Case Studies in Novel Food Processing Technologies: Innovations in Processing, Packaging, and Predictive Modelling

Case studies in novel food processing technologies ß Woodhead Publishing Limited, 2010 Related titles: Food preservat...

Author: Christopher J. Doona | Kenneth Kustin and Florence E. Feeherry (eds.)

470 downloads 1882 Views 4MB Size Report

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

DOWNLOAD PDF

Case studies in novel food processing technologies

ß Woodhead Publishing Limited, 2010

Related titles: Food preservation by pulsed electric fields: from research to application (ISBN 978-1-84569-058-8) Pulsed electric field (PEF) food processing is a novel, non-thermal preservation method that has the potential to produce foods with excellent sensory and nutritional quality and shelf-life. This important book reviews the current status of the technology, from research into product safety and technology development to issues associated with its commercial implementation. Food processing technology: principles and practice (Third edition) (ISBN 978-1-84569-216-2) The first edition of Food processing technology was quickly adopted as the standard text by many food science and technology courses. The publication of a completely revised and updated third edition consolidates the position of this textbook as the best single-volume introduction to food manufacturing technologies available. The third edition has been updated and extended to include the many developments that have taken place since the second edition was published. In particular, advances in microprocessor control of equipment, `minimal' processing technologies, functional foods, developments in àctive' or ìntelligent' packaging, and storage and distribution logistics are described. Technologies that relate to cost savings, environmental improvement or enhanced product quality are highlighted. Additionally, sections in each chapter on the impact of processing on food-borne micro-organisms are included for the first time. Food preservation techniques (ISBN 978-1-85573-530-9) Extending the shelf-life of foods whilst maintaining safety and quality is a critical issue for the food industry. As a result there have been major developments in food preservation techniques, which are summarised in this authoritative collection. The first part of the book examines the key issue of maintaining safety as preservation methods become more varied and complex. The rest of the book looks both at individual technologies and how they are combined to achieve the right balance of safety, quality and shelf-life for particular products. Details of these books and a complete list of Woodhead titles can be obtained by: · visiting our web site at www.woodheadpublishing.com · contacting Customer Services (e-mail: [email protected]; fax: +44 (0) 1223 893694; tel.: +44 (0) 1223 891358 ext. 130; address: Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, UK)


Woodhead Publishing Series in Food Science,Technology and Nutrition: Number197

Case studies in novel food processing technologies Innovations in processing, packaging and predictive modelling

Edited by Christopher J. Doona, Kenneth Kustin and Florence E. Feeherry


Published by Woodhead Publishing Limited, Abington Hall, Granta Park, Great Abington, Cambridge CB21 6AH, UK www.woodheadpublishing.com Woodhead Publishing, 525 South 4th Street #241, Philadelphia, PA 19147, USA Woodhead Publishing India Private Limited, G-2, Vardaan House, 7/28 Ansari Road, Daryaganj, New Delhi ± 110002, India www.woodheadpublishingindia.com First published 2010, Woodhead Publishing Limited ß Woodhead Publishing Limited, 2010 The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publisher cannot assume responsibility for the validity of all materials. Neither the authors nor the publisher, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from Woodhead Publishing Limited. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library. ISBN 978-1-84569-551-4 (print) ISBN 978-0-85709-071-3 (online) ISSN 2042-8049 Woodhead Publishing Series in Food Science, Technology and Nutrition (print) ISSN 2042-8057 Woodhead Publishing Series in Food Science, Technology and Nutrition (online) The publisher's policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp which is processed using acidfree and elemental chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards. Typeset by Godiva Publishing Services Limited, Coventry, West Midlands, UK Printed by TJI Digital, Padstow, Cornwall, UK


Contents

Contributor contact details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xiii

Woodhead Publishing Series in Food Science, Technology and Nutrition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xix

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxvii 1

Non-thermal food pasteurization processes: an introduction . . . P. Chen, S. Deng, Y. Cheng, X. Lin, L. Metzger and R. Ruan, University of Minnesota, USA 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Pulsed electric field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 High hydrostatic pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Ionizing irradiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.5 Ultraviolet radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.6 Non-thermal plasma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.7 Concentrated high intensity electric field . . . . . . . . . . . . . . . . . . . 1.8 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.9 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 2 5 6 8 9 12 12 13

Part I Case studies in high pressure and pulsed electric field processing of food 2

Commercial high pressure processing of ham and other sliced meat products at Esteban EspunÄa, S.A. . . . . . . . . . . . . . . . . . . . . . . . . . . M. Gassiot and P. Masoliver, Esteban EspunÄa, S. A., Spain 2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ß Woodhead Publishing Limited, 2010

21 21

vi

Contents 2.2 2.3 2.4 2.5 2.6 2.7

3

High pressure processing (HPP) equipment . . . . . . . . . . . . . . . . . Commercialized HPP-treated food products . . . . . . . . . . . . . . . . Treatment costs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Company information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

High hydrostatic pressure processing of fruit juices and smoothies: research and commercial application . . . . . . . . . . . . . . . . F. Sampedro and X. Fan, United States Department of Agriculture, Agricultural Research Service (USDA-ARS), USA and D. Rodrigo, CSIC, Spain 3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Fruit composition, high hydrostatic pressure (HHP) treatment and recommended fruit intake . . . . . . . . . . . . . . . . . . . . 3.3 Basic research on high hydrostatic pressure (HHP) processing of fruit juices and derivatives . . . . . . . . . . . . . . . . . . . 3.4 Commercialization of juices treated by high hydrostatic pressure (HHP) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6 Sources of further information and advice . . . . . . . . . . . . . . . . . . 3.7 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23 27 30 31 32 32 34

34 35 37 60 66 66 67 67

4

Pulsed electric field (PEF) systems for commercial food and juice processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 M. A. Kempkes, Diversified Technologies Inc., USA 4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 4.2 Key process parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 4.3 Pulsed electric field (PEF) system overview . . . . . . . . . . . . . . . . 82 4.4 Pulsed electric field (PEF) system trade-offs and optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 4.5 Pulsed electric field (PEF) processing and commercialization status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 4.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 4.7 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

5

The environmental impact of pulsed electric field treatment and high pressure processing: the example of carrot juice . . . . . J. Davis, The Swedish Institute for Food and Biotechnology (SIK), Sweden and G. Moates and K. Waldron, Institute of Food Research, UK 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Goal definition and scoping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3 Inventory of carrot juice processing . . . . . . . . . . . . . . . . . . . . . . . .


103

103 104 106

Contents 5.4 5.5 5.6 5.7 5.8 Part II 6

7

Choice of impact categories and impact assessment methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion and conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

vii 110 110 113 114 115

Case studies in other novel food processing techniques

Industrial applications of high power ultrasonics in the food, beverage and wine industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . D. Bates, Cavitus Pty Ltd, Australia and A. Patist, Cargill Inc., USA 6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2 High power ultrasound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 Process and scale-up parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 Applications and benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Large-scale implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6 Roadmap to successful commercialization . . . . . . . . . . . . . . . . . . 6.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The potential of novel infrared food processing technologies: case studies of those developed at the USDA-ARS Western Region Research Center and the University of California-Davis . . . . . . . . Z. Pan and G. G. Atungulu, University of California-Davis, USA 7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 Effect of infrared (IR) on food molecular constituents . . . . . 7.3 Case studies in novel infrared (IR) technologies for improved processing efficiency and food safety . . . . . . . . . . . . . . . . . . . . . . 7.4 Simultaneous infrared blanching and dehydration (SIRBD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5 Sequential infrared (IR) and freeze-drying of strawberry slices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.6 Infrared (IR) pasteurization of raw almonds . . . . . . . . . . . . . . . . 7.7 Infrared (IR) dry-roasting of almonds . . . . . . . . . . . . . . . . . . . . . . 7.8 An overview of infrared (IR) rough rice drying and disinfestation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.9 Effectiveness of infrared (IR) heating for simultaneous drying and disinfestation of freshly harvested rough rice . . . 7.10 Effectiveness of infrared (IR) heating for disinfestation of stored rough rice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.11 Infrared (IR) radiation heating for tomato peeling . . . . . . . . . . 7.12 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.13 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.14 References and further reading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ß Woodhead Publishing Limited, 2010

119 119 120 121 124 133 136 136 137

139 139 140 141 142 155 170 175 179 180 194 203 204 204 204

viii 8

9

Contents Validation and commercialization of dense phase carbon dioxide processing for orange juice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . K.-L. G. Ho, Chiquita Brands International Inc., USA 8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Dense phase carbon dioxide processing . . . . . . . . . . . . . . . . . . . . 8.3 Better Than Fresh TM (BTF) system . . . . . . . . . . . . . . . . . . . . . . . . . 8.4 Commercialization of the Better Than FreshTM (BTF) system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Progress and issues with the commercialization of cool plasma in food processing: a selection of case studies . . . . . . . . . . . . . . . . . . . P. Sanguansri, K. Knoerzer, J. Coventry and C. Versteeg, CSIRO Food and Nutritional Sciences, Australia 9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2 Case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.3 Case study 1: cascaded dielectric barrier discharge (CDBD) ± cool plasma for the decontamination of packaging materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4 Case study 2: atmospheric gliding arc and blown arc air cold plasma system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.5 Case study 3: atmospheric-based dielectric gas discharge . . 9.6 Case study 4: ultralight dielectric barrier discharge and spot system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.7 Case study 5: microwave vacuum cool plasma generation . 9.8 Case study 6: cool plasma for application in food processing and medical device technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.9 Case study 7: gentle e-ventus Õ disinfection of cereal crop seeds, grain and food . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.10 Conclusions and future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.11 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.12 Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10 Commercial applications of ozone in food processing . . . . . . . . . . . R. G. Rice, RICE International Consulting Enterprises, USA 10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.2 Current commercial examples of ozone in agri-foods industries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.3 Ozone for shellfish and fish processing . . . . . . . . . . . . . . . . . . . . . 10.4 Ozone in breweries and wineries . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.5 Ozone for vegetable processing and storage . . . . . . . . . . . . . . . . 10.6 Ozone washing/packaging of fresh cut salad mixes and fruit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.7 Ozone processing of meats and sushi . . . . . . . . . . . . . . . . . . . . . . .


209 209 211 214 219 222 222 226 226 233 235 237 238 239 241 244 247 250 251 253 258 258 259 259 262 266 269 272

Contents 10.8

Ozone for preparation of fresh (not frozen) microwaveable meals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.9 Cleaning-in-place with ozone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.10 Future prospects for ozone in agri-foods and food processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.11 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Novel technologies for the decontamination of fresh and minimally processed fruits and vegetables . . . . . . . . . . . . . . . . . . . . . . . B. A. Niemira, United States Department of Agriculture, Agricultural Research Service (USDA-ARS), USA 11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Optimization of existing chemical treatments . . . . . . . . . . . . . . 11.3 Antimicrobial treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.4 Adaptation of existing technologies: plasma, phage treatment and bacteria-based biological controls . . . . . . . . . . . . 11.5 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.6 Sources of further information and advice . . . . . . . . . . . . . . . . . . 11.7 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Part III

ix 274 275 277 279 283 283 284 285 289 294 296 297 297

Case studies in food preservation using antimicrobials, novel packaging and storage techniques

12 Use of natamycin as a preservative on the surface of baked goods: a case study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . J. Delves-Broughton, Danisco UK Ltd, UK and L. Steenson, C. Dorko, J. Erdmann, S. Mallory, F. Norbury and B. Thompson, Danisco USA Inc., USA 12.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.2 Natamycin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.3 The problem of mold spoilage in baked goods . . . . . . . . . . . . . 12.4 Trials on the use of natamycin as a surface treatment of baked goods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.5 Considerations and selection of the spraying system . . . . . . . 12.6 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Commercial applications of oxygen depleted atmospheres for the preservation of food commodities . . . . . . . . . . . . . . . . . . . . . . . . . . . . S. Navarro, Food Technology International Consultancy Ltd, Israel 13.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.2 Definitions and uses of oxygen depleted atmospheres . . . . . . 13.3 Effects of modified atmospheres (MAs) on stored-product insects and mites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ß Woodhead Publishing Limited, 2010

303

303 304 307 308 312 317 318 321 321 322 325

x

Contents 13.4

The effect of modified atmosphere (MA) on preventing mold growth and mycotoxin formation . . . . . . . . . . . . . . . . . . . . . 13.5 Effects of modified atmosphere (MA) on product quality . . 13.6 Generation and application of modified atmospheres (MAs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13.7 Types of structures used for modified atmospheres (MAs) . 13.8 Specific applications of modified atmosphere (MA) . . . . . . . . 13.9 Sources of further information and advice . . . . . . . . . . . . . . . . . . 13.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Commercialization of time-temperature integrators for foods . . P. S. Taoukis, National Technical University of Athens, Greece 14.1 Introduction: active and intelligent packaging ± timetemperature integrators (TTIs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.2 History of time-temperature integrators (TTIs) ± definition and principles of operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.3 State of the art time-temperature integrator (TTI) technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.4 Use of time-temperature integrators (TTIs) as tools for food chain monitoring and management . . . . . . . . . . . . . . . . . . . . . . . . . 14.5 Use of time-temperature integrators (TTIs) as shelf-life indicators for consumers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.6 Factors in time-temperature integrator (TTI) commercial success ± industry and consumer attitudes . . . . . . . . . . . . . . . . . . 14.7 Cases of time-temperature integrator (TTI) applications . . . . 14.8 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.9 Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14.10 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Development of a nanocomposite meal bag for individual military rations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C. Thellen, J. A. Ratto, D. Froio and J. Lucciarini, US Army Natick Soldier RD&E Center, USA 15.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.2 Introduction of the Meal Ready-to-Eat TM (MRE) . . . . . . . . . . . 15.3 Research and development of the MRETM nanocomposite meal bag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.4 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15.5 Sources of further information and advice . . . . . . . . . . . . . . . . . . 15.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .


327 328 330 333 337 344 344 351 351 352 354 357 360 361 362 364 365 365 367 367 370 375 385 385 386

Contents Part IV

xi

Innovations in advanced food processing techniques and predictive microbial models: case studies

16 Developments in in-container retort technology: the Zinetec ShakaÕ process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . R. Walden, Zinetec Ltd, UK and J. Emanuel, Utek Europe Ltd, UK 16.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.2 The ShakaÕ process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.3 Product quality and the ShakaÕ process . . . . . . . . . . . . . . . . . . . . 16.4 Commercialization of the ShakaÕ process . . . . . . . . . . . . . . . . . . 16.5 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.6 Sources of further information and advice . . . . . . . . . . . . . . . . . . 17 Industrial microwave heating of food: principles and three case studies of its commercialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . R. F. Schiffmann, RF Schiffmann Associates, Inc., USA 17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.2 Fundamental properties of microwaves . . . . . . . . . . . . . . . . . . . . . 17.3 How microwaves heat materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.4 Industrial microwave equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.5 Case studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17.7 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Irradiation of fresh fruits and vegetables: principles and considerations for further commercialization . . . . . . . . . . . . . . . . . . . X. Fan, United States Department of Agriculture, Agricultural Research Service (USDA-ARS), USA 18.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.2 Technology and dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.3 Application of irradiation on fresh produce . . . . . . . . . . . . . . . . . 18.4 Considerations and challenges for commercialization in the US . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.6 Sources of further information and advice . . . . . . . . . . . . . . . . . . 18.7 Disclaimer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Consumer acceptance and marketing of irradiated meat . . . . . . . R. F. Eustice, Minnesota Beef Council, USA and C. M. Bruhn, University of California-Davis, USA 19.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.2 Time to take a fresh look at irradiation . . . . . . . . . . . . . . . . . . . . . 19.3 History of irradiation of foods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.4 Education: the key to consumer acceptance . . . . . . . . . . . . . . . . 19.5 Future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ß Woodhead Publishing Limited, 2010

389 389 391 402 403 405 406 407 407 408 411 412 415 425 426 427 427 428 430 433 438 439 439 439 442 442 445 449 453 458

xii

Contents 19.6 19.7

Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

20 Comparing the effectiveness of thermal and non-thermal food preservation processes: the concept of equivalent efficacy . . . . . . M. G. Corradini, Universidad Argentina de la Empresa, Argentina and M. Peleg, University of Massachusetts-Amherst, USA 20.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.2 Traditional microbial mortality kinetics and sterility measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.3 Non-linear kinetics of microbial inactivation and deterioration processes involving nutrient or quality losses . 20.4 Equivalence criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.5 Freeware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.7 Disclaimer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20.8 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

460 460 464 464 466 469 474 482 483 486 486

21 A case study in military ration foods: the Quasi-chemical model and a novel accelerated three-year challenge test . . . . . . . . . . . . . . . C. J. Doona, F. E. Feeherry and E. W. Ross, US Army Natick Soldier RD&E Center, USA 21.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.2 Modeling S. aureus growth in intermediate moisture (IM) bread . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.3 Microbial challenge study of Maple-filled French toast . . . . 21.4 Results of the microbial challenge study . . . . . . . . . . . . . . . . . . . 21.5 Conclusions and future trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21.6 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

494 501 505 510 511

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

514


489 489

Contributor contact details

Chapter 1

(* = main contact)

Editors C. J. Doona* and F. E. Feeherry US Army Natick Soldier RD&E Center 15 Kansas Street Natick, MA 01760-5018 USA E-mail: [email protected] K. Kustin Department of Chemistry MS015 Brandeis University PO Box 549110 Waltham, MA 02453 USA

P. Chen, S. Deng, Y. Cheng, X. Lin, L. Metzger and R. Ruan* Center for Biorefining Department of Bioproducts and Biosystems Engineering Department of Food Science and Nutrition University of Minnesota 1390 Eckles Avenue St. Paul, MN 55108 USA E-mail: [email protected]

Chapter 2 M. Gassiot* and P. Masoliver Esteban EspunÄa, S.A. c/ Mestre Turina 39-41 17800 Olot Spain E-mail: [email protected] [email protected]


xiv


Chapter 3 F. Sampedro and X. Fan* Eastern Regional Research Center Agricultural Research Service US Department of Agriculture 600 East Mermaid Lane Wyndmoor, PA 19038 USA E-mail: [email protected] [email protected] D. Rodrigo Institute of AgroChemistry and Food Technology CSIC PO Box 73 46100 Burjassot Valencia Spain E-mail: [email protected]

Chapter 4 M. A. Kempkes Diversified Technologies Inc. 35 Wiggins Avenue Bedford, MA 01730 USA E-mail: [email protected]

Chapter 5 J. Davis* The Swedish Institute for Food and Biotechnology (SIK) Gothenburg Sweden

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

Chapter 6 D. Bates* Cavitus Pty Ltd 32 Spring Gully Rd Crafers, SA 5052 Australia E-mail: [email protected] A. Patist Cargill Research 2301 Crosby Road Wayzata, MN 55391 USA E-mail: [email protected]

Chapter 7 Z. Pan* Processed Foods Research Unit USDA-ARS Western Region Research Center Albany, CA 94710 USA E-mail: [email protected]

E-mail: [email protected]


Contributor contact details Z. Pan and G. G. Atungulu Department of Biological and Agricultural Engineering University of California-Davis One Shields Avenue Davis, CA 95616 USA E-mail: [email protected] [email protected]

xv

Chapter 11 Brendan A. Niemira Produce Safety Research Project US Department of Agriculture, Agricultural Research Service Eastern Regional Research Center 600 E. Mermaid Lane Wyndmoor, PA 19038 USA E-mail: [email protected]

Chapter 8 K.-L. G. Ho Chiquita Brands International Inc. 607 Brunken Avenue Salinas, CA 93901 USA E-mail: [email protected]

Chapter 9 P. Sanguansri*, K. Knoerzer, J. Coventry and C. Versteeg CSIRO Food and Nutritional Sciences 671 Sneydes Road Werribee Australia E-mail: [email protected] [email protected] [email protected] [email protected]

Chapter 10 Rip G. Rice RICE International Consulting Enterprises 1710 Hickory Knoll Road Sandy Spring, MD 20860 USA E-mail: [email protected]

Chapter 12 J. Delves-Broughton Danisco UK Ltd Food Protection/Multiple Food Applications 6 North Street Beaminster Dorset DT8 3DZ UK E-mail: joss.delves-broughton@ danisco.com Larry Steenson*, Cathy Dorko, Jerry Erdmann, Fritz Norbury, Steven Mallory and Brett Thompson Danisco USA Inc. Four New Century Parkway New Century, KS 66031 USA E-mail: larry.steenson/newcentury/ [email protected] cathy.dorko/newcentury/ [email protected] jerry.erdman/newcentury/ [email protected] fritz.norbury/madison/ [email protected] steven.mallory/newcentury/ [email protected] brett.thompson/newcentury/ [email protected]


xvi


Chapter 13 Shlomo Navarro FTIC (Food Technology International Consultancy) Ltd 5 Argaman Street Rishon Letsion 75709 Israel E-mail: [email protected]

Chapter 14 P. S. Taoukis National Technical University of Athens School of Chemical Engineering Division IV ± Product and Process Development Laboratory of Food Chemistry and Technology Iroon Polytechniou 5 15780 Athens Greece E-mail: [email protected]

Chapter 15 Christopher Thellen*, Jo Ann Ratto, Danielle Froio, Jeanne Lucciarini US Army Natick Soldier RD&E Center 15 Kansas Street Natick, MA 01760 USA E-mail: [email protected]

Chapter 16 R. Walden* Zinetec Ltd 22 Highworth Road Faringdon Oxfordshire SN7 7EE UK

John Emanuel Utek Europe Ltd 20 Regents Park Road London NW1 7TX UK E-mail: [email protected]

Chapter 17 R. F. Schiffmann R. F. Schiffmann Associates, Inc. 149 West 88 Street New York, NY 10024-2424 USA E-mail: [email protected]

Chapter 18 X. Fan United States Department of Agriculture Agricultural Research Service Eastern Regional Research Center 600 E. Mermaid Lane Wyndmoor, PA 19038 USA E-mail: [email protected]

Chapter 19 R. F. Eustice* Minnesota Beef Council 2950 Metro Drive 102 Bloomington, MN 55425 USA E-mail: [email protected]

E-mail: [email protected]


Contributor contact details xvii C. M. Bruhn Department of Food Science and Technology University of California-Davis Davis, CA 95616 USA E-mail: [email protected]

Chapter 20 Micha Peleg* Department of Food Science University of Massachusetts Amherst, MA 01003 USA E-mail: [email protected]

Maria G. Corradini Instituto de TecnologõÂa Faculdad de IngenierõÂa y Ciencias Exactas Universidad Argentina de la Empresa Ciudad de Buenos Aires Argentina E-mail: [email protected]

Chapter 21 C. J. Doona*, F. E. Feeherry and E. W. Ross US Army Natick Soldier RD&E Center 15 Kansas Street Natick, MA 01760-5018 USA E-mail: [email protected] [email protected]


Woodhead Publishing Series in Food Science, Technology and Nutrition

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

Chilled foods: a comprehensive guide Edited by C. Dennis and M. Stringer Yoghurt: science and technology A. Y. Tamime and R. K. Robinson Food processing technology: principles and practice P. J. Fellows Bender's dictionary of nutrition and food technology Sixth edition D. A. Bender Determination of veterinary residues in food Edited by N. T. Crosby Food contaminants: sources and surveillance Edited by C. Creaser and R. Purchase Nitrates and nitrites in food and water Edited by M. J. Hill Pesticide chemistry and bioscience: the food±environment challenge Edited by G. T. Brooks and T. Roberts Pesticides: developments, impacts and controls Edited by G. A. Best and A. D. Ruthven Dietary fibre: chemical and biological aspects Edited by D. A. T. Southgate, K. W. Waldron, I. T. Johnson and G. R. Fenwick Vitamins and minerals in health and nutrition M. Tolonen Technology of biscuits, crackers and cookies Second edition D. Manley Instrumentation and sensors for the food industry Edited by E. Kress-Rogers Food and cancer prevention: chemical and biological aspects Edited by K. W. Waldron, I. T. Johnson and G. R. Fenwick Food colloids: proteins, lipids and polysaccharides Edited by E. Dickinson and B. Bergenstahl Food emulsions and foams Edited by E. Dickinson Maillard reactions in chemistry, food and health Edited by T. P. Labuza, V. Monnier, J. Baynes and J. O'Brien The Maillard reaction in foods and medicine Edited by J. O'Brien, H. E. Nursten, M. J. Crabbe and J. M. Ames Encapsulation and controlled release Edited by D. R. Karsa and R. A. Stephenson Flavours and fragrances Edited by A. D. Swift Feta and related cheeses Edited by A. Y. Tamime and R. K. Robinson Biochemistry of milk products Edited by A. T. Andrews and J. R. Varley Physical properties of foods and food processing systems M. J. Lewis


xx


24 Food irradiation: a reference guide V. M. Wilkinson and G. Gould 25 Kent's technology of cereals: an introduction for students of food science and agriculture Fourth edition N. L. Kent and A. D. Evers 26 Biosensors for food analysis Edited by A. O. Scott 27 Separation processes in the food and biotechnology industries: principles and applications Edited by A. S. Grandison and M. J. Lewis 28 Handbook of indices of food quality and authenticity R. S. Singhal, P. K. Kulkarni and D. V. Rege 29 Principles and practices for the safe processing of foods D. A. Shapton and N. F. Shapton 30 Biscuit, cookie and cracker manufacturing manuals Volume 1: ingredients D. Manley 31 Biscuit, cookie and cracker manufacturing manuals Volume 2: biscuit doughs D. Manley 32 Biscuit, cookie and cracker manufacturing manuals Volume 3: biscuit dough piece forming D. Manley 33 Biscuit, cookie and cracker manufacturing manuals Volume 4: baking and cooling of biscuits D. Manley 34 Biscuit, cookie and cracker manufacturing manuals Volume 5: secondary processing in biscuit manufacturing D. Manley 35 Biscuit, cookie and cracker manufacturing manuals Volume 6: biscuit packaging and storage D. Manley 36 Practical dehydration Second edition M. Greensmith 37 Lawrie's meat science Sixth edition R. A. Lawrie 38 Yoghurt: science and technology Second edition A. Y. Tamime and R. K. Robinson 39 New ingredients in food processing: biochemistry and agriculture G. Linden and D. Lorient 40 Benders' dictionary of nutrition and food technology Seventh edition D. A. Bender and A. E. Bender 41 Technology of biscuits, crackers and cookies Third edition D. Manley 42 Food processing technology: principles and practice Second edition P. J. Fellows 43 Managing frozen foods Edited by C. J. Kennedy 44 Handbook of hydrocolloids Edited by G. O. Phillips and P. A. Williams 45 Food labelling Edited by J. R. Blanchfield 46 Cereal biotechnology Edited by P. C. Morris and J. H. Bryce 47 Food intolerance and the food industry Edited by T. Dean 48 The stability and shelf life of food Edited by D. Kilcast and P. Subramaniam 49 Functional foods: concept to product Edited by G. R. Gibson and C. M. Williams 50 Chilled foods: a comprehensive guide Second edition Edited by M. Stringer and C. Dennis 51 HACCP in the meat industry Edited by M. Brown 52 Biscuit, cracker and cookie recipes for the food industry D. Manley 53 Cereals processing technology Edited by G. Owens 54 Baking problems solved S. P. Cauvain and L. S. Young 55 Thermal technologies in food processing Edited by P. Richardson 56 Frying: improving quality Edited by J. B. Rossell 57 Food chemical safety Volume 1: contaminants Edited by D. Watson 58 Making the most of HACCP: learning from others' experience Edited by T. Mayes and S. Mortimore 59 Food process modelling Edited by L. M. M. Tijskens, M. L. A. T. M. Hertog and B. M. NicolaõÈ



xxi

60 EU food law: a practical guide Edited by K. Goodburn 61 Extrusion cooking: technologies and applications Edited by R. Guy 62 Auditing in the food industry: from safety and quality to environmental and other audits Edited by M. Dillon and C. Griffith 63 Handbook of herbs and spices Volume 1 Edited by K. V. Peter 64 Food product development: maximising success M. Earle, R. Earle and A. Anderson 65 Instrumentation and sensors for the food industry Second edition Edited by E. Kress-Rogers and C. J. B. Brimelow 66 Food chemical safety Volume 2: additives Edited by D. Watson 67 Fruit and vegetable biotechnology Edited by V. Valpuesta 68 Foodborne pathogens: hazards, risk analysis and control Edited by C. de W. Blackburn and P. J. McClure 69 Meat refrigeration S. J. James and C. James 70 Lockhart and Wiseman's crop husbandry Eighth edition H. J. S. Finch, A. M. Samuel and G. P. F. Lane 71 Safety and quality issues in fish processing Edited by H. A. Bremner 72 Minimal processing technologies in the food industries Edited by T. Ohlsson and N. Bengtsson 73 Fruit and vegetable processing: improving quality Edited by W. Jongen 74 The nutrition handbook for food processors Edited by C. J. K. Henry and C. Chapman 75 Colour in food: improving quality Edited by D MacDougall 76 Meat processing: improving quality Edited by J. P. Kerry, J. F. Kerry and D. A. Ledward 77 Microbiological risk assessment in food processing Edited by M. Brown and M. Stringer 78 Performance functional foods Edited by D. Watson 79 Functional dairy products Volume 1 Edited by T. Mattila-Sandholm and M. Saarela 80 Taints and off-flavours in foods Edited by B. Baigrie 81 Yeasts in food Edited by T. Boekhout and V. Robert 82 Phytochemical functional foods Edited by I. T. Johnson and G. Williamson 83 Novel food packaging techniques Edited by R. Ahvenainen 84 Detecting pathogens in food Edited by T. A. McMeekin 85 Natural antimicrobials for the minimal processing of foods Edited by S. Roller 86 Texture in food Volume 1: semi-solid foods Edited by B. M. McKenna 87 Dairy processing: improving quality Edited by G. Smit 88 Hygiene in food processing: principles and practice Edited by H. L. M. Lelieveld, M. A. Mostert, B. White and J. Holah 89 Rapid and on-line instrumentation for food quality assurance Edited by I. Tothill 90 Sausage manufacture: principles and practice E. Essien 91 Environmentally-friendly food processing Edited by B. Mattsson and U. Sonesson 92 Bread making: improving quality Edited by S. P. Cauvain 93 Food preservation techniques Edited by P. Zeuthen and L. Bùgh-Sùrensen 94 Food authenticity and traceability Edited by M. Lees 95 Analytical methods for food additives R. Wood, L. Foster, A. Damant and P. Key 96 Handbook of herbs and spices Volume 2 Edited by K. V. Peter 97 Texture in food Volume 2: solid foods Edited by D. Kilcast 98 Proteins in food processing Edited by R. Yada 99 Detecting foreign bodies in food Edited by M. Edwards


xxii


100 Understanding and measuring the shelf-life of food Edited by R. Steele 101 Poultry meat processing and quality Edited by G. Mead 102 Functional foods, ageing and degenerative disease Edited by C. Remacle and B. Reusens 103 Mycotoxins in food: detection and control Edited by N. Magan and M. Olsen 104 Improving the thermal processing of foods Edited by P. Richardson 105 Pesticide, veterinary and other residues in food Edited by D. Watson 106 Starch in food: structure, functions and applications Edited by A.-C. Eliasson 107 Functional foods, cardiovascular disease and diabetes Edited by A. Arnoldi 108 Brewing: science and practice D. E. Briggs, P. A. Brookes, R. Stevens and C. A. Boulton 109 Using cereal science and technology for the benefit of consumers: proceedings of the 12th International ICC Cereal and Bread Congress, 24±26 May, 2004, Harrogate, UK Edited by S. P. Cauvain, L. S. Young and S. Salmon 110 Improving the safety of fresh meat Edited by J. Sofos 111 Understanding pathogen behaviour in food: virulence, stress response and resistance Edited by M. Griffiths 112 The microwave processing of foods Edited by H. Schubert and M. Regier 113 Food safety control in the poultry industry Edited by G. Mead 114 Improving the safety of fresh fruit and vegetables Edited by W. Jongen 115 Food, diet and obesity Edited by D. Mela 116 Handbook of hygiene control in the food industry Edited by H. L. M. Lelieveld, M. A. Mostert and J. Holah 117 Detecting allergens in food Edited by S. Koppelman and S. Hefle 118 Improving the fat content of foods Edited by C. Williams and J. Buttriss 119 Improving traceability in food processing and distribution Edited by I. Smith and A. Furness 120 Flavour in food Edited by A. Voilley and P. Etievant 121 The Chorleywood bread process S. P. Cauvain and L. S. Young 122 Food spoilage microorganisms Edited by C. de W. Blackburn 123 Emerging foodborne pathogens Edited by Y. Motarjemi and M. Adams 124 Benders' dictionary of nutrition and food technology Eighth edition D. A. Bender 125 Optimising sweet taste in foods Edited by W. J. Spillane 126 Brewing: new technologies Edited by C. Bamforth 127 Handbook of herbs and spices Volume 3 Edited by K. V. Peter 128 Lawrie's meat science Seventh edition R. A. Lawrie in collaboration with D. A. Ledward 129 Modifying lipids for use in food Edited by F. Gunstone 130 Meat products handbook: practical science and technology G. Feiner 131 Food consumption and disease risk: consumer±pathogen interactions Edited by M. Potter 132 Acrylamide and other hazardous compounds in heat-treated foods Edited by K. Skog and J. Alexander 133 Managing allergens in food Edited by C. Mills, H. Wichers and K. HoffmanSommergruber 134 Microbiological analysis of red meat, poultry and eggs Edited by G. Mead 135 Maximising the value of marine by-products Edited by F. Shahidi 136 Chemical migration and food contact materials Edited by K. Barnes, R. Sinclair and D. Watson 137 Understanding consumers of food products Edited by L. Frewer and H. van Trijp 138 Reducing salt in foods: practical strategies Edited by D. Kilcast and F. Angus 139 Modelling microorganisms in food Edited by S. Brul, S. Van Gerwen and M. Zwietering



xxiii

140 Tamime and Robinson's Yoghurt: science and technology Third edition A. Y. Tamime and R. K. Robinson 141 Handbook of waste management and co-product recovery in food processing Volume 1 Edited by K. W. Waldron 142 Improving the flavour of cheese Edited by B. Weimer 143 Novel food ingredients for weight control Edited by C. J. K. Henry 144 Consumer-led food product development Edited by H. MacFie 145 Functional dairy products Volume 2 Edited by M. Saarela 146 Modifying flavour in food Edited by A. J. Taylor and J. Hort 147 Cheese problems solved Edited by P. L. H. McSweeney 148 Handbook of organic food safety and quality Edited by J. Cooper, C. Leifert and U. Niggli 149 Understanding and controlling the microstructure of complex foods Edited by D. J. McClements 150 Novel enzyme technology for food applications Edited by R. Rastall 151 Food preservation by pulsed electric fields: from research to application Edited by H. L. M. Lelieveld and S. W. H. de Haan 152 Technology of functional cereal products Edited by B. R. Hamaker 153 Case studies in food product development Edited by M. Earle and R. Earle 154 Delivery and controlled release of bioactives in foods and nutraceuticals Edited by N. Garti 155 Fruit and vegetable flavour: recent advances and future prospects Edited by B. BruÈckner and S. G. Wyllie 156 Food fortification and supplementation: technological, safety and regulatory aspects Edited by P. Berry Ottaway 157 Improving the health-promoting properties of fruit and vegetable products Edited by F. A. TomaÂs-BarberaÂn and M. I. Gil 158 Improving seafood products for the consumer Edited by T. Bùrresen 159 In-pack processed foods: improving quality Edited by P. Richardson 160 Handbook of water and energy management in food processing Edited by J. KlemesÏ, R. Smith and J-K Kim 161 Environmentally compatible food packaging Edited by E. Chiellini 162 Improving farmed fish quality and safety Edited by é. Lie 163 Carbohydrate-active enzymes Edited by K.-H. Park 164 Chilled foods: a comprehensive guide Third edition Edited by M. Brown 165 Food for the ageing population Edited by M. M. Raats, C. P. G. M. de Groot and W. A. Van Staveren 166 Improving the sensory and nutritional quality of fresh meat Edited by J. P. Kerry and D. A. Ledward 167 Shellfish safety and quality Edited by S. E. Shumway and G. E. Rodrick 168 Functional and speciality beverage technology Edited by P. Paquin 169 Functional foods: principles and technology M. Guo 170 Endocrine-disrupting chemicals in food Edited by I. Shaw 171 Meals in science and practice: interdisciplinary research and business applications Edited by H. L. Meiselman 172 Food constituents and oral health: current status and future prospects Edited by M. Wilson 173 Handbook of hydrocolloids Second edition Edited by G. O. Phillips and P. A. Williams 174 Food processing technology: principles and practice Third edition P. J. Fellows 175 Science and technology of enrobed and filled chocolate, confectionery and bakery products Edited by G. Talbot


xxiv


176 Foodborne pathogens: hazards, risk analysis and control Second edition Edited by C. de W. Blackburn and P. J. McClure 177 Designing functional foods: measuring and controlling food structure breakdown and absorption Edited by D. J. McClements and E. A. Decker 178 New technologies in aquaculture: improving production efficiency, quality and environmental management Edited by G. Burnell and G. Allan 179 More baking problems solved S. P. Cauvain and L. S. Young 180 Soft drink and fruit juice problems solved P. Ashurst and R. Hargitt 181 Biofilms in the food and beverage industries Edited by P. M. Fratamico, B. A. Annous and N. W. Gunther 182 Dairy-derived ingredients: food and neutraceutical uses Edited by M. Corredig 183 Handbook of waste management and co-product recovery in food processing Volume 2 Edited by K. W. Waldron 184 Innovations in food labelling Edited by J. Albert 185 Delivering performance in food supply chains Edited by C. Mena and G. Stevens 186 Chemical deterioration and physical instability of food and beverages Edited by L. Skibsted, J. Risbo and M. Andersen 187 Managing wine quality Volume 1: viticulture and wine quality Edited by A. G. Reynolds 188 Improving the safety and quality of milk Volume 1: milk production and processing Edited by M. Griffiths 189 Improving the safety and quality of milk Volume 2: improving quality in milk products Edited by M. Griffiths 190 Cereal grains: assessing and managing quality Edited by C. Wrigley and I. Batey 191 Sensory analysis for food and beverage control: a practical guide Edited by D. Kilcast 192 Managing wine quality Volume 2: oenology and wine quality Edited by A. G. Reynolds 193 Winemaking problems solved Edited by C. E. Butzke 194 Environmental assessment and management in the food industry Edited by U. Sonesson, J. Berlin and F. Ziegler 195 Consumer-driven innovation in food and personal care products Edited by S. R. Jaeger and H. MacFie 196 Tracing pathogens in the food chain Edited by S. Brul, P. M. Fratamico and T. A. McMeekin 197 Case studies in novel food processing technologies: innovations in processing, packing and predictive modelling Edited by C. J. Doona, K. Kustin and F. E. Feeherry 198 Freeze-drying of pharmaceutical and food products T.-C. Hua, B.-L. Liu and H. Zhang 199 Oxidation in foods and beverages and antioxidant applications Volume 1: understanding mechanisms of oxidation and antioxidant activity Edited by E. A. Decker, R. J. Elias and D. J. McClements 200 Oxidation in foods and beverages and antioxidant applications Volume 2: management in different industry sectors Edited by E. A. Decker, R. J. Elias and D. J. McClements 201 Protective cultures, antimicrobial metabolites and bacteriophages of food and beverage biopreservation Edited by C. Lacroix 202 Separation, extraction and concentration processes in the food, beverage and nutraceutical industries Edited by S. S. H. Rizvi 203 Determining mycotoxins and mycotoxigenic fungi in food and feed Edited by S. De Saeger 204 Developing children's food products Edited by D. Kilcast and F. Angus



xxv

205 Functional foods: concept to profit Second edition Edited by M. Saarela 206 Postharvest biology and technology of tropical and subtropical fruits Volume 1 Edited by E. M. Yahia 207 Postharvest biology and technology of tropical and subtropical fruits Volume 2 Edited by E. M. Yahia 208 Postharvest biology and technology of tropical and subtropical fruits Volume 3 Edited by E. M. Yahia 209 Postharvest biology and technology of tropical and subtropical fruits Volume 4 Edited by E. M. Yahia 210 Food and beverage stability and shelf-life Edited by D. Kilcast and P. Subramaniam 211 Processed meats: improving safety, nutrition and quality Edited by J. P. Kerry and J. F. Kerry 212 Food chain integrity: a holistic approach to food traceability, authenticity, safety and bioterrorism prevention Edited by J. Hoorfar, K. Jordan, F. Butler and R. Prugger 213 Improving the safety and quality of eggs and egg products Volume 1 Edited by Y. Nys, M. Bain and F. Van Immerseel 214 Improving the safety and quality of eggs and egg products Volume 2 Edited by Y. Nys, M. Bain and F. Van Immerseel 215 Feed and fodder contamination: effects on livestock and food safety Edited by J. Fink-Gremmels 216 Hygiene in the design, construction and renovation of food processing factories Edited by H. L. M. Lelieveld and J. Holah 217 Technology of biscuits, crackers and cookies Fourth edition Edited by D. Manley 218 Nanotechnology in the food, beverage and nutraceutical industries Edited by Q. Huang 219 Rice quality K. R. Bhattacharya 220 Meat, poultry and seafood packaging Edited by J. P. Kerry 221 Reducing saturated fats in foods Edited by G. Talbot 222 Handbook of food proteins Edited by G. O. Phillips and P. A. Williams 223 Lifetime nutritional influences on cognition, behaviour and psychiatric illness Edited by D. Benton


Preface

In providing a variety of food choices and on-the-go convenience to satisfy the sophisticated palate and needs of today's consumer, food processors and product developers regularly employ an extensive array of interesting processing and packaging technologies to ensure the safety, freshness, nutritive content, quality, and stability of foods. Such diversity may at first seem surprising, but consider the varieties of products available on the market, the nearly year-round availability of some products, and the high level of safety consumers routinely enjoy. Leafy green vegetables (e.g., spinach) are offered raw, frozen, canned, or washed, cut, and premixed as salad blends. Dairy products, too, can be frozen (ice cream), fermented (yogurt and cheeses), or treated to be stored refrigerated (heat pasteurized milk) or at ambient conditions (ultra-high temperature processed milk). At the same time, market forces and creative research and development continue to compel the development of new or enhanced products by advancing the frontiers of food science and furthering the applications of novel processing and packaging technologies; issues such as protecting the environment, reducing energy consumption, and decreasing the usage of water resources are also being given a higher priority. While science has the capacity to lead to safer, more convenient, and healthier foods using more eco-friendly technologies, it is also essential for exciting laboratory developments to find application, implementation, and even commercialization in the food industry. The collection of expert scientists, engineers, and technologists who have contributed chapters to Case studies in novel food processing technologies: Innovations in processing, packaging and predictive modelling demonstrate the continuing proliferation of innovation in the processing, packaging, and safety of foods, while also providing actual examples of real-life experiences involving the commercialization of food products using novel processes and predictive models.


xxviii Preface At a time when the United States has elected another President from Illinois, it is easy to hark back to the first President from that state. As an innovator with a strong interest in new technologies, Abraham Lincoln is the only President to hold a patent, and Lincoln delivered lectures on discoveries and inventions before he became President (available at http://showcase.netins.net/web/creative/ lincoln/education/patent.htm). Man is not the only animal who labors; but he is the only one who improves his workmanship (1858). [patent laws] secured to the inventor, for a limited time, the exclusive use of his invention; and thereby added the fuel of interest to the fire of genius, in the discovery and production of new and useful things (1859). During the Civil War, Lincoln took a personal interest in new technologies such as ironclad ships, observation balloons, breech-loading rifles, and machine guns. It may seem an odd juxtaposition to discuss innovations in weaponry during times of military warfare with case studies in novel food processing technologies, but it is important to remember that many scientific innovations and advances in food preservation came in support of the military. Nicholas Appert's innovations in canning in the early 1800s were developed for Napoleon who, along with Frederick the Great, realized that àn army marches on its stomach.' One chapter in this book presents a unique food safety model and an innovative accelerated 3-year microbial challenge study for a new enrobed breakfast sandwich product for military rations (Chapter 21), and a second chapter presents recent developments in the use of nanotechnology in the development of lighter weight, recyclable packaging for military rations. While Commander-in-Chief Lincoln may not have conceived of foods with 3year shelf-lives or the use of nanotechnology, one cannot help but imagine that Inventor Lincoln would certainly have been impressed by the intriguing scientific progress and technological advances communicated in these and other chapters. Non-thermal food pasteurization processes: an introduction by Chen et al. clearly establishes at the outset the interest that the food industry and consumers have in using non-thermal pasteurization processes (high pressure processing, HPP; pulse electric field, PEF; ionizing irradiation; UV light; non-thermal plasma, NTP; and concentrated high intensity electric field, CHIEF) to generate valueadded products of increased quality and nutrient retention, while being more energy efficient than traditional thermal processes. Part I, Case studies in high pressure and pulsed electric field processing of food, focuses on commercial applications of HPP and PEF, such as HPP-treated meat products by pioneers at Esteban EspunÄa, S.A., and HPP of fruit juices and juice-based products such as smoothies appearing in food markets around the world (Sampedro et al.). This section is further augmented by including Kempkes' chapter on PEF equipment for commercial juice processing, and Davis et al.'s analysis of the environmental impact of HPP and PEF processes using life cycle assessment.


Preface

xxix

Part II, Case studies in other novel food processing techniques, introduces a diverse range of novel and alternative processing techniques spanning a number of food, beverage, and industrial applications that are better taken together under a single, encompassing rubric. Niemira's chapter addresses a range of creative applications of conventional treatments and innovative approaches devised for fresh produce processing. In contrast, Rice presents an ozone-centric view; and why not? Ozone has found a number of commercial applications for more than a century, including its current uses for agri-foods, aqua-foods, and other food processing applications. Additionally, Part II will also help familiarize readers with the advantages and limitations of large-scale commercial applications of ultrasonics, and the commercialization of the first dense phase carbon dioxide processing system in the United States for orange juice pasteurization. Recent developments in novel infrared methods and in novel cool plasma techniques are also presented. Part III, Case studies in food preservation using antimicrobials, novel packaging and storage techniques, also includes a wide range of technologies, from a number of novel commercial applications of oxygen-depleted atmospheres for storing cereal grain commodities, to applications involving spraying natural preservatives onto the surface of finished baked goods. With respect to packaging technologies, Taoukis presents the commercialization of active food packaging using time-temperature integrators to better manage the food chain, to accompany the chapter on the use of nanotechnology for the packaging of military rations mentioned previously. As mentioned above, foods tend to have diverse characteristics, and food processing and packaging technologies are similarly diverse; there is no ònesize fits all' approach in food preservation. Part IV, Innovations in advanced food processing techniques and predictive microbial models: case studies, demonstrates that even established thermal methods will continue to be used, but might tend to evolve with technological advances (what Lincoln might call `Man's ability to improve his workmanship'). Such improvements include modifications to retorting technologies to achieve faster sterilization than is achievable with existing methods. In another chapter, three commercially successful innovative microwave food processes are presented, only one of which is still in operation (even innovation and commercial success are no guarantee of longevity). Major applications of irradiation around the world for the disinfestation of fresh fruits, and the use of irradiation in the beef industry, respectively, are featured in individual chapters. As the first may be last, and the last may be first, there will be no greater impact than Corradini and Peleg's chapter on establishing equivalent efficacy of thermal and nonthermal preservation processes, which is especially important considering the expanding usage of nonthermal processing technologies. Doona et al.'s chapter presents the Quasi-chemical model, a unique mathematical model that has evolved and adapted to evaluate a range of food safety applications. The instant application of the Quasi-chemical food safety model for the inactivation kinetics of Staphylococcus aureus is one major aspect of the chapter, and the other major


xxx

Preface

focus of the chapter is the development of a novel, accelerated 3-year microbial challenge study that saves time, money, and labor while ensuring the safety of military ration foods, themselves having evolved technologically perhaps far beyond anything a nineteenth-century soldier, or inventor-President, ever would have imagined. C. J. Doona K. Kustin F. E. Feeherry


1 Non-thermal food pasteurization processes: an introduction P. Chen, S. Deng, Y. Cheng, X. Lin, L. Metzger and R. Ruan, University of Minnesota, USA

Abstract: The food industry and consumers have significant interest in nonthermal pasteurization processes because they offer better quality and nutrition retention and are more energy efficient than traditional thermal processes. Non-thermal processes may also create value-added products and open new market opportunities. This chapter will provide an overview of several non-thermal processes with the potential for producing valued-added foods, including pulse electric field (PEF), high hydrostatic pressure (HHP), ionizing irradiation, UV light, non-thermal plasma (NTP), and concentrated high intensity electric field (CHIEF). Their respective mechanisms for inactivating microorganisms, technical characteristics, and current status of the application of these processes will be discussed. Key words: non-thermal pasteurization, pulse electric field (PEF), high hydrostatic pressure (HHP), ionizing irradiation, ultraviolet (UV) light, nonthermal plasma (NTP), and concentrated high intensity electric field (CHIEF).

1.1

Introduction

Many consumers enjoy the robust, natural flavor and taste of unpasteurized/raw apple juice or cider. However, due to associated outbreaks of foodborne illnesses, unpasteurized fruit juice has become mostly a thing of the past. In 1998, FDA adopted a regulation that forced fresh juice processors to either pasteurize their products to inactivate 5 logs of pathogenic microorganisms or attach the label `WARNING: this product has not been pasteurized and, therefore, may contain harmful bacteria which can cause serious illness in


2


children, the elderly and persons with weakened immune systems' (FDA, 1998). In 2001, FDA adopted the ruling to implement the Hazard Analysis and Critical Control Point (HAACP) procedures for the Safe and Sanitary Processing and Importing of Juice, effective February, 2002 (FDA, 2001). Apple juice production and consumption in the United States has been in decline for many years, which puts tremendous pressure on the fruit juice industry to boost consumption, while ensuring safety and retaining freshness and nutrients. In order to retain full flavor of their products, some companies have adopted a tight sanitation and HACCP program to achieve a 5-log reduction in production of unpasteurized apple juice/cider. A few producers even accept the warning label on some products, and others have combined `light' or ùltralight' pasteurization with HACCP, thus minimizing the decrement of flavor. However, most of the companies prefer to choose pasteurization to assure the safety of their products. Methods of pasteurization have changed from conventional treatments used in the past. Until recently, thermal processes, especially ultra high temperature (UHT) and high temperature short time (HTST) have been the most commonly used methods in the food industry to increase shelf-life and maintain food safety. However, studies have shown that heat degrades product color, flavor, and nutrients because of protein denaturation and the loss of vitamins and volatile flavors (Processors, 1998). Therefore, there is increasing demand for alternative methods for fresh food pasteurization that ensure safety while decreasing product degradation. Non-thermal methods provide such an option because they reduce overprocessing to result in more fresh-like foods featuring greater retention of color, flavor, and nutrients. Currently, there are several methods having the `nonthermal' claim for liquid food product pasteurization: (1) pulse electric field (PEF), (2) high hydrostatic pressure (HHP), (3) irradiation, and (4) UV light. Two emerging processes; namely, cold or non-thermal plasma (NTP), and concentrated high intensity electric field (CHIEF), are under development. In this chapter, we will provide a brief description of each of their mechanisms of microbial inactivation, technological characteristics, and current application status of these processes. Some of these alternative processes have been studied extensively for at least two decades, but none of these alternative processes is in large-scale commercial practice for fruit juice and milk pasteurization due to technical issues or, more often, economic disadvantages. The high resistance of enzymes and bacterial spores to these processes is a major problem. Efforts are needed to improve these processes or develop new processes. It is also suggested that combinations of these processes and other methods, which are termed `hurdle technology', may present potential benefits and practical uses of these processes.

1.2

Pulsed electric field

High intensity pulsed electric field (PEF) processing (Fig. 1.1) involves the application of short pulse (1±10 s) of high voltage (typically 20±80 kV/cm) to


Non-thermal food pasteurization processes: an introduction

Fig. 1.1

3

Pulsed electric field (PEF) process schematic diagram.

food materials located between two metal (usually stainless steel) electrodes (Qin et al., 1996; Vega-Mercado et al., 1997). Studies of exposure of microorganisms to electric fields have indicated that electric field can cause changes to cell membranes (Pothakamury et al., 1997; Barbosa-CaÂnovas et al., 1999). When a voltage is applied to a cell, a sufficiently high transmembrane potential is induced across the cell membrane, causing the membrane to rupture (direct mechanical damage, the electric breakdown theory), or destabilizing the lipid and proteins layers of cell membranes, resulting in pores (electroporation theory). The damaged cell membrane loses its selective semi-permeability, which allows water to enter the cell, and results in excessive cell volume swelling, and ultimately leads to cell rupture and inactivation of the organism. Some studies have provided microscopic evidence to support this theory (Harrison et al., 1997; CalderoÂn-Miranda et al., 1999). Recent studies showed increased membrane permeability after PEF treatment (Aronsson et al., 2005; GarcõÂa et al., 2007). PEF has been used to process fruit juices (Jin and Zhang, 1999), dairy products (Reina et al., 1998), and eggs (Dunn, 1996). Research found that apple juice processed with PEF at 50 kV/cm, 10 pulses, pulse width of 2 s, and initial temperature controlled at 45 ëC had a shelf-life of 28 d compared to a shelf-life of 21 d for untreated, fresh-squeezed apple juice. PEF processed apple juice showed no physical or chemical changes in ascorbic acid or sugar contents. PEF also demonstrated advantages over heat pasteurization for orange juice in terms of vitamin C, flavor, and color retention without inducing sedimentation like thermal treatments (Yeom et al., 2000). A majority of studies involving PEF focused on its effect on milk and dairy products due to the importance of the dairy industry. Model aqueous suspensions similar to milk ultrafiltrate, pasteurized milk, and raw milk have been used in those studies. Different levels of microbial inactivation were achieved with PEF treatment depending on the type of samples, type of microbe, the field strength, and the number of pulses applied during the process (Martin et al., 1997; Pothakamury et al., 1997; Bai-Lin et al., 1998; Qin et al., 1998). The inactivation of enzymes by PEF is limited, although the effect


4


of PEF on enzymes has been shown to vary with the electric field intensity, the number of pulses applied during the process, and the intrinsic characteristics of the enzyme (Bendicho et al., 2003; Kambiz et al., 2008). There are a limited number of studies on the effect of PEF on the nutrients and sensory quality of milk. Bendicho et al. (2002) found that PEF-treated milk showed no changes in the contents of most vitamins, except for ascorbic acid (Vitamin C), which reduced slightly. Grahl and MaÈrkl (1996) reported that the ascorbic acid content of milk was reduced considerably (90%, data not shown) by PEF treatment, whereas the content of vitamin A and the flavor showed no significant changes. Zulueta et al. (2007) reported that high intensity PEF treatment slightly changed the amounts of total fat, saturated fatty acids, monounsaturated fatty acids, and polyunsaturated fatty acids contained in an orange juice-milk beverage fortified with n-3 fatty acids and oleic acid; however, these changes were not considered significant. They concluded that changes in molecular composition of the orange juice-milk beverage were negligible from a nutritional viewpoint (Zulueta et al., 2007). Other research found that PEF has little influence on the physical, chemical and sensory properties of milk (Qin et al., 1995). Most of the recent studies on PEF tend to adopt an approach combining multiple factors such as the addition of heat (Craven et al., 2008; Noci et al., 2009; Riener et al., 2008; Wouters et al., 1999; Yu et al., 2009), antimicrobial compounds (Sobrino-Lopez and Martin-Belloso, 2008; Sobrino-Lopez et al., 2009), and thermosonication (Noci et al., 2009). Positive synergistic effects on bacteria and enzyme inactivation were demonstrated. Other research has also found PEF useful in facilitating the extraction of juice and other compounds from plant tissues (Bazhal et al., 2001; Fincan et al., 2004; Knorr and Angersbach, 1998; El-Belghiti and Vorobiev, 2005; Schilling et al., 2007). According to a fact sheet posted on The Ohio State University (OSU) Extension website (Ramaswamy et al., 2005), the first commercial scale continuous PEF system is located in OSU's Department of Food Science and Technology. Diversified Technologies Inc. (Bedford, MA) manufactures high voltage, high power pulse modulators, DC power supplies and control systems, builds commercial PEF systems with the PEF treatment chambers supplied by the Ohio State University. However, commercial applications of PEF in food pasteurization have been limited so far. There are a number of drawbacks in the application of PEF technology to foods. First, ohmic (electro-resistive) heating occurs during the PEF discharge, which causes the temperature of the sample to rise, and hence a cooling system has to be in place in order to maintain as closely as possible the initial, lower temperature of liquid samples. Therefore, a significant amount of energy is dissipated by the unwanted heating up and necessary cooling of the liquids. Second, since the electrodes have to be immersed in the liquid, they contribute a major source of contamination to the liquid food due to the erosion of the electrodes that occurs during discharge. Finally, the initial equipment investment is very capital-intensive and presents a major obstacle for the commercial application of PEF technology.



1.3

5

High hydrostatic pressure

High hydrostatic pressure (HHP) refers to the application of hydrostatic pressure ranging from 100 to over 800 MPa to foods for the purpose of inactivating spoilage and pathogenic microorganisms (Sangronis et al., 1997; Spilimbergo et al., 2002; Smelt, 1998). Liquid or solid food, with or without packaging, is placed in the pressure vessel (Fig. 1.2). The closed pressure vessel is filled with a pressure transmitting fluid, which is usually water or a dilute aqueous solution. The liquid is compressed usually by a pump or pressure intensifier, and the hydrostatic pressure distributes uniformly throughout the pressure vessel and equally in all directions of the food surfaces. Treatment times, once constant high pressure levels are achieved, can range from a millisecond pulse to over 20 min, and initial treatment temperatures can range from 0 to 90 ëC. HHP generally produces better results for the pasteurization of foods when combined with initial temperatures around 45±50 ëC.

Fig. 1.2

High hydrostatic pressure (HHP) process schematic diagram.


6


A number of mechanisms of microbial inactivation by HHP have been proposed. HHP is believed to cause pressure sensitive non-covalent bonds (hydrogen, ionic, and hydrophobic bonds) to break. Since non-covalent bonds are present chiefly in large molecules such as proteins, polysaccharides, lipids, and nucleic acids, the breakdown of non-covalent bonds will lead to significant damage to enzymes, membranes, and genetic molecules of microbes, therefore inhibiting the metabolism, growth, and reproduction of microorganisms. The disruption of membranes and inactivation of enzymes, including those responsible for DNA replication and transcription, are believed to be the main mechanisms of pressure-induced microbial injury and death (Hoover et al., 1989). HHP is used widely for the pasteurization of post-processing refrigerated salads and entreÂes, avocado fruit, apple sauce, ham, and oysters. HHP also appears to be a very attractive method for pasteurization of fruit juices and milk products. In fact, HHP processed fruit juices are commercially available in Japan. Some researchers reported 5±7 log reductions of bacteria in apple and other fruit juices (Jordan et al., 2001). Linton et al. (2001) used HHP (200±700 MPa for 15 min at 20 ëC) to process UHT skim milk inoculated with different strains of E. coli. The strains varied in their sensitivity to HHP. The least sensitive strains could be completely inhibited at 600 MPa for 30 min. Gao et al. (2006) demonstrated 6-log cycle reduction of L. monocytogenes in milk at 448 MPa, 41 ëC, and 11 min. In addition to inactivating foodborne microorganisms to prevent spoilage or to ensure food safety, there is also interest in understanding the effects of HHP on proteins in milk. Johnston et al. (1992) studied the extent of conformational and other changes in skim milk proteins caused by the application of HHP at pressures less than 600 MPa. They found that pH and Ca2+ ion activity were unaffected, but the lightness of the color (L*) of milk decreased by HHP treatments less than 300 MPa. HHP treatments caused interior hydrophobic groups to become exposed, indicating irreversible unfolding of proteins. There are also studies to demonstrate the combined effects of HHP, mild heat treatment, and antimicrobial compounds (Patterson and Kilpatrick, 1998; Garcia-Graells et al., 1999; Haiqiang and Hoover, 2003; Gao et al., 2006; Bilbao-Sainz et al., 2009). HHP is less efficient for inactivating bacterial spores in low acid foods, requiring relatively higher initial temperatures (~90 ëC) to achieve sterilization temperatures and inactivate resistant endospores during processing. While HHP preserves the sensory quality and nutritional value of liquid foods to create a significant advantage for commercial products, equipment costs are capital-intensive and are only available as batch processes, making HHP an unlikely alternative to conventional pasteurization methods for low value foods for the near future.

1.4

Ionizing irradiation

Food irradiation involves the use of ionizing radiations to inactivate spoilage and pathogenic microorganisms, control insects and parasites, and inhibit postharvest ripening and sprouting (Sendra et al., 1996; Zehnder, 1988; Burditt,



7

1982). Radiation sources may include radioactive isotopes of cobalt or cesium or beta rays or x-rays from electron accelerators. During the radiation processing of foods, ionizing radiation may directly damage sensitive macromolecules, such as DNA, making the organisms unable to replicate or reproduce. Irradiation may also generate free radicals which react with molecular constituents in cells, and cause irreversible lethal damage to the cells. The effectiveness of food irradiation is dosage dependent. Insects and parasites may be killed at low dosages under 0.1 kiloGray (kGy). Medium doses between 1.5 and 4.5 kGy are needed to inactivate most bacterial pathogens. Inactivating bacterial spores and viruses requires doses higher than 10±45 kGy. The first commercial application of irradiation in food preservation took place in 1957 in Germany, when a spice manufacturer irradiated their products with electrons to improve the hygienic quality. Forty-two countries in the world currently have approved the use of ionizing irradiation on more than 100 foods. In the US, irradiation of some fruits and vegetables, poultry, beef, pork, and lamb has been approved by the FDA. Irradiation of papayas to kill fruit flies in Hawaii is a successful example of the use of irradiation to control the spread of insects via agricultural exports. In the US, irradiation of meat and meat products requires prior approval not only by the FDA, but also the USDA's Food Safety and Inspection Service. Protein rich foods such as milk are not suited for pasteurization by irradiation because irradiation may induce off-flavor, odor, and discoloration. There were earlier studies on the irradiation of liquid milk to extend shelf-life. Naguib et al. (1974) showed that three species of Brucella (Br. abortus, Br. melitensis and Br. Suis) in skim-milk (approximately 109 organisms/mL) were completely destroyed by exposure to gamma irradiation at dosages greater than 200 Krad. An early study by Scanlan and Lindsay (1967) found the irradiation of liquid milk with a dose of 4.5 Mrad promoted extreme browning and caramelization. When the milk was irradiated in the frozen state at ÿ80 and ÿ185 ëC an extremely bitter flavor resulted. All of the irradiated milk samples were regarded as unacceptable by flavor assessment. Fractionating of the irradiated milk separated the bitter flavor and suggested that the bitter component was a protein or a non-dialyzable protein fragment. A more recent study by Naghmoush et al. (1983) showed more favorable results. These researchers treated raw milk from cows, buffaloes or goats with 0.25±0.75 Mrad of gamma-irradiation. The treated samples showed noticeable bacterial count and spore count reduction compared with untreated controls. However, the irradiation treatment decreased the nutritional content (specifically, carotene and vitamin A) and flavor of all three types of milk, and samples became progressively more oxidized with increasing radiation dose; the initially yellowish cows' milk became progressively whiter. Although irradiation has gained substantial media attention and has been approved for use on a broad range of foods, consumers still worry about the safety of irradiated food products, and the acceptance of irradiated foods grows, albeit very slowly. The pace of growth varies by country, as different countries have different prevailing consumer attitudes, regulations, and enforcement.


8


Further research on issues such as the radiation resistance of different microorganisms and strains, irradiation-induced chemical reactions in foods, the influence of irradiation on sensory and nutritional losses, the use of irradiation in combination with other hurdle techniques, the interaction with packaging materials, and consumer education and safety awareness will continue to guide the development of food irradiation.

1.5

Ultraviolet radiation

Ultraviolet (UV) processing uses radiation from the UV region of the electromagnetic spectrum to inactivate microorganisms in foods, water, and packaging materials (Koutchma, 2009). The typical wavelength for UV processing ranges from 100 to 400 nm. UV in the 200±280 nm region is believed to be most effective in inactivating bacteria and viruses. During UV irradiation, DNA molecules absorb UV light, which causes crosslinking between neighboring pyrimidine nucleoside bases in the same DNA strand, and this mutation in the DNA basepairing results in hindered growth and reproduction (Miller et al., 1999). To inactivate microorganisms effectively, a minimum of 400 J/m2 energy in all parts of the product being irradiated must be obtained through UV treatment. The effectiveness of a UV process is a function of `the transmissivity of the product, the geometric configuration of the reactor, the power, wavelength and physical arrangement of the UV source(s), the product flow profile, and the radiation path length' (FDA, 2009). UV light lacks penetration capability, and therefore lends itself best to surface treatments. The treatment is more efficient for liquid foods that are pre-filtered or clarified. To enhance the lethality of UV treatments for inactivating microorganisms, the UV may be used in combination with other alternative processing technologies, including strong chemical oxidizing agents such as ozone and hydrogen peroxide. There is considerable interest in using UV irradiation for the pasteurization of milk (Munkacsi and Elhami, 1976; Caserio et al., 1978; Bodurov et al., 1979; Filipov, 1979, 1981; Giraffa and Carini, 1984; Ibarz et al., 1986; Yu et al., 1999; Smith et al., 2002; Chernyh and Yurova, 2006; Matak et al., 2007). Yu et al. (1999) studied the effects of UV radiation time, distance from the UV source, thickness of the treated milk sample during processing, and temperature. They reported that there was a critical thickness of the liquid milk sample, which is apparently limited by the penetrability of UV light. Their study also showed that varying temperature in the range of 0±37 ëC did not significantly influence the pasteurization process. Smith et al. (2002) reported that samples from dairy bulk tanks of milk showed no bacterial growth after the samples were exposed to UV light (248 nm) at a dose of 12.6 J/cm2. There are some negative effects of UV radiation on milk quality, such as the deterioration of flavor due to an increase in thiobarbituric acid reactive substances and acid degree values which are related to chemical oxidation and hydrolytic rancidity (Matak et al., 2007), delayed



9

rennet coagulation of milk and the development of acidity, and the induction of a slight cooked flavor (Munkacsi and Elhami, 1976). Although there is interest in potentially applying UV radiation to control microbes in liquid milk (Smith et al., 2002), and notwithstanding the fact that FDA has approved the use UV radiation for the treatment of water and food under specific conditions (CFR, 2005), the commercial application of UV radiation for food pasteurization is presently unavailable.

1.6

Non-thermal plasma

Non-thermal plasma (NTP) is electrically energized matter in a gaseous state, and can be generated by passing gases through electric fields (Conrads and Schmidt, 2000). The mean electron energies of NTP, which is about 20 eV, are considerably higher than those of the components of the ambient gas. During NTP generation, the majority of the electrical energy goes into the production of energetic matters rather than into gas heating. The energy in NTP is thus directed preferentially to the electron-impact dissociation and ionization of the background gas to produce NTP species including electrically neutral gas molecules, charged particles in the form of positive ions, negative ions, free radicals and electrons, and quanta of electromagnetic radiation (photons) (Ulrich, 2007). Theses species are very strong oxidizers that can rapidly decompose other inorganic and organic compounds. Plasma may kill both vegetative cells and bacterial endospores. The killing mechanisms of NTP are not well established; however, there are some hypotheses. It is well documented that reactive oxygen species (ROS) such as oxygen radicals can produce profound effects on cells by reacting with various macromolecules (Kelly-Wintenberg et al., 1999; Mounir, 2005). Among the cellular macromolecules altered are membrane lipids (Montie et al., 2000), which exhibit sensitivity probably because of their location at the cell surface and their susceptibility to oxidation by ROS. Altering the cytoplasmic membrane lipids results in a release of intracellular substances and the death of cells. Philip et al. (2002) proposed three basic mechanisms, individually or synergistically, responsible for the inactivation of microbial spores in plasma environments. These mechanisms include: (1) destruction of DNA by UV irradiation, (2) volatilization of compounds from the spore surface by UV photons and (3) erosion, or so-called ètching,' of the spore surface by adsorption of reactive species like free radicals. NTP has been used mostly for water and wastewater treatment, surface sterilization and environmental control (Montie et al., 2000; Mounir, 2005; Ma et al., 2001a, 2001b, 2001c; Ruan and Chen, 2000; Ruan et al., 1999a, 1999b, 2000; Ashikov et al., 2008). There is increasing interest in using NTP to inactivate vegetative foodborne pathogens on different surfaces, such as thin films of agar (Kayes et al., 2007), heat sensitive polyethylene terephthalate (PET) foils (Muranyi et al., 2007), polycarbonate membranes (Yu et al., 2006),


10


culture media (Ashikov et al., 2008), and almond (Deng et al., 2007). Logreductions ranging from 1 to 7 were reported for vegetative pathogens in these studies. There are relatively few reports on the use of NTP for the pasteurization of liquids (Anpilov et al., 2002; Ashikov et al., 2008). Ruan and co-workers began their research on the use of NTP for liquid food pasteurization in the early 2000s (Ruan et al., 2003a, 2003b, 2003c; Montenegro et al., 2002). The dielectric barrier discharge NTP systems they developed (Fig. 1.3) were able to produce up to 6-log reductions of E. coli 25922 in water and juice. These systems use an AC power supply and are cheap to construct. They also designed reactors that incorporate a bubbling mechanism to enhance NTP discharges for treating liquids (Fig. 1.4), and they devised a system for processing solid foods (Figs 1.5 and 1.6). Even with these successes, the use of

Fig. 1.3

Fig. 1.4

Non-thermal plasma (NTP) process schematic diagram.

Non-thermal plasma (NTP) reactor designed for liquid treatment.



11

Fig. 1.5 Non-thermal plasma (NTP) reactor for solid foods (e.g., almond) treatment.

Fig. 1.6

Prototype of a non-thermal plasma (NTP) system for dry fresh almond pasteurization.


12


these systems for milk pasteurization encountered difficulties associated chiefly with the poor penetration of NTP species. These processing systems eventually evolved to become concentrated high intensity electric field systems, as described in the following section.

1.7

Concentrated high intensity electric field

Concentrated high intensity electric field (CHIEF) is a new process developed by researchers at the University of Minnesota (Ruan et al., 2008). The process uses a unique treatment chamber (orifice) and electrode configuration where a high intensity electric field is concentrated within the orifice through which liquid is pasteurized. CHIEF bears characteristics similar to those of a dielectric barrier NTP system, which consists of two electrodes separated by two layers of dielectric materials and driven by AC power. However, the mechanisms of microbial inactivation by CHIEF resemble more closely those of PEF than of NTP. In comparison with PEF technology, CHIEF has some unique characteristics: · it is powered by low and medium frequency alternate current (AC) power instead of high frequency pulsed direct current (DC) power, and thus requires significantly lower capital investment; · it uses a non-metal (dielectric) barrier to limit electric current flow through the liquid to eliminate ohmic (electro-resistive) heating, thereby reducing the temperature rise and avoiding contamination from the oxidation, corrosion, and erosion of metal electrodes, which occurs commonly with conventional PEF methods, and the need to change electrodes periodically; · it uses a unique configuration design which significantly improves energy efficiency by directing voltage (electric field strength) to the treated liquid, instead of dissipating energy in the electrodes and dielectric barriers. Our recent studies have demonstrated a 7-log reduction of E. coli 0157 inoculated in orange juice and a 5-log reduction of E. coli 25922 inoculated in milk. In these processes, the temperature rise is minimal (from 16 to 50 ëC), and there was no significant physical and chemical changes observed. We consider the CHIEF process to be one of the most promising, and perhaps the leading technology for the non-thermal pasteurization of fresh milk.

1.8

Conclusions

Six non-thermal food processes of industrial and academic significance were reviewed. Most of these processes are still under development. Among these processes, ionizing irradiation is the most mature technology. Further efforts to address some technical issues and to increase consumer acceptance through education and government safety regulations are expected to increase the



13

commercial use of food irradiation. Some commercial PEF and HHP systems are available, but the initial capital investments are costly, and sometimes prohibitively costly, limiting the technologies to high value products. UV is limited by its low penetration capacity. No commercial UV system is available for food processing. NTP and CHIEF are emerging processes with great potential. Substantial research is needed to understand the processes and their limitations, to develop cost effective processing procedures and equipment, and to collect scientific data on a broad spectrum of microbes, enzymes and food systems.

1.9

References

ANPILOV, A. M., BARKHUDAROV, E. M., CHRISTOFI, N., KOP'EV, V. A., KOSSYI, I. A.,

and ZADIRAKA, Y. (2002) Pulsed high voltage electric discharge disinfection of microbially contaminated liquids. Letters in Applied Microbiology, 35, 90±94. È NNER, U. and BORCH, E. (2005) Inactivation of Escherichia coli, Listeria ARONSSON, K., RO innocua and Saccharomyces cerevisiae in relation to membrane permeabilization and subsequent leakage of intracellular compounds due to pulsed electric field processing. International Journal of Food Microbiology, 99, 19±32. TAKTAKISHVILI, M. I.

ASHIKOV, Y., GRUSHIN, M., TRUSHKIN, N., KHOLODENKO, V., CHUGUNOV, V., KOBZEV, E., SHIRKOVA, N., IRKHINA, I. and KIREEV, G. (2008) Atmospherics-pressure non-thermal plasma sterilization of microorganisms in liquids and on surfaces. Pure and Applied Chemistry, 80(9), 1953±1958. BAI-LIN, Q., BARBOSA-CANOVAS, G. V., SWANSON, B. G., PEDROW, P. D. and OLSEN, R. G. (1998) Inactivating microorganisms using a pulsed electric field continuous treatment system. Industry Applications, IEEE Transactions on, 34, 43±50. Â NGORA-NIETO, M. M., POTHAKAMURY, U. R. and SWANSON, B. BARBOSA-CAÂNOVAS, G. V., GO (1999) Preservation of foods with pulsed electric fields. In Taylor, S. L. (ed.) Food science and technology international series. San Diego, CA, Academic Press. BAZHAL, M. I., LEBOVKA, N. I. and VOROBIEV, E. (2001) Pulsed electric field treatment of apple tissue during compression for juice extraction. Journal of Food Engineering, 50, 129±139. BENDICHO, S., ESPACHS, A., ARANTEGUI, J. and MARTIN, O. (2002) Effect of high intensity pulsed electric fields and heat treatments on vitamins of milk. Journal of Dairy Research, 69, 113±123. BENDICHO, S., BARBOSA-CANOVAS, G. V. and MARTIN, O. (2003) Reduction of protease activity in milk by continuous flow high-intensity pulsed electric field treatments. Journal of Dairy Science, 86, 697±703. BILBAO-SAINZ, C., YOUNCE, F. L., RASCO, B. and CLARK, S. (2009) Protease stability in bovine milk under combined thermal-high hydrostatic pressure treatment. Innovative Food Science & Emerging Technologies, 10, 314±320. BODUROV, N., FILIPOV, Z. and PETROVA, L. (1979) Changes in vitamin D content of cows' milk irradiated in vitro with ultraviolet rays, and of blood serum of calves fed on irradiated milk. Veterinarnomeditsinski Nauki, 16, 35±40. BURDITT, A. K., JR. (1982) Food irradiation as a quarantine treatment of fruits. Food Technology, 36, 51±54.


14

Case studies in novel food processing technologies and SWANSON, B. G. (1999) Transmission electron microscopy of Listeria innocua treated by pulsed electric fields and nisin in skimmed milk. International Journal of Food Microbiology, 51, 31± 38.

CALDEROÂN-MIRANDA, M. L., BARBOSA-CAÂNOVAS, G. V.

CASERIO, G., GENNARI, M., BERETTA, G., BIANCHI, M. A., BARLUZZI, A. M., CALIGARA, S., MALACRIDA, M. and GHILARDI, A. (1978) Effects of ultraviolet radiation on the nutritive and organoleptic properties of milk for cheesemaking. Industria del Latte, 14, 3±20. CFR (2005) Code of Federal Regulations. CFR. US Government Printing Office, Washington DC. CHERNYH, E. A. and YUROVA, E. A. (2006) Effect of ultraviolet on milk composition and properties. Molochnaya Promyshlennost, 7. CONRADS, H. and SCHMIDT, M. (2000) Plasma generation and plasma sources. Plasma Sources Science and Technology, 441. CRAVEN, H. M., SWIERGON, P., NG, S., MIDGELY, J., VERSTEEG, C., COVENTRY, M. J. and WAN, J. (2008) Evaluation of pulsed electric field and minimal heat treatments for inactivation of pseudomonads and enhancement of milk shelf-life. Innovative Food Science & Emerging Technologies, 9, 211±216. DENG, S., RUAN, R., MOK, C., HUANG, G., LIN, X. and CHEN, P. (2007) Inactivation of Escherichia coli on almonds using nonthermal plasma. Journal of Food Science, 72, M62±M66. DUNN, J. (1996) Pulsed light and pulsed electric field for foods and eggs. Poultry Science, 75, 1133±1136. EL-BELGHITI, K. and VOROBIEV, E. (2005) Modelling of solute aqueous extraction from carrots subjected to a pulsed electric field pre-treatment. Biosystems Engineering, 90, 289±294. FDA (1998) Food Labeling: Warning and Notice Statement: Labeling of Juice Products; Final Rule. FDA (2001) Hazard Analysis and Critical Control Point (HAACP); Procedures for the Safe and Sanitary Processing and Importing of Juice; Final Rule. Federal Register. FDA (2009) Kinetics of Microbial Inactivation for Alternative Food Processing Technologies. In FDA (ed.). FILIPOV, Z. (1979) Effect of ultraviolet irradiation of milk on contents of some macroelements, alkaline phosphatase and vitamin D of blood serum and on bone density in calves. Veterinarnomeditsinski Nauki, 16, 26±33. FILIPOV, Z. (1981) Optimal dose of ultraviolet irradiation of microorganisms in cows' milk. Veterinarnomeditsinski Nauki, 18, 64±68. FINCAN, M., DEVITO, F. and DEJMEK, P. (2004) Pulsed electric field treatment for solid-liquid extraction of red beetroot pigment. Journal of Food Engineering, 64, 381±388. GAO, Y. L., XING-RONG, J. and HAN-HU, J. (2006) Statistical analysis of inactivation of Listeria monocytogenes subjected to high hydrostatic pressure and heat in milk buffer. Applied Microbiology & Biotechnology, 70, 670±678. GARCIA-GRAELLS, C., MASSCHALCK, B. and MICHIELS, C. W. (1999) Inactivation of Escherichia coli in milk by high-hydrostatic-pressure treatment in combination with antimicrobial peptides. Journal of Food Protection, 62, 1248±1254. Â MEZ, N., MANÄAS, P., RASO, J. and PAGA Â N, R. (2007) Pulsed electric fields cause GARCIÂA, D., GO bacterial envelopes permeabilization depending on the treatment intensity, the treatment medium pH and the microorganism investigated. International Journal of Food Microbiology, 113, 219±227.



15

and CARINI, S. (1984) Ultraviolet and infrared utilization for the microbiological safety of milk. Microbiologie Aliments Nutrition, 2, 91±93. È RKL, H. (1996) Killing of microorganisms by pulsed electric fields. GRAHL, T. and MA Applied Microbiology & Biotechnology, 45, 148±157. HAIQIANG, C. and HOOVER, D. G. (2003) Modeling the combined effect of high hydrostatic pressure and mild heat on the inactivation kinetics of Listeria monocytogenes Scott A in whole milk. Innovative Food Science & Emerging Technologies, 4, 25±34. HARRISON, S. L., BARBOSA-CAÂNOVAS, G. V. and SWANSON, B. G. (1997) Saccharomyces cerevisiae structural changes induced by pulsed electric field treatment. Lebensmittel-Wissenschaft und -Technologie, 30, 236±240. HOOVER, D. G., METRICK, C., PAPINEAU, A. M., FARKAS, D. F. and KNORR, D. (1989) Biological effects of high hydrostatic pressure on food microorganisms. Food Technology, 43, 99±107. IBARZ, A., PAGAN, J. and VICENTE, M. (1986) Effect of ultraviolet radiation on raw milk. Alimentacion Equipos y Tecnologia, 5, 95±101. JIN, Z. T. and ZHANG, Q. H. (1999) Pulsed electric field inactivation of microorganisms and preservation of quality of cranberry juice. Journal of Food Processing & Preservation, 23, 481±497. JOHNSTON, D. E., AUSTIN, B. A. and MURPHY, R. J. (1992) Effects of high hydrostatic pressure on milk. Milchwissenschaft, 47, 760±763. JORDAN, S. L., PASCUAL, C., BRACEY, E. and MACKEY, B. M. (2001) Inactivation and injury of pressure-resistant strains of Escherichia coli O157 and Listeria monocytogenes in fruit juices. Journal of Applied Microbiology, 91, 463±469. KAMBIZ, S., VERSTEEG, C., SHERKAT, F. and WAN, J. (2008) Alkaline phosphatase and microbial inactivation by pulsed electric field in bovine milk. Innovative Food Science & Emerging Technologies, 9, 217±223. KAYES, M. M., CRITZER, F. J., KELLY-WINTENBERG, K., ROTH, J. R., MONTIE, T. C. and GOLDEN, D. A. (2007) Inactivation of foodborne pathogens using a one atmosphere uniform glow discharge plasma. Foodborne Pathogens and Disease, 4, 50±59. GIRAFFA, G.

KELLY-WINTENBERG, K., HODGE, A., MONTIE, T., DELEANU, L., SHERMAN, D., ROTH, J., TSAI, P.

and WADSWORTH, L. (1999) Use of a one atmosphere uniform glow discharge plasma to kill a broad spectrum of microorganisms. J. Vac. Sci. Technol, 17, 1539± 1544. KNORR, D. and ANGERSBACH, A. (1998) Impact of high-intensity electric field pulses on plant membrane permeabilization. Trends in Food Science & Technology, 9, 185± 191. KOUTCHMA, T. (2009) Advances in ultraviolet light technology for non-thermal processing of liquid foods. Food & Bioprocess Technology, 2, 138±155. LINTON, M., MCCLEMENTS, J. M. J. and PATTERSON, M. F. (2001) Inactivation of pathogenic Escherichia coli in skimmed milk using high hydrostatic pressure. Innovative Food Science & Emerging Technologies, 2, 99±104. MA, H., CHEN, P. and RUAN, R. (2001a) H2S and NH3 removal by silent discharge plasma and ozone combo system. Plasma Chemistry and Plasma Processing, 21, 611±624. MA, H., CHEN, P., ZHANG, M., LIN, X. and RUAN, R. (2001b) Study of SO2 removal using nonthermal plasma induced by dielectric barrier discharge (DBD). Plasma Chemistry and Plasma Processing, 22, 239±254. MA, H., CHEN, P., ZHANG, M., LIN, X. and RUAN, R. (2001c) Treatment of H2S and NH3 using ozone and silent discharge plasma. Proceeding of International Conference on Agricultural Science & Technology. Beijing, China.


16


and SWANSON, B. G. (1997) Inactivation of Escherichia coli in skim milk by high intensity pulsed electric fields. Journal of Food Process Engineering, 20, 317±336. MATAK, K. E., SUMNER, S. S., DUNCAN, S. E., HOVINGH, E., WOROBO, R. W., HACKNEY, C. R. and PIERSON, M. D. (2007) Effects of ultraviolet irradiation on chemical and sensory properties of goat milk. Journal of Dairy Science, 90, 3178±3186. MILLER, R., JEFFREY, W., MITCHELL, D. and ELASRI, M. (1999) Bacterial responses to ultraviolet light. Am Soc Microbiol, 65(8), 535±541. MONTENEGRO, J., RUAN, R., MA, H. and CHEN, P. (2002) Inactivation of E. coli O157:H7 using a pulsed nonthermal plasma system. Journal of Food Science, 67, 646±648. MONTIE, T. C., KELLY-WINTENBERG, K. and ROTH, J. R. (2000) An overview of research using the One Atmosphere Uniform glow discharge Plasma (OAUGDP) for sterilization of surfaces and materials: Nonthermal medical/biological treatments using electromagnetic waves and ionized gases. IEEE Transactions on Plasma Science, 28, 41±50. MOUNIR, L. (2005) Low temperature plasma-based sterilization: overview and state-of-theart. Plasma Processes and Polymers, 2, 391±400. MUNKACSI, F. and ELHAMI, M. (1976) Effect of ultrasonic and ultraviolet irradiation on the chemical and bacteriological quality of milk. Egyptian Journal of Dairy Science, 4, 1±6. MURANYI, P., WUNDERLICH, J. and HEISE, M. (2007) Sterilization efficiency of a cascaded dielectric barrier discharge. Journal of Applied Microbiology, 103, 1535±1544. NAGHMOUSH, M. R., SHEHATA, T. E., GIRGIS, E. S., KAMAL, T. H. and IBRAHIM, M. K. (1983) Gamma irradiation effect of some bacteriological and other properties of cow's, buffaloes and goat's milk. Egyptian Journal of Dairy Science, 11, 61±69. NAGUIB, M. M., NOUR, M. and SALEM, A. A. (1974) Effect of gamma-irradiation on the survival of Brucella species in milk. Egyptian Journal of Dairy Science, 2, 93±99. NOCI, F., WALKLING-RIBEIRO, M., CRONIN, D. A., MORGAN, D. J. and LYNG, J. G. (2009) Effect of thermosonication, pulsed electric field and their combination on inactivation of Listeria innocua in milk. International Dairy Journal, 19, 30±35. PATTERSON, M. F. and KILPATRICK, D. J. (1998) The combined effect of high hydrostatic pressure and mild heat on inactivation of pathogens in milk and poultry. Journal of Food Protection, 61, 432±436. PHILIP, N., SAOUDI, B., CREVIER, M.-C., MOISAN, M., BARBEAU, J. and PELLETIER, J. (2002) The respective roles of UV photons and oxygen atoms in plasma sterilization at reduced gas pressure: the case of N2-O2 mixtures. IEEE Transactions on Plasma Science, 30, 1429±1436. POTHAKAMURY, U. R., BARBOSA-CANOVAS, G. V., SWANSON, B. G. and SPENCE, K. D. (1997) Ultrastructural changes in Staphylococcus aureus treated with pulsed electric fields/Cambios ultraestructurales en Staphylococcus aureus sometida a campos electricos pulsantes. Food Science and Technology International, 3, 113±121. PROCESSORS, A. H. J. C. (1998) Fresh Unpasteurized Apple Juice/Cider Quality Assurance Plan. QIN, B. L., POTHAKAMURY, U. R., VEGA, H., MARTIN, O., BARBOSA-CANOVAS, G. V. and SWANSON, B. G. (1995) Food pasteurization using high-intensity pulsed electric fields. Food Technology, 49, 55±60. QIN, B. L., POTHAKAMURY, U. R., BARBOSA-CANOVAS, G. V. and SWANSON, B. G. (1996) Nonthermal pasteurization of liquid foods using high-intensity pulsed electric fields. Critical Reviews in Food Science & Nutrition, 36, 603±627. MARTIÂN, O., QIN, B. L, CHANG, F. J, BARBOSA-CAÂNOVAS, G. V.



17

and OLSEN, R. G. (1998) Inactivating microorganisms using a pulsed electric field continuous treatment system. Industry Applications, IEEE Transactions on, 34, 43±50. RAMASWAMY, R., JIN, T. and ZHANG, H. (2005) Pulsed Electric Field Processing Fact Sheet for Food Processors (FSE-2-05) [Online]. Columbus. Available: http:// ohioline.osu.edu/fse-fact/0002.html. REINA, L. D., JIN, Z. T., ZHANG, Q. H. and YOUSEF, A. E. (1998) Inactivation of Listeria monocytogenes in milk by pulsed electric field. Journal of Food Protection, 61, 1203±1206. RIENER, J., NOCI, F., CRONIN, D. A., MORGAN, D. J. and LYNG, J. G. (2008) Combined effect of temperature and pulsed electric fields on apple juice peroxidase and polyphenoloxidase inactivation. Food Chemistry, 109, 402±407. RUAN, R. and CHEN, P. (2000) Non-thermal plasma demo system for odor control. AFTA Spring 2000 Proceedings. EPRI Publication. RUAN, R., HAN, W., NING, A., CHEN, P., GOODRICH, P. and ZHANG, R. (1999a) Treatment of odorous and hazardous gases using non-thermal plasma. J. Advanced Oxidation Technologies, 4, 328±332. RUAN, R., HAN, W., NING, A., DENG, S., CHEN, P. and GOODRICH, P. (1999b) Effects of design parameters of planar silent discharge plasma reactors on gaseous ammonia reduction. Transactions of ASAE, 42, 1841±1845. RUAN, R., CHEN, P., NING, A., DENG, S., MA, H., WANG, Y. and GOODRICH, P. (2000) Non-thermal plasma reactor for decomposition of odorous gases. US Patent. RUAN, R. R., HONGBIN, M., CHEN, P. L., SHAOBO, D. and XIANGYANG, L. (2003a) Method and apparatus for non-thermal pasteurization of living-mammal-instillable liquids. US Patent Application 0180421. RUAN, R. R., HONGBIN, M., MINGLIANG, Z., CHEN, P. L. and OYEN, D. (2003b) Method and apparatus for non-thermal pasteurization. US Patent. RUAN, R. R., HONGBIN, M., MINGLIANG, Z., CHEN, P. L. and OYEN, D. (2003c) Method and apparatus for non-thermal pasteurization. US Patent Application 0026877. RUAN, R., DENG, S., CHEN, L., LIN, X. and METZGER, L. (2008) Dielectric barrier reactor having concentrated electric field. US Patent Application 20080099406. 05/01/2008. SANGRONIS, E., POTHAKAMURY, U., RAMOS, A. M., IBARZ, A., BARBOSA-CANOVAS, G. V. and SWANSON, B. G. (1997) High hydrostatic pressure: an alternative in non-thermal treatment of foods. Alimentaria, 283, 33±43. SCANLAN, R. A. and LINDSAY, R. C. (1967) Observations on the low-temperature irradiation of milk. Journal of Dairy Science, 51, 1967±1968. SCHILLING, S., ALBER, T., TOEPFL, S., NEIDHART, S., KNORR, D., SCHIEBER, A. and CARLE, R. (2007) Effects of pulsed electric field treatment of apple mash on juice yield and quality attributes of apple juices. Innovative Food Science & Emerging Technologies, 8, 127±134. SENDRA, E., CAPELLAS, M., GUAMIS, B., FELIPE, X., MOR-MUR, M. and PLA, R. (1996) Food irradiation ± general aspects. Food Science & Technology International/Ciencia y Tecnologia de Alimentos Internacional, 2, 1±11. SMELT, J. P. P. M. (1998) Recent advances in the microbiology of high pressure processing. Trends in Food Science & Technology, 9, 152±158. SMITH, W. L., LAGUNAS-SOLAR, M. C. and CULLOR, J. S. (2002) Use of pulsed ultraviolet laser light for the cold pasteurization of bovine milk. Journal of Food Protection, 65, 1480±1482. SOBRINO-LOPEZ, A. and MARTIN-BELLOSO, O. (2008) Enhancing the lethal effect of highQIN, B. L., BARBOSA-CANOVAS, G. V., SWANSON, B. G., PEDROW, P. D.


18


intensity pulsed electric field in milk by antimicrobial compounds as combined hurdles. Journal of Dairy Science, 91, 1759±1768. SOBRINO-LOPEZ, A., VIEDMA-MARTINEZ, P., ABRIOUEL, H., VALDIVIA, E., GALVEZ, A. and MARTIN-BELLOSO, O. (2009) The effect of adding antimicrobial peptides to milk inoculated with Staphylococcus aureus and processed by high-intensity pulsedelectric field. Journal of Dairy Science, 92, 2514-2523. SPILIMBERGO, S., ELVASSORE, N. and BERTUCCO, A. (2002) Microbial inactivation by highpressure. The Journal of Supercritical Fluids, 22, 55±63. ULRICH, K. (2007) Twenty years of Hakone Symposia: From basic plasma chemistry to billion dollar markets. Plasma Processes and Polymers, 4, 678±681. VEGA-MERCADO, H., MARTIN-BELLOSO, O., BAI-LIN, Q., FU JUNG, C., GONGORA-NIETO, M. M., BARBOSA-CANOVAS, G. V. and SWANSON, B. G. (1997) Non-thermal food preservation: pulsed electric fields. Trends in Food Science & Technology, 8, 151±157. WOUTERS, P. C., DUTREUX, N., SMELT, J. P. P. M. and LELIEVELD, H. L. M. (1999) Effects of pulsed electric fields on inactivation kinetics of Listeria innocua. Applied & Environmental Microbiology, 65, 5364±5371. WOUTERS, P. C., ALVAREZ, I. and RASO, J. (2001) Critical factors determining inactivation kinetics by pulsed electric field food processing. Trends in Food Science & Technology, 12, 112±121. YEOM, H. W., STREAKER, C. B., ZHANG, Q. H. and MIN, D. B. (2000) Effects of pulsed electric fields on the quality of orange juice and comparison with heat pasteurization. Journal of Agricultural and Food Chemistry, 48, 4597±4605. YU, D. X., ZHANG, Y. Q. and ZHONG, X. L. (1999) Sterilization effect of ultraviolet to milk. Science & Technology of Food Industry, 20, 24±26. YU, H., PERNI, S., SHI, J. J., WANG, D. Z., KONG, M. G. and SHAMA, G. (2006) Effects of cell surface loading and phase of growth in cold atmospheric gas plasma inactivation of Escherichia coli K12. Journal of Applied Microbiology, 101, 1323±1330. YU, L. J., NGADI, M. and RAGHAVAN, G. S. V. (2009) Effect of temperature and pulsed electric field treatment on rennet coagulation properties of milk. Journal of Food Engineering, 95, 115±118. ZEHNDER, H. J. (1988) Food irradiation ± science, technology, practice. Beta gamma, 1, 27±33. ZULUETA, A., ESTEVE, M. J., FRASQUET, I. and FRIGOLA, A. (2007) Fatty acid profile changes during orange juice-milk beverage processing by high-pulsed electric field. European Journal of Lipid Science & Technology, 109, 25±31.


Part I Case studies in high pressure and pulsed electric field processing of food


2 Commercial high pressure processing of ham and other sliced meat products at Esteban EspunÄa, S.A. M. Gassiot and P. Masoliver, Esteban EspunÄa, S.A., Spain

Abstract: The chapter covers the evolution of high pressure processing technology for commercial applications, particularly for the treatment of sliced cooked ham at 400 MPa, and for the treatment of dry cured ham and Tapas products with next generation high pressure equipment at 600 MPa. The characteristics of the high pressure equipment, the treatment of specific products and their intrinsic physico-chemical properties, and the benefits from a commercial viewpoint will be presented in detail. Key words: high pressure processing, meat products, inactivation of pathogens and spoilage organisms, effect of water activity.

2.1

Introduction

During the slicing and packaging processes of sliced meat products, the products inevitably suffer from some type of microbiological recontamination. The growth of microorganisms present from this recontamination can limit the safe preservation and shelf-life of these products. Such preservation problems typically have only moderate impact on dry, cured, sliced products; however, they are a more serious concern for products with high water activities, high pH, and that contain virtually no competing bacterial flora capable of hindering the proliferation of spoilage microorganisms. These problems in the preservation of sliced heat-treated products are a particular hindrance for sliced cooked ham in terms of their commercialization and marketability. The main problems


22


identified involve the presence and growth of Lactobacilli that cause the progressive acidification of the product and limit the product shelf-life to a relatively short period. In view of these problems, some corrective actions were taken to maximize safety and freshness during product shelf-life, such as employing logistics procedures that avoided intermediate warehouses and reducing the expiration dates for sliced cooked products (and sliced cooked ham in particular). However, these changes were insufficient to improve the taste of the cooked ham near the end of its commercial shelf-life, and there was an increase in returns of expired product. Several studies between 1990 and 1997 attempted to resolve these issues for cooked ham. Thermal pasteurization of the cooked, sliced ham in its final packaging was found to be an effective method for enhancing microbiological stability of the product. Heat pasteurization processes work with cooked products and they cannot be used with dry cured products. In these cases, heat pasteurization compromises the organoleptic and sensory characteristics (texture, flavor, color, etc.) of the products. Even with cooked products, thermal pasteurization unavoidably causes the release of juices, protein, and fat from the product, which adversely effects product appearance. These liquids accumulate in the package and have an undesired effect on texture, juiciness, color, etc. An absorbent interleaver was developed in order to absorb excess liquid and improve the presentation of the thermally pasteurized products. While the interleaver improved product appearance, it was not a commercially acceptable solution, because the products became too dry and developed a tough texture during their shelf-life. High pressure processing (HPP) is an emerging nonthermal food processing technology in which the food is subjected to high hydrostatic pressures (200± 700 MPa) by a non-compressible fluid (usually water) at generally moderate temperatures (usually significantly below 100 ëC). HPP has been applied in the food industry since 1992, when the first products treated by HPP were marketed in Japan. By the end of 1995, seven companies were marketing commercial HPPtreated products, such as jam, fruit juice, sauces, rice wine, and rice cake (Hayashi, 1997). By 1996, HPP technology was gaining prominence in the food industry because of its advantages for inactivating microorganisms and enzymes at ambient or relatively low temperatures with less adverse affect on the flavor, color and nutritional constituents of foods compared to thermal-only processes (Hoover et al., 1989; Mertens and Knorr, 1992; Cheftel, 1995; Cheftel and Culioli, 1997). In general, HPP tends not to destroy the covalent bonds between atoms of the constituent molecules, as the energy used during the treatment is relatively low, and the process affects hydrogen bonds and ionic and hydrophobic interactions in macromolecules. Accordingly, HPP treatments are effective against microorganisms and enzymes to ensure food safety and shelf-life stability, respectively, but HPP is less aggressive than heat so that the product tends to retain much of the flavor, texture, nutrients, and quality attributes of the product pre-processing. In assessing alternatives to heat pasteurization of sliced meat products, Esteban EspunÄa, S.A. decided to explore the potential use of HPP. In October


Commercial high pressure processing of sliced meat products

23

1996, pilot tests were conducted in Nantes (France) using HPP treatments up to 400 MPa on various sliced meat products (cooked and cured) to determine the effects of HPP on reducing microorganisms, inhibiting microbial growth after treatment, and on the organoleptic characteristics (such as appearance, texture, color, syneresis, etc.) of the products. HPP treatments of cooked products at 400 MPa significantly reduced microbial population levels, but the HPP treatments at 400 MPa were less effective for inactivating microorganisms with dry, cured products. Results of these pilot tests were encouraging for the application of HPP technology to preserve sliced cooked meat products, to maintain safety and product freshness until the end of their commercial shelf-life, and also to overcome problems induced by thermal pasteurization processes. Esteban EspunÄa, S.A. considered HPP an exciting opportunity for offering consumers traditional meat products featuring improved quality and enhanced microbial safety during the shelf-life of the product. The successful marketing of HPP-treated fruit products (various fruit products in Japan, avocado products in the US, and various fruit juices in France, Portugal, and the UK) helped create a positive market outlook among global consumers toward foods preserved with HPP that encouraged EspunÄa to invest in HPP technology as a marketable alternative to thermal pasteurization for sliced, cooked meat products. Several important factors went into the company making this decision. First, HPP would potentially improve product quality of the cooked meat products and help the company gain market advantage over the competition. Second, being the first sliced meat company to exploit HPP technology while it was undergoing further development would help the company gain significant experience with this technology prior to its competitors. Third, the company also believed equipment manufacturing companies would eventually develop industrial equipment capable of carrying out treatments at the higher pressures needed for cured products (they comprise the main product base of EspunÄa), and prior experience with HPP technology would then be advantageous. Fourth, the use of high pressure would also provide the company with more potential opportunity for developing new products with HPP.

2.2

High pressure processing (HPP) equipment

2.2.1 400 MPA HPP equipment (Fig. 2.1) In July, 1997, Esteban EspunÄa, S.A. bought a prototype horizontal configuration high pressure machine with a capacity of 320 L and maximum working pressure of 400 MPa. The company's production design flows and HACCP system established in the plant suggested that a horizontal system had clear advantages over vertical configuration equipment to avoid the cross-contamination of treated and untreated products. The horizontal configuration requires loading at one end of the machine and unloading at the other end. This separation between the treated and the untreated products helps avoid cross-contamination.


24


Fig. 2.1

400 MPa HPP equipment.

Dimensions Length 5 m, inner diameter 280 mm, external diameter 800 mm, and volume 255 L. Working conditions Treated product sliced cooked ham. Maximum operating pressure 400 MPa, pressurization hold time 10 min, maximum water temperature 15 ëC, production 120 kg/cycle, cycles per day 30, and total cycles performed (1998±2008) 59 000. Laboratory and pilot-scale research for sliced meat products Laboratory and pilot-scale research for sliced meat products (cooked ham, dry cured ham, and bacon) determined the following: · the optimal working parameters of high pressure, temperature, and processing time · the effectiveness of the HPP treatment on the inactivation of pathogens of interest in each sliced meat product · the shelf-life of each HPP-treated sliced meat product · The texture and organoleptic characteristics of the HPP-treated meat products · the appropriateness of the HPP treatment for each sliced meat product.



25

2.2.2 Major operational challenges with the equipment Since the first high pressure unit installed at Esteban EspunÄa, S.A. was a prototype, EspunÄa had to collaborate with the equipment supplier continuously for the first years of operation to resolve any operational issues, such as maintenance costs and equipment repairs, finding suitable baskets for product placement, and defining the drying process of products after HPP treatments. Maintenance costs and equipment and repairs The original pumping system of the prototype did not function properly due to problems with the multiplier seals of the high pressure pumps. The nature of the construction material is important to prevent wear of the seals (gaskets). The first alternative pumping system provided significant improvements, but the problems were not resolved. Different suppliers of multipliers were contacted until the correct equipment for the high pressure pumps was found. Another major problem was detected in the wear of the joints. A large number of different joints were tested in order to find materials capable of withstanding the high pressures of the equipment. The discharge valves still suffer significant wear during operation and must be replaced frequently. Baskets for product placement The design of the baskets for placing product inside the high pressure vessel is crucial, since they determine the amount of product treatable per cycle. They need to be made of a durable material that does not damage the coating material of the interior of the pipe body as they slide during loading and unloading. It is also important that water drains out of them quickly for rapid processing throughput and to conveniently facilitate its recovery. The original baskets were made of perforated metal that allowed the water to drain quickly, but the metal rapidly eroded the coating material of the interior of the pipe body. Special hard plastic baskets had to be developed to solve these problems. Drying the products After HPP of the pre-packaged products, the product is removed from the high pressure vessel and it is covered with water. Cold drying equipment was purchased and implemented under a standard operating procedure. This cold drying process prepares the packaged product for labelling and packing, while also ensuring that the product maintains quality (compared to the effects of heat). 2.2.3 600 MPa HPP equipment (Fig. 2.2) As mentioned above, high pressure treatments at 400 MPa tend not to significantly reduce pathogens and their concomitant risks in dry cured products, and higher pressure capabilities are need for ensuring the safety of these products. The 600 MPa high pressure equipment has a capacity of 318 L and a horizontal configuration to avoid post-processing cross-contamination of treated products.


26


Fig. 2.2 600 MPa equipment.

Dimensions Length 4.5 m, internal diameter 300 mm, external diameter 806 mm, and volume 218 L. Working conditions Treated products dry cured meats. Maximum operating pressure 600 MPa, pressurization hold time 5±10 min (depending on product), maximum water temperature 15 ëC, production 110 kg/cycle, cycles per day 38, total cycles performed (2005±2008) 34 500. Laboratory and pilot-scale research for dry cured meat products Between 2000 and 2005, internal laboratory and pilot-scale studies were undertaken to assess the effectiveness of 600 MPa treatment for dry cured products (especially for sliced cured ham) using pilot-scale equipment. The most noteworthy investigations are listed below: · Determining the effects of HPP at 600 MPa on the microbiology, bioequivalence, biochemical properties, and bioavailability of nutrients; and determining the mutagenic activity of vacuum-packed sliced meat products: cooked ham, dry cured pork ham, and marinated beef loin (GreÁbol, 2002; Garriga et al., 2004; GarcõÂa Regueiro et al., 2002). · Evaluating the inactivation kinetics of L. monocytogenes by HPP (unpublished results). Personal communication of research carried out by



· · · ·

27

SSICA-Stazione Sperimentale per l'Industria delle Conserve Alimentari, Parma, Italia for Esteban EspunÄa, S.A. For related reference, see Gola et al. (2003). Modeling, designing, and optimizing a high pressure-assisted freezing process in food (Arnau et al., 2003). Evaluating the application of HPP for dry-cured ham to improve texture, flavor, and safety (Serra et al., 2007a; 2007b). Determining the efficiency of HPP at 600 MPa to inactivate pathogens of concern in different meat products (JofreÂ et al., 2009). Developing new products treated by high pressure (internal research). The main results from this research work were:

· HPP at 600 MPa and 31 ëC (88 ëF) for 6 min to treat sliced, vacuum-packed dry cured ham samples caused a reduction of at least 2-log cycles of spoilage microorganisms, maintained low levels of survivors during the storage period at refrigerated temperatures, contributed to retaining organoleptic attributes and freshness during prolonged storage periods (investigated up to 120 days), and helped prevent off-flavors, sour taste, and gas formation (Garriga et al., 2002). · Enterobacteriaceae and Escherichia coli were below the detection limit in all HPP-treated and untreated samples (Garriga et al., 2004). · Bioequivalence analysis concluded that HPP-treated (600 MPa for 10 min at 31 ëC) vacuum-packed, cooked ham and dry cured ham were substantially equivalent to their untreated counterparts (GarcõÂa Regueiro et al., 2002). · Showed that an HPP treatment at 600 MPa and 25 ëC for at least 7.5 min was sufficient to obtain 5 decimal reductions of strains of L. monocytogenes isolated from raw ham with water activity (aw) 0.90 (unpublished results). Personal communication of research carried out by SSICA-Stazione Sperimentale per l'Industria delle Conserve Alimentari, Parma (Italy) for Esteban EspunÄa, S.A. For related reference, see Gola et al. (2003). · HPP treatments at 600 MPa and 31 ëC for 6 min reduced > 2-log cycles of Salmonella spp. and L. monocytogenes in dry cured products (for more information, see JofreÂ et al., 2009). · Toxicological evaluation of both HPP-treated and untreated cooked ham and dry cured ham were carried out using an in vitro Ames test (Maron and Ames, 1983) in order to compare the potential mutagenicity. All extracts obtained from the samples were shown to be ineffective as mutation-inducing agents in the experimental conditions.

2.3

Commercialized HPP-treated food products

The specific objectives of Esteban EspunÄa, S.A. were to achieve successful commercialization and marketing of the HPP-treated meat products, sliced cooked ham in particular, by assuring product freshness until the end of its best-


28


Fig. 2.3 Commercial sliced cook ham product (aw > 0.95) treated with HPP (400 MPa, 10 min, 15 ëC) and labeled `Product pasteurized by high pressure: it keeps fresh until it is consumed'.

before date, and ensuring food safety by reducing health risks associated with pathogenic microorganisms. In 1998, Esteban EspunÄa, S.A. pioneered the use of HPP for meat products when it began marketing sliced cooked ham (Fig. 2.3). In mid-2002, the company launched the first phase of its range of tapas products developed with the use of HPP technology (Fig. 2.4). Tapas are mini pork sausages made with Spanish paprika and marinated diced pork that are heat-and-serve products for consumer convenience. 2.3.1 Effect of high pressure on high water activity products (aw > 0.95) As mentioned above, slicing and packaging operations take place after cooking, and preventing cross-contamination from occurring at these points is critical with regard to determining the shelf-life and safety of the products. Under good hygiene/manufacturing practices, the levels of pathogenic and spoilage microorganisms in the products are very low. Because of the high water activity (aw > 0:95) of cooked ham, lactic acid bacteria on the ham coming mainly from cross-contamination during slicing and packaging can quickly grow to 108 CFU/g in untreated products (Fig. 2.5). The pressurized product shows a significant delay in the growth of spoilage microorganisms compared to the untreated product, thereby also contributing to maintaining the sensorial freshness for at least 60 days after treatment (Garriga



29

Fig. 2.4 An example of the `Tapas al minute' range: Pinchitos a las finas hierbas ± diced pork marinated with Spanish herbs (aw > 0.95) and treated with HPP after packaging (400 MPa and 10 min or 600 MPa and 5 min).

Fig. 2.5 Lactic acid bacteria evolution during commercial shelf-life. HPP treatments significantly reduce the population levels and growth of Lactobacilli, which compromise the taste, flavor, and shelf-life of cooked packaged meat products.


30


et al., 2002; JofreÂ et al., 2009). Bioequivalence and sensorial tests also show that HPP induces no differences in biochemical properties, aroma, flavor, color, and texture compared to untreated samples (Garriga et al., 2004; GarcõÂa Regueiro et al., 2002). Thus, HPP treatments produced a commercial packaged cooked sliced ham product (Fig. 2.3) and tapas products (Fig. 2.4) that retain their fresh taste for 60 days (the best-before date) and has widespread consumer appeal (GreÁbol, 2002). 2.3.2 Effect of HPP on lower water activity (aw < 0.92) products Dried cured products (sausage, cured ham, etc.) are products that are generally microbiologically very stable, with the proliferation of spoilage microorganisms limited by the relatively low water activity values of the products (Feiner, 2006). The main concern with these products is ensuring the absence of pathogens, especially Salmonella spp. and L. monocytogenes. European law requires the absence of Salmonella in 25 g of product, and allows only 2-log cycles of Salmonella spp. and L. monocytogenes in dry cured products (JofreÂ et al., 2009). HPP treatments at 600 MPa and 25 ëC for at least 7.5 min effected the inactivation of 5 decimal reductions of strains of L. monocytogenes isolated from raw ham with aw 0.90 (unpublished results). For related references see Gola et al. (2003). Enterobacteriaceae and Escherichia coli were below the detection limit in all HPP-treated and untreated samples (Garriga et al., 2004). Additionally, bioequivalence analysis of vacuum-packed HPP-treated (600 MPa and 31 ëC for 10 min) cooked ham and dry cured ham were essentially equivalent to their untreated counterparts (GarcõÂa Regueiro et al., 2002).

2.4

Treatment costs

The main costs involved in the use of HPP for applications relating to food preservation are the initial capital investment, routine operating costs and maintenance, and subsequent amortization costs. Maintenance costs depend on the reliability of each component in the equipment, and higher pressures tend to cause more wear of the components. The use of higher pressures reduces treatment times and increases production capacity, and these factors help to recover incurred financial costs. The cost per kg of HPP-treated products using the commercial equipment (600 MPa) is half the cost per kg of treated product in the prototype (400 MPa) (see Table 2.1).


Commercial high pressure processing of sliced meat products Table 2.1

31

High pressure operating costs (2007) A 400 MPa

B 600 MPa

Initial investment (USD)

1,233,000

2,055,000

Cycles (2007) Kg treated (2007) Direct labor (USD) Amortization (USD) Maintenance (USD) Total costs (2007 USD)

5,896 449,980 491,930 119,190 31,370 642,490

11,559 975,742 380,293 213,720 83,965 677,978

17,455 1,425,722 872,223 332,910 115,335 1,320,468

Costs/Cycle

A/Cycles 400 MPa

B/Cycles 600 MPa

(A+B)/ (Cycles A + Cycles B)

83.43 20.22 5.32 108.97

32.90 18.49 7.26 58.65

49.97 19.07 6.61 75.65

A/kg treated 400 MPa

B/kg treated 600 MPa

(A+B)/ (kg A + kg B treated)

1.09 0.26 0.07 1.43

0.39 0.22 0.09 0.70

0.61 0.23 0.08 0.93

USD/Cycle USD/Cycle USD/Cycle USD/Cycle

direct labor amortization maintenance total (2007 USD)

Costs/kg treated USD/kg direct labor USD/Kg amortization USD/Kg maintenance USD/Kg total (2007 USD)

2.5

A+B combined

Conclusions

After ten years of operating HPP equipment for food preservation for successfully commercialized meat products, we draw the following conclusions based on our cumulative experience: · The application of HPP was commercially successful for traditional meat products (sliced cooked ham), and its availability also facilitated the development of commercially successful new products, such as the `Minute Tapas' range of products. · HPP treatments at 600 MPa for 6 min. is an efficient method to delay the growth of spoilage microorganisms in packaged sliced cooked ham and dry cured ham (Garriga et al., 2002). · HPP at 600 MPa for 6 min. significantly reduces the safety risks associated with Salmonella and L. moncytogenes in packaged sliced cooked ham and dry cured ham (Garriga et al., 2002). · The use of HPP allows us to sell value-added high-quality (high organoleptic attributed and reduced microbial risks) premium products to certain customers with specific standards and a willingness to purchase premium products at premium prices.


32

2.6


Company information

Esteban EspunÄa, S.A. was founded in 1947 and manufactures meat products from its headquarters in Olot, the capital of the Garrotxa region (in the north of Catalunya, Spain). Over the years the company has combined traditional and modern production methods to provide a variety of high quality meat products. In 1989 Esteban EspunÄa, S.A. launched its sliced meat product range, with the aim of offering its customers innovative, convenient products. This new product range caused a sales increase which motivated the construction of an entirely new slicing plant after only four years. In 1998, Esteban EspunÄa, S.A. pioneered the use of high pressure in the meat products industry by marketing sliced cooked meat products. The commercial success of these products has led to additional commercial successes with the development of innovative new products, such as their range of tapas products also treated with HPP.

2.7

References

Â NDEZ, ARNAU, J., GOU, P., MONFORT, J.M., SANZ, P.D., MOLINA-GARCIÂA, A.D., OTERO, L., FERNA Ä A, X., GREÁBOL, N., MASOLIVER, P., GASSIOT, M., YUSTE, J., P.P., GUAMIS, B., ESPUN

2003. `Procedimiento para la proteccioÂn y estabilizacioÂn del color de carnes y productos elaborados de carne, frescos, marinados o parcialmente deshidratados, tratados por alta presioÂn'. Spanish application number: 200300734. CHEFTEL, J.C. 1995. Review: High-pressure, microbial inactivation and food preservation. Food Science and Technology, International 1, 75±90. CHEFTEL J.C., CULIOLI, J. 1997. Effects of high pressure on meat: A review. Meat Science 46, 211±236. COMMISSION REGULATION (EC) No 2073/2005 of 15 November 2005 on microbiological criteria for foodstuffs. FEINER, G. 2006. Introduction to the microbiology of meat and meat products. In Meat products handbook: Practical science and technology. Cambridge: Woodhead Publishing Limited. Â RRAGA, M.C., HORTO Â S, M., DIÂAZ , I., VALERO, A., RIUS, M.A. 2002. GARCIÂA REGUEIRO, J.A., SA Bioequivalence of meat products treated with high pressure processing, available at http://hdl.handle.net/2072/4750 GARRIGA, M., AYMERICH, T., HUGAS, M. 2002. Effect of high pressure processing on the microbiology of skin-vacuum packaged sliced meat products: cooked pork ham, dry cured pork ham and marinated beef loin, available at http://hdl.handle.net/ 2072/4686 GARRIGA, M., GREÁBOL, N., AYMERICH, M.T., MONFORT, J.M., HUGAS, M. 2004. Microbial inactivation after high-pressure processing at 600 MPa in commercial meat products over its shelf life. Innovative Food Science and Emerging Technologies 5, 451±457. GOLA, S., FRUSTOLI, M., ROVERE, P., MIGLIOLI, L. 2003. Inattivazione di Listeria monocitogenes in prosciutto crudo trattato con la pressione idrostatica. Industria Conserve 78, 441±449. GREÁBOL, N. 2002. Commercial use of high hydrostatic pressure in sliced cooked ham in Spain. In Hayashi, R. (ed.), Trends in high pressure bioscience and technology, pp. QUEVEDO, J.



33

385±388. Amsterdam: Elsevier Science. 1997. High-pressure bioscience and biotechnology in Japan. In Heremans, K. (ed.), High Pressure Research in the Biosciences and Biotechnology, 1±4. Leuven: Leuven University Press. HOOVER, D.G., METRICK, A.M., PAPINEAU, A.M., FARLAS, D-F., KNORR, D. 1989. Biological effects of high hydrostatic pressure on food microorganisms. Food Technology 43, 99±107. JOFREÂ, A., AYMERICH, T., GREÁBOL, N., GARRIGA, M. 2009. Efficiency of high hydrostatic pressure at 600 MPa against food-borne microorganisms by challenge tests on convenience meat products. LWT-Food Science and Technology 42, 924±928. MARON D., AMES, B.N. 1983. Revised methods for the Salmonella mutagenicity test. Mutation Res. 113, 173±215. MERTENS, B., KNORR, D. 1992. Developments of non-thermal processes for food preservation. Food Technology 46, 126±133. HAYASHI R.

Â GARRA, C., GREÁBOL, N., GUA Á RDIA, M.D., GUERRERO, L., GOU, P., MASOLIVER, P., SERRA, X., SA

2007a. High pressure applied to frozen ham at different process stages. 1. Effect on the final physicochemical parameters and on the antioxidant and proteolytic enzyme activities of dry-cured ham. Meat Science 75, 12±20.

GASSIOT, M., MONFORT, J.M., ARNAU, J.

Á RDIA, M.D., GUERRERO, L., GOU, P., MASOLIVER, P., GASSIOT, M., SERRA, X., GREÁBOL, N., GUA Â GARRA, C., MONFORT, J.M., ARNAU, J. 2007b. High pressure applied to frozen ham SA

at different process stages. 2. The effect on the sensory attributes and on the colour characteristics of dry-cured ham. Meat Science 75, 21±28.


3 High hydrostatic pressure processing of fruit juices and smoothies: research and commercial application F. Sampedro and X. Fan, United States Department of Agriculture, Agricultural Research Service (USDA-ARS), USA and D. Rodrigo, CSIC, Spain

Abstract: Several world-wide health organizations have pointed out the importance of increasing the intake of fruits and vegetables in the diet. Consumers are also increasing the demand for more convenient, nutritious, fresh and price-reasonable products. Although thermal pasteurization has been the processing technology of choice to preserve fruit juices, the thermal process damages the nutritional and sensory properties of products. As a result of scientific studies demonstrating the benefits of high hydrostatic pressure (HHP) technology and the advances in the design of process equipment, high quality fruit juices and related products treated by HHP are appearing in food markets around the world. Key words: fruit juices, smoothies, high hydrostatic pressure, microbial safety, food quality, enzymes, bioactive compounds, consumer attitudes, commercial application.

3.1

Introduction

Fruit and vegetable juices and their derivatives, are major world commodities and part of the economic lifeblood of many countries, particularly in the developing world. The perception of the healthy nature of these products is one of the major reasons for their consumption. Owing to their perishable characteristics, however, it is necessary to process them to extend their shelf-lives. To


High hydrostatic pressure processing of fruit juices and smoothies

35

prolong shelf-life, thermal pasteurization is most commonly employed, but losses of representative flavor compounds, color, and vitamins occur (Yeom et al., 2000). In recent years, consumers have increasingly sought so-called `fresh' products stored under refrigeration. The trend of increasing consumption of these products is partly due to the application of emerging non-thermal technologies, such as high hydrostatic pressure (HHP). This is an interesting field for application in fruit juices, because HHP employs cold pasteurization, which preserves the nutritional quality and characteristic flavor of the products (Min and Zhang, 2003; Rivas et al., 2006; Sampedro et al., 2009a,b; SaÂnchez-Moreno et al., 2006). Large amounts of scientific data have shown the advantages and benefits of HHP processing versus thermal pasteurization in the processing of fruit juices. High microbial reduction of spoilage and acid-resistant pathogens, enzymatic stabilization, preservation of bioactive compounds, and positive consumer attitudes toward this technology have been demonstrated in numerous studies (Balasubramaniam and Farkas, 2008; Oey et al., 2008a, 2008b; Rastogi et al., 2007; San Martin et al., 2002; Wright et al., 2007). As a result of the research efforts showing the benefits of HHP technology and the advances in the design of process equipments that satisfy the industrial production and cost requirements, fruit-based products treated by HHP with a competitive price and high nutritional quality are gaining an increasing share in markets around the world, especially in Australia and Europe.

3.2 Fruit composition, high hydrostatic pressure (HHP) treatment and recommended fruit intake 3.2.1 Fruit composition Fruits that contain a wide range of different compounds and show considerable variation in composition and structure play a very significant role in human nutrition. The most important components in fruit and its derivatives can be grouped as follows: water, proteins, carbohydrates, fats, minerals and vitamins. Most of these components are essential nutrients that are needed by the human body. Water is the most abundant component (more than 80%) in fruit, ranging from 82% in grapes to 90% in strawberries (Fourie, 1996). However, the maximum water content varies between stages of maturity and even between individual fruits of the same kind because of structural differences. Proteins usually contribute less than 1% of the fresh weight of fruit. Carbohydrates consist of polysaccharides such as starch, cellulose, hemicellulose and pectic material, and also disaccharides and monosaccharides such as sucrose, fructose, and glucose. The total carbohydrate value varies from 3% in lemons to about 15% in grapes (Fourie, 1996). Dietary fiber makes up a unique component within the total carbohydrate content of fruits and vegetables. Fiber is the


36


structural material of plant cells that are resistant to the digestive enzymes of the human stomach and is essential for human intestinal function. Lipid content of fruit and vegetables is generally below 1%, and they are therefore not a good source of fats. Fruits also contain a variety of essential mineral elements, among which potassium is the most abundant and occurs mainly in combination with various organic acids. Calcium is always present in the pectic material in the cell walls of the fruit and magnesium in the chlorophyll molecules. Phosphorous can play an important part in carbohydrate metabolism. As far as vitamin content, considerable differences are reported between fruit species and varieties, as well as between the same varieties grown under different environmental conditions. Fruits and vegetables are specially known as a source of ascorbic acid. Vitamin A is fat soluble and does not occur as such in fruit, although certain fruit carotenoids can be converted to vitamin A in the body. On the other hand, fruit is a moderate to poor source of the members of the vitamin B and E group. Several components with antioxidant activity naturally occur in fruit. These components include ascorbic acid, tocopherols, betacarotene and other flavonoid components. Fruits are also rich sources of phytochemicals such as phenolics and flavonoids which may reduce the risk of cardiovascular disease, cancer, and other chronic diseases. 3.2.2 Juices, smoothies, and pulps with the potential to be treated by HHP Fruit juices are made from fresh fruit by mechanical squeezing (premium juices), or also from fruit juice concentrates by diluting with water. Premium or direct juices are considered the best candidates for HHP processing due to their high quality requirement for commercial appeal. Smoothies are blended cold drinks consisting of a number of ingredients including fruit (and sometimes vegetables) and fruit juice. Depending on the type of smoothie, crushed ice, sugar or honey, some types of thickener such as milk, soymilk, or yogurt, or other flavor enhancers and stabilizers can be added to create a complex composition. Smoothies have milkshake-like consistencies which are thicker than slush drinks. They are usually sold as a drink, snack or meal alternative, they are available either ready-made or made-to-order, and they are becoming an increasingly popular way of consuming dietary fruits. Often marketed to health-conscious people, smoothies are commonly fortified with `boosts' or ènhancers' (additional vitamins, minerals, herbs amino acids or other nutrients). The fruit pulp is an intermediate product made from fresh fruit, not intended for consumption as such, which also includes whole and large portions of fruit. It can be used as raw material for yogurt and dessert preparations or diluted for consumption in juice. 3.2.3 Fruit intake recommendation Low fruit and vegetable intake is among the top 10 risk factors contributing to attributable mortality (WHO, 2003). Fruits and vegetables as part of the daily



37

diet could help prevent major noncommunicable diseases (NCD) such as cardiovascular diseases and certain cancers. Eating a variety of vegetables and fruits clearly ensures an adequate intake of most micronutrients, dietary fibers and a host of essential non-nutrient substances. Increased fruit and vegetable consumption can also help displace foods high in saturated fats, sugar or salt. A published report of a Joint FAO/WHO Expert Consultation on Diet, Nutrition and the Prevention of Chronic Diseases recommends the intake of a minimum of 400 g of fruits and vegetables per day (excluding starchy tubers such as potatoes) for the prevention of chronic diseases including heart disease, cancer, type 2 diabetes, and obesity (WHO, 2003). Presently, the estimated intake levels of fruits and vegetables varies considerably around the world ranging from less than 100 g/day in less developed countries, to about 450 g/day in Western Europe. However, the intake increases if consumption of fresh or canned juice is taken into account, and it is important, therefore, to achieve high quality beverages and to ascertain their nutritional value. In a systematic review, Ruxton et al. (2006) found that pure fruit and vegetable juices appeared to offer similar health benefits to whole fruits and vegetables, probably because of similarities in antioxidant and/or polyphenol content. Therefore, the nutritional quality of fruit juices is extremely important in order to provide vitamins, minerals and fiber to satisfy dietary recommendations.

3.3 Basic research on high hydrostatic pressure (HHP) processing of fruit juices and derivatives 3.3.1 Aspects related to food safety Among the spoilage microflora encountered in fruit juices, the more common ones include lactic acid bacteria (Lactobacillus and Leuconostoc species), fermentative yeasts (Saccharomyces cerevisiae) and spore-forming molds due to their capability of growing at low pH values (4.4 log), LMM1010 (>4.7 log), LMM1020 (>3.4 log), LMM1030 (NDa) OJ-400 MPa: MG1655 (>4.4 log), LMM1010 (1.5 log), LMM1020 (2.4 log), LMM1030 (>3.4 log)

GarcõÂa-Graells et al. (1998)

Orange

Escherichia coli O157:H7 (NCTC 12079)

400±550 MPa for 5 min at 20 ëC

pH (3.4-4.5):550 MPa for 5 min at 20 ëC (6 log) pH (5.0): 550 MPa for 5 min at 30 ëC (6 log)

Linton et al. (1999a, 1999b)

Orange

Staphylococcus aureus (485 and 765), Listeria monocytogenes (CA), Escherichia coli O157:H7 (933 and 931), Salmonella enteritidis (FDA) and Salmonella typhimurium (E21274)

345 MPa for 5 min at 50 ëC

>8 log for all strains

Alpas and Bozoglu (2000)

Orange and apple

Escherichia coli O157:H7 (C9490), Escherichia coli (ATCC 11775) Listeria monocytogenes (NCTC 11994)

100±500 MPa, 5 min at 20 ëC

AJ: L. monocytogenes 300 MPa (5 log) E. coli O157:H7 500 MPa (5 log), E. coli 350 MPa (5 log) OJ: L. monocytogenes 300 MPa (3 log) E. coli O157:H7 500 MPa (1-2 log), E. coli 350 MPa (5 log)

Jordan et al. (2001)

Table 3.1

Continued


Juice

Microorganism

HHP conditions

Inactivation results

Reference

Orange, apple, grape and carrot

Escherichia coli O157:H7 (ATCC 43895, SEA13B88, 43985) Salmonella strains (S. hartford, S. muenchen, S. typhimurium, S. agona, S. enteritidis)

615 MPa for 1±2 min at 15 ëC

615 MPa for 2 min GJ: E. coli cocktail (8.34 log), Salmonella strains (>8 log) OJ: E. coli cocktail (2.12 log), S. enteritidis (7.67 log), S. typhimurium (6.91 log), rest strains (>8 log) AJ: E. coli cocktail (0.41 log), S. typhimurium (3.92 log), rest strains (5 log) CJ: E. coli cocktail (6.40 log), S. enteritidis (6.67 log), S. typhimurium (5.06 log), S. hartford (5.31 log), rest strains (>7 log)

Teo et al. (2001)

Orange (fresh and concentrated)

Leuconostoc mesenteroides (ATCC 8293)

200±400 MPa for 0±60 min at 20 ëC

Fresh OJ: 350 MPa (D value-2.0 min, z value-137 MPa) Concentrated OJ: 400 MPa (D value-6.1 min, z value-251 MPa)

Basak et al. (2002)

Apple

Alicyclobacillus acidoterrestris (ATCC 49025 and NFPA 1013) (spores)

22, 45, 71 and 90 ëC 207, 414 and 621 MPa for 5 and 10 min

22 ëC: 45 ëC: 71 ëC: 90 ëC:

Lee et al. (2002)

Apple and orange

Alicyclobacillus acidoterrestris (veg. cells)

350 MPa for 20 min at 50 ëC

OJ: 4.4 log AJ: 4.64 log

No inactivation 207 MPa, 10 min (3.5 log) 414 MPa, 10 min (5.5 log) 414 MPa, 1 min (5.5 log)

Alpas et al. (2003)


Apple, apricot, cherry and orange

Listeria monocytogenes

250 and 350 MPa for 5 min at 30 and 40 ëC

350 MPa for 5 min at 30 ëC (>6 log) cherry>orange>apricot>apple

Alpas and Bozoglu (2003)

Apple

Escherichia coli (ATCC 29055)

150±400 MPa for 5 min at 25 ëC

Single pulse 400 MPa at 25 ëC (8 log)

Ramaswamy et al. (2003)

Orange and peach

Listeria monocytogenes (KUEN 136)

300, 400 and 600 MPa for 1±70 min

D values (min) PJ: 6.17-300 MPa 3.39-400 MPa 1.52600 MPa z value: 506 MPa OJ:2.87-300 MPa 1.80-400 MPa 0.87600 MPa z value: 576 MPa

DogÏan and Erkmen (2004)

Orange and peach

Escherichia coli (KUEN 1504)

300±700 MPa for 1±24 min at 25 ëC

D values (min) PJ: 300 MPa-5.38, 400 MPa-3.38, 600 MPa-1.22 z value: 450.1 MPa OJ: 300 MPa-2.42, 400 MPa-1.57 600 MPa-0.68 z value: 558.4 MPa

Erkmen and DogÏan (2004)

Orange and apple

Escherichia coli O157:H7

0.1±250 MPa for 20 min at 25 and 4 ëC RD: Rapid decompression (2 ms) SD: Slow decompression (30 s)

AJ: 25 ëC-250 MPa RD and SD (7 log) 4 ëC-215 MPa (RD) 225 MPa (SD) (7 log) OJ: 25 ëC-250 MPa RD (5 log) SD (3.5 log) 4 ëC-215 RD (6.5 log) 225 SD (6 log)

Noma et al. (2004)

Table 3.1

Continued


Juice

Microorganism

HHP conditions


Reference

Orange (Navel and Valencia var.)

Salmonella strains (S. typhimurium, S. montevideo and S. enteritidis)

300±600 MPa at 20 ëC

Conditions for 5 log reduction Navel OJ: 300 MPa-200 s, 450 MPa-20 s, 600 MPa-5 s Valencia OJ: 300 MPa-369 s, 450 MPa25 s, 600 MPa-4 s

Bull et al. (2005)

Apple, orange, apricot and sour cherry

Staphylococcus aureus, Escherichia coli O157:H7 and Salmonella enteritidis

250±450 MPa for 0-60 min at 25± 50 ëC

350 MPa for 5 min at 40 ëC (8 log)

Bayindirli et al. (2006)

Apple (concentrated at different concentrations)

Alicyclobacillus acidoterrestris (NFPA 1013 and 1101) (spores)

22, 45, 71 and 90 ëC 207, 414 and 621 MPa for 5 and 10 min

22 ëC: No inactivation 45 ëC-17.5 ëBrix: 621 MPa, 10 min (2.5 log) 45 ëC-35ëBrix: No inactivation 71 ëC-17.5 ëBrix: 207 MPa, 5 min (5 log) 71 ëC-35 ëBrix: 621 MPa, 10 min (5 log) 90 ëC-70 ëBrix: No inactivation

Lee et al. (2006)

Banana

Escherichia coli (ATCC 43888) Sighella flexneri (LMG10472) Yersinia enterocolitica (LMG7899) Salmonella typhimurium (LT2)

225±350 MPa for 15 min at 25 ëC combined with hen egg white lysozyme (HEWL) and lambda lysozyme (LaL)

E. coli: HHP (1.2 log), HHP+HEWL (1.7 log), HHP+LaL (6.5 log) S. typhimurium: HHP (2.8 log), HHP+HEWL (3 log), HHP+LaL (6.3 log) Y. enterocolitica: HHP (1.3 log), HHP+HEWL (1.3 log), HHP+LaL (3.5 log) S. flexneri: HHP (0.5 log), HHP+HEWL (0.6 log), HHP+LaL (4.0 log)

Nakimbugwe et al. (2006)

Apple and orange


150±350 MPa for 5 min at 20, 40 and 60 ëC

Orange juice: 248 MPa at 60 ëC (6 log) Apple juice: 203 MPa at 57 ëC (6 log)

MunÄoz et al. (2007)


Apple and orange

Escherichia coli O157:H7 (H1730, E0019, F4546, 994 and cider and E009) Salmonella strains (S. agona, S. baildon, S. gaminara, S. michigan and S. typhimurium )

300 and 550 MPa for 2 min at 6 ëC

E. coli strains: 300 MPa (0.12±0.53 log) 24 h at 6 ëC: 0.44±1.58 log 550 MPa (1.25±4.39 log) 24 h at 6 ëC: 3.24±5.57 log Salmonella strains: 300 MPa (0.26±0.62 log) 24 h at 6 ëC: 0.41±0.86 log 550 MPa (4.88±6.52) 24 h at 6 ëC (5.70±5.97)

Whitney et al. (2007)

Kiwifruit and pineapple

Escherichia coli (ATCC 11775) Listeria innocua (ATCC 33090)

300 MPa for 5 min 300 MPa for 300 s with 1±10 pulses

300 MPa-5min: E. coli and L. innocua (4 log in kiwifruit and 1 log in pineapple) 300 MPa-10 pulses-60 s: E. coli and L. innocua (4.5 log in kiwifruit and 2.8 and 3.5 log in pineapple) 350 MPa-5 min: E. coli and L. innocua (5 log in kiwifruit and 2.5 and 3.5 log in pineapple)

Buzrul et al. (2008)

Cashew apple


250±400 MPa for 1.5±7.5 min at 25 ëC

D values: 250 MPa-16.43 min, 300 MPa11.52 min, 350 MPa-2.42 min, 400 MPa1.21 min z value: 123.46 MPa

Lavinas et al. (2008)

Low-acid orange juice

Yersinia pseudotuberculosis (197) Francisella tularensis (LVS)

300 and 500 MPa for 2±6 min at 10 and 25 ëC

500 MPa for 2 min at 10 ëC (5 log)

Schlesser and Parisi (2009)

a

ND: Non detected

44


al. (2007) found an optimal combination of moderate levels of high pressure (200±250 MPa) and mild temperature (57±60 ëC) for E. coli inactivation (6 log cycles) in OJ and AJ. At these treatment conditions, all natural flora present in the juices were reduced to almost undetectable levels. In a selective environment of a high-pressure fruit juice processing plant, the occurrence of naturally E. coli pressure-resistant strains cannot be ruled out. In this regard, GarcõÂa-Graells et al. (1998) isolated several mutants that were pressure-resistant (800 MPa at 10±40 ëC) from a pressure-sensitive E. coli. The non-treated E. coli mutants were able to survive in acidic substrates (AJ and mango juice pH 3.3±4.0) at refrigeration conditions (8 ëC) for at least 30 days. The survival at refrigeration conditions even in acidic conditions could be explained by the fact of a reduced permeability of cell membrane to protons or reduced metabolic activity at reduced temperatures. However, after pressure treatment (300±500 MPa) and further storage (5 days of storage at 8 ëC) the number of survivors was below the detection limit (high numbers of cells were injured during the treatment, resulting in a reduced resistance to the low pH during storage). This fact illustrates that sublethal injury is a crucial parameter which should be monitored after HHP treatment to detect any recovery of injured cells, particularly over prolonged storage periods. It is known that E. coli O157:H7 strains have been implicated in several outbreaks related to unpasteurized fruit juices due to its high resistance to acidic environment and low infective dose. Several studies have been performed using this strain and HHP treatment in different fruit juices (Table 3.1). In a first study, Linton et al. (1999a, 1999b) studied the survival of E. coli O157:H7 in OJ at different pH levels (3.4 to 5.0). The survival of E. coli in OJ was pH dependent. After 20 and 25 days at 3 ëC at pH 3.4 and 3.6, respectively, no cells were detected. After 25 d at higher pH levels (3.9, 4.5, and 5.0) the reduction was 4.5, 1.3, and 0.6 log units. This fact could risk the occurrence of food poisoning, if OJ became contaminated with E. coli O157:H7. This is particularly true since its survival is longer than the length of time required for juice spoilage to occur. The authors found an optimal treatment (6 log reduction) of 550 MPa for 5 min at 20 ëC at pH levels of 3.4 to 4.5 and 550 MPa at 30 ëC at pH 5.0. Other studies have also shown the higher pressure resistance of the E. coli O157:H7 strain. Jordan et al. (2001) used two E. coli strains (type-strain and O157:H7) in AJ and OJ substrates. Both strains were more resistant to the pressure treatment in OJ than in AJ. Slight pH differences (higher pH in OJ), viscosity or other physicalchemical characteristics seemed to affect the microorganism resistance. A reduction of 1 log in E. coli O157:H7 was achieved in OJ, whereas 5 log was reduced in AJ after a treatment of 500 MPa for 5 min. The type-strain of E. coli was much more pressure-sensitive and after 350 MPa, a 5 log reduction was achieved in both substrates. After storage at 4, 25, and 37 ëC, a further 3.3 and 7 log reduction at 4 and 25±37 ëC temperatures, respectively, was achieved in O157:H7 strains in OJ. This fact also corroborates that refrigeration temperatures seem to protect the pressurized E. coli cells against the acidic environment.



45

In some cases, more than one bacterial strain can be present in the fruit juice and it is interesting to study the pressure resistance of E. coli strains composed of a cocktail. In this regard, Teo et al. (2001) and Whitney et al. (2007) used a cocktail of several E. coli O157:H7 strains related to different outbreaks to inoculate different juices (OJ, AJ, grapefruit (GJ), and carrot (CJ)). Treatment at 615 MPa for 2 min at 15 ëC was able to inactivate more than 5 log cycles in CJ and GJ samples but not in the rest of the substrates. After storage for 24 h, further inactivation was observed (3.2±5.6 log reductions). It seemed that acidresistant strains were also more resistant to pressurization. These differences among strains could be due to differences in their membrane composition and ability to repair membrane damage in acid environments after pressurization. In addition, differences in gene expression related to stress response could also contribute to increased resistance to pressure. Some strains of Salmonella are able to survive acidic conditions and therefore are present in the fruit-based products, if low hygienic conditions are present, raw material is highly contaminated, or the product is recontaminated after processing. Several studies have been conducted to assess the pressure sensivity of these acid-resistant strains (Table 3.1). Teo et al. (2001) and Whitney et al. (2007) have shown an optimal treatment of 2 min at 550±615 MPa and 15 ëC in the inactivation of several strains of Salmonella (S. hartford, S. muenchen, S. agona, S. enteritidis and S. typhimurium) in several fruit juices (AJ, OJ, GJ and CJ) achieving more than 5 log cycles in all samples. Several lactic acid bacteria species can spoil fruit juices in optimal conditions for their growth. Leuconostoc mesenteroides is one important species among them. L. mesenteroides inactivation was studied by Basak et al. (2002) in fresh and concentrated OJ (42ë Brix) after high pressure treatment. The survivor curves showed a biphasic phenomenon. The pressure instantaneously reduced the counts of the microorganism (4.4 log cycles at 400 MPa and 20 ëC), then the survivor curves followed first-order rate destruction. The study also showed that the effectiveness was significantly reduced in the concentrated OJ where both the lower aw and higher soluble solids content protected the microorganism from pressure. Most spore-forming pathogenic bacteria will not germinate or grow in an acidic environment. However, some spoilage spore-forming bacteria such as Alicyclobacillus acidoterrestris have been implicated in acidic beverage and fruit juice spoilage (Lee et al., 2006). A. acidoterrestris is a soilborne and thermoacidophilic microorganism, and can be present in the final product through soil adhering to the surface of fruits during harvest or by the water used during juice processing. Their spores can germinate, grow and cause spoilage in a pH range of 2.5 to 6.0, a level below the typical range for spore-forming bacteria. The germination and growth has been observed in OJ incubated at 44 ëC for 24 h and can occur even at higher temperatures (80 ëC) (Pettipher et al., 1997). Their spores are resistant to the normal pasteurization conditions normally applied to acidic fruit products and can germinate and grow during storage of retail products where spoilage can occur (Jensen, 1999). This


46


exceptional fact is possibly attributable to its unique cellular membrane composition containing ring structures (cyclohexane fatty acids) closely packed leading to a high stabilization of the membrane and retaining more divalent cations than other bacterial spores, explaining their high resistance to demineralization (Lee et al., 2006). The typical spoilage effects are organoleptic taint due the production of guaiacol leading to a `medicinal' or `phenolic' offflavor and light cloudiness (Lee et al., 2002). Some studies have investigated the pressure resistance of A. acidoterrestris spores to high pressure (Table 3.1). It seems that the spores are also baroresistant, primarily due to the dehydrated state of the core. The inactivation of bacterial spores seems to occur in two steps: high pressure germination and inactivation of germinated spores (350 MPa for 20 min at 50 ëC) (Alpas et al., 2003; Lee et al., 2006). Again, the aw of the fruit juice influence the degree of inactivation; HHP is less effective in concentrated juices. The contamination of concentrated fruit juices with A. acidoterrestris and the dilution to produce commercial juice will rapidly spread the microorganism. Molds and yeasts Yeasts such as fermentative Saccharomyces cerevisiae and Zygosaccharomyces bailii can spoil fruit juices. Both yeasts have the ability to produce ascospores. The ascospore protoplast has a structure similar to vegetative cells, but the wall consists of an outer and inner coat (Raso et al., 1998). Spore formation can be induced on fruit surfaces where low concentrations of sugars and ethanol are present. During juice extraction, ascospores may contaminate the juice, eventually causing spoilage. Ascospores are known to be more resistant than vegetative cells to different food processing and chemical and physical agents (Zook et al., 1999). Vegetative cells are easily inactivated by pressure and usually a treatment of 350 MPa at room temperature is enough to instantaneously reduce the initial yeast load in a fruit juice (Table 3.2) (Bang and Swanson, 2008). When studying the microflora of OJ, Parish (1998) found that the inactivation kinetics parameters of juice microbial flora were similar to the ascospores, meaning that a high percentage of the yeasts present in the juice were in the form of spores. This fact points out the need for validating high pressure treatment in a juice with high populations of ascospores. The fact than ascospores are more pressure-resistant than vegetative cells in fruit juices has been corroborated in several studies (Raso et al., 1998; Parish, 1998). In ascospore survivor curves, a shouldering effect, mainly at low pressure values (300 MPa), indicates that at lower pressures, a time threshold was required before substantial inactivation was observed. The pH of fruit juice (3.5±5) seems not to affect ascospores inactivation by HHP. In addition, no differences are observed among different juices and model systems indicating no protective effects (Zook et al., 1999). This is different than the situation with bacteria, where inactivation increases at pH lower than 4.5. This phenomenon corroborated the ability of yeasts to grow at low pH. Comparing the inactivation


Table 3.2

HHP inactivation of molds and yeasts in fruit juices


Juice

Microorganism

HHP conditions


Reference

Apple and cranberry (concentrated)

Byssochlamys nivea (ascospores)

21 and 60 ëC Continuous (C): 689 MPa, 5±25 min Oscillatory (O): 689 MPa, 1±5 pulses

C-21 ëC: No inactivation C-60 ëC: 0-1 log O-21 ëC: No inactivation O-60 ëC-aw 0.98: 4 log O-60 ëC-aw 0.94: 0-1 log

Palou et al. (1998)

Orange

Saccharomyces cerevisiae (veg. cells and ascospores)

350±500 MPa for 1±300 s

Ascospores: 500 MPa (D value-4 s, z value-123 MPa) Veg. cells: 500 MPa (D value-1 s, z value-106 MPa) Microflora: 500 MPa (D value-3 s, z value-103 MPa)

Parish (1998)

Orange, apple, pineapple, cranberry and grape

Zygosaccharomyces cerevisiae (veg. cells and ascospores) (ATCC 36947)

300 MPa for 15±25 min at 25 ëC

Veg. cells: 300 MPa at 25 ëC for 5 min (5 log) Ascospores: 300 MPa at 25 ëC for 30 min (1-3 log) Orange > cranberry > apple > grape > pineapple

Raso et al. (1998)

Orange and apple

Saccharomyces cerevisiae (YM-147) (ascospores)

300±500 MPa for 1 s±30 min

OJ: 500 MPa (D value-0.18 min, z value-117 MPa) AJ: 500 MPa (D value-0.15 min, z value-115 MPa)

Zook et al. (1999)

Table 3.2

Continued


Juice

Microorganism

HHP conditions


Reference

Orange (fresh and concentrated)

Saccharomyces cerevisiae (ATCC 38618)

100±400 MPa for 15±60 min at 20 ëC

Fresh OJ: 250 MPa (D value-5.4 min, z value-135 MPa) Conc. OJ: 400 MPa (D value-23.5 min, z value-287 MPa)

Basak et al. (2002)

Apple

Talaromyces avellanus (veg. cells and ascospores)

200±600 MPa for 10±60 min

Veg. cells: 200 MPa, 17 ëC, 20 min (5 log) Ascospores: 600 MPa, 60 ëC, 50 min (5 log)

VoldrÏich et al. (2004)

Orange and pineapple

Saccharomyces cerevisiae

1±20 pulses, 100± 250 MPa, 25± 45 ëC, 30±120 s

6 pulses of 200 MPa at 45 ëC for 60 s (5 log) 7 pulses of 250 MPa at 45 ëC for 60 s (7 log)

DonsõÁ et al. (2007)

Apple

Saccharomyces cerevisiae (ATCC1664)

138±414 MPa for 30 s

354 MPa for 30 s at 23 ëC (6 log)

Bang and Swanson (2008)

Pinneapple (nectar and juice)

Byssochlamys nivea (ascospores)

550 and 600 MPa for 3±15 min at 20±90 ëC

Nectar: 600 MPa for 15 min at 90 ëC (5.6 log) Juice: 600 MPa for 15 min at 80 ëC (5.7 log)

da Rocha Ferreira et al. (2009)


49

data of several studies on yeasts with those obtained by other authors on bacteria species confirmed that yeasts are more sensitive to pressure than non-sporulating bacteria. In addition, a lower pressure-resistance of yeast ascospores is observed in comparison with bacterial spores (Reyns et al., 2000). The ascospores are larger and higher in lipid and carbohydrate content, and do not have dipicolinic acid and a cortex, which is considered very important in the resistance of the spore due to its ability in maintaining the state of osmotic dehydration of the protoplast. However, when using a concentrated juice (42ë Brix) the aw can affect the S. cerevisiae inactivation. The inactivation rate significantly decreased in the concentrated OJ. The high content of soluble solids or low aw seems to be the main reason for the low inactivation rate (Basak et al., 2002; DonsõÁ et al., 2007). Fruits used for juice processing can be contaminated with molds. Most fungi are heat sensitive and usual pasteurization processes in fruit juices are adequate to inactivate them. However, several molds, especially ascospores producing molds, have high heat-resistance even at elevated temperatures of 90 ëC (Palou et al., 1998). Ascospores of several Ascomycetes such as Byssochlamys nivea have been found to contaminate commercial fruit juice concentrate due to its wide pH growth range (2.0±9.0). Other molds such as Talaromyces avellanus can also be found as natural contamination in fruit products (Voldùich et al., 2004). Regarding the effect of physico-chemical characteristics of substrates on efficacy of HHP technology, the water activity or soluble solids content plays an important role (Table 3.2). This is especially true when HHP treatment is considered for the pasteurization or decontamination of concentrated fruit juices where severe treatment conditions are necessary. 3.3.2 Aspects related to food quality Enzymes Enzymes are a special type of protein with enormous catalytic power and great specificity. Their biological activity arises from active sites brought together by a three-dimensional configuration. They have two important regions; one that recognizes the substrate and the other that catalyzes the reaction once the substrate has been bound (Hendrickx et al., 1998). These two are called the active site and take place in a small part of the enzyme total volume. Changes in the active site interfering in the enzyme-substrate union or protein denaturation can produce an activity loss or functionality variations (Tsou, 1986). In general, covalent bonds are not affected by HHP treatment because the primary structure of the enzyme will not be damaged. The hydrogen bonds are also relatively baroresistant and the secondary structure will not be affected up to pressure values around 700 MPa. However, HHP treatment affects electrostatic and hydrophobic interactions that maintain the tertiary and quaternary structures stability (Ludikhuyze et al., 2002). Within the food quality related enzymes, the most important in fruit juices are the following:


50


· Polyphenoloxidase (PPO) which is responsible for enzymatic browning · Pectinmethylesterase (PME) which is responsible for cloud loss and consistency changes · Peroxidase (POD) which increases the production of undesirable flavors. In fruit juices, enzyme baroresistance is generally higher than the majority of naturally found microorganisms. For that reason, fruit juice preservation treatment is based on the inactivation of the enzymes responsible for its quality deterioration (PME in citrus juices, PPO in apple juice, among others). However, in some cases, no relation is observed between enzyme baroresistance and thermoresistance. Enzyme baroresistance also depends greatly on: · the type of enzyme ± enzymes with different three-dimensional structures containing different percentage of -helix, -sheet, -turn and random coil; · the source of the enzyme ± from native enzyme to purified form extracted from different parts of the plant; · the nature of the system ± from buffer to real food with more complex composition where different interactions can be produced, different physicochemical characteristics such as pH, sugar content, aw and pulp concentration, different fruits, varieties and harvest season will lead to different enzyme structure; and · the process conditions ± combinations of pressure, temperature and time.

A primary purpose of HHP processing is food preservation by maintaining the quality of fresh product and thus inactivation of quality deteriorative enzymes. Among these enzymes, PME has been extensively studied by HHP technology in order to find the optimal inactivation conditions. PME is a texturerelated enzyme mainly associated with the pulp content of citrus juices. It destabilizes the suspension formed by pectin mycelia (cloud loss) leading to a clarified product with low commercial value. Different studies have demonstrated the baroresistance of the enzyme to elevated pressures in OJ. Treatments below 500 MPa at room temperature seem not to affect the native PME activity in OJ (Cano et al., 1997; Nienaber and Shellhammer, 2001). Processing conditions of 600±700 MPa for 1±3 min combined with mild temperatures (50±60 ëC) seems to be effective in inactivating the native PME (Nienaber and Shellhammer, 2001; Polydera, et al., 2004, Sampedro et al., 2008), which stabilizes the OJ between 90 d and 16 wks at 4 ëC (Parish, 1998; Goodner et al., 1999). Low pH seems to enhance the PME inactivation while increasing soluble solids content (concentrated OJ) seems to decrease the effectiveness of HPP showing a protective effect (Basak and Ramaswamy, 1996). Regarding the application of HHP technology to smoothies, Sampedro et al. (2008) studied the influence of HHP processing in a beverage based on a mixture of orange juice and milk (50% of OJ, 20% of milk, 30% of water, 0.3% of pectin and 7.5% of sugar). The authors found optimum conditions for PME inactivation at 90 ëC for 1 min or 700 MPa at 55 ëC for 2 min showing the protective effect of the orange-milk



51

media. In addition, PME was more thermostable and baroresistant in the OJmilk based beverage system than in OJ. POD is an oxidoreductase related to the oxidation of a wide range of natural substances present in fruits, especially those containing aromatic groups. The mode of action involves the generation of free radicals able to abstract hydrogen from such substrates. Usually, hydrogen peroxide or oxygen act as oxidizing agents. POD contributes to phenolic oxidation leading to deteriorative changes in flavor, texture, color and nutrition. In OJ, POD is involved in the loss of flavor quality (EÂlez et al., 2006). Several studies have shown the high baroresistance of POD in different substrates. In one study, Cano et al. (1997) achieved only a 25 and 50% inactivation of POD in strawberry (230 MPa for 15 min at 43 ëC) and OJ (400 MPa for 15 min at 32 ëC). Below or above these conditions, higher enzyme activity was observed. It seemed that the activation was pH dependent and at higher pH, the activation was strong as in the case of strawberry. In this sense, GarcõÂa-PalazoÂn et al. (2004) only observed a 35% inactivation of POD in strawberry after 600 MPa for 15 min at 20 ëC and no more inactivation was achieved with pressures increasing up to 800 MPa, whereas Fang et al. (2008) observed a residual activity of 30% after 600 MPa for 30 min at 50 ëC. PPO is an oxidative enzyme responsible for undesirable color changes, undesirable flavors and nutritional losses. It is mainly related to the browning reactions, catalyzing the hydroxylation of mono-phenols, leading to the formation of di-phenols and the following oxidation of di-phenols to form quinones in the presence of oxygen. Next, the condensation of quinones generates dark substances (melanines) which negatively influence the quality and marketability of commercial fruit juices (Giner et al., 2002). Cano et al. (1997) studied the effects of HHP processing on PPO in strawberry puree. A maximum of 60% of inactivation was achieved combining 285 MPa at room temperature. At higher pressures or temperature levels, higher enzyme activity was observed (enzyme activation). The low pressure and temperature conditions applied in this study could explain the low treatment effectiveness and the activation phenomenon. Palou et al. (1999) studied PPO inactivation in a banana pureÂe. Pressure treatment alone (689 MPa for 10 min at room temperature) was able to reduce enzyme activity by only 20%. Only after a blanching treatment (saturated steam for 7 min) followed by pressure treatment (689 MPa) they were able to reduce the initial activity by more than 95%. In a later study, GarcõÂa-PalazoÂn et al. (2004) showed a high baroresistance of PPO in red raspberries, resulting in a remaining activity of 70% after 800 MPa for 15 min. In contrast, PPO in strawberries was more sensitive to the pressure treatment and 600 MPa for 15 min or 800 MPa for 10 min was enough for a complete inactivation. The authors linked the enzyme stability in raspberries and strawberries with the stability of their anthocyanins and the consequent color loss. Higher PPO activity decreased the stability of anthocyanins. Because raspberries retain high levels of activity of PPO after HPP treatment, anthocyanins in raspberries were more susceptible to HHP treatment.


52


Color The color observed by human beings is the perception of the wavelengths coming from the surface of the object on the retina of the eyes (Tijsken et al., 2001). Food appearance can change depending on the amount of light, the light source, the observer's angle of view, size and background differences. However, standardized instrumental color measurements used in the food industry such as the Hunter color (L*, a* and b*). L* is a measure of brightness/whiteness that ranges from 0 to 100 (white if L* 100, black if L* 0), a* is an indicator of redness that varies from ÿa* to a* (ÿa* green, a* red) and b* is a measure of yellowness that varies from ÿb* to b* (ÿb* blue, b* yellow). The CIELAB system is used as a quality index in fruit juices to assess the conformity to specifications or measure the changes as a result of food processing or storage (Giese, 2000). Maintaining the natural color of fruit juices is a major challenge to the application of HHP technology, because color is the first characteristic that is noticed in food and predetermines consumer perceptions of freshness and expectations of both flavor and quality (Rodrigo et al., 2007). The changes in the natural color of fruit juices are based on the degradation of pigments by enzymatic and non-enzymatic reactions. Compounds such as anthocyanins are responsible for the color in some fruits. Rodrigo et al. (2007) studied the degradation kinetics of strawberry juice color after HHP processing and concluded that the combination of L*, a* and b* parameters in the form of L*x a*/b* was the most accurate way to describe the color degradation in strawberry juice. No differences were found between treated and control samples up to 700 MPa for 60 min at 65 ëC. However, the effect of pH was found to be significant in the strawberry samples. The a* parameter after HHP processing increased as the pH increased from 3.7 to 5. It seems that pelargonidin-3-glucoside anthocyanin is the main compound responsible for the red color in strawberries and it is not stable in a pH range of 5±7. Sensory and consumer studies Some sensory studies have been conducted in fruit juices in order to demonstrate the advantages of HHP processing versus thermal pasteurization on preserving the natural sensory properties. Some studies have used trained sensory panelists in order to compare the freshness and acceptability of samples treated by HHP and traditional thermal pasteurization. However, due to their training, sensory panelists may not be representative of the typical consumers of fruit juices. In these cases, consumer acceptance studies are necessary. In addition, the analytical profile of volatile compounds related to the fruit juices aroma is performed, to try to connect the unique flavor of fruit juices with specific chemical compounds. Baxter et al. (2005) studied changes in the sensory properties and flavor compounds during a 12-wk shelf-life storage of OJ processed by HHP and thermal pasteurization at 4 and 10 ëC. Ten trained sensory panel members were used for the descriptive sensory analysis of samples and 30±40 regular consumers of OJ participated in a consumer acceptability study. Regarding the



53

color, the trained panel did not observe differences among the different samples during the overall period. Increasing the duration of the storage period led to a decrease in the sweetness and strength of orange odors and an increase of aged, artificial and fermented odors. At the end of the storage period, consumers did not differentiate between control, thermally and pressure-treated samples at 4 ëC. However, scores were lower in the HHP for 10 ëC samples and were unacceptable for the thermally processed samples at 10 ëC. Twenty volatile compounds were analyzed in the storage of OJ at 4 and 10 ëC. Considerable reductions were found for most compounds in both HHP and thermally processed juices compared with the control sample at ÿ20 ëC. The compounds showing the greatest reductions were octanal, citral, ethyl butanoate and limonene with the final concentration of compounds 6±38% lower than the initial level. The decrease in the volatile compounds concentration during the storage was produced by a combination of factors. PET bottles used for the OJ storage seemed to absorb some volatile compounds. In addition, oxidation, hydrolysis and acid-catalyzed reactions were responsible for the degradation of volatile compounds. Working with an OJ-milk beverage, Sampedro et al. (2009a,b) studied the volatiles profile after HHP treatment. After HHP treatment (650 MPa for 15 min at 30 ëC), some volatile compounds increased. The authors argued that in a complex matrix such as OJ, with the presence of suspended solids, a portion of analytes could be entrapped in the pulp. HHP treatment could increase the membrane permeabilization and facilitate the release of several compounds from the suspended solids to the liquid phase, facilitating its extraction into the headspace. The average loss of volatile compounds concentration was between ÿ14.2 and 7.5% at 30 ëC and 22.9 and 42.3% at 50 ëC. As for tropical fruit juices, Laboissiere et al. (2007) conducted a study on the effects of HHP processing on the sensory characteristics of a Brazilian yellow passion fruit juice. They found very similar patterns for sensory properties between fresh passion fruit juice and HHP processed juice. The only parameter that differentiated both samples was color. In addition, a panel could distinguish between fresh and HHP treated samples and commercially pasteurized samples. The main sensory attributes that differentiated those samples were the presence of suspended particles, phase separation, natural aroma and flavor, artificial aroma and flavor, cooked aroma and flavor and fermented flavor. Most of these sensory attributes considered as sensory defects possibly resulted from the heat pasteurization, addition of artificial aromas, flavor compounds and stabilizers. Consumers are becoming more conscious about the potentially negative impact of food processing on their health and the environment. Healthy and natural foods are the most important area of research by the majority of food companies (Katz, 2000). `Fresh' remains the most desirable food label claim. Other aspects such as country-of-origin, organic food, local foods and environmental concerns, have continued to rank high in public attention (Deliza et al., 2005). A positive consumer attitude towards the use of HHP technology is necessary to guarantee the success of the product in today's competitive global market, where new food product innovation is required for survival. That means


54


that the role of the consumer in the technology validation process must be taken into account. In this sense, Deliza et al. (2003 and 2005) conducted two studies concerning the Brazilian consumer attitudes toward to the use of high pressure processing. They chose four consumer groups consisting mainly of women and fruit juice consumers. The product chosen for the study was pineapple juice processed by HHP with three different information labels (low, medium and high information). The first statement, pointed out by the majority, was the price as an important attribute during their decision-making process. The study also revealed that most consumers inferred the product taste based on the label information which affected the product expectation and perception. When the information about the technology was presented, three of the four groups perceived the product as having higher quality. The information about the technology had a significant impact on the intention of purchase. However, information about the technology without further explanation led to a negative impact on the consumer intention of purchase. These results lead to important information for the food producers. The information about the technology in the product label is essential for the product to be perceived by consumers as higher quality. In addition, factors influencing fruit juice purchasing include convenience, taste and cost. In a later study, Nielsen et al. (2009) conducted a consumer study using focus groups in six European countries (two from northern Europe and four from eastern Europe). A baby food and fruit juice were used as selected products for focus group discussions. Participants were positive towards HHP products for naturalness, improved taste, and high nutritional value (high vitamin content). Longer shelf-life in comparison to fresh squeezed juice products, high prices and lack of information were seen as negative toward the technology. Environmentally friendly and natural products (no preservatives) were positive towards the technology. Differences were observed among the different cultures. Participants from northern countries were more skeptical about the new technologies and were more worried about the impact on the environment, whereas eastern countries saw the higher price solely as negative. Differences were also seen among the products. Longer shelf-life and higher prices were seen negatively in fruit juices, since consumers are more accustomed to fresh squeezed juices. In contrast, participants saw higher price and longer shelf-life as positives for baby food; however, some participants were negative to HHP baby food, since it is not homemade. In a potential buying situation, quality and especially taste play a critical role in accepting and maintaining the commercial marketability of these novel products. Bioactive compounds Daily intake of fruits and vegetables has been related with the prevention of degenerative processes such as cardiovascular disease and certain cancers. This protective action has been attributed to their bioactive compounds, which have antioxidant properties. HHP treatment is expected to be less detrimental than



55

thermal treatment to low molecular weight food compounds such as flavoring agents, pigments and vitamins as covalent bonds are generally not affected by pressure (Butz et al., 2004). Vitamin C is the most important water-soluble nutrient and is related to the antioxidant capacity of the OJ (CorteÂs et al., 2008). Regarding the effects of HHP treatment on the stability of vitamin C, ascorbic acid (L-AA) and total antioxidant capacity in OJ were studied. Specifically, SaÂnchez-Moreno et al. (2003b, 2005) and Plaza et al. (2006) studied the effects of different high pressure (100±400 MPa for 1±5 min at 30±60 ëC) and thermal treatments (70 ëC, 30 s and 90 ëC, 1 min) in an OJ stored at 4 ëC for 40 d. Around 10% of the vitamin C content was lost after the combined treatment (400 MPa for 1 min at 40 ëC or 100 MPa for 5 min at 60 ëC) but no loss was found at 30 ëC. They argued that the high contents after processing could be attributed to a partial elimination of enzymes (POD and ascorbate oxidase) responsible for L-AA and vitamin C oxidative degradation. At higher temperatures, greater decreases in vitamin C content were observed due to possible thermal degradation. During storage, the losses reached 24 and 32% after thermal and HHP treatments, respectively. The untreated sample only lost 10% of the initial content. These differences were attributed to the different levels of enzyme inactivation (POD and ascorbate oxidase) achieved by the treatments that could degrade the L-AA during storage by an oxidative process. The antioxidant capacity was unaffected by HHP and low pasteurization processes, whereas high pasteurization processing (90 ëC for 1 min) reduced the antioxidant capacity by 6.5%. The authors found a correlation between L-AA, vitamin C and antioxidant capacity, thus the high stability of both compounds after the different treatments also stabilized the antioxidant capacity. On the other hand, no correlation was found between total carotenoids and flavonones and antioxidant capacity, indicating the lack of relevant effect of these compounds on the total OJ antioxidant capacity. The type of packaging used in the storage of the fruit juice after processing can influence the stability of the antioxidant capacity of OJ. Polydera et al. (2003) studied the degradation kinetics of ascorbic acid after a HHP treatment (500 MPa at 35 ëC for 5 min) and thermal treatment (80 ëC for 20 s) using two different packaging materials, an intermediate oxygen barrier (polypropylene bottles) and a high oxygen barrier (polyethylene, aluminum and cellophane) during 1±2 months storage of an OJ at 0, 5, 10, and 15 ëC. When polypropylene bottles were used, the degradation kinetics of L-AA during the storage period seemed to follow a first-order reaction. The rate parameter (k) was lower in the pressure-treated sample than in the thermally-treated sample, indicating HPP juice had a lower degradation rate in L-AA than thermally processed juice during storage. The degradation rate of L-AA increased in both samples as storage temperatures increased. When flexible pouches were used, the degradation rates seemed to have two stages. The first part followed first-order kinetics and the second part followed zero-order kinetics. The more rapid decrease of LAA at the beginning of storage can be attributed to autoxidation, the reaction of L-AA with dissolved oxygen, and then the lower rates could be controlled by the low diffusion of oxygen of the material or by an anaerobic decomposition.


56


Inactivation rates of anaerobic decomposition are usually 2±3 orders of magnitude lower than the oxidative degradation (Gregory, 1996). For these reasons, the inactivation rates were significantly higher in the propylene bottles at both treatments. The authors estimated the shelf-life of OJ based on the vitamin C content (>20 mg/100 mL of vitamin C in the OJ at the expiration date) regulated by the Association of the Industry of Juices and Nectars from Fruits and Vegetables of the European Union. Using polypropylene bottles, the shelflife at 5 ëC was estimated to be 50 and 34 d for the HHP and thermal samples, respectively. When the storage temperature was increased to 15 ëC, the shelf-life period decreased to 20 and 18 d, respectively. On the other hand, using the flexible pouches, the shelf-life was estimated to be 90 and 62 d after HHP and thermal treatment at 5 ëC and 62 and 50 d at 15 ëC, respectively. The lower degradation rates of vitamin C in the HHP-treated samples extended the shelflife compared to thermal pasteurization. Owing to the high oxygen barrier of the flexible pouches, the shelf-life was increased with respect to the polypropylene bottles. The sensory evaluation during the shelf-life indicated higher scores for the pressure-treated samples when comparing with the thermal ones. The degradation of water soluble vitamins (vitamin C, B1 and B6) after high pressure treatments were studied by Sancho et al. (1999) using a model system and strawberry smoothie to check the effects of HHP treatment as well as thermal pasteurization (76 C for 20 s) and sterilization (120 C for 20 min) on water soluble vitamins. In the model system, the vitamin C content varied from 10 to 12% after 200±600 MPa for 30 min at room temperature. Changes in B1 and B6 vitamins after the pressure treatment were insignificant. In the strawberry system, changes in vitamin C content were not significant after the HHP processing and thermal pasteurization. However, a 33% loss was observed after the sterilization process. Carotenoids are among the most abundant bioactive compounds in fruits and have diverse biological functions and actions. Provitamin A activity carotenoids and xantophylls are known to provide protection against macular degeneration. They also are potent antioxidant and free-radical scavengers and modulate the pathogenesis of cancers and coronary heart disease (Torregrosa et al., 2005). Studies conducted by de Ancos et al. (2002) and SaÂnchez-Moreno et al. (2003a) showed an increase in the total carotenoid content by 23 and 43% after a pressure treatment at 100 and 350 MPa, respectively. Regarding the individual carotenoids, -carotene increased by 50%, -carotene by 60%, -cryptoxanthin by 42% and -cryptoxanthin by 63% after 350 MPa for 5 min at 30 ëC. Differences in the oxygen and hydrocarbon carotenoids in the intracellular locations of juice vesicles could lead to variability in the release of carotenoids after HHP treatment. The higher carotenoid content could be explained by the release of carotenoids from the food matrix (orange cloud) after denaturation of protein-carotenoid complexes induced by pressure. The carotenoids extraction was pressure-dependent and increased at higher pressure levels. These changes could increase the amount of antioxidant carotenoids available, improving the bioavailability and absorption in HHP treated juices. The effect of treatment



57

time was checked at 350 MPa at 30 ëC. The carotenoids content was increased by 43% after 5 min, but increasing the treatment time to 15 min did not show any further improvement. The effect of temperature (30 and 60 ëC) was studied at 100 MPa. Results showed that an increase in the temperature did not exhibit any improvement in the carotenoid content. The carotenoid degradation is mainly due to oxidation and geometric isomerization. The isomerization of carotenoids goes through covalent bonding rupture and it seems that pressure does not significantly affect covalent bonding. During the refrigerated storage of OJ after different treatments, no losses were found after 15 d at 4 ëC. At the end of storage (30 d) losses in the 50 MPa and control samples were 17 and 42%, respectively, whereas the 350 MPa sample had 72% higher content than the fresh one. Long treatment times increased carotenoid content after storage (30 d) with 68, 72 and 100% increasing after 2.5, 5 and 15 min, respectively. The sample at the higher treatment temperature (60 ëC) showed higher losses during the storage (28%) with respect to the sample at 30 ëC. Vit-A carotenoid content increased (52 and 45%) after treatments at 100 and 300 MPa. Neither increasing the time nor temperature increased the carotenoid content. During the storage, Vit-A carotenoid content increased in the 200 and 350 MPa samples (36 and 63%) but decreased in the control, 50 and 100 MPa samples (9, 42 and 24%). It seemed that at lower pressure, Vit-A carotenoid losses were higher possibly due to the residual POD activity that survived the lower pressure treatments. Flavonoids belong to a group of natural substances with variable phenolic structures and are found in different quantities in fruit juices. They possess antiinflamatory, antiallergic, anti-viral, hypocholesterolemic, and anticarcinogenic properties (SaÂnchez-Moreno et al., 2003a). Orange juice is a dietary source of flavonoids, mainly flavanones. SaÂnchez-Moreno et al. (2003a, 2005) studied the effects of HHP treatment at different temperatures and thermal pasteurization (70 ëC for 30 s and 90 ëC for 1 min) on the main flavanones (hesperidin and naringenin) in OJ. Just after pressure treatments at 350 and 400 MPa, the naringenin content increased by 13 and 12%, respectively, and by 34 and 22%, respectively, for hesperidin content. They argued that some structural changes and permeabilization of cell walls of OJ sacs could release phenols from proteins and increase the extraction of flavanones. Pasteurization processes led to diminishing naringenin content (16.0%) but hesperidin content was unaffected. During the storage of pressure treated samples, the naringenin content increased around 11% for 350 and 400 MPa samples and no differences were observed between the 100 MPa samples and the control. Regarding hesperidin content, an increase was observed for 350 and 400 MPa samples (19 and 21%) with no differences among the 100 MPa sample and the control. The authors argued that the remaining activities of POD and PPO suggested for the degradation of polyphenols in the 100 MPa sample could explain the lower content of flavanones at that processing condition. Flavanones tend to precipitate at low pH from the soluble fraction to the cloud in OJ, leading to an increase in the proportion of flavanones in the cloud after processing. For that reason, a lower release of naringenin after processing during extraction could occur.


58


Folates are hematopoietic vitamins with special importance during pregnancy. Regarding the nutritional needs of humans, folate deficiency occurs frequently, probably due to poor diet selection and losses during food processing (Sauberlich et al., 1987). Fruits are an important source of folate. For example, a 200 g portion of fresh oranges contains nearly 50% of the recommended daily intake (Butz et al., 2004). Very little data has been published in relation to the effects of HHP treatment and folate stability. One study conducted by Butz et al. (2004) studied the stability of three main types of folates found in OJ (tetrahydrofolate, 5-methyltetrahydrofolate and 5-formyltetrahydrofolate) after HHP treatment (600 MPa at 25 and 80 ëC). An orange juice model containing ascorbic acid was used to check its protective effect against pressure in the folates content. At 25ëC, the losses ranged from 10 to 40% and at 80 ëC from 25 to 95% after a 24 min treatment. The pressure sensivity was as follows: 5methylfolate>5-formylfolate>tetrahydrofolate. To check the thermal effects on the folates content, the authors performed a thermal treatment at 80 ëC, observing that heat alone decreases folates content by 20, 50 and 80% in 5methylfolate, 5-formylfolate and tetrahydrofolate, respectively. Combinations of HHP and thermal treatment did not increase the losses in 5-methylfolate but increased the losses in 5-formylfolate and tetrahydrofolate by 30 and 20% respectively. This indicates that small molecules such as vitamins are stable to HHP treatment and do not undergo cleavage of covalent bonds, certain reactions are accelerated by pressure. This is the case with 5-formylfolate where the formation of a 5, 10-methenylfolate derivative is accelerated by pressure. When comparing the model juice with fresh squeezed orange juice, the authors observed that natural folates in OJ were more stable than in the model system after the high pressure treatment. As the model and fresh juice had the same LAA, the authors suggested that the presence of other substances such as vitamin C and flavonoids in the natural juice protected folate. Anthocyanins play an important role in the antioxidant and antiradical capacities of fruits. Twenty anthocyanins are known, but only six (pelargonidin, cyanidin, peonidin, delphinidin, petunidin and malvidin) are important in food (Zabetakis et al., 2000). They are known for their brilliant red and purple colors and products containing anthocyanins are more susceptible to color changes during processing and storage. The degradation of anthocyanins can be influenced by different parameters such as temperature, enzymes, oxygen, and sugar content. Grape juice is known for its high content of polyphenolics and anthocyanins. Talcott et al. (2003) and del Pozo-Insfran et al. (2007) studied the effect of HHP (600 MPa for 15 min) and thermal (90 ëC for 15 min) treatments on anthocyanin content of a grape juice. The fortification of L-AA is a common practice in the food industry to protect against oxidation, but its combination with anthocyanins may be mutually destructive in the presence of oxygen. Treatment at 400 MPa caused the highest loss (70%) of anthocyanin content due to the activation of PPO activity, whereas at 600 MPa or thermal treatment achieved low losses (3± 5%). The presence of L-AA enhanced the losses (12.4 and 18.1%) after thermal



59

and HHP treatments, respectively. In addition, the formation of hydrogen peroxides from the oxidation of L-AA may contribute to the degradation of anthocyanin. Furthermore, these peroxides may activate residual POD, which further degrades L-AA and anthocyanins. After 21 d of storage at 25 ëC, the anthocyanins content were reduced by 28±34% for all the samples. By-products from the degradation of L-AA and/or monosaccharides, such as furfuraldehydes, could contribute to anthocyanin degradation during storage. The individual anthocyanins differed in their resistance to the HHP treatment. The anthocyanins containing Èõ-diphenolic groups were less stable to HHP processing, possibly because they were more susceptible to the enzymatic oxidation. Strawberry juice is also a medium in which thermal treatment can damage the natural color by the degradation of anthocyanins, thus the use of HHP technology can be a challenge. Zabetakis et al. (2000) focused on the pelargonidin derivatives, 3-glucoside and 3-rutinoside, which are the main anthocyanins in strawberry. At 4 ëC, the losses the of 3-glucoside derivative were similar for all pressure treatments (400±600 MPa at 22 ëC) and the control sample lost 20% at the end of 9 d of storage. However, at 400 MPa the losses were higher and reached 40% after 9 d. The behavior of 3-rutinoside was similar, with higher losses after 400 MPa (50% after 9 d) and similar losses after 600 MPa and control sample (25%). However, at 200 and 800 MPa, the losses were lower (10%). At 20 ëC, there were no differences in the losses of 3-glucoside of all samples at the end of storage (50%), whereas 3-rutinoside treated at 200 and 800 MPa had a higher anthocyanin content (25% losses). At 30 ëC, there were no differences in 3-glucoside in all samples (70±80% losses after 9 d), whereas no differences (60±80%) in 3-rutinoside were found except for 200 and 800 MPa with losses of 42%. The authors explained this behavior by the residual enzyme activity after the different pressure treatments. Three main enzymes are involved in the oxidation of the anthocyanins: POD, PPO, and -glucosidase. These enzymes have lower activity at lower temperatures, thus the losses of anthocyanins were lower at 4 ëC in comparison with 20 and 30 ëC. In addition, glucosidase seems to be activated at pressures around 400 MPa, which could explain the higher losses at that pressure. At higher pressures (800 MPa) the activity of these enzymes is irreversibly reduced, corresponding to lower losses in anthocyanin content. The differences in the degradation of both derivatives could be explained by the differences in substrate specificity of -glucosidase which was higher for 3-glucoside than 3-rutinoside, and the breakdown would therefore be more extensive and losses higher for 3-glucoside. Blackcurrants contain high levels of flavonoids with anthocyanins as the most important group. They are present in the skin of the berries and are responsible for the characteristic aroma and color. Kounaki et al. (2004) studied the effects of pressure (200±800 MPa for 15 min at room temperature) on two important anthocyanins in black currant, delphinidin-3-rutinoside and cyanidin-3-rutinoside during the storage at 5, 20 and 30 ëC for 7 d. At 5 ëC, the losses of anthocyanins were the lowest because they were mediated by enzymatic action (PPO activity) and the enzyme activity would be low at that temperature. Losses of delphinin-3-


60


rutinoside and cyanidin-3-rutinoside contents reached 58 and 40%, respectively, after 7 d of storage at 4 ëC. Treatment at 600 MPa seemed to retain higher levels of both anthocyanins. At 20 ëC, the losses reached 58% for both anthocyanins. The sample treated at 200 MPa counted for the highest losses (55±60%). At 30 ëC, the losses were around 70±75%, and the samples treated at 200 MPa had the lowest anthocyanin content with less than 20%. The authors found a relation between LAA content and anthocyanin degradation, where the rapid loss of L-AA appeared to contribute to the lower rate of anthocyanin loss. Also, the higher loss after 200 MPa could respond to an oxidative enzyme activation.

3.4 Commercialization of juices treated by high hydrostatic pressure (HHP) 3.4.1 Key drivers to employ novel rather than conventional processing technologies High hydrostatic pressure processing provides a unique opportunity for food processors to develop generations of new, value-added food products having a superior quality than those produced by conventional thermal methods. These processes can help meet the challenges of producing innovative products from natural sources without compromising biologically active compounds, while ensuring foods with low microbial counts of spoilage organisms and safe from pathogens. Further, HHP can preserve food products without heat or chemical preservatives to significantly extend refrigerated shelf-life has opened new market opportunities, particularly in the area of `natural' preservative-free, wholesome products. Tropical fruit juices are another area of interesting consideration with regard to HHP processing. Tropical fruits have gained popularity in the last several years due to their unique flavors, aromas, and colors and their annual production is unlimited by seasonality. Some Latin American countries such as Brazil, Colombia, Ecuador, Mexico and Costa Rica are significant producers of tropical fruits. However, in some cases, the production of these local fruits is lost due to the lack of a means of commercialization. HHP technology could be a viable alternative to process these fruits in the form of juice and pulps, to retain the original sensory properties while extending their shelf-lives. This could improve trade exportation to demanding markets such as the EU and USA. The Food and Drug Administration (FDA) has recently accepted pressureassisted thermal-sterilization (PATS) as a process for commercial application in low acid foods. It combines heat with high pressure to produce commercially sterile, low acid food, which improves the quality compared to thermally processed foods, while simultaneously eliminating the food safety risks associated with pathogenic bacterial spores such as C. botulinum and its toxins. Although most juices have a pH below 4.5 and are considered as àcidic foods', the pH varies in relation to juice composition. Undoubtedly, this new regulation will help increase the number of commercial HHP applications available.



61

3.4.2 Commercial application HHP technology has become a commercially implemented technology in fruit juice processing, spreading from its origins in Japan, followed by USA and Europe, and now Australia with worldwide utilization increasing almost exponentially since 2000 (Norton and Sun, 2008). In the US, Genesis Juice Corp. processes eight types of organic juices by HHP including apple, carrot, apple-ginger, apple-strawberry, ginger lemonade, strawberry lemonade, a herbal tea beverage, and apple- and banana-based smoothies. European companies presently employing this technology in fruit juice processing include smoothies by Invo in Spain, orange and grapefruit juices and a mixture of strawberryorange juice by UltiFruit in France, FrubacËa manufacturing different fruit-based beverages in Portugal, Juicy Line-Fruity Line in Holland, Beskyd Frycovice, a.s manufacturing mixtures of broccoli-apple-lemon and broccoli-orange-lemon in Czech Republic, and ATA S.P.A manufacturing carrot and apple juices in Italy, and Puro commercializing smoothies in the UK (Table 3.3). Some interesting information and general trends can be drawn from Table 3.3. Most of these companies incorporating HHP technology did so before the establishment of the international standards of GMP (Good Manufacturing Practice) for raw material suppliers and processes. Manufacturers of HPP juices are usually small to medium sized companies comprising less than 100 employees. Significant collaborations with national food technology research bodies have been developed in order to accomplish shelf-life, consumer studies, and to satisfy legal and regulatory requirements. Regarding the processing conditions, treatments are optimized at a pressure level of 600 MPa in combination with moderate heat. In addition, due to the special characteristics of fruit juices, small productions are achieved ranging from 50 to 200 kg/h to satisfy consumer demand. Shelf-lives are estimated at ca. 10±35 d of refrigeration conditions, depending on the type of juice. Products are sold at supermarkets chains, specialty and gourmet stores, and food services providing fruit preparations and dressings. Two main packaging formats are used, a small volume of 250 mL quantity corresponding to a single portion, and a bigger format of 1 L. Market prices are around ¨3 or $4.5 per 250 mL serving. Marketing is a key instrument used by companies to highlight the benefits of their HHP products compared to the competition, and advertising generally emphasizes that the fruit products are natural, supplemented with vitamins, and can even be considered as sport beverages. Case study: HHP fruit juice processing in Australia In the last few years, Australia has become a leader in the developing of HHPtreated juices and derivatives. The Australian Research organization (CSIRO) and their joint venture Food Research Group, Food Science Australia have been developing high pressure processing systems in Australia for over a decade. Several companies are using HHP technology in their processes. A questionnaire was compiled and sent from the authors of the present chapter to the fruit juice manufacturing companies to obtain the relevant data


Table 3.3

Relevant information of HHP fruit juice manufacturing companies


Company

Company data and products

Treatment/ production

Shelf-life

Outlets/ points of sale

Invo (Spain)

Trade mark: Invo HHP smoothies: Orange+banana, strawberry+banana, pineapple+banana, blackberry+apple and 4 oranges 250 and 750 mL bottles; ¨3 per 250 mL

600 MPa

5 ëC

Gourmet stores

Beskyd Frycovice, a.s. (Czech Republic)

Trade mark: Refit HHP juices: Broccoli+apple+lemon, broccoli+orange+lemon 300 mL bottle (pH < 4.2)

±

10 d 5 ëC

Supermarket

Juicy Line-Fruity Line (The Netherlands)

100 employees Trade mark: Juicy-Line and Walking fruit Juices and smoothies

600 MPa

28 d

±

Pressure Fresh Australia Pty Ltd. (Australia)

30±100 employees; 13 y in operation Trade mark: Austchilli Thermal pasteurized fruit juices, vegetable and herb products. HHP products: Avocado, guacamole, pomegranate juice and fruit pureÂes Bags and bottles

50±200 kg/h

20-40 d 5 ëC

Food service and supermarkets

Preshafood Ltd. (Australia)

40 d 5 ëC

Supermarket chains, specialty stores, delis and restaurants


FrubacËa-Cooperativa de Hortofruticultores, C.R.L. Commercialized by: GL ImportacËaÄo ExportacËaÄo S.A. (Portugal)

Cooperative of 26 producers; 60 employees Juices (trade marks): Copa, Fresco, Pingo Doce 7 flavors (6 apple based and 1 orange); 750 and 250 mL Smoothies (trade mark): Sonatural 4 flavors; 125 and 250 mL

600 MPa

35 d 5 ëC

Supermarket

Ulti Fruit (France)

21 employees Trade mark: Ulti 2 lines production: fresh and pressurized (40%) HHP juices: orange (70%), grapefruit and strawberry+orange 250 mL and 1 L

±

16 d 5 ëC

Supermarket and hypermarket

ATA S.P.A. (Italy)

Trade mark: Gustivivi HHP juices: carrot, apple pH = 4.2

Carrot: 600 MPa/3 min Apple: 600 MPa/5 min

21 d 5 ëC

Puro (Northern Ireland, UK)

Trade mark: Puro HHP Smoothies 250 mL; £1.89±2.5 per 1 L

Genesis Juice Corp. (USA)

Trade mark: Genesis Organic HHP organic juices: apple, carrot, apple-ginger, applestrawberry, ginger lemonade, strawberry lemonade, herbal tea beverage and apple and banana smoothie based 240 mL

15±21 d 5 ëC

Supermarket

Supermarket

64


about the industrial experience of using HHP technology. General trends were drawn from the questionnaire. Companies were small to medium in size (from 30 to 100 employees). Although most of companies have been in operation for over a decade, the introduction of HHP processing occurred more recently. In addition, thermal processing is still used in many of the companies for the production of vegetable and herb products, stabilized fruit preparations for yogurt, ice-creams (ripples) and fruit ices, diced chicken (cooked) and egg and mayonnaise salad for sandwiches. The main reasons reported by companies for choosing HHP technology were their efforts to provide a product with a clearly perceptible improvement and the ability to achieve a premium price for a premium product. Companies added that HHP products are superior in taste, texture and color to current competitor products, and consumers have to be informed about these differences. In addition, the product must be value-added to be able to support the extra processing costs. Supply chains must be in place or developed for the delivery of high quality raw materials, to ensure that the products are microbiologically safe. Optimum market survival is improved by creating new products that have never been made before. Fresh processed juices, stabilized fruit preparations, chilled packaged salads, whole and value-added egg products and processed diced chicken products for sandwich bars, chilled cooking sauces and curry pastes and guacamole, pomegranate juices and fruit pureÂes are the main products processed by companies using the HHP technology. When asked if sensory tests were performed, companies answered that small group sensory tests were conducted based on preference testing to confirm that the product flavor and color benefits could be translated to the consumer's preference and purchase intent. In addition, changes in the product labeling were introduced after HHP processing, highlighting the benefits of HPP as a cold pasteurization process that produces the most natural, freshest tasting, and most nutritious juice on the market (Fig. 3.1). The main clients of HHP fruit juice products were supermarket chains, and specialty and deli stores. No special legal approvals were required for the domestic products. Regarding consumer attitudes, no negative attitudes, apart from the higher price, were observed, and mostly there were difficulties explaining the technology and its benefits to the consumer without becoming too technical and either losing their interest or confusing them. One of the problems mentioned by HHP fruit juice processors was to develop a PET bottle for the unit that was extremely space-efficient to allow for maximum utilization of the cylindrical processing cavity to maximize throughput. A clever bottle design with a local supplier solved this issue by creating an immediately differentiated and individual wedge design. On average (due to differences in packaging size), the cost comparisons for the actual processing step was approximately 40% more expensive than traditional thermal pasteurization. The recovery of the investment was around 5± 10 yr. The packaging materials used in the HHP process included: blow molded PET bottles, standup high barrier laminated spout pouches, high barrier


ß Woodhead Publishing Limited, 2010 Fig. 3.1

HHP fruit juice labeling (Preshafood LtdÕ).

66


laminated bag in box systems, and high barrier laminated preformed bags, which were thermally sealed. Changes in the product formulation, such as the performance of hydrocolloid and polysaccharide stabilizers, post HPP treatment, meant that it was possible in some cases to actually reduce the amount of stabilizers required and still achieve the same viscosity, texture and mouthfeel.

3.5

Future trends

Taking into account the limitations of traditional thermal processes and the trends in consumer demand, the future of food preservation is moving towards ala-carte processing, consisting of a specifically designed process treatment for each type of food. There are foods for which traditional technologies are, and will continue to be the most efficient processing option. However, some market niches have appeared for emerging technologies to be viable for producing food products that are healthier, retain more of their fresh-like character, and, most importantly, are safe for the consumer from a microbiological point of view. Thus, there are cases where HHP is the most appropriate technology to meet consumer demand, and the use of HHP for juices and derivative products will likely continue to grow as costs decline and food manufacturers identify new applications where HHP can deliver product quality improvements that consumers demand and appreciate.

3.6

Sources of further information and advice

Equipment manufacturers Avure Technologies Inc. (www.avure.com) Elmhurst Research, Inc. (www.elmhurstresearch.com) Engineered Pressure Systems Inc. (www.epsi-highpressure.com) Epsi Inc. (www.epsi-highpressure.com) Kobelco (www.kobelco.co.jp) Mitsubishi Heavy Industries (www.mhi.co.jp) NC Hyperbaric (www.nchyperbaric.com) Resato International (www.resato.com) Stansted Fluid Power Ltd (www.sfp-4-hp.demon.co.uk) Uhde Hockdrucktechnik (www.uhde-hpt.com) Fruit juice manufacturers Invo (Madrid, Spain) (www.invo.es) Beskyd Frycovice, a.s. (HornõÂ Cerkev, Czech Republic) (www.beskyd.cz/index.php) Juicy Line-Fruity Line (Ochten, Holland) (www.fruity-line.com) Pressure Fresh Australia Pty Ltd (Bundaberg Qld, Australia) (www.austchilli.com.au/devindexpf.aspx)



67

Preshafood Ltd. (Victoria, Australia) (www.preshafood.com.au) FrubacËa (Leiria, Portugal) (www.glsa.pt) Ulti Fruit (Vigneux-sur-Seine, France) ATA S.P.A. (Catanzaro, Italy) Puro-www.barefruitproducts.com/about/index.html Genesis Juice Corporation (Oregon, US) (www.genesisorganicjuice.com)

3.7

Acknowledgements

Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture. The authors want to express their gratitude to Bradley Wardrop-Brown from BOI Food Tech and Packaging, Andrew Gibb from Preshafood Ltd, Trent DePaoli from Pressure Fresh Australia Pty Ltd and Sheldon Rubin from Toby's Family Foods, LLC/Genesis Organic Juice for their interest and willingness in participating in this chapter. The authors are also grateful to Carole Tonello from NC Hyperbaric S.A. for providing useful information.

3.8

References

and BOZOGLU F (2000), `The combined effect of high hydrostatic pressure, heat and bacteriocins on inactivation of foodborne pathogens in milk and orange juice', World J Microbiol Biotech, 16, 387±392. ALPAS H and BOZOGLU F (2003), Èfficiency of high pressure treatment for destruction of Listeria monocytogenes in fruit juices', FEMS Immun Med Microbiol, 35, 269± 273. ALPAS H, ALMA L and BOZOGLU F (2003), Ìnactivation of Alicyclobacillus acidoterrestris vegetative cells in model system, apple, orange and tomato juices by high hydrostatic pressure', World J Microbiol Biotech, 19, 619±623. BALASUBRAMANIAM V M and FARKAS D (2008), `High-pressure food processing', Food Sci Tech Int, 14, 413±418. BANG W-S and SWANSON B G (2008), `Scanning electron microscopy studies of Saccharomyces cerevisiae structural changes by high hydrostatic pressure treatment', Food Sci Biotech, 17, 1102±1105. BASAK S and RAMASWAMY H S (1996), Ùltra high pressure treatment of orange juice: a kinetic study on inactivation of pectin metyl esterase', Food Res Int, 29, 601±607. BASAK S, RAMASWAMY H S and PIETTE J P G (2002) `High pressure destruction kinetics of Leuconostoc mesenteroides and Saccharomyces cerevisiae in single strength and concentrated orange juice', Innov Food Sci Emerg Tech, 3, 223±231. BAXTER I A, EASTON K, SCHNEEBELI K and WHITFIELD F B (2005), `High pressure processing of Australian navel orange juices: sensory analysis and volatile flavor profiling', Innov Food Sci Emerg Tech, 6, 372±387. BAYINDIRLI A, ALPAS H, BOZOGÏLU F and HIZAL M (2006), Èfficiency of high pressure ALPAS H


68


treatment on inactivation of pathogenic microorganisms and enzymes in apple, orange, apricot and sour cherry juices', Food Control, 17, 52±58. BULL M K, SZABO E A, COLE M B and STEWART C M (2005), `Toward validation of process criteria for high-pressure processing of orange juice with predictive models', J Food Prot, 68, 949±954. Â NDEZ-GARCIÂA A, DIETERICH S, LINDAUER R, BOGNAR A and BUTZ P, SERFERT Y, FERNA TAUSCHER B (2004), Ìnfluence of high-pressure treatment at 25 ëC and 80 ëC on folates in orange juice and model media', J Food Sci, 69, 117±121. BUZRUL S, ALPAS H, LARGETEAU A AND DEMAZEAU G (2008), Ìnactivation of Escherichia coli and Listeria monocytogenes in kiwifruit and pineapple juices by high hydrostatic pressure', Int J Food Microbiol, 124, 275±278. Â NDEZ A and DE ANCOS B (1997), `High pressure and temperature effects CANO M P, HERNA on enzyme inactivation in strawberry and orange products', J Food Sci, 62, 85±88. CORTEÂS C, ESTEVE M J and FRIÂGOLA A (2008), Èffect of refrigerated storage on ascorbic acid content of orange juice treated by pulsed electric fields and thermal pasteurization', Eur Food Res Tech, 227, 629±635. DA ROCHA FERREIRA E H, ROSENTHAL A, CALADO V, SARAIVA J and MENDO S (2009), `Byssochlamys nivea inactivation in pineapple juice and nectar using high pressure cycles', J Food Eng, 95, 664±669. DE ANCOS B, SGROPPO S, PLAZA L and CANO P (2002), `Possible nutritional and healthrelated value promotion in orange juice preserved by high-pressure treatment', J Sci Food Agric, 82, 790±796. DEL POZO-INSFRAN D, DEL FOLLO-MARTIÂNEZ A, TALCOTT S T and BRENES C H (2007), `Stability of copigmented anthocyanins and ascorbic acid in muscadine grape juice processed by high hydrostatic pressure', J Food Sci, 72, 247±253. DELIZA R, ROSENTHAL A and SILVA A L S (2003), `Consumer attitude towards information on non conventional technology', Trends Food Sci Technol, 14, 43±49. DELIZA R, ROSENTHAL A, ABADIO F B D, SILVA C H O and CASTILLO C (2005), Àpplication of high pressure technology in the fruit juice processing: benefits perceived by consumers', J Food Eng, 67, 241±246. Ï AN C and ERKMEN O (2004), `High pressure inactivation kinetics of Listeria DOG monocytogenes inactivation in broth, milk, and peach and orange juices', J Food Eng, 62, 47±52. DONSIÁ G, FERRARI G and MARESCA P (2007), `Pulsed high pressure treatment for the inactivation of Saccharomyces cerevisiae: the effect of process parameters', J Food Eng, 984±990. Â I and MARTIÂN-BELLOSO O (2006), Ìnactivation of orange juice peroxidase EÂLEZ P, AGUILO by high-intensity pulsed electric fields as influenced by process parameters', J Sci Food Agric, 86, 71±81. Ï AN C (2004), `Kinetic analysis of Escherichia coli inactivation by high ERKMEN O and DOG hydrostatic pressure in broth and foods', Food Microbiol, 21, 181±185. FANG L, JIANG B and ZHANG T (2008), Èffect of combined high pressure and thermal treatment on kiwifruit peroxidase', Food Chem, 109, 802±807. FOURIE P C (1996), `Fruit and human nutrition', in Arthey D and Ashurst P R, Fruit Processing, Glasgow, Blackie Academic & Professional, 20±39. GARCIÂA-GRAELLS C, HAUBEN K J A and MICHIELS C W (1998), `High-pressure inactivation and sublethal injury of pressure-resistant Escherichia coli mutants in fruit juices', Appl Environ Microbiol, 64, 1566±1568. GARCIÂA-PALAZOÂN A, SUTHANTHANGJAI W, KADJA P and ZABETAKIS I (2004), `The effects of



69

high hydrostatic pressure on -glucosidase, peroxidase and polyphenoloxidase in red raspberry (Rubus idaeus) and strawberry (Fragaria x ananassa)', Food Chem, 88, 7±10. GIESE J (2000), `Color measurement in foods as a quality parameter', Food Technol, 54, 62±65. GINER J, ORTEGA M, MESSEGUEÂ M, GIMENO V, BARBOSA-CAÂNOVAS G V and MARTIÂN O (2002), Ìnactivation of peach polyphenoloxidase by exposure to pulsed electric fields', J Food Sci, 67, 1467±1472. GOODNER J K, BRADDOCK R J, PARISH ME and SIMS C A (1999), `Cloud stabilization of orange juice by high pressure processing', J Food Sci, 64, 699±700. GREGORY III J F (1996), `Vitamins', in Fennema O R, Food chemistry, New York, Marcel Dekker, 559±568. HENDRICKX M, LUDIKHUYZE L, VAN DEN BROECK I and WEEMAES C (1998), Èffects of high pressure on enzymes related to food quality', Trends Food Sci Technol, 9, 197± 203. JENSEN (1999), Àlicyclobacillus ± a new challenge for the food industry', Food Aus, 51, 33±36. JORDAN S L, PASCUAL C, BRACEY E and MACKEY B M (2001), Ìnactivation and injury of pressure-resistant strains of Escherichia coli O157:H7 and Listeria monocytogenes in fruit juices', J Appl Microbiol, 91, 463±469. KATZ F (2000), `Research priorities move toward healthy and safe', Food Technol, 54, 42± 44. KOUNIAKI S, KAJDA P and ZABETAKIS I (2004), `The effect of high hydrostatic pressure on anthocyanins and ascorbic acid in blackcurrants (Ribes nigrum)', Flavour Fragance J, 19, 281±286. LABOISSIERE L H E S, DELIZA R, BARROS-MARCELLINI A M, ROSENTHAL A, CAMARGO L M A Q

and JUNQUEIRA R G (2007), Èffects of high hydrostatic pressure (HHP) on sensory characteristics of yellow passion fruit juice', Innov Food Sci Emerg Technol, 8, 469±477. LAVINAS F C, MIGUEL M A L, LOPES M L M and VALENTE MESQUITA V L (2008), Èffect of high hydrostatic pressure on cashew apple (Anacardium occidentale L.) juice preservation', J Food Sci, 73, 273±277. LEE S Y, DOUGHERTY R H and KANG D H (2002), Ìnhibitory effects of high pressure and heat on Alicyclobacillus acidoterrestris spores in apple juice', Appl Environ Microbiol, 68, 4158±4161. LEE S Y, CHUNG H J and KANG D H (2006), `Combined treatment of high pressure and heat on killing spores of Alicyclobacillus acidoterrestris in apple juice concentrate', J Food Prot, 69, 1056±1060. LINTON M, MCCLEMENTS J M J and PATTERSON M F (1999a), Ìnactivation of Escherichia coli O157:H7 in orange juice using a combination of high pressure and mild heat', J Food Prot, 62, 277±279. LINTON M, MCCLEMENTS J M J and PATTERSON M F (1999b), `Survival of Escherichia coli O157:H7 during storage in pressure-treated orange juice', J Food Prot, 62, 1038± 1040. LUDIKHUYZE L, VAN LOEY A, OEY I and HENDRICKX M (2002), `High pressure processing of fruit and vegetables', in Hendrickx M and Knorr D, Ultra High Pressure Treatments of Foods, New York, Kluwer Academic, 168±171. MILLER L G and KASPAR C W (1994), Èscherichia coli 0157:H7 acid tolerance and survival in apple cider', J Food Prot, 57, 460±464.


70


and ZHANG Q H (2003) Èffects of commercial-scale pulsed electric field processing on flavor and color of tomato juice', Food Chem Tox, 1600±1606. Â NCHEZ-MORENO C and CANO M P (2007), Èffects of high pressure Ä OZ M, DE ANCOS B, SA MUN and mild heat on endogenous microflora and on the inactivation and sublethal injury of Escherichia coli inoculated into fruit juices and vegetable soup', J Food Prot, 70, 1587±1593. NAKIMBUGWE D, MASSCHALCK B, ANIM G and MICHIELS C W (2006), Ìnactivation of Gramnegative bacteria in milk and banana juice by hen egg white and lambda lysozyme under high hydrostatic pressure', Int J Food Microbiol, 112, 19±25. MIN S

NIELSEN H B, SONNE A-M, GRUNERT K G, BANATI D, POLLAK-TOTH A, LAKNER Z, OLSEN N V,

and PETERMAN M (2009), `Consumer perception of the use of highpressure processing and pulsed electric field technologies in food production', Appetite, 52, 115±126. NIENABER U and SHELLHAMMER T H (2001), `High-pressure processing of orange juice: kinetics of pectinmethylesterase inactivation', J Food Sci, 66, 328±331. NOMA S, TOMITA C, SHIMODA M and HAYAKAWA I (2004), `Response of Escherichia coli O157:H7 in apple and orange juices by hydrostatic pressure treatment with rapid decompression', Food Microbiol, 21, 469±473. NORTON T and SUN D-W (2008), `Recent advances in the use of high pressure as an effective processing technique in the food industry', Food Bioprocess Technol, 1, 2±34. OEY I, VAN DER PLANCKEN I, VAN LOEY A and HENDRICKX M (2008a), `Does high pressure processing influence nutritional aspects of plant based food systems?', Trends Food Sci Tech, 19, 300±308. OEY I, LILLE M, VAN LOEY A and HENDRICKX M (2008b), Èffect of high-pressure processing on colour, texture and flavour of fruit- and vegetable-based food products: a review', Trends Food Sci Tech, 19, 320±328. Â PEZ-MALO A, BARBOSA-CAÂNOVAS G V, WELTI-CHANES J, DAVIDSON P M and PALOU E, LO SWANSON B G (1998), Èffect of oscillatory high hydrostatic pressure treatments on Byssochlamys nivea ascospores suspended in fruit juice concentrates', Letters Appl Microbiol, 27, 375±378. Â PEZ-MALO A, BARBOSA-CAÂNOVAS G V, WELTI-CHANES J and SWANSON B G (1999), PALOU E, LO `Polyphenoloxidase activity and color of blanched and high hydrostatic pressure treated banana puree', J Food Sci, 64, 42±45. PARISH M E (1998), Òrange juice quality after treatment by thermal pasteurization or isostatic high pressure', Lebens Wiss Technol, 31, 439±442. PETTIPHER G L, OSMUNDSON M E and MURPHY J M (1997), `Methods for the detection and enumeration of Alicyclobacillus acidoterrestris and investigation of growth and production of taint in fruit juice and fruit juice-containing drinks', Letters Appl Microbiol, 24, 185±189. Â NCHEZ-MORENO C, EÂLEZ-MARTIÂNEZ P, DE ANCOS B, MARTIÂN-BELLOSO O and CANO PLAZA L, SA M P (2006), Èffect of refrigerated storage on vitamin C and antioxidant activity of orange juice processed by high-pressure or pulsed electric fields with regard to low pasteurization', Eur Food Res Technol, 223, 487±493. POLYDERA A C, STOFOROS N G and TAOUKIS P S (2003), `Comparative shelf-life study and vitamin loss kinetics in pasteurized and high pressure processed reconstituted orange juice', J Food Eng, 60, 21±29. POLYDERA A C, GALANOU E, STOFOROS N G and TAOUKIS P S (2004), Ìnactivation kinetics of pectin methylesterase of greek navel orange juice as a function of high hydrostatic ZONTAR T P



71

pressure and temperature process conditions', J Food Eng, 62, 291±298. and IDZIAK E (2003), `High-pressure destruction kinetics of E. coli (29055) in apple juice', J Food Sci, 68, 1750±1756. Â N M L, GO Â NGORA M M, BARBOSA-CAÂNOVAS G V and SWANSON B G (1998), RASO J, CALDERO Ìnactivation of Zygosaccharomyces cerevisiae in fruit juices by heat, high hydrostatic pressure and pulsed electric fields', J Food Sci, 63, 1042±1044. RASTOGI N K, RAGHAVARAO K S M S, BALASUBRAMANIAM V M, NIRANJAN K and KNORR D (2007), Òpportunities and challenges in high pressure processing of foods', Cri Rev Food Sci Nutr, 47, 69±112. REYNS K M F A, SOONTJENS C C F, CORNELIS K, WEEMAES C A, HENDRICKX M E and MICHIELS C W (2000), `Kinetic analysis and modelling of combined high-pressure-temperature inactivation of the yeast Zygosaccharomyces bailii', Int J Food Microbiol, 56, 199±210. Â NOVAS G V and RODRIGO M (2006), Èffect of RIVAS A, RODRIGO D, MARTIÂNEZ A, BARBOSA-CA PEF and heat pasteurization on the physical-chemical characteristics of blended orange and carrot juice', Lebens Wiss Technol, 39, 1163±1170. RODRIGO D, VAN LOEY A and HENDRICKX M (2007), `Combined thermal and high pressure colour degradation of tomato puree and strawberry juice', J Food Eng, 79, 553± 560. RUXTON C H S, GARDNER E J and WALKER D (2006), `Can pure fruit and vegetable juices protect against cancer and cardiovascular disease too? A review of the evidence', Int J Food Sci Nutr, 57, 249±272. SADO P N, JINNEMAN K C, BUSBY G J, SORG S M and OMIECINSKI C J (1998), Ìdentification of Listeria monocytogenes from unpasteurized apple juice using rapid test kits', J Food Prot, 58, 1199±1202. SAMPEDRO F, RODRIGO D and HENDRICKX M (2008), Ìnactivation kinetics of pectin methyl esterase under combined thermal-high pressure treatment in an orange juice-milk beverage', J Food Eng, 86, 133±139. SAMPEDRO F, GEVEKE D J, FAN X and ZHANG Q H (2009a), Èffect of PEF, HHP and thermal treatment on PME inactivation and volatile compounds concentration of an orange juice-milk based beverage', Innov Food Sci Emerg Technol, 10, 463±469. SAMPEDRO F, GEVEKE D J, FAN X, RODRIGO D and ZHANG Q H (2009b), `Shelf-life study of an orange juice-milk based beverage after PEF and thermal processing', J Food Sci, 74, 107±112. SAN MARTIÂN M F, BARBOSA-CAÂNOVAS G V and SWANSON B G, (2002), `Food processing by high hydrostatic pressure', Crit Rev Food Sci Nutr, 42, 627±645. Â NCHEZ-MORENO C, PLAZA L, DE ANCOS B and CANO P (2003a), Èffect of high-pressure SA processing on health-promoting attributes of freshly squeezed orange juice (Citrus sinensis L.) during chilled storage', Eur Food Res Technol, 216, 18±22. Â NCHEZ-MORENO C, PLAZA L, DE ANCOS B and CANO P (2003b), `Vitamin C, provitamin A SA carotenoids, and other carotenoids in high-pressurized orange juice during refrigerated storage', J Agric Food Chem, 51, 647±653. Â NCHEZ-MORENO C, PLAZA L, EÂLEZ-MARTIÂNEZ P, DE ANCOS B, MARTIÂN-BELLOSO O and CANO SA P (2005), Ìmpact of high pressure and pulsed electric fields on bioactive compounds and antioxidant activity of orange juice in comparison with traditional thermal processing', J Agric Food Chem, 53, 4403±4409. Â NCHEZ-MORENO C, PLAZA L, DE ANCOS B and CANO P (2006), Ìmpact of high-pressure SA and traditional thermal processing of tomato puree on carotenoids, vitamin C and antioxidant activity', J Sci Food Agric, 86, 171±179. RAMASWAMY H S, RIAHI E


72


and NARBONNE J-F (1999), Èffect of ultra-high hydrostatic pressure on hydrosoluble vitamins', J Food Eng, 39, 247±253. SAUBERLICH H E, KRETSCH M J, SKALA J H, JOHNSON H L and TAYLOR P C (1987), `Folate requirements and metabolism in nonpregnant women', Amer J Clin Nutr, 46, 1016±1028. SCHLESSER J and PARISI B (2009), Ìnactivation of Yersinia pseudotuberculosis 197 and Francisella tularensis LVS in beverages by high pressure processing', J Food Prot, 72, 165±168. SHARMA R (2005), `Market trends and opportunities for functional dairy beverages', Aus J Dairy Technol, 60, 196±199. TALCOTT S T, BRENES C H, PIRES D M and DEL POZO-INSFRAN D (2003), `Phytochemical stability and color retention of copigmented and processed muscadine grape juice', J Agric Food Chem, 51, 957±963. TEO A Y-L, RAVISHANKAR S and SIZER C E (2001), Èffect of low-temperature, high-pressure treatment on the survival of Escherichia coli O157:H7 and Salmonella in unpasteurized fruit juices', J Food Prot, 64, 1122±1127. TIJSKENS L M M, SCHIJVENS E P H M and BIEKMAN E S A (2001), `Modeling the change in colour broccoli and green beans during blanching', Innov Food Sci Emerg Technol, 2, 303±313. TORREGROSA F, CORTEÂS C, ESTEVE M J and FRIÂGOLA A (2005), Èffect of high-intensity pulsed electric fields processing and conventional heat treatment on orange-carrot juice carotenoids', J Agric Food Chem, 53, 9519±9525. TSOU C L (1986), `Location of active sites of some enzymes in limited and flexible molecular regions', Trends Biochem Sci, 67, 3058±3062. Â SÏ J, TICHAÂ L, CÏERÏOVSKY Â TKA Â J (2004), `Resistance of vegetative Â and KRA VOLDRÏICH M, DOBIA cells and ascospores of heat resistant mould Talaromyces avellanus to the high pressure treatment in apple juice', J Food Eng, 61, 541±543. WHITNEY B M, WILLIAMS R C, EIFERT J and MARCY J (2007), `High-pressure resistance variation of Escherichia coli O157:H7 strains and Salmonella serovars in triptic soy broth, distilled water, and fruit juice', J Food Prot, 70, 2078±2083. WHO/FAO EXPERT CONSULTATION (2003), `Diet, nutrition and the prevention of chronic diseases', WHO Tech Report Ser, 916. WRIGHT A O, CARDELLO A V and BELL R (2007), `Consumer evaluations of high pressure processed Foods', in Doona C J and Feeherry F E, High pressure processing of foods, Institute of Food Technologists Series, Weinheim, Wiley-VCH. YEOM H W, STREAKER C B, ZHANG Q H and MIN B (2000), Èffects of pulsed electric fields on the quality of orange juice and comparison with heat pasteurization', J Agric Food Chem, 48, 4597±4605. ZABETAKIS I, LECLERC D and KAJDA P (2000), `The effect of high hydrostatic pressure on strawberry anthocyanins', J Agric Food Chem, 48, 2749±2754. ZOOK C D, PARISH M E, BRADDOCK R J and BALABAN M O (1999), `High pressure inactivation kinetics of Saccharomyces cerevisiae ascospores in orange and apple juices', J Food Sci, 64, 533±535. SANCHO F, LAMBERT Y, DEMAZEAU G, LARGETEAU A, BOUVIER J-M


4 Pulsed electric field (PEF) systems for commercial food and juice processing M. A. Kempkes, Diversified Technologies, Inc., USA

Abstract: This chapter describes the key PEF treatment parameters, and the specialized equipment required to implement this process in liquid food processing. It also describes the interactions between the process parameters and the electrical design and performance of PEF systems, as a guide for potential adopters of this technology. Finally, this chapter presents initial commercial applications of PEF processing, and guidelines for its future adoption. Key words: non-thermal, pasteurization, PEF, high voltage, juices, bacteria, disinfection

4.1

Introduction

Pulsed electric field (PEF) processing is a low temperature, non-thermal, nonchemical, low impact process that achieves very high kill rates, making PEF processing a practical alternative to thermal pasteurization. PEF processing involves the application of short duration (1±20 s), very high voltage pulses that create a high voltage field (approximately 20±50 kV/cm) across a liquid food to kill resident bacteria, molds, and other microorganisms via electroporation of the cell membrane (Fig. 4.1). The pulses are so frequent that all of the liquid in a pipe can be treated as it flows through the treatment chamber. By using multiple treatment chambers to apply pulses to a stream of fluid in a continuous flow process, kill ratios of 5±9 log reductions, similar to those resulting from pasteurization, have been achieved. Unlike pasteurization, however, the food is not heated during PEF processing, so its taste and


74


Fig. 4.1 Cell electroporation resulting from PEF treatment.

nutritional value remain essentially indistinguishable from fresh, untreated product ± while maintaining the level of food safety associated with pasteurization. PEF processed products simply taste fresher than pasteurized products, yet have equivalent safety and shelf life. Multiple experiments have demonstrated that the shelf life of PEF processed food is comparable to that yielded by pasteurization, with no or very minimal impact on the taste, color, or nutritional value of the food. PEF processing is particularly beneficial to fresh juices, beer, and other foods that are susceptible to changes in their flavor caused by the heat of pasteurization. The commercial debut of PEF processing occurred in 2005, with Genesis Juice's introduction of PEF processed juices to the consumer market. This represents the first known commercial introduction of PEF processed foods. Since that time, there has been considerable interest in the adoption of PEF processing, and research into process scale-up. In applications other than foods, PEF processing can also improve the performance of industrial processes such as the removal of water from sludge, or the extraction of sugars and starches from plants, because the ruptured cells release their intracellular liquids more easily into their surroundings. This chapter will describe the key PEF parameters, and the specialized equipment required to implement this process in liquid food processing. It will describe the interactions between the process parameters and the electrical design and performance of PEF systems, as a guide for potential adopters of this technology. 4.1.1 PEF utility The origins of PEF processing are deep and varied. Researchers demonstrated that voltage fields could disrupt biological cells as early as 1960, with microbial inactivation initially demonstrated in 1967. Only since the mid-1990s, however, has there been the necessary confluence of applied research, developments in high voltage equipment, and commercial interest in non-thermal processes to


PEF systems for commercial food and juice processing

75

move PEF technology from the laboratory and into commercial operation. Three key steps in this development were the Dunn and Pearlman patent on the PEF process for disinfection (4,695,472) in 1987, the development of the co-field flow treatment chambers at The Ohio State University (OSU) in 1997, and the initial use of solid state, high voltage pulse modulators for PEF in 2000 (built by DTI for OSU). This combination of advances laid the groundwork for the substantial movement of PEF processing in the last decade from laboratory to commercialization. PEF development has occurred in three primary, loosely related areas over this period: medical treatment, tissue disintegration, and disinfection. Research in these areas has been conducted separately, but with considerable crossfertilization of results and processes. Among these areas, the use of high voltage fields for electroporation is probably most mature in the medical world. A Google search yields over one million hits for the term èlectroporation', with the vast majority related to medical and microbiological uses of this technique at the cellular level, and dozens of companies selling equipment specifically designed for electroporation. The well established method of medical electroporation is to apply pulsed voltage fields across living cells and open pores in cellular membranes for a short period as a means to extract or insert specific chemicals into the cell (for chemotherapy or genetic manipulation) without permanently damaging the cells themselves. The equipment required is typically very small in terms of voltage, power, and capacity, since medical electroporation is typically applicable to small samples (from a micro-liter to milliliters). This line of activity has developed relatively independently of the other two major applications ± disinfection and tissue disintegration. PEF has also been investigated (and is currently being commercialized) as a technique for permeabilization of plant and animal tissues for a variety of purposes ± typically focused on extraction, drying, or pre-processing of the tissue for subsequent chemical or microbial processing. In all cases, the electroporation-induced rupture of the plant or animal tissue cells opens these cells to allow the exchange of materials between the internal and external environment, such as the simplified release of water (enhanced drying) or internal materials (extraction of sugar from sugar beet), or the entrance of chemicals or microbes into the internal structure of the cell (accelerated fermentation). Several commercial installations using PEF have been reported, including the use of PEF for wastewater treatment (prior to anaerobic digestion of wastewater sludge), extraction, and drying. In general, the focus of these activities is process acceleration, reduction of energy costs, or both. Disinfection is the primary application of PEF related to food products ± using PEF to induce electroporation and kill microbes in liquid or pumpable foods. Generally, the desire has been to achieve disinfection levels similar to thermal pasteurization, but without the damage to taste, nutrition, and other characteristics of unprocessed liquids which can be significantly degraded by heating the food (i.e., thermal processing). In contrast to medical electroporation, disinfection requires that the voltage field be applied at a high enough


76


intensity, and for a sufficient duration, such that the electroporation process does not just open the pores in the cell membranes, it ruptures the cell membrane irreversibly. In addition, this field must be applied to larger fluid volumes, and kill all of the microbes in those volumes, to achieve required disinfection levels. Substantial research over the last 20 years has clearly demonstrated that PEF can achieve disinfection of liquids across a wide range of microbes and products, at relatively low temperatures compared to pasteurization (i.e., 20± 50 ëC). Literally thousands of studies, and a number of books, have been published showing the impact of PEF on microbes and foods in various combinations, under a wide range of treatment protocols. In general, these studies have focused on two related areas: showing microbial kill, and assessing the impact of PEF on food quality (taste, nutrition, etc.) across a wide range of pumpable products ranging from juices to semi-solids, such as sausages. A sampling of these reports show PEF research against a range of microorganisms, including pathogens, spoilage organisms, and surrogates for pathogens, including: · · · · · · · · · · · · · · ·

Escherichia coli Pseudomonas flourescens Bacillus subtillus Saccharomyces cerevisae Listeria monocytogenes Lactobacillus plantarum Yersinia enterocolitica Salmonella typhimurium Bacillus megaterium Candida utilis Clostridium welchii Listeria innocua Salmonella senftenberg Sarcin lutea Yeasts and molds.

A primary result of these studies is that it is possible to achieve high microbial kill levels (as high as 9 logs, or only one survivor out of one billion initial microbes), with proper treatment conditions and careful processing. Given that the standard for pasteurization is typically a 5-log reduction in microbial survival, PEF is clearly capable of replacing pasteurization as a disinfection step. As this research moves towards commercial applications, a key area of interest has been balancing the competing factors of high levels of disinfection, with minimal impact on the taste of the food. One key finding from these studies is that, to the best of our knowledge, PEF is only effective in killing vegetative microbes, yeasts, and molds ± PEF appears ineffective against spores or viruses. This is not surprising, given that PEF is believed to work through the mechanism of electroporation, and spores and viruses do not have active membranes that the electric field can impact. For this



77

reason, the preponderance of PEF studies focused on food disinfection have targeted acidic products, such as citrus juices and tomato sauces, where spore regeneration is not an issue. Researchers are still investigating the use of PEF, either by itself or in combination with other treatments, to kill resistant bacterial spores, but without notable success to date. This work, if ultimately successful, would significantly expand the applicability of PEF disinfection to a larger range of liquid foods. There has also been considerable research on the applicability of PEF processing to a variety of liquid foodstuffs. The majority of this research has focused on fruit and vegetable juices, but successful PEF processing has been applied to products from beer to salad dressings and salsa, and even to sausage. As a result of this range of PEF application, there are several key product parameters that can be identified to select a potential candidate for PEF processing: 1. The product typically must be pumpable, but can be highly viscous. Treatment is applied as the product is flowing (i.e., pre-packaging). 2. It must be possible to remove bubbles from the liquid, either through deaeration or pressure, to prevent arcing. 3. There can be particulates in the fluid (fruit and vegetable chunks, for example), so long as the treatment chambers are large enough to prevent clogging. 4. Acidified products are typically necessary if spore-forming microorganisms represent a pathogen or spoilage organism of concern. In summary, PEF processing has made significant advances based on approximately the past 20 years of focused R&D using a wide range and variety of pathogens and spoilage organisms, potential products for PEF treatment, and equipment required to apply the PEF treatment. PEF is FDA approved for foods in the US, and the remaining barriers to PEF commercialization are primarily economic, relating to developing new products with increased value in terms of nutrition, taste, and quality) achieved through PEF processing, compared to the initial high costs of implementing PEF.

4.2

Key process parameters

Any discussion of PEF system design must be based upon an effective PEF treatment protocol. In designing a PEF system, this protocol (or at least its boundaries) must be known. The PEF treatment protocol must be known for a particular liquid food, targeting its associated potential pathogens and spoilage organisms (or other organisms of interest). The protocol is developed from experiments involving multiple trials at different electrical field strengths and durations, and followed by microbiological assessment of the inoculated food after PEF treatment. For example, a typical treatment protocol might require the application of a 35 kV/cm field for a minimum of 50 s to yield a bacterial


78


reduction (typically, a 5-log reduction) of a target organism (such as E. coli) in a liquid food substrate. Numerous protocols have been developed and published for a wide range of liquid foods, organisms, and PEF systems. For the purposes of designing a PEF system, we assume that the boundaries of the protocol are known. From that point, the food's characteristics, and the desired processing capacity are the two major remaining considerations to be accounted for in the system design. These are typically expressed as the liquid conductivity and flow rate. These two factors, when combined with the desired protocol, form the basis for designing the PEF system. 4.2.1 Pulse shape One critical issue in assessing PEF system performance is the pulse shape of a given PEF system. PEF is generally believed to work on a voltage threshold, and only exposure time above that critical field strength is believed to have the desired effect on killing target microorganisms. In an ideal PEF pulse, there would be zero rise and fall time, and the flattop of the pulse would be concentrated at a constant voltage. All of the pulse time and energy, therefore, would be at the desired voltage. In practice, however, this pulse shape is not attainable. All pulses have finite rise and fall times, where the full voltage is not present, and it is not possible to achieve a perfect stationary flattop. Characterizing the actual pulse voltage and time is therefore subject to arbitrary estimates (or, more commonly, ignored all together). To illustrate this, Fig. 4.2 shows three normalized voltage waveforms for typical real pulses: a rectangular, exponential, and half-sine wave. These are simplified versions of three of the most common modulator pulse shapes. They

Fig. 4.2 Three normalized PEF voltage pulses ± square wave, half-sine wave, and decaying exponential. Although each pulse has the same nominal peak voltage and pulsewidth, the total energy and time above any voltage threshold voltage vary substantially.



79

Table 4.1 Comparison of energy and time above 80% Vmax for an ideal pulse and the three voltage nominal pulses in Fig. 4.2

Total energy Energy above 80% Vmax Time above 80% Vmax

Ideal

Square

Exponential

Half-sine

100% 100% 100%

87% 84% 84%

37% 14% 16%

50% 36% 40%

all have the same peak voltage (Vmax 100), and total pulsewidth (with the exponential cut off at approximately 25% Vmax). In this simplified example, researchers using each of these three waveforms could report results assuming these pulse characteristics (V 100, t 100) were the same. However, their total energy and time above any arbitrary threshold voltage varies dramatically. The total energy delivered, and the treatment time above a threshold voltage, could vary by more than a factor of six, depending on the pulse shape each researcher used. Given this disparity, it is not surprising that there are wildly different biological results reported by different researchers, for apparently similar to identical conditions. More specifically, compared to the ideal pulse, each of these pulse shapes has the characteristics shown in Table 4.1 (based on a threshold of 80% of Vmax). If it were possible to run for longer pulsewidths, these differences would begin to converge, but this is typically not the case with PEF systems. For shorter pulsewidths, the disparity between the ideal and actual pulse shapes becomes even greater, since more of the total pulsewidth consists of the rise and fall times, which, in essence, translates into wasted power that does not contribute to the PEF treatment. It is critical, therefore, to approximate the ideal waveform as closely as possible. Defining treatment protocols, therefore, requires knowledge of both the pulse shape and measurement thresholds. Figure 4.3 shows PEF voltage and current pulses from a Diversified Technologies, Inc. pilot-scale PEF system processing orange juice. The DC voltage setting is 24 kV, and this peak voltage is maintained for just under 2 s in this example. The commanded pulsewidth (at the input pulse command) is approximately 2.5 s, and the 0.5 s difference is attributable to the pulse risetime. The total pulsewidth, taking into account both rise- and fall-times, could be interpreted to be as long as 3 s. If 20 pulses are applied to each element of the juice, the reported treatment time could vary between 35 and 60 s. Similarly (but less significantly for this case), the peak voltage could be reported as several kV above or below the nominal 24 kV, depending on where the measurement is made during the pulse. How these parameters are reported significantly impacts the results when attempting to replicate protocols on different PEF systems, with different pulse shapes. There is a strong need within the PEF research community to standardize the reporting of PEF results, allowing data from different organizations, using different PEF systems and waveforms, to be compared and assessed on a common basis. At a minimum, the voltage pulse shape should be


80


Fig. 4.3 PEF voltage and current waveforms from DTI's pilot PEF system into orange juice. Pulse parameters are 2 s pulsewidth, 24 kV, 45 A, but could be reported as up to 3 s in pulsewidth, and 20±26 kV in voltage, depending on how the pulse is characterized.

shown, with measurement points for peak voltage and pulsewidth. This would allow other PEF researchers and process developers to make their own adjustments between systems. 4.2.2 Conductivity/flow rate The power required to apply a given protocol is determined by the conductivity of the fluid, and the desired flow rate. Conductivity is a measure of the electron mobility within a volume of material ± a measure of how easily it passes electrical current. Conductivity is expressed as Siemens/meter (S/m), or, to get to whole numbers for typical fluids, mS/cm. It is the reciprocal of resistivity, which is the electrical resistance of a volume of material. In most PEF research, conductivity () is used as a measure because it is readily measured, applicable to the liquid itself, and allows calculation of the current that will pass through a volume of fluid through application of Ohm's law. Sample conductivities (ms/ cm) are shown in Table 4.2. The electric field (in V/m or more commonly, kV/cm) is set by the treatment protocol, so the energy required to deliver this field to a volume of liquid is a direct function of the fluid conductivity. Since power is energy over time, the energy required per liter, and the number of liters to be processed per unit time, give the total average power required in the PEF system. A 100 kW average power system can process five times the volume of product per hour as a 20 kW system, for the same protocol. The power required increases linearly with flow rate and conductivity for a given protocol, and by the square of the required field


PEF systems for commercial food and juice processing Table 4.2

81

Conductivities of common liquids

Liquid

Conductivity (ms/cm)

Apple juice Water (tap, distilled) Liquid eggs Orange juice Beer Milk Yogurt

1.75 0.0011 5.88 3.3±5.5 1.5 3.2±5.8 6.0

* Published and DTI measurements

strength ± making field strength the most critical parameter to establish correctly, in order to minimize electrical power requirements in the PEF process. Adapting to high flow rate is the key to commercial scale-up of PEF systems. A laboratory PEF system typically processes liters per hour (or less), and a pilot plant typically operates at tens to hundreds of liters per hour. Commercial systems, however, must be capable of processing thousands to tens of thousands of liters per hour ± representing two to three orders of magnitude higher throughput than a pilot system. Fortunately, research at Ohio State University (OSU) and other organizations has demonstrated that the PEF process itself operates independently from flow rate ± so long as the field strength and dose (total treatment time) are maintained. This allows treatment protocols developed in laboratory and pilot scale systems to readily transition to commercial scale PEF operations. Flow rate determines several other major PEF system characteristics. For the most common treatment chamber design, the co-field flow chamber (discussed below), the diameter of the treatment chamber must be sized to pass the desired flow at reasonable pressure drops. The presence of particulates and `chunks' in the flow can also impact the sizing of the chamber. At the same time, to achieve uniform field strength within the treatment chamber, the gap across which the voltage is applied must increase with pipe diameter (maintaining the gap at more than 1.2 the diameter provides reasonably uniform fields). Larger gaps require lower fluid pressure in the treatment chambers, but require higher absolute voltages to maintain a given field strength. For example, doubling the treatment chamber diameter allows four times the flow at a given pressure, but requires twice the length and peak voltage (and also twice the average current) to maintain the same field strength and treatment time. The final parameter to be assessed in the design of a PEF system is the required pulse frequency. The total energy delivered to a liter of fluid is known from the treatment protocol. As more liquid is processed, the average power goes up linearly. In conventional pulsed power systems, increasing the average power is typically achieved by increasing the pulsewidth, allowing the pulse frequency to remain at reasonable levels. For PEF systems, however, it is generally not possible to run at pulsewidths longer than approximately 10 s


82


before arcing occurs. Without the ability to increase the pulsewidth, the only remaining options are to increase the energy in each pulse (which directly increases the cost of the PEF system), or to operate with more pulses per second (increased pulse frequency). This need for higher pulse frequencies becomes critical as PEF systems are scaled to commercial flow rates, as discussed in the next section.

4.3

Pulsed electric field (PEF) system overview

There are three unique elements to a PEF system, in comparison to a traditional pasteurizer. First, a DC power supply (Fig. 4.4) transitions the AC power available from the utility into high voltage, DC power. The second major element of the PEF system is the pulse modulator, which transforms that DC power into short, high peak power pulses. Finally, there is the treatment

Fig. 4.4

150 kW DC power supply for the PEF system shown in Fig. 4.15.



83

chamber, where the high voltage pulses are applied to the liquid itself. The next sections describe each of these subsystems, with general assessments of the alternatives available. 4.3.1 Power supplies A DC power supply converts the AC power available from the utility into high voltage, DC power. DC power supplies are typically rated in terms of their average power (in Watts). There are three basic DC power supply architectures. The simplest is a transformer-rectifier, which operates at line frequency (50± 60 Hz). These supplies are the least expensive, but can be very large at high power, and have significant drawbacks in terms of voltage regulation, voltage control, and their impact on the overall plant power system. They are, however, very inexpensive on a `$ per Watt' basis, especially at high power levels (hundreds of kW). The second basic power supply architecture is a switching power supply. In this design, the input power is rectified and `chopped' at high frequency (10± 50 kHz) into a transformer rectifier. Since the size of the transformer decreases as the chopping frequency increases, it is possible to build very compact, high power DC supplies in this way. Switching supplies provide highly regulated and rapidly adjustable output voltage, which supports tight control of the PEF process parameters, independent of the modulator architecture. Switching power supplies are typically used in applications requiring up to ~500 kW, which supports PEF processing up to flow rates of approximately 10 000 l/hr ± sufficient for most anticipated commercial lines. The drawback of switching power supplies is their cost, which can be 2±5 times higher than simple transformer rectifiers of similar power. At very high power levels (500 kW and above), the optimal solution is often the combination of a transformer-rectifier (for unregulated power) with a high frequency voltage regulator, such as a buck regulator, on the output to provide the final voltage control. This class of power supplies appears to be most applicable to very high throughput PEF processing, such as for large-scale commercial food processing and wastewater treatment. 4.3.2 Modulators The design and construction of PEF pulse modulators builds on over 50 years of R&D into pulse modulators for other applications, including radar and particle accelerators. Ideally, the modulator for a PEF system will provide pulses of very consistent voltage, with fast rise and fall times, at any desired pulsewidths and pulse frequencies needed for the desired PEF process parameters. The desirability of a modulator design is typically based on how well it meets these criteria. All pulse modulators are based on the ability to switch electricity very rapidly. This can be done mechanically, although this is slow and difficult to


84


Fig. 4.5

PFN modulator, with closing switch.

operate at any frequency, and not compatible with PEF processing. It can be done by creating a controlled short circuit, in a device such as a spark gap, thyratron, or using a solid-state device such as a Silicon Controlled Rectifier (SCR). This is referred to as a closing switch (Fig. 4.5). Closing switches must be able to remain open, holding off the full input voltage, until they are commanded to close. When they are closed, current will continue to flow through the switch until the input power is dissipated. Typically, closing switches must wait until there is zero current (or a reverse voltage) in order to open again, and prepare for the next pulse. This switching can also can be accomplished by allowing current to alternately flow and be interrupted, using a vacuum tube (such as a tetrode) or a power transistor, such as an Insulated Gate Bipolar Transistor (IGBT). In these opening and closing switches, current can be turned both on and off at any time, but the switch must be able to withstand the stresses of opening under full current, as well as holding the full voltage when open. Along with the two classes of switch (closing, and opening and closing) there are three fundamental modulator designs available ± pulse forming networks (PFNs, Fig. 4.5), `hard switches' (Fig. 4.6), and transformer coupled systems (Fig. 4.7). All of these designs have their origins in the days of vacuum tubebased modulators, but they have transitioned to solid-state switches in place of vacuum tubes over the past decade. The pulse forming network (PFN) shown in Fig. 4.5 holds a predetermined amount of energy in its capacitors, and creates a shaped pulse through the combination of capacitors and inductors in the network. This allows modification of the pulse parameters, such as risetime and pulsewidth, but only by manually changing the values of the capacitors and inductors themselves. For best performance, PFNs and transformer coupled systems must be impedance matched to the load ± meaning that the voltage and current must have a single



Fig. 4.6

85

Hard switch modulator, with opening and closing switch.

optimal relationship. Finally, after a pulse, the PFN must be completely recharged from a power supply, which can limit the frequency of pulses that can be generated with this design. The alternative is to use a `hard switch', capable of switching the full voltage on and off directly (Fig. 4.6). This has a number of benefits to PEF processing ± it allows flexibility in both pulsewidth and pulse frequency, and, since these hard switches are typically low impedance, allows the modulator to provide a consistent, repeatable output voltage over the range of peak currents required as the food conductivity varies. Solid-state switches are ideally suited to both of these requirements. The trade-off is that the power supply must also operate at this full voltage. DTI has pioneered the use of solid-state hard switches for PEF and other applications (Fig. 4.8), with other designs emerging in recent years. The alternative to full voltage switching is a pulse transformer coupled design. In this design, the switching required to create a pulse occurs at relatively low voltage (typically less than 20% of the desired output voltage), and the resulting low voltage pulse is passed through a pulse transformer, which increases the pulse voltage by a factor related to the ratio of the primary to

Fig. 4.7 Transformer coupled modulator with opening and closing switch. Transformer coupled modulators are also possible with PFNs on the primary, rather than an opening and closing switch.


86


Fig. 4.8 60 kV Bi-polar solid state PEF system built by Diversified Technologies, Inc. (DTI) for OSU in 1999 under the Dual Use Science & Technology (DUST) PEF Consortium. This was the first commercial scale PEF system in the world. The power supplies and pulse modulators are in the large tank, with the treatment chambers in the smaller `sidecar' tank. Controls are located above the unit.

secondary turns in the transformer itself. This is a critical simplification when it is difficult or impossible to switch the high voltage directly ± to create a 20 kV pulse, it is possible to use only a 2 kV switch, and a 10:1 turns ratio pulse transformer. The trade-off, however, is that the same energy must be present at both the primary and the secondary of the pulse transformer. This means that to create a 20 kV, 100 A pulse (2 MW peak power), the 2 kV switch must handle 1000 A to provide the same 2 MW of peak power. Transformer-coupled modulators have traditionally been built using a closing switch and a pulse forming network on the primary of the pulse transformer (the combination of Figs 4.5 and 4.7). Alternatively, there is the `hybrid' modulator, which combines a solid state hard switch at lower voltage with a pulse transformer (Fig. 4.7). In another design, multiple switches operate in parallel on multiple primaries, with each corresponding transformer secondary connected in series, to achieve high voltage. These designs eliminate several drawbacks of the PFN (e.g., fixed pulsewidth, limited frequency). Independent of the primary pulse generation approach, the pulse transformer itself has a couple of drawbacks for PEF processing. First, the transformer core must be `reset' between pulses to avoid core saturation (in a mono-polar



87

system). This can require a second, reset pulser which must operate between high voltage pulses, limiting the pulse frequency available. Second, the pulse transformer typically wastes approximately 10% of the total power in the PEF system ± turning it into unusable heat. The most critical factor, however, is that in PEF systems the load impedance can vary considerably. The liquid being processed is the load, and is therefore an integral part of the circuit. Its conductivity can vary by over an order of magnitude across different foods, and even a single food type, such as orange juice, can vary considerably in conductivity due to changes in the raw materials. The simulated pulse shapes shown in Fig. 4.9 illustrate this effect. The PEF protocol and overall system design is shown in Table 4.3. In this case, we are applying 35 kV/cm to a treatment chamber with a gap (cell length) of 0.6 cm, which requires an applied voltage of 21 kV. There are a total of two chambers (four treatment zones). The risetime of the pulses are modeled at (from left to right) conductivities of 1, 3, and 5 mS/cm. In the upper set of pulses, we use a hybrid modulator based on pulse transformer ratio of 6:1, meaning that a 3.5 kV pulse is required on the primary of the pulse transformer for 21 kV pulse on the secondary. As the pulse-shapes at differing conductivities clearly show, changing the conductivity, significantly changes the pulse waveform. At 1 mS/cm, the pulse is reasonable, but by 5 mS/cm, it may never be possible to get to full voltage within a single pulse before the onset of arcing. A hard switch (lower three pulses), on the other hand, shows much less variation, and maintains a reasonable pulse shape across the range of conductivities, even though the switch performance (risetime, etc.) is the same in both cases. This pulse variability, unfortunately, often eliminates impedance matched modulator designs (using pulse transformers and/ or pulse forming networks) from consistent performance in a PEF system. Table 4.4 summarizes many of these key architecture and performance relationships. This is intended to highlight the relative performance differences between each type of architecture, rather than serve as an exhaustive comparison. Table 4.3

Example PEF protocol

Flow rate Cell diameter Cell length Velocity Transit time

300 liter/hr 0.5 cm 0.6 cm 424 cm/s 1.4 m/s

Minimum rep rate

1768 Hz

Field Vsec Conductance Resistance Ncells Net resistance

35 kV/cm 21 kV 5 ms/cm 611 Ohms 4 153 Ohms


Fig. 4.9 Modeled pulse waveforms for 6 : 1 transformer coupled (a±c) and hard switch (d±f) modulator designs. Pulse performance into conductivities of 1 mS/cm (a and d), 3 mS/cm (b and e), and 5 mS/cm (c and f) are shown. The variation in risetime of the hard switch modulators due to the changes in load impedance are relatively small, and result in reasonable PEF pulses, with risetimes at or below 1 s. The transformer coupled system, on the other hand, essentially multiplies the changes in load impedance by the transformer ratio squared (or 36 in this example). The result is that the pulse risetime may exceed total pulse width, preventing the full voltage from being applied. ß Woodhead Publishing Limited, 2010


Fig. 4.9

Continued


89

90


Table 4.4 Qualitative assessment of potential PEF modulator architecture and switches, rated as H(igh), M(edium), and L(ow) performance on each parameter Architecture

Voltage Rise/ flattop fall time

PFN/Pulse transformer Spark gap M SCR M Thyratron M Hybrid modulator IGBT H Tetrode H (vacuum tube) Hard switch IGBT H Tetrode H (vacuum tube)

Variable Pulsewidth load flexibility impedance

High frequency

Reliability/ lifetime

H H H

L L L

L L L

L M L

L H L

M M

M M

M M

M M

H M

H H

H M

H H

H M

H M

4.3.3 Treatment chambers The third major element of a PEF system is the treatment chamber (Fig. 4.10) where the high voltage pulses are applied to the food. The key attributes of the treatment chamber are its ability to minimally impact the fluid flow, while ensuring a consistent electric field is applied to all elements of the flow. Unfortunately, these two attributes are often in conflict, making design of the treatment chamber a critical exercise in system optimization. It is critical that the treatment chamber design produce a consistent field over the gap, and be as immune as possible to arcing (electrical breakdown). The two basic factors affecting treatment chamber design and operation are the physical configuration of the chamber, and the electrode and insulator materials themselves. There are many chamber designs that have been developed and patented over the last 20 years, but the prevailing approach is the co-field flow chamber design, developed and patented by OSU. This design has been shown to provide the optimal balance between the flow and field requirements. One critical attribute of this design, however, is that to maintain consistent field strengths, the gap over which the field is applied must be proportional to the pipe diameter. Therefore, larger pipe diameters, which support higher flow rates, require proportionally higher pulse voltages to maintain the same field strength. The co-field flow design is best utilized at 5 cm pipe diameters and below, which translates to ~200 kV pulses (at 40 kV/cm), the nominal limit for solid-state, hard switched modulators. For larger pipe diameters, alternative modulator or treatment chamber designs are likely to be required. In most PEF systems, the voltage is applied to the fluid across multiple treatment chambers, which are used in (fluid flow) series as the fluid passes through the PEF system. This allows the desired treatment time to be applied to



91

Fig. 4.10 Cutaway view of OSU co-field flow chamber built under license by DTI. The food flows through the small hole in the center of the chamber diameter, and the high voltage pulses are applied to the center conductor, establishing two treatment areas within the chamber itself.

a faster flow rate, at lower pulse frequencies, and helps ensure that every element of the flow receives the desired field strength and duration, for more consistent treatment. All of the PEF treatment chambers have two primary elements: the electrodes, which are conductors that pass the electrical voltage into the liquid being processed, and the insulators between these electrodes, which maintain the `gap' between electrodes, where the voltage field is applied (Fig. 4.11). Insulators are the simplest aspect of this design. Substantial research has been conducted in the high voltage world, for radar, power distribution, and other applications, to characterize insulator capabilities and failure mechanisms. Multiple insulators exist from this known population that are food-grade materials, allowing their use in PEF treatment chambers. The key is to utilize a material which can withstand both the electrical stresses within the treatment chamber as well as the mechanical stresses imposed by rapidly pulsing the electrodes (due to the electromagnetic force applied during each pulse). Electrodes, on the other hand, are still an active area of research. Early PEF electrodes were built from carbon (graphite), with the intent of avoiding any contamination of the treated product by non-food materials. Unfortunately, these early electrodes had very short lifetimes. The electrical current would erode them rapidly, and deposits would develop on the electrode surfaces, which acted


92


Fig. 4.11 Commercial-scale treatment chambers, at 1 cm diameter and 1.25 cm gap. Two chambers, with four treatment zones, are included in this design. The insulators are the four white discs, with the HV connections between the top two and bottom two insulators. Quick disconnects allow rapid assembly and disassembly of the treatment chambers for cleaning and electrode replacement.

as electrical insulators. These early studies reported electrode life ranging from tens to hundreds of hours. When an electrode needs to be replaced, the cost incurred is not only the cost of the electrode, but also the cost of the system downtime for the electrode replacement and the cost of cleaning the entire process line before operation can be resumed. Since those early PEF systems in the 1980s, considerable research has been performed to develop treatment chambers with extended operational lifetimes, supporting extended, continuous operation of the PEF systems in both R&D and commercial applications. There are three major factors that impact electrode life: erosion, cathodization, and deposition. Erosion relates to physical wear of the electrode as the fluid passes through it. Many foods, especially those that are acidic or highly particulate, will cause electrode erosion even in the absence of high voltage pulses. This is not a significant problem, or it would be experienced in every liquid processing system, including pasteurization systems. More critically, the application of high voltage pulses leads to cathodization of the negative electrode, where electrode ions are transported out of the electrode by the voltage itself. Minimizing this effect, while maintaining food safety, is therefore a major area of research. Figure 4.12 shows the relative cathodization of different materials in OSU's experiments, and shows that titanium (Ti) and



93

Fig. 4.12 OSU data showing the erosion levels of PEF treatment chamber electrodes (Ti = titanium; Pt = platinum, Fe = iron, B = boron) in re-circulated media. Higher concentrations show that more electrode material has been transferred into the media. Note that for typical PEF applications, the food passes through the treatment chambers only once.

platinum (Pt) have the best performance. The key is that minuscule amounts of material are introduced into large volumes of fluid, and these very low concentrations are acceptable in every known country, allowing use of these materials in PEF system. DTI and others have worked with a variety of materials, including combinations of materials, in electrodes with excellent results to date. Wear on the electrode is also caused by arcing and corona discharge within the electrode. This is not a normal product of PEF processing. It occurs when the voltage is not well controlled, or when dissimilar particulates within the fluid cause areas of field stress beyond the breakdown voltage of the liquid. DTI's solid-state, high voltage systems are specifically designed to maintain voltage control within the system. They identify and respond to arcing very quickly by removing voltage from the chambers, limiting arc energy and the pitting at the arc site on the electrode (Fig. 4.13). This capability alone appears to minimize the effect of arcing on the electrode, as demonstrated in multiple DTI systems. Another potential cause of electrode wear has been reported in the literature as a result of DC or low frequency currents ± primarily leakage currents between high voltage pulses. Every solid-state high voltage switch has some leakage current when the switch is off ranging from micro-amps to milliamps. Based on this data, preventing this wear requires that the leakage current must be prevented from flowing through the treatment chambers themselves. In PEF systems with pulse transformers, and in bi-polar systems, this leakage current is shunted around the treatment chambers, and therefore has no impact. For hard switched, mono-polar systems, it is possible to include a bypass inductor in parallel with the treatment chamber. This inductor appears as a short circuit between high voltage pulses (which will waste energy), and as an open circuit


94


Fig. 4.13

Electrode erosion on the inner diameter of the PEF treatment chamber.

during the short PEF pulses. At the same time, however, the erosion of the electrodes is almost exclusively related to the total amount of charge flowing through them. In a commercial PEF system, however, operating at hundreds of amps of peak current, it would take decades of leakage current to equal the charge transport represented by one hour of pulsed operation. The final factor related to treatment chamber design and lifetime is anodization/deposition of material on the positive electrode. This is the major argument for bi-polar pulsing: alternating the polarity of the pulses applied in the treatment chamber may prevent molecules from depositing on the positive electrode (since it becomes the negative electrode in the next pulse). There are two sources of potential deposition on the electrode, the cathode and the liquid itself. It is difficult to envision that true anodization can be an issue in a PEF system, since the liquid flow in the chamber will transport any ions from the cathode out of the chamber before they deposit on the anode. Since PEF is a single pass system, there is little opportunity for these ions to reattach. A more likely effect is that proteins and other ionized molecules within the liquid deposit on the electrodes, especially in areas of high field strength. The anecdotal evidence for this is widespread, but is primarily associated with electrodes with very rough surfaces (such as the graphite electrodes used in early PEF systems). Again, recent data shows that operating with shorter pulses at higher pulse frequencies minimizes this effect. There are, however, still issues to be addressed with specific products (such as milk), which seem to create deposits on the electrodes relatively rapidly.



95

4.4 Pulsed electric field (PEF) system trade-offs and optimization As the previous sections have demonstrated, there are many factors that impact the design of a PEF system. There are technical constraints on all levels of the equipment, operational requirements that must be met, and financial criteria that must be achieved to make PEF processing a profitable endeavor. Balancing these competing criteria is key to the optimization of PEF systems for commercial processes. Our approach is to begin with the fixed parameters facing every PEF system. All of the key PEF system design parameters trace back to the following three elements: · required process protocol (field strength, treatment time) · product characteristics (conductivity, viscosity) · desired throughput/flow rate. The critical, and often least defined, parameters come from the treatment protocol itself. A starting point for these conditions can typically be found in the PEF research literature, but often requires specific experimentation to define the specific process conditions required. As discussed earlier, it is critical that the pulse shapes in the pilot system are as similar as possible to those of the commercial system to allow the results to be readily scaled. These parameters alone allow calculation of the average power required in the PEF system, which determines the overall system size. The system voltage and treatment chamber size (which are directly related) are the first parameters to be selected. A 35 kV/ cm field strength requirement, for example, can be satisfied by a 35 kV pulse power system, with a 1 cm gap distance, or a 17.5 kV system with 0.5 cm gaps, or any other combination which results in the desired field strength. The treatment chamber size, in turn, dictates both the current required in each chamber for each pulse (due to the chamber volume and fluid conductivity), and the chamber diameter (which, with viscosity, determines the pressure drop in each treatment chamber). Finally, the number of chambers, pulsewidth, and pulse frequency can be adjusted to ensure that the fluid receives the desired total treatment time. All of these parameters are interactive, and must be simultaneously assessed against the cost and complexity (and often, even the feasibility) of the required pulse modulator. At DTI, we perform these optimizations regularly, and have developed software tools to support these interactive design decisions. They are not, however, generalizable. They are based on our approach to pulse modulator design, and other modulator designs would lead to different optimizations. Three potential system designs are illustrated in Table 4.5 for the same protocol and conductivity, based on the requirement for processing 5000 liters/ hour of juice (conductivity 5 ms/cm) at 35 kV/cm for 40 s total treatment time. Each column represents a different approach to the PEF system design, with the major variable being the number of treatment chambers used (1±4). The only common parameter is average power, which is wholly dependent on the desired


96


Table 4.5 PEF system sizing examples, with differing numbers of treatment chambers Protocol was 35 kV/cm, 40 s treatment time, with a conductivity of 5 ms/cm. Pressure was not considered in developing these examples No. of chambers Gap distance (cm) Gap diameter (cm) Voltage (kV) Peak current (A) Peak power (MW) Pulsewidth (s) Pulse frequency (Hz) Delta T/chamber (ëC) Avg. power (kW) Fluid velocity (m/s)

1

2

4

2.4 2 84 550 46.2 6.7 553 52 170 4.4

1 0.85 36 200 7.2 6.7 3600 26 170 24.5

0.72 0.6 25 200 5 5 6882 13 170 49

treatment time and field strength. The Delta T/chamber, multiplied by the number of chambers, is a constant, so the protocol determines the total temperature rise. With multiple chambers, the temperature rise is lower in each chamber, introducing the opportunity to minimize temperature excursions by cooling the product between treatment chambers. This is not possible in the single chamber design. The peak power required from the modulator decreases considerably as the gap distance is reduced (and more chambers are added). For solid-state modulators, the major determinant of cost is typically the peak power required, which is clearly lowest in the four chamber design. The selection of a specific design for this set of conditions, however, may hinge on two major factors: the ability of the modulator to operate at the higher pulse frequency, and the pump pressure required to maintain the fluid velocity through the four chambers. Operating at high pulse frequencies increases the switching losses in the modulator proportionately, and can create thermal problems in the switches if not carefully designed and cooled. Pump pressure (not shown, but calculable from the fluid velocity and viscosity, number of chambers, and chamber diameter) is typically limited for a commercial system, as it affects not just the pumps, but fittings, pipes, and other components, including the treatment chambers themselves. Optimization of a PEF system for any given application clearly requires a close interaction between the design of the pulse power system, and the capabilities of the rest of the plant in terms of pumping, cooling, etc. The final factor in system optimization cost is often the most critical. For commercial systems, the cost that matters is made up of two elements: the capital cost for a level of capacity (in $/liter/hour), and the cost of the electricity needed to provide the PEF treatment itself. The electricity cost is made up of two parts: the protocol (which is typically not subject to change), and the efficiency of the PEF system itself. As described earlier, the closer the pulse



97

Fig. 4.14 Commercial-scale PEF system, rated at up to 300 kW average power. The modulator and treatment chambers are shown. One 150 kW power supply used with this system is shown in Fig. 4.4. This system can treat up to 10 000 liters/hour.

shape of the PEF system is to the ideal pulse (zero rise and fall-time, perfect stationary flattop), the higher the system efficiency. For a commercial system, operating nearly continuously, improvements in system efficiency (even at the expense of higher initial system costs) typically yield the lowest overall processing cost. Finally, PEF system capacity directly affects the capital cost (on a per-liter basis). Moving to higher capacity means increasing the pulse voltage and average current in the system, which directly translates to higher PEF hardware costs. These costs do not increase linearly with the added capacity ± so the overall per-liter cost goes down, making larger systems more economical than pilot systems on that basis. As an example, the commercial PEF system shown in Fig. 4.14 has a process capacity over 10 the product per hour as the pilot system shown in Fig. 4.16, but costs only 5 as much, which is half the cost in $/liters/hour. Beyond a certain size, however, the sheer scale of the equipment (in terms of voltage and average power) begins to make this cost increase faster than the additional capacity. At that point, it makes more sense to increase plant


98


capacity by duplicating smaller systems, rather than building ever larger systems. DTI's current assessment is that this crossover point is reached around 500 kW of average power, but this size is increasing over time, as PEF (and other pulsed power system) designs mature.

4.5 Pulsed electric field (PEF) processing and commercialization status PEF processing has been the subject of focused R&D for over 20 years. In its earliest days, the emphasis was on validating and defining the ability of high voltage pulses to kill bacteria, molds, and other potential pathogens. In the early 2000s, programs in both the US and Europe focused on scaling the results discovered in the laboratory to commercially viable processes and systems. Since that time, several key developments have brought PEF processing to the brink of wide-scale commercial adoption. First, Genesis Juice Corporation introduced the first FDA approved, PEF processed juices to the US market in August, 2005 (Fig. 4.15). Genesis Juice Company was formed in 1977 to

Fig. 4.15 PEF treated Genesis Juice on sale in Oregon, 2006.



99

produce and sell organic, unpasteurized fruit juices in the Portland, OR market. As a result of a 2001 change in FDA policy concerning ùnpasteurized' juices, Genesis received a `warning letter' of non-compliance from the US Food and Drug Administration on 13 November 2003. Genesis customers removed the company's products from their shelves by February, 2004. Genesis undertook a search for alternatives to pasteurization, which led the company to PEF processing. PEF processing offered the promise of retaining the fresh flavor and nutrients at the heart of Genesis' products, while allowing the company to meet the new FDA regulations. In April 2004, Genesis reached an agreement with OSU to conduct a cooperative research effort on Genesis' juices using PEF processing. By August, 2005, Genesis had met the FDA's requirements using PEF processing, and was able to re-enter the juice market with a line of PEF processed juices, thereby representing the first FDA approved use of PEF for safe, wholesome commercial food products in the US. Consumer acceptance was very high, and Genesis attempted to expand its market reach in the Pacific Northwest, supported in part by the extended shelf life of PEF processed juice, which conferred additional time for shipment and distribution beyond Genesis' traditional, local markets. Unfortunately, the financial impact of no sales for over 18 months, combined with the capital requirements of rapid growth, proved fatal to Genesis. The company was forced to cease operation before selling its brand name in mid 2007. At the time, however, market acceptance of PEF processed juice was strong and growing. At approximately the same time, DTI introduced a standard pilot-scale PEF system (Fig. 4.16), capable of treating 100±500 liters/hour of juices or other products, allowing researchers to process a variety of products for both R&D and pre-production process definition. Since that time, these pilot systems have been deployed to research facilities around the world, including the US, Europe, and Australia. As a result, Genesis and DTI were awarded the Institute of Food Technologists' Food Technology Industrial Achievement Award in 2007. Finally, the world's first known commercial deployment of large-scale PEF technology occurred in 2006 (Fig. 4.14) in Mesa, Arizona. This system, rated at 10 000 liters/hour capacity, was deployed for wastewater treatment, rather than food processing, but contains all of the same system elements. In over 30 months of operation to date, this system has demonstrated the ability to meet the same operational concerns that would impact a food processing plant ± efficiency, reliability, electrode life, and unattended operation. There are at least three companies, including DTI, building PEF systems, at both the pilot and commercial scale, so commercial PEF systems are now readily available to food processors.

4.6

Conclusions

Multiple researchers have shown PEF processing to be equivalent to pasteurization in terms of pathogen reduction for a wide range of liquid foods.


100


Fig. 4.16 25 kV, 25 kW pilot PEF system built by DTI. The power supply and pulse modulator are in the rack on the left, while the treatment chambers are in a separate stainless enclosure on the right.

For foods that are heat sensitive, there are considerable benefits in taste, color, and nutritional value from the non-thermal PEF process. The application of PEF to other industrial processes builds directly on the research in food processing, and new applications of PEF are emerging at a significant pace. The use of solid-state, high voltage pulsed power systems for PEF processing is the key to these commercial applications. Solid-state technology allows this PEF to scale from small laboratory systems to large-scale processing facilities. There are three identified areas where further system development is required. The first is achieving common definitions for critical PEF parameters, such as field strength, pulsewidth, and treatment time. While the definition of these parameters may seem trivial, the ambiguities in their application prevent both researchers and commercial operators from developing an accepted and available set of treatment protocols for differing foods. Developing common definitions and processes for PEF protocols and systems will help researchers, system developers, regulators, and food processors to clearly communicate their needs and constraints, and allow all sides to achieve their objectives more economically.



101

Second, while there has been considerable research into treatment chamber designs and materials, the PEF community has limited experience with extended, commercial operations. Genesis Juice, and DTI's experience in wastewater processing, demonstrate that it is possible to build robust, longlasting treatment chambers, but we do not have an ideal solution. Considerable work remains to fully optimize these treatment chamber designs, especially for high protein foods. Finally, the overall design of the pulsed power systems for PEF is maturing. New pulse power technologies are allowing PEF systems to become smaller and less expensive. The entire community is gaining experience in specifying, building, and operating PEF systems. Commercial PEF processing is now a viable alternative to pasteurization for a number of heat-sensitive products.

4.7

Bibliography

BARBOSA-CANOVAS, G., GOULD, G.,

2000, Innovations in Food Processing, Technomic. 2001, Pulsed Electric Fields in Food Processing,

BARBOSA-CANOVAS, G., ZHANG, Q.H.,

Technomic.

BARBOSA-CANOVAS, G., ZHANG, Q.H.,

C.H.I.P.S.

2002, Pulsed Electric Fields in Food Processing,

1999, Preservation of Foods with Pulsed Electric Fields, Academic Press. BARBOSA-CANOVAS, G., TAPIA, M., CANO, M., 2004, Novel Food Processing Technologies, C.H.I.P.S. CLARK, J., 2006, Pulsed electric field processing, Food Technology, Jan., pp. 66±67. DUNN, J.E., PEARLMAN, J.S., 1987, Methods And Apparatus For Extending The Shelf Life Of Fluid Food Products, US Patent 4,695,472. GAUDREAU, M., HAWKEY, T., KEMPKES, M., PETRY, J., ZHANG, Q.H., 2001, A Solid-state Pulsed Power System for Food Processing, International Food Technology Conference. GAUDREAU, M., HAWKEY, T., PETRY, J., KEMPKES, M., 2004, Design Considerations for Pulsed Electric Field Processing, Proceedings of the European Pulsed Power Conference. GAUDREAU, M., HAWKEY, T., PETRY, J., KEMPKES, M., 2006, Scaleup of PEF Systems for Food and Waste Streams, 3rd Innovative Food Centre Conference. JEYAMKONDAN, S., JAYAS, D.S., HOLLEY, R.A., 1999, Pulsed electric field processing of foods. J. Food Prot. 62(9), 1088±1096. KEMPKES, M., CASEY, J., GAUDREAU, M., HAWKEY, T., ROTH, I., 2002, Solid-State Modulators for Commercial Pulsed Power Systems, Pulsed Power Conference. LI, S.Q., ZHANG, Q.H., 2005, Electrode Erosion Under High Intensity Pulsed Electric Fields, Proceedings of the International Food Technology Conference. MIN, S., JIN, T., ZHANG, Q.H., 2003, Commercial scale pulsed electric field processing of tomato juice, J. Agric. Food Chem. 51 (11), 33383344. MORREN, J., ROODENBURG, B., DE HAAN, S.W.H., 2003, Electrochemical reactions and electrode corrosion in pulsed electric field (PEF) treatment chambers, Innovative Food Science and Emerging Technologies 4, 285±295. RASO, J., HEINZ, V., 2006, Pulsed Electric Fields Technology for the Food Industry, Springer. BARBOSA-CANOVAS, G., GONGORA-NIETO, M., POTHAKAMURY, U., SWANSON, B.,


102


2002, Modeling of Pulsed Electric Field (PEF) Processes, Proceedings of the International Food Technology Conference. TOEPFL, S., 2006, Pulsed Electric Field (PEF) for Permeabilization of Cell Membranes in Food and Bioprocessing ± Applications, Process and Equipment Design and Cost Analysis, PhD Thesis, Technical University of Berlin. USDA, CENTER FOR FOOD SAFETY AND APPLIED NUTRITION, 2000, Kinetics of Microbial Inactivation for Alternative Food Processing Technologies: Pulsed Electric Fields, 2 June. YEOM, H.W., STREAKER, C.B., ZHANG, Q.H., MIN, D.B., 2000, Effects of Pulsed Electric Fields on the Activities of Microorganisms and Pectin Methyl Esterase in Orange Juice. Ohio State University. YIN, Y., ZHANG, Q.H., SASTRY, S.K., 1997, High voltage pulsed electric field treatment chambers for the preservation of liquid food products. US Patent 5,690,978. SALENGKE, S.K., ZHANG, Q.H.,


5 The environmental impact of pulsed electric field treatment and high pressure processing: the example of carrot juice J. Davis, The Swedish Institute for Food and Biotechnology (SIK), Sweden and G. Moates and K. Waldron, Institute of Food Research, UK

Abstract: The environmental impact of novel processing techniques has been investigated using carrot juice as a case study. Pulsed electric field (PEF) treatment and high pressure (HP) processing have been compared to mild thermal pasteurizing of carrot juice using life cycle assessment (LCA) to ascertain environmental hotspots along the chain from cultivation to packaging and delivery of the product. From this analysis, it can be seen that the contribution of the pasteurization step is quite small in relation to the overall impact. Packaging, however, has a significant impact and should be considered when making changes to the process or product. Key words: carrot juice, life cycle assessment (LCA), high pressure and pulsed electric field processing, environmental impact.

5.1

Introduction

Novel processing (NP) is the collective name for a variety of technologies including high pressure (HP) processing, pulsed electric field (PEF), ohmic heating, microwave heating and others. These technologies propose to confer improved quality attributes (e.g., improved texture and freshness) to a variety of products as well as the potential for reduced energy usage (Toepfl, 2006a). Such


104


technologies have already been adopted around the world on a commercial scale for foodstuffs including cold meats and fruit juices, and they currently provide the focus of a concerted research effort in the European Union (NovelQ, 2009). This study uses the technique of life cycle assessment (LCA) to quantify the environmental impact associated with three methods for producing carrot juice: (1) conventional processing (mild thermal treatment), (2) HP-processed and (3) PEF-treated.

5.2

Goal definition and scoping

The aim of the study has been to compare the environmental impact of thermally pasteurized carrot juice with the impact of carrot juice produced by either HP or PEF treatments. The aim has been to identify the component parts of the life cycle that are important in terms of environmental impact, so that these aspects can be taken into account when considering or designing an NP production system. 5.2.1 Functional unit The functional unit of the study was 1 litre of carrot juice at the point of sale for each of the three different types of carrot juice. 5.2.2 Description of products The three different types of carrot juice compared were: (1) carrot juice produced with mild pasteurization with conventional heating technology, (2) carrot juice pasteurized with PEF, and (3) carrot juice pasteurized with HP. All three juices were assumed to be packaged in a 250 mL polyethylene (PE) bottle (based on an example of HP packaging available on the market), and to require chilling after production. 5.2.3 Data sources and quality of data The data for processes in the core system were collected from a number of sources which are specified in the inventory section. In order to evaluate the conventional process, a visit to a factory in Sweden was made to collect data. The specific details of the processing are confidential, but the energy required and storage conditions are documented in the inventory section. Data on the HP and PEF processes were gathered from literature and personal communications with experts. Data on the cultivation of carrots was taken from a recent Swedish study (Davis et al., 2010). For all the processes in the background system, e.g. use of energy and emissions from transport, fuel production and combustion, production of packaging materials, etc., data from the Ecoinvent database (Ecoinvent Centre, 2007) has been exploited.


Environmental impact of PEF treatment and high pressure processing

105

5.2.4 System boundaries The study includes all the steps from cultivation (including inputs to the farm) up to the point of sale (see Fig. 5.1; `T' in the figure stands for transport). Infrastructure is included in the processes for which data are taken from the Ecoinvent database (i.e., production of diesel, electricity and heat, plastics, as well as waste treatment and transport). Thus the environmental burden for these processes also includes the burden for producing and maintaining buildings, industrial plants, vehicles, roads, etc. For other processes, data concerning infrastructure, such as the environmental burden of producing and maintaining the industrial equipment for processing the carrot juice, irrespective of whether it is processed with conventional equipment, HP or PEF, has not been included in the analysis due to a lack of information. In this respect, Frischknecht et al. (2007) have explored the relevance of capital goods for a number of products and services. For agricultural products, the capital goods are very significant when it comes to ecotoxicity and energy use, although this is not the case for other impact categories. Since the agricultural components of all three processes studied will be similar, excluding infrastructure from the agricultural processing should have little effect on the comparison between the products, even though it means that the total impact is slightly underestimated in terms of energy use (use of energy for producing the agricultural machinery). However, in the discussion,

Fig. 5.1

Processes included in the study.


106


we have tried to assess whether there might be significant differences between the environmental burden of the different processing equipment, taking into account the life span and capacities of the equipment. Since most of the bottles of this type (250 mL) will be sold at smaller shops (e.g., newsagents), we assume that people will buy them en route to another destination, for example on their way home from work. Since juice that will be consumed at home is likely to be bought in a larger packaging unit (1 L) we have not included any consumer transport in the analysis.

5.3

Inventory of carrot juice processing

Data on carrot cultivation has been taken from a recent study on Swedish production of vegetables (Davis et al., 2010), which incorporates data from six carrot growers. This has enabled the impact of cultivation on an average farm to be derived. Data on conventional juice production is taken from one carrot juice manufacturer in Sweden. The same variety of carrots is assumed to be used in all three production systems and the production of juice is assumed to take place in the same geographic location independent of technology (i.e., the location of the juice producer visited in the project). 5.3.1 Carrot cultivation The data used for the carrot cultivation is given in Table 5.1. These represent average data from six Swedish carrot growers. We have assumed that the cultivation is on ground comprising 5% peat soil and 95% mineral soil (cultivation on peat soil generates significant emissions of CO2 due to losses of soil carbon). 5.3.2 Transport of carrots and bottles The carrots are transported from the farm to the juice factory by lorry (280 km) and ferry (117 km), with a load of 20 tonnes. The return journey for the empty lorry is 280 km. Ecoinvent data (Ecoinvent Centre, 2007), have been used to model the transport (lorry type: 40 tonnes (t), 100% load factor and empty, respectively; ferry type: barge). The PE granulate for the bottles is transported on average 1100 km to the bottle producer. From there the bottles are transported Table 5.1

LCI data for carrot cultivation

Environmental impact

Per kg carrots at farm gate

Acidification (g SO2e/kg) Eutrophication (g PO4e/kg) Climate change (g CO2e/kg) Primary energy demand (MJe/kg), 49% fossil


0.32 0.29 86.7 1.44


107

a further 150 km to the juice producer. We have assumed a 40 t lorry with a load factor of 70% (Ecoinvent Centre, 2007). 5.3.3 Production of conventional carrot juice The production details for the conventionally produced carrot juice have been collected from a manufacturer in Sweden. The product is treated with a very mild pasteurization to keep the product fresh. After production, it keeps for 14 days in chilled storage, and after it has been opened it should be consumed within three days. The producer uses the Nantes variety of carrot, a cylinder shaped carrot with a sweet, mild flavor and deep orange color. The process starts with the carrots being scrubbed, washed and sorted (any damaged carrots are removed). The carrots are shredded and the juice is pressed out of the shredded carrots. The juice is pumped to a storage tank until the whole batch of carrots has been pressed. From there, the juice is pumped to the pasteurizer and a light pasteurization is performed. The juice is then pumped to a chilled buffer tank, where it is kept until it is pumped to the filling machine for the filling of plastic bottles (0.25 L). The bottles are kept chilled until they are collected by the distributers. The juice is pressed 14 days before the best before date on the bottles. The producer states that the juice is typically consumed within 8 days. The energy used for making carrot juice has been derived by measuring the total energy used at the plant and employing a mass-based allocation by allocating a specific proportion according to the proportion of carrot juice produced out of the total production of all juices (the plant produces a number of different juices). Since production of the different juices at the plant is very similar, mass allocation was considered to be the best method to allocate the burden to the different juices. All the energy used is taken into account, for production, storage of raw materials and products, heating of buildings, ventilation, offices, etc. The data used for the analysis is given in Table 5.2. The use of cleaning agents for the cleaning and sanitizing of equipment was not taken into account due to a lack of data on resource use and emissions produced from cleaning agents. 5.3.4 Carrot waste to animal feed The carrot waste is picked up at the producer by nearby farmers who feed it to their dairy cows. Since this waste stream is free for the farmers (i.e., it is of no economic value to the producer), we have not allocated any environmental burden, or credit, to the carrot waste. In this way, the environmental burden of the juice is not underestimated. 5.3.5 Non-conventional processing For the non-conventional processing we have assumed the same recipe as in the conventional processing case: 1.6 kg carrots per litre of juice. We assume that all


108


Table 5.2

Data for conventional carrot juice production

Inputs and emissions

Per L juice

Inputs: Carrots (kg) Water (litres) Electricity (kWh), for pasteurization

1.60 2.90 0.03

Electricity (kWh), all other

0.23

Natural gas (kWh)

0.13

Outputs: Carrot juice (L) COD (g) Carrot waste (kg) Solid waste to incineration (kg)

1 4.49 0.60 0.04

Corrugated board to recycling (kg)

0.005

Wood waste to incineration (kg)

0.037

Metal waste to recycling (kg)

0.003

Plastics to recycling

0.0002

Ecoinvent LCI data/comments

Èlectricity, medium voltage, at grid/SE' Èlectricity, medium voltage, at grid/SE' `Natural gas, burned in industrial furnace > 100 kW RER'

Sold to farmers `Municipal solid waste to incineration/CH'* Èurope corrugated board base paper, wellenstoff, at plant/RER'* `Disposal, packaging cardboard to incineration/CH'* `Steel, electric, chromium steel 18/ 8, at plant/RER'* Data from article: LCA of thermoplastic recycling (Garrain et al., 2007)

* These recycling and incineration processes from Ecoinvent have been modified to include the avoided burden of replaced material/energy.

novel processing takes place in the same plant as the conventional processing, i.e. the transport distances for inputs to and delivery from the factory are the same in all scenarios. We also assume the same procedure in the plant for pressing juice from carrots, it is only the pasteurization step that is different. Accordingly, we assume the same energy use for all the other steps, apart from the pasteurization process itself, in the production of the juice in all three scenarios. Energy use for pasteurization by HP Data on energy use for pasteurization using HP have been derived by Houska (pers. comm., Food Research Institute Prague, 2009). It is assumed that the bottles are placed in a basket and enclosed in the pressure chamber. The chamber volume is 125 L, with a diameter of 400 mm, and height of 1000 mm. The real chamber capacity is 75 L of juice filled in 300 PE bottles, each having the capacity of 250 mL. The energy needed for pressurizing 1 kg of water (75 L of product, the rest of the pressure chamber volume occupied by working medium) is determined as



109

the work performed by the piston in the cylinder compressing the fluid of bulk modulus Ev and density (for water: Ev 2100 MPa, density 1000 kg.mÿ3). The HP pasteurization of juice is typically done at P 400 MPa with holding time of 10 min. E

p2 400 106 40 000 J/kg 2Ev 2 1000 2 109

Relating the energy consumption to the product alone yields 40 000 125/75 66 666 J/kg. Some energy is also needed for filling and emptying the chamber, so in the analysis we have assumed a total energy use for the HP pasteurization of 100 kJ/L of carrot juice. Energy use for pasteurization by pulsed electric fields Data on energy use for pasteurization using PEF have been taken from Hoogland and de Haan (2007) and is based on an industrial pilot plant with the following parameters: flow rate 5000 L/h, process conductivity 0.2±0.7 S/m, field strength 2.5±3.5 kV/mm, pulse duration 2 s, number of pulses 5, and total power 75 kW. This gives an energy use of 54 kJ/L (75 60 60=5000), which has been used in this LCA study. Data for PEF use can, however, cover quite a wide range Toepfl et al. (2006b) estimated total costs of PEF preservation based on two specific energy inputs of 50 kJ/kg and 700 kJ/kg respectively. One reason why the energy use may vary is because the energy requirement depends on the conductivity of the product, which differs between products (Ruhlman et al., 2001). 5.3.6 Packaging For all three products, a 250 mL HDPE bottle has been chosen. Data on the weight of the bottle and cap is based on an HP-treated product available on the market. The weight of the bottle is 19.5 g HDPE and the cap 2.9 g HDPE. Life cycle inventory (LCI) data from Ecoinvent has been used for the production of virgin HDPE and blow moulding of the bottles. The bottles are wrapped in PE plastic in packs of four, 7 g of plastic sheet per pack of four, based on data from the conventional producer. We assume that this secondary packaging is incinerated with energy recovery. Regarding the bottles, after consumption, 30% of the bottles are sent to material recycling and 70% to municipal waste incineration with energy recovery (FTI, 2009). 5.3.7 Transport from juice manufacturer to point of sale The product is sold all over Sweden, in all kinds of shops, such as small newsagents and large supermarkets. We assume an average distance of 400 km by chilled lorry (Ecoinvent 40 t, load factor 70%) from the producer to the point of sale.


110


5.3.8 Point of sale We assume that the majority of the juice in this packaging unit is sold at small convenience shops (e.g., newsagents). Data on the energy use (0.007 kWh electricity/L) from storing food at the retailer is based on a survey conducted at Swedish retailers (Carlsson, 2000). No specific data on smaller shops was available. Hence, we assume that all three products are similar in terms of storage time and storage condition (refrigeration).

5.4 Choice of impact categories and impact assessment methods The impact categories included in the analysis are the use of primary energy, contribution to global warming potential, eutrophication potential, and acidification potential. These are important when considering the environmental impact from production of food products. Another important impact category in food LCAs is the contribution to ecotoxicity as a consequence of the use of pesticides. However, due to a lack of standardized methods for assessing the toxicity of the pesticides, this impact category has not been included in the analysis. The method used for the environmental impact assessment of the selected categories is CML in SimaPro (PreÂ, 2006), amended so that it uses the latest characterization factors for GWP from IPCC 2007. For primary energy use, the method Cumulative Energy Demand (CED) in SimaPro was used.

5.5

Results

The following processes/activities are included in the labels in Figs 5.2±5.5. Cultivation of carrots

Juice production

Pasteurisation Packaging Transport Retail Total

Cultivation of carrots including production of farm inputs (NPK fertilizer, pesticides and diesel), combustion of diesel and use of electricity for watering and cold storage of carrots Use of energy for washing, scrubbing, sorting, shredding and juicing harvested carrots (i.e., all processes other than pasteurization), and the treatment of waste material (other than carrot waste, e.g. cardboard), COD to water Production of electricity used for pasteurization (by heat, HP, or PEF) Production and recycling/incineration of packaging All truck and boat transport in the system including transport of packaging Storage of juice at retailer All processes/activities.



Fig. 5.2

111

Use of primary energy for the juices produced with different techniques (MJ-equivalents/L juice).

Figure 5.2 shows the primary energy demand for the different juices. Production and waste treatment of packaging demands the most energy, followed by transport, carrot cultivation and juice production. In the overall scheme of things, the pasteurization step requires relatively little energy (this is the case for all three technologies). The global warming potential data in Fig. 5.3 gives a very similar picture to that of use of energy, since the emissions are closely linked to the energy use. However, Swedish electricity production generates very little greenhouse gas emissions because it is mainly non-fossil-fuel based. Hence, the steps that solely require electricity (pasteurization and retailer) make a very small contribution to climate change. Also, some of the energy used at the juice producer step is electrical so the impact here is slightly smaller than for energy use. Of course, this may be different in other countries that use fossil fuels for electricity generation. When it comes to eutrophication (Fig. 5.4), the leakage of nutrients from cultivation makes the most important contribution, followed by emissions of NOx from combustion of natural gas at the juice manufacturer, transport, and packaging. For acidification in Fig. 5.5, it is mostly emissions of SO2 and NOx from the tractor at the farm, transport, and packaging production that make the main contributions.


112


Fig. 5.3 Global warming potential for the juices produced with different techniques (kg CO2-equivalents/L juice).

Fig. 5.4

Eutrophication potential for the juices produced with different techniques (g PO4-equivalents/L juice).



Fig. 5.5

5.6

113

Acidification potential for the juices produced with different techniques (g SO2-equivalents/L juice).

Discussion and conclusions

From the analysis we can see that for this type of product, for which the environmental burden from agriculture is moderate in relation to the total impact (this holds true for all studied impacts except eutrophication), the role of packaging is very important. In contrast, the energy use for pasteurization is so small in comparison to the total life cycle energy use for the products, a significant difference in energy use for pasteurization does not make any difference between the products overall. In a previous study (Davis et al., 2009), the production of salsa using different technologies was compared. In that study, the conventional salsa was cooked longer, and hence the difference between the technologies was more evident. In the current carrot juice study, only a light pasteurization process was performed, so therefore the differences in energy use were more marginal. For products where more energy-intensive processing is undertaken, these novel technologies might prove to be more beneficial in terms of overall energy savings than those found here. The environmental impact from production and waste handling of the equipment used in the industrial processing is not included in the analysis, due to a lack of information. However, it is estimated that for the conventional processing, the equipment (pasteurizer, pipes, etc.) are used for a number of years producing many millions of liters of juice over the life span of the equipment. As a consequence, the burden split per manufactured unit (e.g., per liter of juice) is expected to be very small. For the other types of processing, HP and PEF, it is


114


expected that the equipment would also be used for quite large production volumes over a number of years (shorter or longer than the conventional equipment is hard to say), so would not expect to contribute significantly to the environmental impact of each produced volume of juice. Furthermore, since the equipment is mostly constructed from large pieces of metal, which are fairly easy to recover and recycle (and can most likely also be made out of recycled metal) this further limits the burden associated with the equipment. In this study, we have assumed that all three products being compared are stored at similar temperatures and times after processing. It is possible that the HP and PEF products may have longer shelf lives and reduce the wastage of the product at the retailer compared with the conventional product. Since we have limited knowledge on the relationship between shelf life and product wastage, it is difficult to analyze the influence of the longer shelf life on the environmental impact of the products. However, a decreased wastage is generally beneficial for the environmental profile of a product. One drawback of this study is that data on thermal pasteurization has been gathered from a running industrial plant, whereas the data on the HP and PEF pasteurization processes are based on literature values and calculations. This is because there are currently a limited number of industrial-scale plants using these novel technologies and access to data has been difficult. Further research is needed to acquire actual information on energy and resource use from industrial operations using these novel technologies. It has not been possible to compare the price of the three different products, due to a lack of information, but this would have been interesting. Generating greater economic value without increasing the environmental impact is often called increased eco-efficiency, which in turn is very much in line with sustainable development. Sustainable development is partly about decreasing the environmental impact per given benefit; economic value is a common measurement for benefit. Hence, if the quality aspects of the HP and PEF products could motivate a higher price, this would give a positive environmental profile for these products. In summary, when exploring alternative technologies for juice production, the packaging choice is one that must be carefully considered, in order not to outweigh possible environmental benefit of the technology with a packaging solution that is associated with heavy environmental burden. Furthermore, this study highlights that the increased quality aspects that novel technologies can offer is possible without any increased environmental burden.

5.7

Acknowledgements

This work has been part-funded by the Biotechnology and Biological Sciences Research Council (BBSRC), UK, and the Commission of the European Communities, Framework 6, Priority 5 `Food Quality and Safety', Integrated Project NovelQ FP6-CT-2006-015710. This study does not necessarily reflect the views of the Commission and its future policy in this area.



5.8

115

References

(2000), Livscykelinventering av butiker ± Data och metoder foÈr att beraÈkna butikens roll vid LCA av livsmedel (Life cycle inventory of retail stores ± Data and methods to calculate the contribution at the retail stage in an LCA of food, in Swedish), SIK-Rapport Nr 676, SIK ± the Swedish Institute for Food and Biotechnology, Gothenburg, Sweden. DAVIS J., MOATES G.K. and WALDRON K.W. (2009), High-pressure processing: a step toward sustainability?, Food Safety Magazine, 15, 12±15. DAVIS J., WALLMAN M., EMANUELSSON A. and SUND V. (2010), Emissions of Greenhouse Gases from Production of 19 Fruits, Vegetables and Plants sold in Sweden ± Part one: Analysis of current production (project report to be published in 2010), SIK ± the Swedish Institute for Food and Biotechnology, Gothenburg, Sweden. ECOINVENT CENTRE (2007), Ecoinvent data v2.0. Ecoinvent reports No. 1-25, Swiss Centre for Life Cycle Inventories, DuÈbendorf, 2007. CARLSSON K.

FRISCHKNECHT R., ALTHAUS H.-J., BAUER C., DOKA G., HECK T., JUNGBLUTH N., KELLENBERGER

D. and NEMECEK T. (2007), The environmental relevance of capital goods in life cycle assessments of products and services, Int J LCA, 12, 7±17. Ê sa SohleÂn at FTI ± FoÈrpacknings- och FTI (2009), personal communication with A tidningsinsamlingen, Stockholm, Sweden (www.ftiab.se). GARRAIN D., MARTINEZ P., VIDAL R. and BELLES M. (2007), LCA of thermoplastics recycling, 3rd International Conference on Life Cycle Management, ZuÈrich, 27±29 August 2007. HOOGLAND H. and DE HAAN W. (2007), Economic aspects of pulsed electric field treatment of food, In: H L M Lelieveld, S Notermans and S W H de Haan (eds), Food preservation by pulsed electric fields: From research to application, Woodhead Food Series No. 144, Woodhead Publishing, Cambridge. HOUSKA M. (2009), personal communication, Head of Department of Food Engineering, Food Research Institute, Prague, Czech Republic. IPCC (2007), Intergovernmental Panel on Climate Change 2007, IPCC Fourth Assessment Report, The Physical Science Basis (http://www.ipcc.ch/ipccreports/ar4-wg1.htm). NovelQ, Integrated Project NovelQ FP6-CT-2006-015710 (www.novelq.org). PREÂ CONSULTANTS (2006). SimaPro Software 7.0, Amersfoort. The Netherlands (www.pre.nl). RUHLMAN K., JIN Z. and ZHANG Q. (2001), Physical properties of liquid foods for pulsed elecric fields treatment. In: G. Barbosa-Canovas, Q. Zhang and G. TabiloMunizaga (eds), Food Preservation Technology Series: Pulsed Electric Fields in Food Processing Fundamental Aspects and Applications, Technomic Publishing Company, Lancaster, PA. TOEPFL S., MATHYS A., HEINZ V. and KNORR D. (2006a), Review: potential of high hydrostatic pressure and pulsed electric fields for energy efficient and environmentally friendly food processing, Food Reviews International, 22, 405± 423. TOEPFL S., HEINZ V. and KNORR D. (2006b), Applications of pulsed electric fields technology for the food industry. In: J. Raso and V. Heinz (eds), Pulsed Electric Fields for the Food Industry: Fundamentals and Applications, Springer Verlag, Heidelberg.


Part II Case studies in other novel food processing techniques


6 Industrial applications of high power ultrasonics in the food, beverage and wine industry D. Bates, Cavitus Pty Ltd, Australia and A. Patist, Cargill Inc., USA

Abstract: Since the late 1990s, high power ultrasound has become an alternative food processing technology applicable to large-scale commercial applications for emulsification, homogenization, mass transfer, enhanced heat transfer, anti-fouling, extraction, crystallization, de-aeration, fermentation, enhanced food functionality, de-foaming, inactivation of enzymes, particle size reduction, extrusion, and both temporary and permanent viscosity alterations. This can be attributed to significant improvements in equipment design and efficiency, as well as in the application of the technology. This chapter presents an introduction to the technology and discusses several examples of ultrasonic applications in the food, beverage and wine industry that have realized successful commercialization, and includes advantages and limitations of ultrasonics. Key words: ultrasonics in food, beverage, wine, high power ultrasound, industrial applications of ultrasound.

6.1

Introduction

High power ultrasonics has been investigated at the laboratory bench for many years in academia and industry; however, major advances have been made in the last 10±15 years transforming this previously laboratory-based prototype technology into currently fully operational commercial processes used throughout the world. The applications of high power ultrasound in a food and beverage


120


operation range from enhancing existing processes by retro-fitting high power ultrasonic technology, to developing processes previously not considered possible with conventional energy sources. Discussion will include the principal mechanism of ultrasound, some key process parameters, a list of applications in the food industry, and several examples of ultrasonic applications that have made it to the commercial scale. A `roadmap' comprising some basic steps to successfully scaling up an innovative technology in general is also presented.

6.2

High power ultrasound

The main effect of ultrasound on a fluid is to impose an acoustic pressure (Pa) in addition to the hydrostatic pressure already acting on the medium. The acoustic pressure is a sinusoidal wave dependent on time (t), frequency (f), and the maximum pressure amplitude of the wave, Pa,max (Muthukumaran et al., 2006): Pa Pa;max sin 2ft

6:1

The maximum pressure amplitude of the wave (Pa,max) is directly proportional to the power input of the transducer. At low intensity (amplitude), the pressure wave induces motion and mixing within the fluid called acoustic streaming (Leighton, 1994). At higher intensities, the local pressure in the expansion phase of the cycle falls below the vapor pressure of the liquid, and causes tiny bubbles to grow (created from existing gas nuclei within the fluid). A further increase generates negative transient pressures within the fluid that enhances bubble growth and produces new cavities by the tensioning effect on the fluid (Mason, 1998). During the compression cycle, the bubble shrinks, and its contents are absorbed back into the liquid. However, since the surface area of the bubble is now larger, not all vapor is absorbed back into the liquid, and the bubble grows over a number of cycles. Within a critical size range, the oscillation of the bubble wall matches that of the applied frequency of the sound waves, causing the bubble to implode during a single compression cycle (Moholkar et al., 2000). This process of compression and rarefaction of the medium particles and the consequent collapse of the bubbles comprises the well-known phenomenon of cavitation, the most important effect in high power ultrasonics. The conditions within these collapsing bubbles can be dramatic, with temperatures of 5000 K and pressures of up to 2000 atmospheres (Suslick, 1988; Laborde et al., 1998). It is the combination of these factors (heat, pressure and turbulence), which is used to accelerate mass transfer in chemical reactions, create new reaction pathways, breakdown and dislodge particles (when cavitation is in proximity of a solid surface) or even generate different products from those obtained under conventional conditions (Suslick, 1988). When sound waves reflect on a solid surface or an air±water interface, a standing wave can be formed. The acoustic pressure at the nodes is equal to zero, and at the anti-node, the acoustic pressure fluctuates from a maximum to a minimum. Leighton (1994) and Laborde et al. (1998) explain that bubbles


Industrial applications of high power ultrasonics 121 smaller than the resonance size accumulate at the anti-node, whereas bubbles larger than the resonance size accumulate at the node and consequently coalesce as they collide. This process of bubble transport and growth at the nodes and anti-nodes is called microstreaming and is the main mechanism for ultrasonic degassing. Ultrasound can be divided into three frequency ranges: · power ultrasound (16±100 kHz) · high frequency ultrasound (100 kHz±1 MHz) · diagnostic ultrasound (1±10 MHz). Power ultrasound (20±100 kHz) is used for most sonochemical applications, but because cavitation can be produced using sound at frequencies from within the audible range to frequencies as high as 2 MHz, the frequency range used for sonochemistry applications is expanding. However, most reaction processes will operate at their optimum of 17±24 kHz, as this is the frequency at which the maximum (cavitational) energy can be attained. The use of ultrasonics in industrial processes has two main requirements; a liquid medium (even if the liquid element forms only 5% of the overall medium, yet the medium is still `pumpable') and a source of high-energy vibrations (the ultrasound). The vibrational energy source is called a transducer and there are two main types; piezoelectric and magnetostrictive. Piezoelectric transducers are the most commonly used in commercial scale applications due to their scalability (i.e., the maximum power per single transducer is generally higher than that of magnetostrictive transducers). The technology in the area of commercial ultrasonic equipment is developing at a great pace and no novel process for the application of ultrasound in the industry is possible without ultrasonic equipment manufacturers willing to build new designs according to the requirements of customers. Figure 6.1 shows an example of two potential flow cell designs for a commercial flow-through application. The shaded areas represent the sonotrode (i.e., ultrasonic probe). The design on the left allows for a highly concentrated area of energy, whereas the design on the right allows for larger flows, including multiple systems in series without a significant pressure drop.

6.3

Process and scale-up parameters

6.3.1 Energy and intensity Ultrasonic liquid processing can be described by the following parameters: amplitude, back pressure, temperature, viscosity, and concentration of solids. The experimental outcome (e.g., percent improved extraction yield and/or rate) is a function of: 1. Energy ± the energy input per volume treated material (in kWh/L); 2. Intensity ± the actual power output per surface area of the sonotrode (in Watts/cm2).


122


Fig. 6.1 Possible continuous flow cell designs for a commercial application. The shaded areas represent the ultrasonic probe. Note that the generator is not shown for the system on the right.

The energy input is a function of power and flow rate and can be described as follows for continuous applications: Power of Sonotrode (W) 6:2 Winput (kWh/L) Wspec Q (L/min) 60 (min/hr) 1000 (W/kW) and for batch applications: Winput (kWh/L) Wspec

Power of Sonotrode (W) treatment time (s) 3:6E6 (J/kWh) Volume of treated materials (L) 6:3

The energy intensity can be calculated by: Power of Sonotrode (W) Wdensity (W/cm2 ) Surface area of the Sonotrode (cm2

6:4

Both energy and intensity are independent of scale, and any ultrasonic process will be scaleable using these two parameters (Hielscher, 2005). A very general relationship between flow rate and energy for several ultrasonic applications is shown in Fig. 6.2 and indicates how the flow rate vs. energy relationship depends on the application. Ultrasonic pasteurization in combination with mild heat (> 50 ëC), for example, requires a lot of energy per volume of treated


Industrial applications of high power ultrasonics 123

Fig. 6.2 General relationship of flow rate (liters per hour, Lph) vs. energy (kilowatts, kW) for several ultrasonic applications.

material, and hence the maximum flow rate achievable is relatively small (i.e., per kW of energy). On the other hand, de-gassing requires a very small amount of energy, and much higher flow rates can be treated per kW of ultrasonic energy. 6.3.2 Pressure Increasing the pressure (as controlled by the back pressure of the flow line) increases the cavitation threshold and reduces the number of cavitation bubbles (Muthukumaran et al., 2006). On the other hand, increasing the external pressure increases the pressure in the bubble at the moment of collapse and results in a more violent collapse (Lorimer and Mason, 1987). Therefore, increasing the back-pressure can be an effective tool in intensifying the process without having to increase the amplitude (Hielscher, 2005). In a typical industrial ultrasonic application, back pressure is set between 1 and 5 bars, depending on the application. 6.3.3 Temperature and viscosity Temperature affects the vapor pressure, surface tension, and viscosity of the liquid medium (Muthukumaran et al., 2006). While increased temperature increases the number of cavitation bubbles, the collapse is `cushioned' or `dampened' by the higher vapor pressure. Cavitation bubbles form less easily in a highly viscous environment, and increased temperature decreases the viscosity and allows for more violent collapse. Thus, there is an optimum temperature at which the viscosity is low enough to induce formation of enough cavitation bubbles, yet the temperature is still low enough to avoid the dampening effect related to the higher vapor pressure. The parameters discussed above make it clear that there is no off-the-shelf solution for every ultrasonic application. It takes time to develop the process with the goal of achieving the optimal result with a minimum amount of energy and number of transducers required for commercial applications.


124


6.4

Applications and benefits

6.4.1 Summary of applications Ultrasonics has shown benefits in a wide variety of applications. Five key areas highlighted in this chapter are summarized in Table 6.1. For an informative and comprehensive list of ultrasonic applications, the reader is referred to Patist and Bates (2008). 6.4.2 Extraction The extraction of organic compounds from agricultural products, plants or seeds has conventionally been based on a combination of solvent, heat and agitation. This can be significantly improved by the use of ultrasound, as the energy generated from collapsing cavitation bubbles provides greater penetration of the solvent into the cellular material and improves mass transfer to and from interfaces. At higher ultrasonic intensities (i.e., Watts/cm2), extraction processes can be further improved with the disruption of cell walls and the release of Table 6.1 Examples of large high power ultrasound (HPU) applications in the food industry (references in the appropriate subsections) Application

Mechanism

Benefit

Extraction

Increased mass transfer of solvent, release of plant cell material through cavitational cell rupture Cavitation resulting in high shear micro-streaming

Increased extraction rate and yield in solvent, aqueous or supercritical systems

Emulsification/ homogenization Viscosity alteration

Defoaming

Cleaning and sanitation

Reversible and non-reversible structural modification via cavitation induced high-shear micro-streaming. Sono-chemical modification involving crosslinking and molecular restructuring and alignment Airborne pressure waves causing foam bubble collapse

Increased heat transfer through high shear. Direct cavitational damage to microbial cell membranes

Cost effective emulsion formation: reduced level of emulsifier, small particle size, and narrow distribution Non-chemical modification for improved processing traits, reduced additives, unique functionality

Increased production throughput, reduction or elimination of antifoam chemicals, as well as reduced wastage in bottling lines Enzyme inactivation adjunct at lower temperatures for improved quality attributes, enhanced food safety, contamination removal


Industrial applications of high power ultrasonics

125

cellular materials (Knorr, 2003; Zhang et al., 2003; Vinatoru, 2001; Li et al., 2004; Vilkhu et al., 2006). Ultrasonic extraction has been shown to be highly effective on a number of products (Mason, 1998): · Extraction of sugar from sugar beet. When ultrasound was applied to sugar beet, cavitation resulted in cell disruption and subsequent release of cellular material into the bulk medium. The microstreaming effects (high velocity liquid that results from the collapse of cavitation bubbles and creates microcurrents) that occurred resulted in enhanced mass transfer. These combined effects provided a more efficient method for sugar extraction from sugar beets (unpublished results). · Extraction of rennin. Ultrasound increased both the yield and activity of rennin (the enzyme used to assist the coagulation of casein in the production of cheese) compared to conventional extraction technologies (Zayas 1986). · Extraction of protein from defatted soy beans. A continuous, particularly efficient process was developed for the sonication of soy protein using a 2.2 kW ultrasonic probe operating at a frequency of 20 kHz. This resulted in increased yields and significantly reduced process times compared to other technologies (Karki et al., 2010) · The extraction of tea solids, the starting materials for instant tea, can be improved by the use of ultrasound. Mason and Zhao (1994) reported a 20% increase in the extraction of solids from tea leaves by incorporating ultrasound at 60 ëC into the process, compared to steeping in hot water at 90± 100 ëC (same time period), which impacts adversely on some of the volatile components. The application of ultrasound also allowed a reduction in process time as the majority of the material was extracted in the first 10 min of sonication. High power ultrasound (HPU, discussed in this chapter) offers an innovative and new alternative to existing extraction processes and overcomes the limitations of low power ultrasound from a technical, market opportunity and economic perspective. While low power ultrasound was limited to small batch processes, HPU can be integrated into large volume continuous applications in both high value flow streams, commodity flow streams and even waste flow streams with potential payback on capital investment in less than two years. A summary of application and the potential of HPU is listed in Table 6.2. The benefits of HPU can be summarized as follows: · · · · · ·

bolt-on technology to existing extraction process nonthermal process low energy and maintenance cost increased productivity rate increased yield opportunity for aqueous extraction (instead of solvent) and therefore premium quality product · improved health and wellness.


126


Table 6.2 Examples of ultrasonic extraction systems and their benefits in terms of yield and rate Product

Extract

Grape must

Color Anthocyanin Tannins Palm oil Corn oil Citrus oil Color Coffee compounds Tea compounds Polyphenols Alpha/Beta carotene

Palm fruit Corn germ Citrus fruit Blueberry Coffee beans Tea Carrot

Yield improvement (%) 10±40 10±20 5±15 5±10 ± 10±30 10±30 5±20 15±50 15±20 10±30

Increase in production rate (%)

600

Cavitus Pty Ltd (2007), have developed commercial extraction systems using high power ultrasound in the food and beverage industry. Patist et al. (2006), for example, showed that ultrasonics can be used to enhance the extraction of peel oil from citrus fruits. The fully commercial system uses several 16 kW systems in series to treat flow rates in excess of 36 m3/hr. A second example in the wine industry uses a 48 kW unit to treat 50 m3/hr of must for the extraction of grape color and anthocyanin during the fermentation process. More detail on the use of ultrasonics in the wine industry is provided in the next section. Ultrasonics in the wine industry The extraction of color and flavor from grapes is an important process which determines the final composition of the wine. It is generally accepted that an increase in grape and wine color density is correlated with an increase in aroma intensity and wine quality. Red color and flavor compounds are located in the cells of grapes skins. The release of color and flavor is facilitated by mechanical action (crushing, pressing), death of tissues and cells in the absence of oxygen, heat and temperature and presence of alcohol. The addition of pectinolytic enzymes during cold soaking (cold maceration) also helps to release color. According to Peynaud (1981), the maximum extraction of anthocyanins present in the grape in the best of conditions is 30% of the 200±500 mg/L of the anthocyanins present in young red wines, with significant variation between varieties. Growing conditions are also responsible for color differences, and grapes from warm-to-hot regions generally contain less anthocyanin pigments. During vinification, color diminishes because anthocyanins become adsorbed on to seeds, skins, tannins, stems, and yeast lees. Color is also lost during cold stabilization and/or barrel maturation. Color stability decreases over an extended period as anthocyanin molecules polymerize with themselves and with other phenolic compounds to form insoluble precipitates.


Industrial applications of high power ultrasonics

127

Preliminary studies by Cavitus (2007) on peeled skins of Pinot Noir and an American red table grape, and on Cabernet Sauvignon must showed a significant increase in anthocyanins and red color density following ultrasonic treatment. The following study is an extension of the preliminary trials, in order to obtain more precise analytical data on the effects of HPU for anthocyanin retention and red color density. During one test, Cabernet Sauvignon grapes from Yalumba's Wrattonbully vineyards were crushed and inoculated with an active dry wine yeast. The must was then divided into 20 L lots in plastic fermenters with air-locks and kept in temperature-controlled room. HPU treatment of individual lots were carried out by passing the must through a flow cell at flow rates of 7 or 25 L/min and coupled to a 2 kW ultrasonic unit. Variations in amplitude of the ultrasound (25, 50 or 100%) were also applied. Six lots of musts were treated ultrasonically on the first day (Day 1) only, and one lot was treated once daily on Days 1±4 of the study. The caps were gently plunged twice daily. Fermentation was complete by Day 11. The young wines are presently undergoing further treatment until they are bottled. Musts, fermenting musts, and wines were sent to The Australian Wine Research Institute for color profiling and chemical analyses. Sensory and color analyses of the bottled wines will be carried out immediately after bottling and at 3-monthly intervals. The average gain in anthocyanin concentration following treatment of the musts with HPU immediately after crushing was 27% over the untreated (control) wine. The overall gain in color density over the control for the same wines ranged from 23±32% (average 29%). The results of the trial clearly demonstrate that HPU can significantly improve the extraction of anthocyanins from red musts, as well as improve red color density. The data presented in Figs 6.3 and 6.4 relate to a must (Lot C1 and C2) treated with HPU on the first day (Day 1) at a flow rate of 25 L/min (1500 L/ hour) and an amplitude of 50%. Figure 6.3 shows the measured changes in anthocyanin concentrations over time of the untreated (control) must and must treated with HPU. Anthocyanin extraction occurred rapidly over the first 4 days for the untreated and treated musts. The increase in anthocyanin extraction by the treated must versus the control ranged from 18% on Day 1 to 50% on Day 2 and 25% on Day 13. Figure 6.4 shows the changes in red color density of the untreated and treated musts over time. Red color density increased rapidly over the first 4 days of fermentation, and by Day 13 the young wine showed an overall 19% gain in color density compared to the untreated. The study shows that HPU technology can serve as a powerful tool to enhance the extraction of color and flavor from must and can bring about marked benefits for process cost efficiency and wine quality, as options available to the winemaker to achieve the color intensity required in the final wine, such as cold soaking (cold maceration), warm to hot fermentation/ maceration, extended/post-fermentation maceration and enzyme additions to the unfermented musts are either time-consuming and/or costly, and may even lower the quality of the wine.


128


Fig. 6.3

Changes in anthocyanin concentrations in the control and HPU-treated musts over time, and the percent gain by the treated must over the control.

Fig. 6.4

Changes in red color density of the control and HPU-treated musts over time, and the percent gain in color by the treated must over the control.


Industrial applications of high power ultrasonics 129 6.4.3 Emulsification/homogenization The shock wave resulting from a collapsing cavitation bubble provides enough energy in terms of shear for efficiently mixing of two immiscible liquids. Relatively low energy input can result in the formation of very fine, highly stable emulsions (Canselier et al., 2002; Freitas et al., 2006). This prospect is currently being developed in-line for food products such as fruit juices, mayonnaise and tomato ketchup (Wu et al., 2000). Little, if any, additional emulsifier is required to maintain the stability of the system. For applications such as mayonnaise, an excellent white color is produced, which reflects the narrow dispersion of small particle sizes (unpublished results) One benefit of the ultrasonic emulsification process is that it can be installed in-line within the existing plant. In one particular commercial application the traditional homogenizer was replaced by ultrasonics, which allowed for a 50% reduction in emulsifier (in this case the most expensive ingredient). As a bonus the shelf life was extended by several months (unpublished results). 6.4.4 Viscosity alteration Many food systems exhibit complex flow behavior and the viscosity is often determined by multiple factors such as pH, molecular weight of the protein, pectin or polysaccharide, hydrogen bonding, and other inter- and intramolecular forces. The reduction and control of the viscosity of food and beverage products has usually been based on the use of a combination of chemical modifiers or heat. Ultrasound can be applied to either increase or decrease the viscosity, and, dependent on the intensity, temporary (lasting form minutes to hours) or permanent. In the case of thixotropic fluids, cavitation causes shear that temporarily causes a decrease in viscosity. However, if enough energy is applied, the molecular weight may be decreased, causing a permanent viscosity reduction (Seshadri et al., 2003). Conversely, Bates et al. (2006) showed that ultrasound allowed for better penetration of moisture into the fibre network of tomato pureÂe and caused an increase in the viscosity. This application was commercialized on a full industrial scale, and it is unique in that it alters product functionality without reformulation. Temporary reductions in the viscosity of food and beverage flow streams have been found to be beneficial for improving the efficiencies and performance of downstream processing technologies, including the following: · · · · · · · · · ·

increased flux rate during membrane and other filtration methods higher DS in spray drying homogenizing heat transfer kinetics in retorting reduced fouling in evaporators and heat exchangers cooling transfer kinetics in cooling tunnels enhanced performance during packaging and filling operations improved de-aeration during packaging operations improved pasteurization and food safety benefits spray coating.


130


Table 6.3 benefit

Examples of ultrasonic viscosity reduction applications and its processing

Product

Downstream processing improvement

Grape concentrate Dairy Yeast biomass Canned soups Food coatings

Membrane filtration Spray drying Spray drying Retorting Spray coating

Confectionery products Beverages

Depositing into moulds Packaging and filling operations

Chocolate Soya protein isolate

Chocolate production Spray drying

Viscosity reduction (%) 50 30 30 40 70

30±60 50 5 60

Benefit

Flux rate doubled Increased DS Increased DS Reduced retort time Ability to spray coat, smaller droplet size, improved uniformity, reduced coating use Increased brix, reduced drying time, reduced waste Improved filling level control, filling line speed increase, reduced aeration Reduced cocoa butter Increased DS

Table 6.3 shows a list of ultrasonic viscosity reduction applications and their benefit. 6.4.5 De-foaming A foam is the dispersion of a gas in a liquid, with the density approaching that of the gas. However, its mechanical behavior can be similar to that of a solid, depending on the type of foam. In food manufacturing, problems with foaming can result in: · · · · · · ·

reduced working capacity of vessels product quality problems, sometimes leading to product losses production delays and even shutdowns obstruction of ducts and exhaust valves wetting of outlet air filters reduced cleaning efficiencies malfunction of control instruments.

Foam is most often controlled by the use of mechanical breakers or by the addition of chemical anti-foaming agents. With the use of chemicals becoming more restricted and mechanical breakers not always being effective, HPU may provide an alternative solution. High power ultrasonics can be used to break foam through a combination of fluctuating high pressures, bubble resonance, cavitation, radiation pressure and sonic wind (Gallego-JuaÂrez, 1998; Morey et al., 1999). For this application, a compact, stepped plate, high energy transducer is used to transmit airborne ultrasound. The transducer has no


Industrial applications of high power ultrasonics 131 moving parts, no airflow, does not interfere with the process, is readily sterilizable, and can be easily installed onto existing process lines. It has been successfully applied to control of foam excess produced on high speed canning lines and in the dissipation of foams in fermenters (see Fig. 6.5). In canning lines, the airborne ultrasonic radiation is focused on the work area to quickly dissipate the foam and avoid liquid losses. The new system was successfully applied to control foam in the filling operation of cans with a well-known commercial beverage company at a speed of more than 20 cans per second. Two focused transducers working at 20 kHz were used in parallel for efficient and quick operation, in order to widely cover the can surface with high intensity sound pressure levels (165 dB). The power applied to each transducer was only 150 W. The energy consumption was only 5 mWh/can. De-foaming systems have also been developed for the treatment of foams in reactors. This technique was developed in a beer fermenter ± the rate of foam breaking was 200 L/min with an input power on the transducer of 300 W and energy consumption of about 30 Wh/m3.

6.4.6 Ultrasonic cleaning and sanitation in the wine industry Barrels are the highest cost element in wine making, excluding the cost of grape production. There are more than seven million barrels in wineries around the world. Barrels used for aging red wine can be re-used at least once, but, for a variety of reasons, they are usually discarded after about four years. Some used wine barrels are purchased by other wineries and are re-used to age lower quality wine, and others are re-used to age whiskey. There is also a small market for used barrels as garden decorations or planters. Used barrels become contaminated with spoilage microbes, such as Lactobacillus and Brettanmyces/Dekkera, which are capable of infiltrating 8± 10 mm into the wood ± the same depth that the wine can penetrate ± making the common practice of scraping the surface to expose fresh wood ineffective for disinfection purposes. Further complicating this challenge is the precipitation of tartrates as wine matures. Tartrate deposits are difficult to re-dissolve and block the pores of the wood. It is essential during the aging process for a small amount of ambient oxygen to diffuse through the barrel and into the wine, while some alcohol diffuses out of the wine and evaporates. (The loss of alcohol is greater from more potent spirits, such as brandy and whiskey.) This loss of alcohol is commonly referred to, and not necessarily fondly, by the owners of the diminishing asset, as `the angels' share. Some common attempts at barrel cleaning and sanitation include: · · · · ·

low/high pressure cold and hot water chemicals (caustic, citric acid, sulfur dioxide, ozone) shaving the wood dry ice blasting microwaves.



Schematic representation of Cavitus' airborne ultrasonic defoaming system in a tank and on top of a canning line.

Industrial applications of high power ultrasonics 133 The common practices of cleaning/disinfecting barrels are not completely effective. Steam can disinfect barrels, but only for about 2 mm into the wood. Many chemicals require 24±48 h of contact time to be effective. Sulfur dioxide is often added to a closed barrel, if it is not to be filled immediately, and it reduces contaminating yeasts, but not microbes that are protected in the wood. Sulfur dioxide has another beneficial effect by reacting with compounds in the wood and wood char (barrels are normally lightly charred before use) to form desirable flavor compounds that are extracted into the wine as it ages. Ozone appears to be an effective anti-microbial, but requires good cleaning and some winemakers are concerned that reaction products of ozone with the wood may contribute unwanted off-flavors to the wine. Shaving and re-charring the interior surfaces of the barrels can only be done a few times and removes 3±5 mm of wood surface without completely sterilizing the barrel. Dry ice blasting can remove surface contamination but does not affect embedded yeasts. A French company uses high pressure water, alkaline cleaners, acidified solutions and microwaving to clean and disinfect barrels for the secondary market. Cavitus (2007) have studied the effects of ultrasound on staves from used barrels and were able to demonstrate that a few minutes of sonication in water with 400 W or 1 kW were able to remove tartrate deposits. Jiranek et al. (2008) recently discussed the use of ultrasonics in managing wine microbiology, with ultrasonics proving to be an effective tool in killing spoilage microorganisms located deep in the pores of the wood. In separate studies (Yap et al., 2007) demonstrated that sonication killed up to 99.9% of the cells of suspensions of Brettanomyces/Dekkera. Cavitus (2007) has launched a patented ultrasonic barrel cleaning and disinfection system for the wine industry following 18 months of in-house development and independent pilot trials. The Cavitus cleaning system is fully automated and designed to retrofit existing wine cleaning operations. Used wine barrels are filled with water at a pre-determined optimum temperature, and a sonotrode inserted through a bung hole and sonic energy is applied for a few minutes. Afterwards, the barrel is emptied, maybe filled with sulfur dioxide, and sealed until needed. The cleaning result can be seen in Fig. 6.6, and the amount of microbial kill as a function of treatment time is shown in Fig. 6.7.

6.5

Large-scale implementation

HPU is rapidly becoming a significant food-processing technology with the capability for large commercial scale-up and good return on capital investment for the following reasons: 1. The commercialization of HPU equipment has focused on the design of large continuous-flow treatment chambers (flow cells) that reduce the cost per volume of treated material. A typical large flow cell module provides 2± 16 kW of power, with amplitudes ranging from 1 to 150 micron peak-to-peak displacement, and flows ranging from 1 to 1000 L/min, depending on the


134


Fig. 6.6 Oak wine barrel after traditional high pressure water cleaning (left) compared to ultrasonic cleaning (right) using Cavitus' proprietary ultrasonic cleaning system.

2.

3.

4.

5.

application. Larger flow rates would require multiple systems in series or parallel (see Fig. 6.1). The efficiency of ultrasonic generators and transducers has improved over the years, thereby reducing internal heating (and the need for subsequent expensive cooling systems), which often cause system failures. Current systems have energy efficiency around 90±95%, which means simply that most of the power sent to the transducer is actually transferred to the medium. The technology has been designed and engineered for easy installation as either a stand-alone system or bolt-on attachment to an existing process, without requiring any major modifications to the plant operation. The units are compact and occupy a small footprint in a processing plant. If necessary, soundproof cabinets are available to reduce the noise generated by the cavitation. Although the technology has been termed HPU, the energy consumption is generally very competitive with other types of food processing technologies. Depending on the application, the amount of energy per liter material treated (defined in units of kWh/L, see Section 6.3.1) required is comparable to other units operating in the industry (for example, homogenization, milling, heat shock, etc.). One of the main benefits of ultrasonic technology is the absence of moving parts and therefore low maintenance costs. The lack of rotors, seals, grease, etc., makes these systems particular robust. The only part which requires



Fig. 6.7

Microbiological reduction of Dekkera/Brettanomyces by ultrasonics (HPU) and conventional method (high pressure/hot water sprays).

replacement is the sonotrode (probe) which is in direct contact with the medium. Depending on the amplitude and the abrasiveness of the medium, the lifetime of a sonotrode ranges from 3 to 24 months. Examples of several commercial applications and their business case are represented in Table 6.4. The payback (defined here as investment cost over the benefit) occurs in generally less than one year. Corporations generally use more sophisticated tools, such as net present value (NPV), internal rate of return (IRR), and return on investment (ROI) to evaluate the business case (Brealey et al., 2006). Table 6.4 Business case examples of commercialized ultrasonic applications (due to confidentiality reasons, the application details are generalized) Application

Description

Extraction Emulsification

Yield increase Reformulation and improved shelf-life Increased production capacity and reduced energy Increased production capacity in tanks Enhanced bottling production and reduced waste

Viscosity reduction De-foaming De-foaming

Flow rate (m3/hr)

Power (kW)

Benefit (k$/yr)

Payback time

50 8

48 32

7000 500

< 2 years 1 year

70

48

600

< 2 years

0.3

2000

< 1 year

0.6

500

< 2 years

10 1200 cans/min


136


6.6

Roadmap to successful commercialization

Based on the authors' experience, the following steps provide a `roadmap' to successful commercialization of HPU technology: 1. Ultrasonic technology has to show potential payback (return on investment) when first considered. 2. Ensure a good understanding of economics (total cost, payback, etc.) compared to current practices and all the potential alternatives. 3. Build the right project team (who is responsible for what, e.g., sponsor vs. stakeholder). 4. Develop a project charter, which includes objectives, budget, timelines and resources required. This helps manage expectations and ensures that senior management understands what it takes to commercialize the technology. A good approach is the so-called Stage-GateTM process (Cooper, 2001), which focuses on doing projects right and doing the right projects according to a staged project management process from ìdeation' to `launch' (note that Stage-GateTM can be used for new products and processes). 5. Perform lab-scale tests. The goal is to prove the concept, which can be completed in as short as a few days. 6. Perform pilot testing. This involves continued treatment of a bypass (slip) stream of the full-scale operation to optimize parameters such as back pressure, probe and flow cell design, and energy intensity. This may take 1±2 weeks, after which economics and payback can be refined. During this stage it is important to consider alternative technology options. 7. Integrated pilot test. This step includes a long-term test (2±6 months) during which wear and tear and variability in the feed stock are evaluated. The deliverable of this step is a solid understanding of the overall economics to justify a full-scale roll-out. 8. Commercialization. This includes the full commercial installation and capturing of lessons learned.

6.7

Conclusion

Over the last 10±15 years, HPU has grown from a laboratory-based prototype technology into fully operational commercial processes throughout the world. Owing to the improved efficiency of the equipment itself, its scalability, ease to retrofit, and low maintenance costs due to its design having few moving parts, the payback is usually less than two years. This implies that while the technology holds great promise, the drawback is that it will have to be carefully developed and customized uniquely for every single application. In this chapter, we have presented examples in which the application of ultrasonics fits niches and provides unique value compared to alternative technologies often considered conventional.



6.8

References

and BRIDGES M W (2006), `Method of treatment of vegetable matter with ultrasonic energy', US patent application 20060110503. BREALEY R A, MYERS S C and ALLEN, F (2006), Principles of Corporate Finance, 8th edn, New York: McGraw-Hill. CANSELIER J P, DELMAS H, WILHELM A M and ABISMAIL B (2002), Ùltrasound emulsification ± an overview', J. Dispersion Science and Technology, 23, 333±349. CAVITUS PTY LTD (2007), Àpplying high power ultrasonics to food and beverage processing', Adelaide, Australia, www.cavitus.com, Australian patent application AU2007001958. COOPER R (2001), `Winning at new products', in The New Product Process: The StageGateTM Game Plan, 3rd edn. Cambridge, MA: Perseus Publishing, pp. 113±153. FREITAS S, HIELSCHER G, MERKLE H P and GANDER B. (2006), `Continuous contact and contamination free ultrasonic emulsification ± a useful tool for pharmacuetical development and production', Ultrasonics Sonochemistry, 13, 76±85. Â REZ J A (1998), `Some applications of air-borne power ultrasound to food GALLEGO-JUA processing', in Povey M J W and Mason T J (eds), Ultrasound in Food Processing. London: Blackie Academic & Professional, pp. 127±143. HIELSCHER T (2005), Ùltrasonic production of nano-size dispersions and emulsions', paper presented at 1st Workshop on Nano Technology Transfer, ENS Paris, 14±16 December, Paris, France. JIRANEK V, GRBIN P, YAP A, BARNES M and BATES D (2008), `High power ultrasonics as a novel tool offering new opportunities for managing wine microbiology', Biotechnol Lett, 30 (1), 1±6. KARKI B, LAMSAL B P, JUNG S, VAN LEEUWEN J, POMETTO A L, GREWELL D and KHANAL S K (2010), Ènhancing protein and sugar release from defatted soy flakes using ultrasound technology', J. of Food Engineering, 96, 270±278. KNORR D (2003), Ìmpact of non-thermal processing on plant metabolites' J. of Food Engineering, 56, 131±134. LABORDE J L, BOUYER C, CALTAGIRONE J P and GERARD A (1998), Àcoustic bubble cavitation at low frequencies', Ultrasonics, 36, 589±594. LEIGHTON T G (1994), The Acoustic Bubble, San Diego, CA: Academic Press. LI H, PORDESIMO L and WEISS J (2004), `High intensity ultrasound assisted extraction of oil from soybeans', Food Research International, 37, 731±738. LORIMER J P and MASON T J (1987), `Sonochemistry Part 1. The physical aspects', Chem. Soc. Rev., 16, 239±274. MASON T J (1998), `Power ultrasound in food processing ± the way forward', in Povey M J W and Mason T J (eds), Ultrasound in Food Processing. London: Blackie Academic & Professional, pp. 103±126. MASON T J and ZHAO Y (1994), Ènhanced extraction of tea solids using ultrasound', Ultrasonics, 32, 375±377. MOHOLKAR V S, REKVELD S and WARMOESKERKEN M M C G (2000), `Modeling of the acoustic pressure fields and the distribution of the cavitation phenomena in a dual frequency sonic processor', Ultrasonics, 38, 666±670. MOREY M D, DESHPANDE N S and BARIGOU M (1999), `Foam destabilization by mechanical and ultrasonic vibrations', J. Coll. and Interface Sci., 219, 90±98. MUTHUKUMARAN S, KENTISH S E, STEVENS G W and ASHOKKUMAR M (2006), Àpplication of ultrasound in membrane separation processes: a review', Rev. Chem. Eng., 22, BATES D M, BAGNALL W A


138


155±194. and BATES D (2008), 'Ultrasonic innovations in the food industry: from the laboratory to commercial production', Innovative Food Science and Emerging Technologies, 9, 147±154 PATIST A, MINDAYE T T and MATHIESSEN T (2006), `Process and apparatus for enhancing peel oil extraction', US patent application 20060204624. PEYNAUD E (1981), Knowing and Making Wine, New York: John Wiley & Sons, pp. 35± 52. SESHADRI R, WEISS J, HULBERT G J and MOUNT J (2003), Ùltrasonic processing influences rheological and optical properties of high methoxyl pectin dispersions', Food Hydrocolloids, 17, 191±197. SUSLICK K S (1988), `Homogeneous sonochemistry', in Suslick K S (ed.), Ultrasound: Its Chemical, Physical, and Biological Effects. New York: VCH Publishers, pp. 123± 163. VILKHU K, MAWSON R, SIMONS S and BATES D (2006), Àpplications and opportunities for ultrasound assisted extraction in the food industry ± a review', Innovative Food Science and Emerging Technologies, 9, 161±169. VINATORU M (2001), Àn overview of the ultrasonically assisted extraction of bioactive principles from herbs', Ultrasonics Sonochemistry, 8, 303±313. WU H, HULBERT G J and MOUNT J R (2000), Èffects of ultrasound on milk homogenization and fermentation with yogurt starter', Innovative Food Science & Emerging Technologies, 1, 211±218. YAP A, JIRANEK V, GRBIN P, BARNES M and BATES D (2007), `Studies on the application of high power ultrasonics for barrel and plank cleaning and disinfection' The Australian and New Zealand Wine Industry Journal, 22(3), 95±104. ZAYAS J F (1986), Èffect of ultrasonic treatment on the extraction of chymosin', J. of Dairy Science, 69(7), 1767±1775. ZHANG R, XU Y and SHI Y (2003), `The extracting technology of flavonoids compounds', Food and Machinery, 1, 21±22. PATIST A


7 The potential of novel infrared food processing technologies: case studies of those developed at the USDA-ARS Western Region Research Center and the University of California-Davis Z. Pan and G. G. Atungulu, University of California-Davis, USA

Abstract: Infrared (IR) radiation heating has been considered as an alternative to current food and agricultural processing methods for improving product quality and safety, increasing energy and processing efficiency, and reducing water and chemical usage. As part of the electromagnetic spectrum, IR has the capacity to provide high heating and heat transfer rates. This chapter reports several IR-based processing technologies that have recently been developed to take advantages of IR for the blanching and dehydration of fruits and vegetables, roasting and pasteurization of almonds, disinfestation and drying of rice, and peeling of tomatoes. The development and commercialization of IR-based food processing technologies could open new avenues to delivering safe and value-added foods desirable to consumers, while reducing the consumption of natural resources during processing. Key words: infrared heating, food processing, drying, blanching, pasteurization, safety, roasting, peeling, emerging technologies, commercialization.

7.1

Introduction

One of the primary objectives of the food industry involves the transformation of raw agricultural materials by a series of operations into foods suitable for consumption. Processing, in general, has become more sophisticated and diverse


140


in response to the growing consumer demand for improved food quality while ensuring food safety. Consumer expectations of convenience, variety, adequate shelf-life, nutritive content, reasonable cost, and environmental soundness have required modifications to existing food processing practices, including the adoption of novel processing technologies. This chapter covers the use of infrared (IR) technology for food processing involving novel engineering approaches and applications of IR radiation heating to process food and agricultural products to meet consumer needs. Case studies which target specific concerns of the food processing industry are addressed that emphasize technologies such as simultaneous IR blanching and dehydration (SIRBD) of fruits and vegetables; sequential IR freeze-drying (SIRFD); sequential IR hot air (SIRHA) roasting and pasteurization of raw almonds; simultaneous IR heating of rough rice for drying and disinfestation applications; and IR radiation heating for tomato-peeling. The specific merits and economic benefits of the above-mentioned IR technologies which employed catalytic IR (CIR) emitters are described. This fundamental information and significant database of research references on IR application for food and agricultural processing will provide value and impact to food process engineers, food processing companies, education and research institutes, and quality control and safety managers in food processing and food manufacturing operations.

7.2

Effect of infrared (IR) on food molecular constituents

When IR radiation impinges upon a food surface, the IR energy is absorbed at discrete frequencies corresponding to intra-molecular transitions between energy levels according to the nature of the chemical bonds present. The wavelengths of IR fall in the spectrum of 0.76±1000 m and can be typically categorized into near infrared (NIR) (0.76±2 m), medium infrared (MIR) (2± 4 m) and far infrared (FIR) (4±1000 m). Foodstuffs absorb MIR and FIR energy in the range 2.5±100 m (Rosenthal, 1992; Sakai and Hanazawa, 1994) most efficiently through stretching modes of vibrations, which leads to the radiative heating process. For agricultural and food product processing, the high temperatures corresponding to NIR radiation could cause product discoloration and quality deterioration, and temperature needs to be carefully controlled when NIR is used. FIR is associated with low temperature and energy emission. If temperature is too low, the energy emitted may not be enough to meet the energy requirements of food processing. Useful temperature of IR radiation may be in the range of 150±2200 ëC which corresponds to the IR peak wavelengths of 7±1.2 m. The IR absorption band characteristics of chemical groups relevant to the heating of foods are summarized in Table 7.1 (Rosenthal, 1992). Even though complete information is not available, the approximate values for the strong absorption bands of major food constituents are proteins at 3±4 m and 6±9 m;


The potential of novel infrared food processing technologies

141

Table 7.1 The infrared absorption band characteristics of chemical groups relevant to the heating of food Chemical group

Absorption wavelength (m)

Hydroxyl group (OÐH) Aliphatic carbon-hydrogen bond Carbonyl group (C=O) (ester) Carbonyl group (C=O) (amide) Nitrogen-hydrogen group (ÐNHÐ) Carbon-carbon double bond (C=C)

2.7±3.3 3.25±3.7 5.71±5.76 5.92 2.83±3.33 4.44±4.76

Relevant food component Water, carbohydrates Fats, carbohydrates, proteins Fats Proteins Proteins Unsaturated fats

lipids at 3±4 m, 6 m, and 9±10 m; and sugars at 3 m and 7±10 m. The four principal absorption bands of liquid water are 3, 4.7, 6, and 15.3 m (Sandu, 1986). Superimposing the absorption bands of the principal food constituents and those of liquid water shows significant overlap in the absorption spectra of these food components (Sandu, 1986), and it remains a challenge for practical applications to use differential or selective heating efficiently for targeting water without heating other molecular components in a food material. The absorption properties of foodstuffs depend mainly on three factors: water content, thickness, and physicochemical nature of the product. Most foodstuffs show a high transmissivity at wavelengths less than 2.5 m (Sandu, 1986). Studies also showed that the transmissivity of foodstuffs in the NIR range increases abruptly when water content is lowered, while fresh and dry apples showed similar spectral absorptivity at wavelengths above 3 m (Krust et al., 1962; Ginzberg, 1969). Ginzberg (1969) also showed that as the thickness of foodstuff increases, a simultaneous decrease in transmissivity and increase in absorptivity occur. For different applications, the optimal thickness of foodstuff and the selected radiation wavelength could be different based on varying transmissivity and absorptivity. Thin slices could be preferred for processes using IR energy, since the high transmissivity would result in higher heating rates. On the other hand, NIR has advantages over MIR and FIR with its larger transmissivity. However, the temperature of NIR radiation could be too high for processing certain food and agricultural products to maintain the high quality of these products. In addition, the decrease in absorptivity and increase in transmissivity of NIR during drying could also be a problem to thin materials. As materials dry, the shrinkage of thin materials can result in low absorption of NIR energy, since most of the radiation energy can be reflected and transmitted through the thin layer.

7.3 Case studies in novel infrared (IR) technologies for improved processing efficiency and food safety The following case studies introduce our novel IR technologies for improved food processing efficiency and safety using catalytic IR (CIR) emitters. In the


142


CIR emitters, natural gas or propane combines with air across a platinum catalyst and reacts by oxidation-reduction to yield a controlled bandwidth of IR energy and small amounts of CO2 and water vapor. Because the CIR energy is generated without the use of flames, the process is safe and hazard-free. The bandwidth of radiant energy is in the FIR range of approximately 3±7 m, which is in the range that water absorbs energy very efficiently (3, 4.5, and 6 m). IR also has certain penetration capabilities that facilitate the fast heating of food materials. To take the advantage of the high heating rate of IR, we developed several processes for different applications, including drying, desinfestation, blanching, roasting, and peeling.

7.4 Simultaneous infrared blanching and dehydration (SIRBD) In the food industry, blanching has become a very important unit operation step to inactivate enzymes, modify food texture, preserve food color, flavor, and nutritional value, and to remove trapped air prior to freezing, canning, and drying of fruits or vegetables. Recent requirements for energy conservation and waste reduction have motivated the need to improve the design of blanching equipment. Pan and McHugh (2004) developed a new method that uses IR radiation energy to simultaneously perform dry blanching and dehydration of fruits and vegetables. This ìnfrared dry-blanching' (IRDB) technology is intended to replace current steam, water, and/or microwave blanching methods, to produce many kinds of value-added dried, refrigerated, frozen, and dehydrofrozen fruit- or vegetable-based products. The merits of IRDB technology and equipment include: 1. Uniform heating to enhance energy efficiency and limit the product damage from over-heating. 2. Capability of zone heating to address differential density. 3. Ability to treat large or small lots with the same piece of equipment. 4. A safe process with no harmful side-effects to humans or the environment. 5. Portability. The following case study illustrates the performance of a new industrial scale IR heating system which has been built and tested for IR dry-blanching (IRDB) and simultaneous infrared dry-blanching and dehydration (SIRDBD) of fruits and vegetables. 7.4.1 Equipment The design of the mobile IR unit built to demonstrate on an industrial scale the efficacy of the continuous IR heating system to accomplish IRDB and SIRDBD of various fruits and vegetables is illustrated in detail in Figs 7.1 and 7.2 (courtesy of Catalytic Industry Group Inc., Independence, KS, USA). The newly developed



Fig. 7.1

143

Technical drawing of the IR equipment (side and top views).

pilot-scale mobile IR heating unit is equipped with an automatically controlled variable speed conveyor belt and catalytic IR emitters powered with natural gas (Fig. 7.3) and can be used for processing various vegetables and fruits. The equipment has an effective heating area of 1:5 4:6 m (the height, overall length and width of the new mobile IR equipment are 2, 6 and 2 m,


144


Fig. 7.2

Fig. 7.3

Technical drawing of the IR equipment (rear view).

Mobile infrared heating equipment for processing vegetables and fruits.

respectively and overall width of the equipment, including the control panel, is 2.4 m) and weighs approximately 2000 kg. The unit consists of 8 emitters in 4 imaginary zones as shown in Fig. 7.4. Each zone is equipped with 2 emitters. Zones 1 and 3 together and zones 2 and 4 together are hereinafter referred to as section 1 and 2, respectively. Each emitter has a dimension of 0:6 1:5 m. Emitters can be positioned at different distances from the belt to vary the IR intensity. The angular alignment of the emitters was for achieving optimized performance of catalytic IR emitters. Unless otherwise stated, during the equipment testing the distance between the emitters and the belt was 0.08 m at the lower end and 0.13 m at the higher end. Whenever necessary to provide very



Fig. 7.4

145

Side view of part of IR equipment with imaginary zones and imaginary sections.

high heat flux to the product, the distance between the emitters and the product can be adjusted to less than 0.05 m. The intensity of the IR reaching the products could be adjusted by varying the gas supply. In this equipment, the highest IR intensity is achieved by setting the gas supply valves fully open (100%), which provides energy of 607 714 kJ/h for its eight emitters. Similarly, the lowest IR intensity can be achieved by setting the values at the lowest position (0%, actually corresponds to 50% gas flow to the emitters) which provides energy of 303 857 kJ/h for the eight emitters. Adjusting the setting from 0 to 100% at the panel changes the actual gas flow from 50 to 100%, respectively. In the current study, we reported the actual gas flow rates rather than the settings at the panel. The unit had 4 type-J thermocouples (2 on top of zone 1 and 2) for monitoring the air temperature within the unit itself. The IR emitters were only run on natural gas. However, they can be changed to operate with propane. A fan was installed on top of the unit and was used to remove air with high moisture from the heating chamber, when necessary. It is recommended that the fan should be turned off during blanching to obtain high relative humidity levels in the chamber and to minimize the moisture loss, when desirable. All electrical components in the IR unit operated at 208V. During the equipment tests, the speed of the belt varied from 0.270 m/min to 1.225 m/min, which corresponded to the total resident times of approximately 4± 17 min. The speed of the belt in the unit was controlled by using a variable speed motor. The relationship between frequency (Hz) of motor and actual speed of the conveyor belt was established and is given by the relationship (7.1):


146

Case studies in novel food processing technologies vb 0:1551 f 0:9996

7:1

in which vb is the speed of belt (ft/min) and f is the frequency (Hz) and the correlation coefficient r2 was 0.9996. The belt specifications were thus: flat flex 42 0:062 60} wide 25 sp, sle, ss. The stainless steel belt has a mesh opening of 2:25 0:25}. 7.4.2 Dry-blanching and dehydration of potatoes Russet potatoes were used as samples in tests conducted to study the efficacy of the new equipment for dry-blanching and dehydration. The potatoes were purchased from a local grocery store and sliced into three different thicknesses (2:89 0:34, 6:42 0:36, and 9:03 0:32 mm). In another dry-blanching and dehydration test, diced potatoes with dimensions of 9:5 8:3 mm were also used. The following procedure was adopted to measure the moisture content of the potato samples. Frozen potatoes were brought from the processing facility and allowed to thaw in plastic pouches. The entire content of the plastic pouches was emptied into a blender and ground. At least 10 g of ground sample were placed in metallic dishes and dried in a vacuum oven at 70 ëC for 24 h. Experiments were done in triplicate and average values were reported. Initial moisture content of potatoes was 81:21 2:21% on wet basis. In order to assess the optimal processing conditions to achieve blanched products, tests under different conditions were conducted (Table 7.2). The residence time of direct exposure to IR is the time that the potato slices were passing directly through the top/bottom heating section, whereas the total residence time is the time that the potato slices remained in the IR unit. This means that in the case of tests with Section 2 off, the potato slices only experienced heating in Section 1. However, the total residence time count is based on the total time for samples to go through both Sections 1 and 2. The load rates of potato slices varied from 2.10 to 6.42 kg/m2 for the 2.89 and the 9.02 mm thick samples, respectively. The load rate is function of the slice thickness and the load area and is established as LR 0:7089 Th 0:1468

7:2

Table 7.2

Test conditions for dry-blanching and dehydration of potato slices

Test condition

Gas supply Section Section 1 2

1 2 3 4 5 6

100% 100% 100% 100% 100% 100%

OFF OFF OFF 50% OFF 50%

Conveyor speed m/min

Fan

0.5 1.0 0.7 1.1 0.8 1.2

ON OFF OFF OFF OFF OFF


Residence time (s) Exposure Total to IR 272 142 185 256 164 224

544 283 370 256 328 224


147

in which LR is the load rate (kg/m2) and Th is the product thickness (mm) and the correlation coefficient r2 was 0.9933. To determine the final temperature of the product exiting the IR dry-blancher, we measured the surface and center temperatures of the food products using a hand-held IR thermometer and type-K thermocouple, respectively. To monitor the temperature change during the entire residence time within the unit, the center and surface temperatures of slices were measured and recorded using type-K thermocouple and Omega HH147 Data Logger Thermometer (Omega, Connecticut, USA). The temperature data provided important information for avoiding over- or underheating products and improving the design of the IR heating equipment and the configuration of the emitters. We also measured enzymatic activity, product quality, and final moisture content. The procedures for testing the residual activity of polyphenol oxidase (PPO) were provided by Deb Dihel (Director R&D, Product Development, ConAgra Foods, personal communication). It is known that if PPO is not inactivated, browning reactions take place within 30±60 minutes. Thus, after IR dry-blanching, the color change of potatoes was visually examined. 7.4.3 Results of IR dry blanching and dehydration of potatoes The test results under different equipment settings and operation conditions are given in Table 7.3. In general, the weight reduction increased with a decrease of thickness which ranged from 10 to 54%. However, the surface temperature of the products increased as the product thickness increased, which could be due to cooling effect at the exiting point when the temperatures were measured. The largest weight reduction occurred for Test 1 when the fan was on and the heating time was relatively long. In the case of diced potatoes, it was noted that after 4 minutes exposure to IR radiation, the weight of diced potatoes was reduced by 51.06% and the moisture content decreased to 66.31%. In order to regain moisture lost during IR dry-blanching, the diced potatoes were dipped in water for 1 minute after blanching, and the moisture content increased to 70.2%. After 30 minutes, slight browning was observed in control samples, and no browning was observed in blanched samples (Fig. 7.5). Therefore, dipping in after IR dry blanching improved the final appearance of diced potatoes. The blanching time could be shortened significantly by optimizing the configuration of the emitters. During our experiments, we observed that the best conditions for blanching 2.89 mm thick samples were found in Test 2, as can be seen in Fig. 7.6(a). Both Test 3 and 4 had similar effects on the inactivation of PPO and were suitable for blanching the 6.42 and 9.03 mm thick samples ± the 2.89 mm samples showed charring in some regions (Figs 7.6(a) and (b). Test 3 used a much shorter heating time (185 seconds) compared to Test 4 (256 seconds), which means less energy consumption for the Test 3 conditions. It is seen in Fig. 7.6(c) that the color change in the 9.03 and 6.42 mm thick samples was significant. This could be due to the fact the temperature profiles during IR blanching of the thicker samples in Test 2 were not enough to inactivate PPO in the center of the thick slices.


148


Table 7.3 Final moisture content, percentage weight reduction and surface temperature of potato slices after infrared dry-blanching Test condition*

Slice thickness (mm)

Final moisture content (%MC)

Weight reduction (%)

Temperature (ëC)

Test 1

2.89 0.34 6.42 0.36 9.03 0.32

59.52 74.81 77.00

53.57 25.38 18.29

52.8 5.1 55.8 2.0 57.2 1.5

Test 2

2.89 0.34 6.42 0.36 9.03 0.32

73.29 78.17 79.34

29.63 13.91 9.04

48.5 3.3 51.7 0.7 56.6 2.7

Test 3

2.89 0.34 6.42 0.36 9.03 0.32

62.42 76.57 78.31

50.00 19.79 13.38

49.7 2.3 51.1 2.1 52.3 1.0

Test 4

2.89 0.34 6.42 0.36 9.03 0.32

69.93 76.56 77.67

37.50 19.83 15.85

93.8 3.1 90.6 3.0 90.6 2.5

Test 5

2.89 0.34 6.42 0.36 9.03 0.32

64.85 77.65 79.05

46.54 15.93 10.29

42.1 1.8 53.0 3.1 53.7 2.6

Test 6

2.89 0.34 6.42 0.36 9.03 0.32

66.97 76.75 78.42

43.10 19.20 12.94

99.2 0.2 92.9 6.0 78.9 1.4

* Table 7.2 summarizes the test conditions for dry-blanching and dehydrating potato slices

Fig. 7.5 Diced potatoes 30 minutes after dry-blanching.



149

Fig. 7.6 Images of potato slices of different thicknesses: (a) 30 minutes after blanching (Test 4); (b) 45 minutes after IR blanching (Test 3); (c) 1 hour after blanching (Test 2).


150


Fig. 7.6 Continued

However, for the 2.89 mm thick samples, it can be assumed that PPO was inactivated, since no browning reactions occurred. The heating profile of the 6.42 and 9.03 mm thick potato slices were measured during the IR dry-blanching process. Figure 7.7 shows the temperature profile and the percentage of remaining PPO activity for the 6.42 mm thick slices at 0.8 m/min belt speed and emitters on OFF mode in Section 2 (Test 5). During IR heating (Test 5 and 6 conditions were used for the 6.42 mm slices, and Test 3 and 6 conditions were used for the 9.02 mm slices), the surface temperature of the potato slices rose more rapidly than the center temperature. However, after exiting the region with the first two emitters (after 50 seconds in Fig. 7.7(a)) of zones 1 and 3, the surface temperature decreased due to evaporative cooling and the center temperature remained almost constant. Upon entering the region with the second two emitters in Section 1 (zone 1 and 3) the surface and center temperatures of potato slices began to rise again, albeit not as significantly as in the case of the region with the first two emitters. The temperature of the slices in Section 2 started to decrease gradually until exiting the unit. Although there was no heating in Section 2, the section still prevented the product from rapid cooling. To estimate the PPO activity in samples, the decimal reduction time (Dvalue) of 5 min at 65 ëC with a z-value of 8 ëC was used (Anthon and Barret, 2002). Based on the results indicated in Fig. 7.7(b), the inactivation of PPO starts when the center temperature of the slices reaches close to 65 ëC which



151

Fig. 7.7 (a) Temperature profile and (b) percentage of remaining PPO for 6.42 mm thick slices at 0.8 m/min belt speed and emitters on OFF mode in Section 2 (Test 5).

corresponded with the location of products exiting the first top and bottom emitters in zone 1 and 3. This shows that the significant temperature increase in the first top and bottom emitters in zone 1 and 3 triggered the inactivation of PPO. In the area of the Fig. 7.7(b) labeled as number 3, the slices were traveling between the first and the second emitters in Section 1. Upon entering the second top and bottom emitters in Section 1, there was little PPO left to be inactivated. Thus, the results indicate that before exiting the second top and bottom emitters in Section 1, the PPO was completely inactivated. The same experiment was also repeated with a slightly higher residence time for potato slices having 9.03 mm thickness, and similar temperature profiles and inactivation results were obtained. For the first section, the temperature profile of Test 6 (Fig. 7.8(a)) was very similar to the Test 5 (Fig. 7.7(a)), but the temperature increase was slower because the speed of the conveyor belt was almost 1.5 times faster. In Section 2


152


Fig. 7.8 (a) Temperature profile and (b) percentage of remaining PPO for 6.42 mm thick slices at 1.2 m/min belt speed and 50% natural gas flow in Section 2 (Test 6).

of the equipment, because the emitter was set at low intensity with 10.55 kW of heat flux per emitter, the temperature of the potato slices continued to increase. However, the significant difference between the surface and center temperatures was not expected. This might have been due to the fact that the thermocouple tip protruded outside the surface of the potato slice and was exposed to IR heating directly. The PPO inactivation started when the surface temperature of slices reached 65 ëC upon exiting the second top and bottom emitters in zone 1 and 3, whereas the center temperature of the slices reached above 65 ëC upon exiting the first top and bottom emitters in zone 2 and 4 (first top and bottom emitters in Section 2). When inactivation of PPO was complete at the surface, about 30% PPO still remained at the center region. Interestingly, although the overall energy in this case was higher compared to Test 5, 100% inactivation of PPO could not be achieved. Therefore, to quickly heat up the product to temperatures for PPO inactivation, using high IR intensity in the initial stage is necessary.



153

7.4.4 Energy consideration In order to avoid unnecessary energy use and to obtain high quality of either blanched or partially or fully dehydrated products, specific adjustment of the equipment setting based on each product is necessary. To perform an energy analysis of the processing system, the following considerations were made: 1. Energy loss for the conversion of natural gas to IR radiation was based on a conversion rate of approximately 80%; 2. Heat losses by natural convection and radiation were evaluated based on the material characteristics of the equipment; and 3. The temperatures of the walls of the building surrounding the equipment were assumed to be the same with the ambient temperature (20 ëC). Energy analysis results for a continuous process (Test 2) with a total processing time of 283 s and 4 running emitters during IR dry blanching of potato are summarized in Table 7.4 (natural gas supply at Section 1 was set at 100% and at Section 2 the emitters were turned off; the conveyor speed was set at 1 m/min, the fan was set in the off mode; the exposure time to IR was 142 s, and the total product residence time in the unit was 283 s). The results indicated that the IR blanching process had relatively high energy efficiency. 7.4.5 Comments on continuous and intermittent modes of operation In general, IR equipment can be designed and operated in two different heating modes, continuous or intermittent heating. During continuous heating, the radiation intensity is maintained constant by retaining a continuous supply of natural gas to the CIR emitter. Intermittent heating is normally achieved by keeping product temperature constant by turning the natural gas or electricity Table 7.4 Summarized energy and cost information for a selected condition of new mobile IR dry blancher during the dry blanching of Russet potato slices (Test 2)a Parameter

Amountb

Power consumption Energy losses Heat loss by natural gas to IR Heat loss by natural convection Heat loss by radiation Total heat loss Operation efficiency Total natural gas consumption Cost per lb of tomato

42.2 kW 4777.3 kJ 1074.4 kJ 140 kJ 5986.7 kJ 74.9% 0.64 m3 0.9 c/lb

a

Selected test conditions for potato processing are thus: Gas supply at Section 1 equals 100% and at Section 2 is off; Conveyor speed is set at 1 m/min (3.175 ft/min); Fan is set at off mode; The exposure time to IR is 142 s and total product residence time in unit 283 s. b The calculations are based on a continuous process with the total processing time of 283 s and 4 running CIR emitters.


154


supply on and off. A photo of our lab-scale IR unit that can achieve continuous and intermittent heating with the capabilities of automatic data acquisition, controlling and recording of various operation parameters, such as gas flow rate, emitter temperature, and product temperature is shown in Fig. 7.9. For the mobile IR unit, both heating modes could be achieved, because of the arrangement of IR emitters in different sections. A continuous or intermittent mode of design and operation should be chosen based on processing needs. Generally, both heating modes have their own advantages and disadvantages. An appropriate heating mode and appropriate processing conditions need to be determined based on the application and the property of the materials. For quick heating or enzyme inactivation, continuous

Fig. 7.9 Photo of lab-scale double-sided catalytic infrared dryer/blancher.



155

heating is advantageous since it delivers a constant high energy to the surface of product. For certain fruits and vegetables, moisture removal may not be desired during blanching. In this case, quick blanching with limited moisture reduction is necessary. Continuous heating may be beneficial for such an application. However, our previous studies have shown that prolonged continuous heating can cause severe surface discoloration (Zhu and Pan, 2009; Zhu, 2007), which should be avoided. For certain applications, drying is needed after proper blanching, such as the production of dehydrofrozen products. In such cases, intermittent heating works best in the drying stage, since it tends not to cause severe surface darkening by regulating the product temperature (Sandu, 1986, Zhu et al., 2010). Advantages of intermittent heating have also been recognized in terms of energy savings and improved product quality, since the desired processing temperature can be maintained (Chua and Chou, 2003). 7.4.6 Conclusions on IR dry blanching and dehydration of potatoes IR heating can be used for achieving simultaneous dry blanching and dehydration of sliced or diced potatoes through the use of appropriate equipment settings and operating conditions. The inactivation of PPO for the 2.89 mm thick potato slices was achieved when emitters in Section 1 were on and the belt speed was 1 m/min corresponding to residence time of 283 s. However, to achieve full blanching of thicker slices (6.42 and 9.03 mm) the belt speed was lowered to 0.7 or 0.8 m/min, corresponding to residence times of 370 and 328 s, respectively. Alternatively, the belt speed can be increased to 1.1 m/min (256 s residence time) when the emitters in the second section of the equipment were run at the lowest heating level. The weight loss varied from 29.63% for the 2.89 mm thick samples to 13.38% for the 9.03 mm thick slices. Using high heat in the very first stages to heat the slices to enzyme inactivation temperatures was essential for obtaining high quality blanched product and for reducing energy consumption. The moisture loss during the blanching and dehydration can be controlled by selecting appropriate processing conditions, or by replenishing water lost by blanching by dipping in water or spraying water onto the product post-blanching. The optimized equipment settings and operational conditions need to be determined based on each specific product and the quality requirements of the final product.

7.5 Sequential infrared (IR) and freeze-drying of strawberry slices Combining IR radiation and freeze-drying (FD) sequentially is a relatively new approach that has shown great potential for industrial applications. The sequential IR radiation and freeze-drying (SIRFD) is a two-step process involving IR pre-dehydration followed by regular FD. Traditionally, hot air drying and FD have been used in the food industry to dry fruits and vegetables. However, hot air drying is a time and energy consuming process. Its low heat transfer rate to


156


the product results in low energy efficiency, and the associated lengthy process may cause undesirable changes that compromise product quality (Nowak and Lewicki, 2004). Therefore, the industry has been using FD to yield premium products, despite the increased costs incurred. Typically, FD minimizes the negative impacts of drying and could produce the highest quality food product among any drying methods. The predominant factors are: the low drying temperature that helps maintain product structure during sublimation (Singh and Heldman, 1993); the rigidity of the solid matrix structure of dried food that prevents it from collapsing; the porous structure of the product that facilitates rapid and almost complete rehydration when water is added at a later time (Mujumdar, 1995); the decreased shrinkage of foods (Shishehgarha et al., 2002); the retention of aroma, flavor, and nutrients in the finished product; the crispy texture of certain products which is desirable for many food applications; and the ability to store the product at ambient temperature (Baker, 1997). However, FD is an expensive process for dehydration of foods because of the high capital and operating costs (energy-intensive), and the lengthy time required (i.e., slow drying rate). On an industrial scale, the operating cost of FD processes is on the order of 4±5 times higher than that of the spray-drying technique, and 8±10 times higher than that of the single-stage evaporator (Flink, 1977). Accordingly, FD is usually used only for high value products whose market value can justify the high manufacturing costs. Combining IR and FD to accomplish quick moisture removal while maintaining the product quality has gained interest, especially for snacks or add-ins in breakfast cereals (Pan et al., 2008a, 2008b; Shih et al., 2008; Lin et al., 2007; Pan, 2006). Because IR heating is an efficient drying method that significantly shortens drying time, the energy saving of the SIRFD method could be significant. The following case study illustrates some successful accomplishments using SIRFD to produce crispy dried fruits with reduced drying time, improved energy efficiency, and improved product quality (Shih et al., 2008). 7.5.1 Samples and experiment design Fresh strawberries (variety Camarosa) obtained from Frozsun Foods, Inc. (Oxnard, Cal.) were used in this study. They were washed with water, destemmed, and then sliced into pieces 4:10 0:10 mm thick using a food processor (model FP 200, Hobart Corp., Troy, Ohio). To determine the moisture content of strawberries, 10±15 g samples were placed in pre-weighed aluminum weighing dishes and dried according to AOAC methods (AOAC, 1994) in a vacuum oven (model V01218A, Lindberg/Blue, Asheville, NC). The dishes were removed and weighed using a balance with an accuracy of 0.01 g (model 602, Denver Instrument Co., Arvada, Colo.). The slices were then dried using various methods including IR, hot air, and FD. The moisture content of fresh strawberries ranged from 89.9 to 91.0% wet basis (w.b). The purpose of the study was to determine the drying characteristics and product quality of sliced strawberries by pre-dehydration to remove 30, 40, or



157

50% of their initial weight using each of the three IR intensities (3000, 4000, and 5000 W mÿ2) before freeze drying. The corresponding target moisture contents were 86, 83, or 80%, respectively, after the pre-dehydration. The weight changes were measured every minute during the drying process using a digital balance. For comparison, samples with similar weights of moisture removed were produced using hot-air drying. When hot air was used for the pre-dehydration step, the process is called sequential hot-air freeze-drying (SHAFD). Both sets of samples were frozen at ÿ18 ëC before being exposed to FD to reach a final moisture content of about 5% (w.b.) for quality evaluation. The dried products were evaluated for color, thickness shrinkage, rehydration ratio, crispness, and firmness. The samples were also dried for different durations to determine the drying rate of FD. 7.5.2 IR and hot-air pre-dehydration methods An IR dryer/dehydrator equipped with two IR emitters powered by natural gas was used in the IR pre-dehydration tests. Figure 7.10 schematically illustrates the equipment set-up during the experiments. The average IR intensities were measured with an Ophir FL205A thermal excimer absorber head (Ophir Optronics, Inc., Wilmington, MA) with 3% accuracy. An automatic data acquisition and control system developed in our laboratory controlled and recorded various operation parameters. Strawberry slices were arranged in a single layer on the drying tray (metal screen), which was sprayed with PAM cooking spray (ConAgra Foods, Inc., Omaha, Neb.) to prevent the slices from sticking to the tray. The drying tray was placed between the two IR emitters in a position parallel to the emitter face. Strawberry slices were heated from both top and bottom. Slices were placed within the confines of the waveguard at a loading of approximately 1.33 kg mÿ2 (or about 240 g for each batch). The

Fig. 7.10

Schematic diagram of catalytic infrared dryer setup.


158


change in sample weight during the drying process was measured using a digital balance until the target weight reduction was reached. Type-T thermocouples (0.15 s response time) placed at the center of the strawberry slices were used to measure the product temperature. For comparison, a hot-air cabinet dryer (product code 062, Proctor and Schwartz, Inc., Horsham, Pa.) was also used to dry the samples, to obtain a drying curve, and for quality evaluation. The dryer was set at 62.8 ëC based on common industrial practice, and the sample weight changes were also measured using a digital balance during drying. The drying rates of IR and hot-air drying were calculated based on the weight change and expressed as weight of water (g)/{initial weight (g) time (min)}. After the IR or hot-air pre-dehydration, the slices were transferred to wax paper by flipping the drying tray, and then transported to a large-scale air-blast freezing system at a temperature of ÿ18 ëC. 7.5.3 Freeze-drying method The frozen, pre-dehydrated samples and the control samples were removed from the freezer and placed as a single layer in a pilot-scale Ultra\VirTual Series EL freeze-dryer (VirTis Co., Gardiner, NY). The freeze-dryer was operated in shelfdriven mode, which was controlled based on shelf temperature, and run with programmed procedures. To determine the drying characteristics of the strawberry slices during FD, the pre-dehydrated and control samples were dried for various times (0.0167, 1, 2, 4, 6, 8, 16, 22, and 29 h). Control samples were also dried for 50 h. The samples were weighed at the end of each drying period, and the moisture contents were calculated. All samples used for quality evaluation had about 5% moisture content (w.b.). 7.5.4 Quality evaluation Thickness The thicknesses of the unprocessed samples and dried samples were measured using a Cen-tech electronic digital caliper (Harbor Freight Tools, Camarillo, CA). Shrinkage was determined based on the difference between initial and final thicknesses and recorded as a percentage of initial thickness. Color Color values L, a, and b were measured using a Minolta CR-200 reflectance colorimeter (Minolta, Japan). The colorimeter (illuminant D65, 2 observer) was calibrated against a standard ceramic white tile (Y 94:4, x 0:3159, y 0:3333). Because the color varied on the surface of the strawberry slices, the dried samples were ground to powder using a small-scale blender to obtain representative colors. A 1 g sample of strawberry powder was put in a 5 cm diameter plastic Petri dish. The lens of the colorimeter, covered with plastic wrap, was placed directly on the strawberry powder to measure the color values.



159

Re-hydration ratio Since dried products such as strawberry slices may be used in cereals, a rehydration test was performed. Five samples of dried strawberry from each drying trial were placed in whole milk for 3 min. They were then removed from the milk, gently dried by blotting with paper, and re-weighed. The re-hydration ratio was calculated by dividing the final weight by the original weight (Lin et al., 1998). Crispness Crispness was evaluated using a TA.XT2 texture analyzer (Texture Technologies Corp., Scarsdale, NY). Dried strawberry samples were tested using a 6.4 mm (0.25 in.) diameter ball probe and its accompanying chip/cracker fixture (TA-101). A `pipe' cylinder with an outside diameter of 25 mm and an inside diameter of 18 mm was mounted on the plate component of the TA-101 to support the strawberry samples. The values of initial slope (crispness, g mmÿ1) of the force curve were measured and calculated. Microstructure To evaluate the structural change of slices dried with different methods and to understand the mechanism of water transport during drying, scanning electronic microscopy (SEM) studies of cross-sections of the dried slices were performed. Selected samples were carefully cut using sharp razor blades (Ted Pella, Inc., Redding, CA) to expose a cross-section surface. Specimens were mounted on aluminum stubs using double-coated carbon tabs (Ted Pella, Inc., Redding, CA), sputter-coated with gold-palladium using a Denton Desk II sputter-coating unit (Denton Vacuum, Moorestown, NJ), and photographed in a Hitachi S-4700 field emission scanning electron microscope (Hitachi, Japan) at 2 kV. 7.5.5 Results of sequential IR and freeze-drying of strawberry slices Moisture IR pre-dehydrated strawberry slices took less time to reach a specific moisture content (regular FD or control) than the slices without pre-dehydration (Table 7.5). The samples with 40% weight reduction pre-dehydrated at IR intensity of 4000 W mÿ2 took only 29 h to achieve the final moisture content of approximately 5% (w.b.), compared to 50 h for the control. The SIRFD method saved about 42% of FD time, which indicates a significant energy saving potential. Moisture content did not change much at the early stage of FD. The FD process involves three stages: (1) the freezing stage, (2) the primary drying stage, and (3) the secondary drying stage. In the freezing stage, the temperature of the strawberry slices was lowered from ÿ18 ëC to ÿ40 ëC in 2 h. Drying of the foodstuff took place in the primary drying stage when the drying chamber was evacuated and its pressure was reduced to a value that would allow frozen water to sublime. Therefore, moisture loss did not take place until after the 2 h freezing stage.



Table 7.5

Average moisture contents of samples with different treatments during freeze-drying

Time (h)

Control

0.0167 1 2 4 6 8 16 22 29 50

90.10 89.63 90.36 88.34 74.09 38.18 21.18 11.66 8.93 5.90

30% Weight reduction



3000 W mÿ2

4000 W mÿ2

5000 W mÿ2

3000 W mÿ2

4000 W mÿ2

5000 W mÿ2

3000 W mÿ2

4000 W mÿ2

5000 W mÿ2

84.95 84.03 85.09 80.58 70.23 33.64 9.91 6.96 5.76

85.25 84.22 85.37 81.83 70.87 34.13 10.59 7.42 6.61

85.46 84.31 85.40 81.98 71.52 33.94 12.44 10.85 6.69

81.75 80.90 81.82 78.69 59.94 30.70 8.70 6.77 4.98

82.44 80.91 82.47 78.89 60.68 32.44 8.59 6.62 5.43

82.71 80.91 82.48 78.99 61.16 34.86 9.87 8.64 5.94

79.74 78.38 79.89 76.39 56.37 28.08 6.94 5.07 4.39

79.95 78.92 79.89 76.55 56.97 29.27 6.36 5.40 4.66

80.16 78.92 80.32 77.48 62.45 31.85 9.14 7.12 5.68


161

It was found that pre-dehydration resulted in faster drying by FD, which may be related to amount of the water in the product. Based on statistical analysis of the results, IR intensity did not have a significant effect on the FD process (p > 0:05), but the level of weight reduction by pre-dehydration did (p < 0:05). For example, at the end of a 29 h FD period, the moisture contents of SIRFD samples dried at an IR radiation intensity of 5000 W mÿ2 were 5.76, 4.98, and 4.39% for 30, 40, and 50% weight reduction during catalytic IR drying, respectively. This showed that a high level of pre-dehydration can significantly reduce the FD time. Product quality is another factor that needs to be considered in determining an appropriate level of pre-dehydration. Shrinkage Ketelaars et al. (1992) found that shrinkage during drying is attributed to moisture removal and to stresses developed in the cell structure during drying. Shrinkage was evident for all drying methods and conditions used in this study, with an extent dependent on the methods and conditions (Fig. 7.11). Since regular FD samples were not pre-dehydrated, structural rigidity was created during the freezing stage of the FD process that prevented collapse of the solid matrix during drying (Mujumdar, 1995). In general, SIRFD samples showed slightly more shrinkage in thickness than FD samples, but less shrinkage than SHAFD samples. The thickness shrinkages were 5.0% for regular FD samples. For SIRFD, more shrinkage was observed in relation to more weight reduction by pre-dehydration. For example, samples pre-dehydrated under 5000 W mÿ2 had 11.6, 19.14, and 20.8% thickness shrinkages for the 30, 40, and 50% weight

Fig. 7.11 Thickness shrinkage of dehydrated strawberry slices dried with different methods and conditions. Reg FD: regular freeze-drying; SIRFD: sequential infrared radiation and freeze-drying; and SHAFD: sequential hot-air and freeze-drying.


162


reductions, respectively, because of the moisture loss during pre-dehydration stage and also because of the longer drying times required to obtain 50% weight reductions compared to 30 or 40% reductions. Product shrinkage also depended on radiation intensity: thickness shrinkage decreased as radiation intensity increased. For instance, shrinkage decreased from 16.6% for a sample dried under 3000 W mÿ2 to 11.6% for a sample dried under 5000 W mÿ2 to achieve a 30% weight reduction. The differences in shrinkage of the slices can be explained by the drying rate. Drying at higher radiation intensities requires shorter times to achieve the target weight reduction, and the heat exposure time for the slices was therefore shorter compared to drying at lower radiation intensity, and, consequently, causes less deterioration in the cell structure and matrix. Compared to SHAFD samples, SIRFD samples experienced much less thickness shrinkage in the dried product. The results showed that with 50% weight reduction, a SHAFD sample had 34.13% thickness shrinkage as compared to 20.85% for the SIRFD sample dried under 5000 W mÿ2. This might be due to the longer hot-air drying time causing more cells to collapse. For products that tend to exhibit less shrinkage, it is recommended that higher radiation intensity, such as 5000 W mÿ2, be used during predehydration. Color Color measurement results were significantly different (p < 0:05) for all drying methods and conditions. The weight-reduction level in pre-dehydration had more influence on the L and a values of strawberry slices than the other process variables. Figure 7.12 shows the L values of fresh and dried strawberry samples. In general, drying significantly increased whiteness (increased L value), with the lightness of FD and SHAFD samples greater than the SIRFD samples, resulting in a lighter color tone of the dried products. The increased whiteness was similar to results found in the study of Li and Ma (2003), where the brightness/ whiteness of sliced strawberries increased after FD. In fact, the water content of the fresh and dried products also affected their appearance. For the SIRFD samples, L values decreased as weight reduction increased at the same radiation intensity. This could be due to the corresponding longer IR drying time causing darkening of the strawberry slices. The color measurements showed that the redness of SIRFD samples was generally stronger than the redness of fresh, FD, or SHAFD strawberry samples, resulting in a dark-red color in the dried products (Fig. 7.13). This phenomenon could be attributed to the water loss effectively increasing the concentration of red pigments (anthocyanins) in the dried product (Hammami and ReneÂ, 1997). Compared to SHAFD, the SIRFD samples experienced a higher drying temperature in pre-dehydration, which led to greater a values. This could be caused by the acceleration of non-enzymatic browning with temperature (Jamradloedluk et al., 2007). With IR pre-dehydration, the temperature of the strawberry slice increases faster than with hot-air drying, and SIRFD strawberries were more reddish than SHAFD strawberries.



163

Fig. 7.12 Color value L of strawberry samples with different drying methods. Reg FD: regular freeze-drying; SIRFD: sequential infrared radiation and freeze-drying; and SHAFD: sequential hot-air and freeze-drying; SB: strawberry (samples with different letters are significantly different at p < 0:05).

The desired color of the finished product may have a hue angle value similar to the fresh sample, which has a hue of 22.3 (orange-red color). Based on statistical analysis of the results, the hue angles were significantly different (p < 0:05) among all samples (Fig. 7.14). The SIRFD and SHAFD samples had hue angle values lower than fresh samples, but higher than FD samples. Two of

Fig. 7.13 Color value a of strawberry samples with different drying methods. Reg FD: regular freeze-drying; SIRFD: sequential infrared radiation and freeze-drying; and SHAFD: sequential hot-air and freeze-drying; SB: strawberry (samples with different letters are significantly different at p < 0:05).


164


Fig. 7.14 Hue angle of strawberry samples from all drying tests. Reg FD: regular freeze-drying; SIRFD: sequential infrared radiation and freeze-drying; and SHAFD: sequential hot-air and freeze-drying; SB: strawberry (samples with different letters are significantly different at p < 0:05).

the SIRFD samples (3000 W mÿ2 with 50% weight reduction level, and 4000 W mÿ2 with 30% weight reduction) resembled the fresh strawberry product. The SHAFD samples also had hue angle values closer to those of the fresh samples, but with their increased L value, the samples were observed to have a light orange-reddish color. The FD samples, on the other hand, had the lowest hue angle value (19.3), but with the greater increase in L value and smaller a value, appeared pinkish. From visual observation, the darker tone of SIRFD strawberries resulting from non-enzymatic browning and effective concentration of the anthocyanins is more desirable than the light tone of FD samples and the light orange-reddish color tone of SHAFD samples (Fig. 7.15). Baysal et al. (2003) found that hue angles were not significantly different among raw, hot air, microwave, and IR dried samples in their study drying carrots and garlic. For SIRFD samples, hue angle values changed with radiation intensity and weight reduction. When the radiation intensity was low, the increased weight reduction level increased the hue angle value. However, when radiation intensity was high, the increased weight reduction decreased the hue angle value. Based on the statistical analysis results, radiation intensity had a more significant effect on hue angle value than weight reduction level. Microstructure of strawberries sliced by cross-section During drying, water in the berry could be transported via several possible pathways (Tyree, 1970). In the first pathway, water passes from one cell to the next via cytoplasmic strands (plasmodesmata). In the second, water enters and leaves successive cells along its pathway by passing through plasmalemma boundaries. The most important pathway for water movement through plant tissues is through the cell wall.



165

Fig. 7.15 Effect of drying method on appearance of dried strawberries. (a) Regular FD strawberry, (b) SIRFD strawberry, and (c) SHAFD strawberry samples. Reg FD: regular freeze-drying; SIRFD: sequential infrared radiation and freeze-drying; and SHAFD: sequential hot-air and freeze-drying.


166


Fig. 7.16 SEM of cross-section of strawberry slices dried under different drying methods. (a) Regular FD, (b) SIRFD, (c) SHAFD. Reg FD: regular freeze-drying; SIRFD: sequential infrared radiation and freeze-drying; and SHAFD: sequential hot-air and freeze-drying.



167

It was evident that the FD strawberry structure (Fig. 7.16(a)) had uniform, small pores with little or no damage or disruption of cell walls at the slice surface. When a strawberry slice is dried by IR, it heats rapidly. Water vapor expands the cellular walls and develops large pores within the material (Jamradloedluk et al., 2007). This unique microstructure could enhance crunchiness or crispness. Accordingly, the SIRFD sample (Fig. 7.16(b)) showed cells collapsed at the surface layer, with a dense layer or crust at the surfaces and a porous structure in the interior slices. This result was expected, since the surfaces of strawberry slices were exposed to rapid surface heating during IR drying. Mositure evaporating from the surface caused at a rate comparable to or higher than the rate at which moisture migrated from the interior to the surface caused cells to collapse. Consequently, a dense layer formed at the surface of the SIRFD sample, and large pores (intercellular spaces) were seen in slices from the center region of the strawberry, which could be due to water vapor created during catalytic IR drying. Unlike the case with IR heating, the temperature of samples dried with hot air increased gradually from ambient temperature to the drying temperature. As the moisture in the materials was released, the vapor pressure caused by internal evaporation of moisture was less than in the case of IR drying. Therefore, the SHAFD sample (Fig. 7.16(c)) revealed severe structural damage of the cell walls. In particular, the cell walls of the center region completely collapsed, attributable to the long drying time required by hot-air drying. As a result, the hot-air dried samples were not as crispy as the SIRFD samples. Rehydration ratio The different rehydration capacities of strawberries dried by different methods at different conditions are shown in Fig. 7.17. In general, the SIRFD samples had a lower rehydration ratio than the FD samples; however, the SIRFD samples had a higher rehydration ratio than SHAFD samples. The fact that SIRFD strawberry slices had lower rehydration capacity than FD samples could be explained by the crust formation in the SIRFD samples that could have slowed down the penetration of milk into the dried sample during rehydration, whereas the more porous structure of FD samples facilitated rapid rehydration in milk. It is generally believed that the degree of rehydration is dependent on the degree of cellular and structural disruption (McMinn and Magee, 1997). Since the FD sample did not experience high-temperature heating, the cell structure was not damaged, and structural rigidity was maintained that created a porous structure. As for the SHAFD samples, the cellular structure was completely collapsed due to the long heating time, thereby making rehydration of the SHAFD samples more difficult compared to the SIRFD and FD samples. In comparing the pre-dehydrated samples, SIRFD samples showed better rehydration ratios (1.03±1.71) than SHAFD samples (0.92±1.06). Apparently, product shrinkage from the collapse of cellular tissues caused by severe heating and/or prolonged drying made it more difficult to rehydrate SHAFD samples. In


168


Fig. 7.17 Rehydration ratio of strawberry slices dried with different drying methods after 3 min soaking. Reg FD: regular freeze-drying; SIRFD: sequential infrared radiation and freeze-drying; and SHAFD: sequential hot-air and freeze-drying (samples with different letters are significantly different at p < 0:05).

a review by Sakai and Hanazawa (1994), the rehydration capability of Welsh onions dried with far-IR radiation under vacuum was greater than for those dried with hot air. Similar results were also reported by Kumar et al. (2005) for IR and hot-air drying of onions. The rehydration ratio of the SIRFD samples decreased as the weight reduction by IR-drying increased, since more shrinkage was observed with higher weight reduction. For example, the rehydration ratio of the 5000 W mÿ2 SIRFD samples during soaking for 3 min decreased from 1.71 to 1.03 as the weight reduction increased from 30 to 50%. As a result, the strawberry slices could not easily soak up milk. Crispness Texture is an important sensory attribute for many cereal-based foods. A crisp food should be firm and snap easily when deformed, emitting a crunchy sound. Tests of mechanical compression have been used to correlate crispness to a physical parameter in a force-deformation curve (Krokida et al., 2001). The crispness of strawberry slices dried by different methods is shown in Fig. 7.18. Statistical analysis indicated that drying method had a significant effect (p < 0:05) on the crispness of the final products. The samples processed with SIRFD had higher crispness than those processed by FD or SHAFD. Crispness was mainly related to the crust/dense layer formation at the surface and structural changes. The SIRFD sample had a modest crust and large porous structure in the central region, resulting in a high-crispness product. SIRFD samples dried with a radiation intensity of 5000 W mÿ2 were crisper than samples dried at lower intensities. Drying temperature may contribute to the



169

Fig. 7.18 Crispness comparison of strawberry slices dried with different drying methods. Reg FD: regular freeze-drying; SIRFD: sequential infrared radiation and freezedrying; and SHAFD: sequential hot-air and freeze-drying (samples with different letters are significantly different at p < 0:05).

effect of crispness; the increased temperature at the higher radiation intensity may also remove moisture faster in the strawberry slices. As mentioned earlier, the pores in the central region of the SHAFD sample collapsed. Because the membranes of all the cells were completely disrupted in the SHAFD sample, the middle lamella practically disappeared, indicating breakdown of pectins in the middle lamella of the cell walls and loss of binding force between cells (Alvarez et al., 1995). Thus, the SHAFD sample was a less crisp product. 7.5.6 Concluding remarks on sequential IR and freeze-drying of strawberry slices IR drying provided a much higher drying rate than hot-air drying, and the rate increased remarkably with increases in radiation intensity. The recommended radiation intensity for IR drying was 5000 W mÿ2 to achieve a weight reduction of 30±40%. Excessive water loss and weight reduction during IR predehydration may result in increased shrinkage. SIRFD strawberry samples exhibited slightly more shrinkage than FD strawberry samples, but less shrinkage than SHAFD samples. Product firmness of SIRFD treated samples was higher compared to FD samples. Color data show an increase of whiteness and redness in the dried samples compared to fresh strawberries. The strawberry samples treated with SIRFD exhibit rehydration ratios somewhere below the values of FD samples and above the values for SHAFD samples. Strawberry chips dried with SIRFD had a darker red color and crisper texture than the FD or SHAFD samples. It is likely that the SIRFD processing method can also be used


170


for producing other crispy fruit and vegetable pieces with improved product quality and increased processing efficiency to reduce energy consumption.

7.6

Infrared (IR) pasteurization of raw almonds

Owing to the outbreaks of salmonellosis associated with whole raw almonds, the almond industry is pursuing a mandatory pasteurization plan that takes aggressive measures to prevent outbreaks of salmonellosis. Several different technologies have been used or are under consideration for raw almond pasteurization and the inactivation of Salmonella enterica serovar Enteritidis on raw almond kernels, including propylene oxide (PPO) fumigation, FMC JSP-I almond surface pasteurization technology, Ventilex steam pasteurizer, vacuum steam, radio frequency, cold plasma, and IR heating. The following case study focuses on the efficacy of IR heating and holding for pasteurization, optimization of the pasteurization procedures, and maintenance of quality of raw almonds. The key deliverables in this case study are: (1) measures of IR pasteurization effectiveness and product quality under different combinations of IR heating temperature, holding temperature and time; and (2) optimized processing conditions of IR pasteurization methods for commercial implementation by the almond industry. 7.6.1 Approaches to study pasteurization of raw almonds In order to benefit from the high heat flux provided by IR technology to reduce the heating-up time, almonds were heated to 100, 110 and 120 ëC using IR, then held in a custom-designed holding device at 70, 80 or 90 ëC for different periods of time. In order to investigate the effect of each heat treatment step on the possible quality change of the raw almonds, qualitative and quantitative assays were conducted. Qualitative assays were based on observing changes in flavor of raw almonds after treatments, whereas quantitative assays were based on measuring the color of almonds in Lab color space and examining the overall color change (E) and overall hue angle change (Hueë) values. Sensory analysis was conducted with 80 panelists for raw almonds treated with conditions that provided over 4-log bacterial reductions of Pediococcus (the surrogate for Salmonella enterica serovar Enteritidis). 7.6.2 Results on IR pasteurization of raw almonds In the preceding discussion, the total reduction of Pediococcus is actually the cumulative value contributed by each stage (heating by IR, cooling, then holding) of the process. The total reduction value of a treatment can be expressed as ReductionTotal Reductionheating

by IR

Reductioncooling Reductionholding



171

As the target temperature of the IR heating decreased, the bacterial reductions at different holding times and temperatures also decreased. Heating almonds to 100, 110 or 120 ëC by exposure to IR for less than 1 min reduced the Pediococcus population by 0.320, 0.583 and 0.620 logs, respectively (Table 7.6). During the subsequent cooling period in ambient air, the bacterial reduction increased 1 to 1.7-log, due to the fact that the surface temperature of the almond was over 70 ëC for 28 s until it reached 120 ëC, but it took 165 s for the surface to cool from 120 to 70 ëC. Therefore, almonds experience longer durations of temperatures capable of inactivating Pediococcus during the cooling stage than they do during the IR heating period. In general, when the almond surface temperatures were increased to 100, 110 and 120 ëC by IR heating, cooled to 90 or 80 ëC, then held for 5±15 or 22±30 min, respectively, more than 4-log reductions of Pediococcus were achieved. When the almonds were initially heated to 120 ëC by IR, the holding times at 80 and 90 ëC achieved over 4-log reductions. In the case where almonds were heated to 110 ëC by IR, all of the holding times at 90 ëC and only the 30 min holding time at 80 ëC provided over 4-log reduction of Pediococcus. In most of these conditions, except {IR 110 ëC ± 80 ëC ± 22 min}, {IR 100 ëC ± 80 ëC ± 22 min}, and {IR 100 ëC ± 90 ëC ± 5 min}, our results demonstrated that IR heating combined with holding effectively pasteurizes the almonds and meets the industrial pasteurization requirements of a minimum 4-log bacterial reduction. In general, during ambient cooling of IR-heated almonds to holding temperatures of 70, 80 or 90 ëC, an additional 0.5±1.0 log reduction occurred which increased the total Pediococcus reduction around 0.8±1.8 logs. Holding the almonds at 70, 80, or 90 ëC provided an additional 1.4±7.5 log bacterial reduction. At the end of the whole process (IR heating, cooling and holding), it was clear that the holding temperature of 70 ëC did not provide the required 4log reduction, whereas holding at 90 and 80 ëC met the pasteurization requirements, with the exceptions above noted. The combined heating, cooling, and hold times considered in this study only slightly changed the Hue values of the skin of almond and its flesh (Tables 7.7 and 7.8). These changes, compiled in the tables, are around 3ë, and may not be visually distinguishable. Thus, the color of treated almonds will be perceived to be the same as the raw almonds. Under most treatment conditions, the E values of almond's skin and flesh varied in the range of 0.7±3.0 and 3.1±5.9, respectively. The variation in E value of almond's skin was probably due in part to the variation of the raw almond's color components, rather than the heat treatment, since there was no observable monotonic increase/decrease correlated with either the holding time or the holding temperature. However, the increase in E value of almond flesh could be due to the slight increase of b* value (yellow) caused by the heat. During sensory analysis, over 80% of the panelists found no difference between treated and untreated almonds in terms of appearance. Sensory panelists observed that treatments heating the almonds to 100 ëC and holding at 90 ëC for 10 min, or heating to 110 ëC and holding either at 90 ëC for 10 min or 80 ëC for 30 min, or heating to 120 ëC and holding at 80 ëC


Table 7.6

Log reduction value of Pediococcus population

Treatment


IR IR IR IR IR IR IR IR IR IR IR IR

heating to 100 ëC heating to 110 ëC heating to 120 ëC 120 ëC + Cooling 120 ëC + Cooling 120 ëC + Cooling 110 ëC + Cooling 110 ëC + Cooling 110 ëC + Cooling 100 ëC + Cooling 100 ëC + Cooling 100 ëC + Cooling

IR120 ëC IR110 ëC IR100 ëC IR120 ëC IR110 ëC IR100 ëC IR120 ëC IR110 ëC IR100 ëC IR120 ëC IR110 ëC IR100 ëC a

± ± ± ± ± ± ± ± ± ± ± ±

70 ëC 70 ëC 70 ëC 70 ëC 70 ëC 70 ëC 70 ëC 70 ëC 70 ëC 70 ëC 70 ëC 70 ëC

± ± ± ± ± ± ± ± ± ± ± ±

to to to to to to to to to

15 15 15 30 30 30 45 45 45 60 60 60

90 ëC 80 ëC 70 ëC 90 ëC 80 ëC 70 ëC 90 ëC 80 ëC 70 ëC

mina min min min min min min min min min min min

Total reduction

Stage reduction

Treatment

0.320 0.583 0.620 1.350 1.430 1.670 1.140 1.280 1.480 0.790 0.930 0.950

0.320 0.583 0.620 0.730 0.810 1.050 0.557 0.697 0.897 0.470 0.610 0.630

IR120 ëC IR110 ëC IR100 ëC IR120 ëC IR110 ëC IR100 ëC IR120 ëC IR110 ëC IR100 ëC

± ± ± ± ± ± ± ± ±


± ± ± ± ± ± ± ± ±

15 15 15 22 22 22 30 30 30

2.429 2.327 2.050 3.271 2.999 2.929 3.692 3.438 2.908 3.846 3.692 3.234

1.379 1.430 1.420 2.221 2.102 2.299 2.642 2.541 2.278 2.796 2.795 2.604

IR120 ëC IR110 ëC IR100 ëC IR120 ëC IR110 ëC IR100 ëC IR120 ëC IR110 ëC IR100 ëC

± ± ± ± ± ± ± ± ±


± ± ± ± ± ± ± ± ±

Format of the notation is IR heating temperature ± holding temperature ± holding time

Total reduction

Stage reduction

min min min min min min min min min

3.643 3.497 2.928 4.109 3.909 3.665 6.838 6.105 4.989

2.833 2.800 2.318 3.299 3.212 3.055 6.028 5.408 4.379

5 min 5 min 5 min 10 min 10 min 10 min 15 min 15 min 15 min

5.700 4.143 3.810 7.629 6.460 5.678 8.258 7.779 5.981

4.970 3.586 3.340 5.087 5.903 5.208 7.528 7.222 5.511

Table 7.7

Color parameters (L*a*b*, Hueë and E) of almond's skin

Treatment


Raw almond IR heating to 100 ëC IR heating to 110 ëC IR heating to 120 ëC IR120 ëC ± 70 ëC ± 15 min IR110 ëC ± 70 ëC ± 15 min IR100 ëC ± 70 ëC ± 15 min IR120 ëC ± 70 ëC ± 30 min IR110 ëC ± 70 ëC ± 30 min IR100 ëC ± 70 ëC ± 30 min IR120 ëC ± 70 ëC ± 45 min IR110 ëC ± 70 ëC ± 45 min IR100 ëC ± 70 ëC ± 45 min IR120 ëC ± 70 ëC ± 60 min IR110 ëC ± 70 ëC ± 60 min IR100 ëC ± 70 ëC ± 60 min IR120 ëC ± 80 ëC ± 15 min IR110 ëC ± 80 ëC ± 15 min IR100 ëC ± 80 ëC ± 15 min IR120 ëC ± 80 ëC ± 30 min IR110 ëC ± 80 ëC ± 30 min IR100 ëC ± 80 ëC ± 30 min IR120 ëC ± 90 ëC ± 15 min IR110 ëC ± 90 ëC ± 15 min IR100 ëC ± 90 ëC ± 15 min

L*

a*

b*

Hueë

E

48.92 2.34 49.69 2.12 49.05 2.49 48.27 1.96 49.75 0.86 49.53 1.24 49.57 0.42 48.92 0.72 48.74 0.33 49.82 0.52 48.33 1.24 49.01 1.63 48.65 1.41 49.23 2.90 48.00 0.23 48.39 0.18 47.00 0.65 47.58 0.33 47.44 0.95 49.45 0.81 49.68 0.67 49.62 0.52 47.08 0.47 47.76 0.45 46.16 0.83

16.80 0.88 16.60 0.69 17.03 0.58 17.27 0.63 16.71 0.29 17.44 0.51 16.88 0.29 17.23 0.21 17.28 0.08 17.04 0.05 17.20 0.29 17.04 0.24 17.19 0.54 17.39 0.95 17.79 0.19 17.59 0.11 17.27 0.51 17.02 0.28 17.58 0.46 17.36 0.15 17.41 0.46 17.25 0.07 17.85 0.20 17.69 0.22 17.42 0.23

31.91 2.23 32.77 1.55 31.94 2.25 31.23 2.03 32.65 0.77 32.57 0.46 32.19 0.91 32.36 0.60 32.67 0.96 33.28 0.83 32.46 1.11 33.30 1.13 33.80 1.55 32.60 2.54 33.06 0.10 32.63 0.59 30.34 0.72 30.16 0.34 31.20 1.87 32.90 0.60 33.31 0.09 33.07 0.89 30.60 0.12 30.99 0.93 29.82 0.66

N/A 1.34 0.90 1.37 1.13 1.80 1.15 0.68 0.54 0.75 0.84 0.22 0.24 0.49 0.23 0.69 0.18 0.62 0.39 0.35 0.41 0.86 0.67 0.77 0.46 1.26 0.85 0.70 0.23 0.80 0.63 2.08 0.67 1.89 0.50 1.86 0.85 0.48 0.29 0.43 0.31 0.46 0.17 2.74 0.30 2.18 0.69 2.74 0.85

N/A 2.38 1.46 2.89 1.66 2.48 1.71 1.05 0.71 1.37 0.33 0.89 0.24 0.90 0.37 1.11 0.45 1.25 0.88 1.56 0.84 1.91 0.69 2.29 1.19 2.04 0.73 1.79 0.30 1.28 0.26 2.97 0.92 2.64 0.44 2.52 1.49 1.13 0.60 1.50 0.12 1.25 0.76 2.87 0.44 2.16 0.78 3.21 0.71

Table 7.8

Color parameters (L*a*b*, Hueë and E) of almond's flesh

Treatment

L*


Raw almond IR heating to 100 ëC IR heating to 110 ëC IR heating to 120 ëC IR120 ëC ± 70 ëC ± 15 min IR110 ëC ± 70 ëC ± 15 min IR100 ëC ± 70 ëC ± 15 min IR120 ëC ± 70 ëC ± 30 min IR110 ëC ± 70 ëC ± 30 min IR100 ëC ± 70 ëC ± 30 min IR120 ëC ± 70 ëC ± 45 min IR110 ëC ± 70 ëC ± 45 min IR100 ëC ± 70 ëC ± 45 min IR120 ëC ± 70 ëC ± 60 min IR110 ëC ± 70 ëC ± 60 min IR100 ëC ± 70 ëC ± 60 min IR120 ëC ± 80 ëC ± 15 min IR110 ëC ± 80 ëC ± 15 min IR100 ëC ± 80 ëC ± 15 min IR120 ëC ± 80 ëC ± 30 min IR110 ëC ± 80 ëC ± 30 min IR100 ëC ± 80 ëC ± 30 min IR120 ëC ± 90 ëC ± 15 min IR110 ëC ± 90 ëC ± 15 min IR100 ëC ± 90 ëC ± 15 min

79.68 0.43 76.80 0.56 80.75 0.40 80.59 0.53 78.13 1.07 79.50 0.16 80.70 1.30 77.79 4.64 76.37 1.90 80.67 0.55 79.93 0.59 80.76 0.59 78.57 1.79 77.78 0.38 78.30 0.64 78.88 2.50 80.39 0.88 80.49 1.86 75.66 3.71 78.61 0.53 79.22 0.92 78.46 0.70 80.31 0.66 77.66 1.24 80.83 1.01

a* 1.04 0.29 0.90 0.31 ÿ0.07 0.20 ÿ0.15 0.21 ÿ0.35 0.16 ÿ0.53 0.45 0.61 0.19 ÿ0.05 0.40 0.44 0.21 0.62 0.34 ÿ0.05 0.28 0.01 0.28 0.60 0.17 ÿ0.44 0.12 0.00 0.25 ÿ0.43 0.94 ÿ0.49 0.33 ÿ0.38 0.39 0.30 0.17 ÿ0.23 0.11 0.55 0.10 0.96 0.10 ÿ0.40 0.13 ÿ0.06 0.36 0.47 0.17

b*

Hueë

E

19.89 0.53 22.99 0.48 23.34 0.30 22.85 0.79 23.50 0.62 24.19 0.57 23.53 1.23 24.13 2.11 24.71 0.27 22.93 0.54 23.93 0.44 23.09 1.41 22.79 1.04 24.14 0.45 23.75 0.21 24.91 0.37 23.62 1.04 23.46 0.38 24.38 0.82 24.07 0.02 24.03 0.34 24.18 0.26 23.22 0.58 25.41 0.35 23.73 0.35

N/A 0.82 0.65 3.17 0.49 3.37 0.53 3.84 0.38 4.25 1.10 1.52 0.41 3.16 0.99 1.97 0.49 1.43 0.86 3.10 0.67 2.95 0.69 1.48 0.45 4.05 0.31 3.00 0.59 4.00 2.17 4.17 0.79 3.91 0.94 2.29 0.38 3.53 0.26 1.70 0.22 0.73 0.24 3.96 0.30 3.12 0.81 1.87 0.41

N/A 4.25 0.65 3.80 0.31 3.39 0.56 4.27 0.55 4.62 0.39 4.05 0.64 6.04 2.35 6.00 1.30 3.25 0.71 4.23 0.46 3.63 1.19 3.32 1.60 4.90 0.48 4.25 0.46 5.62 0.44 4.20 0.79 4.21 0.59 5.75 1.32 4.52 0.12 4.26 0.40 4.51 0.20 3.74 0.43 6.07 0.51 4.15 0.07


175

for 30 min or 90 ëC for 5 min did not cause significant overall change of almond quality at a significance level of 0:10. 7.6.3 Concluding remarks on IR pasteurization of raw almonds The use of IR heating is a promising technology for the surface pasteurization of raw almonds without significantly compromising the raw almond quality attributes. Based on bacterial reduction, and preservation of sensory quality, any of the following three processing conditions is recommended to almond processors for best results: 1. IR heating 120 ëC, holding at 90 ëC for 5 min 2. IR heating 110 ëC, holding at 90 ëC for 10 min 3. IR heating 100 ëC, holding at 90 ëC for 10 min. Any of the above three treatments provides over 5.5 log reductions of Pediococcus. Since the required minimum bacterial reduction is 4-log, the above recommendations can be further optimized to a lower temperature and time.

7.7

Infrared (IR) dry-roasting of almonds

Dry-roasting is a thermal process used by the almond industry. At present, the typical dry roasting process uses hot air, which is achieved via a continuous conveyor roaster or rotary roaster. The continuous conveyor roaster can be single-stage or multiple-stages operating at a variety of temperatures. Common temperatures used for hot air roasting range from 130 to 154 ëC (265 to 310 ëF). At the lower temperature, it may take 40±45 minutes to obtain a light to medium roasted product, while at the higher temperature, it may take 10±15 minutes to obtain a light to medium roasted product. There are two concerns about the current dry roasting processes. Firstly, they may not ensure pasteurization of the product, particularly with respect to a minimum 4-log reduction of Salmonella Enteriditis PT 30 (SE PT 30). Secondly, they require a relatively long processing time, thereby also increasing processing costs. The industry has a desire to develop new processing methods that can produce a safely pasteurized product in a shorter time to achieve cost-savings. This case study outlines research that was also supported by the California Almond Board as the raw almond pasteurization. 7.7.1 Approaches to study IR dry-roasting of almonds IR heating was studied and applied to improve the safety and processing efficiency for dry-roasting almonds. The key deliverables in this case study include: · the appropriate IR heating conditions to achieve the desired product temperatures with minimum heating/roasting time;


176


· the pasteurization efficacy of IR compared to sequential IR radiation and hot air roasting (SIRHA) and to traditional hot air roasting; · the quality of the almond kernels so produced; and · recommendations about the technology for scaling up for commercial applications. Figure 7.19 shows the pilot scale IR equipment with double-sided heating (Catalytic Industrial Group Inc., Independence, KS) that was used in this study. The pilot scale IR device had four IR emitters, two at the top and two at the bottom, for a total heating area of 269 61 cm. When placed on the metal screen tray, the almond samples are located at distances of 16 and 12.5 cm from the top and bottom emitters, respectively. In order to reach the desired kernel temperature with a minimum heating time, the equipment could be operated at an IR intensity of 11 080 W mÿ2 corresponding to 3 inch-water pressure of natural gas supply for this equipment. Almonds were roasted at 130, 140, and 150 ëC with three different methods, IR heating, SIRHA heating, and traditional hot air heating. The heating rates and

Fig. 7.19

Pilot-scale catalytic infrared heating equipment.



177

color changes of the almonds by the different heating methods and temperatures were evaluated. The value of overall color change was used as the indicator of degree of roasting. The degree of roasting of commercial products represented by the overall color change was quantitatively described by using E computed with the equation by Ozdemir and Devres (2000). Standard commercial medium and heavily roasted kernels had E values of 11.5 and 21.4, respectively. Pediococcus spp. NRRL B-2354 was used as a surrogate of Salmonella Enteriditis PT 30 in evaluating the pasteurization efficacy of different processing methods and conditions. 7.7.2 Results of IR dry-roasting of almonds The roasting times for producing medium and heavily roasted almonds by using different dry-roasting conditions are listed in Table 7.9. The overall color changes of roasted almonds obtained under various conditions in this study are shown in Fig. 7.20. For hot air heating, 34, 18, and 13 min were required to reach medium roasting at 130, 140, and 150 ëC, respectively. For SIRHA, the corresponding times of hot air heating were reduced to 21, 11 and 5 min, excluding the additional IR preheating time of 39, 44, and 53 s, respectively. For IR roasting, the times of 11, 6 and 4 min were the shortest among the three methods at the same roasting temperature and the same final level of roasting using E values, corresponding to time reductions of 68, 67, and 69% compared to hot air, and 38, 39, and 62% compared to SIRHA roasting, respectively. The time required to heat almonds to 150 ëC was less than 1 min using IR, compared to about 15 min with hot air heating.

Fig. 7.20

Overall color changes (E) of almonds under different roasting conditions.



Table 7.9 Roasting times and time reductions for producing medium and heavily roasted almonds with different roasting methods and conditions Roasting method

Hot air

Infrared

Sequential IR and hot air

Roasting temperature (ëC)

130

140

150

130

140

150

130

140

150

Medium

Roasting time (min) Time reduction (%)a

34 ±

18 ±

13 ±

11 68

6 67

4 69

21 38

11 39

5 62

Heavily

Roasting time (min) Time reduction (%)a

72 ±

30 ±

19 ±

20 72

14 53

7 63

52 28

24 20

12 37

a

Time reduction (%) (Time of hot air roasting ÿ Time of IR or SIRHA roasting)/(Time of hot air roasting) 100%


179

Table 7.10 Reductions in Pediococcus population size on medium roasted almonds under different conditionsa Treatment temperature (ëC) Hot air treatment Infrared treatment Sequential IR and hot air treatment

130

140

150

3.58a AB 2.94a B 4.1a A

4.62a B 3.21a C 5.82b A

5.39a B 4.12b B 6.96c A

a The same letters in lower case in the same row mean no significant difference at P 0:05; the same letters in upper case in the same column mean no significant difference at P 0:05.

Table 7.10 shows the reductions in Pediococcus on medium roasted almonds under different conditions. When SIRHA roasting was used for producing medium roasted almonds, 4.10-, 5.82- and 6.96-log bacterial reductions were achieved by using the respective roasting temperatures of 130, 140 and 150 ëC and roasting times of 21, 11 and 5 min. When IR heating alone was used to produce medium roasted almonds, a 4.12-log bacterial reduction was achieved at 150 ëC for 4 min, compared to 13 min with hot air at 150 ëC. Hot air roasting at 140 and 150 ëC resulted in 4.62- and 5.39-log bacterial reductions, which required 18 and 13 min of roasting, respectively, and is much longer than the IR or SIRHA roasting. 7.7.3 Concluding remarks on IR dry-roasting of almonds SIRHA roasting is a substantially faster method for producing pasteurized roasted almonds with tremendous potential to reduce costs associated with the longer roasting times of current hot air methods. The roasting process can be easily implemented in the industry by adding an IR pre-heating device in front of regular hot air roasters. The roasting using IR alone is recommended only for pasteurization that targets 4-log bacterial reduction.

7.8 An overview of infrared (IR) rough rice drying and disinfestation Nearly all rice produced in the USA is dried by conventional convection drying which has low processing and energy efficiencies. This drying method has high production costs and lowers product quality (Kunze and Calderwood, 1985; Stipe et al., 1972). In order to mitigate low head rice yield (HRY) and milling quality, current practices normally use multiple drying passes that remove a relatively small amount of moisture (2±3%) in each pass but expose the rice to a relatively low temperature (up to 54 ëC for 15±20 min) to minimize the moisture gradient generated during the drying process. After each drying pass, the rice is tempered to allow the moisture inside the rice kernels to equilibrate before it is further dried. It has been reported that a reduction in the amount of head rice is influenced by the amount of moisture removed within a time interval, rather than


180


by the temperature of the drying air, which indicates that a certain amount of moisture can be quickly removed at a high temperature without significantly lowering the head rice yield. IR radiation offers advantages over conventional drying methods under similar drying conditions, including its high heating rate and energy efficiency (Pan et al., 2008b; Sharma et al., 2005; Das et al., 2004a, 2004b; Zhu et al., 2002; Afzal and Abe, 1997, 1998; Masamure et al., 1998; Ginzberg, 1969; Bilowicka, 1960). Because IR does not heat up the medium, the temperature of the rice kernel is not limited by the wet bulb temperature of the surrounding air, and the rice kernel can be quickly heated to high temperatures. Additionally, IR radiation heating achieves fast and relatively uniform heating due to the heat penetration of the rice kernel resulting in quick moisture removal, reduces the moisture gradient in the rice kernels during heating and drying, and improves milling quality. IR drying/heating technology may also provide the potentials to achieve disinfestation. Because the agricultural and food industry is currently facing the pressure of losing the use of methyl bromide, farmers and processors are seeking environmentally sound alternative methods for the disinfestation of rice. The following two case studies involve the use of IR for drying and disinfestation of rough rice: 1. Effectiveness of IR heating for simultaneous drying and disinfestation of freshly harvested rough rice. 2. Effectiveness of IR heating for disinfestation of stored rough rice. These studies focus on the application of IR drying/heating technology for processing rice with the aim of improving the processing and energy efficiency, producing finished products with improved quality and disinfestating dried rice.

7.9 Effectiveness of infrared (IR) heating for simultaneous drying and disinfestation of freshly harvested rough rice The goal of this study was to investigate the drying characteristics, milling quality, and effectiveness of disinfestation of rough rice under conditions of IR radiation heating (Pan et al., 2008b). The specific objectives were as follows: 1. To study the drying and milling characteristics of rice with high and low harvest moisture content (MC) undergoing single-layer heating using IR heating, followed by tempering and cooling treatments. 2. To determine the most effective IR heating conditions for disinfesting and the technical feasibility of simultaneously drying and disinfestating. 7.9.1 Approaches to accomplish IR drying and disinfestations Materials and equipment Freshly harvested medium grain rice, variety M202 was obtained from the Farmers' Rice Cooperative (West Sacramento, CA) and used for the IR drying



181

and disinfesting tests. The MC of rough rice at harvest was 25:0 0:3% (high MC). Rough rice was divided equally into two portions, one portion retained the high MC, and the second portion was dried slowly to a MC of 20:6 0:2% in ambient at a temperature ranging from 17 to 20 ëC. The thickness of the rice bed on the floor was less than 5 cm. During the slow drying, the rice was mixed frequently to ensure uniform drying. It took about 3 days to reach MC 20.6%. All reported MC determinations are on a wet weight basis and done according to the air oven method (130 ëC for 24 h, see ASAE, 1995). Both rice samples were kept in polyethylene bags and sealed to prevent moisture loss. The rice samples were divided into 250 g samples with a sample divider at test time. Four days before the disinfestation tests, some of the 250 g rice samples were infested with 100 adult lesser grain borers (Rhizopertha dominica) and 50 adult angoumois grain moths (Sitotroga cerealella), the most common insects in rough rice. The infested rice samples contained both adult insects and their eggs at the time of testing. The drying and disinfestation tests were separately conducted using noninfested and infested samples. A catalytic emitter provided by Catalytic Industrial Group (Independence, Kansas) was used as the IR radiation source. The dimensions of the emitter were 30 60 cm. An aluminum box with dimensions of 65 cm (length) 37 cm (width) 45 cm (height) was installed around the emitter as a waveguide to achieve uniform IR intensity of the rice bed surface. A 250 g rice sample was placed on the drying bed as a single layer with a corresponding calculated loading rate of 2 kg mÿ2. The rice bed was located 5 cm below the bottom edge of the waveguide. The average IR intensity at the rough rice bed surface was 5348 W mÿ2, which was measured using an Ophir FL205A Thermal Excimer Absorber Head (Ophir, Washington, MA). The drying bed was made with an aluminum plate of 3 mm thickness as its high reflectivity minimized the radiation energy loss through the drying bed. The reflected radiation energy could also be used to heat the bottom of the rice kernels. A piece of plywood was installed beneath the aluminum plate to reduce the energy loss through conduction. To measure the drying characteristics and milling quality, 16 non-infested rice samples were heated for each of four durations (15, 40, 60 or 90 s) with an initial drying bed surface temperature of 35 ëC. The rice sample weights were measured using a balance with two-decimal accuracy before and after heating to calculate the moisture removal by heating. For the disinfestation tests, heating for less than 15 s was too little to kill the insects, so 8 infested rice samples were heated for durations of 25, 40, 60, and 90 s. The grain temperature during the 25 s of heating was also determined. To compare milling quality, control samples were produced by drying the high and low MC rough rice samples, using room air, to 13.6% MC. 7.9.2 Tempering and cooling treatments In order to study the effects of tempering on moisture loss during cooling, disinfestation, and milling quality, tempered and non-tempered samples were


182


prepared. Half of the heated rice samples (8 non-infested and 4 infested samples) were tempered, and the rest of the samples were cooled without tempering. The tempering was conducted by placing rice samples in closed containers in an incubator for 4 h immediately following the heating. During incubation, the temperature in the incubator was set to be the same as the heated rice. For noninfested rice samples, four thin layers (about 1 cm thick) samples were each cooled using natural cooling (slow cooling) or forced air cooling at room temperature of 20±24 ëC. For natural cooling, the thin layer of rice was placed on a laboratory bench for about 30 min. For forced air cooling, the samples were placed on mesh trays and cooled by blowing room air through the bed with air velocity of 0.1 m/s for 5 min. After the natural and forced air cooling processes, the temperature of the rice samples were close to ambient. The weight changes during cooling were used to calculate the moisture loss. The cooled samples were stored in polyethylene bags before they were further dried to 13:3 0:2% MC using room air. Two 250 g samples of each treatment were combined (for a total weight of more than 400 g) for milling quality and disinfestation evaluation. The samples were stored in Ziplock bags at room temperature for about one month before milling. In order to avoid losing insects during handling, the infested rough rice samples during disinfestation tests were cooled only with natural cooling after heating or tempering. 7.9.3 Milling quality The most important rice milling quality indicators are total rice yield (TRY), head rice yield (HRY), and degree of milling (Whiteness Index, WI). To evaluate the effects of the different treatments, non-infested rice samples (400 g) were dehulled and milled using a Yamamoto Husker (FC-2K) and Yamamoto Rice Mill (VP-222N, Yamamoto Co. Ltd, Japan). The rice samples were milled three times to achieve well-milled rice as defined by the Federal Grain Inspection Service (USDA FGIS, 1994). The settings of Throughput and Whitening were 1 and 4, respectively, during the first two millings, and 1 and 5 during the third milling. HRY was determined with Graincheck (Foss North America, Eden Prairie, MN), and the WI was determined with the Whiteness Tester, C-300 (Kett Electronic Laboratory, Tokyo, Japan). A high index number indicates whiter milled rice. 7.9.4 Effectiveness of disinfestation treatment After the IR heating or tempering treatments, all naturally cooled, infested rice samples were transferred to glass jars with screened lids to allow moisture and oxygen exchange, and all jars were kept in incubators at 28 2 ëC with 64 3% relative humidity (RH) to allow development of the surviving insects and eggs (Kirkpatrick, 1975). The populations of the surviving and emergent live adult insects were visually counted one day after the treatment and then every several days over a 35-day period that covered more than one life cycle of the insects.



183

All adult insects were removed from the rice samples after each examination. The average numbers of live adult insects in the two samples under each treatment at different storage times are reported. Because each sample was obtained by combining two original samples, the original numbers of insects were doubled in each incubated sample. 7.9.5 Results of IR drying and disinfestation Data of the rice milling quality were statistically evaluated (p < 0:05) in Excel using the t test with the assumption of equal variances. TRY, HRY and WI of IR dried rice and control samples were statistically compared. Because the TRY and HRY of rice dried using IR heating followed by non-tempering or forced air cooling treatment were significantly lower than the corresponding values of the control samples (only statistical results of rice dried with IR heating followed by tempering and natural cooling are reported). The values with letter a were not significantly different from the control samples at p < 0:05. Moisture removal for different heating durations After the 20.6% and 25.0% MC rough rice samples were heated for 15, 40, 60, and 90 s, they reached corresponding temperatures of 42.8, 54.3, 61.2, and 69.4 ëC, and 42.8, 55.5, 59.1, and 68.0 ëC, respectively. The low MC rice samples rose to slightly higher temperatures than the high MC rice sample during 60 and 90 s of heating. The maximum difference in the temperatures of the samples with different initial MC under the same heating duration was 2.2 ëC, which is relatively small. Therefore, the average temperatures of low and high MC rice samples at different heating durations are presented in Fig. 7.21. A high correlation between the average rice temperature and heating time was obtained using a power model. The model can be used to predict the temperature change of the rice under the tested moisture range for a known heating time and bed temperature. If it is necessary to reduce the heating time, the method of preheating the drying bed to a relatively high temperature could be considered. The trend of high moisture removal for the high MC rice samples is clearly shown in Fig. 7.22, even though the difference between the low and high MC rice samples was relatively small. With 90 s heating (average temperature of 68.7 ëC), the moisture removal was 2.8 and 2.5 percentage points for the high and low MC rice samples, respectively. It is important to note that the average drying rates of rice samples with initial MCs of 25.0 and 20.5% MCs were 2.4, 1.8, 1.7, and 1.7 percentage points per minute at the moisture removal levels of 0.6, 1.2, 1.7, and 2.6 percentage points by each drying pass. The high drying rate at relatively high moisture removal levels by each drying pass, for example of 1.7 percentage points per minute at 1.7 and 2.6 percentage points MC removal, was much higher than that of the current commercial, conventional heated air drying of 0.1±0.2 percentage points per minute, due to the low air temperature used (Kunze and Calderwood, 1985). The high drying rate was achieved by using IR heating alone, without counting the moisture loss during cooling.


184


Fig. 7.21

Relationship between rice temperature and heating time.

Fig. 7.22 Moisture removals of rice samples with different initial moisture contents after heating to various temperatures.



185

Moisture removal under different tempering and cooling treatments The clear trends of tempering vs. non-tempering and natural cooling vs. forced air cooling are seen in Figs 7.23 and 7.24, respectively. For low MC rice, the moisture removal from the tempered rice samples under natural cooling and

Fig. 7.23 Moisture removal of rice with initial MC of 20.6% under different cooling methods with and without tempering (T ± tempering, NT ± no tempering, NC ± natural cooling, FAC ± forced air cooling).

Fig. 7.24 Moisture removal of rice with initial MC of 25.0% under different cooling methods with and without tempering (T ± tempering, NT ± no tempering, NC ± natural cooling, FAC ± forced air cooling).


186


forced air cooling were 0.6±1.3 and 1.1±1.9 percentage points, respectively, in the tested temperature range from 42.8 to 69.4 ëC. In contrast, non-tempered rice had 0.4±0.8 and 0.7±0.9 percentage point moisture loss under natural cooling and forced air cooling, respectively. Tempering resulted in 0.2±0.5 percentage points more moisture loss than non-tempering. The forced air cooling also removed up to 0.9 percentage points more moisture than natural cooling in the tested temperature range. However, at the high heating temperature of 69.4 ëC without tempering, similar amounts of moisture loss were achieved with both natural cooling and forced air cooling. This was due to the formation of moisture gradients after more than 2.5 percentage points of moisture were lost, such that moisture diffusion in the rice kernels became the factor limiting further improvement of the drying rate by forced air cooling. The high MC rice had moisture loss trends similar to those of the low MC rice during cooling, even though more moisture was removed compared to the low MC rice. The tempered rice had moisture removals of 1.6±2.2 percentage points for forced air cooling and 0.8±1.5 percentage points for natural cooling, compared to 1.1±1.3 percentage points for forced air cooling and 0.4±1.1 percentage points for natural cooling of non-tempered rice in the tested temperature range. The tempering treatment resulted in more moisture removal than the nontempering treatment with natural cooling and forced air cooling. The tempering process reduced the moisture gradient in the rice kernels and allowed moisture to equilibrate before the rice kernels were cooled. Without tempering, there was a significant moisture gradient in the rice kernels and a low MC near the surface, which resulted in less total moisture removal during cooling. In general, reduced moisture gradient in the tempered rice kernels and forced air cooling increased moisture removal during the cooling process. Therefore, the tempering process is a critical step in increasing moisture removal during cooling. In order to achieve high moisture removal during cooling, a combination of tempering and forced air cooling could be used, even though excessive moisture removal could cause rice fissures and lower the rice milling quality. The trend of total moisture removal at different temperatures with different tempering and cooling treatments was more or less parallel to the moisture removal caused by heating only (Figs 7.25 and 7.26). The highest total MC removals from the rice were 1.7±4.4 and 2.2±4.8 percentage points for low and high MC rice samples, respectively, which were achieved with tempering and forced air cooling as the treatments. The lowest total MC removal generally occurred for rice with no tempering and a natural cooling treatment. For rice treated with tempering and natural cooling, the total moisture removal was 1.4, 2.4, 3.2 and 4.3 percentage points for the high MC rice and 1.3, 2.0, 2.7 and 3.8 percentage points for the low MC rice over the tested temperature range. The moisture removals were the second highest among the treatments when the temperatures were above 55 ëC. These numbers indicated that 2.7±3.2 percentage points of moisture were removed with 1 min heating followed by tempering and natural cooling. The drying rates were much higher than the 2 to 3 percentage point moisture removal with 15 to 20 min heating of the current



187

Fig. 7.25 Total moisture removal of rice with initial MC of 20.6% under different cooling methods with and without tempering (T ± tempering, NT ± no tempering, NC ± natural cooling, FAC ± forced air cooling).

conventional heated air drying. For total moisture removal, the moisture removed due to sensible heat during cooling was a very significant portion. For example, 37 and 44% of total moisture removal occurred during cooling when the low and high MC rice samples, respectively, were heated for 60 s (to about

Fig. 7.26 Total moisture removal of rice with initial MC of 25.0% under different cooling methods with and without tempering (T ± tempering, NT ± no tempering, NC ± natural cooling, FAC ± forced air cooling).


188


60 ëC), followed by tempering and natural cooling. Because no additional heating energy is needed during the cooling, the high moisture removal could further improve the energy efficiency of the IR drying process. The exact amounts of energy saving and consumption are subjects of future research. Rice milling quality In general, for both the high and low initial MC rice samples, IR dried rice with tempering followed by natural cooling had similar TRY that was higher than the controls (Figs 7.27 and 7.28). On average, the total rice yields (TRYs) of low and high MC rice dried using IR followed by natural cooling were 68.% and 68.1%, respectively, which were 0.3 and 0.7 percentage points more than the controls. In particular, the rice dried at about 60 ëC with natural cooling had the highest TRYs of 68.4% for low MC rice and 68.6% for high MC rice, compared to 67.7 and 67.4% for the respective controls. This meant that the TRYs of IR dried rough rice were 0.7 to 1.2 percentage points higher than the controls. However, samples treated by other methods had much lower TRYs than the controls, especially the rice with low MC dried at high temperature. Similar trends were also observed for the head rice yields (HRYs) (Figs 7.29 and 7.30). The low MC rice samples dried using IR with tempering and natural cooling had significantly higher HRY (0.6±1.9 percentage points) than the control, and the highest HRY of 65.2% was obtained at a rice temperature of 61.2 ëC. For the high MC rice, the rice dried followed by tempering and natural cooling had the same HRY (63.6%) at 58.8 ëC as the control and slightly lower HRY at 42.8 ëC and 55.5 ëC than the control. All other post-heating treatments resulted in much lower HRYs.

Fig. 7.27 Total rice yields of rice with 20.6% initial moisture content and different drying treatments (T ± tempering, NT ± no tempering, NC ± natural cooling, FAC ± forced air cooling).



Fig. 7.28

189

Total rice yields of rice with 25.0% initial moisture content and different drying treatments.

When the results for the WI of the milled rice were examined, it could be seen that the IR dried rice generally had higher WI values than the controls, especially for the low MC rice, even though the differences between the controls and some of the treated rice samples were not significant (Figs 7.31 and 7.32). The results indicated that most of the IR dried rice with tempering followed by natural cooling had a similar milling degree to the control. It seems that there is

Fig. 7.29 Head rice yields of rice with 20.6% initial moisture content and different drying treatments.


190


Fig. 7.30 Head rice yields of rice with 25.0% initial moisture content and different drying treatments.

a trend that WI increased with an increase in the rice drying temperature for the non-tempering treatments, especially for the low MC rice. This could be due to the difference in the hardness of rice subjected to different treatments and/or the contribution of broken kernels to the color.

Fig. 7.31 Whiteness of milling rice with 20.6% initial moisture content and different drying treatments.



Fig. 7.32

191

Whiteness of milling of rice with 25.0% initial moisture content and different drying treatments.

Based on the milling quality results, it can be concluded that rough rice can be dried using IR followed by tempering and natural cooling to improve rice milling quality. Rice temperature with IR heating should be controlled at close to or below 60 ëC. For the current rice drying practice, the drying temperature or heated air temperature is controlled significantly below 60 ëC to avoid creating fissures and lowering the HRY. The high temperatures generated with IR heating did not damage the rice quality, probably because the relatively uniform heating resulted in lower moisture gradients compared to conventional heated air drying. The results indicate that the rice milling quality can be uncompromised and a relatively large amount of moisture can be removed with a single drying pass at a high drying rate, because the rice is heated quickly and uniformly, thereby minimizing the moisture gradient. When a large amount of moisture is removed by IR heating, tempering becomes increasingly important to reestablish the moisture equilibrium in the rice kernels. The study also showed that the cooling method following the tempering was important. Rapid cooling using forced air can significantly lower the rice milling quality. Because a relative large amount of moisture was removed during forced air cooling, the cooling might re-generate significant moisture and temperature gradients causing fissures. Based on the glass transition hypothesis, the temperature and moisture at the rice surface were lowered first, and the starch reached a glassy state during cooling (Cnossen et al., 2000). At the same time, the temperature and moisture at the centers of the rice kernels were still relatively high, and the starch remained in a rubbery state. The differences in the thermo-mechanical properties of the starch at different stages would generate stresses and fissures, resulting in breakage during milling and a lower rice milling quality. Therefore, controlled slow cooling will be very important for


192


high temperature rice drying. Since the natural cooling effectively preserved the quality, controlled slow cooling could be accomplished by low rates of air flow through a bin of rice. Effectiveness for disinfestations The disinfestation results clearly showed adult beetles were more heat resistant than adult moths (Tables 7.11 and 7.12). The 60 and 90 s heating times resulted in the deaths of moths in all stages in the rice at both initial MCs tested. With only a few adult moths surviving the low temperature treatments. It was also observed that some adult moths developed from the eggs or first-stage larvae during the incubation of the low MC rice that had a 25 s heating treatment. For beetles, 90 s of heating, regardless of tempering, and 60 s of heating with tempering achieved a near 100% kill rate, although a total of four inactive beetles were found in all the samples under such treatments. With the low temperature treatments, significant numbers of live adult beetles were discovered during the first week of the incubation, which were believed to be adult beetles that survived the treatments. The results obtained agreed with reported results that the time to death of the insects was less than 1 min when they were heated to a temperature above 62 ëC (Banks and Fields 1995; Fields and Muir, 1996). Non-tempered samples, especially at low temperatures, had fewer insects developing during incubation than tempered samples. This could be due to Table 7.11 Numbers of live moths in the rice samples with different drying treatmentsa Harvest MC (%)

Heating time (s)

Rice Tempering temperature ( ëC)

Days of storage after treatment 1b

5

8

15

27

32

34

20.6%

90 90 60 60 40 40 25 25

69.4 69.4 61.3 61.3 54.3 54.3 49.0 49.0

Yes No Yes No Yes No Yes No

0 0 0 0 0 0.5 0 0

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 8 2.5 0 0 0 0.5 17.5 3.5 0 0.5 0

25.0%

90 90 60 60 40 40 25 25

68.0 68.0 59.1 59.1 55.5 55.5 49.0 49.0


0 0 0 0 0.5 0 1 0

0 0 0 0 0.5 0 1 0

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0

a Numbers are the average numbers of insects recovered from two samples at each treatment condition b Numbers of insects that survived the thermal treatment


The potential of novel infrared food processing technologies Table 7.12 Harvest MC (%)

193

Numbers of live beetles in rice samples with different drying treatmentsa

Heating time (s)

Rice Tempering temperature ( ëC)

Days of storage after treatment 1b

5

8

15

27

32

34

20.6%

90 90 60 60 40 40 25 25

69.4 69.4 61.3 61.3 54.3 54.3 49.0 49.0


0 0 0 1 0 0 0 0.5 26 51 0.5 0 45.5 54.5 50.0 44.5

0.5 0 0 0 1 0 3.5 2

0 0 0 0 1 0 1.5 0.5

0.5 0 0 0 0 0 0.5 0.5

0 0 0 0 0.5 0 0 0

0 0 0 0 0 0 0 0

25.0%

90 90 60 60 40 40 25 25

68.0 68.0 59.1 59.1 55.5 55.5 49.0 49.0


0 0 0 0 0 0 2 4.5 26 51 0 0 58.5 67.5 29.5 48.5

0 0 0 0.5 1 0 2.5 0.5

0 0 0 0 1.5 0 1.5 1

0 0 0 0 0 0 2 1

0 0 0 0 0.5 0 0 0

0 0 0 0 0 0 0 1

a Numbers are the average numbers of insects recovered from two samples at each treatment condition b Numbers of insects that survived the thermal treatment

cooling shock in the non-tempered samples that reduced the survival capability of the insects after IR treatment, which needs to further studied. Based on the disinfestation results, heating rice to 60 ëC followed by tempering will achieve complete disinfestation of moths and beetles. However, rice samples heated to 60 ëC followed by tempering also had a high rice milling quality, and we conclude that IR heating appears to be useful for simultaneous drying and disinfestation of freshly harvested rough rice. 7.9.6 Concluding remarks on IR drying and disinfestation of freshly harvested rough rice High drying temperatures of rice can be achieved in a relatively short heating time using a catalytic IR emitter with a single layer of rough rice. The moisture removal during heating increased with an increase in rice temperature. It took only 60 s to achieve a rice temperature of about 60 ëC and removal of 1.7 and 1.8 percentage points MC during IR heating alone for the low and high MC rice, respectively. The tempering process after the rapid IR heating is essential to achieve high rice milling quality and improve the amount of moisture removal during cooling. Natural cooling following the tempering treatment can be used to remove a significant amount of moisture while retaining a high rice milling quality, but forced air cooling following heating or tempering can lower rice


194


milling quality, which is not recommended. The recommended conditions for simultaneous drying and disinfestation of freshly harvested rice are a 60 ëC rice temperature followed by tempering and slow cooling.

7.10 Effectiveness of infrared (IR) heating for disinfestation of stored rough rice The goal of this part of the study was to develop rapid, non-chemical, safe alternative methods to eliminate insect pests from stored rough rice while retaining high rice quality. The specific objectives were as follows: 1. To determine the effectiveness of IR heating treatments on the disinfestation of stored rough rice. 2. To investigate the effects of IR heating treatments on moisture loss and rice milling quality. The tests were conducted with two different approaches: thick-layer heating and single-layer heating using IR dryers. 7.10.1 Approaches to study effectiveness of IR heating for disinfestation of stored rough rice Material and methods of quality evaluation in the thick-layer heating treatment Two different stored California medium grain M202 samples were used for this study. A rice sample naturally infested with Angoumois grain moth (Sitotroga cerealella) and 12.9% MC was used for disinfestation tests. The moisture loss of rice samples under different IR treatments were determined. A rice sample with 13.5% MC was obtained from Farmer's Rice Co-operative (West Sacramento, CA) and used for milling quality evaluation. MC and milling quality were determined using standard Federal Grain Inspection Service methods. The evaluated quality indicators were TRY, HRY, and WI. The catalytic IR heating device (Fig. 7.33) was used for the tests. Since the IR radiation directly heats the rice without heating the surrounding air, the air temperature inside the heating chamber was significantly lower than the heated rice. Therefore, the rice bed temperature was used to control temperature. Rice temperature was measured using two thermocouples inserted in the middle of the rice bed, and the average of the thermocouple readings was used to control the natural gas supply to the IR emitter (switched on or off) by a control system that compared the average bed temperature with a pre-determined set point. The heating times needed to reach the set point temperatures were recorded. Then the samples were kept in the heating chamber for the desired time periods. The sample size for the disinfestation treatments was 2 kg per batch, which corresponded to about a 2 cm rice bed layer thickness. After completing the desired treatment time, the samples were taken out of the heating chamber and saved for moisture and disinfestation evaluation. The experimental design is



Schematic diagrams and set-up of (a) catalytic vibro-bed infrared dryer and (b) conventional heated air dryer for rice drying.

196


Table 7.13 Experimental design of infrared heating treatment of thick-layer rice Temperature (ëC) 45 50 60 70

Heating time (min) 1 1 1 1

5 5 5 5

10 10 10 10

shown in the Table 7.13. Based on the results of the disinfestation tests, four treatment conditions (50 ëC for 1 or 5 min, and 60 ëC for 1 or 5 min) were used to produce samples for the milling quality evaluation. After IR heating, all rice samples were transferred to plastic containers or glass jars with screen on lids to maintain moisture and oxygen exchange for surviving insects, larvae or eggs to grow. These containers were kept inside an incubator at 80% RH and 28 ëC for an observation period of up to about 42 days. Insect populations at each observation period were determined by counting the number of emerging adults in each rice sample (both treated and control) every 2±3 days of the entire observation period. Cumulative numbers of emerging adults as a function of time were then calculated and reported. The results showed the disinfestation effectiveness of the treatments on stored rice infested with Angoumois grain moth. If no live adult insects were observed after 1 or 2 insect life cycles (about 21 days for each cycle), the treatment conditions were considered as effective. After each observation and counting, all adult insects were removed. Materials and methods for single-layer heating treatment Stored rough rice, medium grain rice, M202, with MC of 11.0% was obtained from Pacific International Rice Mills, Inc. (Woodland, CA). Rice samples of 250 g were infested with 100 adult lesser grain borers (beetles), Rhizopertha dominica, and 50 adult angoumois grain moths, Sitotroga cerealella, at 18 and 6 days before the thermal treatment to produce larvae and eggs of the insects in the samples. At 18 days before the IR treatment, the adult insects were mixed with the rice samples, kept for two days, then manually removed by sifting and hand picking. It was expected that the eggs laid by the adult insects during the two days would become larvae at the time of thermal treatment. At six days before the treatment, the same numbers of adult insects were put into the infested rice samples and kept until the IR treatment. The infested rice was kept in an incubator for more insects to emerge. In order to reduce the moisture loss during disinfestation treatment, the infested, stored rice samples were heated as single-layer using an IR emitter with the radiation intensity of 5300 W mÿ2 and five exposure times from 10 to 30 s. The rice load rate was 2 kg mÿ2. To reduce the heating time, the drying bed was pre-heated to temperatures close to the target rice temperatures before sample loading. The final temperature of the heated rice was in the range of 46±67 ëC,


The potential of novel infrared food processing technologies Table 7.14

197

Experimental design of infrared heating treatment of single-layer rice

Heating time (s)

Rice temperature ( ëC)

Holding time (min)

46 53 60 62 67

0, 5, 20, 60, 180 0, 5, 20, 60, 180 0, 5, 20, 60, 180 0, 5, 20 0

10 15 20 25 30

which was measured using an IR temperature sensor. After the heating treatments, the samples were held at the heated temperature for various times up to 3 h, and then cooled gradually in a closed container to the room temperature, about 23 ëC. The detailed experimental design is shown in Table 7.14. The disinfestation evaluation method was the same as the method used for thicklayer heating. The moisture losses of rice samples caused by heating were calculated based on the weight loss from the initial MC. Based on the disinfestation results, uninfested rice samples heated to temperatures of 46, 53, and 60 ëC were produced for milling quality evaluation. The quality evaluations were conducted at Pacific International Rice Mills, Inc. (Woodland, CA) and Farmer's Rice Cooperative (West Sacramento, CA) based on the methods used for the thicklayer rice heating treatment. 7.10.2 Results of IR disinfestation of stored rough rice Results of IR disinfestation under thick-layer treatment When rice samples were heated in the heating chamber, it took about 2, 3, 4 and 5 minutes to reach 45, 50, 60, and 70 ëC, respectively. The heating was quite rapid and could be further improved, if a thinner layer is used, which may reduce the moisture loss during the treatment (Table 7.15). Moisture losses were in range of 0.59±2.86% under the tested conditions. When the treatment was 50 ëC and 1 min, the moisture loss was about 1%. Table 7.15 Moisture content of thick-layer rice sample treated with infrared at different conditions (% w.b.) Temperature (ëC) Control 45 50 60 70

0 12.92

Treatment time (min) 1 5 12.33 11.87 11.71 10.91

11.87 11.75 11.24 10.87


10 11.78 11.45 10.79 10.06

198


Fig. 7.34 Emerging adult insects in infrared treated thick-layer samples (no insects found for the samples treated at 50 ëC or above).

The disinfestation results showed that only control and treated samples at 45 ëC had emerging adult insects (Fig. 7.34). No insects were found in the rest of samples after two insect life cycles. These results indicate that temperatures above 50 ëC effectively kill all forms of Angoumois grain moth (Sitotroga cerealella). The minimum treatment under the test conditions was 50 ëC and 1 min. The total treatment time including heating was about 4 min with about 1% moisture loss. The milling quality of rice samples treated with IR at 50 ëC for 1 or 5 min was not affected compared to the control sample (Table 7.16). No difference in whiteness was observed between milled rice samples treated at 50 ëC and the control, however, significant quality loss occurred for the rice samples treated at 60 ëC. Through optimization of the treatment conditions, the IR heating could be an effective method for stored rice disinfestation without quality loss, and the moisture loss could be minimized. Table 7.16 Moisture change and milling quality of infrared treated California medium grain M202 rice samples Condition Control 50 ëC ± 1 50 ëC ± 5 60 ëC ± 1 60 ëC ± 5

MC (%) min min min min

13.5 12.3 12.0 11.9 11.4

0.1 0.0 0.0 0.1 0.1

TRY (%) 68.5 69.2 69.5 69.8 70.3

0.4 0.4 0.2 0.4 0.4

HRY (%) 54.1 53.8 54.6 52.4 44.9


0.7 0.6 0.1 0.0 0.4

WI 44.4 45.1 44.7 45.1 44.5

0.3 0.3 0.3 0.4 0.2


199

Results of IR disinfestation under single-layer treatment Since the single-layer heating was used, the required heating time to reach certain temperature was significantly less compared to thick-layer treatments. It took only 20 s to reach 60 ëC, which meant that the heating rate was very high (Fig. 7.35). Due to the reduced heating time, the moisture loss was also significantly reduced. For example, the moisture loss was only 0.53% when the rice sample was heated to the temperature of 60 ëC. In the tested temperature range, 46±67 ëC, the moisture losses were in the range of 0.28±0.76%. The results meant that single-layer heating was a better method for reducing the moisture loss caused by IR disinfestation treatment compared with the thick-layer heating. The disinfestation results of stored rice are shown in Tables 7.17 and 7.18. No live adult moths were found in all treated samples during the first 14 days. For storage times of 21 days or longer, live moths appeared for all treatments at 46 ëC or 53 ëC with no holding and with 5 min holding, which may indicate that some insect eggs survived the thermal treatments at those conditions. For beetles, it was clear that treatment temperatures at 53 ëC or below could not completely kill the adult beetles. It seems that 60 ëC treatment was effective even though the treatment with 5 min holding recovered two unhealthy live beetles in the three samples. Very few live beetles from the 46 ëC treated samples were recovered during incubation. Such results may indicate that adult beetles were more heat resistant than the insect in other forms, such as eggs and larvae, which was different from the moths. The disinfestation results also showed that the beetles were more heat resistant than the moths.

Fig. 7.35 Stored rice temperature and moisture loss after infrared single-layer heating treatment.


200


Table 7.17 Numbers of live moths in single-layer rice samples treated with infrared heatinga Rice temperature (ëC) 67 62 62 62 60 60 60 60 60 53 53 53 53 53 46 46 46 46 46

Holding time (min)

1b

0 0 5 20 0 5 20 60 180 0 5 20 60 180 0 5 20 60 180

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Days of storage after treatment 14 21 27 31 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3 0 0.7 2.0 0.3

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3 4.3 6.7 3.0

0 0 0 0 0 0 0 0 0 0.3 0.3 0 0 0 1.3 8.0 4.3 4.7 1.7

35 0 0 0 0 0 0 0 0 0 0 0 0 0 0 13.0 9.3 2.7 4.3 1.0

a

Numbers are the average numbers of insects recovered from three samples at each treatment condition b Numbers of insects that survived the thermal treatment

The milling qualities of rice samples treated with the temperatures from 46 to 60 ëC with 0±3 h holding times are shown in Figs 7.36±7.38. The IR treatments reduced the TRY 0.2 to 1.0 percentage points, but the HRY slightly increased, except for the treatment of 60 ëC without holding. Since the WI of treated rice samples were 0.5±0.7 unit higher than the control, the TRY of treated samples and control could be very similar, if the samples were milled to the similar whiteness. Therefore, it is reasonable to believe that IR disinfestation treatments did not significantly affect the rice milling quality, except for the treatment of 60 ëC without holding. The holding was necessary for the 60 ëC treatment to reduce the quality losses. 7.10.3 Concluding remarks on IR disinfestations of stored rough rice IR heating could be used to disinfest stored rough rice. For thick-layer treatment, the required temperature and time for killing all moths were 50 ëC and holding for 1 min, since the heating took 3 min. Under such treatments, rice milling quality was unaffected, and there was about a 1% moisture loss. For single-layer treatment, the minimum treatments were 53 ëC with 20 min holding for moths and 60 ëC with 20 min holding for beetles. For stored rice, IR disinfestation


The potential of novel infrared food processing technologies Table 7.18 heatinga

Numbers of live beetles in single-layer rice samples treated with infrared

Rice temperature (ëC) 67 62 62 62 60 60 60 60 60 53 53 53 53 53 46 46 46 46 46

201

Holding time (min)

1b

0 0 5 20 0 5 20 60 180 0 5 20 60 180 0 5 20 60 180

0 0 0 0 0 0.7 0 0 0 2.3 2 3.7 0 2 67.0 64.7 52.7 60.0 69.3

Days of storage after treatment 14 21 27 31 0.3 0 0 0 0 0 0 0 0 0 0 0 0.7 0 0.7 1.3 2.0 3.7 4.0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3 2.0 1.0 1.7 0

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.3 0 0.3 0

0 0 0 0 0 0 0 0 0 0.3 0 0 0 0 0 0 0.7 0.3 0

35 0 0.3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.7 0 0.7

a

Numbers are the average numbers of insects recovered from three samples at each treatment condition b Numbers of insects that survived the thermal treatment

Fig. 7.36

Total rice yields of single-layer rice treated at different temperatures with and without holding.


202


Fig. 7.37 Head rice yields of single-layer rice treated at different temperatures with and without holding.

caused about 0.53% moisture loss. The 60 ëC temperature of single-layer stored rice can be achieved with 20 s of heating when the drying bed was pre-heated to the targeted temperature. The IR treated stored rice had similar milling qualities compared to the corresponding control samples.

Fig. 7.38 Whiteness index of single-layer rice treated at different temperatures with and without holding.



7.11

203

Infrared (IR) radiation heating for tomato peeling

Hot lye peeling and steam peeling are the commercial techniques adopted by the fruit and vegetable processing industry. Lye peeling is the widely industrialized method for producing high quality peeled fruit and vegetable products, however, lye peeling adversely impacts the environment and related industries. The associated burden related to salinity management of wastewater and escalated cost and threat to long-term supply of water have made this technology unattractive to processors. On the other hand, steam peeling results in inferior products such as losses in product appearance, firmness and yield (Pan et al., 2009c). Owing to the heating characteristics of IR derived from its limited penetration capability, IR could be a suitable method for peeling fruits and vegetables. Hart et al. (1970) and Sproul et al. (1975) studied IR peeling of white potatoes and peaches to significantly reduce peeling losses, wastewater generation, and the use of caustic lye. Pan et al. (2009c) described the development of an IR peeling method for tomatoes as an alternative to lye and steam peeling in a cost-effective and environmentally friendly way. Their IR heating system is equipped with catalytic IR emitters powered by natural gas. Pan et al. (2009c) reported the improvements of IR dry-peeling over lye, sequential lye-IR or sequential enzyme-IR peeling methods. Table 7.19 shows a sample of their findings comparing the IR dry-peeling and lye peeling of tomatoes. IR dry-peeling of tomatoes significantly reduced peeling losses compared to hot lye peeling. The treatment using lye and enzyme as the pretreatment of IR peeling did not provide any advantageous synergistic effects over IR peeling alone. The sequential enzyme-IR peeling indicated easier peeling, but resulted in much higher peeling losses and longer treatment times compared to those obtained with IR peeling alone. The sequential lye-IR peeling also had Table 7.19 Effects of heating time on tomato peeling with lye and infrared heating for tomato Sun6366 Methods and conditions Lye10 Lye10 Lye10 Lye10 IR12 IR12 IR12 IR12

± ± ± ±

± ± ± ±

30 45 60 75

30 45 60 75

s s s s

s s s s

Peelability (cm2/g)

Ease of peeling

Peeling loss (%)

0.004 0.008 0.004 0.003

3.5a 4.1a,b 4.7b 4.9c

11.46 11.68 13.37 13.55

0.020a 0.004b 0.002b 0.002b

1.6a 3.0b 4.1c 4.6d

7.64 6.11 7.74 9.41

Peeled firmness (kg)

Surface temperature (ëC)

1.5a 1.3b 1.4b 1.3b

95 95 95 95

1.8a 1.8a 1.5a,b 1.6b

57.9a 66.2b 70.1c 76.8d

Note: Subscript 10 of Lye10 stands for the concentration of the lye solution. Subscript 12 of IR12 stands for the gap of the two IR emitters. Mean separation was via Duncan's Multiple Range Test. Means with a different letter (a b c d) in each section are significantly different at the 0.05 level.


204


significantly higher peeling losses compared to IR dry-peeling under the same conditions. The results indicate that IR dry-peeling has the promising potential to be an alternative to lye use and the water resource crisis for tomato processors.

7.12

Future trends

Consumer and industry demand to improve the quality of agricultural products and to achieve more economic processing operations has brought much attention to the replacement of conventional operations with a number of novel IR food processing and preservation methods. It has become a priority in the food industry to develop novel and sustainable processing technologies to reduce wastewater disposal and chemical usage, address long-term water supply problems, and improve energy efficiency, while at the same time delivering high quality and safe processed food products. Consequently, IR-based technologies have emerged as potential solutions to a number of drawbacks of conventional methods, as illustrated above. In this chapter, we have succinctly covered several novel IR processing technologies we have recently developed, such as simultaneous IR dry blanching and dehydration of fruits and vegetables; combined and sequential IR freeze drying; sequential IR and hot air roasting of almonds; IR pasteurization of raw almonds; simultaneous IR rough rice drying and disinfestation; stored rice disinfestation; IR heating for tomato peeling. These new IR-based technologies have the potential to significantly and positively impact the food processing industry. Commercial equipment manufacturing is urgently needed to move these technologies from limited research and pilot scale applications to the commercial food processing industry to benefit consumers, the environment, and natural resources. Therefore, educating relevant stakeholders of the merits of these new technologies over the conventional counterparts should be given a great priority, to deliver these novel technologies, with their inherent technical advantages, to the marketplace.

7.13

Acknowledgements

The authors are extremely grateful to their co-workers and students, including Dr Tara McHugh, Dr Gokhan Bingol, Connie Shih, Dr Yi Zhu, and Xuan Li whose contributions and expertise were invaluable to the research and preparation of this chapter.

7.14

References and further reading and ABE, T. (1997) Combined convection and far-infrared radiation drying of rough rice. ASAE Paper No. 9760972. St. Joseph, MI.

AFZAL, T.M.



205

and ABE, T. (1998) Diffusion in potato during far infrared radiation drying. J. Food Eng. 37(4): 353±365. ALVAREZ, C.A., AGUERRE, R., GOMEZ, R., VIDALES, S., ALZAMORA, S. and GERSCHENSON, L. (1995) Air dehydration of strawberries: effects of blanching and osmotic pretreatments on the kinetics of moisture transport. J. Food Eng. 25(2): 167±178. ANTHON, G.E. and BARRETT, D.M. (2002) Kinetic parameters for the thermal inactivation of quality related enzymes in carrots and potatoes. J. Agric. Food Chem. 50, 4119±4125. AOAC (1994) Official Methods of Analysis, 14th edn. VA: Association of Official Analytical Chemists. ASAE STANDARDS (1995) S352.2: Moisture measurements-Unground grain seeds. Moisture relationships of grains (42nd edn). St. Joseph, MI: ASAE. BAKER, C.G.J. (1997) Industrial Drying of Foods. New York: Chapman & Hall. BANKS, H.J. and FIELDS, P.G. (1995) Physical methods for insect control in stored grain ecosystems. In Stored Grain Ecosystems. New York: Marcel Dekker, pp. 353±409. BAYSAL, T., ICIER, F., ERSUS, S. and YILDIZ, H. (2003) Effects of microwave and infrared drying on the quality of carrot and garlic. European Food Res. and Tech. 218(1): 68±73. BILOWICKA, E. (1960) Research concerning the drying of small seeds by infrared radiation. International drying conference. Warsaw, Poland: Institute of Mechanization and Electrification of Agriculture. CHENG, L.M. (1992) Food machinery for the production of cereal foods, snack foods and confectionary. New York: Ellis Horwood Series in Food Science and Technology. CHUA, K.J. and CHOU, S.K. (2003) Low-cost drying methods for developing countries. Trends Food Sci Technol. 14 (12): 519±528. CNOSSEN, A.G., SIEBENMORGEN, T.J., YANG, W. and BAUTISTA, R.C. (2000) The glass transaction temperature concept in rice drying and tempering: effect on milling quality. Trans. ASAE 43(6): 1661±1667. DAGERSKOG, M. and OSTERSTROM, L. (1997) Infra-red radiation for food processing I. A study of the fundamental properties of infra-red radiation. Lebensmittel Wissenschaft Technologie ± Food Science Technology 12(4): 237±242. DAS, I., DAS, S.K. and BAL, S. (2004a) Determination of mixing index of paddy grains under vibrating conditions. J. Food Process Eng. 26(1): 121±133. DAS, I., DAS, S.K. and BAL, S. (2004b) Specific energy and quality aspects of infrared (IR) dried parboiled rice. J. Food Eng. 62(1): 129±133. EPRI (1993) Technology Guidebook for Electric Infrared Process Heating, CMF Report No. 93-2. ERTEKIN, C. and YALDIZ, O. (2004) Drying of eggplant and selection of a suitable thin layer drying model. J. Food Eng. 63(3): 349±359. FIELDS, P.G. and MUIR, W.E. (1996) Physical control. In Integrated Management of Insects in Stored Products. New York: Marcel Dekker. FLINK, J. (1977) Energy analysis in dehydration processes. Food Technol. 31, 77±84. GINZBERG, A.S. (1969) Application of Infrared Radiation in Food Processing. London: Leonard Hill Books. HAMMAMI, C. and RENEÂ, F. (1997) Determination of freeze-drying process variables for strawberries. J. Food Eng. 32(2): 133±154. HART, M.R., GRAHAM, R.P., HUXSOLL, C.C. and WILLIAMS, G.S. (1970) An experimental dry caustic peeler for cling peaches and other fruits. J. Food Sci. 35(6): 839±841. INCROPERA, F.P. and DEWITT, D.P. (2002) Introduction to Heat Transfer, 4th edn. New York: John Wiley & Sons. AFZAL, T.M.


206


and PRACHAYAWARAKORN, (2007) Influences of drying medium and temperature on drying kinetics and quality attributes of durian chip. J. Food Eng. 78(1): 198±205. KETELAARS, A., JOMAA, W., PUIGALLI, J. and COUMANS, W. (1992) Drying shrinkage and stress. In Mujumdar, A.S. (ed.), Drying '92, Part A. Amsterdam: Elsevier, pp. 293± 303. KIRKPATRICK, R.L. (1975) Infrared radiation for control of lesser grain borers and rice weevils in bulk wheat. J. Kansas Entomol. Soc. 48(1): 100±104. KROKIDA, M.K., OREOPOULOU, V., MAROULIS, Z.B. and MARINOS-KOURIS, D. (2001) Effect of pre-treatment on viscoelastic behaviour of potato strips. J. Food Eng. 50(1): 11±17. KRUST, P.W., MCGLAUCHLIN, L.D. and MCQUISTAN, R.B. (1962) Elements of Infra-red Technology. New York: John Wiley & Sons. KUMAR, D.G.P., HEBBAR, H.U., SUKUMAR, D. and RAMESH, M.N. (2005) Infrared and hot-air drying of onions. J. Food Proc. and Preservation 29(2): 132±150. KUNZE, O.R. and CALDERWOOD, D.L. (1985) Rough rice drying. In Rice: Chemistry and Technology. St. Paul, MN: American Association of Cereal Chemists, pp. 233± 263. LI, G. and MA, Z. (2003) Vacuum freeze-drying process of strawberry. Food and Machinery (3): 18±19. LIN, T.M., DURANCE, T.D. and SCAMAN, C.H. (1998) Characterization of vacuum microwave, air and freeze dried carrot slices. Food Res. Int. 31(2): 111±117. LIN, Y., TSEN, J. and KING, V.A. (2005) Effects of far-infrared radiation on the freeze-drying of sweet potato. J. Food Eng. 68: 249±255. LIN, Y., LEE, T., TSEN, J. and KING, V.A. (2007) Dehydration of yam slices using FIR-assisted freeze drying. J. Food Eng. 79(4): 1295±1301. MCMINN, W. and MAGEE, T. (1997) Physical characteristics of dehydrated potatoes: Part II. J. Food Eng. 33(1): 49±55. MASAMURE, A., SADO, H., HODA, T., SHIMIZU, M., NABETANI, H., NAKAJIMA, M., ET AL. (1998) Drying of potato by far infrared radiation. Nippon Shokuhin Kogyo Gakkaishi 35(5): 309±314. MILLER, D.D. (1998) Food Chemistry: A Laboratory Manual. New York: John Wiley & Sons. MUJUMDAR, A.S. (1995) Handbook of Industrial Drying, 2nd edn. New York: Marcel Dekker. NIELSEN, S.S. (ED.) (1998) Food Analysis, 2nd edn. Gaithersburg, MD: Aspen Publishers, Inc. NOWAK, D. and LEWICKI, P.P. (2004) Infrared drying of apple slices. Innov. Food Sci. Emerg. Technol. 5, 353±360. OZDEMIR, M. and DEVRES, O. (2000) Analysis of color development during roasting of hazelnuts using response surface methodology. J. Food Eng. 45: 17±24. PAN, Z. (2006) Strawberry dehydration using sequential infrared radiation and freezedrying method. Dissertation, Shih Y. C., Biological and Agricultural Engineering, University of California, Davis. PAN, Z. and MCHUGH, T.H. (2004) Novel infrared dry-blanching (IDB), infrared blanching, and infrared drying technologies for food processing. In Pending, US Patent Application. 20060034981. Filed 8/13/2004, published 2/16/2006. PAN, Z., SOLAR, M.L. and YOKOHAMA, W.H. (2004) Rice utilization and product development. Annual comprehensive research report. PAN, Z., KHIR, R., GODFREY, L.D., LEWIS, R., THOMPSON, J.F. and SALIM, A. (2008a) Feasibility JAMRADLOEDLUK, J., NATHAKARANAKULE, A., SOPONRONNARIT, S. S.



207

of simultaneous rough rice drying and disinfestations by infrared radiation heating and rice milling quality. J. Food Eng. 84(3): 469±479. PAN, Z., SHIH, Y.C., MCHUGH, T. and HIRSCHBERG, E. (2008b) Study of banana dehydration using sequential infrared radiation heating and freeze-drying. LWT ± Food Sci. Technol. 41(10): 1944±1951. PAN, Z., BINGOL, G. and MCHUGH, T. (2009a) Test Results and Performance of Mobile Infrared Heating Equipment for Processing Fruits and Vegetables. USDA-ARSWRRC Report. PAN, Z., YANG, J., BINGOL, G. and MCHUGH, T. (2009b) Infrared Heating for Raw Almond Pasteurization. Final Report for Almond Board of California. PAN, Z., LI, X., BINGOL, G., MCHUGH, T.M. and ATUNGULU, G.G. (2009c) Development of Infrared Radiation Heating Method for Sustainable Tomato Peeling. Applied Engineering in Agriculture (Journal of American Society of Agricultural and Biological Engineering), 25(6): 935±941. RATTI, C. and MUJUMDAR, A.S. (2007) Infrared drying. In Handbook of industrial drying, vol. 1, 3rd edn. New York: Marcel Dekker, pp. 423±437. ROSENTHAL, I. (1992) Electromagnetic radiations in food science. Adv. Series Agric. Sci. 19. SAKAI, N. and HANAZAWA, T. (1994) Applications and advances in far-infrared heating in Japan. Trends Food Sci. Technol. 5: 357±362. SANDU, C. (1986) Infrared radiative drying in food engineering: a process analysis. Biotechnol. Progr. 2(3): 109±119. SHARMA, G.P., VERMA, R.C. and PATHARE, P.B. (2005) Thin-layer infrared radiation drying of onion slices. J. Food Eng. 67(3): 361±366. SHIH, C. and PAN, Z. (2006) Strawberry dehydration using sequential infrared radiation and freeze-drying method. Dissertation, Biological and Agricultural Engineering, University of California, Davis. SHIH, C., PAN, Z., MCHUGH. T., WOOD, D. and HIRSCHBERG, E. (2008) Sequential infrared radiation and freeze-drying method for producing crispy strawberries. Trans. ASABE 51(1): 205±216. SHISHEHGARHA, F., MAKHLOUF, J. and RATTI, C. (2002) Freeze-drying characteristics of strawberries. Drying Technol. 20(1): 131±145. SIEGEL, R. and HOWELL, J.R. (2001) Thermal Radiation Heat Transfer, 4th edn. Philadelphia, PA: Taylor and Francis, 419±429. SINGH, K.K. (1994) Development of a small capacity dryer for vegetables. J. Food Eng. 21: 19±30. SINGH, R.P. and HELDMAN, D.R. (1993) Introduction to Food Engineering, 2nd edn. San Diego, CA: Academic Press. SPROUL, O., VENNES, J., KNUDSON, W. and CYR, J.W. (1975) Infrared dry caustic vs. wet caustic peeling of white potatoes. Environmental Protection Technology Series. Corvallis, OR: National Environmental Research Center, Office of Research and Development, US Environmental Protection Agency. STIPE, D.R., WRATTEN, F.T. and MILLER, M.F. (1972) Effects of various methods of handling brown rice on milling and other quality parameters. Louisiana Agricultural Experiment Station Annual Program Rep., Rice Exp. Stn 113. TILLER, F.M. and GARBER, H.J. (2002) Infrared radiant heating. Ind. Eng. Chem. 34(7): 773± 781. TYREE, M.T. (1970) The symplast concept: a general theory of symplastic transport according to the thermodynamics of irreversible processes. J Theoret Biol. 26: 181±214.


208


(1994) Rice Inspection Handbook. Washington, DC: Agricultural Marketing Service. ZHU, Y. (2007) Processing and Quality Characteristics of Apple Slices under Simultaneous Infrared Dry-blanching and Dehydration (SIRDBD). PhD dissertation submitted in partial satisfaction of the requirements for the degree of doctor of philosophy in food science and technology in the office of graduate studies of the University of California Davis. ZHU, Y. and PAN, Z. (2009) Processing and quality characteristics of apple slices under simultaneous infrared dry-blanching and dehydration with continuous heating. J. Food Eng. 90(4): 441±452. ZHU, K., ZOU, J., CHU, Z. and LI, X. (2002) Heat and mass transfer of seed drying in a two pass infrared radiation vibrated bed. Heat Transfer ± Asian Research 3(12): 141±147. ZHU, Y., PAN., Z., MCHUGH, T.H. and BARRETT, D. (2010) Processing and quality characteristics of apple slices processed under simultaneous infrared dry-blanching and dehydration with intermittent heating. J. Food Eng. 97(1): 8±16. USDA FEDERAL GRAIN INSPECTION SERVICE


8 Validation and commercialization of dense phase carbon dioxide processing for orange juice K.-L. G. Ho, Chiquita Brands International Inc., USA

Abstract: Dense phase carbon dioxide is a non-thermal processing alternative with many advantages for the pasteurization of juice. This chapter discusses the validation, scale-up, and commercialization of the first dense phase carbon dioxide processing system in the United States. Process validation and shelf life studies showed that dense phase carbon dioxide processing is capable of maintaining the freshly squeezed quality of orange juice while meeting the 5-log reduction on pertinent pathogens as mandated by the Food and Drug Administration Juice Hazard Analysis Critical Control Point Regulation. Results of the pilot model, the prototypes, and the commercial system demonstrated that the dense phase carbon dioxide system is scalable in terms of size and performance. Key words: dense phase carbon dioxide, orange juice, process validation, system scale-up.

8.1

Introduction

More and more consumers from all over the world are looking for healthy beverages such as freshly squeezed juices or 100% juice blends. According to the United States Department of Agriculture's (USDA's) 2008 report, between 1970 and 2001 there was a 33% increase in fruit-juice consumption per-capita. Fresh juices are becoming increasingly popular among consumers because juices taste fresher and contain the vitamins and minerals originally in the fruit to give consumers energy, nutrition, and various health benefits. Some common


210


health benefits of juices are vitamin C and folic acid in orange juice (Dubost 2008), antioxidants in pomegranate juice (Seeram et al., 2008), the digestive health benefits of prune juice (Stacewicz-Sapuntzakis et al., 2001; Piirainen et al., 2007), the prevention of bladder infection from cranberry juice (Regal et al., 2006), and the prevention of age-related diseases such as atherosclerosis and diabetes by Goji juice (Potterat and Hamburger, 2008). Commercially, thermal processes such as pasteurization and canning are the prevalent methods used to improve the shelf life of juices. These conventional food processes use heat to inactivate spoilage microorganisms and indigenous enzymes that cause spoilage. Thermally processed juice, while stable during the target shelf life, shows dramatically decreased sensory attributes and nutritional content compared to freshly squeezed orange juice due to the thermal treatment. In the early 1990s, juice producers began manufacturing and marketing more and more un-pasteurized raw juice and apple cider to meet a growing consumer trend of seeking improved organoleptic qualities in foods. However, the consumption of unpasteurized cider and juice resulted in a number of outbreaks of Salmonella and Eschericia coli O157:H7. According to the US Food and Drug Administration Final Rule to Increase Safety of Fruit and Vegetable Juice about 16 000 to 48 000 estimated cases of illnesses were related to unpasteurized juice each year (FDA, 2001a). In response to the rise in outbreaks across North America associated with the consumption of un-pasteurized juices and cider, FDA issued a Juice Hazard Analysis and Critical Control Point (HACCP) regulation (21 CFR 120) designed to improve the safety of juice products. In this FDA (2001b) `Procedures for the safe and sanitary processing and importing of juice; Final Rule' juice processors are required to analyze the manufacturing process and decide whether there are any microbiological, chemical, or physical hazards that could contaminate their products. When a potential hazard is identified, the processor is required to implement control measures to prevent, reduce, or eliminate the hazard. In terms of product safety, processors are also mandated to use processes that consistently produce at least a 5-log10 reduction of the pertinent microorganisms for a period that is greater or equal to the shelf life of the product stored under normal and moderate temperature abuse conditions. Alternative non-thermal processes such as high-pressure (HP) processing, dense phase carbon dioxide (DPCO2) processing, pulse electric field (PEF) processing, and ultraviolet (UV) radiation processing have been gaining considerable interest to be used as means of processing fresh squeezed juice to increase flavor and nutrient retention, while also inactivating enzymes, spoilage microorganisms, and pathogens that can limit shelf life or compromise food safety. To implement these innovative technologies in the juice industry, process validation from bench top to pilot model, from pilot model to prototype, and from prototype to commercial-scale systems is key to ensuring safe, wholesome juice production (Koutchma et al., 2005).


Dense phase carbon dioxide processing for orange juice

8.2

211

Dense phase carbon dioxide processing

Carbon dioxide (CO2) has a long history of applications in the food industry because CO2 is nontoxic, nonflammable, inexpensive, and odorless. CO2 has generally recognized as safe (GRAS) status (Select Committee on GRAS Substance Database, 2006). Figure 8.1 is a pressure-temperature phase diagram of carbon dioxide (Allam et al., 2003). It shows that when gaseous or liquid CO2 is heated and compressed above its critical point (31 ëC and 72.6 atm) it becomes a dense and highly compressible fluid called supercritical CO2 that demonstrates properties of both liquid and gas. Dense phase carbon dioxide (DPCO2) is a collective term for liquid CO2 and supercritical CO2. The low viscosity of DPCO2 allows it to penetrate efficiently into tiny pores and crevices. This property enables DPCO2 to have a higher diffusion coefficient and to act as a better solvent than gaseous CO2 (Mathews et al., 2001). 8.2.1 Dense phase carbon dioxide (DPCO2) microbial and enzymatic inactivation efficacy Research has shown that DPCO2 possesses anti-microbial activities (Daniels et al., 1984; Kamihira et al., 1987; Haas et al., 1989; Dillow et al., 1999). It is effective in killing vegetative pathogens like E. coli (Ballestra et al., 1996; Kim et al., 2007; Liao et al., 2007, 2008), Salmonella typhimurium (Garcia-Gonzalez et al., 2009), and Listeria monocytogenes (Lin et al., 1994) and other microorganisms present in the juice (Arreola, 1991a; Lin et al., 1992; Ho, 2005). One

Fig. 8.1

Pressure-temperature phase diagram for CO2. DPCO2 is a collective term for liquid CO2 and supercritical CO2 (Allam et al., 2003).


212


Fig. 8.2

Schematic diagram showing the migration of CO2 into the cytoplasm and the reactions of carbon dioxide and water inside the cell.

of the most widely accepted mechanisms that are used to explain the bactericidal effect of DPCO2 is cell membrane damage. When CO2 diffuses into the cell membrane, it increases the fluidity of the membrane and causes the formation of pores in the membrane (Hong et al., 2001; Garcia-Gonzalez et al., 2007). These changes are irreversible and they generally result in the leakage of cytoplasmic materials and death of the cell. The second generally accepted mechanism involves the lowering of intracellular pH. The permeability of the membrane to CO2 allows migration of significant amounts of CO2 into the cytoplasm. Inside the cell, as shown in Fig. 8.2, CO2 reacts with water to form carbonic acid and its conjugate ions. The environment of the cytoplasm favors the dissociation of carbonic acid molecules into bicarbonate ions and hydrogen ions. In order to avoid the hydrogen ions from lowering the pH of the cytoplasm, the cell devotes large amounts of energy in pumping the excess hydrogen ions out of the cell. As the influx of the CO2 continues, the cell can not keep up with the energy requirement for exporting the hydrogen ions, thus resulting in the accumulation of hydrogen ions and eventual lowering of the pH of the cytoplasm. This pH change significantly hinders the metabolic activities and key intracellular enzymatic systems of the microorganisms (Hong and Pyun, 1999; Hong et al. 1999). In addition, studies have shown that DPCO2 is effective in inactivating various enzymes that are present in foods such as lipoxygenase (Liao et al., 2009), pectinesterase (Balaban et al., 1995; Truong et al., 2002; Zhi et al., 2008; Zhou et al., 2009), peroxidase, and polyphenol oxidase (Liu et al., 2008). The DPCO2 effects on these types of enzymes are critical to the quality of juice, as off-flavors, enzymatic browning, and sedimentation may result from the activity of the indigenous enzymes that are present in the liquid food. The mechanisms for inactivating these enzymes are not as well investigated as those relating to the bactericidal efficacy of DPCO2. Some research suggested that DPCO2 inactivates extracellular enzymes by altering their molecular properties. Specifically, Liao et al. (2009) observed with transmission electron microscopy



213

the aggregation of lipoxygenase molecules caused by DPCO2, with the DPCO2treated lipoxygenase molecules showing a dramatic decrease in their alpha-helix content. 8.2.2 DPCO2 processing patents and systems Numerous patents involving the use of DPCO2 in treating liquid food have been issued. Friedrich et al. (1985) and Christianson and Friedrich (1985) patented processes using pressurized CO2 to inactivate lipoxygenase in soy and peroxidase in corn germ. US Patent No. 5,704,276 by Osajima et al. (1997) indicates a method for the continuous inactivation of enzymes in liquid food, using supercritical CO2. US Patent No. 5,393,547 and 6,723,365 by Balaban et al. (1995) and Balaban (2004), respectively, describe the reduction of microbial populations and inactivation of indigenous enzymes in liquid food by a batch or continuous DPCO2 processing system. Sims (2001) patented a DPCO2 processing system that destroyed microorganisms and enzymes in juices (US Patent No. 6,331,272). The system involves a membrane with minute pores that allows the separate flow paths of the DPCO2 and the liquid food to contact each other in a non-dispersive manner. Presently, there are only a few pilot-model scale DPCO2 processing systems available for treating liquid food continuously. The most recently built system in the United States was made by the Food Safety Intervention Technologies Unit team of the Eastern Regional Research Center of the USDA in Wyndmoor, Pennsylvania, US. The system is equipped with a gas-liquid porous metal contactor that has been reported to enhance the mixing of DPCO2 with the liquid food under pilot-plant testing (Yuk et al., 2008). In 2003 Mitsubishi Kakoki Co. (Tokyo, Japan) manufactured a pilot-scale DPCO2 processing system based on the patents owned by Shimadzu Co. (Kyoto, Japan). The maximum flow rate of the incoming DPCO2 and liquid food is 3.0 kg/h and 20 kg/h, respectively, and the system is restricted to laboratory studies only (Osajima et al., 1997; Shimoda et al., 1998). In 2002, PoroCrit LLC (Berkeley, CA, USA) based on the Sims patent (2001) built a DPCO2 processing system that is equipped with a membrane contactor made of several hollow fiber membrane modules for treating liquid foods. Despite the aforementioned companies and organizations, Praxair Inc. (Burr Ridge, IL, USA) is the recognized industrial leader dedicated to systematically scaling up and commercializing DPCO2 processing. Using US Patent Nos. 5,393,547 and 6,723,365 from the University of Florida, Praxair Inc. is the first processor to commercialize a 151 L/min (units) continuous DPCO2 processing system for treating liquid foods in 2004 (Praxair Inc., 2003b, 2003c). The commercial scale system is a scale-up from their 1.94 L/min prototype units and it was commercialized under the trademark name `Better Than FreshTM' (BTF) (Higgins, 2002; Praxair Inc., 2003c).


214


8.3

Better Than FreshTM (BTF) system

Praxair Inc. licensed the DPCO2 processing technology from the University of Florida and built a pilot model of the BTF system in 1998 (Higgins, 2002). The continuous BTF system, similar to other DPCO2 systems, consisted of three major regions; the pressurized mixing region, the reaction region, and the depressurization region. The pressurized mixing region is the area where the DPCO2 is combined with a pressurized flow of liquid food. The pressure in the flow regions is maintained primarily to keep CO2 in a continuous fluid phase. The operating pressure of DPCO2 processing is around 34.5 MPa, which, by comparison, is much lower than the operating pressures of HP processing (234± 600 MPa). The reaction region is the location in which the inactivation of the required levels of harmful microorganisms and enzymes takes place. After the reaction region, the mixture flow passes into the depressurization region, in which the pressure is decreased sufficiently to vaporize and separate the CO2 from the liquid food (Fig. 8.3) (Balaban et al., 1995; Ho and Connery, 2004). Two conditions exempted DPCO2 from being considered as a food additive in the DPCO2-treated juice, and thus making regulatory clearance less complicated. The first condition is that CO2 will be removed from the treated juice at the end of the process. Based on the definition of food additive as stated in CFR 21 section 201(s) a food additive is àny substance the intended use of which results or may reasonably be expected to result, directly or indirectly, in its becoming a component or otherwise affecting the characteristics of any food including any substance intended for use in packing, packaging, producing, manufacturing, processing, preparing, treating, transporting, or holding food; and including any source of radiation intended for any such use.' As CO2 is being isolated from the juice at the end of the process, it will not be a part of the treated juice and thus not an additive for the process. The second condition is the

Fig. 8.3

Schematic diagram of DPCO2 processing for freshly squeezed orange juice.



215

GRAS status of CO2, since `GRAS' substances are included in the categories of substances that are exempted from the food additive according to the definition of the Act. Fresh squeezed orange juice has been chosen in this case study as the liquid food to demonstrate the efficacy of the BTF system. Before the commercialization of a novel process for orange juice, that process must be validated at the pilot-scale level and pass at least two critical criteria. The first criterion is microbial safety. The microbial safety aspect of orange juice is regulated by FDA (2001b), as published in Title 21 CFR 120 `Procedures for the Safe and Sanitary Processing and Importing of Juice' on Jan. 19, 2001. The second criterion is quality. The applied process needs to provide a product with high consumer acceptance for an intended shelf life without compromising the physical attributes, nutritional content, and sensory characteristics of the freshly squeezed juice. 8.3.1 BTF system pilot model and microbial validation The BTF system pilot model was validated for effectiveness against real pathogens at a process authority facility, the Illinois Institute of Technology Research Institute (IITRI) in Chicago, IL (Praxair Inc. 2003c). Based on the history of orange juice outbreaks, the pertinent, or primary target, pathogen of concern for freshly squeezed orange juice was identified as Salmonella. However, in order to broaden the application spectrum for the BTF system, two other common liquid food vegetative pathogens, E. coli O157:H7 and L. monocytogenes, were also included in the microbial validation studies. A fivestrain `cocktail' was used for each of the vegetative pathogens in the challenge test. The strains of each pathogen (Table 8.1) (Ho and Connery, 2004) were selected based on their history with juice outbreaks. The use of the five-strain cocktail expanded the probability of including strain(s) that were more resistant to the DPCO2 processing, and thus provided a more conservative estimate of the efficacy of the process. All the vegetative pathogen strains were harvested at their early stationary phase, `stress-adapted' by growing in pH 3.8±4.0 medium broth, and cooled for 18 h at 4 ëC prior to the inoculation. The `stress-adapted' procedures were intended to simulate the acidic and low temperature environment typical of orange juice, thereby minimizing shock and stress on the inoculated bacteria during spiking. Complete clean-in-place (CIP) sanitation of the BTF system was mandated before and after each trial run to minimize background contamination. The BTF system pilot model was not equipped with a built in CIP system, and sanitation was achieved by circulating hot sanitizer throughout the whole system. To validate cleanliness of the system residual rinse water was collected for microbial enumeration after sanitation. At the beginning of the test run, the BTF system pilot model was primed with cooked orange juice (90 ëC, 30 min) until steady state was reached. Cooked orange juice was used for the challenge studies to ensure that the challenge test was performed in the target liquid food


216


Table 8.1 Log10 reduction of vegetative pathogens in orange juice resulted from the treatment of the BTF system pilot model

Salmonella chloraesuis cocktaila Listeria monocytogenes cocktailb Eschericia coli O157:H7 cocktailc a

b c

Untreated spiked orange juice (cfu/mL)

DPCO2-treated spiked orange juice (cfu/mL)

Log10 Reduction

1:4 108 4:5 107 1:5 108

7.7 7.8

S. chloraesuis subsp chloraesuis Weldin serotype Enteritidis (ATCC 4931), S. chloraesuis subsp chloraesuis Weldin serotype Agona (ATCC 51957), S. chloraesuis subsp chloraesuis Weldin serotype Enteritidis (ATCC 31194), S. chloraesuis subsp chloraesuis Weldin serotype Muenchen (ATCC 8388), S. chloraesuis subsp chloraesuis Weldin serotype Enteritidis (ATCC 13076), and S. chloraesuis subsp chloraesuis Weldin serotype Typhi (ATCC 19430) L. monocytogenes ATCC 51414, ATCC 51775, ATCC 43257, ATCC 51778, ATCC 13932, and ATCC 15313 E. coli O157:H7 ATCC 35150, ATCC 43894, ATCC 43890, ATCC 43895, ATCC 700599 and ATCC 700377

environment and the tests were not interfered with by microorganisms indigenous to the orange juice. Non-inoculated orange juice samples were collected at the feed tank and at the exit port to document the absence of background microbial contamination from the juice and from the system. The cooked juice in the feed tank was spiked with high levels (107±108 cfu/ mL) of `stress-adapted' 5-strain cocktails of pathogen inoculum. The high level of microbial load was used to enable a measurable level of residual cells for the comparisons of the efficacy of various process variables and to demonstrate >5 log10 reduction on the tested pathogens. The spiked-cooked orange juice was gently mixed in the feed tank for 5 min in order to ensure homogeneous distribution of the inoculated culture throughout the juice. Samples were collected from the feed tank at the beginning and at the end of all the test runs. Results showed that there were less than 0.05 log differences among the various before and after test-run feed tank samples, indicating that microbial reduction in the spiked-cooked orange juice was caused by the DPCO2 process treatment. Pathogen challenge studies with the BTF system pilot model showed that there were 7.7-, 8.1-, and 8.2-log10 reductions of the five-strain cocktails of L. monocytogenes, Salmonella, and E. coli O157.H7, respectively (Table 8.1). No recovery of injured pathogens in the DPCO2 processing treated juice samples were detected during the 30 days of storage at 4 ëC. These inactivation results of vegetative pathogens coincided with findings of other scientists. Lin et al. (1994) showed an 8-log10 reduction of L. monocytogenes, and Sims and Estigarribia (2002) demonstrated an 8.8-log10 reduction of E. coli as a result of DPCO2 processing. Lower levels of (102±103 cfu/mL) inoculum were employed in repeating the challenge studies. These lower inoculum levels were used to reflect the typical levels of actual microbial contaminants generally found in the juice and to explore possible tailing effects occurring from the inactivation



217

process. No residual cells were detected with the lower inoculum challenge tests (Ho, 2003a; Ho and Connery, 2004). All of the validation work involving pathogens has to be done in a controlled environment (e.g., a biosafety level 2 or 3 containment facility). In order to minimize biological hazardous risks to the handler and cross-contamination to the facility, non-pathogenic `surrogate' bacteria were selected for studies during the scale-up process. The major criteria used in choosing the surrogate strains were its intrinsic resistance to DPCO2 treatment in comparison to the pertinent pathogens and the presence of a marker (e.g., natural antibiotic resistance) to aid in differentiating the surrogate strains from the indigenous microbial flora of the product. Based on the above criteria Listeria innocua ATCC 33090 (L. innocua) and Escherichia K-12 with streptomycin resistance ATCC 25253 (E. coli K-12) were chosen as the surrogate of the Gram-positive and the Gram-negative vegetative pathogens, respectively. Surrogate challenge studies replicated with the BTF system pilot model showed that there were 5.6- and 6.1-log10 reductions of L. innocua and E. coli K-12, respectively, indicating that they were more resistant to the DPCO2 process than their pathogenic counterparts (Ho, 2005). 8.3.2 BTF system prototype model Following the completion of the validation study with pathogens, four BTF system prototype models were built. The scalable machine used off-the-shelf components instead of customized parts and had an integrated CIP system, a critical feature for sanitation and for minimizing background contamination of product. Assembling the CIP system for the prototype model was challenging because a majority of off-the-shelf sanitary components cannot sustain high pressures, and not a lot of high-pressure components have sanitary versions. After thorough research, the prototype was built with sanitary components that could withstand the pressure and work under a one-button CIP system. Sanitizer solutions were circulated under pressure throughout the system at a controlled velocity, temperature, and residence time, and sanitation verification procedures, as described for the pilot model, were used to show that the inner surfaces of the prototype were thoroughly cleaned and safe (Higgins, 2002). The capacity of the BTF-system prototypes was 1.9 L/min, a 3-fold increase from the 0.7 L/min rate demonstrated by the pilot model (Praxair, 2003a, 2003b). The flow rate of the prototypes was still far away from that of a commercial model but the programming logic and instrumentation incorporated into the model was identical to that of the commercial unit. The control panel was a PLC from Allen-Bradley equipped with the Wonderware's InTouch 7.11 control interface that allows remote access for diagnostics and troubleshooting 24 hours a day (Higgins, 2002). This provided a lot of convenience in terms of technical support, especially when the system was at beta sites for testing. Each prototype demonstration model was mounted on a skid to enable easy shipping to juice processing plants for microbial, quality, sanitation, and equipment validation. The idea was to allow potential users to envision a larger version of


218


the BTF system and to establish confidence in the commercial scale BTF system. Surrogate challenge tests using L. innocua and E. coli K-12 were carried out with all the scale-up prototype models prior to putting them to work. The purpose was to demonstrate that these prototype models mirrored the performance of the original pilot-scale model. Results indicated that the BTF system prototype models were able to deliver 5.8- and 6.0-log10 reductions on L. innocua and E. coli K-12, respectively (Ho, 2003a, 2005; Ho and Connery, 2004). This demonstrated that the prototype models had similar performance to that of the BTF system pilot model. 8.3.3 Quality and shelf life validation Besides efficacy in reducing contaminating pathogenic microorganisms, maintaining the quality of the freshly squeezed orange juice with processing and inactivating indigenous microorganisms and enzymes responsible for spoilage and quality deterioration are critical for novel technologies in their application to orange juice processing. Subsequent to achieving surrogate validation, shelf life studies with freshly squeezed orange juice were carried out for 70 days at 4 ëC with the BTF system prototype model. The quality of the DPCO2 treated juice was compared to that of the untreated freshly squeezed orange juice stored under the same conditions. During the shelf life studies, samples were retrieved weekly and evaluated for indigenous microorganisms, nutritional content, and indigenous enzymes. The indigenous microflora populations were estimated using standard plate counts (total aerobic plate counts) and yeast and mold population. After 2 weeks of storage at 4 ëC, the untreated freshly squeezed orange juice showed increases from 4.8 to 5.8 logs and 3.0 to 4.2 logs in standard plate counts and yeast and mold populations, respectively. The DPCO2 treated juice, on the other hand, demonstrated a very stable and low indigenous microbial population throughout the entire 70-day storage period at 4 ëC. Standard plate counts remained at 8.04 >6 3.2±5.5

Lettuce Baby carrot

E. coli O157:H7

1.00 ?

15 ?

80

2.31 3.08

Apple

E. coli O157:H7

18.0

10

90±95 3.8±>7.0

Du et al. (2003)

Blueberry Strawberry Raspberry

Salmonella Salmonella Salmonella

8.0

120

2.44±3.67 3.76±4.41 1.54

Sy et al. (2005a)

Cabbage

Salmonella E. coli O157:H7 L. monocytogenes

4.1

30.8 20.5 29.3

4.42 3.13 3.60

Sy et al. (2005b)

Carrot


4.1

30.8 20.5 29.3

5.15 5.62 5.88

Lettuce


4.1

30.8 20.5 29.3

1.58 1.57 1.53

Reference Han, Sherman, et al. (2000a) Han, Linton, et al. (2000b) Han, Linton, et al. (2001) Du et al. (2002) Singh et al. (2002)

Table 11.1 Continued


Produce

Microorganism

ClO2 (mg/l)

Time (min)

Apple Tomato Onion Peach

Salmonella Salmonella Salmonella Salmonella

4.1 4.1 4.1 4.1

25 25 20 20

4.21 4.33 1.94 3.23

Apple

Allicyclobacillus acidoterrestris

4.32

60

>5

Lee et al. (2006)

Blueberry

L. monocytogenes Salmonella E. coli O157:H7

4

720

99.9

3.94 3.62 4.25

Popa et al. (2007)

Strawberry

E. coli O157:H7 L. monocytogenes Salmonella enterica

5

10

90±95

4.6 4.7 4.3

Mahmoud et al. (2007)

Lettuce

E. coli O157:H7 S. enterica

5

10

90

3.9 2.8

Mahmoud and Linton (2008)

Melon

E. coli O157:H7 L. monocytogenes Salmonella Poona

5 5 5

10 10 6

90±95

4.6 4.3 5

Mahmoud et al. (2008)

Experiments were performed at 20±23 ëC. Table adapted from Gomez-Lopez et al. (2009). Used with permission.

RH (%)

log Reduction

Reference

Novel technologies for the decontamination of fresh fruits and vegetables

289

commonly used food processing technologies for shelf-stable food products. However, heat can be extremely damaging to the quality and stability of fresh and minimally processed fruits and vegetables. For this reason, thermal treatments have been restricted to a narrow range of applications for fruits and vegetables. Unripened or incompletely ripened fruits can be given a thermal treatment to kill and/or sterilize insect pests. Even at this more advanced stage of maturity, thermal treatments can compromise produce quality, when applied improperly. Recently, the use of a precise thermal treatment as an antimicrobial step for cantaloupe has gained attention from industry. A one minute submersion in water heated to 70 ëC of intact melons with surface contamination by Salmonella reduced the pathogen population from 4.6 to 0.8 log cfu/cm2, a reduction of 3.8 log cfu/cm2 (Ukuku et al., 2004). A treatment of two minutes at 76 ëC reduced Salmonella and total microflora by 3 log cfu/cm2. Cut fruit pieces prepared from the treated melons had higher quality which lasted throughout 28 days of storage at 4 ëC (Fan et al., 2006, 2008). Solomon et al. (2006) obtained reductions of 4.6 log cfu/cm2 for Salmonella on melons treated for one minute at 85 ëC. In that work, thermal penetration profiles indicated that the internal temperature of treated melons was such that edible flesh 10 mm below the surface was unaffected by thermal conduction from the heated rind. The optimal treatment temperature for cantaloupe appears to be approximately 74±76 ëC. Salmonella cannot survive in wash water maintained at this temperature. In actual practice, thermal wash treatments would be combined with conventional chemical rinses and mechanical brushing. When combined with a mechanical brushing step, a 20-second treatment at 75 ëC resulted in a 3 log cfu reduction of E. coli (Fallik et al., 2007). Also, the speed of cooling of the thermally treated product can be via relatively slow air-cooling, or, as in the case of the preparation of cut fruit, by the more rapid and expedient method of removing the heated rind.

11.4 Adaptation of existing technologies: plasma, phage treatment and bacteria-based biological controls The preceding sections introduced some key areas of research in which conventional treatments are being modified either to achieve improved levels of effectiveness or to be used in entirely new ways. This section will consider produce processing technologies which were originally developed for entirely different applications. In many ways, the adaptation of an existing set of tools can be as challenging as de novo development. To take technologies which are mature in their respective context of ink adhesion, electronics manufacture, animal husbandry or field pathology, and adapt them for use to improve the safety of fresh produce, requires setting aside many existing principles and practices. The promise in leveraging the existing body of knowledge is the potential to speed the development of an effective food processing tool. The challenge is in


290


bringing together the information and perspectives from disparate fields of inquiry, and doing so with the openmindedness necessary to making the technology work under a wholly new set of operational constraints. The examples discussed in this section of the chapter are generally less mature and not ready for commercial implementation compared to many of those discussed earlier. Nevertheless, they hold significant potential for widespread usage as effective tools. 11.4.1 In-package plasma Cold plasma is a novel sanitizing technology which has shown promise for use on fresh produce. Plasma technologies, and associated terminology, are not particularly common in the context of food processing. Although the technologies used to create plasmas are varied, the underlying mechanisms involve similarities of energy transfer. As energy is added to materials, they change state, going from solid to liquid to gas, with large-scale inter-molecular structure breaking down. As additional energy is added, the intra-atomic structures of the components of the gas break down, yielding plasmas ± concentrated collections of ions, radical species and free electrons (Gadri et al., 2000; Niemira and Gutsol, 2009). Although technically it is a distinct state of matter, for all practical purposes, cold plasma may be regarded as an energetic form of gas. For food processing, it is useful to specify that the term `cold plasma' refers to operation at nonthermal temperature ranges, rather than requiring refrigeration as part of the system. This clarification further serves to distinguish cold plasma in the context of produce sanitization from unrelated applications in other areas such as textiles, plastics and electronics manufacturing and processing. Most cold plasma technologies used for food processing rely on the application of the plasma directly to the food product, or indirectly via a forced air stream. An example of this approach is the gliding arc plasma system (Niemira and Sites, 2008). Treatments of three minutes effectively reduced human pathogens applied to the surfaces of Golden Delicious apples. At the optimal gas flow rate, reductions of Salmonella were 3.4 log cfu/ml, while E. coli O157:H7 was reduced by 3.5 log cfu/ml (see Fig. 11.1). These treatments resulted in minimal color or textural changes to the treated produce. A modified design of cold plasma emitters offers the potential for inpackage treatment processing (Schwabedissen et al., 2007). Electrically conductive labels are affixed to the inside surface of the container. By inducing a voltage through the packaging, cold plasma may be generated on the label's edges, generating ozone and other sanitizing plasma species inside the package. A ten-minute treatment using this approach generated ozone concentrations of approximately 2000 ppm within a container. This was sufficient to effect a 4-log cfu/ml reduction of Bacillus subtilis on agar within the package. The process is undergoing optimization by critical analysis and adjustment of the cold plasma generating discharge labels. Factors such as their method of application (screen-printed, vs. applied or bonded) their shape,



291

Fig. 11.1 Cold plasma inactivation of Salmonella Stanley (top graph) and E. coli O157:H7 (lower graph) on golden delicious apples. Feed gas is air, flow rate 10 liters/ min (circle), 20 liters/min (square), 30 liters/min (triangle), or 40 liters/min (diamond). Different letters for each treatment time indicate significant differences (P < 0:05) among flow rates. Bars standard error. (Adapted from Niemira and Sites, 2008).

material, electrical conductivity, etc., all influence the efficacy of plasma generation. A different type of external plasma generation system used external electrodes to generate ozone within the package (Klockow and Keener, 2009). The voltage applied externally led to the creation of ozone in the 3 mm thick plasma field inside the plastic bag, between the pinched electrodes. The resultant ozone concentrations within the bag were 1.6 and 4.3 mg/L for bags filled with air and oxygen gas, respectively. Spinach leaves inoculated with E. coli


292


Table 11.2 E. coli O157:H7 surviving populations and corresponding ozone concentrations at refrigeration (5 ëC) Treatment time (min)

Gas type

Storage time (h)

Survivor population (log10 CFU/leaf)

Ozone (by volume) (mg/L)

0 5 5 5 5 5 5

Oxygen, air Oxygen Oxygen Oxygen Air Air Air

0.5/2/24 0.5 2 24 0.5 2 24

8.2 0.4a 6.3 0.6b,c 5.6 1.4c 2.4 0.5d 7.1 0.4b 5.9 0.7c 3.6 1.7d

N/A 2.7 0.5a 1.0 0.3b 0.0 0.0d 0.8 0.4bc 0.3 0.2cd 0.0 0.0d

Average population of untreated, unstored, inoculated samples was 7.6 0.6 log10 CFU/leaf. The microbial detection limit for E. coli O157:H7 (6460) was 2.0 log10 CFU/leaf. Average ozone concentrations after 5 min treatment in air and oxygen were 1.6 0.2 mg/L and 4.3 1.0 mg/L, respectively. Values in survivor population column with different letters are significantly different (P < 0:05). Values in ozone column with different letters are significantly different (P < 0:05). Reprinted from Klockow and Keener (2009). Used with permission.

O157:H7 demonstrated significant reductions in microbial populations following these treatments, ranging from 3 to 5 log cfu/leaf (Table 11.2). It should be noted that although the treatment was effective in reducing the pathogen, it also had a negative impact on sensory quality. The degree of discoloration was related to the concentration of ozone, with oxygen packaging having a greater negative impact than air packaging. Cold plasma is a rapidly developing technology, and holds significant potential for operational application to fruits and vegetables. As with all processing technologies, however, the retention of quality of the treated produce is a fundamental requirement for any antimicrobial treatment. For this reason, a clearer understanding of the sensory impact of efficacious levels of plasma treatment will be an essential part of establishing protocols for commercial use. 11.4.2 Phage treatments Bacteriophages are viruses which infect and kill bacteria such as L. monocytogenes, E. coli O157:H7 or Salmonella. Bacteriophages are commonly referred to in the food science community simply as phages. They are regarded as a targeted, self-replicating bio-based antimicrobial tool. The advantage of phage treatments is that they will use the cellular machinery of the pathogenic bacterial host to reproduce, thus amplifying the concentration of viral particles in the presence of the bacterial threat agent. The phage-host interaction is strain specific, with a given isolate of bacteriophage being effective against a single isolate of bacteria, or, at most, against a narrow range of isolates (Sharma et al., 2005). This specificity implies that real-world applications would rely on cocktails of phages to broaden their utility as a food treatment. The USFDA recently approved phage treatments as a means for suppressing and/or



293

eliminating L. monocytogenes from packaged ready-to-eat meat and poultry products (Lang, 2006). A liquid culture of the Listeria-specific phage is applied to the meat or poultry product immediately prior to packaging. The regulations related to foods treated with such phage preparations require the ingredients label on the product to contain the phrase `bacteriophage preparation' (Lang, 2006). Additional descriptive labeling may also be required, depending on the composition of the product. Regulations are not yet in place to allow for phage treatments of fresh produce. However, it is clear that phage-based treatments for fruits and vegetables would most likely be applied as a dip or spray, and possibly in combination with other antimicrobial treatments. A phage treatment reduced L. monocytogenes on apples and melons by 0.4 and 5.3 log cfu/g, respectively. A nisin treatment reduced L. monocytogenes on apples and melons by 0.9±2.0 and 3.0 log cfu/g, respectively. Combining the nisin and phage treatment reduced L. monocytogenes on apples and melons by 1.5±2.3 and 6.4 log cfu/g, respectively (Leverentz et al., 2003). Phage KH1 reduced E. coli O157:H7 attached to stainless steel coupons by 1±2 log cfu, and cells living in mature, protective biofilms were not significantly reduced (Sharma et al., 2005). Abuladze et al. (2008) demonstrated the efficacy of a bacteriophage cocktail against E. coli O157:H7 on tomato, spinach and broccoli. A 5 minute treatment with phage cocktail preparations of 108, 109, and 1010 PFU/ml resulted in ~1.2±3.0 log cfu/g reductions on the vegetable commodities, levels comparable to those obtained with similarly inoculated inert surfaces or with beef products. Reductions were enhanced by increasing the concentration of phage particles applied. The optimization of phage treatments for fresh produce is a matter of ongoing research, in advance of regulatory approval. Possible applications include preharvest (i.e., in-field) application, or post-harvest (i.e., during processing or packaging). The economics of scale come into play with respect to any intervention that is proposed for a pre-harvest application. Even for leafy vegetables such as lettuce or spinach, only a portion of the plant is harvested. Therefore, in applications to plants growing in the field, a measurable proportion of the phage cultures applied would be as a prophylactic measure, rather than curative. It remains to be seen if such a methodology could be made practical, even with phage cocktails capable of targeting a broad pathogen range. 11.4.3 Bacterial-based biological control Research on the use of bacterial biological controls has been going on for a number of years. These preparations may involve a single isolate or a mixed culture, defined or natively derived. This approach has shown some success with applications to control fungal phytopathogens such as Alternaria alternata (Wang et al., 2008). Competitive exclusion has found applications in altering the intestinal microflora of poultry and swine to prevent the establishment of Salmonella (Atterbury 2009). This provides chicks and immature pigs with beneficial gut microflora weeks or months before they would have acquired it


294


otherwise. The heightened disease resistance of a mature GI microflora makes it virtually impossible for Salmonella to multiply. This reduces Salmonella in the environment, in general, while protecting the specific animal being treated. While the use of biocontrol and competitive exclusion is being used in animal systems and against phytopathogens, the development of an effective biocontrol for plant-contaminating enteric human pathogens has been more challenging. The goal is to develop a readily applied isolate or cocktail of isolates that has specific bactericidal or bacteriostatic potential. The microflora on and in fresh produce can range from 102 to 109 cfu/g. Interactions among the bacteria, yeasts and fungi which make up this population can exert positive or negative influences on human pathogen growth and/or survival (Fett, 2006; Liao, 2008). Native microflora derived from alfalfa seeds and from baby carrot effectively inhibited the enteric pathogens Salmonella, E. coli and L. monocytogenes when inoculated on bell pepper disks (Liao, 2007). One of the primary species in this suppressive microflora is a strain of Pseudomonas fluorescens designated Pf 279. This strain was originally isolated as a biocontrol agent of a phytopathogen that attacks the roots of wheat plants. It has since been found that Pf 2-79 effectively suppresses enteric pathogens on sprouting seeds. The environment required for sprout production, with ample nutrition in a humid, almost aqueous environment, is considered to be one of the primary reasons for a series of sprout-related food borne illness outbreaks in the last 15 years. Using this biocontrol agent as a pre-treatment, Salmonella growth was retarded by 2±3 log cfu/g relative to the control (Liao, 2008). The work to scale up effective biocontrol and competitive exclusion treatments has been a difficult process. Lessons have been learned from the successes of this type of intervention in animal cultivation and in phytopathogen suppression. However, the relatively low population densities and sporadic distributions of enteric pathogens on fresh produce make it a challenge to completely eliminate them with bioactive methods.

11.5

Future trends

Despite the difficulty of accurately projecting trends, an observer of a few decades ago would have been able to foresee some of the developments of recent years. Efforts have long been underway to harmonize international regulations, in support of increasing globalization of the food supply chain. The obverse of these efforts has been the occasional use of regulations by various nations to achieve unilateral trade balance goals. The competition from international and overseas suppliers of fresh produce has become more intense; at the same time, global partnerships in production and supply chains have broadened the number of products available in domestic markets. The increased demand for convenient processed foods has led to new categories of food products. However, the increased complexity of production of these multi-component convenience foods has magnified the potential for difficulties in ensuring food



295

safety and in component tracebacks compared to raw or unprocessed commodities. From a food safety perspective, it has become increasingly clear that fresh and fresh-cut produce is comparable to meat, poultry and seafood in terms of the research attention it deserves. An extensive summary of specific areas of key research has recently been presented (Niemira et al., 2009). Regulators, producers and processors have been working with researchers to improve sanitation controls, detection, traceback and epidemiology. The goal of this section is to identify the major factors that will influence the future of produce processing. Rather than an ungrounded attempt to predict which food processing technology will be the `next big thing', the discussion will be oriented on extrapolating existing trends towards likely future directions. 11.5.1 Tolerances Compared to standard practices of the past, fresh produce today must meet exacting metrics for handling and safety. The visual and `hand-on-pallet' load inspections, which may have once sufficed for both suppliers and purchasers, have been replaced by microbiological testing reports, automated temperature data loggers, and computerized process controls. The increasingly widespread use of rigorous HACCP plans signify the specific tolerances for sensory quality and microbiological safety at every step in the chain. Buyers are requiring tighter controls from processors, who in turn have more stringent standards of their grower suppliers. Guidance for the appropriate standards for each phase of an operation ± irrigation water quality, worker hygiene practices, flume water amendment protocols, packing line swab testing, etc. ± comes from various sources. Industry trade groups regularly issue recommendations. Scientific bodies such as the Institute of Food Technologists and the International Association for Food Protection convene panels of experts from industry, government and academia to review the science and offer guidance. Regulators such as the FDA serve by fostering and supporting these discussions, and in implementing guidelines based on the sound science that arises from them. Finally, independent testing service providers play a key role in applying the relevant science when reviewing the facilities and practices of growers and processors. Integration and coordination of these activities will be the means by which the industry will meet ever more stringent tolerances. Weak links in the chain will be identified and addressed, not as a one-and-done approach, but as a continual process of improvement. 11.5.2 Traceback Part of the tighter tolerances of the future will be to have structures in place which will facilitate traceback. In the past, a traceback exercise may have led back to an individual grower or supplier. Currently, some, but not all, supply chains can be traced back to a particular field and specific date of harvest. It is


296


common to see barcoded inventory supply labels on pallets and individual boxes, ready to be cross-referenced to supply and delivery manifests. The data management structures in use today serve the needs of normal commerce. Particularly in a commodity environment such as tomatoes, where repack from several suppliers is not unusual, traceback of the source of individual items is not a simple task. The trend in the fresh and fresh-cut produce industry is to adjust product coding and data management so as to enhance efficiency of the normal commerce, but also to better serve the needs of the traceback process. These improvements may derive from developments in technology, such as the use of RFID tags on pallets and boxes, or by better use of existing inventory management and tracking tools. In part, recent changes in country of origin labeling (COOL) will better serve this process. Supply chain control and validation will be at the center of development. 11.5.3 Technology While new technologies for epidemiological analysis will serve traceback needs, and new inventory tracking tools will serve supply chain management, the primary technological drivers for the fresh produce industry will be in the areas of communications and systems integration. This trend is already visible in the increasing use of common standards for microbiological quality at various points of growing and processing. Information about where and when a particular commodity load was grown and harvested will be most useful and valuable when it is readily available. It may be that the entire history of each pallet or box, from planting date to harvesting date, may accompany it through the supply chain. Much more information will be shared, evaluated and used on a more proactive basis than ever before. It is a widely cited truism that the world is getting smaller and more `talkative' through increased communications. It is as true in the area of fruit and vegetable production and processing as it is in every other sphere of life. The accelerating pace of technology development will ensure that this trend will continue, and will allow for coordination and cooperation among involved industry partners.

11.6

Sources of further information and advice

This present work notwithstanding, many valuable reference materials exist in electronic form as continually updated web sites, RSS outlets, news aggregator feeds, etc. The online sources listed below, current as of March 2010, are some of the means by which producers and consumers can benefit from recent advances in connectivity and communications. The realities of the global market environment means that real-time access to information will be a cornerstone of rapid response to product recalls and compliance issues. The ability to share, discuss, evaluate and act on information will allow the industry to maximize efficiency and control for every stage of production and distribution. While



297

neither exhaustive nor unchangeable, the following list of sources are a valuable entry point into the global network of communication related to the safety and security of fresh produce, and food processing technologies. · http://twitter.com/FoodSafety (USDA ± National Agricultural Library) · http://twitter.com/USDA_nass (USDA ± National Agricultural Statistics Service) · http://twitter.com/USDAFoodSafety (USDA ± Food Safety Inspection Service) · http://www.fsis.usda.gov/News_&_Events/Feeds/index.asp (USDA ± FSIS RSS feeds, podcasts and open blogs) · http://twitter.com/FDArecalls (FDA ± Recalls, Market Withdrawals and Safety Alerts) · http://www.fda.gov/oc/rss/ (FDA ± RSS feeds) · http://twitter.com/CDCemergency (US Centers for Disease Control and Prevention) · http://bites.ksu.edu/ (BITES Food Safety Network listserv summary and archives) · http://twitter.com/FoodProcessing (Industry trade journal)

11.7

Acknowledgements

The author would like to thank Ms L. Cheung for technical assistance in preparation of this manuscript, and Drs X. Fan and Y. Liu for their critical reviews. Mention of trade names and commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture.

11.8

References

and SULAKVELIDZE A (2008), `Bacteriophages reduce experimental contamination of hard surfaces, tomato, spinach, broccoli, and ground beef by Escherichia coli O157:H7', Appl Environ Microbiol, 74(20), 6230±6238. ATTERBURY R J (2009), `Bacteriophage biocontrol in animals and meat products', Microbial Biotechnol, DOI: 10.1111/j.1751-7915.2009.00089.x BEUCHAT L R, PETTIGREW C A, TREMBLAY M E, ROSELLE B J and SCOUTEN A J (2004), `Lethality of chlorine, chlorine dioxide, and a commercial fruit and vegetable sanitizer to vegetative cells and spores of Bacillus cereus and spores of Bacillus thuringiensis', J Food Prot, 67, 1702±1708. DU J, HAN Y and LINTON R H (2002), Ìnactivation by chlorine dioxide gas (ClO2) of Listeria monocytogenes spotted onto different apple surfaces', J Food Prot, 19, 481±490. DU J, HAN Y and LINTON R H (2003), Èfficacy of chlorine dioxide gas in reducing Escherichia coli O157:H7 on apple surfaces', Food Micro, 20, 583±591. ABULADZE T, LI M, MENETREZ M, DEAN T, SENECAL A


298


and VINOKUR Y (2007), `Hot water rinsing and brushing technology for the fresh-cut industry', Acta Hort (ISHS), 746, 229±236. FAN X, ANNOUS B A, SOKORAI K J, BURKE A M and MATTHEIS J P (2006), `Combination of hot water surface pasteurization of whole fruit and low dose irradiation of fresh-cut cantaloupe', J Food Prot, 69, 912±919. FAN X, ANNOUS B A, BEAULIEU J and SITES J (2008), Èffect of hot water surface pasteurization of whole fruit on shelf life and quality of fresh-cut cantaloupes', J Food Sci, 73, M91±98. FETT W F (2006), Ìnhibition of Salmonella enterica by plant-associated pseudomonads in vitro and on sprouting alfalfa seed', J Food Prot, 69, 719±728. FALLIK E, RODOV V, HOREV B, SELA S, ALKALAI-TUVIA S

GADRI R B, ROTH J R, MONTIE T C, KELLY-WINTENBERG K, TSAI P, HELFRITCH D J, FELDMAN P,

and CHEN Z (2000), `Sterilization and plasma processing of room temperature surfaces with a one atmosphere uniform glow discharge plasma (OAUGDP)', Surface Coatings Technol, 131, 528±542. Â MEZ-LO Â PEZ V M, RAJKOVIC A, RAGAERT P, SMIGIC N and DEVLIEGHERE F (2009), `Chlorine GO dioxide for minimally processed produce preservation: a review', Trends Food Sci Tech, 20(1), 17±26. HAN Y, SHERMAN D M, LINTON R H, NIELSEN S S and NELSON P E (2000a), `The effects of washing and chlorine dioxide gas on survival and attachment of Escherichia coli O157:H7 to green pepper surfaces', Food Micro, 17, 521±533. HAN Y, LINTON R H, NIELSEN S S and NELSON P E (2000b), Ìnactivation of Escherichia coli O157:H7 on surface-uninjured and -injured green bell pepper (Capsicum annuum L.) by chlorine dioxide gas as demonstrated by confocal laser scanning microscopy', Food Micro, 17, 643±655. HAN Y, LINTON R H, NIELSEN S S and NELSON P E (2001), `Reduction of Listeria monocytogenes on green peppers (Capsicum annuum L.) by gaseous and aqueous chlorine dioxide and water washing and its growth at 7 C', J Food Prot, 64, 1730± 1738. JIFSAN (JOINT INSTITUTE FOR FOOD SAFETY AND APPLIED NUTRITION) (2007), Tomato Safety Research Needs Workshop. Available from: http://www.jifsan.umd.edu/events/ event_record.php?id=18 (accessed 5 March 2010). KLOCKOW P A and KEENER K M (2009), `Safety and quality assessment of packaged spinach treated with a novel ozone-generation system', LWT ± Food Science and Technology, 42(6), 1047±1053. KOSEKI S, ISOBE S and ITOH K (2004), Èfficacy of acidic electrolyzed water ice for pathogen control on lettuce', J Food Prot, 67, 2544-2549. LANG L (2006), `FDA approves use of bacteriophages to be added to meat and poultry products', Gastroenterology, 131(5), 1370. LEE S Y, DANCER G I, CHANG S S, RHEE M S and KANG D H (2006), Èfficacy of chlorine dioxide gas against Alicyclobacillus acidoterrestris spores on apple surfaces', Int J Food Microbiol, 108, 364±368. SHERMAN D M, KARAKAYA F

LEVERENTZ B, CONWAY W S, CAMP M J, JANISIEWICZ W J, ABULADZE T, YANG M, SAFTNER R

and SULAKVELIDZE A (2003), `Biocontrol of Listeria monocytogenes on fresh-cut produce by treatment with lytic bacteriophages and a bacteriocin', Appl Env Microbiol, 69, 4519±4526. LIAO C H (2007), Ìnhibition of foodborne pathogens by native microflora recovered from fresh peeled baby carrot and propagated in cultures', J Food Sci, 72, M134±M139. LIAO C H (2008), `Growth of Salmonella on sprouting seeds as affected by the inoculum



299

size, native microbial load and Pseudomonas fluorescens 2-79', Lett Appl Microbiol, 46, 232±236. MAHMOUD B S M and LINTON RH (2008), Ìnactivation kinetics of inoculated Escherichia coli O157:H7 and Salmonella enterica on lettuce by chlorine dioxide gas', Food Micro, 25, 244±252. MAHMOUD B S M, BHAGAT A R and LINTON RH (2007), Ìnactivation kinetics of inoculated Escherichia coli O157:H7, Listeria monocytogenes and Salmonella enterica on strawberries by chlorine dioxide gas', Food Micro, 24, 736±744. MAHMOUD B S M, VAIDYA N A, CORVALAN C M and LINTON R H (2008), Ìnactivation kinetics of inoculated Escherichia coli O157:H7, Listeria monocytogenes and Salmonella Poona on whole cantaloupe by chlorine dioxide gas', Food Micro, 25, 857±865. NIEMIRA B A and GUTSOL A (2009), `Non-thermal plasma as a novel food processing technology', in Zhang H Q, Barbosa-CaÂnovas G, Balasubramaniam V M, Dunne P, Farkas D and Yuan J, Non-thermal Processing Technologies for Food, Blackwell Publishing, Ames, IA. NIEMIRA B A and SITES J (2008), `Cold plasma inactivates Salmonella Stanley and Escherichia coli O157:H7 inoculated on golden delicious apples', J Food Prot, 71, 1357±1365. NIEMIRA B A, FAN X, GRAVANI R B, DOONA C J and FEEHERRY FE (2009), `Research needs and future directions', in Fan X, Niemira B A, Doona C J, Feeherry F E and Gravani R B (eds), Microbial Safety of Fresh Produce: Challenges, Perspectives and Strategies, Blackwell Publishing, Ames, IA, pp. 421±426. PARK E J, ALEXANDER E, TAYLOR G A, COSTA R and KANG D H (2008a), `The decontaminative effects of acidic electrolyzed water for Escherichia coli O157:H7, Salmonella typhimurium, and Listeria monocytogenes on green onions and tomatoes with differing organic demands', Food Micro, 26(4), 386±390. PARK E J, ALEXANDER E, TAYLOR G A, COSTA R and KANG D H (2008b), Èffects of organic matter on acidic electrolysed water for reduction of foodborne pathogens on lettuce and spinach', J Applied Micro, 105(6), 1802±1809. POPA I, HANSON E J, TODD E C D, SCHILDER AC and RYSER E T (2007), Èfficacy of chlorine dioxide gas sachets for enhancing the microbiological quality and safety of blueberries', J Food Prot, 70, 2084±2088. SAPERS G M, WALKER P N, SITES J E, ANNOUS B A and EBLEN DR (2003), `Vapor-phase decontamination of apples inoculated with Escherichia coli', J Food Sci, 68, 1003± 1007. SCHWABEDISSEN A, LACINSKI P, CHEN X and ENGEMANN J (2007), `PlasmaLabel ± a new method to disinfect goods inside a closed package using dielectric barrier discharges', Contrib Plasma Phys, 47, 551±558. SETLOW P, DOONA C J, FEEHERRY F E, KUSTIN K, SISSON D and CHANDRA S (2009), Ènhanced safety and extended shelf-life of fresh produce for the military', in Fan X, Niemira B A, Doona C J, Feeherry F E and Gravani R B (eds), Microbial Safety of Fresh Produce: Challenges, Perspectives and Strategies, Blackwell Publishing, Ames, IA, pp. 263±288. SHARMA M, RYU J H and BEUCHAT L R (2005), Ìnactivation of Escherichia coli O157:H7 in biofilm on stainless steel by treatment with an alkaline cleaner and a bacteriophage', J Appl Microbiol, 99, 449±459. SINGH N, SINGH R K, BHUNIA A K and STROSHINE RL (2002), Èfficacy of chlorine dioxide, ozone, and thyme essential oil or a sequential washing in killing Escherichia coli O157:H7 on lettuce and baby carrots', Lebensmittel-Wissenschaft und


300

Case studies in novel food processing technologies -Technologie, 35, 720±729.

and TAUXE R V (2004), `Fresh produce: a growing cause of outbreaks of foodborne illness in the United States, 1973 through 1997', J Food Prot, 67, 2342±2353. SOLOMON E B, HUANG L, SITES J E and ANNOUS B A (2006), `Thermal inactivation of Salmonella on cantaloupes using hot water', J Food Sci, 71(2), M25±M30. SY K V, MCWATTERS KH and BEUCHAT LR (2005a), Èfficacy of gaseous chlorine dioxide as a sanitizer for killing Salmonella, yeasts and molds on blueberries, strawberries, and raspberries', J Food Prot, 68, 1165±1175. SY K V, MURRAY M B, HARRISON M D and BEUCHAT L R (2005b), Èvaluation of gaseous chlorine dioxide as a sanitizer for killing Salmonella, Escherichia coli O157:H7, Listeria monocytogenes, and yeasts and molds on fresh and fresh-cut produce', J Food Prot, 68, 1176±1187. UFPA (UNITED FRESH PRODUCE ASSOCIATION) (2007), Leafy Greens Food Safety Research Conference. Available from: http://www.unitedfresh.org/newsviews/ leafy_greens_food_safety_research (accessed 5 March 2010). UKUKU D O, PILIZOTA V and SAPERS G M (2004), Èffect of hot water and hydrogen peroxide treatment on survival of Salmonella and microbial quality of whole cantaloupe and fresh-cut cantaloupe', J Food Prot, 67, 432±437. SIVAPALASINGAM S, FRIEDMAN C R, COHEN L

VANDEKINDEREN I, DEVLIEGHERE F, VAN CAMP J, KERKAERT B, CUCU T, RAGAERT P, BRUYNE J

DE and MEULENAER B DE (2009), Èffects of food composition on the inactivation of foodborne microorganisms by chlorine dioxide', Int J Food Micro, 131(2±3), 138± 144. WANG H, FENG H and LUO Y (2006), `Dual-phasic inactivation of Escherichia coli O157:H7 with peroxyacetic acid, acidic electrolyzed water and chlorine on cantaloupes and fresh-cut apples', J Food Safety, 26, 335±347. WANG Y, BAO Y, SHENA D, FENG W, YU T, ZHANG J and ZHENG X D (2008), `Biocontrol of Alternaria alternata on cherry tomato fruit by use of marine yeast Rhodosporidium paludigenum Fell & Tallman', Int J Food Micro, 123(3), 234±239. WU V C H and KIM B (2007), Èffect of a simple chlorine dioxide method for controlling five foodborne pathogens, yeasts and molds on blueberries', Food Micro, 24(7±8), 794±800.


Part III Case studies in food preservation using antimicrobials, novel packaging and storage techniques


12 Use of natamycin as a preservative on the surface of baked goods: a case study J. Delves-Broughton, Danisco UK Ltd, UK and L. Steenson, C. Dorko, J. Erdmann, S. Mallory, F. Norbury and B. Thompson, Danisco USA Inc., USA

Abstract: This chapter describes the use of natamycin preparations applied as a post-baking spray to the surface of baked goods as means of preservation, delaying or preventing the growth of surface mold. The physical and chemical properties of natamycin and its use and safety as a food preservative, and the problem of mold spoilage are reviewed. Natamycin trials carried out with bread loaves and information on the selection and design of suitable spraying systems that can be used in a bakery are presented. Key words: mold spoilage, natamycin, preservation, baked goods, surface spray systems.

12.1

Introduction

Demand for food with a long shelf life, but, at the same time, free of synthetic chemical preservatives, is an expanding area of research in novel uses for natural preservatives such as natamycin. One such novel use for natamycin is as a surface treatment of baked goods to prevent or delay yeast and mold spoilage as a replacement for propionate and sorbate. The development of such a new application provides an interesting case history in that it requires a multidisciplinary approach with contributions from food microbiologists, bakers, engineers and process technologists. Information is presented here on natamycin and its uses as a food preservative: trials for its use as a preservative on the


304


surface of baked goods; and the requirements, design and selection of suitable spraying and conveyor systems that could be installed in a modern bakery.

12.2

Natamycin

Natamycin, previously sometimes known as pimaracin or tennectin, is a polyene macrolide antimycotic produced by the actinomycete Streptomyces natalensis and other closely related Streptomyces spp. Natamycin is active against yeasts and molds, and shows no activity against bacteria. 12.2.1 History Natamycin was first isolated in 1955 from a culture filtrate of a Streptomycetes isolated from a soil sample in South Africa (Struyk et al., 1959; Brik, 1981). Natamycin is produced by fermentation of S. natalensis in a medium containing a carbon source (e.g., starch or molasses) and a fermentable nitrogen source (e.g., corn steep liquor, casein, soya bean meal). Fermentation is aerobic and mechanical agitation and antifoaming agents can aid the process. The temperature range is 26±30 ëC and pH range of 6±8. Owing to its low solubility natamycin will accumulate mainly as crystals and these can be extracted following separation of the biomass by solvent extraction (Struyk and Waivisz, 1975). Natamycin preparations have been used for several years as a preservative protecting foods and beverages against yeast and mold spoilage. Many applications are in bacteria fermented foods prone to yeast or mold spoilage as the preservative has a selective action against yeasts and molds with no action against bacteria. Commercial preparations available are NatamaxÕ (Danisco, Denmark), DelvocidÕ (DSM, Holland) and Silver Elephant Natamycin (Zheijiang Silver Elephant Bio-Engineering, China). The natamycin content of most preparations is 50% with the incipient being lactose, glucose, or sodium chloride. Preparations are also available that contain food grade polymers that aid the adherance of natamycin for surface treatments of foods (DelvesBroughton et al., 2006). 12.2.2 Physical and chemical properties Natamycin belongs to a group of antifungals known as polyene macrolides. The structure (Fig. 12.1) was first determined by Ceder (1964) and the stereo structure by Lancelin and Beau (1995). It has a molecular weight of 665.7 Daltons, is amphoteric and has an isoelectric point of 6.5. Natamycin is a white to cream-colored crystalline powder with no taste and little odor. It is stable in powder form if stored at room temperature, but in aqueous solutions it is less stable, particularly if exposed to acidic conditions, light, certain oxidants and heavy metals (Raab, 1972). Natamycin has low solubility in water


Use of natamycin as a preservative on the surface of baked goods 305

Fig. 12.1 The structure of natamycin.

(approximately 40 g/mL). This low solubility is an advantage in the surface treatment of foods because it ensures that the preservative remains on the surface of the food where it is needed, rather than migrating into the foods. Increased solubility occurs with a range of solvents (Delves-Broughton et al., 2005). Raab (1972) reports on the effect of pH on stability of natamycin solutions. It is more stable in the pH range 4.5±9, and in pHs above and below this range becomes significantly less stable. 12.2.3 Antimicrobial spectrum Natamycin is effective against a wide range of yeasts and molds. The preservative is usually effective at concentrations between 1 and 10 g/mL. In general, yeasts are more sensitive than molds, the minimum inhibitory concentrations (MIC) of yeasts usually less than 5 g/mL, whereas that of molds can be 10 g/ mL or higher. 12.2.4 Mode of action The mode of action of natamycin involves an interaction between natamycin and ergosterol, an essential component of membranes of yeasts and molds. Originally it was proposed that this interaction resulted in increased membrane permeability efflux of cellular material. However, recent research by Te Welscher et al. (2008) and van Leeuwen et al. (2009) has shown that the action of natamycin does not increase permeability of the cytoplasmic membrane but more likely prevents cell growth, spore germination, and inhibits membrane associated enzyme activity. Penicillium discolor, Verticillium cinnabarinum, and Botrytis cinerea, are three molds with reduced ergosterol content in their cell membranes, and ergosterol-deficient mutants of Aspergillus nidulans have demonstrated reduced natamycin sensitivity (Ziogas et al., 1983). De Boer and Stolk-Horsthuis (1977) and De Boer et al. (1979) compared the sensitivity of


306


yeasts and molds from cheese and sausage factories where natamycin had been used for several years and where it had never been used. There were no differences in the sensitivities to natamycin of yeasts and molds between these sites. 12.2.5 Method of assay Shirk et al. (1962) developed an agar diffusion bioassay using Saccharomyces cerevisiae as indicator organism. However, High Performance Liquid Chromatography (HPLC) is the preferred method of assay (Anon., 2007). Surface natamycin can be extracted from the surface of foods using methanol. The limit of detection for the HPLC assay is 0.5 g/g. Various other methods have been described such as ultraviolet spectrophotometry (CapitaÂn-Vallvey et al., 2000) and enzyme immunoassay (Maertlbauer et al., 1990). 12.2.6 Uses of natamycin in foods The primary applications of natamycin are for the surface treatment of cheeses and fermented sausages to prevent the growth of yeasts and molds. These two product applications have wide regulatory approval. The three main methods of surface treatment of cheese are by spraying, dipping, or by applying the natamycin in polyvinyl acetate (PVA) suspension coatings. Fermented sausages are prone to mold spoilage during the ripening process. As the sausages ripen the pH drops and this reduces the water-holding capacity of the sausages, resulting in a decrease in moisture content, which provides ideal conditions for the growth of yeasts and molds. The use of natamycin for the surface treatment of cheeses and sausages is allowed in the EU and many other countries at a maximum level of 1 mg natamycin/dm2 with a penetration depth of no more than 5 mm. In the USA, natamycin is not approved in meats but is approved in cheese at a maximum level of 20 g/g, and in other foods such as non-standardized yogurt, cottage cheese, sour cream, non-standardized dressing, and marinades and sauces (Thomas and Delves-Broughton, 2001; Delves-Broughton et al., 2005). Natamycin is approved in the USA at levels in bread up to 14 mg/kg, tortillas and English muffins up to 20 mg/kg, and cakes and US style muffins at 7 mg/kg. In China it can be used on the surface of moon cakes and baked goods when applied by spraying or dipping in a suspension of concentration of 200±300 mg/ kg, providing that the residue in the treated product is less than 10 mg/kg. 12.2.7 Safety and tolerance Natamycin was last extensively reviewed in 2003 by the Joint Expert Committee on Food Additives, JECFA (2003), which confirmed that the previously established Acceptable Daily Intake (ADI) of 0±0.3 mg/kg body weight was satisfactory. The European Union (EU) has not yet set an ADI, and use in the EU is restricted to the surface of cheeses and dried fermented sausages. The


Use of natamycin as a preservative on the surface of baked goods 307 intravenous route is the path by which polyene macrolide antimicrobials are most toxic and oral administration is less toxic (Hamilton-Miller, 1973). There is apparently no adsorption of up to 500 mg/day natamycin from the human intestinal tract after 7 days administration (Brik, 1981). Laboratory feeding studies to determine the ADI were carried out by Levinskas et al. (1966) and are summarized by Delves-Broughton et al. (2005).

12.3

The problem of mold spoilage in baked goods

Bread, which normally is about 0.95 water activity (aw), has a short shelf life. In some cases bread and other baked goods can be sufficiently moist inside to permit the growth of spore-forming bacteria, producing `ropey spoilage'. However, it is considered that molds are the far more serious problem (Anon., 1998). Baked goods primarily comprise raw agricultural commodities; therefore mold spores are continually introduced into the process environment of the bakery (Poisson, 1975; Rogers and Hesseltine, 1978; Mislivec et al., 1979; Seiler, 1986; Eyles et al., 1989). Mold spores associated with these commodities are spread throughout the processing environment during normal operations such as cleaning and mixing (Legan, 1993; Gemeinhard and Bergman, 1977; Spicher 1967, 1980). Baked goods are vulnerable to mold spoilage, the nature and incidence of which is related to the moisture content of the food. High aw products such as bread, cakes, muffins and pastries spoil rapidly, usually due to the growth of species of Penicillium such as P. roqueforti, P. brevicompactum and P. chrysogenum, as well as Aspergillus, Wallemia, Eurotium, Rhizopus, Mucor and Chrysonilia sitophila (Dragoni et al., 1980, 1989; Spicher 1984; Spicher and Isfort 1987; Pitt and Hocking, 1999). During bread baking, the internal temperature of a bread loaf approaches 100 ëC. As the crumb approaches 98 ëC, the optimal baking time has been reached (Stear, 1990) and all fungi and vegetative cells and mold spores are destroyed, but mold spores can recontaminate the product in the post bake area (Ponte and Tsen, 1987). Manufacturers need to find solutions to the problem of making good tasting, moist baked goods such as bread, muffins, cakes, and the like, which require a long mold-free shelf life. Mold inhibitors commonly used in baked products are often synthetic chemical preservatives such as sorbate and propionate. As the pH of most of these bakery products is a minimum of pH 6, these organic acid preservatives may be ineffective. Further problems associated with these preservatives include negative taste impact (Seiler, 1964; Pyler, 1973), and the public preference for natural rather than chemical preservatives and microbial resistance (Pitt and Hocking, 1999). Additionally, Monascus ruber is a mold resistant to propionic acid that produces red spots on certain breads (Spicher and Isfort 1988) and some species of Penicillium are able to degrade sorbate (Pitt and Hocking, 1999; Daley et al., 1986). Recently, natural antimycotics have started to appear more readily in bakery products. These natural compounds are produced by a variety of different


308


microorganisms during controlled fermentation processes (Anon., 1984; Nemensky et al., 1978). Many of these compounds are commonly labeled based on the fermentation substrate the bacterium is propagated. This would include products like cultured dextrose, cultured wheat starch, and cultured whey, to name a few. Many of these natural compounds are also pH dependent. This is a significant hurdle to overcome, because many baked goods exhibit a pH in the range between 6 and 8, which makes many of them largely ineffective. This problem can be overcome by acidifying the baked goods to make the antimycotics more effective, or by significantly increasing the dosage. Both of these solutions can lead to flavor differences, off-odors, and may result in performance issues for yeast leavened products. The latter problem of yeast inhibition during leavening can be overcome by encapsulating the preservative, increasing the level of yeast, or by applying the preservative on the surface of the baked good at a high level, but which equates to a low level based on the total weight. One such natural preservative that meets the last criterion is natamycin.

12.4 Trials on the use of natamycin as a surface treatment of baked goods Various studies on the application of natamycin as a mold preservative to heat processed foods and to cheeses and meats have shown that surface applications are far more effective than incorporating the preservative throughout the food matrix. One reason is that molds are aerobic and tend to grow on the surface and rarely inside the food. Another reason is that the low solubility of natamycin means that it can be concentrated on the surface of the food and does not migrate inwards where it is not required. Furthermore adding natamycin to the dough would inhibit the yeast fermentation. The bakery industry has recognized that the determination of the mold-free shelf life of baked goods is best and most conveniently carried out by careful daily examination of products for the appearance of individual mold colonies during incubation at the chosen test temperature (Seiler, 1964). This method gives a far more realistic and relevant result than determination of yeast and mold numbers using microbiological agar plate-counting techniques. Baked goods were sprayed using a pilot-plant spray equipment at the Danisco laboratories at New Century in Kansas, US. Figure 12.2 shows a diagrammatic picture of the pilot spray unit. A feature not shown is the reservoir for the natamycin suspension that is constantly recirculated to keep the natamycin in aqueous suspension. A trial was conducted on muffins to explore the feasibility of spraying a natamycin suspension on the surface to increase the shelf life. Muffins are a popular snack product in the British Isles, Australia, New Zealand, and North and South America. Its popularity is expanding to other countries. Muffins can be described as a highly aerated, soft textured baked cereal product, generally



Fig. 12.2 Diagramatic picture of pilot spray system for application of natamycin to baked goods.

round and 20±30 mm thick. They are usually reheated, and eaten hot often with butter, jam, or cheese, and traditionally as an afternoon snack (Campbell-Platt, 1987). The muffins were produced using the following formula: 2000 g flour; 300 g starch; 150 g glucose; 2000 g sugar; 2500 g whole egg; 30 g baking powder; 50 g whey powder; 40 g sodium bicarbonate; 26 g sodium chloride; 1500 g cooking palm oil; 500 g water; 150 g of GRINDSTEDÕ FSB 270 emulsifier and stabilizer system (propylene glycol esters of fatty acids, beet fiber, mono- and diglycerides of fatty acids, sodium stearoyl-2-lactate); and GRINDSTEDÕ LBG 246 (locust bean gum). This resulted in a muffin with a surface water activity between 0.855 and 0.878 and pH value of approximately 8. A control set of muffins were produced with no antimicrobial, and sprayed with either water alone or with natamycin suspension to yield a final level of 4±5 g/cm2. The natamycin and water sprays were applied to the surface of the muffin shortly after exiting the oven. All muffins were incubated at 25 ëC. The untreated muffins had visual mold at 7 days, and the muffins sprayed with water at 11 days. This difference can not be explained. In contrast, the natamycin-treated muffins did not develop any signs of visual mold growth over 68 days' storage. Another trial was conducted to investigate the use of natamycin suspension applied to the surface of bread loaves. The bread was produced using the following formula: 2000 g flour; 40 g salt; 40 g sugar; 80 g shortening; 40 g


310


instant dry yeast; 1320 g water; 0.3 g GrindamylTM PowerBake 920 (bakery enzyme); 0.2 g GrindamylTM Maxlife 65 (amylolytic enzyme); 6 g Dimodan PH 320-M; 6 g Grindsted SSL P 55 Veg (sodium stearoyl lactylate); 0.2 g Ascorbic acid; 0.2 g GrindamylTM A1000 (alpha-amylase). The breads were sprayed with 2.5 g/cm2 of natamycin on each loaf shortly after exiting the oven. The control loaves were sprayed with water at the same point in the process as the natamycin. The breads were packed in clear bread bags, stored at 30 ëC, and observed over time for the presence of visible mold growth. All control products had visual mold within 12 days. The treated samples were mold-free at the end of the 60 days of the study. The above study was repeated except the bread loaves were allowed to cool before the natamycin was applied to the surfaces. The control was treated with water and target natamycin levels were 2±4 g/cm2. The loaves were stored at ambient temperatures and observed over time for visual mold. The control loaves were moldy within 7 days and the treated loaves remained mold free for 10±17 days. Additional trials were conducted by applying a natamycin solution rather than a suspension to the surface of bread loaves. The advantage of using a natamycin solution is that the need for constant agitation of the suspension is eliminated. In addition to the 0.02% natamycin and 0.02% lactose carrier, the solutions contained either 10% ethanol ethanol and 89.6% glycerine or 50% propylene glycol or 49.6% glycerine. The control breads were sprayed with water and the treated samples were sprayed with natamycin solution to achieve 2 g/cm2 across the surface. The spray was applied just after the bread exited the oven, products were packed in clear bread bags, incubated at 22±24 ëC and 66% relatively humidity, and observed for visual mold throughout the study. The control product started molding on day 6, but the treated samples were still mold free at the end of the 16-day study. The study was repeated using the same parameters but in a different facility. The control samples displayed visible mold in 4 days, and the treated samples started molding on day 11. In addition to the above trials, a commercial pan bread trial was conducted to investigate the use of natamycin in combination with other antimicrobial products. The natamycin suspension in this trial was applied to the surface of the bread shortly after the depanning process by a state-of-the-art spraying system. Results are shown in Fig. 12.3. There were 32 loaves per treatment. The loaves containing vinegar alone started molding by day 8 and all 32 loaves had visible mold by day 21. When natamycin was added at 14 ppm in combination with the vinegar, the mold-free shelf life increased to 13 days and only 3 loaves were moldy by the end of the study. The addition of cultured wheat flour (CWF) at 1% and 2% to the vinegar increased the mold-free shelf-life by 5 and 11 days, respectively. Both of these variables had less than 10 loaves showing mold after 30 days of ambient storage. The addition of all three variables resulted in only two loaves with visible mold (1% CWF) and no moldy loaves (2% CFW) at 30 days. The described trials were investigative in nature and were all conducted with rudimentary handheld spray systems. For optimal results, it is important that



Commercial bakery natamycin spray trial.

312


natamycin be applied over the entire product to achieve an even distribution of the spray. It is important to note that the natamycin suspensions required constant agitation to prevent natamycin from settling. The bread trials were all conducted using the same formula throughout all studies. All studies were uninoculated and relied on contamination by mold spores naturally found in the production facilities. The natamycin content in the suspensions or solutions, or on the bakery products, were analyzed via HPLC using a published methods (Anon., 2007). Sensory testing on all bread products showed no significant difference in attributes between the treated and untreated loaves. In all cases natamycin showed a significant increase in shelf-life over the control product. One of the most important aspects of applying natamycin is reaching a consistent, full, and even coverage across the surface of the product of interest. Because of the importance of getting good coverage, various types of spraying systems can be evaluated. This evaluation can be done by conducting trials with each system and analyzing resultant natamycin residual levels and shelf life data. In the case of pan bread, several locations across the entire loaf can be examined to determine if target levels are being achieved. It is essential that adequate numbers of sample locations be selected to assist in identifying potential distribution issues across a loaf. Different surface locations of the bread can be analyzed. Natamycin can be extracted from the sample and the natamycin levels can be monitored by HPLC (Anon., 2007). In addition to the sample locations, it is important that adequate numbers of loaves are analyzed to obtain sufficient data for statistical analysis. There are a several approaches to evaluating the susceptibility of bakery products to mold growth. One approach is to perform a shelf life study. This type of study relies on mold contamination during normal processing to inoculate the product. The product is held at a given temperature, monitored on some frequency, and mold (counts or visual appearance) recorded throughout the study. Challenge studies are also utilized for evaluating bakery products for mold growth. In a challenge study, the product is first inoculated with a specified level of mold spores, typically the type most commonly contaminating the product. Mold isolates can be obtained through environmental sampling or from spoiled products. Mold spores are prepared and transferred to a suitable substrate, such as flour, dextrose, or maltodextrin. Once transferred to the substrate, the mold spores can be standardized to a given level and inoculated to the product surface. Mold growth is then monitored over the shelf life, either by appearance of visible mold, or plate-counting.

12.5

Considerations and selection of the spraying system

12.5.1 Demands/challenges facing surface applied mold inhibitors When considering systems for surface application of mold inhibitors, it is appropriate to compare any new methodology to current methodology.



Presently, mold inhibitors are typically added to each batch of dough as part of the formula and the inhibitor is evenly dispersed throughout the product. The surface application of a mold inhibitor is typically executed as the products exit the oven by spray application to individual products. As bakery products typically exit at high speed, up to 200 units/min, it is apparent that the challenge is to achieve uniform surface coverage. 12.5.2 Criteria to consider when choosing a non-recirculating system for surface applied mold inhibitor · Natamycin is a proven effective mold inhibitor. In commercial production, natamycin will only be effective if the application is uniform and consistent. Natamycin application must be evenly distributed over the surface area, including crevices of every loaf, to achieve the desired performance. This criterion is a high priority when comparing application systems from different equipment suppliers. Comparative testing involves collecting post-application samples from a significant number of locations on the loaf surface. The samples are analyzed to determine for surface levels of natamycin. Ideally, the natamycin is applied to surfaces within narrow tolerances. The application system should include automatic self-monitoring of application levels. Monitoring can be as simple as a visual, or audible signal or it can be more complicated, such as automatic switching to redundant components or systems. · Non-recirculating systems are preferred for system design. Past experience has indicated that, with prolonged operation of recirculating systems, bacterial contamination becomes highly probable. With a non-recirculating system, the possibility of bacterial contamination in the solution is reduced significantly. · With non-recirculating systems, any overspray that does not hit the target is lost. Minimizing overspray becomes a major design focus. Minimizing overspray losses requires product recognition that allows the application to take place only when product is in the target zone. Other considerations are proper positioning and numbers of application nozzles and optimum positioning of the loaf on the application conveyor. · A successful application system should be highly predictable when parameters change and adjustments are made. For example, if the speed of the application conveyor is reduced with no change in spray application settings, the resultant natamycin application should show a proportional increase. If the results are not predictable when various settings are changed systematically it is extremely difficult for the user to manage the application system. · Owing to low water solubility, the application of natamycin is in aqueous suspensions. The application system must be designed to prevent settling of the solids in the suspension. The most effective method to ensure that natamycin does not settle is to keep the suspension agitated. To monitor effective agitation, samples are collected from the suspension over time and analyzed for natamycin content.


314


· Typically, natamycin is diluted in water at the rate of 1000:1 for application. In most bakeries, floor space is considered a premium. To minimize system space requirements, the application system includes a reservoir for concentrated natamycin. The concentrate is automatically proportioned with water to supply the application reservoir on demand. The application reservoir and the concentration reservoir can both be as small as 10 gallons each. This design eliminates the need for a large application reservoir. As long as the natamycin concentrate is sufficient to create a suspension that resists separation, the concentration level can be chosen so that replenishing it is only necessary once each day. · Mechanical reliability, longevity and maintenance cost should be a major consideration. One effective method to focus on with respect to this criterion is to ask the equipment supplier for a long-term warranty or guarantee. · When mechanical failures do occur, simplicity and time to repair are of prime importance. A successful application system cannot require equipment supplier personnel for emergency repairs ± training personnel in advance will help mitigate these risks. · Although the system may contain technically complicated components, the system should be designed so that the average plant operator can easily be trained to operate and make necessary adjustments. · The system should meet food plant GMP design and choice of materials. Some latitude can be exercised on this criterion, based on individual customer preferences. · Bakeries are seldom designed for water washdown. The application system solution reservoirs and liquid circulation circuit should be designed for Clean-in-Place (CIP) cleaning with regard to the ease of cleaning solution disposal. · It is of primary importance to choose an equipment supplier that has global representation. Taking into account various regulatory standings, natamycin has the potential for global use and should not be restricted by regional equipment representation. · Last but not least, system cost must be a consideration. A great deal is being asked of this system for surface application of mold inhibition. An equipment supplier must be chosen that can meet the requirements outlined above yet also provide pricing that is acceptable to the customer. Depending on the user's objectives for natamycin, considerable savings or revenue improvement can be realized. It is beneficial to help the user identify and quantify these savings, to justify the investment in the natamycin application system. 12.5.3 Optimum system choice for a surface applied natamycin mold inhibitor Surface application technologies considered or tested · Liquid constant pressure atomization: this is the most common form of liquid application to a target. A simple example would be a garden hose spraying flowers or a car wash wand.


Use of natamycin as a preservative on the surface of baked goods 315 · Combination constant liquid and air atomization: this application combines liquid under hydraulic pressure in combination with pressurized air. A nozzle designed to blend the two is used for delivery. Misting for cooling purposes or foam generation for fighting fires would be two examples of this technology. · Liquid pressure atomization in combination with `poppet' on/off frequency: this technology is very similar to a `pop off' or `safety' valve on a boiler or water heater in the home. Liquid under hydraulic pressure is delivered to a nozzle with check type blockage. The check may be spring loaded or incorporate some other type of adjustable loading. The liquid supply pressure overcomes the check resistance momentarily forcing the valve open and releasing a `shot' of liquid. Pressure drops as the liquid is released, allowing the check to close. This process is repeated at a high frequency, appearing to be a constant flow. · Liquid constant pressure atomization in combination with electronic over air on/off frequency: liquid under hydraulic pressure is delivered to a valve nozzle combination. The valve cycles on/off at a very high frequency. The mechanical action of the valve is controlled by electrical frequency in combination with air pressure. The frequency can be adjusted in a wide range. Different nozzle configurations can be used with the same frequency drivers. · Ultrasonic atomization in combination with air shaping and delivery: liquid under hydraulic pressure is delivered to a nozzle equipped with an ultrasonic passage. As the liquid passes through the ultrasonic passage it is vaporized to a very fine mist. The mist is shaped and directed toward the intended target by low pressure air. · Electrostatic charging of target in combination with hydraulic liquid pressure. A conventional nozzle as described in the first bullet of this section is used to atomize and direct the liquid toward the target. The nozzle is equipped with a device to electrically charge the liquid particles. The intended target is also electrically charged with opposite polarity than the liquid particles. Opposite charges attract each other with the intended result of effective coverage. Danisco's solution application choice After careful consideration and extensive testing, Danisco chose Spraying Systems, Co. of Weaton, Illinois, USA as the supplier of the system to deliver the surface applied natamycin. Spraying Systems Co. is a multinational company with representation and support in most major countries throughout the world. Their application system was considered an optimum choice for the following reasons: · · · · · ·

Repeatability of sprayed liquid volume from loaf to loaf. Even distribution of the natamycin across the surfaces of the product. Accurate flow rate compensation to accommodate conveyor speed changes. Spray validation for each spray cycle of each nozzle. Cost of solution implementation. Efficacy of the process proven by laboratory results.


316


Selection criteria and characteristics of the chosen spray system When considering an optimal solution for applying a mold inhibitor like natamycin, achieving proper application rates under variable processing conditions can prove challenging. Implementing oversimplified spraying systems can result in expensive lessons learned. Understanding the importance of liquid parameters and the complexity of variable processing conditions at the outset can reduce exposure to these types of mistakes. When approaching any spray application, characteristics of the liquid to be sprayed are critical. Questions that should always be considered include: · · · · · ·

What is the viscosity of the liquid? Is the liquid abrasive? Is the liquid homogeneous or does it contain particulates? What are these particle sizes? What flow rate/application rate is required? Are there special safety concerns when handling or spraying the liquid?

Knowing the answers to these questions for natamycin helped determine the best technology for the application. Because of the large particle size present in the natamycin suspension and the low application rate requirement, Spraying Systems Co. recommended `Pulse Width Modulated' (PWM) flow control to achieve optimal results. A multiple nozzle arrangement using flat fan nozzles provided 360ë coverage of the product when passing through the `spray zone.' Because a large nozzle orifice size was required to accommodate the particulates in the natamycin, PWM technology was necessary to provide fast cycle speeds. Extremely fast cycling of the nozzles achieved the low flow rates that were required without affecting spray pattern integrity. All of the system performance requirements were met using Spraying Systems Co.'s electrically controlled PulsaJetÕ automatic spray nozzles and AutoJetÕ spray control. The natamycin suspension is mixed in a concentrated form. Therefore, it must be diluted further before application onto the product to maintain appropriate concentration levels. To ensure that the suspension is mixed accurately on a continual basis, Spraying Systems Co. designed an automatic dilution and recirculation system for the solution. The diluted product is constantly agitated and is also circulated throughout the entire system to ensure that particulates do not fall out of suspension. The automated dilution and mixing of the suspension also alleviates human intervention from the process to reduce possible errors without adding additional labor costs. Droplet size, pattern, and positioning of the nozzles ensure efficient transfer of liquid from the nozzle to the product. When line speed changes occur in the process, the system maintains accurate dosing by changing both the cycling rate of the nozzles and liquid pressure adjustment if required. Sensors are also used to measure and detect the length of each product. The algorithm written in the control logic activates the sprays only when the product is within the `spray zone,' which limits overapplication of liquid when not required. To ensure natamycin is applied from each nozzle within the `spray zone,' spray check


Use of natamycin as a preservative on the surface of baked goods 317 sensors are mounted on each nozzle to provide spray validation each time the nozzle is triggered. When the sensor does not confirm a spray cycle, the control logic will alarm and notify plant personnel. These alarms can be configured either to provide visual and audible alarms or to shut down the production process altogether. In summary, accuracy in applying natamycin ensures proper mold inhibition. Precision spray control by Spraying Systems Co. provides a robust, accurate, and efficient system to achieve this. 12.5.4 Importance of conveyor design and target positioning Equally if not more important than the selection of spraying equipment is the positioning of the bread loaf at it passes through the target zone. The best chosen spray application equipment will not deliver satisfactory performance if the bread loaves are not uniformly and consistently positioned as they pass through the target application zone. A successful application system consists of three basic components: (1) the suspension application equipment, (2) the conveyor for transporting the loaves through the target zone that allows spraying on all surfaces including the bottom and (3) a device for ensuring that the bread is positioned uniformly and consistently as it passes through the target application zone. Component 3 should be located at the entrance of the treatment conveyor. It should be designed in such a manner, that it can receive loaves with any spacing and configuration in any manner and re-position them on the treatment conveyor to establish uniform configuration and equal spacing in the treatment zone. Without an effective loaf positioning device, the benefits of natamycin as a preservative and well chosen solution application equipment will not reach their full potential.

12.6

Future trends

The low solubility of natamycin is both an advantage and disadvantage in the surface application of foods. The advantage is that the low solubility prevents migration of the natamycin into the food so that it remains on the surface at a relatively high concentration. As mold spoilage of foods requires oxygen and molds grow predominantly on the surface of foods, this is an advantage. The disadvantage is that a suspension of natamycin requires constant agitation/ stirring to prevent settling, and suspended natamycin can cause problems such as the blockage of spray nozzles. There is interest in combining natamycin with a carrier molecule that would increase solubility and thus allow more even spraying on the surface of baked goods. Other potential benefits are increased stability, protection against UV degradation, and also possibly increased antifungal activity. Cyclodextrins are one potential encapsulation partner for natamycin. These cyclic oligosaccharides contain numerous glucose monomers, the most common


318


of which contain 6±12 monomers. The specific coupling and conformation of the glucose units provide the cyclodextrin molecule with a rigid, conical structure, having a hollow, hydrophobic interior of a specific volume (Masters et al., 2009). The unique shape and physicochemical properties of the cavity enable the cyclodextrin molecules to absorb (form inclusion complexes with) organic molecules, or parts of organic molecules, capable of fitting into the cavity. Natamycin is an organic molecule known to undergo this form of complexation with cyclodextrins (Koontz and Marcy, 2003; Koontz et al., 2003; Cevher et al., 2008). To accomplish this, the hydrophobic end of the peptide binds with the hydrophobic internal cavity of the carrier molecule (cyclodextrin) thereby forming a partial encapsulation, and leaving both ends of the complex's exterior polar. As water is a polar solvent, the inclusion complex becomes inherently soluble. As it pertains to the baking industry, a natamycin-based product containing cyclodextrins could potentially shield natamycin from ultraviolet breakdown, eliminate the need for agitation, promote a more homogeneous natamycin application, and eliminate the plugging of spray nozzles (thereby reducing line stoppages). These potential improvements could improve the convenience and ease of application of natamycin to baked goods.

12.7

References

(1984). Effective and natural cultured whey ingredient inhibits mold. Baker's Digest 58, 24. ANONYMOUS (1998). Cereals and cereal products. In: Microorganisms in Foods, 6, Microbial Ecology of Food Commodities. International Commission on Microbiological Specifications for Foods of the International Union of Biological Societies (ICMSF), Blackie, London, pp. 313±355. ANONYMOUS (2007). Cheese, cheese rind and processed cheese ± determination of natamycin content ± Method by molecular absorption spectrophotometry and by high-performance liquid chromatography. International Standard ISO 9233-2. BRIK, H. (1981). Natamycin. In: Flory, K. (ed.), Analytical Profiles of Drug Substances, Academic Press, New York, p. 513. CAMPBELL-PLATT, G. (1987). Fermented Foods of the World. A Dictionary and Guide, Butterworths, London. Â N-VALLVEY, L.F., CHECA-MORENO, R. and NAVAS, N. (2000). Rapid ultraviolet CAPITA spectrophotometric and liquid chromatographic methods for the determination of natamycin in lactoserum matrix. Journal of AOAC International, 83, 802±808. CEDER, O. (1964). Pimaracin. VI. Complete structure of the antibiotic. Acta Chemica Scadinavica, 18, 126±134. CEVHER, E., SENSOY, D., ZLOH, M. and MULAZIMOGLU, L. (2008). Preparation and characterization of natamycin:gamma-cyclodextrin inclusion complex and its evaluation in vaginal mucoadhesive formulations. Journal of Pharmaceutical Sciences, 10, 4319±4335. DALEY, N.M., LLOYD, G.T., RAMSHAW, E.H. and STARK, W. (1986). Off-flavours related to the use of sorbic acid as a food preservative. CSIRO Food Research Quarterly, 46, 59±63. ANONYMOUS


Use of natamycin as a preservative on the surface of baked goods 319 and STOLK-HORSTHUIS, M. (1977). Sensitivity to natamycin (pimaricin) of fungi isolated in cheese warehouses. Journal of Food Protection, 40, 533±536. DE BOER, E., LABOTS, H., STOLK-HORSTHUIS, M. and VISSER, J.N. (1979) Sensitivity to natamycin of fungi in factories producing dry sausage. Fleischwirtsch, 59, 1868. DELVES-BROUGHTON, J., THOMAS, L.V., DOAN, C.H. and DAVIDSON, P.M. (2005). Natamycin. In: Davidson, P.M., Sofos, J.N., Branen, A.L. (eds), Antimicrobials in Foods, CRC Press, Taylor and Francis Group, Boca Raton, FL, pp. 275±290. DELVES-BROUGHTON, J., THOMAS, L.V. and WILLIAMS, G. (2006). Natamycin as an antimycotic preservative on cheese and fermented sausages. Food Australia, 58, 19±21. DRAGONI, I., ASSENTE, G., COMI, G., MARINO, C. and RAVENNA, R. (1980). Sull'ammuffimento del pane industriale confezionate: Monilia (Neurospora) sitophila e alter specie responsabili. Tecnologia Alimentaria, 3, 17±26. DRAGONI, I., BALZARETTI, C. and RAVARETTO, R. (1989) [Mycoflora seasonal variability in a confectionery production line]. Industrie Alimentari (Pinerolo, Italy), 28, 481±486, 491. EYLES, M.J., MOSS, R. and HOCKING, A.D. (1989). The microbiological status of Australian flour and the effects of milling procedures on the microflora of wheat and flour. Food Australia, 41, 704±708. GEMEINHARD, H. and BERGMANN, I. (1977). Zum Vorkommen von Schimmelpilzen in Backereistauben. Zentralblatt fuÈr Bakterologie Parasitenkunde, Infektionskrankheiten und Hygiene, Abt II, 132, 44±45. HAMILTON-MILLER, J.M.T. (1973). Chemistry and biology of the polyene macrolide antibiotics. Bacteriological Reviews, 37, 166±196. JOINT EXPERT COMMITTEE ON FOOD ADDITIVES, JECFA (2003), Safety evaluation of certain food additives and contaminants, www.inchem.org/documents/jecfa/jecmono/ v48je06.htm KOONTZ, J.L. and MARCY, J.E. (2003). Formation of natamycin; cyclodextrin inclusion complexes and their characterization. Journal of Agricultural and Food Chemistry, 51, 7100±7110. KOONTZ, J.L., MARCY, J.E., BARBEAU, W.E. and DUNCAN, S.E. (2003). Stability of natamycin and its inclusion complexes in aqueous solution. Journal of Agricultural and Food Chemistry, 51, 7111±7114. LANCELIN, J-M. and BEAU, J.M. (1995). Stereostructure of glycosylated polyene macrolides: the example of pimaracin. Bull. Soc. Chim. Fr., 132, 215±223. LEGAN, J.D. (1993). Mould spoilage of bread: the problem and some solutions. International Biodeterioration & Biodegradation, 32, 33±53. LEVINSKAS, G.J., RIBELIN, W.E. and SCHAFFER, C.B. (1966). Acute and chronic toxicity of pimaracin. Toxicology and Applied Pharmacology, 8, 97±131. MAERTLBAUER, E., ALI, H., DIETRICH, R. and TERPLAN, G. (1990). Enzyme immunoassay for the detection of natamycin in cheese rind. Archive fuÈr Lebensmittelhygien, 41, 112±114. MASTERS, J.G., PAYNE, R., SZELES, L.H., XIAOYAN, L. and WILLIAMS, M. (2009). Antiplaque oral composition containing enzymes and cyclodextrins. US Patent No. 7,601,338. MISLIVEC, P.B., BRUCE, V.R. and ANDREWS, W.H. (1979). Mycological survey of selected health foods. Applied and Environmental Microbiology, 37, 567±571. NEMENSKY, J.V., STONE, P. and LEE, S. (1978). Dried cultured wheat flour is a natural mold inhibitor. Baking Industry, 145, 16±17. PITT, J.I. and HOCKING, A.D. (1999). Fungi and Food Spoilage, 2nd edn, Aspen Publishers, Gaithersburg, MD. DE BOER, E.


320


(1975). La microflore des farines. La Meuniere FrancËaise, 314, 10±19. and TSEN C.C. (1987). Bakery products. In Beuchat, L.R. (ed.), Food and Beverage Mycology, Van Nostrand Reinhold, New York, pp. 233±267. PYLER, E.J. (1973). Baking Science and Technology, Siebel, Chigago, IL, pp. 210±221. RAAB, W. (1972). Natamycin (Pimaracin). Its Properties and Possibilities in Medicine. Georg Thieme Publishers, Stuttgart. ROGERS, R.F. and HESSELTINE, C.W. (1978). Microflora of wheat and wheat flour from six areas of the United States. Cereal Chemistry, 55, 889±898. SEILER, D.A.L. (1964). Factors affecting the use of mould inhibitors in bread and cake. In: Molin, N. (ed.), Microbial Inhibitors in Food, Almquivst and Wiksell, Stockholm, pp. 211±220. SEILER, D.A.L. (1986). The microbial content of wheat and flour. International Biodeterioration Supplement, 22, 35±40. SHIRK, R.J., WHITEHALL, A.R. and CLARK, W.L. (1962). The bioassay of natamycin and its binding effect in orange juice. Journal of Food Science, 27, 605±608. SPICHER, G. (1967). Causes and control of mold contamination of bakeries. The Bakers Digest, 41, 30±36, 81. SPICHER, G. (1980). Zur AufklaÈrung der Quellen und Wege der Schimmelkontamination des Brotes im Grossbackbetreib. Zentralblatt fuÈr Bakteriologie Parasitenkunde Infektionskrankheiten und Hygiene. 1 Abt. Original Reiheb Hygiene Betriebshygiene Praeventive Medizin, 170, 508±528. SPICHER, G. (1984). Die Erreger der Schimmelbildung bei Backwaren. I. Die auf verpackten Schnittbroten auftretenden Schimmelpilze. Getreide Mehl und Brot, 38, 77±80. SPICHER, G. and ISFORT, G. (1987). Die Erreger der Schimmelbildung bei Backwaren. IX. Die auf vorgebackenen BroÈtchen, Toast- und Weichbrotchen auftretenden Schimmelplize. Deutsche Lebensmittel Rundschau, 83, 246±249. SPICHER, G. and ISFORT, G. (1988). Die Erreger der Schimmelbildung bei Backwaren. X. Monascus ruber, ein nicht alltaÈglicher Schimmelerreger des Brotes. Getreide Mehl und Brot, 42, 176±181. STEAR, C.A. (1990). Handbook of Breadmaking Technology, Elsevier Science Publishers, Barking. STRUYK, A.P. and WAISVISZ, J.M. (1975). Pimaricin and process of producing same. United States Patent No. 3,892,850. STRUYK, A.P., HOETTE, I., DROST, G., WAIVISZ, J.M., VAN EKK, T. and HOOGERHEIDE, J.C. (1959). Pimaracin, a new fungal antibiotic. Antibiotics Annual 1957±1958, 878±885. POISSON, J.

PONTE, J.G.

TE WELSCHER, Y.M., TEN NAPEL, H.H., BALAQUE, M. M., SOUZA, C.M., RIEZMAN, H., DE KRUIFF, B.

and BREUKINK, E. (2008). Natamycin blocks fungal growth by binding specifically to ergosterol without permeabilizing the membrane. Journal of Biological Chemistry, 283, 6393±6401. THOMAS, L.V. and DELVES-BROUGHTON, J. (2001). Applications of the natural food preservative natamycin. Research Advances in Food Science, 2, 1±10. VAN LEEUWEN, M.R., GOLOVINA, E.A. and DIJKSTERHUIS, J. (2009). The polyene antimycotics nystatin and filipin disrupt the plasma membrane, whereas natamycin inhibits endocytosis in germinating conidia of Penicillium discolor. Journal of Applied Microbiology, 106, 1908±1918. ZIOGAS, B.N., SISLER, H.D. and LUSBY, W.R. (1983). Sterol content and other characteristics of pimaricin-resistant mutants of Aspergillus nidulans. Pesticide Biochemistry and Physiology, 20, 320±329.


13 Commercial applications of oxygen depleted atmospheres for the preservation of food commodities S. Navarro, Food Technology International Consultancy Ltd, Israel

Abstract: Oxygen depleted modified atmospheres (MAs) generated by a variety of different methods were used successfully to replace fumigants for insect control and for the quality preservation of a number of stored products. Flexible hermetic structures suitable for long-term, large-scale storage or for intermediate storage of grain in bags or in bulk have been applied using biogenerated MAs or for limited sizes using vacuum. Applications of these flexible chambers for cereals, nuts, dry fruits, cocoa and coffee, narcissus bulbs or museum artefacts are currently practised. Key words: quality preservation, storage insect control, methyl bromide alternatives, flexible storage structures, vacuum.

13.1

Introduction

There is an increasing demand for quality food uncontaminated by molds, insects, and insecticide residues. In developed countries, the loss of quality is particularly important. In developing countries, poor handling and storage methods under warm and humid climatic conditions promote rapid deterioration of the stored foodstuffs. Postharvest losses of food grain in developing countries have been conservatively estimated during the 1980s at 10±15% by the Food and Agriculture Organization's (FAO) Special Action Program for the Prevention of Food Losses. For example, losses of corn due only to insects in farmers' stores in Nigeria, Swaziland, and Kenya were of the order of 6±10%.


322


Increased public concern over the adverse effects of pesticide residues in food and the environment has led to the partial substitution of the use of contact pesticides (typically organophosphates and pyrethroids) and fumigants by alternative control methods. It is worth noting that of the 14 fumigants listed some 25 years ago by Bond (1984), only one remains today in regular use worldwide, namely, phosphine and methyl bromide, which is used only in developing countries. Methyl bromide kills insects relatively quickly, but because of its contribution to stratospheric ozone depletion (UNEP, 2002), it was phased out in developed countries by 2005, and it is scheduled to be phased out in developing countries (UNEP, 2006) by 2015. In contrast, phosphine remains popular, particularly in developing countries, because it is easier to apply than methyl bromide. However, many insects have developed resistance to phosphine over the last decade. Food commodities can be stored for extended periods, provided that there is no insect infestation and that their water activity is low enough to prevent microbial growth. However, quantitative or qualitative losses still occur. Qualitative losses, for example, may consist of changes in physical appearance, in nutritional degradation due to oxidation and increase in free fatty acids, the presence of insects or their fragments, or contamination by molds or the presence of mycotoxins. Some of these are difficult to detect visually. If the moisture content is maintained sufficiently low, insects remain the main concern for the quality preservation of durable agricultural commodities. Therefore, in this chapter, the major emphasis is placed on technologies based on oxygen depleted atmospheres for the quality preservation of agricultural commodities and the control of insect pests. Oxygen depleted atmospheres are a kind of modified atmosphere (MA) that offers a safe and environmentally benign alternative to the use of conventional residue producing chemical fumigants for controlling insect pests attacking stored grain, oilseeds, processed commodities, and packaged foods.

13.2

Definitions and uses of oxygen depleted atmospheres

The objective of oxygen depleted atmospheres is associated with the application of MA treatment that aims to attain a composition of atmospheric gases rich in CO2 and low in O2, or a combination of these two gases at normal or altered atmospheric pressure within the treatment enclosure for the exposure time necessary to control the storage pests and preserve the quality of the commodity. Terms used in reference to MA storage for the control of storage insect pests or for the preservation of food have appeared in the literature as controlled atmosphere (CA, to be defined below), as sealed storage, or atmospheres used at high or low pressures to define the same method of treatment but using different means. Therefore, an attempt is made here to propose definitions that will add clarity to the available methods for controlling insects during storage, whether at normal atmospheric pressure or under altered atmospheric pressure.


Oxygen depleted atmospheres for the preservation of food commodities

323

MA is proposed to serve as the general term, including all cases in which the composition of atmospheric gases or their partial pressures in the treatment enclosure has been modified to create in it conditions favorable for the control of insects during storage and preserve the quality of the commodity. In an MA treatment, the atmospheric composition within the treated enclosure may change during the treatment period. This term will comprise all the following designations. 13.2.1 MAs under normal atmospheric pressure Controlled atmosphere (CA) CA is a modified gas composition, usually produced artificially, and maintained unchanged by additionally generating the desired gases (CO2 or N2) or by further purging the storage atmosphere with these gases, supplied from pressurized cylinders or otherwise (Fig. 13.1). This supplementary introduction

Fig. 13.1 Application of carbon dioxide-based MA on a silo bin and the schematic presentation of the application process.


324


of gases is carried out when their concentration in the sealed container falls to below the desired level. The CA method is intended to rectify changes caused by possible leaks of gases (that cause the increase of O2 or decrease of CO2 content in the enclosure), which are almost impossible to avoid. Thus, the term CA, although commonly employed as the one describing the entire subject, actually has its own limited and specific meaning. Hermetic storage Hermetic storage is a type of MA that can be applied for the protection of grain. It is also called `sealed storage' or àir-tight storage' or `sacrificial sealed storage'. This method takes advantage of sufficiently sealed structures that enable insects and other aerobic organisms in the commodity or the commodity itself to generate the MA by reducing the O2 and increasing the CO2 concentrations through respiratory metabolism. Assisted hermetic storage Assisted hermetic storage is another type of hermetic storage that uses exothermic gas generators, catalytic oxygen converters, or respiration gases of plant material. In this type of hermetic storage, the atmosphere has been modified by the supply of an atmosphere generated externally from the storage container, so that a gas composition of low-O2 ( (accessed 12 February 2010). MARLER BLOG, 2005±2010. Available at http://www.about-ecoli.com/ecoli_outbreaks/ view/jack-in-the-box-e-coli-outbreak (accessed on 12 February 2010). MARTIN T and ALBRECHT JA, 2003. Meat irradiation education program. Journal of the American Dietetic Assoc. 103: A30. MCKINNEY M, 2009. Family of E. coli victim sues Cargill for $100 million. StarTribune.com, December 4, 2009. Available at http://www.startribune.com/ lifestyle/health/78529222.html?elr=KArks:DCiUo3PD:3D_V_qD3L: c7cQKUiD3aPc:_Yyc:aUU (accessed on 12 February 2010). MOLINS, RA, 2001. Introduction: Historical notes on food irradiation. In RA Molins (ed), Food Irradiation Principles and Applications, Wiley-Interscience, New York. MOSS M, 2009a. Safety of beef processing method is questioned. In: New York Times, December 30, 2009. Available at ?tf="PS2B42">http://www.nytimes.com/2009/ 12/31/us/31meat.html?_r=1&scp=1&sq=e.%20coli+vaccine&st=cse (accessed 10 January 2010). MOSS M, 2009b. Woman's shattered life shows inspection flaws, New York Times, October 3, 2009. Available at http://www.nytimes.com/2009/10/04/health/04meat.html (accessed on 15 January 2010). NATIONAL CATTLEMEN'S BEEF ASSOCIATION AND THE CATTLEMEN'S BEEF BOARD, 2002. Irradiation: Consumer Perceptions. Available at http://www.beefresearch.org/ CMDocs/BeefResearch/Irradiation%20Consumer%20Perceptions.pdf (accessed on 12 February 2010). NATIONAL CENTER FOR HEALTH STATISTICS, 2004. HEALTH, UNITED STATES, 2004. Available at http://www.cdc.gov/nchs/data/hus/hus04trend.pdf#027 (accessed on 12 February 2010). NATIONAL HEALTH MUSEUM, 2009. Access Excellence Classic Collection. Radioactivity: Historical Figures. Available at http://www.accessexcellence.org/AE/AEC/CC/ historical_background.php (accessed on 10 February 2010). NEWMAN W, 2009. After delays E. Coli vaccine approved. New York Times, December 4, 2009. Available at http://topics.nytimes.com/top/reference/timestopics/subjects/f/ food_safety/index.html?scp=2&sq=vaccine+beef&st=cse (accessed 13 January 2010). POHLMAN A, WOOD O, and MASON A, 1994. Influence of audiovisuals and food samples on consumer acceptance of food irradiation. Food Technol. 48: 46±49. TAUXE RV, 2001. Food safety and irradiation: protecting the public from foodborne infections. Emerging and Infectious Diseases 7(3): 515±521. TAUXE RV, 2009. Advancing Food Safety through Irradiation. Presented at Council of State and Territorial Epidemiologists Annual Convention, Buffalo, NY, 10 June, 2009. USDA FOOD AND NUTRITION SERVICE, 2003. Ìrradiation of raw meat and poultry: Questions and answers.' Available at http://ublib.buffalo.edu/libraries/e-resources/ebooks/ images/eev9878.pdf (accessed on 12 February 2010). USDA/FSIS, 2006. Celebrating 100 Years of the Federal Meat Inspection Act. Available from http://www.fsis.usda.gov/About_FSIS/100_Years_Timeline/index.asp (accessed 10 February 2010). USDA/FSIS, 2009. Analysis of Raw Ground Beef and Raw Ground Beef Component JOHNSON A, ESTES RA, JINRU C,


Consumer acceptance and marketing of irradiated meat

463

Samples of E. coli O157:H7. Available at http://www.fsis.usda.gov/Science/ Ecoli_Raw_Beef_Testing_Data_YTD/index.asp (accessed on 15 December 2009). US HOUSE OF REPRESENTATIVES, COMMITTEE ON ENERGY AND COMMERCE, SUBCOMMITTEE

2008. `Regulatory failure: Must America live with unsafe food?' Testimony of Daniel Wegman on March 12, 2008. Available at http://energycommerce.house.gov/index.php?option=com_content&view= article&id=630&catid=31&Itemid=58 (accessed on 12 February 2010). WALTAR AE, 2004. Radiation and Modern Life, Fulfilling Marie Curie's Dream. Prometheus Books, Amherst, NY. WEISS B, 2009. Personal communication. Corporate Communications, Omaha Steaks, 16 Mar 2009. ON OVERSIGHT AND INVESTIGATIONS,


20 Comparing the effectiveness of thermal and non-thermal food preservation processes: the concept of equivalent efficacy M. G. Corradini, Universidad Argentina de la Empresa, Argentina and M. Peleg, University of Massachusetts-Amherst, USA

Abstract: New non-thermal preservation technologies have created the need to establish equivalence between the lethality of conventional thermal processes and that of their alternatives. Since microbial inactivation need not follow first-order kinetics, and if it does the D-value's temperature dependence need not be log-linear, the traditional `F0 value' can rarely be directly converted into a survival ratio. For nonlinear inactivation, a model like the Weibullian-Log logistic (WeLL) can translate dynamic survival ratios recorded in thermal, non-thermal or combined processes into an equivalent-time curve at a chosen reference temperature. This can be done in real time or in the analysis of industrial, experimental or simulated data. Key words: survival curves, inactivation kinetics, sterilization, pasteurization, Weibull distribution, microbial mortality, chemical changes, quality loss.

20.1

Introduction

The term `food preservation' covers any of the numerous physical, chemical or biological methods to eliminate, stop or slow down natural deteriorative processes in foods and extend their shelf-life as edible and safe products. In the technical literature, however, the term frequently refers exclusively to methods


Effectiveness of thermal and non-thermal food preservation processes

465

that eliminate spoilage microorganisms and pathogens and/or stop enzymatic activity. Some of the methods designed to protect foods from microbial spoilage have a long history. Notable examples are drying, smoking, salting, fermentation (and their combinations), and, in northern climes, cold storage. Other methods, like deep freezing, irradiation (ionizing radiation, electron beam and UV), ultra high hydrostatic pressure treatment (HHP), pulsed electric fields, and the use of chemical antimicrobials (such as benzoates or sorbates) or antibiotics (like nisin) are obviously modern. Thermal preservation is perhaps a class of its own. It includes processes like pasteurization at relatively low temperatures, and sterilization at high temperatures, ultra-high temperatures (UHT), and, more recently, combinations of high temperature and ultra high pressure. The main purpose of all these methods, and several others not mentioned here, is to destroy spoilage organisms and pathogens. However, food deterioration need not be caused by the presence of microorganisms only. Rancidity caused by lipid oxidation, non-enzymatic browning, emulsion separation, and powders caking are examples of spoilage in which microorganisms are not involved. They are usually avoided or their severity reduced by the addition of antioxidants, emulsifiers or anti-caking agents, and/or by processes like deaeration, homogenization or agglomeration. Although in the broader sense these methods are means of food preservation, too, the focus of this chapter will be on processes specifically aimed at destroying microorganisms and inhibiting enzymatic activity. To a lesser extent, the chapter will also address nutritional losses and chemical or textural changes that thermal processes may cause. In many instances, a food's safety and shelf-life are determined not only by the preservation method, but also by the package integrity, its intrinsic properties, and how they change in time. These aspects will also not be discussed. Certain preservation processes produce such a dramatic change in the food's appearance and properties that the result is only remotely associated with the food's raw form, if at all. For example, very few people will identify wine as preserved grape juice, raisins as preserved grapes, pickles as preserved vegetables, cheese as preserved milk, or a sausage as preserved meat. Consequently, such products should be considered foods based on their own unique properties. What originally made them last longer, the alcohol content, sugar concentration, pH, water activity, and/or the presence of salts can therefore be treated like any other component of their composition. Much of the discussion in the chapter will focus on the kinetics of microbial inactivation by preservation processes, and other changes that might occur during food preservation, and their mathematical characterization. The emphasis will be on how kinetic models can be used to assess the equivalence of preservation processes, whether they of the same kind (e.g., a comparison of high temperature short time, HTST, versus low temperature long time, LTLT) or of different kinds (e.g., heat versus HHP or chemical treatments). This chapter will also address issues such as how one process affects two different target organisms (e.g., the destruction of Clostridium botulinum spores versus the destruction of Bacillus sporothermodurans spores), or how a process affects nutrient retention


466


and other quality attributes. The underlying assumption is that the inactivation of vegetative microbial cells, bacterial spores, and enzymes and the thermal degradation of vitamins are governed by very different mechanisms at the cellular or molecular level, and that their temporal manifestation at the `population level,' as determined by counts, biochemical activity, or concentration can be quantified by the same or similar kinetic models. Thus according to this notion, differences in kinetics, are manifested in the kinetic model's mathematical structure and the magnitude of its coefficients. The latter, in turn, can be translated into the relative magnitude of the processes' efficacy and the time scale on which they operate. Most of the current theories of microbial mortality were originally developed for thermal processing, canning in particular. They were subsequently applied to the heat inactivation of enzymes and microbial destruction by non-thermal preservation methods (FDA, 2000). Chemical and biochemical degradation processes have been derived primarily from kinetic models borrowed from physical chemistry. Most, if not all, of the models were originally derived to represent simple reactions in a controlled environment. Some of the fundamental mechanistic assumptions underlying these classic kinetic models may need revision, especially when applied to food systems involving active enzymatic systems or whole organisms. The processes in such systems are frequently complex, multistage and interactive rather than simple, and the molecular environment continuously changes (Peleg et al., 2004). Therefore, at least part of what follows will be a departure from the traditional manner of interpreting experimental kinetic data, whether intended for microbial mortality, enzymatic inactivation or the chemical synthesis or degradation of nutrients and other molecular food components.

20.2 Traditional microbial mortality kinetics and sterility measures The traditional methods to assess the efficacy of thermal processes taught in most food science, microbiology and engineering programs are based on the assumption that the inactivation of vegetative cells and bacterial spores follows linear, or first-order, kinetics (e.g., Jay, 1996). Expressed mathematically, it is assumed that under isothermal conditions there is a universal log-linear relationship between the survival ratios, St, and time, t, i.e., loge St ÿkTt

20:1

where St N t=N0 , N t and N0 are the momentary and initial number of survivors, respectively, k the temperature dependent (exponential) destruction rate constant and t is the exposure time. Equation 20.1 is frequently presented in the form: t 20:2 log10 St ÿ DT



467

where DT, the `D value', is the time needed to increase or decrease the survival ratio by a factor of ten. This assumption has been extended to the thermal inactivation of enzymes (e.g., Toledo, 1999) and to other methods of preservation, notably HHP where DT has been replaced by Dp P (e.g., Smelt et al., 2002). The temperature dependence of the `D value' has been assumed to be loglinear too, i.e., DT T ÿ Tref 20:3 log10 z DTref where DTref is the `D value' at a reference temperature, Tref, and z, known as the `z value', is the temperature span needed to raise or lower the `D value' by a factor of ten. According to this model, the lethality of an isothermal heat process can be measured in terms of the number of decades reduction in the target organisms' survival ratio. Thus, canned low acid foods are considered to have achieved `commercial sterility' when the temperature-time combination at the can's coldest point is sufficient to reduce the spores of C. botulinum, had they been present, by twelve orders of magnitude (12D). According to the traditional theory of heat inactivation, the efficacy of a `dynamic' heat process, i.e., having non-isothermal `temperature profile' is expressed by its `F0 value' calculated by: Z t 10TtÿTref =z dt 20:4 F0 t 0

where Tt is the `temperature profile', i.e., the time-temperature relationship of the product at its coldest point. The reference temperature for low acid canning traditionally has been 121.1 ëC (250 ëF) and the `z-value' used for the calculation is that of C. botulinum spores. According to Eq. 20.4, all processes having the same final `F0 value' also have the same lethality, regardless of the temperature profile, Tt, and the process's actual duration. In other words, processes having the same `F0 value' produce the same number of decades reduction in the target cells or spores and hence the same final survival ratio. This stems from the assumed equality (see Peleg, 2006): log10 St ÿ

F0 t DTref

20:5

Where true, the choice of the reference temperature would be immaterial because: 10TÿTref 2 =z 10TÿTref 2 =z 10TÿTref 1 =z T ÿT =z DTref 2 DTref 1 DTref 2 10 ref 1 ref 2

20:6

Therefore, as long as the target cells or spores' inactivation follow Eqs 20.2 and 20.3, the `F0 value' and final survival ratio can be used interchangeably as a


468


thermal process's efficacy measure. By definition, the survival ratio (and its logarithm) is dimensionless, while the `F0 value' has time units. Thus conversion of one to the other requires the inclusion of DTref , which has time units ± see Eq. 20.5. Experimental evidence has revealed that in many cases, even when the isothermal survival curves could be considered log-linear (i.e., follow Eq. 20.1 or 20.2), the temperature dependence of the `D value' can not. Consequently, the log-linear model, Eq. 20.3, has been frequently replaced by the Arrhenius equation: kT Ea 1 1 ÿ 20:7 ÿ ln R T Tref kTref where kT and kTref are the exponential inactivation rate constants at T and Tref, expressed in degrees Kelvin, Ea is the ènergy of activation' and R the Universal gas constant. Because Eqs 20.7 and 20.3 are mutually exclusive mathematical models (Lewis and Heppel, 2000), the equality expressed in Eq. 20.5 does not hold when the Arrhenius equation is used. In that case, the reference temperature choice does affect the calculated process efficacy and the simple and direct relation between the survival ratio and `F0 value' is lost. This, too, has been well known, see Datta (1993) and Nunes et al. (1993), for example, who showed that the discrepancy between the actual survival ratio and its value as deduced from the `F0 value' varies continually with the difference between the process temperature and the one chosen as a reference. This fact by itself is a sufficient reason to discard the `F0 value' as a thermal process's efficacy measure. But the Arrhenius equation, when applied to microbial inactivation and complex biochemical processes, has its own serious problems (Peleg et al., 2004; Peleg, 2006). One is that the ènergy of inactivation' is expressed in kJ/mole. What is a `mole' in the context of microbial cells or spores inactivation is unclear, and the relevance of the Universal gas constant, R, to heat induced biophysical processes that operate at the cellular level has never been properly explained. Other drawbacks of the Arrhenius model would not be eliminated, even if these two issues could be set aside. (For example, by rewriting Eq. 20.7 in the form lnkT=kTref A1=T ÿ 1=Tref , A, which replaces the term Ea/R and has temperature units, becomes a measure of the temperature effect on the exponential rate free from any link to gases and other simple chemical reactions.) The first problem with the Arrhenius model application to microbial inactivation is that the rationale for the temperature scale's compression by converting units of T from ëC to ëK in the reciprocal is unclear, and the same can be said about replacing the rate constant k by ln k. The second problem, also shared by the log-linear model (Eq. 20.3), is that neither Eq. 20.3 nor Eq. 20.7 makes a qualitative distinction between lethal and non-lethal temperatures. At temperatures lower than about 80 ëC, many bacterial spores, especially those of Bacilli species, are hardly destroyed, if at all, a fact exploited when one wants to produce spores suspensions or lyophilized powders free of vegetative cells.



469

Also, both Eq. 20.3 and Eq. 20.7 are based on the assumption that heat inactivation indeed follows strictly first-order kinetics, which entails that the exponential inactivation rate is only a function of temperature, but is constant in time. If true, then at 115 ëC, for example, the exponential inactivation rate must be exactly the same, if the spores have just reached this temperature (after heating their suspensions from 25 ëC in 2 minutes) or cooled to this temperature after being held at 125 ëC for 2 hours! The number of surviving spores will be very different, of course, but not the exponential rate constant! The same will be true if the rate constant is defined in terms of another fixed-order kinetics. All the above also pertains to non-thermal methods of microbial inactivation, enzymatic inhibition and chemical degradation of nutrients or other modes of quality loss. Whenever conventional kinetic models are adapted for new applications, by replacing the temperature with high hydrostatic pressure, chemical disinfectant concentration, or radiation intensity, for example, one or more of the above problems will almost certainly arise. Therefore, meaningful comparison between the efficacies of different preservation processes must be based on the survival ratio itself or on a lethality measure that can be directly converted into a survival ratio. The same is true for complex biochemical and chemical degradation processes, except that the survival ratio would be defined in terms of concentration, for example, instead of cells or spores counts.

20.3 Non-linear kinetics of microbial inactivation and deterioration processes involving nutrient or quality losses 20.3.1 The Weibullian model Survival curves, by definition, are the cumulative form of the temporal distribution of mortality events (or expiration, failure or breakage, etc., in the non-living world). Thus, most microbial survival curves can be described by the Rosin-Rammler distribution function (Schubert et al., 1984) better known as Weibull's (Peleg and Cole, 1998; van Boekel, 2002). Nevertheless, a survival curve's slope has 1/time, i.e., rate units, and hence the connection to kinetics (Peleg, 2006). For kinetic calculations involving microbial survival, writing the basic Weibullian model in the form of a power law equation seems to be the most convenient (Peleg and Cole, 1998; Peleg and Penchina, 2000; Peleg, 2006). Thus, for isothermal inactivation it can be written as: log10 St ÿbTt nT

20:8

where bT and nT are temperature dependent coefficients. Notice that according to Eq. 20.8, upper concavity, or `tailing', is expressed by nT < 1 and downward concavity, indicating `damage accumulation', by nT > 1. The traditionally assumed `first-order kinetics' is simply the case of Eq. 20.8 with nT 1 for all temperatures (see Fig. 20.1). In thermal, chemical and ultrahigh pressure inactivation, the model can be simplified by assigning a fixed



The various shapes of semi-logarithmic survival curves. Notice that log-linear survival (`first order kinetics') is just a special case of the Weibullian model.


471

representative value to the exponent, i.e. making nT n. (For further details, examples and explanation, see Peleg and Penchina, 2000; van Boekel, 2002; Mafart et al., 2002; Corradini and Peleg, 2004, and Peleg, 2006.) For dynamic (i.e., non-isothermal) Weibullian survival, the model needs to be written as a rate (differential) equation, i.e.: nTtÿ1 d log 10 St log10 St nTt 20:9 ÿbTtnTt ÿ dt bTt where bTt and nTt are the momentary values of these coefficients, i.e., at the momentary temperature Tt. Although cumbersome in appearance, Eq. 20.9 is an ordinary differential equation (ODE), which can be solved numerically with almost any modern mathematical software, such as MathematicaÕ, MapleÕ or MatlabÕ, for almost any conceivable `temperature profile', Tt. This model equation can also be converted into a `difference equation' and solved with general-purpose software like MS ExcelÕ (Peleg et al., 2005; Peleg, 2006) ± see below. The Weibullian model (Eqs 20.8 and 20.9) has been used successfully for thermal and non-thermal microbial inactivation kinetics, and the description and prediction of vitamin losses during thermal preservation and storage. Unlike the traditional survival models (Eqs 20.1±20.7), the Weibullian model correctly accounts for the fact that microorganisms' momentary exponential inactivation rate, or that of a chemical compound's degradation, need not be a function of the momentary temperature only but also of the population or system's thermal history. The same is true for situations where the lethal agent is not heat, in which case the momentary exponential inactivation rate will depend on the pressure or lethal chemical agent's concentration, for example (Peleg, 2006). 20.3.2 The role of temperature Published experimental data suggests that the temperature dependence of the Weibullian `rate parameter', bT, regardless of whether nT is fixed or variable, is best described by the log-logistic model (e.g., Corradini and Peleg, 2004; Peleg, 2006): bT loge f1 exp kT ÿ Tc g

20:10

where k and Tc are constants, characteristic of the organism, its growth history and the medium in which it is treated. According to this model, when T Tc , bT 0 and when T Tc , bT kT ÿ Tc , i.e., rises almost linearly with temperature (see Fig. 20.2). Notice that Eq. 20.10, makes a clear distinction between the lethal temperature regime, T Tc , and the non-lethal, T Tc and that Tc serves as the marker of the lethality's onset. This model equation's construction also eliminates the unnecessary `logarithmic compression' of bT, which rarely, if ever, changes by a factor larger than ten in the pertinent lethal temperature range. At the sub-lethal temperature range and below, bT ! 0, and its value can drop by several orders of magnitude, like the rise of the `D


472


Fig. 20.2 Schematic view of the log-logistic temperature dependence of the Weibullian rate parameter, bT, of microorganisms or spores having different heat sensitivities.

value'. But at these low temperatures, its absolute magnitude is so small that it has no practical consequences as far as lethality is concerned. Combining Eqs 20.9 and 20.10 produces the Weibullian Log-logistic (WeLL) model: d log 10 St ÿloge f1 exp fkTt ÿ Tc gg nTt dt nTtÿ1 nTt log10 St ÿ loge f1 exp fkTt ÿ Tc gg

10:11

Or, where nTt n (i.e., constant): d log 10 St ÿloge f1 exp fkTt ÿ Tc gg n dt nÿ1 n log10 St ÿ loge f1 exp fkTt ÿ Tc gg

10:12

As already mentioned, Eqs 20.11 and 20.12 can be easily solved numerically by mathematical and even general-purpose software ± see below. (For linear temperature profiles, Eq. 20.12 can be even solved analytically (Peleg et al., 2003; Peleg, 2006).) According to the modeling approach that has produced Eqs 20.11 and 20.12, the description of microbial mortality needs at least three survival parameters,



473

and not two as in the traditional models (e.g., D, z or kTref , and Ea). In the WeLL model with a constant exponent (Eq. 20.12), the three parameters are n, k and Tc. Moreover, Eq. 20.12 can be used not only to simulate and correctly predict dynamic inactivation patterns (e.g., Corradini and Peleg, 2004; Periago et al., 2004; Pardey et al., 2005; Peleg, 2006), but also to serve as a model for extracting the survival parameters, n, k and Tc, from experimental survival curves determined under non-isothermal conditions (Peleg and Normand, 2004; Peleg, 2006; Peleg et al., 2008a; Corradini et al., 2008, 2009a). A modified version of the WeLL model, could also be used to correctly predict dynamic survival patterns in chemical disinfection and there is evidence that it might be applicable to HHP as well (Corradini and Peleg, 2003b; Doona et al. 2007). Although the WeLL model seems to depict and predict correctly most of the observed patterns of microbial survival, it need not be unique, let alone universal. It can be shown that alternative mathematical models might be required, as in the case of clearly sigmoidal isothermal survival (Peleg, 2003a, 2006), the existence of an activation shoulder (Peleg, 2002; Corradini and Peleg, 2003a) or prolonged tailing in the survival curve (Peleg, 2006). It has also been demonstrated that alternative models can provide predictions of the same quality as those rendered by the WeLL model, especially if they are based on four instead of three survival parameters. However, for any survival model to be truly predictive, it ought to be based on the notion that the momentary logarithmic inactivation rate is the isothermal inactivation rate at the momentary temperature at the time that corresponds to the momentary survival ratio. In other words, any general survival model must be consistent with the fact that the logarithmic inactivation rate need not be a function of temperature only, but also a function of the population's thermal history. This concept, as already mentioned, can be extended to enzymatic inactivation and nutrient losses as well as to complex biochemical degradation reactions and processes. The general concept can be extended to processes and reactions involving `growth' or àccumulation' as in the case of non-enzymatic browning, the development of off-flavors, etc. In such cases, the right side of Eq. 20.8 could probably be used after dropping the minus sign. Or, it may be replaced altogether by an expression that accounts for an asymptotic concentration or sigmoid (logistic) growth, when appropriate. Notable exceptions are biological and chemical processes whose product shows a peak concentration and oscillatory reactions. Such reactions need a slightly different mathematical treatment. But again, the underlying principle ought to be that the momentary reaction or process's rate is the isothermal rate at the momentary temperature at the time that corresponds to the system's state. Conventional fixed-order kinetics models might be applicable to a large variety of relatively simple reactions. But they might be inadequate for complex processes having alternative and interacting pathways, especially when they occur in an active and changing chemical environment, such as occurs in a food undergoing a heat treatment.


474

20.4


Equivalence criteria

When it comes to microbial safety of foods by a thermal process, the requirement is that the coldest point in the product receives adequate treatment to eliminate the target microorganism. The equivalence of two processes can therefore be established by comparing the thermal histories of their cold points. The situation is different when it comes to nutrient losses or off-flavor development. In such cases, it is unclear what constitutes equivalent losses or damage. Is it the total over the whole volume? Is it the mean? If so, is it the linear, geometric or logarithmic mean? If not the mean, is it the loss or damage at the worst point? Addressing these issues is outside the scope of this chapter. Suffice it to say that if there were agreed upon standard `loss or damage criteria', their quantification would require the incorporation of heat transfer considerations into the kinetic model. For many products, the heat transfer calculations will require knowledge of the particular food's relevant physical constants (e.g., heat transfer coefficients) and their temperature dependence ± information that might not be readily available or easy to obtain. For this reason, we limit our discussion to establishing equivalent lethality, and only provide some comments on their potential implications for quality. The issue of equivalence can arise in different contexts but most commonly in: 1. Comparison of a dynamic process's efficacy in relation to that at a standard reference temperature, chemical agent concentration, or hydrostatic pressure, etc. 2. Comparison of a dynamic process's efficacy against that of another dynamic process of the same kind. 3. Comparison of the performance of one preservation method versus another, like disinfection with a dissipating chemical agent vis-aÁ-vis isothermal or dynamic heat pasteurization processes. 20.4.1 The equivalent time curve The equivalent isothermal time curve is a plot of the time at a lethal reference temperature, tequiv that produces the same survival ratio as the time, tproc, in the actual process be it static or dynamic (Corradini et al., 2006). The curve construction is shown schematically in Fig. 20.3. The plots in the figure are based on the assumption that the hypothetical organism's isothermal inactivation follows a Weibullian pattern with nT n < 1. The principle, however, is just as applicable to any monotonic inactivation pattern, i.e. regardless of whether the semi-logarithmic survival curve at the reference temperature is linear, concave upward or downward or sigmoid of type I or II (see Fig. 20.1). Suppose that the dynamic survival curves, which corresponds to a given temperature profile (Fig. 20.3(a)) was experimentally recorded or simulated by a computer using Eq. 20.11 or 20.12 as the survival model. Either way, if the organism's or spore's isothermal survival curve at the reference temperature is



The equivalent time curve construction. The equivalent time at a reference temperature, tequiv , is defined as the isothermal time at this reference temperature that produces the same survival ratio as the actual process at time tproc (after Corradini et al., 2006).

476


known (Fig. 20.3(b)), then the equivalent time curve (Fig. 20.3(c)) could be constructed graphically in the manner shown in the figure. If, however, the organism's or spore's isothermal survival curve is known a priori to be Weibullian with an exponent n and a rate parameter bTref , then tequiv could be calculated with the formula (Corradini et al., 2006; Peleg, 2006): tequiv

ÿlog10 Stproc bTref

1=n

20:13

Using Eq. 20.13, construction of the tequiv vs. tproc relationship from any experimental or simulated survival curve becomes straightforward. The same is true for any survival pattern whose primary model's equation has an analytical inverse, i.e. the time can be expressed algebraically as a function of the survival ratio, or its logarithm. Most, if not all, observed isothermal microbial survival patterns, including sigmoidal, can be described by a mathematical model that has an analytical inverse. Theoretically, however, an exception could arise that cannot be ruled out. The survival curve of a mixed microbial population having two inflection points is a possible example where the isothermal survival model's equation has no analytical inverse. In that case, tequiv would have to be written as a term that would yield its value through a numerical solution of the equation (Peleg, 2003b). Writing such a term in the syntax of MathematicaÕ is not difficult and the term, once defined, will be treated by the program as a standard function like Exp[x] or Cos[x]. Similar terms can be defined in the syntax of other advanced mathematical programs. Once expressed in this way, this `numeric tequiv' can be plotted against the process time in the same manner as that calculated algebraically with the analytical inverse. Three examples of dynamic survival curves converted into equivalent time curves are shown in Fig. 20.4. As seen in the figures, they are based on three different isothermal survival patterns and demonstrate that the methodology is applicable to different types of survival curves. In contradistinction to the `F0 value' calculation, that of the equivalent time, tequiv, does not have any prerequisites. This specifically eliminates the requirement that all the isothermal experimental survival curves must be log-linear, and that the D value's temperature dependence must be log-linear too. Consequently, tequiv can be considered a model-independent lethality index that is equally applicable to any monotonic survival patterns and not exclusively to log-linear kinetics. The derivation of tequiv does not require instituting a sterility criterion like the `12D C. botulinum cook' and extrapolating the survival curve to ratios several orders of magnitude below the detection level. Suppose now that it is known from experience (inoculated pack and incubation and storage studies, for example) that an isothermal process, which drives the logarithmic survival ratio below ÿ8, say, is safe, and that it takes 3 minutes to reach this level of inactivation at a given reference temperature. If an actual dynamic process had a tequiv of 6 minutes, say, this process would be at least as safe as the aforementioned isothermal process. The use of tequiv also allows setting a safety factor. For



Schematic view of the equivalent time curves of organisms or spores having linear and non-linear isothermal semi-logarithmic survival curves and different heat sensitivities.

ß Woodhead Publishing Limited, 2010 Fig. 20.5 Schematic view of how to construct the equivalent time curve for disinfection with a dissipating chemical agent.


479

example, if 3 minutes at the reference temperature is known to be sufficient, we may require that a commercial process should be twice as long, say, for added security. The introduction of a safety factor (2 in the above example) does not require extrapolation to survival ratios that have never been determined experimentally and is model independent (Corradini et al., 2009b). The destruction of microbial populations by non-thermal processes may not always reach 8±10 orders of magnitude (let alone twelve). Still, the method of assessing the efficacy of these non-thermal processes in terms of establishing an equivalent time can be as useful as in thermal processing. An example is given in Fig. 20.5, in which the efficacy of a chemical disinfection process using a dissipating agent is presented as equivalent time under a constant `reference' concentration. Notice that in the equivalent time derivation of this case, the lethal agent's concentration plays the same role as that of the lethal temperature in thermal processing. 20.4.2 Equivalent lethality of different preservation methods The most straightforward way to express a preservation process's adequacy is in terms of the number of decades reduction in the survival ratio that it achieves. Thus, comparing the final survival ratios of any two processes, whether of the same kind or different, will show their relative efficacy. For Food Processors and Food Technologists accustomed to relating equivalent time at a reference temperature as a safety measure, plots of the kinds shown in Figs 20.4 or 20.5 serve the same purposes. The middle two plots (see figures) represent the familiar isothermal heat treatment at a chosen reference temperature. The two plots on the left side represent the progress of hypothetical treatments by other preservation methods, such as HHP, chemical disinfection, irradiation, etc. Obviously, the time scales can be very different in the thermal and non-thermal treatments. Still, it is plausible that several hours of exposure of a pathogen to a dissipating chemical disinfectant can have the same lethal effect on a targeted pathogen as 2 minutes of pasteurization at 65 ëC, for example. The method can be extended to assess a process's efficacy with respect to two or more target organisms or spores, of C. botulinum and B. sporothermodurans. An example is given in Fig. 20.6. These plots were created with Eq. 20.12 as a model and using survival parameters derived from published data for these spores (Corradini et al., 2006; Periago et al., 2004). Results showed that the thermal process that would reduce the C. botulinum spores by 8 decades would reduce those of the B. sporothermodurans by only 4, which corresponds to equivalent times of 15 and 20 minutes at 121.1 ëC, respectively. Similarly, the procedure can be used to evaluate the performance of different temperature profiles on the process's adequacy. As shown in Fig. 20.7, this option allows evaluation of the effect of temperature variations within a retort or between retorts on the product's safety. The variations in the process's temperature uniformity here are manifested in the number of decades reduction range and/or the equivalent time at the chosen reference temperature. The


480


Fig. 20.6 Comparison of the equivalent time curves of C. botulinum and B. sporothermodurans spores subjected to two hypothetical heat treatments. The WeLL model's survival parameters used to generate the plots were derived from published data (after Peleg et al., 2008a).



481

Fig. 20.7 Schematic view of how variations in a heat treatment's temperature profile are manifested in the survival and equivalent time curves of a microorganism or spore.


482


accurate assessment of nutrient losses, flavor deterioration, or discoloration should be based on actual temperature±time histories at local regions within the product's entire volume, and not simply at the `coldest point' during processing. It should be noted, that when it comes to quality attributes, `coldest points' might actually provide the best-case scenario not the worst. Using the kinetics parameters of quality loss (the thermal degradation rate constant of thiamine, for example) one could, in principle, estimate the equivalent time in minutes that a process would effect the same amount of loss as would occur at a given reference temperature.

20.5

Freeware

Interested readers can implement the concepts described in this chapter and the publications on which it is based by using freeware written by Mark D. Normand that has been posted on the Web in two forms. The first is a series of downloadable programs written in MS ExcelÕ [available at: http://wwwunix.oit.umass.edu/~aew2000/GrowthAndSurvival/GrowthAndSurvival.html]. They are set for thermal pasteurization and sterilization and for chemical disinfection. Each program comes in two versions. In one, the `profile' is generated by a mathematical formula with parameters entered or modified by the user. The other allows the user to paste experimental time-temperature or time-concentration data. In both versions, the user can enter the target organism's survival parameters and choose the reference temperature. All the program calculations are done with the WeLL model (Eqs 20.12 and 20.13) or its concentration equivalent. Examples of the screen appearance are given in Fig. 20.8. An additional spreadsheet allows the user to plot simultaneously up to five different temperature profiles and/or target organism combinations, and watch the corresponding survival and tequiv vs. tproc time curves. In the PC version of the programs, the survival and equivalent time curves can be seen plotted simultaneously in real time. The second freeware form consists of five programs that can be downloaded from Wolfram Demonstration Project's Website [available at: http://wwwunix.oit.umass.edu/~aew2000/WolframDemoLinks.html#inactivation] sponsored by Wolfram Research Inc. (Champaign, IL). The first of the five `demonstration programs' generates and plots the temperature profile; the second generates and plots the bT vs. T relationship; the third generates and plots the survival curves that corresponds to the chosen set of temperature profile parameters; the fourth demonstration the corresponding tequiv vs. tprocess curve; and the fifth generates and plots variations in the Weibullian rate parameter bTt during the chosen thermal process. These five programs provide userfriendly convenience to adjust the temperature profile parameters, the target organism survival parameters, and the reference temperature by moving sliders on the screen (Fig. 20.9). Thus the user can readily examine a large number of hypothetical scenarios and their theoretical consequences and the effect of



483

Fig. 20.8 A computer screen after running the free downloadable MS ExcelÕ program that generates the temperature profile and calculates the corresponding survival and equivalent time curves. Notice that the profile characteristics, the organism or spores' survival parameters and the reference temperature are set by the user. For more details see text and Peleg et al. (2005) or Peleg (2006).

variations in the temperature profile and the target organism survival parameters on the process efficacy. The MS ExcelÕ and the MathematicaÕ programs can be used in academic and industrial training. In principle, the MS ExcelÕ version can be incorporated into a food plant's process monitoring system to translate the continuously logged time±temperature histories into equivalent time at a chosen reference temperature in real-time. This equivalent time, or the corresponding survival ratio, can also be used to control the thermal process (with the needed safety factor) in order to guarantee a safe product.

20.6

Conclusions

Despite its known deficiencies, the most widely used microbial lethality measure in the canning industry is still the `F0 value'. This chapter shows that


484


Fig. 20.9 An example of three Wolfram Demonstrations. Shown above are the temperature profile and its sliders settings, and the corresponding survival curve with its settings (which include both the profile's and the organism's survival parameters) and, on the facing page, the equivalent time curve and its settings (which include the temperature profiles, the organism's survival parameters and the reference temperature). Notice that all the generation parameters can be set by moving sliders with the mouse or entered numerically. For more details see text and Peleg et al. (2008b).



485

Fig. 20.9 Continued

equivalent lethality can be established without the key assumptions on which the `F0 value' concept is based, i.e., that microbial inactivation always follow firstorder kinetics and that the exponential rate constant has log-linear temperature dependence. These assumptions are avoided by allowing the isothermal survival curve to be treated as a process following non-linear kinetics and the temperature dependence of its parameter by ad hoc empirical models. The Weibull-Log logistic (WeLL) model is a notable example of a resulting model that has been able to predict correctly the dynamic inactivation patterns of several microorganisms. As shown in this chapter, free user-friendly software to do sterility calculations is now available and it renders the inactivation rate equations solution an almost trivial task. The same can be said about nonthermal preservation processes. Their efficacy too can be easily calculated and compared in terms of the accomplished final survival ratio and/or equivalent time under reference conditions. This allows the food microbiologist, technologist or engineer to establish (or refute) the equivalent lethality to thermal or other preservation processes of known efficacy. An issue not yet fully resolved is how to model the efficacy of combined processes, such as HHP at high temperatures or chemical disinfection at an elevated or oscillating temperature (Peleg, 2006). However, once the survival curve corresponding to the combined process is recorded experimentally, it can


486


be converted into an equivalent time under any set of reference conditions, regardless of whether the treatment is isothermal, non-isothermal, or nonthermal. Estimating the equivalent isothermal time curves for nutrient degradation and other kinds of quality losses like texture and flavor is also still a problem, although the mathematical tools to do the calculations already exist. The solution to this problem will require either a method to quantify the heterogeneous distribution of the loss within the container's total volume or to identify a representative `coldest point.' In heat-processed foods, the solution will most probably come in the form of models that combine microbial inactivation kinetics with heat transfer theories. And finally, a word of caution is in order here. As demonstrated in this chapter, kinetic models and the calculations based on them can be very useful in establishing the equivalency of different processes. However, the models should only be used as a screening device. While they can clearly identify risky process conditions, an actual process's safety should only be established and validated experimentally, i.e., by microbiological testing and other protocols.

20.7

Disclaimer

The mention of any commercial product in this chapter does not imply its endorsement over other products by the authors or the institutions with which they are affiliated.

20.8

References

and PELEG, M. (2003a) A theoretical note on estimating the number of recoverable spores from survival curves having an àctivation shoulder'. Food Research International, 36, 1007±1013. CORRADINI, M. G. and PELEG, M. (2003b) A model of microbial survival curves in water treated with a volatile disinfectant. Journal of Applied Microbiology, 95, 1268± 1276. CORRADINI, M. G. and PELEG, M (2004) Demonstration of the Weibull-Log logistic survival model's applicability to non-isothermal inactivation of E. coli K12 MG1655. Journal of Food Protection, 67, 2617±2621. CORRADINI, M. G., NORMAND, M. D. and PELEG, M. (2006) On expressing the equivalence of non-isothermal and isothermal heat sterilization processes. Journal of the Science of Food and Agriculture, 86, 785±792. CORRADINI, M. G., NORMAND, M. D. and PELEG, M. (2008) Prediction of an organism's inactivation patterns from three single survival ratios determined at the end of three non-isothermal heat treatments. International Journal of Food Microbiology, 126, 98±111. CORRADINI, M. G., NORMAND, M. D., NEWCOMER C., SCHAFFNER, D. W. and PELEG, M. (2009a) Extracting survival parameters from isothermal, isobaric and ìso-concentration' inactivation experiments by the `three end points method'. Journal of Food Science, 74, R1±R11. CORRADINI, M. G.



487

and PELEG, M. (2009b) Direct calculation of the survival ratio and isothermal time equivalent in heat preservation processes. In: Simpson, R. (ed.) Engineering Aspects of Thermal Processing, Boca Raton, FL, CRC Press, pp. 211±230. DATTA, A. K. (1993) Error estimates for approximate kinetic parameters used in food literature. Journal of Food Engineering, 18, 181±199. DOONA, C. J., FEEHERRY, F.E., ROSS, E.W., CORRADINI, M.G. and PELEG, M. (2007). The quasichemical and Weibull distribution models of nonlinear inactivation kinetics of E. coli ATCC 11229 by high pressure processing. In: Doona, C. J. and Feeherry, F. E. (eds), High Pressure Processing of Foods, Ames, IA, IFT Press and Blackwell Publishing, pp. 115±144. FDA (2000) Kinetics of microbial inactivation for alternative food processing technologies. www.cfsan.fda.gov/~comm/ift-over.html JAY, J. M. (1996) Modern Food Microbiology, New York, Chapman and Hall. LEWIS, M. and HEPPEL, N. (2000) Continuous Thermal Processing of Foods: Pasteurization and UHT Sterilization, Gaithersburg, MD, Aspen Publishers. MAFART, P., COUVERT, O., GAILLARD, S. and LEGUERINEL, I. (2002) On calculating sterility in thermal preservation methods: application of the Weibull frequency distribution model. International Journal Food Microbiology, 72, 107±113. NUNES, R. V., SWARTZEL, K. R. and OLLIS D. F. (1993) Thermal evaluation of food processes: the role of a reference temperature. Journal of Food Engineering, 20, 1±15. PARDEY, K. K., SCHUCHMANN, H. P. and SCHUBERT, H. (2005) Modelling the thermal inactivation of vegetative microorganisms. Chemie Ingenieur Technik, 77, 841±852. PELEG, M. (2002) A model of survival curves having an àctivation shoulder'. Journal of Food Science, 67, 2438±2443. PELEG, M. (2003a) Calculation of the non-isothermal inactivation patterns of microbes having sigmoidal isothermal semi-logarithmic survival curves. Critical Reviews in Food Science and Nutrition, 43, 645±658. PELEG M. (2003b) Microbial survival curves: Interpretation, mathematical modeling and utilization. Comments on Theoretical Biology, 8, 357±387. PELEG, M. (2006) Advanced Quantitative Microbiology for Food and Biosystems: Models for Predicting Growth and Inactivation. Boca Raton, FL, CRC Press. PELEG, M. and COLE, M. B. (1998) Reinterpretation of microbial survival curves. Critical Reviews in Food Science and Nutrition, 38, 353±380. PELEG, M. and NORMAND, M. D. (2004) Calculating microbial survival parameters and predicting survival curves from non-isothermal inactivation data. Critical Reviews in Food Science and Nutrition, 44, 409±418. PELEG, M. and PENCHINA, C. M. (2000) Modeling microbial survival during exposure to a lethal agent with varying intensity. Critical Reviews in Food Science and Nutrition, 40, 159±172. PELEG, M., NORMAND, M. D. and CAMPANELLA, O. H. (2003) Estimating microbial inactivation parameters from survival curves obtained under varying conditions ± the linear case. Bulletin of Mathematical Biology, 65, 219±234. PELEG, M., CORRADINI, M. G. and NORMAND, M. D. (2004) Kinetic models of complex biochemical reactions and biological processes. Chemie Ingenieur Technik, 76, 413±423. PELEG, M., NORMAND, M. D. and CORRADINI, M. G. (2005) Generating microbial survival curves during thermal processing in real time. Journal of Applied Microbiology, 98, 406±417. CORRADINI, M. G., NORMAND, M. D.


488


and TER STEEG, P. M. (2008a) Estimating the heat resistance parameters of bacterial spores from their survival ratios at the end of UHT and other heat treatments. Critical Reviews in Food Science and Nutrition, 48, 634±648. PELEG, M., NORMAND, M. D. and CORRADINI, M. G. (2008b) Interactive software for estimating the efficacy of non-isothermal heat preservation processes. International Journal of Food Microbiology, 126, 250±257. PELEG, M., NORMAND, M. D., CORRADINI, M. G. VAN ASSELT, A., DE JONG, P. C.

PERIAGO, P. M., VAN ZUIJLEN, A., FERNANDEZ, P. S., KLAPWIJK, P. M., TER STEEG, P. F., CORRADINI, M. G. and PELEG, M. (2004) Estimation of the non-isothermal inactivation patterns of Bacillus sporothermodurans IC4 spores in soups from their isothermal survival data. International Journal of Food Microbiology, 95, 205±218. SCHUBERT, H., WACHLER, E. and KRUG, H. (1984) Erich Rammler ± a pioneer of particulate technology. Particulate Science Technology, 2, 2±17. SMELT, J. P., HELLEMONS, J. C. and PATTERSON, M. (2002) Effects of high pressure on vegetative microorganisms. In: Hendrick, M. E. G. and Knorr, D. (eds), Ultra High Pressure Treatment of Foods, Food Engineering Series, New York, Kluwer Academic/Plenum Publishers. TOLEDO, R. T. (1999) Fundamentals of Food Process Engineering, 2nd edn, Gaithersburg, MD, Aspen Publishers. VAN BOEKEL, M. A. J. S. (2002) On the use of the Weibull model to describe thermal inactivation of microbial vegetative cells. International Journal of Food Microbiology, 74, 139±159.


21 A case study in military ration foods: the Quasi-chemical model and a novel accelerated three-year challenge test C. J. Doona, F. E, Feeherry and E. W. Ross, US Army Natick Soldier RD&E Center, USA

Abstract: Maple-filled French toast (MFFT) enrobed sandwiches are intermediate moisture (IM) bi-layered foods formulated with `hurdles' to prevent the growth of Staphylococcus aureus for product shelf-life (3 years at T 80 ëF 27 ëC). A novel approach to accelerate microbial challenge testing was investigated. The Quasi-chemical model's secondary model (`growth-nogrowth boundary') was used to guide the formulation of the MFFT. The novel 3-year test protocol used a five-strain cocktail of S. aureus to challenge MFFT with variations in storage temperature, product formulation, and packaging atmosphere. The inactivation kinetics were evaluated with the modified Quasichemical model and determined to be more rapid at T 35 ëC than at T 25 ëC. Removing the oxygen scavenger did not promote S. aureus growth. These results provide a basis for standardizing accelerated microbial challenge tests for new varieties of IM enrobed sandwiches in the pipeline to save time, money, and labor, while ensuring food safety. Keywords: Quasi-chemical model, S. aureus growth/death kinetics, Maplefilled French toast intermediate moisture enrobed sandwich, accelerated 3year microbial challenge test.

21.1

Introduction

Minimally processed intermediate moisture (IM) enrobed sandwiches are convenient on-the-move, eat-out-of-hand (no utensils) nutrient-rich meals that require no end-consumer preparation steps. These `pocket' sandwiches are


490


Fig. 21.1 Many new varieties of `pocket' sandwiches are in development, as shown in (a) and (b), and are popular among consumers for their good taste and convenience as eaton-the-move foods.

popular among consumers of military rations because they satisfy consumer demand for safe, more fresh-like foods, while also featuring high quality, and improved organoleptic attributes. One well-known example is the popular barbecue chicken `pocket' sandwich (ABC News ± World News Tonight, 2002; BBC News, 2002; CNN.com, 2002; Cook, 2002; Fabricant, 2002; GrahamRowe, 2002; Yahoo! News, 2002). Military ration systems are steadily increasing the types and varieties of enrobed sandwiches meeting the extended shelflife requirements (Fig. 21.1). Many types of these enrobed sandwiches are derivatives based on IM pouch bread ± a loaf of pouch bread packaged in trilaminate foil with an oxygen scavenger. A relatively recent and popular enrobed sandwich item is Maple-filled French toast (MFFT), a new variety of high quality, minimally processed, IM breakfast sandwich being developed to meet consumer demand (Fig. 21.2). IM foods are defined (Karel, 1973) conventionally as having Aw levels ranging from 0.7 to 0.9, and moisture content ranging from 20 to 50%. Minimally processed IM foods, in general, tend to feature more fresh-like character, retain increased sensory attributes such as flavor, texture, and color, have higher nutritive content, and receive higher consumer acceptance than their thermally treated counterparts. It is important that minimally processed IM foods feature not only improved organoleptic attributes and consumer acceptance, but that they are safe microbiologically. Staphylococcus aureus is a notorious pathogen that thrives in the ecological niche (low Aw, low moisture content) characteristic of IM foods (Lotter and Leistner, 1978). The potential microbial hazards associated with S. aureus in IM enrobed sandwiches are the possibilities of post-processing contamination (the sandwiches do not undergo an end-treatment to act as an effective kill step to eliminate pathogenic microorganisms), survival for extended periods in low Aw foods such as powdered dry milk (see Dow Jones/Agence France Presse English, 2000), and for recovery from sub-lethal injury during the protracted 3-year storage period (the `Phoenix phenomenon,' see Jay, 1996) in microdomains of relatively high


A case study in military ration foods: the Quasi-chemical model 491

Fig. 21.2 Maple-filled French toast enrobed breakfast sandwich (a) top view and crosssectional view showing moist maple-flavored filling, and (b) diagram of cut interface for inoculation with 5-strain S. aureus cocktail across the cut surface to include bread, filling, and the bread/filling interface.

moisture content and Aw created by moisture migration from the filling to the bread and dispersed inhomogeneously throughout the microstructure of the bread matrix. A factor confounding the microbiological stability of complex, bi-layered foods such as enrobed sandwiches, cream-filled cakes, and pastries is the potential for moisture transfer to occur between the two separate phases during storage that may render a previously stable phase unstable (Tatini, 1973; Rajkowski et al., 1994). For example, dried salami slices stable at Aw 0.85 did not support S. aureus growth, however, storage for 48 h of these meat slices in contact with cheese at Aw 0.91 increased the Aw of these salami slices to Aw 0.91, rendering them capable of supporting S. aureus growth (Rajkowski et al., 1994). Another factor to consider with regard to moisture migration from the filling to the bread during storage for 3 years at ambient temperatures is potential sogginess. Enrobed sandwiches for military rations must retain physico-chemical stability and high quality without the degradation of texture. Formulating and designing IM foods to attain stability over shelf-life with respect to preventing microbiological growth and/or degradation reactions is generally referred to as `hurdle technology' (Leistner et al., 1981). Hurdle technology encompasses a number of methods to impart food stability (microbiological and physico-chemical) by controlling intrinsic factors of the food (such as pH, salinity, and Aw by adjusting the levels of various humectants), by adding anti-microbial compounds or the presence of competing microflora, and by controlling extrinsic factors (atmospheric conditions such as storage temperature, relative humidity, and ambient pressure). Achieving this with IM foods generally requires imposing hurdles of low Aw, low pH, and other factors that influence microbial survivability. It is therefore essential to formulate


492


enrobed sandwiches to be nutritious, have high consumer ratings, and be detrimental to the survivability of potential microbial hazards such as S. aureus. Predicting pathogen growth in foods constrained by `hurdles' (Leistner, 1994; Leistner et al., 1981) has enormous cost-savings potential for food product development by generating guidelines for HACCP plans and reducing the need for extensive and time-consuming microbiological challenge testing. Predictive mathematical models are tools that provide convenient methods for assessing and predicting the criteria that constrain the growth of pathogens or spoilage microorganisms in foods and ensuring food safety or quality (McMeekin et al., 1993). Accordingly, it is essential to know how bacterial populations grow and die in response to the factors that characterize a food product, and to develop appropriate secondary models. Feeherry et al. (2003) examined the ability of military IM pouch bread to support S. aureus growth over the Aw range 0.836± 0.909 at pH 5.2±5.5, then used these results to define conditions that prevent the growth of S. aureus in pouch bread formulations and other IM foodstuffs. Using the four-parameter logistic equation to model S. aureus growth kinetics as functions of food properties (Aw and pH), Feeherry et al. (2003) noted two things: first, in cases where death-only kinetics occurred, the logistics equation could not model the data without first making some significant modifications; and second, after growth to a stationary phase, the data began to show a decline. This data had to be subjectively removed to be consistent with conventional growth studies. Clearly, a new, more comprehensive model was needed to account for continuous growth-death kinetics, and the model had to be versatile and convenient for use in conditions producing (nonlinear) inactivation kinetics without requiring end-user modifications. The Quasi-chemical model is a unique and versatile model suitable for fitting continuous growth-death kinetics and also for fitting nonlinear death-only kinetics without modifying the fundamental structure of the model (Doona et al., 2005; Ross et al., 2005; Taub et al., 2003; Feeherry et al., 2001). Other kinetics models, such as the logistic equation, are sigmoidal functions that, despite their success, do not generally model growth from a lag phase through an asymptotic maximum and followed by death. In fact, experimental conditions are generally configured to isolate growth from inactivation kinetics, and distinct models are applied to evaluate each type of kinetics separately (Gianuzzi et al., 1999). By fitting the growth-death kinetics data, the Quasi-chemical model is a primary model that accurately determines certain useful kinetics parameters (growth rate, lag time). The Quasi-chemical modeling approach was used to interrelate these kinetics parameters with the environmental conditions (Aw, pH, T) and create a unique secondary model called the growth/no-growth boundary line that defines food properties that prevent S. aureus growth based on statistical analysis of the data (Taub et al., 2003). This secondary model was used to predict food formulations for MFFT and other IM foods safe from the growth of S. aureus (Taoukis and Richardson, 2007). To ensure the safety of foods, microbial challenge studies are designed and carried out in accordance with written guides (Swanson, 2005; US Department


A case study in military ration foods: the Quasi-chemical model 493 of Health and Human Services, Food and Drug Administration, 2001; NSF International, 2000; Powers et al., 1999) for commercial products with substantially shorter shelf-life requirements than military ration components. Microbial challenge studies should be conducted under conditions (temperature packaging, environment, etc.) as similar as possible to those the actual product experiences for durations of (1±1.3) the product shelf-life to ensure consumer safety. Since pouch bread-based enrobed sandwiches are formulated or designed not to support the growth nor survival of target pathogens or other microorganisms, microbial challenge studies of military ration foods typically involve sampling at appropriate intervals over 6 month storage spans at the temperature of 25 ëC to ensure safety. The `Phoenix phenomenon' involves the possible re-growth of sub-lethally injured or non-recoverable organisms after long-term storage (Jay, 1996). Given the long-term survival of S. aureus in dry milk powder (Dow Jones/Agence France Presse English, 2000), the potential for injured cells to recover and grow in the nutrient-rich environment of the MFFT, despite its low Aw and pH that could increase with moisture transfer between layers during 3 years of ambient storage-growth, is conceivable. However, a microbial challenge test of 3±4 years duration, while technically feasible, is uneconomical and impractical from a typical product development standpoint. As new enrobed, IM food products are developed to meet military requirements for high consumer acceptance, stability, and food safety during prolonged storage at non-refrigerated temperatures (3 years at T 80 ëF 27 ëC), an accelerated microbial challenge test protocol is needed to ensure the safety of these foods while reducing the duration of these studies (microbial challenge tests usually last for product shelf-life, see below). Analogous to how trained sensory panels evaluate foods using accelerated temperatures (6 months at T 100 ëF 38 ëC) to expedite the development process, we have recently studied the effects of temperature on microbiological challenge tests of enrobed bi-layered sandwiches undergoing long-term storage and developed an accelerated temperature microbiological challenge test to shorten the time while ensuring the safety of IM pocket sandwiches with respect to S. aureus, a pathogen notorious for growing (or surviving) in low Aw foods. We compare these results to samples stored for 3 years and evaluate kinetics data using the Quasi-chemical model adapted for kinetics data showing tailing (Doona et al., 2008; Feeherry et al., 2005). Predictive microbiological models help assess whether additional validation testing is needed to prevent the potential growth of the target organism. In this case, the modified Quasi-chemical model provides another tool to demonstrate that the present formulations of MFFT enrobed sandwiches do not support S. aureus growth over 3-year storage. We apply the Quasi-chemical model to a case study of MFFT and evaluate a novel approach to accelerating the duration of microbial challenge study, using a 5-strain cocktail of S. aureus to challenge MFFT with variations in storage temperature, product formulation, and packaging atmosphere, while ensuring that S. aureus does not survive and re-grow in response to moisture migration during prolonged 3-year storage (the `Phoenix phenomenon'). Results of this case study


494


show that MFFT IM bi-layered breakfast sandwiches controlled by appropriate hurdles (Aw 0.851, pH 5.36) receive high Hedonic ratings by consumer panels, retain their physico-chemical stability during storage, and maintain microbiological safety with respect to S. aureus for the protracted 3-year storage. Modified atmosphere packaging (MAP) helps inhibit S. aureus growth, but removing the O2 scavenger did not promote S. aureus growth in these circumstances. In fact, only inactivation kinetics of S. aureus were observed, and no re-growth was observed in a 3-year shelf-life study. More surprisingly, the inactivation kinetics of S. aureus were more rapid at T 35 ëC (near the growth optimum of S. aureus) than at T 25 ëC, which may reflect the greater difficulty for injured S. aureus cells to meet their metabolic energy demand near the optimum of their microbial ecological nicheÂ than are required at sub-optimal storage conditions. These results may lead to a standardized, accelerated protocol for microbial challenge testing of enrobed, IM foods that can ensure reliable results faster, cheaper, and with less labor than current practices.

21.2 Modeling S. aureus growth in intermediate moisture (IM) bread Feeherry et al. (2003) conducted a conventional growth study of S. aureus to a stationary phase maximum using standard pouch bread with adjustments to the Aw to specific levels by desorptive (varying the amounts of glycerol added to the dough) or adsorptive (equilibrating in closed chambers of certain relative humidities using saturated salt solutions) techniques, and with adjustments to the pH by adding acidulant to the dough. Growth data of S. aureus in bread with variations in Aw and pH were plotted as colony counts versus time and fit with the 4-parameter logistic function (Eq. 21.1), which in all cases gave excellent fits (R2 0:939±0.996). log Nt

log Nf =Ni log Ni 1 exp btm ÿ t

21:1

in which Nt is the survivors at time t; Ni is the enumerated inoculum level; Nf is the survivors at the final measured time; tm is the time of the maximal growth rate; and b is the maximum growth rate, determined as the slope of the line tangent to the curve at tm (Fig. 21.3). In the nutrient-rich environment of food, bacteria and other organisms tend to exhibit a characteristic pattern called the microbial lifecycle (McMeekin et al., 1993). In this lifecycle, the bacteria or other organisms can survive or be stimulated to grow after an initial relatively quiescent initial period with a relatively slow growth rate that generally produces only a small or modest increase in population (called the lag phase). As the bacteria metabolize and reproduce, the growth rate increases dramatically, leading to a relatively sudden increase in population size (called the exponential growth phase). The rate of growth then declines asymptotically to zero, and the population increases in



Fig. 21.3 S. aureus growth in bread as functions of (a) Aw varied by adjusting the glycerol level in dough (l = 0.0% added glycerol and Aw = 0.909; n = 3.0% and Aw = 0.891; s = 6.3% and Aw = 0.866; and u = 9.0% and Aw = 0.839 and pH 5.5); (b) pH adjusted by adding glucono-delta-lactone (GDL) (l = 0.0% added GDL and pH 5.38; n = 0.05% and pH 5.36; and s = 0.1% and pH 5.31 and Aw 0.86).

number until the population density reaches an approximately constant maximum stationary value (called the asymptotic or stationary phase). This so-called stationary phase is not a true steady-state, rather it is indicative of a microbial population in which the rates of growth and death approximately cancel. As the bacterial population ages further, nutrients deplete and excreted metabolic waste products accumulate in the surroundings, and eventually the population density declines due to natural effects (called the death phase and represents dead and injured, non-culturable cells). Conventional methods involve actually adjusting the experimental conditions to isolate growth (lag through stationary phases) kinetics or inactivation kinetics of the target microorganism, depending on the purpose of the study, and


496


distinctly different predictive models are applied to evaluate the kinetics of each type of response. In the growth study of S. aureus in IM bread using the logistic function, Feeherry et al. (2003) revealed some limitations in the conventional approach to growth studies, notwithstanding the success of the research. After the growth in S. aureus counts reached a maximum, the data started to decline, in a manner consistent with the microbial lifecycle. The logistic function cannot fit such continuous growth-death kinetics data, and data had to be removed subjectively (in accord with convention) to ensure the most accurate fit of the data to the model. However, excluding the declining or so-called death data influences the estimated values of growth rates that are determined from the growth curves and that are used to estimate product shelf-life. Since variations in physico-chemical properties were explored to define the limits of conditions that would support S. aureus growth in IM pouch bread, another limitation was noted for some variations in experimental conditions when only death kinetics were observed. The logistics equation expressed above cannot model death-only kinetics, although it is adaptable to a form capable of describing death-only kinetics to overcome that issue. Recognizing the limitations of the logistic equation, a new type of predictive model was needed that could reduce the subjectivity in handling continuous growth-decline data and that was versatile enough to evaluate death-only data without requiring modification by the enduser of the program. We therefore developed and applied a 5-parameter variant of the Quasi-chemical kinetics model that accounts for tailing (Doona et al., 2008; Feeherry et al., 2005). This 5-parameter version is not a fully mechanistic model, but only one of the simplest mathematical forms capable of representing tailing and the potential for recovery of cells from sub-lethal injury that could render cells dead or non-culturable. 21.2.1 The Quasi-chemical model The Quasi-chemical kinetics model for growth-death kinetics accounts for all four phases of the microbial lifecycle and has been used to evaluate a variety of nonlinear kinetics characterizing the growth and/or death of pathogenic microorganisms in foods stabilized by hurdles (Doona et al., 2005; Ross et al., 2005; Taub et al., 2003; Feeherry et al., 2001) or treated with the emerging nonthermal processing technologies of high pressure processing (Doona et al., 2005, 2008; Feeherry et al., 2005). As demonstrated below, the model fits continuous growth-death kinetics of S. aureus in IM bread in various conditions of Aw, pH, and temperature, and these results were used to define a unique secondary model called a `growth/no-growth boundary line' based on interrelating variations in environmental conditions of Aw, pH, and temperature with kinetics parameters (maximum growth rate). Continuous modeling of microbial growth-death kinetics in actual foods advances predictive modeling that conventionally separates growth and death models. The Quasi-chemical mathematical model (Taub et al., 2003) is derived from a proposed 4-step hypothetical mechanism involving four entities, and the steps



Fig. 21.4 Schematic mechanism of the Quasi-chemical model denoting cells in the stages of: metabolizing (M), multiplying (M*), sensitization to death (M**), and dead and injured (D), and the hypothetical antagonist (A), and proceeding to the next stages with rate constants denoted as k.

proceed according to their respective rates and align essentially as a reaction network involving autocatalytic growth coupled with negative feedback through the activity of a postulated antagonistic metabolite (Fig. 21.4). Analogous to describing the rates of the elementary steps in a chemical reaction system, each step is treated as an individual kinetics process that is described by a corresponding rate expression, and the rate expressions derive a set of ordinary differential equations (ODEs ± see Table 21.1) that impart the Quasi-chemical model with several unique and advantageous features. The mathematics of the Quasi-chemical model have been characterized extensively (Ross et al., 2005), including determining relationships among individual rate parameters that impart the model with the ability to model continuous growth-death kinetics or nonlinear death kinetics in foods controlled by hurdles such as Aw, pH, temperature, and in foods subjected to treatments by HPP. The Quasi-chemical model successfully fits the growth-death kinetics or nonlinear death-only kinetics for a generalized set (Doona et al., 2005; Taub et al., 2003) of pathogenic microorganisms (S. aureus, Escherichia coli, and Listeria monocytogenes) in a variety of actual IM food substrates (bread crumb, turkey meat, ham, and cheeses), and as functions of different hurdles, such as Aw Table 21.1 model

Mechanism, rate equations, and ODEs compromising the Quasi-chemical

Mechanism

Rate equations

Differential equations

i) ii) iii) iv) v)

v1 v2 v3 v4 v5

dM/dt v1 dM*/dt v1 + v2 v3 v4 dA/dt v2 v3 dD/dt v3 + v4 (no tailing) dD/dt (v3 + v4) v5 (with tailing)

M ! M* M* ! 2 M* + A M* + A ! D M* ! D D ! M*

k1 M k2M* k3M*A k4M* k5 D


498


Fig. 21.5 Data and fitted curves of S. aureus kinetics in bread as functions of (a) Aw (l = 0.91; s = 0.87; and u = 0.84) at pH = 5.4, T 35 ëC; (b) pH (curves in descending order with pH = 5.38, 5.36, 5.31, 5.19, and 4.97) at Aw 0.86 and T 35 ëC; and (c) temperature T 15±40 ëC (l = 15 ëC, n = 20 ëC, t = 25 ëC, k = 30 ëC, s = 40 ëC, and u = 35 ëC) at Aw = 0.90 and pH = 5.23.

(determined adsorptively and desorptively using the humectant glycerol), pH (manipulated through the addition of the acidulant glucono-delta lactone), and storage temperature, and in the presence of anti-microbial compounds commonly found in foods (lactate and plum pureeÂ). Figure 21.5 shows fits of the Quasi-chemical C model for S. aureus growth in bread with variations in Aw, pH, and temperature, respectively. Figure 21.6 shows the plot of experimentally determined CFU/mL versus time data (filled circles) and three methods to estimate the maximum growth



Fig. 21.5 Continued

Fig. 21.6 Fitted curves comparing Quasi-chemical model and Gompertz model for estimating maximum growth rate (solid line = Quasi-chemical model of all data); hashed line = Gompertz fit of growth-only data; and heavy dashed line = Quasi-chemical model fit of growth-only data).

rate. The solid line depicts the Quasi-chemical model fit using all of the growthdeath data, the dotted line represents the Gompertz model fit using the growth phase only data; and the dashed line characterizes the Quasi-chemical model fit using the growth-only data. All of the fits agreed with the growth phase data, with some differences in the estimation of the growth rate (Table 21.2). The


500


Table 21.2 models

Comparison of estimated growth rates for the Quasi-chemical and Gompertz

Model

Quasi-chemical (all data) Gompertz (growth only data, t 0±7) Quasi-chemical (growth only data, t 0±7)

1.33 1.82 1.48

Gompertz function estimated a higher value of (1.82) than the corresponding value determined with the Quasi-chemical model ( 1:48), or than the Quasichemical model fit estimated for the continuous growth-death data ( 1:33). The growth/no-growth boundary line is estimated by applying a statistical approach that interrelates the parameters Aw, pH, and . Figure 21.7 shows a plot of the (Aw, pH) plane and the observed response of the S. aureus (growth filled circles, no-growth filled triangles) at each combination of pH and Aw (at T 35 ëC) used in the characterization of the growth kinetics. The calculated boundary line divides the plane into `growth' (right of the diagonal line) and `nogrowth' (left of the diagonal) domains that is concordant with the experimental data with the exception of one point (a fail-safe). The boundary line can be used as a guide for facilitating the formulation and design of pouch bread-based products (e.g., IM enrobed sandwiches) to be in the no-growth domain and safe from supporting the growth of S. aureus (Taoukis and Richardson, 2007).

Fig. 21.7 `Growth/no-growth' boundary line is a diagonal line though the (Aw, pH) plane and defining the conditions of pH and Aw in domains of growth (right side of diagonal) and no-growth (left side of diagonal) for S. aureus in bread. Filled circles (l) represent growth, t denote death, and and r depict validation points.



21.3

Microbial challenge study of Maple-filled French toast

Microbial challenge studies ensure the safety of foods such as enrobed sandwiches relying on hurdles (Aw, pH, MAP) to control relevant microorganisms (e.g., S. aureus). There is no single universal method for microbial challenge testing of S. aureus or other organisms in all food products, storage conditions, and processing or handling situations. Rather, the variations in microbial challenge testing conducted by food industry and academia, government research laboratories, and other facilities reflect differences in various factors relating to the specific food product and the purpose of the test. To design an appropriate microbial challenge study of the MFFT enrobed IM breakfast sandwich that yields reliable and accurate results of safety, we used L. monocytogenes Challenge Study `How To' Guidelines (Swanson, 2005), Evaluation and Definition of Potentially Hazardous Foods ± Ch. 6 Microbiological Challenge Testing (US Department of Health and Human Services, Food and Drug Administration, 2001), Non-potentially Hazardous Foods (NSF International, 2000), and Effect of Water Activity on the Microbiological Stability of Mobility-Enhancing Ration Components (Powers et al., 1999). When designing a microbiological challenge test, a standard set of several critical factors must be taken into account, and the microbial challenge test for MFFT must take into account the unique attributes of MFFT and the purpose of this test to meet long-term military storage requirements. The particular points that needed to be taken into special consideration in addressing enrobed, IM, bilayered MFFT breakfast sandwiches involve study duration, moisture transfer between layers, the potential for the Phoenix phenomenon to occur, and how to accelerate the time of the study like sensory studies to save time, money, and labor, while determining the safety of the MFFT. Since microbiological challenge tests should last (1±1.3) product shelf-life at T 25 ëC, a study length of 36±48 months would be needed to satisfy military shelf-life requirements. The potential for moisture migration occurring during storage may induce changes in the physico-chemical properties of regions of the food to become supportive of S. aureus growth or allow a small number of injured cells to recover and grow in the product during long-term storage (the so-called Phoenix phenomenon ± see Jay, 1996) is a particular concern in connection with the outbreak of S. aureus in dry milk powder (Dow Jones/Agence France Presse English, 2000). It is also important to monitor the physico-chemical properties of both layers of the enrobed MFFT sandwiches (Aw and pH) during storage. The factor under consideration for accelerating the microbial challenge test here is temperature. Since the observed growth optimum for S. aureus in bread is T 35 ëC (Fig. 21.5(c)) and MFFT is a pouch bread-based product, the question therefore arises whether T 35 ëC is an optimum growth condition for S. aureus growth in MFFT that can reliably shorten the duration of the microbial challenge study while rigorously ensuring food safety to save time, money, and labor. MFFT samples for this microbial challenge test were prepared either in-house at Natick or by a commercial food product manufacturer. One difference,


502


Table 21.3 Ingredients of MFFT enrobed sandwiches Modified MRE bread

Moist maple filling

Flour Water Shortening Glycerol 4.63% Yeast Salt Sucrose ester Gum arabic Calcium sulfate Xantham gum Sorbic acid Maple flakes Cinnamon flakes French toast flavor Yellow lake

Maple syrup HFCS Staley Isosweet 100 Corn Syrup, Staley 1300 National 104 Ultra Tex 4 Kelgum Dextrose Glycerol 4.70% Water Maple flavor

however, was in the ingredients. The in-house samples used calcium sulfate, and the commercial manufacturer prepared MFFT using appropriate amounts of calcium carbonate, although both products satisfied military specification requirements. The ingredients of the commercial product are compiled in Table 21.3. The relatively high consumer acceptance ratings over accelerated storage for 6 months at T 35 ëC are shown in Table 21.4. Ratings were consistently high (> 7) for all of the considered attributes at time zero, and decreased only relatively slightly over accelerated storage. These high ratings indicate widespread consumer acceptance among diverse panelists, and also indicate an ability to resist rapid degradation reactions that could compromise the product quality and consumer acceptance ratings of other foods during the abusive temperatures of accelerated storage. The variations in formulation between the in-house samples and the commercial manufacturer rendered some corresponding different characteristics in Aw and pH of the end-product. Table 21.5 shows the physical properties of the Table 21.4 Consumer acceptance ratings of MFFT stored at T 35 ëC using a 9-pt Hedonic scale Time/attribute Appearance Odor Flavor Texture Overall

0

1 month

3 months

6 months

7.2 7.3 7.1 7.0 7.2

6.9 6.8 6.8 6.8 7.0

6.6 6.5 6.4 6.8 6.6

6.3 6.0 6.0 6.4 6.1


A case study in military ration foods: the Quasi-chemical model 503 Table 21.5

Initial Aw and pH values of MFFT components prior to inoculation pH

In-house Commercial

Aw

Filling

Bread

Filling

Bread

4.93 5.36

5.07 5.24

0.833 0.851

0.787 0.811

MFFT enrobed sandwich components, the modified bread formulation, and the moist, maple-flavored jelly/viscous liquid filling. According to the written guides, each component of a non-potentially hazardous (non-sterile and non-refrigerated) food should have pH 4.6, have Aw 0.85, and not support the growth of infectious microorganisms like S. aureus. According to the growth/no-growth boundary line for S. aureus in IM pouch bread (Fig. 21.7), the Aw and pH values of the bread component (Table 21.5) place it significantly in the no-growth regime, indicating that S. aureus growth would not be supported. The filling has higher Aw and pH values slightly above the boundary line and in the growth domain. Noting from the formulation of the MFFT (Table 21.3) the comparable humectant content of the layers with respect to glycerol content suggests this similarity helps harmonize the Aw of the respective layers and mitigates moisture migration during storage to some extent. In these circumstances, however, the issue of whether moisture transfer from the filling to the bread is significant enough to create conditions in spatially inhomogeneous microdomains capable of supporting growth of S. aureus during the 3-year shelf life is still a plausible concern. Figure 21.8(a) shows a model bread-filling (barbecue chicken) interface, and Fig. 21.8(b) shows the equilibration of Aw between the filling and bread during storage (within 23 days, and faster at 120 ëF 49 ëC than at room temperature). The MFFT microbial challenge test at two different temperatures (T 25 ëC and T 35 ëC) is a novel approach that used a 5-strain cocktail of S. aureus isolated from foodborne illnesses on two different MFFT formulations (in-house and commercial vendor) with slightly different physical properties of Aw and pH reflecting minor differences in some of the ingredients commonly used in manufacturing within the allowable constraints of military specifications. The temperatures selected are ambient conditions (T 25 ëC) required of the microbial challenge study and the optimal growth temperature for S. aureus (T 35 ëC), to determine if the optimal growth temperature expedites the challenge study results to a more convenient timeframe while still acting as a guarantor of food safety. The following critical factors were considered and handled as discussed below. Steps d, e, and f involved innovations particular to the microbial challenge study for MFFT, as detailed below. a. Selection of challenge organisms. A 5-strain cocktail of S. aureus was selected to account for variability in strain responses from strains that were


504


Fig. 21.8 Moisture migration from filling to bread in (a) model barbecue chicken sandwich, and (b) equilibration of Aw between the filling (s and t at 120 ëF 49 ëC and room temperature, respectively) and bread (l and n at 120 ëF 49 ëC and room temperature, respectively) during storage. Equilibration occurred within 23 days (faster at 120 ëF 49 ëC than at room temperature).

isolated from cases of foodborne illnesses, including S. aureus A-100, ATCC 14458, 993, ATCC 13567, and ATCC 27154. b. Inoculum level. Samples were cut into 25 g squares and inoculated to 104 cfu/ g (Fig. 21.2(b)). c. Inoculum preparation. S. aureus cultures maintained on slants of tryptic soy agar supplemented with 0.5% yeast extract were transferred individually to trypticase soy broth and incubated for 18 h at 35 ëC. The process was repeated for another 18 h, after which 0.3 mL of inoculum from the trypticase soy broth were spread-plated on Plate Count Agar (PCA) and incubated for 18 h



d.

e.

f.

g.

h.

at 35 ëC. Cells were harvested from the PCA and suspended in Butterfield's phosphate buffer to give a Klett54 reading of about 125±135 units (109 cells/ mL). The cultures were mixed in equal proportions and diluted appropriately to produce an inoculum level of approximately 104 CFU/g. Inoculation method. The inoculation method used was innovative for the MFFT. Samples were cut into 25 g squares and inoculated across the cut surface to include bread, filling and the bread/filling interface (see Fig. 21.2(b)). The sample was held in place inside a 400 mL sterile Stomacher bag during inoculation. To try and minimize the introduction of any local areas of higher Aw that could support the growth of the inoculum, the sandwich sample was hand-mixed to some extent to mix the inoculum into the filling, and the sample placed in a Stomacher for 2 min to flatten the sandwich sample. Storage conditions. The storage conditions were varied in a unique manner for the MFFT. After mixing, the inoculated sample was rolled up in the stomacher bag and placed inside a standard trilaminate MRE pouch with a commercial oxygen scavenger, and sealed (not under vacuum). For comparative purposes, one sample set was vacuum-sealed without an oxygen absorber. Samples were held in independent temperature-controlled incubators set at T 25 and 35 ëC, respectively. Duration of study. The microbial challenge study was unique for the MFFT to accommodate military shelf life requirements and requirements for conducting such studies (1±1.3 product shelf-life, which is 3 years or 1095 days at T 25 ëC for military ration food items). Sample analysis and enumeration. Samples were enumerated in duplicate. Series dilutions were made in Butterfield's phosphate buffer and plated on Baird-Parker Agar supplemented with egg yolk-tellurite and incubated at 35 ëC for 48 h. Uninoculated controls were assessed for Aw, pH, aerobic plate counts (APC), E. coli and Coliforms (EC), Yeasts and molds (YM), and inhabitant S. aureus. Pass/fail evaluation criteria. Product passes if there is 1 log of growth for each sample by the endpoint. Product fails if growth > 1 log occurs for any sample at 2 time points or at the endpoint.

21.4

Results of the microbial challenge study

Over the period of study, uninoculated control samples were evaluated using standard microbiological quality control techniques for APCs, EC, YMs, and S. aureus. From days 0±270, the counts in all cases were very low and represented no potential hazard to the consumer (Table 21.6). The two types of MFFT enrobed sandwiches (either made in-house or by a commercial manufacturer) were evaluated for changes in Aw and pH during 1095 days (3 years) of ambient storage (T 25 ëC). In both cases, the Aw and pH roughly equilibrated within the first 30 days of storage (Table 21.7). For the


506


Table 21.6 Incidental microflora (APCs, EC, YMs, and S. aureus) in uninoculated control MFFT samples Time (days)

0 30 130 270

APC (cfu/g)

E. coli coliforms (cfu/g)

Y&M (cfu/g)

SA (cfu/g)

Case Studies in Novel Food Processing Technologies: Innovations in Processing, Packaging, and Predictive Modelling

Case Studies in Novel Food Processing Technologies: Innovations in Processing, Packaging, and Predictive Modelling (Woodhead Publishing Series in Food Science, Technology and Nutrition)

Innovations in Food Packaging

Thermal Technologies in Food Processing

Green Technologies in Food Production and Processing

Minimal Processing Technologies in the Food Industry

Proteins in Food Processing

Handbook of Frozen Food Processing and Packaging

Hydrocolloids in Food Processing

Proteins in Food Processing

Ozone in Food Processing

Novel Food Processing: Effects on Rheological and Functional Properties (Electro-Technologies for Food Processing Series)

Innovations in Food Packaging (Food Science and Technology International)

Improving Traceability in Food Processing

Food Safety in Shrimp Processing

Handbook of Frozen Food Processing and Packaging (Contemporary Food Engineering)

Food biochemistry and food processing

Food Biochemistry and Food Processing

Food Biochemistry And Food Processing

Novel Food Packaging Techniques

Novel Food Packaging Techniques

Measurement and Control in Food Processing

Measurement and Control in Food Processing

Enzymes in Food and Beverage Processing

Encapsulation Technologies for Active Food Ingredients and Food Processing

Challenges in Natural Language Processing (Studies in Natural Language Processing)

Innovative Food Processing Technologies: Advances in Multiphysics Simulation

Innovative Food Processing Technologies: Advances in Multiphysics Simulation

Biotreatment, Downstream Processing and Modelling

Case Studies in Archaeological Predictive Modeling

Food Processing Operations Modeling

Case Studies in Novel Food Processing Technologies: Innovations in Processing, Packaging, and Predictive Modelling