2006 8:32 AM

5

Pressurized Low Polarity Water Extraction of Biologically Active Compounds from Plant Products Juan Eduardo Cacace and Giuseppe Mazza Agriculture and Agri-Food Canada

CONTENTS 5.1 5.2

Introduction ..................................................................................................135 Pressurized Low Polarity Water Extraction Process ...................................136 5.2.1 Basic Concepts.................................................................................136 5.2.2 Equipment ........................................................................................138 5.3 Applications of PLPW Extraction ...............................................................142 5.3.1 Effect of the Extraction Temperature and Pressure ........................142 5.3.2 Fractionation of Compounds of Different Polarity .........................147 5.4 Modeling of PLPW Extraction of Bioactives from Plant Materials...........150 5.5 Conclusions ..................................................................................................152 References..............................................................................................................152

5.1 INTRODUCTION Plants synthesize many classes of organic chemical compounds ranging from simple structures to complex molecules as part of their normal metabolic processes. These compounds are broadly characterized as: (1) primary metabolites which encompass substances such as nucleic acids, proteins, lipids, and polysaccharides that are the fundamental biologically active chemical units of living plant cells, and (2) secondary metabolites which typically have larger, more complex chemical architectures that incorporate one or more primary metabolites into their structures. Various types

135

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 136 Tuesday, July 18, 2006 8:32 AM

136

Functional Food Ingredients and Nutraceuticals

of secondary metabolites synthesized by plants are commonly referred to as phytochemicals and include carotenoids, phenolics, alkaloids, terpenes, sterols, saponins, nitrogen-containing compounds, and organosulfur compounds.1 The most studied of the phytochemicals are the phenolics and carotenoids. Phenolics are compounds possessing one or more aromatic rings with one or more hydroxyl groups and generally are categorized as phenolic acids, flavonoids, stilbenes, coumarins, tannins, isoflavones and lignans. The flavonoids are further classified as flavonols, flavones, flavanols (catechins), flavanones, and anthocyanins. It is known that many phytochemicals can significantly affect human metabolism and health, and therefore, there is considerable interest in extraction of these compounds for their incorporation into functional foods, nutraceuticals, and pharmaceutical products. Also, certain classes of phytochemicals (terpenes and pyrazines) are useful for the production of flavor and fragrances and for incorporation into topical preparations. Phytochemicals typically are not soluble in water under ambient conditions due to their organic nature and the preponderance of nonionic bonds in their architectures. However, they are readily soluble in various organic solvents such as aliphatic alcohols, hexanes, dioxanes, acids, ethers, methylene chloride, trichloroethylene, and acetonitrile. Numerous methods are known for extracting phytochemicals from plant materials; most are based on sequential extraction processes incorporating one or more organic solvents in combination with washing steps.2–7 Some methods teach the use of alkali or alkaline solvents in combination with organic solvents for increased extraction efficiency.8–10 Starting plant materials are usually physically disrupted by means of grinding, shredding, chopping, pulverizing, compressing, or macerating in order to improve extraction efficiency. Phytochemical extracts have to be further processed to remove all trace of the organic solvents, and/or to separate and purify individual phytochemicals. More complex techniques such as hydrodistillation, pressurized solvent extraction, microwave-assisted extraction, supercritical CO2 extraction, and pressurized low polarity water (PLPW) extraction have been used in an attempt to reduce the use of organic solvents and increase the efficiency of the extraction process.2,5,11–13

5.2 PRESSURIZED LOW POLARITY WATER EXTRACTION PROCESS 5.2.1 BASIC CONCEPTS Pressurized low polarity water (PLPW) extraction, also known as subcritical water extraction, is a technology that modifies the properties of water by increasing the temperature up to 374ºC and keeping the pressure high enough to maintain the water in the liquid state to improve its extraction ability. It is known that the physical and chemical properties of water within sealed systems can be manipulated by concurrently controlling the temperature and pressure, whereby the water remains in the liquid state even though its temperature may be significantly increased above its atmospheric boiling point of 100ºC. In this condition, pressurized low polarity water can be maintained in the liquid form up to a temperature of 374ºC and a pressure

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 137 Tuesday, July 18, 2006 8:32 AM

Pressurized Low Polarity Water Extraction

1000

°C

4.5

70

3

0

50

3

m

kg/ 400

m

/ kg

1.5

60 0

kg /

3.5

3.0

4.0

0

kg

m3

3

300

3

/m

8

kg

/m

kg

0 90

2.5

kg / m3

2.0

95 0

/m

00

°C

20

0°C

T

3

500

1000

2000

40 0°C

0

=0 ° 20 C ° 40 C ° 60 C ° 80 C 1 0 °C 0°C

10000

137

50

10.0

360°C 340°C 320°C 300°C 280°C 260°C

1.0

Pressure (bar)

100

Density = 1000 kg/m 3 Entropy = 0.5 kj/kg K

200

240°C 220°C 200°C 180°C 160°C 140°C

= 90%

80%

70%

60%

50%

40%

30%

T = 100°C

Quality

Qu

alit y

1.0

20%

=1

0%

120°C

80°C

60°C

0.1

0

1000 Enthalpy (kj/kg)

2000

FIGURE 5.1 Pressure-enthalpy chart of water (adapted from Haar et al.14).

of 22.1 MPa (221 bars) after which it becomes supercritical water. The low polarity water can be considered in a region from normal atmospheric pressure and room temperature to 22 MPa and 374ºC (Figure 5.1). The polarity, viscosity, surface tension, and dissociation constant of subcritical water are significantly lower compared to water at ambient temperature and pressure conditions, thereby significantly altering its chemical properties to approximate those of organic solvents. Thus, increases of water temperature from 25ºC to 200ºC, for instance, would decrease its dielectric constant from 79 to 35, reaching values similar to those for ethanol (24) or methanol (33). A pressure of 5 MPa (50 bars) would be high enough to keep the

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 138 Tuesday, July 18, 2006 8:32 AM

138

Functional Food Ingredients and Nutraceuticals

water in the liquid phase during this increase of temperature (Figure 5.1). A higher extraction pressure would be detrimental for the process because of a slight increase in the dielectric constant and a considerable increase in the cost of the equipment. At 25ºC, the dielectric constant of water increases from about 79 to 93 when the pressure increases from 10 to 600 MPa (100 to 6000 bars).14 The polarity of pure water would be comparable to water-methanol or water-acetonitrile mixtures at 25ºC.15 Consequently, pressurized low polarity water can easily solubilize organic compounds such as phytochemicals, which are normally insoluble in ambient water.16 Extraction using PLPW has been compared with and reported to be superior to conventional extraction techniques including hydrodistillation,17–20 solid-liquid extraction,13,18,21 and supercritical CO2 extraction.22,23 Some of the benefits that have been mentioned include higher selectivity, cleanliness, speed, and cost savings of both raw material and energy. The heat advantage of PLPW extraction over hydrodistillation has been reported at 20 times per kg of water, although 10 kg more water would be used.24 While methods based on sequential extraction processes with organic solvents are useful for extraction and purification of small quantities of phytochemicals for research purposes, they are difficult to scale up to commercial volumes because of problems associated with cost and safe removal and recovery of the organic solvent from the extracts and spent plant materials. Furthermore, the types and concentrations of organic solvents must be carefully selected in order to avoid structural changes to the target phytochemicals during extraction that may adversely affect one or more of their desirable physical, chemical, or biological properties.

5.2.2 EQUIPMENT The most commonly used equipment in PLPW systems (Figure 5.2) consists of storage tank (T) for pure water connected to a high-pressure pump (P), which is connected to a valve (V1) and to a heating coil (HC) housed within a temperaturecontrolled chamber (O). The pressure in the line is displayed by a pressure gauge installed outside the oven. The heating coil is connected to an extraction vessel (EV), which is also mounted inside the temperature-controlled chamber. The chamber is equipped with programmable temperature controller and recorder (TC). The extraction vessel is connected to a cooling coil (CC), which in turn is connected to the inlet of a pressure regulation valve or backpressure regulator (BPR). The pressure regulation valve is connected to a collection vessel (CV). The collection vessel can be changed periodically to provide a plurality of collection volumes, thereby separating and individually collecting multiple eluant fractions. A source plant material is loaded into the extraction vessel and it is maintained inside the vessel during extraction by fitting adequate frits in the inlet and outlet of the vessel. Water is pumped from the storage tank into the extraction vessel until a desired in-line pressure is achieved, usually in the range of about 4 to 7 MPa. Liquid chromatography (LC) pumps and gas chromatography (GC) ovens have been used most frequently for pumping and heating the water (Table 5.1). The pressurized water within the extraction vessel is then heated by raising the temperature within the extraction chamber while in-line pressure is maintained at a

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 139 Tuesday, July 18, 2006 8:32 AM

Pressurized Low Polarity Water Extraction

139

V1 TC P

CC T

T

BPR HC

O

EV CV

FIGURE 5.2 Diagram of pressurized low polarity water extractor: T, water tank; P, water pump; HC, heating coil; TC, temperature controller; EV, extraction vessel; O, oven; CC, cooling coil; BPR, backpressure regulator; CV, collection vial.

desired level by the pressure regulation valve. Stainless steel tubing coils of various lengths have been used as heating and cooling elements. Varied pressure regulators have been used from very simple home tubing restrictors to automatic backpressure regulators (Table 5.1). Self-contained equipments have been also used for pressurized liquid or accelerated liquid extraction. Flow-control devices taking action on the pump achieve precise flow rates of subcritical water through the extraction vessel. The temperature within the chamber is set and maintained by an automatic controller during an extraction procedure. Components of the system between the pump and the pressure regulation valve must be connected by tubing resistant to high pressure (8 to 10 MPa) required for PLPW extraction. Extraction of bioactives using PLPW has been usually performed using pure water. Water may be further processed by distillation or filtration, and optionally, could be purged with nitrogen to remove all dissolved oxygen prior to its use. Purified water typically has a pH in the range of 5.9 to 6.2. However, if so desired, the pH of purified water can be adjusted into a range of 3.5 to 9.5 with acids or bases, prior to its use, to enhance solubilization and extraction of various phytochemicals. Extraction of bioactives from flaxmeal was optimized by modifying the pH of subcritical water. Thus, buffered water at either high or low pH was used to improve extraction of bioactives.25 Yields of proteins and the lignan secoisolariciresinol diglucoside (SDG) were maximized at alkaline pH 9. The pH was the factor that defined the equilibrium yields of SDG irrespective of the temperature. At pHs 4 and 9, extraction of SDG from flaxmeal at 160ºC and 190ºC reached the same equilibrium yield (Figure 5.3). It is known that alkaline pH helps to solubilize protein.26,27 At the same time, pH could have helped to extract lignans by breaking phenolic-protein interaction and thus releasing the lignans from the plant matrix. Alkaline pH may

© 2007 by Taylor & Francis Group, LLC

Targeted phytochemicals Essential oils Essential oils Essential oils Oxygenates with AO4 activity Fragrance and flavor compounds Anthocyanins Phenolics from whole flaxseed and flaxseed meal

1

Cell

Pump

80 × 3 ID2 mm Frit 2 µm 150 × 11 ID mm (14 mL) Frit 2 µm (10.4 mL) Frit 0.5 µm

LC3 Pump (0.5–3.0 mL/min) LC pump (2.0 mL/min) LC pump (1–3 mL/min) Isco 100 D (1 mL/min) LC pump

50 × 9.4 ID mm (3.47 mL) 0.5 µm and 2 µm frits (10.4 mL) 14.8 ID mm (55 mL) 100 100 100 200

× × × ×

7 0 ID mm 9.4 ID mm 19.3ID mm 7.7 ID mm

(24 mL/min) LC pump (0.3–4 mL/min)

Heater Coil/oven coil/oven 3 m × 0.76 mm ID coil/oven 3 m coil/oven 1 m coil/oven Coil/oven 3 m × 0.76 mm ID coil/oven

Backpressure regulator; 2 Inside diameter; 3 Liquid chromatoraphy; 4 antioxidant

© 2007 by Taylor & Francis Group, LLC

Cooler

BPR1

Coil/cooling recirculation bath (25°C) 1 m coil/cooling recirculation bath (25°C) 1 m coil/ice water bath

Home variable restrictor (2–20 MPa)

40 cm coil/water bath (25°C) Coil/cooling water bath 1 m × 0.76 mm ID coil/water bath (room temperature)

Reference 13 2,17,18,21

Pressure control valve (2–9 MPa) BPR (6–7 MPa) 11 cm to 100 µm restrictor Micro metering valve (4 MPa) BPR (5.2 MPa cartridge)

43,46 23,31 24,22 48 40,25

2441_C005.fm Page 140 Tuesday, July 18, 2006 8:32 AM

TABLE 5.1 Components of PLPW Extraction Systems Used for Extraction of Plant Bioactives

2441_C005.fm Page 141 Tuesday, July 18, 2006 8:32 AM

141

25

119

20

95

15

71

Yield (%)

SDG (mg/g meal)

Pressurized Low Polarity Water Extraction

48

10 160°C pH 4 160°C pH 9 190°C pH 4 190°C pH 9

5

24

0

0 0

100

200

300

400

500

Time or volume (min or mL)

FIGURE 5.3 Pressurized low polarity water extraction of SDG from flaxmeal at 160ºC and 190ºC and pH 4 and 9.

have also hydrolyze d complex polymeric phenolics, reducing them to more available and easily extracted compounds. Whatever mechanisms pH affected, high pH raised SDG and protein yields. Thus, pH effect overcame the increase of solubility and yield obtained by temperature. Carbohydrate extraction from flaxmeal was improved using PLPW at pH 4.25 Also, high-temperature pressurized liquid extraction (80°C to 100°C) using acidified water was as effective as acidified 60% methanol in extracting anthocyanins from grape skins.28 Another modification of water to improve its extraction efficiency has been the addition of sulfur dioxide. Sulfured PLPW (containing 1400 g/mL sodium metabisulfite) extractions of grape skin phenolics over a temperature range of 100ºC to 160°C were compared with conventional hot water or aqueous 60% (v/v) methanol extractions (50°C, 1 h). The PLPW extracts from modified sulfured water had higher levels of total anthocyanins and total phenolics than extracts that used pure water. Furthermore, subcritical water and subcritical sulfured water extracts had comparable or higher levels of anthocyanins than extracts obtained using conventional hot water or 60% methanol.29 A modification of the heating system in the subcritical water extraction could be the replacement of the oven with a water heater interconnected between the pump and the extraction vessel. In such a system the extraction vessel would be provided with a jacket or a surface heater, thereby maintaining the temperature of pressurized low polarity water flowing through the extraction vessel. The jacket may be integral to the extraction vessel, or alternatively, be mountable onto the exterior surface of the vessel.30

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 142 Tuesday, July 18, 2006 8:32 AM

142

Functional Food Ingredients and Nutraceuticals

5.3 APPLICATIONS OF PLPW EXTRACTION PLPW has been applied to extract a variety of biochemicals from a wide range of plant species (Table 5.2). It has been used to extract antioxidants from rosemary31 and Taiwan yams;32 anthocyanins from red grape skin;28,29 ginsenosides from American ginseng,33 catechins and epicatechin from tea leaves and grape seeds,34 oil from cedarwood;35 essential oil from oregano,17 anthraquinones (antibacterial, antiviral, and anticancer compounds) from roots of Morinda citrifolia,36 tanshinone I and IIA from Salvia miltiorrhiza used in Chinese medicine;37 flavones, anilines, and phenols from orange peels;38 kava lactones from kava roots;39 lignans from whole flaxseed40 and flaxseed meal,25 and saponins from cow cockle (Saponaria vaccaria) seed.41 Plant material for the extraction can include homogenous samples; or mixtures of whole plant parts such as seeds, flowers, leaves, stems, and roots; and also, with source plant materials disrupted and processed by methods including one or more of grinding, shredding, chopping, pulverizing, compressing, and macerating. The biomass can be fresh hydrated plant materials or plant materials may be dehydrated prior to extraction, or alternatively, processing by one or more of the methods described above prior to extraction. The plant materials may be packed into an extraction vessel either alone or in combination with inert physical substrates such as glass wool, glass beads, resin beads, silica sand, stainless steel wire cloth, and other like substrates whereby the inert substrates maintain spacing and distribution of the source plant materials throughout the vessel during the course of the extraction procedure thereby facilitating mass transfer while preventing migration and packing of the plant material against the outlet frits.30 Recently, Ho et al.25 showed that extraction of lignans, proteins, and carbohydrates with PLPW was positively affected by the addition of glass beads to flaxseed meal for packing the cell. The overall lignan yield increased from 10% to 50% at all temperatures tested with copacking the meal with glass beads.

5.3.1 EFFECT

OF THE

EXTRACTION TEMPERATURE

AND

PRESSURE

The extraction of bioactives from plant material with the PLPW process is clearly affected by the temperature. Thus, SDG, the major lignan in flaxseeds, along with two other phenolic compounds, p-coumaric acid glucoside and ferulic acid glucoside, were extracted with varied success at different temperatures in a PLPW system. Extraction yields of lignans and other phenolic compounds from flaxseed increased from 10 % at 100ºC up to approximately 90% at 140ºC to 160ºC.40 Temperature also affected PLPW extraction of lignan SDG from flaxmeal. Yield of flaxmeal SDG increased from 30% at 130ºC to a maximum of almost 100% at 190ºC.25 In PLPW extraction of phenolic compounds from grape seeds at 150ºC, catechin increased by 32% and epicatechin increased by 99% over the recovery at 50ºC extraction.42 Saponin yields of PLPW extracts increased with temperature and time. While only 33.2% (w/w) of total saponins was extracted at 125°C in 3 h, 60.2% (w/w) was recovered at 175°C in the first 15 min extraction from cow cockle ground seeds.41 The temperature has such a remarkable effect on the extraction and composition of the final extract that a study of temperature effect is a prerequisite in any new PLPW

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 143 Tuesday, July 18, 2006 8:32 AM

Pressurized Low Polarity Water Extraction

143

extraction research. The effect of temperature on extraction of bioactives using PLPW has also been reported in extractions of lactones from kava root,39 phenolics from grape skin,28 and essential oils from peppermint and savory,23 fennel,21 rosemary,24,31 thyme;43 clove,44 Origanum micranthum,45 Rosa canina,46 and Origanum onites.47 It is clear from these and others studies (Table 5.2) that temperature has to reach a minimum value for the water to gain the properties required to extract a given bioactive. Below the required temperature the extraction does not occur or results in a very low yield. A very good example of the effect of temperature on compound yield is the extraction of volatile compounds from peppermint (Figure 5.4). Four groups of compounds were extracted: the first included carvone, pulegone, and piperitone; the second eucalyptol, menthone, neomenthol, and menthol; the third menthyl acetate; and the fourth β-caryophyllene.23 Only one line per group has been plotted in Figure 5.4 to simplify the plot. As can be observed, extraction yields increased with temperature from values as low as zero to as high as 100%, depending on the group of compounds evaluated. At temperatures below 50ºC, only compounds of the first two groups were extracted, and yields varied depending on the extraction time. Menthyl acetate and β-caryophyllene were not extracted or their yields were very low because their solubility in water at those temperatures was likely very low. In order to begin extraction of menthyl acetate and β-caryophyllene, temperatures above 50ºC and 150ºC had to be used. Increasing the extraction time did not considerably increase the yield of these compounds when the extraction temperature was kept below 50ºC and 150ºC, respectively. Yield points represented in Figure 5.4 are yields of 15 min extraction at each temperature. Thus 80% yield at 175ºC was calculated by adding yields of five extractions of 15 min each. However at 50ºC, menthyl acetate yield after 15 min extraction was 1% or 2%; if the extraction time were increased five times, the yield would be at most 5% to 10%. Effect of temperature has been attributed to an increase on the solubility of the compounds.24 It is also possible that the temperature affects the extraction by breaking interactions between the analytes and the plant matrix.42 Because the extraction yields increase with temperature and also temperature has to reach a minimum value for the bioactives to be extracted with water, the extraction temperature must be selected for the targeted compounds. There is an optimal temperature at which the yield is maximized for every bioactive. Ideally this temperature would be related to the modified properties of the water required for the extraction in the PLPW system, which in turn are linked to the properties of the bioactive that has to be extracted. Thus, the most efficient extraction of the lignan SDG and other phenolics from flaxseed occurred in the temperature range from 140ºC to 160ºC40 and from flaxmeal at 190ºC.25 Recommended temperatures for PLPW extraction of other phytochemicals are: 150ºC for phenolics compounds from grape seeds;42 120ºC for anthocyanins from berries;48 100ºC for volatile flavor and fragrance compounds from Rosa canina,46 and 150ºC for essential oils from Achillea monocephala.49 Most of these extraction temperatures are relatively high, and may be detrimental when heat sensitive bioactives are present in the raw material. Thus, at 160ºC the total SDG measured in flaxseed extracts, solvent wash, and extracted seed residues

© 2007 by Taylor & Francis Group, LLC

Target Phytochemicals

Plant Material

Temperature/Pressure

Major Compounds

Eucalyptol, linalool, terpinen-4-ol, αterpineol, geraniol, ETMCb Essential oils Laurel 50ºC to 200ºC (150ºC)/5 MPa 1,8-cineole, 2 unidentified peaks Essential oils Clove 125ºC, 250ºC/2.4, 5, 10, 17 MPa Eugenol, eugenyl acetate Essential oils Thyme 100ºC, 125ºC, 150ºC, 175ºC (150ºC)/2, 6, 9 MPa α-pinene, p-cymene, γ-terpinene, limonene, E-3-caren-2-ol, thymol, carvacrol, caryophyllene Essential oils Oregano 100ºC–175ºC (150ºC) Thymol Essential oils Achillea monocephala 100ºC, 125ºC, 150ºC, 175ºC/6 MPa 1,8-cineole,camphor, α-campholenal, borneol, terpinen-4-ol and manymore Essential oils Origanum micranthum 100ºC, 125ºC, 150ºC, 175ºC/4 to 8 MPa α-terpineol, linalool, borneol, terpinen-4ol, and many more Essential oils Fennel 50ºC to 200ºC (150ºC)/2 MPa α-pinene, mircene, limonene, camphor, phelandrene, anethol Phenolic antioxidants Sage 70ºC, 100ºC or 150ºC/10 MPa Rosmarinic acid, carnosol, carnosic acid, methyl carnosate Antioxidant compounds Rosemary leaves 25°C, 100°C, 150°C, 200°C and 100°C, 150°C, Scutellarein,rosmanol, genkwanin, 200°Cc/6 MPa carnosol, carnosic acid, NI 1 Fragrance and flavor Peppermint 150°C, 175°C and 50°C, 100°C, 125°C, 150°C, Carvone,pulegone, eucalyptol, menthone, compounds 175°C, 200°Cc/6 MPa neomenthol, menthol, menthyl acetate Fragrance and flavor Savory 100°C, 150°C, 175°C and 50°C, 100°C, 125°C, p-cymene, thymol, carvacrol, linalool, compounds 150°C, 175ºC, 200ºCc/6 MPa borneol, thymoquinone Fragrance and flavors Rosemary 125ºC, 150ºC, 175ºC α-pinene, limonene, camphene, camphor, 1,8-cineole, borneol Essential oils

Marjoram

© 2007 by Taylor & Francis Group, LLC

100ºC – 175ºC (150ºC)a/2–20 MPa

Reference 13 18 44, 22 43

17 49 45 21 50 31 23 23 24

2441_C005.fm Page 144 Tuesday, July 18, 2006 8:32 AM

TABLE 5.2 Applications of PLPW Technology to the Extraction of Plant Bioactives

Rosa canina

50ºC, 100ºC, 150°C/2.5, 5, 7.5 MPa

Eldelberry, raspberry, bilberry, chokeberry

110°C–160°C (100°C–120°C)

Anthocyanins and phenolics Phenolics/lignans

Grape skin

100°C to 160°C by 10°C

Whole flaxseed

100°C, 120°C, 140°C, 160°C/5 MPa

Lignans, proteins

Flaxseed meal

130°C, 160°C, 190°C/5 MPa

Saponins Kava lactones

Cow cockle Kava root

125°C, 175°C/5 MPa 25°C, 50°C, 100°C to 200°C by 25°C/6–7 MPa

a

Benzaldehyde, benzyl alcohol, phenylethyl 46 alcohol, eicosane, and more Phenolic acids, ellagic acid, catechin, 48 epicatechin, resveratrol, quercetin, kaempferol, anthocyanins Monoglucoside and acylated anthocyanins 29 SDG d, p-coumaric acid glucoside, ferulic acid glucoside SDGd proteins, carbohydrates Saponins, cyclopeptides Dihydrokavain, kavain, desmethoxyyangonin, yangonin, dihydromethysticin

40 25 41 39

Selected temperature in brackets; b ethenyl-α,α-4trimethyl-3-(1-methylethenyl)-cyclohexanemethanol; c sequential equal time extractions at listed temperatures of one sample load; d SDG Secoisolariciresinol diglucoside.

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 145 Tuesday, July 18, 2006 8:32 AM

Fragrance and flavor compounds Anthocyanins

2441_C005.fm Page 146 Tuesday, July 18, 2006 8:32 AM

146

Functional Food Ingredients and Nutraceuticals

100

Yield (%)

80

60

40

Carvone Menthol Menthyl acetate β-caryophyllene

20

0 50

100 150 Temperature (ºC)

200

FIGURE 5.4 Effect of temperature on pressurized low polarity water extraction of volatile compounds from peppermint (adapted from Kubátová et al.23).

was lower than the amount achieved at 140ºC; and there was a lower SDG percentage in the seed residue at 160ºC, which may be attributed to degradation of phenolics. At temperatures higher than 160ºC degradation of SDG also occurred in PLPW extractions of bioactives from flaxmeal. However, the detrimental effect of temperature could be alleviated when a high rate of extraction removes most of the bioactives in a relatively short time. Thus, SDG yields of 96% to 98% were reached at temperature as high as 190ºC when the rate of the extraction was increased by the increase of pH to 9.25 Equilibrium time and maximum recovery were reached in less than 100 min in extractions avoiding degradation of SDG. Elevated temperature increases the extraction rate, which in turn reduces the volume of water required for the extraction and the time to reach maximum recovery. The increase of the extraction temperature increased the extraction rate of lignans from flaxmeal, thus the volume of water required for the extraction and the time to reach maximum recovery were reduced. A raise of temperature from 160ºC to 190ºC reduced the extraction time by half.25 Increased extraction rates produced by temperature raises have been reported for the extraction of kava lactones from kava roots39 and flavor and fragrance compounds from savory and peppermint,23 rosemary,24 and clove.44 However, for some plant material and targeted compounds the raise in temperature cannot be unlimited. The combination of high temperatures and long time of extraction must be avoided when temperature sensitive bioactives are being extracted. Ju and Howard28 reported that extraction temperatures higher than 100°C resulted in anthocyanin degradation in pressurized liquid extraction, which was especially marked at 140°C. Reductions from 24% to 40% of the total anthocyanins content have been reported with pure and modified PLPW in the range of temperatures from 110ºC to 160°C.29 It has been demonstrated that most phenolic compounds react easily at high

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 147 Tuesday, July 18, 2006 8:32 AM

Pressurized Low Polarity Water Extraction

147

temperature (65ºC) when they are in contact with air. However, when higher temperatures are applied under nitrogen atmosphere, there are no degradations, since the degradation process for phenolic compounds is an oxidative process requiring the presence of oxygen.42 At high temperatures, dark brown PLPW extracts with strong burning smell and increased viscosity have been found in extractions from parsley,30 flaxmeal,25 Origanum micranthum,45 oregano,17 and leaves and flowers of Achillea monocephala.49 This has been attributed to the presence of browning reaction products such as furfural, acetylfuran, and 5-methylfurfural in the extracts of flowers of Achillea monocephala at 175ºC.49 However, the degradation in PLPW extractions is lower than the effects measured at atmospheric pressure extractions even at lower temperatures. Thus, PLPW allows for the use of temperatures higher than those used in conventional extraction techniques, probably because PLPW requires shorter extraction times and an oxygen-reduced environment. The effect of pressure in PLPW extraction of bioactives from diverse plant material has also been reported. Rovio et al.44 studied the effect of four pressures at two temperatures in PLPW extraction of flavor and fragrance from clove. No significant differences were found for the recoveries of eugenol and eugenyl acetate from clove at 25, 50, 100, and 175 kg/cm2 (2.4, 4.9, 9.8, and 17.2 MPa). Similar responses have been reported on PLPW extractions from Rosa canina,46 oregano,17 Origanum micranthum,45 and marjoram.13 These results are in agreement with the change of water dielectric constant with an increase of pressure. As it has been mentioned above, an increase of pressure of 590 MPa (from 10 to 600 MPa at 25°C) results in a small increase of dielectric constant from 79 to 93.14

5.3.2 FRACTIONATION

OF

COMPOUNDS

OF

DIFFERENT POLARITY

The extraction performance of a compound in a PLPW system is related to the polarity of the compound. Extraction yields of major phenolics present in flaxseed were not markedly different, indicating that they have a similar polarity. This suggests that fractionation of the major phenolic compounds in flaxseed using a PLPW system may not be viable. However, fractionation of phenolic compounds has been achieved in black currants and parsley.30 A sequential-temperature extraction of frozen black currant particles was performed in a PLPW system at temperatures from 80ºC to 240ºC. The total phenolic concentration of the extracts decreased from 80ºC to 120ºC and the yield of the extraction increased until it reached a plateau. However, when the temperature was further increased up to 240ºC, the concentration and the yield increased continuously with the temperature (Figure 5.5). The concentration of anthocyanins decreased continuously to zero and the yield increased to reach equilibrium and remained at those values even after further temperature increase up to 240ºC. HPLC chromatograms of samples collected at 80ºC, 120ºC, and 200ºC (Figure 5.6) indicate that high polarity compounds were extracted at the initial lower temperatures, and their content in the extraction cell decreased with further extraction. Although the identification of the compounds being extracted was not pursued, the increase in yield at the highest temperatures could be attributed to the extraction of newly generated high polarity compounds by hydrolysis of low polarity polymeric compounds or to the extraction of different molecules that

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 148 Tuesday, July 18, 2006 8:32 AM

148

Functional Food Ingredients and Nutraceuticals

1000

5000

Concentration (mg/L)

4000

TPhen Anthoc Yield (mg/100 g)

600

3000

TPhen Anthoc

400

2000

200

1000

0

Yield (mg/100 g)

Concentration (mg/L)

800

0 50

100

150 200 Temperature (ºC)

300

250

FIGURE 5.5 Total phenolic and anthocyanic yield and extract concentration in pressurized low polarity water sequential-temperature extractions from frozen black currant at 80ºC to 240ºC.

(C) 200°C

(B) 120°C

(A) 80°C

0

10

20

30 40 50 Retention time (min)

60

70

80

FIGURE 5.6 HPLC chromatograms at 280 nm of frozen black currant extracts collected at 80ºC, 120ºC, and 200ºC in pressurized low polarity water sequential-temperature extractions.

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 149 Tuesday, July 18, 2006 8:32 AM

Pressurized Low Polarity Water Extraction

149

otherwise would not be extracted. Extractions of those compounds are indicated by the appearance of new peaks at the beginning of the 280 nm chromatograms of the extracts collected at 200ºC and 240ºC (Figure 5.6). On the other hand, anthocyanins were extracted continuously and completely with extractions at 80ºC, 120ºC, and 200ºC and no new compounds that absorb at 525 nm were extracted with further increase of temperature. These observations clearly suggest that by suitably controlling the extraction temperature and the collection of the extracted fractions, PLPW extraction can be used for the recovery or fractionation of compounds of different polarity. Extractions of phenolics from grape seeds using PLPW42 and antioxidant compounds from rosemary produced similar results.31 In these studies, depending upon the temperature used, there were large differences in both the identity and recovery rate of the phenolic compounds extracted. In grape seed extraction at 150ºC, catechin increased by 32% and epicatechin increased by 99% over the recovery from the 50ºC extraction. There were also some compounds that were not detected in the extracts produced at 50ºC and 100ºC but which were detected in the extracts obtained at 150ºC. It was considered that the use of higher extraction temperature helped in the breaking of bonds between the analytes and the matrix, increasing yields of some compounds or extracting compounds that otherwise would not be extracted. In rosemary extractions, when water was heated to 200ºC, the dielectric constant of water was reduced and the solubilities of less polar compounds increased by several orders of magnitude, changing the composition of the extracts.31 Similarly, Palma et al.42 reported that the behavior of the analytes depended on the matrix, and the analytes in grape seeds were more strongly bonded to the matrix than in grape skin. Therefore, for grape skins shorter extraction times and lower temperatures would result in significant increase of yield and fractionation of phenolics.42 The temperature of the water affects the extraction in two ways: first by changing the dielectric constant of the water and thus the solubility of targeted compounds, and second by breaking the interactions between the analytes and the matrix. Production of essential oils richer in oxygenated fragrance compounds of more value and with less contaminant monoterpenes and sesquiterpenes from peppermint can be produced by adequate selection of extraction temperature (Figure 5.4). Oxygenated compounds extract at substantially lower water temperatures than nonoxygenated compounds.23 Extraction at 50ºC to 100ºC will result in extracts richer on valuable oxygenated compounds with a low fraction (< 5%) of terpenes such as pcymene or β-caryophyllene. Similarly, desirable carnosol-rich rosemary extracts can be produced by PLPW extraction of rosemary leaves at 100ºC. Furthermore, by using such a procedure it is possible to obtain enriched extracts with very high antioxidant activity.31 Also, continuous PLPW extraction of oil from marjoram resulted in extracts richer in odoriferous and hence more valuable oxygenated compounds than hydrodistilled oils, which contained 11 to 22 times larger amounts of monoterpenes.13 Therefore, extracts produced by PLPW extractions can give a higher-quality and higher-value product with more intense characteristic natural aroma.

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 150 Tuesday, July 18, 2006 8:32 AM

150

Functional Food Ingredients and Nutraceuticals

5.4 MODELING OF PLPW EXTRACTION OF BIOACTIVES FROM PLANT MATERIALS Two simple models have been applied to describe the extraction profiles obtained with PLPW extraction.51 The first of these models is based on the thermodynamic distribution coefficient (KD), which assumes that analyte release from the matrix is rapid compared to elution; and the second model is a two-site kinetic model which assumes that the extraction rate is limited by the analyte release rate from the matrix, and is not limited by the thermodynamic (KD) partition that occurs during elution. The two models are defined by the following equations51 Thermodynamic model: Sa (1– ) Sb Sa 0 S = + S0  KDm  S0  (Vb − Va )d + 1  

(5.1)

ST = 1 − [ Fe− k 1t ] − [(1 − F )e− k 2 t ] S0

(5.2)

Kinetic model:

Sa: cumulative mass of the analyte extracted after volume Va (mL) Sb: cumulative mass of the analyte extracted after volume Vb S0: initial total mass of analyte in the matrix. Sb /S0 and Sa /S0: cumulative fraction of the analyte extracted by the fluid of the volume Vb and Va ST: mass of the analyte removed by the extraction fluid after time t KD: distribution coefficient; concentration in matrix/concentration in fluid F: fraction of the analyte released quickly k1 and k2: first order rate constant (mL–1) for the quickly and slowly released fractions d: density of extraction fluid at given condition (g/mL) m: mass of the extracted sample (g) The kinetic model does not include solvent volume, but relies solely on extraction time. Therefore, doubling the extractant flow rate should have little effect on the extraction efficiency per unit time if the extraction efficiency is controlled by the kinetics of the initial desorption step (assuming the other extraction parameters remain constant). On the contrary, the thermodynamic model is only dependent on volume of extractant used. Therefore, the mechanism of thermodynamic elution and desorption kinetic can be compared simply by changing the flow rate in PLPW extraction. If the concentration of bioactive compounds increases proportionally with increase in flow rate at certain extraction time, the extraction mechanism can be

© 2007 by Taylor & Francis Group, LLC

2441_C005.fm Page 151 Tuesday, July 18, 2006 8:32 AM

Pressurized Low Polarity Water Extraction

151

explained by the thermodynamic model. However, if the increase of flow rate has no significant effect on the extraction of the bioactive compounds, with the other extraction parameters kept constant, the extraction mechanism can be modeled by kinetic diffusion.40,51 In this case, plots for several flow rates of the extracted amount or yield of the analyte as a function of the extraction time must lie on the same line indicating no effect of the flow rate. The mechanism of control and therefore the model valid for PLPW extraction may be different depending on the raw material, the targeted analyte, and extraction conditions. Thus, phenolic extractions from flaxseed performed at flow rates from 1 to 4 mL/min were affected by the flow rate, which indicates that the mass transfer of the solute from the surface of the solid into the bulk of the water regulated most of the extraction process in a similar way to the thermodynamic model. However, there was no difference in the extraction rate among phenolics extractions performed at 1, 0.5, and 0.3 mL/min, which were not affected by the flow rate. Thus, extractions at low flow rates would have been controlled by the diffusion in the seeds, as the two-site kinetic model establishes above. It has been suggested that at low flow rates (KS

© 2007 by Taylor & Francis Group, LLC

(10.2)

2441_C010.fm Page 241 Thursday, July 20, 2006 10:01 AM

Microbial Modeling as Basis for Bioreactor Design

µ≈

µˆ SS First order appropriation when SS